{"repo": "curl", "sha": "c8df3defd9664cd276adc33754210ee07b5bbf71", "date": "2026-08-20T16:06:17+02:00", "subject": "lib: new easy option string storage", "body": "Change the storage of easy handle option strings from an array sized for\nall possible options to a hash set to reduce memory footprint.\n\nGive the hash set initially room for 4 strings, with first allocation\nhappening when it goes beyond that. Measurements without test suite\nand a forced fail on growing the set gives:\n\nSize Result\n2 1261 tests out of 1951 reported OK: 64%\n4 1792 tests out of 1951 reported OK: 91%\n8 1944 tests out of 1951 reported OK: 99%\n16 1949 tests out of 1951 reported OK: 99%\n32 single fail of 3211, unit test for u8_strset\n\nAdd u8_strset that keeps the tuples (uint8_t id, char *str)\nand allows set/unset by `id`. Add that as data->set.strings.\n\nDefine MACROS\n* CURL_EASY_STR(data, id) for access\n* CURL_EASY_STR_SET(data, id, s) for setting, making a copy\n* CURL_EASY_STR_SETN(data, id, s) for setting, no copy\n* CURL_EASY_STR_CLEAR(data, id) for unsetting\n* CURL_EASY_STR_CLEAR0(data, id) for unsetting and zero-ing value\n\nAdd `data->set.str_copypostfields` to handle former `STRING_COPYPOSTFIELDS`\nstring that was not always a string and could carry NUL bytes.\n\nAdd unit tests to test3211.\n\nCloses #22628", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "6141e709f9cd79466af5cd5f3713b724b974dbe5", "date": "2026-08-20T16:11:16+02:00", "subject": "ci: add Alan as a zone.js reviewer", "body": "This updates the zone.js owners and adds alan-agius4 to compiler approvers.\n\n(cherry picked from commit 555369960a9cac219d442cfce371ffbcf7656adf)", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "angular", "sha": "555369960a9cac219d442cfce371ffbcf7656adf", "date": "2026-08-20T16:11:16+02:00", "subject": "ci: add Alan as a zone.js reviewer", "body": "This updates the zone.js owners and adds alan-agius4 to compiler approvers.", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "curl", "sha": "aed01dd1f09b5df32fd212ea06187df14ccf5212", "date": "2026-08-20T16:28:25+02:00", "subject": "CVE-2026-11856: cross-origin Digest auth state leak", "body": "Cherry-picked from 7f5b0d014ce907ab4352ca88", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "a5e4f9ec90347a61f239461bc18de99796b81289", "date": "2026-08-20T16:28:25+02:00", "subject": "CVE-2026-11856: cross-origin Digest auth state leak", "body": "Reimplemented without the creds system", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "a2ff519103c266b6eb3b549638c6a9209eb623eb", "date": "2026-08-20T16:48:12+02:00", "subject": "OSSL_FN: Add OSSL_FN_clear_bit()", "body": "The OSSL_FN counterpart of BN_clear_bit(): clear a bit by absolute\nindex, failing with OSSL_FN_R_RESULT_ARG_TOO_SMALL when the index is\nout of range (OSSL_FN is fixed-size, so the operand cannot be grown to\nreach it). Control flow branches only on the operand's public width\nand the caller-chosen index, never on limb values, matching the leak\nprofile of the BN counterpart.\n\nThe bit index is size_t rather than the BN counterpart's int; a\nnegative index was never meaningful (an error for clearing, a\nread-as-0 for testing). OSSL_FN_is_bit_set() changes along, to keep\nthe bit-index family consistent.\n\nRelated-to: doc/designs/fixed-size-large-numbers.md\nIssue: https://github.com/openssl/project/issues/1654\nAssisted-by: Pi:moonshotai/kimi-k3\nSigned-off-by: Richard Levitte \nReviewed-by: Matt Caswell \nReviewed-by: Jakub Zelenka \nMerge-date: Tue Sep 22 04:56:58 2026\nMerged-from: https://github.com/openssl/openssl/pull/32454", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "1a77631ea31b912b912dfdd48dc1724a908ea5da", "date": "2026-08-20T16:50:40+02:00", "subject": "OSSL_FN: Add OSSL_FN_MONT_CTX_set_locked()", "body": "The OSSL_FN counterpart of BN_MONT_CTX_set_locked(): thread-safe lazy\ninitialization of a shared Montgomery context cache. The context\nconstruction runs without any lock held, so concurrent lazy inits on\nthe same slot duplicate the work rather than serialize on it.\nPublication takes no lock at all where native atomics exist; the lock\nargument remains only as the fallback for the rest. Unlike the BN\ncounterpart this takes no BN_CTX, since OSSL_FN_MONT_CTX_new() needs\nno scratch context.\n\nRelated-to: doc/designs/fixed-size-large-numbers.md\nIssue: https://github.com/openssl/project/issues/1654\nAssisted-by: Pi:moonshotai/kimi-k3\nSigned-off-by: Richard Levitte \nReviewed-by: Matt Caswell \nReviewed-by: Jakub Zelenka \nMerge-date: Tue Sep 22 04:57:01 2026\nMerged-from: https://github.com/openssl/openssl/pull/32454", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "41868e8aaa6cfff63b1c8010e1356e01d86a1326", "date": "2026-08-20T11:26:11-04:00", "subject": "Remove redundant SvOK() tests in plperl.", "body": "I noticed that some places in our code test \"SvOK(sv) && SvROK(sv)\"\nwhile others check just SvROK(sv). On investigation, it's clear\nthat SvROK implies SvOK so testing both is pointless. While removing\nthese extra checks seems very unlikely to make any performance\ndifference, it does make the code more consistent and intelligible.\n\nAlso mop up a couple of places where there wasn't a null-pointer\ncheck before a SvOK() test. I think these are unreachable cases,\nbut in the name of consistency let's do it the same everywhere.\n\nAuthor: Tom Lane \nReviewed-by: Andrey Rachitskiy \nDiscussion: https://postgr.es/m/2034677.1787172042@sss.pgh.pa.us", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "2ecbc5e021be2cad14f837539b1661dd0c870c07", "date": "2026-08-20T11:36:17-04:00", "subject": "Rewrite tsqueryout() to use StringInfo to build the output string.", "body": "This patch started with noticing that tsquery.c's infix() function\nuselessly multiplies the length of each operand string by\npg_database_encoding_max_length() + 1, where multiplying by 2\nwould be sufficient. Unlike the related thinko in tsvectorout(),\nthis doesn't seem to risk integer overflow, since we're multiplying\nonly a rather short per-operand length. Still, it's incorrect and\nmisleading.\n\nHowever, I then started to question why tsqueryout() is using a\nhand-rolled implementation of an expansible string buffer in the\nfirst place. Replacing that by using the StringInfo infrastructure\nwould make the code noticeably shorter and eliminate not only this\nmistake but a bunch of other mistake-prone arithmetic. It might\neven be faster, given that we've spent effort on micro-optimizing\nStringInfos; but even if it's slower, it's hard to visualize a\nworkload where tsqueryout() is a performance bottleneck. So\nthat's what this patch does.\n\nI also got rid of the overly-creative approach of deparsing a\nbinary operator's right operand into a separate buffer, then\ndeparsing the left operand into the main output buffer, then\ncopying the right operand's text back to the main buffer.\n(Why the operands are stored in reverse order in the first place\nseems lost in the mists of time, but I suppose we're stuck with\nthat choice.) We can easily emit the desired text in-order by\nmoving the \"curpol\" next-item pointer around.\n\nAuthor: Tom Lane \nReviewed-by: Ayush Tiwari \nDiscussion: https://postgr.es/m/3428771.1786635967@sss.pgh.pa.us", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "2dd6e9d6a2482069cfe9d12a09f761c5713f246b", "date": "2026-08-20T17:40:42+02:00", "subject": "refactor(drawline.c): flatten condition #41382", "body": "", "style": "conventional", "type": "refactor", "scope": "drawline.c", "breaking": false} {"repo": "openssl", "sha": "e7f3a42c237558214d0ca47b80e0465acd2fe9a7", "date": "2026-08-20T09:46:32-06:00", "subject": "Apply DNS name constraints to the subject CN only when it may match", "body": "Now that the subject DN is not consulted for host checks by default,\ncall NAME_CONSTRAINTS_check_CN() on the leaf only under\nX509_CHECK_FLAG_ALWAYS_CHECK_SUBJECT - matching the verify logic\n\nReviewed-by: Tomas Mraz \nReviewed-by: Andrew Dinh \nMergeDate: Wed Aug 26 15:21:53 2026\n(Merged from https://github.com/openssl/openssl/pull/31982)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "961c95fea6e097ef4dbc19ee996e5127e58acbac", "date": "2026-08-20T09:03:11-07:00", "subject": "openssl: fix DER buffer leak in Apple SecTrust verification", "body": "`ossl_chain_get_der()` allocates a DER encoding of each peer certificate\nvia `i2d_X509()`, but `Curl_vtls_apple_verify()` only copies it into a\nCFData and never frees the original. This leaks per certificate, per\nhandshake, whenever USE_APPLE_SECTRUST is used with the\nOpenSSL/LibreSSL/BoringSSL backend.\n\nFix frees the buffer inside openssl.c itself, so the GnuTLS backend\n(which borrows rather than allocates) is unaffected.\n\nCloses #22631", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "bdb82ec8c24184844d561e2c76ef02cc1b551213", "date": "2026-08-20T09:25:34-07:00", "subject": "Ensure all pg_locale.h APIs work with collate_is_c.", "body": "Safer for callers, so that checking collate_is_c is only needed if the\ncaller wants to optimize for that case.\n\nBackpatch to avoid creating a hazard for other backpatches in this\narea.\n\nSuggested-by: Andres Freund \nSuggested-by: Heikki Linnakangas \nDiscussion: https://postgr.es/m/v36ssaygf7grb3qzfsjhtdzi7kqd45ds56nyuf7gi5qjml4qbb@ezmfqzmhlrs2\nBackpatch-through: 18", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "03d2554638cb5c95819e1e720e15672630a3cbdd", "date": "2026-08-20T12:52:06-04:00", "subject": "Convert DCH_to_char() to use a StringInfo as destination.", "body": "Formerly, we allocated an output char array with 12 bytes per\nbyte of the format string. That's usually far more than enough,\nand yet we had to add assorted code to defend against cases where\nit isn't enough. Switch to using a StringInfo expansible buffer\ninstead. This is considerably more robust, shortens the code\nnoticeably, and eliminates various edge-case failure conditions.\n\nWhile at it, const-ify some parameters that can be const.\n\nAuthor: Tom Lane \nReviewed-by: Heikki Linnakangas \nDiscussion: https://postgr.es/m/3451175.1786641871@sss.pgh.pa.us", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "28c5d0724875591ff55dd85a065207badc2fa8ad", "date": "2026-08-20T12:55:13-04:00", "subject": "Restructure formatting.c's NUM_processor() to add some const-sanity.", "body": "NUM_processor() is an under-documented mess. Depending on the value\nof is_to_char, it either reads \"inout\" and writes \"number\", or\nthe other way around. That makes it impossible to label either\nstring \"const\", which seems like a minimum expectation in 2026.\nIt also complicates replacing the output buffer with a StringInfo:\nwe'd have to convert both strings, which is pretty pointless for\nthe input side.\n\nMoreover, we're buying very little code savings by doing it like this,\nsince a large majority of the function's code has to be wrapped inside\n\"if (is_to_char)\" tests.\n\nTo fix, split the function into NUM_processor_from_char() and\nNUM_processor_to_char(), and rename/const-ify arguments as\nappropriate. Both paths now read from \"const char *input\"\nand write to \"char *output\".\n\nThis patch doesn't intend to make any algorithmic changes, it's\njust mechanical code rearrangement and symbol renaming. The\nonly real exception is that I got rid of passing \"input_len\"\nto NUM_processor_from_char()'s subroutines in favor of storing\n\"const char *input_end\" in NUMProc.\n\nAuthor: Tom Lane \nReviewed-by: Heikki Linnakangas \nDiscussion: https://postgr.es/m/3451175.1786641871@sss.pgh.pa.us", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "ec352b37d88a12f0256526ff0a51bd86d587d604", "date": "2026-08-20T12:57:25-04:00", "subject": "Replace formatting.c's fixed-size output buffers with StringInfos.", "body": "In both versions of NUM_processor(), use a StringInfo for the\noutput string, eliminating the need to guess an upper bound\nfor the output length, and removing the need to truncate some\nstrings of unpredictable length.\n\nAuthor: Tom Lane \nReviewed-by: Heikki Linnakangas \nDiscussion: https://postgr.es/m/3451175.1786641871@sss.pgh.pa.us", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "44f96c8a49963b9b1bff9305b1831893006240bf", "date": "2026-08-21T00:59:48+08:00", "subject": "site: fix code-box style with Badge", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "38337ee639b1825597dee8cc02cc9c0aaa83feb1", "date": "2026-08-20T13:01:28-04:00", "subject": "Micro-optimize appendStringInfo[VA].", "body": "In the loop in appendStringInfo, avoid a useless assignment to\nerrno during the first (and usually only) iteration. This\nmight make a noticeable difference depending on how efficiently\nthe platform deals with thread-local variables.\n\nTry to make the compiler inline appendStringInfoVA into\nappendStringInfo.\n\nInstead of having appendStringInfoVA go through pvsnprintf, make\nit call vsnprintf directly. pvsnprintf adds little except an\nint-versus-size_t impedance mismatch. We do have to duplicate its\nerror handling for the nprinted < 0 case, but we don't need to\nduplicate its check for MaxAllocSize overrun, because\nenlargeStringInfo can handle that just as easily. Also, its\ninsistence on adding one to nprinted is not needed here, since\nenlargeStringInfo expects the number of data bytes to add.\n\nIn combination, these changes seem to about halve the penalty for\ngoing through appendStringInfo rather than directly to snprintf.c.\nThis is enough to buy back the performance loss incurred in\nformatting.c by the preceding three patches, and even a little more.\nIt should help other usages too.\n\nAuthor: Tom Lane \nReviewed-by: Heikki Linnakangas \nDiscussion: https://postgr.es/m/3451175.1786641871@sss.pgh.pa.us", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "05c7c725a534f4ea14c718503966423133474bbf", "date": "2026-08-20T22:12:03+05:00", "subject": "fix(compiler): scope animations declared in minified nested rules", "body": "The regular expressions in `_scopeAnimationRule` expect an `animation` or `animation-name` property to be preceded by whitespace or a semicolon, and its value to end at a semicolon. Minified CSS breaks both assumptions. Inside an at-rule the property follows a `{`, and the last declaration of a block has no trailing semicolon, so the closing `}` lands inside the captured value. The keyframe name is then left unscoped while the `@keyframes` rule itself is renamed, so the animation does nothing in a production build.\n\nAccept `{` as a leading boundary and stop the value at `}`. The prefix is written back unchanged, and a declaration value cannot contain an unescaped `}`.\n\nFixes #70316\n\n(cherry picked from commit 58b0cb47358861de0a512711addb8e29f984a57e)", "style": "conventional", "type": "fix", "scope": "compiler", "breaking": false} {"repo": "angular", "sha": "58b0cb47358861de0a512711addb8e29f984a57e", "date": "2026-08-20T22:12:03+05:00", "subject": "fix(compiler): scope animations declared in minified nested rules", "body": "The regular expressions in `_scopeAnimationRule` expect an `animation` or `animation-name` property to be preceded by whitespace or a semicolon, and its value to end at a semicolon. Minified CSS breaks both assumptions. Inside an at-rule the property follows a `{`, and the last declaration of a block has no trailing semicolon, so the closing `}` lands inside the captured value. The keyframe name is then left unscoped while the `@keyframes` rule itself is renamed, so the animation does nothing in a production build.\n\nAccept `{` as a leading boundary and stop the value at `}`. The prefix is written back unchanged, and a declaration value cannot contain an unescaped `}`.\n\nFixes #70316", "style": "conventional", "type": "fix", "scope": "compiler", "breaking": false} {"repo": "angular", "sha": "f58f5fd1b4ab5d2882c849147b8c251bc2846322", "date": "2026-08-20T10:28:38-07:00", "subject": "refactor(core): loosen return type of ɵɵFactoryDeclaration to any", "body": "When a derived class inherits from a base class, TypeScript enforces that static\nproperties on the derived class are assignable to those on the base class.\nBecause `ɵɵFactoryDeclaration` had a return type of `T`, static `ɵfac`\nmembers across inheritance hierarchies could result in type incompatibilities\n(for example, when dealing with generics or differing class shapes where the\nderived factory return type is not compatible with the base class factory).\n\nUpdating the return type of `ɵɵFactoryDeclaration` from `T` to `any` avoids\nthese static type conflicts across inheriting classes.\n\n(cherry picked from commit 0d7ae1635074ad84d09ce6378aef4630e111576b)", "style": "conventional", "type": "refactor", "scope": "core", "breaking": false} {"repo": "angular", "sha": "0d7ae1635074ad84d09ce6378aef4630e111576b", "date": "2026-08-20T10:28:38-07:00", "subject": "refactor(core): loosen return type of ɵɵFactoryDeclaration to any", "body": "When a derived class inherits from a base class, TypeScript enforces that static\nproperties on the derived class are assignable to those on the base class.\nBecause `ɵɵFactoryDeclaration` had a return type of `T`, static `ɵfac`\nmembers across inheritance hierarchies could result in type incompatibilities\n(for example, when dealing with generics or differing class shapes where the\nderived factory return type is not compatible with the base class factory).\n\nUpdating the return type of `ɵɵFactoryDeclaration` from `T` to `any` avoids\nthese static type conflicts across inheriting classes.", "style": "conventional", "type": "refactor", "scope": "core", "breaking": false} {"repo": "angular", "sha": "7709f2c86d9cea744c5d2584bb9ea621c32a9158", "date": "2026-08-20T11:42:51-07:00", "subject": "refactor(core): widen ɵɵclassProp value parameter type to any", "body": "When Angular type-checks host bindings and template class bindings\n(`[class.foo]=\"expr\"`), the Type Check Block (TCB) emits the binding\nas a standalone expression statement (e.g. `(expr);`). This verifies that\nthe expression itself is syntactically and semantically valid (e.g. properties\nexist on the component instance), but does not constrain the\nexpression to `boolean` because Angular evaluates class bindings using standard\nJavaScript truthiness.\n\nIn classic `ngtsc`, the runtime Ivy instructions (`ɵɵdefineComponent`) were\ngenerated only during the JS emit phase, after TypeScript type-checking had\ncompleted. Thus, `tsc` never validated the arguments passed to `ɵɵclassProp`.\n\nUnder standalone commpilation, Ivy definitions are generated\ndirectly into the TypeScript AST and type-checked by `tsc`. This causes `tsc`\nto check the emitted `ɵɵclassProp('foo', expr)` call against the instruction's\ndeclared signature. Because `ɵɵclassProp` was strictly typed as\n`boolean | undefined | null`, any valid truthy non-boolean expression\n(e.g., `1`, `items.length`, or non-empty strings) produces a `TS2345` compiler error.\n\nAt runtime, `ɵɵclassProp` delegates to `checkStylingProperty`, which evaluates\nthe value via truthiness (`!!value`). Widening the parameter type to `any`\naligns the instruction's type signature with Angular's binding semantics and\nprevents type-checking failures during in-place compilation.\n\n(cherry picked from commit 779f67777d127d1dd707387a8b05f50ceffb9294)", "style": "conventional", "type": "refactor", "scope": "core", "breaking": false} {"repo": "angular", "sha": "779f67777d127d1dd707387a8b05f50ceffb9294", "date": "2026-08-20T11:42:51-07:00", "subject": "refactor(core): widen ɵɵclassProp value parameter type to any", "body": "When Angular type-checks host bindings and template class bindings\n(`[class.foo]=\"expr\"`), the Type Check Block (TCB) emits the binding\nas a standalone expression statement (e.g. `(expr);`). This verifies that\nthe expression itself is syntactically and semantically valid (e.g. properties\nexist on the component instance), but does not constrain the\nexpression to `boolean` because Angular evaluates class bindings using standard\nJavaScript truthiness.\n\nIn classic `ngtsc`, the runtime Ivy instructions (`ɵɵdefineComponent`) were\ngenerated only during the JS emit phase, after TypeScript type-checking had\ncompleted. Thus, `tsc` never validated the arguments passed to `ɵɵclassProp`.\n\nUnder standalone commpilation, Ivy definitions are generated\ndirectly into the TypeScript AST and type-checked by `tsc`. This causes `tsc`\nto check the emitted `ɵɵclassProp('foo', expr)` call against the instruction's\ndeclared signature. Because `ɵɵclassProp` was strictly typed as\n`boolean | undefined | null`, any valid truthy non-boolean expression\n(e.g., `1`, `items.length`, or non-empty strings) produces a `TS2345` compiler error.\n\nAt runtime, `ɵɵclassProp` delegates to `checkStylingProperty`, which evaluates\nthe value via truthiness (`!!value`). Widening the parameter type to `any`\naligns the instruction's type signature with Angular's binding semantics and\nprevents type-checking failures during in-place compilation.", "style": "conventional", "type": "refactor", "scope": "core", "breaking": false} {"repo": "angular", "sha": "151a8f2d0f3254ca4192c135c4fcd117e369a5aa", "date": "2026-08-20T12:08:07-07:00", "subject": "refactor(core): Decouple ɵɵFactoryDeclaration and ɵɵInjectableDeclaration from type parameter T.", "body": "When compiling under standalone compilation, generated runtime static declarations (static ɵfac and static ɵprov) are emitted into preprocessed TypeScript files and visible to the compiler during semantic typechecking.\n\nWhen a subclass extends a base class where the subclass is not structurally subtype-compatible with the superclass (such as differing generic type constraints, contravariant method parameters, or EventEmitter), TypeScript's class static side heritage check (TS2417) fails because ɵɵFactoryDeclaration and ɵɵInjectableDeclaration structurally referenced the instance type T.\n\nThis change updates ɵɵFactoryDeclaration to return any instead of T, and sets factory return and value types in ɵɵInjectableDeclaration to any. This decouples static side inheritance from T, resolving TS2417 errors across subclassed components and injectables while preserving .d.ts metadata indexing and assignability to ɵɵdefineInjectable. This brings ɵfac and ɵprov into alignment with other Ivy declarations (ɵcmp, ɵdir, ɵpipe, ɵinj), which already treat their generic parameters as phantom metadata.\n\n(cherry picked from commit 66d505e287d9f67066028286db8aadcf04e4e574)", "style": "conventional", "type": "refactor", "scope": "core", "breaking": false} {"repo": "angular", "sha": "66d505e287d9f67066028286db8aadcf04e4e574", "date": "2026-08-20T12:08:07-07:00", "subject": "refactor(core): Decouple ɵɵFactoryDeclaration and ɵɵInjectableDeclaration from type parameter T.", "body": "When compiling under standalone compilation, generated runtime static declarations (static ɵfac and static ɵprov) are emitted into preprocessed TypeScript files and visible to the compiler during semantic typechecking.\n\nWhen a subclass extends a base class where the subclass is not structurally subtype-compatible with the superclass (such as differing generic type constraints, contravariant method parameters, or EventEmitter), TypeScript's class static side heritage check (TS2417) fails because ɵɵFactoryDeclaration and ɵɵInjectableDeclaration structurally referenced the instance type T.\n\nThis change updates ɵɵFactoryDeclaration to return any instead of T, and sets factory return and value types in ɵɵInjectableDeclaration to any. This decouples static side inheritance from T, resolving TS2417 errors across subclassed components and injectables while preserving .d.ts metadata indexing and assignability to ɵɵdefineInjectable. This brings ɵfac and ɵprov into alignment with other Ivy declarations (ɵcmp, ɵdir, ɵpipe, ɵinj), which already treat their generic parameters as phantom metadata.", "style": "conventional", "type": "refactor", "scope": "core", "breaking": false} {"repo": "postgres", "sha": "ae78826476499aa5e155cd1bc030e89fdd063edf", "date": "2026-08-20T15:32:55-04:00", "subject": "doc: change \"allowed in->on the standby server.\"", "body": "Reported-by: phamnguyenducduong@gmail.com\n\nBackpatch-through: master", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "aa655f88f71bc747083a4c7e04b6294ff6d19da6", "date": "2026-08-20T13:54:56-07:00", "subject": "hashtext: fix fragile code.", "body": "Previously, in the path for non-deterministic collations, the code\nassumed that bsize==rsize. That assumption seems to be true for ICU,\nand all non-deterministic collations are ICU, so it's not known to be\nan actual bug.\n\nThe only known place where bsize may not equal rsize is in the libc\nprovider, where strxfrm() can return an upper bound of the size needed\nto store the result. That means the initial call to determine the\nbuffer size (with dest==NULL, n==0) could return a larger number than\nthe actual call with an adequate dest buffer. That's OK, because libc\nlocales are always deterministic.\n\nCommit 679c5084cf2 partially fixed the assumption, but missed this\npart. Fix it, and add a more prominent documentation note.\n\nReviewed-by: Haibo Yan \nDiscussion: https://postgr.es/m/CABXr29Hb31nkj1g2Jmk+1BhAm=3ecGs_pWy4tU++j8CQBnbMxQ@mail.gmail.com\nBackpatch-through: 16", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "81920f24a47e9580af8e25a72f6b5b660b69fee1", "date": "2026-08-20T15:56:19-05:00", "subject": "Simplify autovacuum's TOAST-to-main-relation reloptions map.", "body": "The ar_relid and ar_hasrelopts members of the av_relation struct\ncan be removed. ar_relid isn't used by anything, and we only need\nar_hasrelopts because we add an entry to the map even when the main\ntable has no reloptions set. If we instead only add entries for\ntables with reloptions set, we can take the presence of an entry to\nmean the main table has some.\n\nReviewed-by: Michael Paquier \nReviewed-by: Sami Imseih \nTested-by: Greg Burd \nTested-by: solai v \nDiscussion: https://postgr.es/m/aFRxC1W_kZU9OjJ9%40nathan", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "2e2c426e7609de551dafb6360e6373333a278c4e", "date": "2026-08-20T21:11:02Z", "subject": "fix(compiler-cli): deduplicate deferred imports across multiple blocks", "body": "When @Component.deferredImports is defined as an object mapping block\nnames to dependency arrays, deduplicate entries when aggregating block\nimports into the component-level flattened scope. This prevents\nduplicate directive/component collisions in DirectiveMatcher when\nmultiple @defer blocks share dependencies.", "style": "conventional", "type": "fix", "scope": "compiler-cli", "breaking": false} {"repo": "angular", "sha": "c8eb7f00564beceb21445b1f0374c93854b513d1", "date": "2026-08-20T15:00:19-07:00", "subject": "fix(forms): use dot-access for readonly rule configuration", "body": "The readonly rule previously used 'when' in configOrLogic to determine if a configuration object was provided. Under property renaming/minification, the string literal property lookup fails and causes the rule to fall back to being permanently readonly.\n\nThis change switches to the dot-access form (configOrLogic?.when), matching hidden() and ensuring property renaming works correctly.\n\n(cherry picked from commit 1e35de536d8332dfcbb3d89d01ca84b50aa56bd2)", "style": "conventional", "type": "fix", "scope": "forms", "breaking": false} {"repo": "angular", "sha": "1e35de536d8332dfcbb3d89d01ca84b50aa56bd2", "date": "2026-08-20T15:00:19-07:00", "subject": "fix(forms): use dot-access for readonly rule configuration", "body": "The readonly rule previously used 'when' in configOrLogic to determine if a configuration object was provided. Under property renaming/minification, the string literal property lookup fails and causes the rule to fall back to being permanently readonly.\n\nThis change switches to the dot-access form (configOrLogic?.when), matching hidden() and ensuring property renaming works correctly.", "style": "conventional", "type": "fix", "scope": "forms", "breaking": false} {"repo": "postgres", "sha": "497de6394e687e18b007f571ede7505465c6ee21", "date": "2026-08-21T07:05:13+09:00", "subject": "test_aio: Fix broken error recovery assertions in 001_aio", "body": "The three error recovery checks in `test_handle()` used \"qr/^|ok$/\" to\nlook for the marker \"ok\" in psql's output. '^' matches every string, so\nthe assertions passed no matter what psql printed.\n\nSpelling the regex correctly as \"qr/^ok\\|$/\" exposed that the explicit\nxact case was actually failing, reporting an incorrect \"current\ntransaction is aborted\" instead of showing that an AIO handle can be\nacquired again after an error. This is rewritten with a ROLLBACK,\nsimilarly to the subxact counterpart.\n\nWhile on it, the subxact case had no marker column in its query, so add\none there for consistency, and reformat to use same pattern.\n\nAuthor: Jelte Fennema-Nio \nDiscussion: https://postgr.es/m/DKSU6GI1YLG5.3VF6M4IRKQ7XE@jeltef.nl\nBackpatch-through: 18", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "85c8829ac189b0f47b865b036dad98905e576dc9", "date": "2026-08-21T00:20:18+02:00", "subject": "fix(migrations): preserve registerLocaleData calls in standalone bootstrap migration", "body": "Currently the standalone migration only copies symbols referenced from the NgModule metadata into the main file. Top-level `registerLocaleData` calls, and the default imports they depend on, are dropped silently when the module file is pruned, which breaks locale resolution at runtime with a \"Missing locale data\" error.\n\nCopies top-level `registerLocaleData` calls from the file of the bootstrapped module into the main file, and adds support for carrying over default imports which were previously skipped silently by the reference resolution.\n\nFixes #50886\n\n(cherry picked from commit bdc09e8183a3280d25bfc93b4d5ae4452cc551ec)", "style": "conventional", "type": "fix", "scope": "migrations", "breaking": false} {"repo": "angular", "sha": "bdc09e8183a3280d25bfc93b4d5ae4452cc551ec", "date": "2026-08-21T00:20:18+02:00", "subject": "fix(migrations): preserve registerLocaleData calls in standalone bootstrap migration", "body": "Currently the standalone migration only copies symbols referenced from the NgModule metadata into the main file. Top-level `registerLocaleData` calls, and the default imports they depend on, are dropped silently when the module file is pruned, which breaks locale resolution at runtime with a \"Missing locale data\" error.\n\nCopies top-level `registerLocaleData` calls from the file of the bootstrapped module into the main file, and adds support for carrying over default imports which were previously skipped silently by the reference resolution.\n\nFixes #50886", "style": "conventional", "type": "fix", "scope": "migrations", "breaking": false} {"repo": "postgres", "sha": "d39762b8c79a8a825be4af0a22b048a18483afa3", "date": "2026-08-21T07:51:16+09:00", "subject": "Improve error for RETURNING with system columns under rules", "body": "When a DML query's RETURNING list references the system column of a\ntarget table, and the query was rewritten by a rule with its own\nRETURNING clause, ReplaceVarsFromTargetList() failed with an internal\nerror \"could not find replacement targetlist entry for attno -1\",\nsince the rule's RETURNING list only has entries for user columns.\n\nSuch queries cannot work: the rewritten query need not scan the original\ntarget relation at all, so a rule's RETURNING list has no way to provide\nvalues for that relation's system columns.\n\nThis is a life improvement change as the error reported is just more\nrelevant for the user and this has never worked since RETURNING is\nreported. No backpatch is done, based on the lack of complaints over\nthe ages.\n\nReported-by: Zheng Wang \nAuthor: Zsolt Parragi \nDiscussion: https://postgr.es/m/19632-9155d9baec763c8c@postgresql.org", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "c83f7e727f20a680a5bb874e2e4d699a0f3ee68f", "date": "2026-08-21T08:28:07+09:00", "subject": "ecpg: Add missing NULL check in ecpg_store_input() following OOM failure", "body": "ecpg_store_input() called strlen() directly on the result of\nPGTYPESnumeric_to_asc() without checking for a NULL value, state\npossible when the function fails due to an out-of-memory failure.\n\nThis error is unlikely going to be hit in practice, so no backpatch is\ndone.\n\nAuthor: Gladyshev Ilya \nDiscussion: https://postgr.es/m/ed5a633b469e685f3a1aef9982cff8e1c50e5883@linux.dev", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "0ebeec35fc3c3bf617ac447b36f1b158b076f6e7", "date": "2026-08-20T19:47:32-04:00", "subject": "Fix snapshot import xmin ProcArrayLock bug.", "body": "ProcArrayInstallImportedXmin verifies that the source transaction (the\ntransaction whose snapshot we're importing) is still running, and then\ninstalls the caller's imported xmin. These steps have to be atomic.\nBut it was just about possible for VACUUM to fail to observe the\nimported xmin in either the source proc or the importing one. This\ncould result in VACUUM pruning away deleted tuples that were still\nvisible to the imported snapshot.\n\nTo fix, take ProcArrayLock in exclusive mode while importing an exported\nsnapshot's xmin within ProcArrayInstallImportedXmin. That guarantees\nthat a concurrent VACUUM's OldestXmin cannot advance past the xmin (one\nproc or the other always advertises an xmin that holds it back).\n\nAuthor: Chee Wooson \nReviewed-by: Peter Geoghegan \nDiscussion: https://postgr.es/m/20260730042128.714201-1-chee.wooson@gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "f15737f9fa58e130f40cb3f1c961053a99714812", "date": "2026-08-21T08:05:41+08:00", "subject": "site: restore relative URLs for component LLMs links", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "e2bede96134f757aad5c5b33ac9be055022dbfc8", "date": "2026-08-21T05:51:46+05:30", "subject": "chore: enable pnpm trustPolicy: no-downgrade", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "postgres", "sha": "7f1054b51bc742b7fa27ce8e8c1f3d61cf7349e2", "date": "2026-08-21T12:28:43+12:00", "subject": "Attempt to stabilize plan of self-join test in tidscan.sql", "body": "The test that checks the expected plan for this self-join test has been\nknown to have failed in the past due to badly timed VACUUMs causing\nsmall variations in row estimates on one of the tables, resulting in a\nswapped join order. Currently, failures have only been seen in v14, and\nseemingly due to 74388a1ac and 4496020e6 the failures have not been seen\nin more recent versions.\n\nHere we shrink down the number of matching rows on one side of the join\nto make the alternative join order's costs more expensive relative to\nthe cheapest join order. Previously the alternative order had the same\ncost.\n\nWe do this in all supported versions to reduce the chances of future\nchanges reintroducing stability issues with these queries.\n\nReported-by: Alexander Lakhin \nAuthor: David Rowley \nDiscussion: https://postgr.es/m/f5d1f4c2-6224-4797-be17-c86e77f96c9c@gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "03ec1b37d328dbbe47a161c79da873ee7add96c7", "date": "2026-08-20T20:55:48-04:00", "subject": "fix(vim-patch): :redrawtabpanel regexp, delay grep on hunk", "body": "Too many suggestions in\nhttps://github.com/neovim/neovim/commit/bcf116cc2f2f59e6c0f5ce5a644f832476f77384\n\nMulti-line grep on the hunk contents before filtering by hunk header\nprevents vim-patch.sh from detecting v8.1.2195 and v8.1.2219 as N/A.\n\nRevisit multi-line parsing for `runtime/doc/*.txt`\nwhen Vim v9.2+ has new runtime/doc/ patches\nabout features that require \"has()/exists()\" guards.", "style": "conventional", "type": "fix", "scope": "vim-patch", "breaking": false} {"repo": "ant-design", "sha": "482ccda2279656fff5c669e847de198a9fda545b", "date": "2026-08-21T09:26:06+08:00", "subject": "fix(Descriptions): preserve item state when key is 0", "body": "* fix(Descriptions): preserve item state for zero key\n\n* refactor(Descriptions): extract merged item key", "style": "conventional", "type": "fix", "scope": "Descriptions", "breaking": false} {"repo": "postgres", "sha": "e5aff2aa63a72e4329405dd1404c184f7d439107", "date": "2026-08-20T21:35:26-04:00", "subject": "doc: fix wording in SELECT docs", "body": "Reported-by: Garry Polley\n\nAuthor: Garry Polley\n\nDiscussion: https://postgr.es/m/CAPj_C4wRh_JSi_RSFZLPUDVoVeOjFhos5RG+DggtK38wzakjEQ@mail.gmail.com\n\nBackpatch-through: master", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "f3d38c4dd0fd8e29851c120a143e503933f9da9d", "date": "2026-08-21T04:44:48+03:00", "subject": "feat(Cascader, Select, TreeSelect): improve clear button accessibility", "body": "* feat(Select): improve clear button accessibility\n\n* Add clear button tests\n\n* Add missing \"clear\" tranlsation\n\n* Add clear label support to Cascader and TreeSelect\n\n* Bump @rc-component/select\n\n* Add keyboard tests\n\n* Revert non-English and non-Chinese locale changes\n\n* Fix after review\n\n* fix: support clear icon if only label is passed", "style": "conventional", "type": "feat", "scope": "Cascader, Select, TreeSelect", "breaking": false} {"repo": "vite", "sha": "1f5b5392cbfa0c222fd5006be34b3b18a78f604e", "date": "2026-08-21T10:20:40+08:00", "subject": "test: eliminate eslint warnings", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "vite", "sha": "e79c4711181712321bd8c918a66332be0fe2c3ec", "date": "2026-08-21T10:20:59+08:00", "subject": "chore: replace prettier comment with oxfmt", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "postgres", "sha": "609f969f61bbe0cab817a12c4b9f2c1c0fa3297f", "date": "2026-08-21T14:50:18+12:00", "subject": "Reduce memory overheads for storing a Memoize tuple", "body": "This uses 16 fewer bytes per tuple stored in each MemoizeEntry. 24 bytes\ncan be saved by not doing palloc_object(MemoizeTuple), and instead we\ncan store the pointer to the next tuple by requesting that\nExecCopySlotMinimalTupleExtra() allocate an extra 8 bytes per\nMinimalTuple, a net saving of 16 bytes per cached tuple, which equates\nto 33% less memory in the most extreme case (assuming very few entries\nwith many tuples and the narrowest possible MinimalTuple). Real-world\ncases will save less, as we're not saving anything on the MemoizeEntry\nitself, only tuples stored within it.\n\nSaving memory in Memoize allows more entries to exist concurrently and\nreduces the chances that an entry will be evicted from the cache due to\nmemory pressure. That increases the chances of getting cache hits.\n\nAuthor: David Rowley \nReviewed-by: Chao Li \nDiscussion: https://postgr.es/m/CAApHDvoNurQzpR2nZ4aLq1v_4gstsSdimZP-m34PMTBPptn+zQ@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "fd9bb365db5b83249b97b788df4af4e7877fc1e0", "date": "2026-08-21T12:36:01+09:00", "subject": "psql: Do not let invalid \\getresults affect the next query", "body": "In pipeline mode, an invalid \\getresults argument could previously\naffect the next SQL command in the same pipeline. For example, after\nreporting an error for \\getresults -1, psql could treat the following\nSQL command as a request to read pending pipeline results instead of\nsending it to the server, making the command appear to be skipped or\ncausing missing results.\n\nThis happened because psql marked \\getresults as a request to read\npipeline results before validating its optional argument. When\nvalidation failed, psql reported the error without running the normal\ncleanup path that clears the request.\n\nFix this by validating the \\getresults argument before marking the\ncommand as a request to read pipeline results. After an invalid\nargument, psql now reports the error and sends the following SQL command\nnormally.\n\nBackpatch to v18, where psql pipeline meta-commands were introduced.\n\nAuthor: Fujii Masao \nReviewed-by: Anthonin Bonnefoy \nDiscussion: https://postgr.es/m/CAHGQGwGkgM2HKZeig5hobEgUCjn7MMJux4UCcN=OfOzOEvRT0A@mail.gmail.com\nBackpatch-through: 18", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "6bacc956df5a76cc5653b9de4493453b953439fd", "date": "2026-08-21T11:02:44+07:00", "subject": "fix(config): close bundles when generation fails", "body": "", "style": "conventional", "type": "fix", "scope": "config", "breaking": false} {"repo": "vite", "sha": "91ae134699f78fddc1a05dd974f85e6d5c875a96", "date": "2026-08-21T13:18:47+09:00", "subject": "ci: setup PR-driven release", "body": "", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "d2864abf70036f8d2cf1e710725d26ba32577eb0", "date": "2026-08-21T13:43:06+08:00", "subject": "chore: make verify-commit work in linked worktrees", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "curl", "sha": "4f8dabcec2087b731ad00c9f18f8a04e9f5802a4", "date": "2026-08-21T08:41:49+02:00", "subject": "GHA: install MSYS2 Perl manually for latest GitHub runners", "body": "As the MSYS2 update propagates into more GitHub runner images, we need\nto install Perl to avoid native Windows (aka Strawberry) Perl breaking\njobs that are running tests.\n\nFor others jobs, this patch leaves the new default Strawberry Perl. It's\nfine to use it a little bit more.\n\nAlso:\n- libssh 32-bit was dropped upstream, replace with libssh2, and move\n libssh to the 64-bit job running tests.\n https://github.com/msys2/MINGW-packages/commit/f6ae36581d8dd1f6878442575cc8471830b863ed\n https://github.com/msys2/MINGW-packages/pull/30777\n\nRefs:\nhttps://github.com/actions/runner-images/issues/14562\nhttps://github.com/actions/runner-images/pull/14541#issuecomment-5294304716\nhttps://github.com/msys2/MSYS2-packages/commit/636489312cbc55011cb7ec493550598125e7716f\nhttps://github.com/msys2/MSYS2-packages/pull/6491\nhttps://github.com/msys2/MSYS2-packages/pull/6479\n\nBug: https://github.com/curl/curl/pull/22537#issuecomment-5359154818\nFollow-up to d854ab4673c2f9d8048c7f0f6d164b7e4d5e0865 #22580\n\nCloses #22632", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "31c002aa89787e4bbe237e36f51dead96b674998", "date": "2026-08-21T15:51:57+09:00", "subject": "Collect quals for outer-join reduction on demand", "body": "To prove that a forced-null Var is non-nullable in every matching row,\nreduce_outer_joins() uses the strict quals that hold for every row a\nsubtree emits. The first pass currently gathers those quals into each\nsubtree's state, so that the second pass need not re-walk the jointree\nat decision time.\n\nThat duplicates find_subquery_safe_quals(), which already performs\nthis traversal for query_outputs_are_not_nullable(), and it charges\nevery query containing an outer join for a list build at every\nFromExpr and every inner or semi join, whether or not any upper qual\nturns out to force a Var null. Since each level copies what its\nchildren collected, the space required grows quadratically with the\ndepth of nested inner joins. What it buys is small even then: the\nwalk it avoids is cheap enough to be called on demand.\n\nSo rename find_subquery_safe_quals() to find_safe_quals(), export it,\nand drop safe_quals from the first pass's state in favor of the\nsubtree's jointree node. The second pass calls find_safe_quals() at\nthe point where it needs the proof.\n\nAlso stop treating a semijoin there like a left join: semijoins are\nreachable from the new caller, and a semijoin emits only matched\nleft-hand rows, so its ON quals do hold for every output row.\n\nAuthor: Richard Guo \nDiscussion: https://postgr.es/m/CAMbWs4-6FazY7+vCOvR0-vqNuaXOZic8O4D4ZrwMm2woi+Q3uA@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "407e74a46e443aca881e7c2fc66a1fab6e8ffbab", "date": "2026-08-21T10:02:23+03:00", "subject": "fix(docs-infra): repair redirects pointing at removed pages", "body": "Five redirects sent people to the home page instead of a guide, because\ntheir target no longer exists: the four `guide/ngmodules/*` entries point\nat `/guide/ngmodules`, which has no route, and\n`guide/animations/transitions-and-triggers` had a plural in a target that\nis registered as `transition-and-triggers`.\n\nTwo more resolved only after a second hop, redirecting to a path that is\nitself a redirect.\n\nThe spec only checked that a `redirectTo` starts with a slash, which all\nseven satisfied, so it now also checks that the target is a real page and\nthat it is not another redirect.\n\n(cherry picked from commit 39da06138b78753dbd8c372eed22fd622debda43)", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "39da06138b78753dbd8c372eed22fd622debda43", "date": "2026-08-21T10:02:23+03:00", "subject": "fix(docs-infra): repair redirects pointing at removed pages", "body": "Five redirects sent people to the home page instead of a guide, because\ntheir target no longer exists: the four `guide/ngmodules/*` entries point\nat `/guide/ngmodules`, which has no route, and\n`guide/animations/transitions-and-triggers` had a plural in a target that\nis registered as `transition-and-triggers`.\n\nTwo more resolved only after a second hop, redirecting to a path that is\nitself a redirect.\n\nThe spec only checked that a `redirectTo` starts with a slash, which all\nseven satisfied, so it now also checks that the target is a real page and\nthat it is not another redirect.", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "64e02d82c68e3758855de8c882af2897f6103f9e", "date": "2026-08-21T10:04:28+03:00", "subject": "docs: point the LLM indexes at current pages, not redirects", "body": "`llms-list.md` feeds the generated `llms-full.txt`. Three of its DI\nentries were left behind by a page rename: two named files that no longer\nexist and one named a copy that had been superseded. The generator warns\non an unreadable file and carries on, so the build stayed green while\n`llms-full.txt` quietly lost three DI guides and shipped a stale fourth.\n\n`llms.txt` linked the same old URLs, which only reach their pages through\na redirect, plus `guide/hybrid-rendering`, which redirects to `/guide/ssr`\nand duplicates the line above it.\n\nAlso removes `creating-injectable-service.md`, left behind by that rename\nand reachable from nothing.\n\n(cherry picked from commit 51cb07e98081ab7e4e84a9e0949266a8e19cca84)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "51cb07e98081ab7e4e84a9e0949266a8e19cca84", "date": "2026-08-21T10:04:28+03:00", "subject": "docs: point the LLM indexes at current pages, not redirects", "body": "`llms-list.md` feeds the generated `llms-full.txt`. Three of its DI\nentries were left behind by a page rename: two named files that no longer\nexist and one named a copy that had been superseded. The generator warns\non an unreadable file and carries on, so the build stayed green while\n`llms-full.txt` quietly lost three DI guides and shipped a stale fourth.\n\n`llms.txt` linked the same old URLs, which only reach their pages through\na redirect, plus `guide/hybrid-rendering`, which redirects to `/guide/ssr`\nand duplicates the line above it.\n\nAlso removes `creating-injectable-service.md`, left behind by that rename\nand reachable from nothing.", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "a0854d7917a62dfb1d631599ff707e023f3af55f", "date": "2026-08-21T09:21:50+02:00", "subject": "build: remove scorecard workflow", "body": "The scorecard workflow has been replaced at the org level so these changes remove it from our workflows.\n\n(cherry picked from commit 42f6641120daf520ade71b38d327e15343c016b6)", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "angular", "sha": "42f6641120daf520ade71b38d327e15343c016b6", "date": "2026-08-21T09:21:50+02:00", "subject": "build: remove scorecard workflow", "body": "The scorecard workflow has been replaced at the org level so these changes remove it from our workflows.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "openssl", "sha": "b9b1f48da310634256ad164b8d01200af6864899", "date": "2026-08-21T09:25:44+02:00", "subject": "add windows-arm64 job into OS Zoo workflow", "body": "Reviewed-by: Nikola Pajkovsky \nReviewed-by: Milan Broz \nReviewed-by: Norbert Pocs \nMerge-date: Tue Sep 15 08:59:56 2026\nMerged-from: https://github.com/openssl/openssl/pull/32453", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "f31d6fbc31d3b6901022e011942ac680d347d46a", "date": "2026-08-21T16:33:36+09:00", "subject": "Use pg_parse_lsn() for server-supplied LSNs", "body": "Commit d6bf0ab170 introduced pg_parse_lsn() to validate LSNs given on\nthe command line of pg_waldump, pg_recvlogical, and pg_receivewal.\nThe remaining sscanf(\"%X/%08X\") call sites under src/bin parse LSNs\nthat arrive in server responses, timeline history files, and\nbackup_label files. sscanf() accepts several forms that pg_lsn input\nrejects and can silently continue with a different location than the\ninput text: a first component wider than eight hex digits wraps\naround, a wider second component is truncated, and leading\nwhitespace, signs, \"0x\" prefixes, and trailing characters are\nconsumed or ignored.\n\nConvert those call sites as well. The two call sites that read a\nlocation out of a longer line isolate it by temporarily terminating\nthe string at the next whitespace character, so that they can use\npg_parse_lsn() like the others. Each tool keeps its existing error\nmessage.\n\nMalformed metadata now fails with each tool's existing error instead\nof silently proceeding with a different location. Two error paths\nshift: pg_rewind's history-file parser now rejects trailing\ncharacters attached to a switchpoint, which used to be ignored, and a\nmalformed backup_label location now fails pg_combinebackup's \"could\nnot parse\" check rather than its \"improper terminator\" check.\n\nAuthor: Zexin Li \nReviewed-by: Fujii Masao \nDiscussion: https://postgr.es/m/CAAP6ZkS_3OH3yhhAGK6vu+2V1C2Hv4K6SpRuZL415R-gxjdTSg@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "1bbc278c2744639a8ef608e9e5660d0334a5088b", "date": "2026-08-21T17:55:10+10:00", "subject": "fix(browser): preserve trace popover state (fix #10906)", "body": "", "style": "conventional", "type": "fix", "scope": "browser", "breaking": false} {"repo": "ant-design", "sha": "1b10eb3e7f7dcf8ec04a8f68797906251dc90f0d", "date": "2026-08-21T16:45:37+08:00", "subject": "feat(Checkbox): support classNames and styles in Checkbox.Group", "body": "* feat(Checkbox): support classNames and styles in Checkbox.Group\n\n* fix(Checkbox): provide merged disabled state to Group semantic callbacks", "style": "conventional", "type": "feat", "scope": "Checkbox", "breaking": false} {"repo": "ant-design", "sha": "6632feb44a0235fa4032b4e6f919310263ca4679", "date": "2026-08-21T16:45:58+08:00", "subject": "feat(Radio): support classNames and styles in Radio.Group", "body": "* feat(Radio): support classNames and styles in Radio.Group\n\n* fix(Radio): provide merged props to Group semantic callbacks", "style": "conventional", "type": "feat", "scope": "Radio", "breaking": false} {"repo": "postgres", "sha": "c63b210ed34d9153e82c86abba3160a569b86f72", "date": "2026-08-21T17:50:01+09:00", "subject": "Fix typos in comments.", "body": "Author: Etsuro Fujita \nReviewed-by: Bharath Rupireddy \nDiscussion: https://postgr.es/m/CAPmGK14otpa7XZyBO5GxDABUK9OPWFLsAwoQP3ugNdXq%3DK4J1g%40mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "80748383af79ec9edfc716aea806e3a5d298592d", "date": "2026-08-21T10:54:54+02:00", "subject": "OSSL_FN: Add static storage machinery for compile-time constants", "body": "Add OSSL_FN_HEADER_FIELDS, OSSL_FN_STATIC_DECLARE, and\nOSSL_FN_STATIC_DEFINE to include/crypto/fn_intern.h, enabling an OSSL_FN\nto be statically initialized in .rodata via a union with a fixed-size\narray member.\n\nstruct ossl_fn_st's header fields are now spelled through the\nOSSL_FN_HEADER_FIELDS macro, keeping the runtime struct and the static\nunion shape in lockstep by construction.\n\nThese macros are the support infrastructure for giving BN_FLG_STATIC_DATA\nBIGNUM constants an OSSL_FN backing, so bn_get_ossl_fn() can return a\nvalid view for them without runtime materialization.\n\nRelated-to: doc/designs/fixed-size-large-numbers.md\nIssue: https://github.com/openssl/project/issues/1654\nAssisted-by: Pi:moonshotai/kimi-k3\nSigned-off-by: Richard Levitte \nReviewed-by: Matt Caswell \nReviewed-by: Daniel Kubec \nMerge-date: Tue Sep 22 05:06:23 2026\nMerged-from: https://github.com/openssl/openssl/pull/32460", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "56e9ca2dbefabe648037ab469462fde617858be0", "date": "2026-08-21T10:02:03+01:00", "subject": "screen-redraw: erase only cells a pane owns, not its raw rectangle", "body": "image_redraw_start() was called with a pane's full nominal geometry\nrather than the cells it actually owns in the current scene. When a\nfloating pane occluded part of that rectangle, the erase blanked the\nfloating pane's on-screen content, and the following text redraw\ncorrectly skipped those cells (not owned by this pane), leaving them\nblank with nothing to restore them.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "f42cc5d95612d6f15f78809e5c73a66a2a9a5f42", "date": "2026-08-21T10:03:11+01:00", "subject": "regress: fix contradictory terminal-features assertions in image-support.sh", "body": "Three adjacent blocks shared one 'set -as terminal-features' setting\nwith contradictory requirements - fallback text, then real SIXEL, then\nfallback text again in a single client attach. terminal-features is\nonly detected at client attach, not live, so this needs detach/reattach\nbetween blocks rather than toggling the option in place.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "8cb872e7fb65b03f6068923c6aa7fcf3e71baf21", "date": "2026-08-21T17:03:52+08:00", "subject": "fix(dev): run closeBundle after buildEnd failure", "body": "", "style": "conventional", "type": "fix", "scope": "dev", "breaking": false} {"repo": "vite", "sha": "616296895bd135386d35069a479a5f188c7de298", "date": "2026-08-21T18:06:20+09:00", "subject": "fix(hmr): handle `import.meta.hot.invalidate` in virtual module", "body": "", "style": "conventional", "type": "fix", "scope": "hmr", "breaking": false} {"repo": "vite", "sha": "67a680767317f8e2cb28b6b0500192f993a567cf", "date": "2026-08-21T18:07:38+09:00", "subject": "refactor: remove HmrUrl concept", "body": "", "style": "conventional", "type": "refactor", "scope": "", "breaking": false} {"repo": "vite", "sha": "a500deeb6f52d93ca501a0fc612a5392b939f2f5", "date": "2026-08-21T17:07:57+08:00", "subject": "feat(cli): support naming the CPU profile via --profile [name]", "body": "", "style": "conventional", "type": "feat", "scope": "cli", "breaking": false} {"repo": "curl", "sha": "2f1dda96911c41c725747eb1775da1759ef9bc95", "date": "2026-08-21T11:08:23+02:00", "subject": "u8_strset: start with 8 default entries", "body": "The previous 4 entries were exhausted on simple command lines, depending\non environment/build. Grow the initial size to 8 to avoid early allocs.\n\nCloses #22633", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "b78e2f1bc1cba404c4bd9faf518d26ec85e89fc7", "date": "2026-08-21T05:08:53-04:00", "subject": "feat: support subpath imports in dynamic import statements", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "tmux", "sha": "e63e4fe0d068e6ef0eca493a594b8f971390eed2", "date": "2026-08-21T10:10:46+01:00", "subject": "regress: add failing test for border-status redraw bug", "body": "window_make_pane_status() gates physical redraw of border-status text on\na logical content diff, not physical damage - if something else disturbs\nthose cells, the text stays blank because tmux believes nothing changed.\nThis test currently fails and is expected to keep failing until the\ngeneral damage-tracking redraw work lands (see IMAGE-REDRAW-PLAN.md).", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "8d83bb2df065260e0c532489a8e7c0cac4a89728", "date": "2026-08-21T17:11:57+08:00", "subject": "perf(runtime-vapor): trim v-for hot-path allocations", "body": "", "style": "conventional", "type": "perf", "scope": "runtime-vapor", "breaking": false} {"repo": "vite", "sha": "8156684572bdcf73e9d8568ed67971f0467fab60", "date": "2026-08-21T18:12:00+09:00", "subject": "feat(css): minify style tag", "body": "", "style": "conventional", "type": "feat", "scope": "css", "breaking": false} {"repo": "vite", "sha": "924997a4bdda9115faee9bdb622fcec4fc8357f0", "date": "2026-08-21T18:13:28+09:00", "subject": "feat(worker): remove worker chunk if it's detected that it's not referenced", "body": "", "style": "conventional", "type": "feat", "scope": "worker", "breaking": false} {"repo": "vite", "sha": "e17d2d565b0288f169c7995adb2b192f917548e7", "date": "2026-08-21T10:15:18+01:00", "subject": "feat: add closeServer and closePreviewServer hooks", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "vite", "sha": "5164b6a979da2f0889fbb2fcb6ea07ac5add9b7a", "date": "2026-08-21T18:27:56+09:00", "subject": "test(hmr): skip virtual module `import.meta.hot.invalidate` test in bundled-dev", "body": "", "style": "conventional", "type": "test", "scope": "hmr", "breaking": false} {"repo": "openssl", "sha": "e9d1c2ca4f8b8b4bd9fe22ec057bdb3df8f313b9", "date": "2026-08-21T11:28:20+02:00", "subject": "OSSL_FN: Back ossl_bn_inv_sqrt_2 with a static OSSL_FN constant", "body": "Move the ground truth of inv_sqrt_2 (1/sqrt(2) * 2^256, used for\nFIPS 186-5 auxiliary-prime sizing) into crypto/fn/fn_inv_sqrt_2.c as\nan OSSL_FN_STATIC_DEFINE'd constant, declared in the new internal\nheader include/crypto/fn_constants.h.\n\ncrypto/bn/bn_rsa_fips186_5.c's ossl_bn_inv_sqrt_2 BIGNUM becomes a\nthin view over the union storage: .data points at the OSSL_FN member,\n.d points at the fixed array member. This is the template for the\nother global BN_FLG_STATIC_DATA constants.\n\nRelated-to: doc/designs/fixed-size-large-numbers.md\nIssue: https://github.com/openssl/project/issues/1654\nAssisted-by: Pi:moonshotai/kimi-k3\nSigned-off-by: Richard Levitte \nReviewed-by: Matt Caswell \nReviewed-by: Daniel Kubec \nMerge-date: Tue Sep 22 05:06:24 2026\nMerged-from: https://github.com/openssl/openssl/pull/32460", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "1b5cfe3d3777d4ceb7f35fcee9d3c4279316a084", "date": "2026-08-21T17:34:52+08:00", "subject": "feat: accept Rolldown watch options in `server.watch`", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "openssl", "sha": "eb4b7e6ba2200bb4831d4d5353c19ed98ea5bf27", "date": "2026-08-21T11:39:29+02:00", "subject": "OSSL_FN: Back the DH, FFDHE, and MODP constants with static OSSL_FN", "body": "Move the limb ground truth of crypto/bn/bn_dh.c's group parameter\narrays (dh1024/2048, ffdhe2048-8192, modp1536-8192, and the literal 2)\ninto crypto/fn/fn_dh_consts.c as OSSL_FN_STATIC_DEFINE'd constants,\ndeclared in include/crypto/fn_constants.h.\n\ncrypto/bn/bn_dh.c's ossl_bignum_* BIGNUMs become views over the union\nstorage via the reworked make_dh_bn macro, following the same pattern\nas ossl_bn_inv_sqrt_2.\n\nRelated-to: doc/designs/fixed-size-large-numbers.md\nIssue: https://github.com/openssl/project/issues/1654\nAssisted-by: Pi:moonshotai/kimi-k3\nSigned-off-by: Richard Levitte \nReviewed-by: Matt Caswell \nReviewed-by: Daniel Kubec \nMerge-date: Tue Sep 22 05:06:26 2026\nMerged-from: https://github.com/openssl/openssl/pull/32460", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "4ce0f88584e9536af88c1fbb502a167aaf77b29d", "date": "2026-08-21T11:44:14+02:00", "subject": "OSSL_FN: Back the SRP group constants with static OSSL_FN", "body": "Move the limb ground truth of crypto/bn/bn_srp.c's group prime and\ngenerator arrays (bn_group_1024-8192 and bn_generator_2/5/19) into\ncrypto/fn/fn_srp_consts.c as OSSL_FN_STATIC_DEFINE'd constants,\ndeclared in include/crypto/fn_constants.h.\n\ncrypto/bn/bn_srp.c's ossl_bn_group_* and ossl_bn_generator_* BIGNUMs\nbecome views over the union storage via a make_srp_bn helper macro\nthat takes both the exported BIGNUM name and the union name.\n\nRelated-to: doc/designs/fixed-size-large-numbers.md\nIssue: https://github.com/openssl/project/issues/1654\nAssisted-by: Pi:moonshotai/kimi-k3\nSigned-off-by: Richard Levitte \nReviewed-by: Matt Caswell \nReviewed-by: Daniel Kubec \nMerge-date: Tue Sep 22 05:06:28 2026\nMerged-from: https://github.com/openssl/openssl/pull/32460", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "7e7ee16dd3a6e36029fd41c595724c6614176bcb", "date": "2026-08-21T11:50:33+02:00", "subject": "ws: fix write callback error handling", "body": "When websocket writes out data thought the registered write callback and\nthat returns an error, clear the write buffers, so a sub-sequent flush\nwill not attempt another write. The transfer is failed at that point and\nshould no longer invoke the callback.\n\nFixes #22627\nReported-by: Scott Talbert\nCloses #22634", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "af4b72ddea29811fe65c2853c0fd8582c9768f9a", "date": "2026-08-21T11:52:33+02:00", "subject": "OSSL_FN: Test bn_get_ossl_fn() views on static constants", "body": "Add test_static_const_views to test/fn_internal_test.c, checking that\nbn_get_ossl_fn() on ossl_bignum_const_2, ossl_bignum_ffdhe2048_p, and\nossl_bn_inv_sqrt_2 returns usable OSSL_FN operands with the expected\nwidths and values.\n\nRelated-to: doc/designs/fixed-size-large-numbers.md\nIssue: https://github.com/openssl/project/issues/1654\nAssisted-by: Pi:moonshotai/kimi-k3\nSigned-off-by: Richard Levitte \nReviewed-by: Matt Caswell \nReviewed-by: Daniel Kubec \nMerge-date: Tue Sep 22 05:06:29 2026\nMerged-from: https://github.com/openssl/openssl/pull/32460", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "696a16cca27408f7d073d097efa4f763b614aef9", "date": "2026-08-21T09:53:04Z", "subject": "Use correct X position when combining characters, GitHub issue 5511 from Francesc Rosas.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "936bb6e7cdff4fc7700a4f36a6b828b87d9f8241", "date": "2026-08-21T10:53:59+01:00", "subject": "Regress for GitHub issue 5511 from Francesc Rosas.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "517b97f57ab9473e7417da856eb641d76870a56e", "date": "2026-08-21T12:10:36+02:00", "subject": "feat: searched params attached to workers are now preserved", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "vite", "sha": "a6c08e10a624bd89b78683ff1b0e8cfa1d89aa45", "date": "2026-08-21T19:11:28+09:00", "subject": "refactor: exclude postfix from `__VITE_ASSET__`", "body": "", "style": "conventional", "type": "refactor", "scope": "", "breaking": false} {"repo": "vite", "sha": "4366ac468343252df6d5706361a6348afa66f9cc", "date": "2026-08-21T19:18:21+09:00", "subject": "feat: use `import.meta.ROLLDOWN_FILE_URL_*` for assets in JS", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "vitest", "sha": "3069e59279d14aab6612af92761ea1c8cbf8c18f", "date": "2026-08-21T19:34:40+09:00", "subject": "feat(ui): persist trace view selection in URL", "body": "", "style": "conventional", "type": "feat", "scope": "ui", "breaking": false} {"repo": "openssl", "sha": "f7143f957e47da1efddb92f5cc38532da7da15ed", "date": "2026-08-21T12:54:13+02:00", "subject": "OSSL_FN: Update FIPS module manifests for the static constants", "body": "The result of running 'make update-fips-checksums'\n\nRelated-to: doc/designs/fixed-size-large-numbers.md\nIssue: https://github.com/openssl/project/issues/1654\nAssisted-by: Pi:moonshotai/kimi-k3\nSigned-off-by: Richard Levitte \nReviewed-by: Matt Caswell \nReviewed-by: Daniel Kubec \nMerge-date: Tue Sep 22 05:06:32 2026\nMerged-from: https://github.com/openssl/openssl/pull/32460", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "e38f29ee48bea5ea3178faec5b78708e86f38afb", "date": "2026-08-21T20:14:28+09:00", "subject": "feat: use `import.meta.ROLLDOWN_FILE_URL_*` for other plugins", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "3f8a0ce1832fb30053f55856088371b097bbf7e1", "date": "2026-08-21T11:16:49Z", "subject": "fix(microservices): handle non-json mqtt response packets gracefully", "body": "ClientMqtt.createResponseCallback parsed every incoming reply with an\nunguarded JSON.parse. A non-json payload on a subscribed reply topic threw\ninside the async mqtt message listener, producing an unhandled rejection,\nand the pending request observable never settled.\n\nEvery other transport already guards this parse: ClientRedis falls back to\nthe raw content with a debug log, and the mqtt, redis and rmq servers all\nparse defensively. This applies the same guard ClientRedis uses, passing\nthe raw content to the deserializer.", "style": "conventional", "type": "fix", "scope": "microservices", "breaking": false} {"repo": "vite", "sha": "ce69ea67dc1483dd5df4e94b2c7fa6dab8bf6125", "date": "2026-08-21T20:17:53+09:00", "subject": "docs: plugin asset emitting", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vite", "sha": "0291408b8443129ce6f6d1d440be8facabe9683b", "date": "2026-08-21T20:25:24+09:00", "subject": "test: add `renderBuiltUrl` change changes hash", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "vite", "sha": "92bd2a7f325ed102349cdc6c1ad4b5cd25e1d72f", "date": "2026-08-21T20:25:45+09:00", "subject": "refactor: use `urlId` of `import.meta.ROLLDOWN_FILE_URL` in wasm plugin", "body": "", "style": "conventional", "type": "refactor", "scope": "", "breaking": false} {"repo": "vite", "sha": "75196079fcbbd432887e01844371f2c238a7ed85", "date": "2026-08-21T20:31:19+09:00", "subject": "test: add asset path replacement in `typeof` test", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "postgres", "sha": "c16a1b4f2ce8e88ec40d1465c01e327ab8721524", "date": "2026-08-21T13:55:58+02:00", "subject": "Fix error code for null FOR PORTION OF target", "body": "When the target expression of FOR PORTION OF (...) evaluated to NULL,\nExecInitModifyTable raised an error without an errcode, so clients got\nthe internal error code XX000 for a user-reachable condition.\nOversight in commit 8e72d914c52.\n\nTo fix, report ERRCODE_NULL_VALUE_NOT_ALLOWED, and reword the message\nto \"FOR PORTION OF target must not be null\", matching similar executor\nmessages such as \"frame starting offset must not be null\".\n\nBug: #19630\nReported-by: Zheng Wang \nReported-by: Yanjie Zhao\nReported-by: Yiyang Liu\nAuthor: Zsolt Parragi \nDiscussion: https://www.postgresql.org/message-id/flat/19630-9f10ca28426295fa%40postgresql.org", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "6c61f14c4b24778e66b020fb6ee32e7426e02a05", "date": "2026-08-21T13:59:46+02:00", "subject": "CVE-2026-19931: Negotiate ambient user conn reuse", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "axios", "sha": "9e51e031b453d7b766eaa9410b8b338f811e2cbf", "date": "2026-08-21T14:58:18+02:00", "subject": "feat!: require Node.js 20 for v2", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": true} {"repo": "openssl", "sha": "4b2c1aa778352e8f443c0040de36624f5752d3fa", "date": "2026-08-21T15:15:28+02:00", "subject": "OSSL_FN: Make OSSL_FN_mod_exp_mont() constant-time w.r.t. the exponent", "body": "This is the OSSL_FN counterpart of bn_mod_exp_mont_fixed_top(): fixed-width\nwindows over the exponent's full allocation width, a strided power table,\nand value-masked table selection, replacing the sliding-window path that\nbranched on the exponent's bits and could not be used for secret exponents.\nThe leak profile matches the BN counterpart's: only operand widths and the\nmodulus may leak; there is no early exit on the exponent's value. An\nexponent of value zero in a non-zero-width container flows through the\nloop and yields 1 (table entry 0 holds mont(1)), so the old early exit on\nOSSL_FN_num_bits(p) == 0 survives only for the zero-width case.\n\nOne deliberate divergence from the BN counterpart: bn_mod_exp_mont_fixed_top()\nfalls back to the non-constant-time BN_mod_exp_mont() above\nBN_CONSTTIME_SIZE_LIMIT. OSSL_FN is constant-time by default, so this\nfunction errors out when the power table would not fit in an OSSL_FN (the\nlimb count is an int) rather than silently regress the leak profile.\n\nThe arena size for OSSL_FN_mod_exp_mont_ctx_size() grows with the wider\npower table: 2^window entries at the modulus width (window up to 6, vs.\nat most 16 entries in the sliding-window table) plus alignment slack.\nThe sizing derives the window from the exponent's width only, so it\nmatches the runtime choice exactly.\n\nRelated-to: doc/designs/fixed-size-large-numbers.md\nIssue: https://github.com/openssl/project/issues/1654\nAssisted-by: Pi:moonshotai/kimi-k3\nSigned-off-by: Richard Levitte \nReviewed-by: Matt Caswell \nReviewed-by: Jakub Zelenka \nMerge-date: Tue Sep 22 05:09:26 2026\nMerged-from: https://github.com/openssl/openssl/pull/32455", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "832d77453de18a9f70b517b65fca8118da3eeaeb", "date": "2026-08-21T15:15:41+02:00", "subject": "test: extend OSSL_FN_mod_exp_mont coverage for the constant-time path", "body": "New cases: a zero-valued exponent in a multi-limb container (exercises\nthe loop-yields-1 path that replaced the early exit on num_bits == 0),\nan exponent much wider than the modulus, and an exponent much narrower\nthan the modulus. Strengthen the existing peak-usage validation of\nOSSL_FN_mod_exp_ctx_size() to check peak numbers and limbs, not just\nframes.\n\nAssisted-by: Pi:moonshotai/kimi-k3\nSigned-off-by: Richard Levitte \nReviewed-by: Matt Caswell \nReviewed-by: Jakub Zelenka \nMerge-date: Tue Sep 22 05:09:27 2026\nMerged-from: https://github.com/openssl/openssl/pull/32455", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "openssl", "sha": "f82a14f1e10a7917ee1cb8072706d3a32a308c57", "date": "2026-08-21T15:28:54+02:00", "subject": "OSSL_FN: Use the accelerated Montgomery gather path in mod_exp_mont", "body": "bn_mod_exp_mont_fixed_top() has an OPENSSL_BN_ASM_MONT5 path built on\nbn_mul_mont_gather5() / bn_scatter5() / bn_gather5() / bn_power5() /\nbn_get_bits5(). These operate purely on limb arrays -- OSSL_FN_ULONG is\nBN_ULONG -- so they are usable from OSSL_FN code as-is; OSSL_FN is\nconstant-time w.r.t. values, so the width-based gating (window == 5,\nml > 1, ml <= BN_SOFT_LIMIT) preserves the constant-time profile. The\npath is compile-time gated, so the portable fixed-window path remains as\nthe fallback, and sizing accounts for the extra cache-local copy of the\nmodulus limbs that the accelerated path parks after the power table.\n\nThe SPARC T4 path is not carried over; it gates on runtime CPU\ncapability registers that the OSSL_FN layer has no business probing.\nThe RSAZ paths likewise remain on the BN side for now: they gate on\nexact operand widths plus a runtime CPU check, which deserves its own\nconsideration.\n\nAssisted-by: Pi:moonshotai/kimi-k3\nSigned-off-by: Richard Levitte \nReviewed-by: Matt Caswell \nReviewed-by: Jakub Zelenka \nMerge-date: Tue Sep 22 05:09:29 2026\nMerged-from: https://github.com/openssl/openssl/pull/32455", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "dd1a8fbea34b0a830011c2113e08f4781d711f48", "date": "2026-08-21T23:31:30+08:00", "subject": "ci: set up bot workflow and ai policy", "body": "", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "postgres", "sha": "63979e5bea14c533dadc8f44cee0ec7a2899ccf5", "date": "2026-08-21T11:50:39-05:00", "subject": "Ensure TOAST storage parameters have unsettable defaults.", "body": "A follow-up commit will teach autovacuum and manual VACUUM to fill\nin a TOAST table's unset storage parameters from its main table's.\nTo be able to determine whether a parameter is set, it must default\nto a value that users cannot set. Of all the parameters that can\nbe set on TOAST tables, log_autovacuum_min_duration is the only one\nthat violates this rule. This commit changes its default to -2 so\nthat we can tell whether it is set. This shouldn't produce any\nuser-visible behavior changes.\n\nWhile at it, add assertions to ensure all storage parameters that\ncan be set on TOAST tables continue to follow this rule.\n\nReviewed-by: Michael Paquier \nReviewed-by: Sami Imseih \nReviewed-by: Greg Burd \nTested-by: solai v \nDiscussion: https://postgr.es/m/aFRxC1W_kZU9OjJ9%40nathan", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "64236573525c257ecd7e268b255571328d4871c8", "date": "2026-08-21T13:42:17-04:00", "subject": "feat(cmdatom)!: eliminate the need for vim-repeat #41414", "body": "Problem:\n- `lhs` is not fully realized. E.g. for a \"payload\" mapping\n `lhs` omits the `getchar()` payload during a mapping (vim-surround\n `ds'` reports `lhs=\"ds\"`). This means plugins like vim-repeat are\n still needed...\n- The \"delta\" fields of a CmdAtom are calculated too late.\n - `` and `changed` check whatever (wrong) buffer a command\n (\":bnext\") might land in.\n - CTRL-W_w between two windows on the same buffer reports type=\"motion\".\n\nSolution:\n- `CmdOrigin` samples (buf/win/cursor/changedtick) at each \"scope\" entry\n (CmdFrame, composite, Visual session, insert session).\n- `ddl` reports `changed=true` for the buffer it edited, regardless\n of where the cursor ends up.\n- New fields:\n - `pos`: cursor position at command start.\n - `moved`: indicates whether the cursor moved (in same buffer).\n - `undoseq`: undo state at settlement.\n- lhs now includes the payload: \"ds)\" reports lhs=\"ds)\" instead of \"ds\".\n - Easy for users to \"replay\" any atom.\n- Rename: type \"command\" => \"normal\", \"ex\" => \"excmd\"; `arg` => `cmdarg`\n- Drop `cascade` field (no reason to expose it)", "style": "conventional", "type": "feat", "scope": "cmdatom", "breaking": true} {"repo": "neovim", "sha": "47cd769ed5a62db81f4c1d836902bf6c4c43ebd5", "date": "2026-08-21T20:04:24+02:00", "subject": "feat(cmdatom): mappings capture continuation", "body": "Problem:\nA mapping that ends mid-operation (`nnoremap ,D d`) emits a content-free\n\"mapping\" atom plus a `pending` field, and the \"continuation\" motion\narrives as a sibling atom. Consumers must stitch the two together (which\nhas broken cases, e.g. Insert-opening mappings (\",i\") lose their session\nentirely).\n\n- \",i\": the session atom is dropped bc the mapping RHS is consumed\n before the session starts (typebuf_maplen()==0), so\n atom_is_user_input()=false.\n- \":normal\"-in-opfunc: the opfunc internal \"v..y\" session (a) became\n kVatomTyped just because the deferred composite was open, masking the\n real operator capture via atom_captures, and (b) its nested frames\n re-derived the outer redo.\n- \"Motion\" based on `moved=true`, has false negatives.\n- `CmdAtom.remap` is unnecessary, and clutters the docs/usage.\n\nSolution:\n- Introduce `frame_id` to identify CmdFrames.\n- Classify `type=motion` better, via `NV_MOTION` flag on the `nv_cmds` table.\n- Drop `CmdAtom.pending`, `CmdAtom.remap`.\n- Defer atom_composite_end() at the clock edge while an operator is\n pending, Visual is active, or `restart_edit` is set: the composite\n keeps collecting, so the continuation is captured in the mapping atom.\n- \",i\": Now an open composite counts as user input.\n- \":normal\"-in-opfunc: Now handled correctly.\n- `remap` is now decided by `composite.payload || 0 subatoms`.\n atom_payload_mark() records the read the resolution never captures.\n- `toplevel` is now decided by `CmdFrame.parent == NULL`.\n\nbefore/after:\n\n INPUT BEFORE AFTER\n ---------------------------------------------------------------------\n ,D w {mapping lhs=,D pending=operator} {operator lhs=,Dw keys=dw}\n + {operator keys=dw}\n ysiw\" {mapping lhs=ys pending=operator} {operator lhs=ysiw\" keys=g@iw\"}\n + {operator lhs=g@iw\"}\n ,v d {mapping pending=visual} {visual lhs=,vd keys=viwd}\n + {visual lhs=viwd}\n ,i X", "style": "conventional", "type": "feat", "scope": "cmdatom", "breaking": false} {"repo": "angular", "sha": "d2e3bac33b022679e596ed388eaefa316a0837e9", "date": "2026-08-21T11:38:19-07:00", "subject": "refactor(router): add support for blocking router resources", "body": "Extends router resource integration to support blocking resources during navigation transitions.\n\n(cherry picked from commit fa2aca969f18f41c1d6cfd275d84d0ecac5b9d79)", "style": "conventional", "type": "refactor", "scope": "router", "breaking": false} {"repo": "angular", "sha": "fa2aca969f18f41c1d6cfd275d84d0ecac5b9d79", "date": "2026-08-21T11:38:19-07:00", "subject": "refactor(router): add support for blocking router resources", "body": "Extends router resource integration to support blocking resources during navigation transitions.", "style": "conventional", "type": "refactor", "scope": "router", "breaking": false} {"repo": "postgres", "sha": "4ee79f3b464f97999caa44624f393199a69850a5", "date": "2026-08-21T13:43:15-05:00", "subject": "Move the StdRdOptions parse table to file scope.", "body": "This is preparatory work for a follow-up commit that will walk the\ntable from another function. This just moves code around; it\nshouldn't change any behavior.\n\nReviewed-by: Michael Paquier \nReviewed-by: Sami Imseih \nDiscussion: https://postgr.es/m/aFRxC1W_kZU9OjJ9%40nathan", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "babel", "sha": "1da3cfa1a25503c7ecba7304f8738933e6d16dca", "date": "2026-08-22T03:41:04+08:00", "subject": "Enable `no-useless-computed-key` rule", "body": "no-useless-computed-key", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "cbd18197b5e47945fdf2aed4cf53f0c93a0443de", "date": "2026-08-21T19:59:05Z", "subject": "include: include when building for modern Linux.", "body": "Fix curl compilation errors when building on Linux with LLVM-libc\n(libc.llvm.org). It provides fd_set type (and associated macro) in\n according to POSIX.\n\ncurl uses fd_set type (and associated FD* macro) that are provided by\n according to POSIX. In practice, some libc\nimplementations (notably, glibc on Linux) transitively include this\nheader from , but other implementations with stricter POSIX\ncompliance may not do that (see example of LLVM-libc above).\n\nFix curl compilation under LLVM-libc by including in\npublic when building for Linux and _POSIX_C_SOURCE is set (by\nthe user or by system libc) to be at least 200112L, referencing the\nPOSIX standard that added / fd_set.\n\nCloses #22635", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "6c04b424bd0a32c9e4a0bd9c5ba3d290360190b7", "date": "2026-08-22T00:06:42+02:00", "subject": "VERSIONS.md: document Rock-solid curl releases", "body": "While they are not curl releases done by the Open Source project or\nprovided openly, these version numbers might appear in the wild so let's\ndocument them in this separate list.\n\nCloses #22636", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "7788fcf39bf98c3403f052f3edd73d86fe1501a7", "date": "2026-08-22T11:38:30+12:00", "subject": "Fix failing Memoize test on 32-bit machines", "body": "609f969f6 reduced memory overheads for storing a tuple in Memoize. That\nresulted in a Memoize eviction test no longer evicting anything from the\ncache on 32-bit machines.\n\nHere we fix that by adding another column to store in the cached\nMinimalTuple.\n\nReported-by: Michael Paquier \nDiscussion: https://postgr.es/m/aof8KwoXaqOBFR_D@paquier.xyz", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "961c540fdedd54162a99652a09fcc20906bfa05e", "date": "2026-08-22T02:58:34+03:00", "subject": "refactor(docs-infra): remove images no longer referenced by any doc", "body": "Most of these came over from the angular.io docs and were orphaned as adev\nrewrote or dropped the guides that used them. Two are newer: a devtools\nscreenshot that was added but never referenced, and the logo left behind\nwhen the Firebase Studio launcher was removed. All of them are still\ncopied into the build and served from angular.dev.\n\nNo markdown, template, stylesheet, TypeScript file or build target\nreferences any of them, by filename or through a path built at runtime,\nand the images are globbed into the build rather than listed, so nothing\nelse needs updating. Removes 78 files, 5.15 MB in total, and empties\nnineteen directories.\n\n(cherry picked from commit 355a9d83e3c3d8fabae2eeff2c812ac40eb22224)", "style": "conventional", "type": "refactor", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "355a9d83e3c3d8fabae2eeff2c812ac40eb22224", "date": "2026-08-22T02:58:34+03:00", "subject": "refactor(docs-infra): remove images no longer referenced by any doc", "body": "Most of these came over from the angular.io docs and were orphaned as adev\nrewrote or dropped the guides that used them. Two are newer: a devtools\nscreenshot that was added but never referenced, and the logo left behind\nwhen the Firebase Studio launcher was removed. All of them are still\ncopied into the build and served from angular.dev.\n\nNo markdown, template, stylesheet, TypeScript file or build target\nreferences any of them, by filename or through a path built at runtime,\nand the images are globbed into the build rather than listed, so nothing\nelse needs updating. Removes 78 files, 5.15 MB in total, and empties\nnineteen directories.", "style": "conventional", "type": "refactor", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "c867eed0c99c98c66bca3e8ce90838ec590facbf", "date": "2026-08-22T03:05:34+03:00", "subject": "docs: repair the AST link in the AOT compiler guide", "body": "The link to the Wikipedia article on abstract syntax trees spelled the\nunderscores as asterisks, which 404s. The same link two paragraphs\nearlier in this guide is correct.\n\n(cherry picked from commit 3074fdb3ab5ec2ba1ead4d35ae74bec4b3850683)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "3074fdb3ab5ec2ba1ead4d35ae74bec4b3850683", "date": "2026-08-22T03:05:34+03:00", "subject": "docs: repair the AST link in the AOT compiler guide", "body": "The link to the Wikipedia article on abstract syntax trees spelled the\nunderscores as asterisks, which 404s. The same link two paragraphs\nearlier in this guide is correct.", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "postgres", "sha": "e8c6d114ea112ddaa8ab226092c46e3d68f63a83", "date": "2026-08-22T12:51:22+12:00", "subject": "Doc: fix typo", "body": "Author: Jochen Bandhauer \nDiscussion: https://postgr.es/m/69ecfa04-9177-42fd-8d5d-9f375669fc5b@jbitc.de\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "39a7b685395db67dc7f8abafa32f14369ddc8a65", "date": "2026-08-21T22:42:45-04:00", "subject": "feat(vim-patch): ignore vim-patch tokens from commit list", "body": "Sort vimpatch_commit_ignore.txt based on \"git log\" order\nelse \"diff\" fails to filter out reverted vim-patch tokens.\nPut latest commit to revert on top of file.", "style": "conventional", "type": "feat", "scope": "vim-patch", "breaking": false} {"repo": "vite", "sha": "97ad042170f4c71b518239723b733dd98e8e3e76", "date": "2026-08-22T15:36:54+09:00", "subject": "chore: use oxfmt `sortImports`", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "postgres", "sha": "617636328f5285fa2e0b7e42f4ae3cb0d82f8b5e", "date": "2026-08-22T16:23:16+09:00", "subject": "Retire execdebug.h and its debugging facility", "body": "execdebug.h includes \"pretty old and crufty\" code used to print\ndebugging information in the executor. As reported recently, this code\nhas not compiled in the last 10 years or so, and there are better\nalternatives available, such as elog().\n\nThere is no indication that this facility is still being used, so remove\nit along with its associated debugging macros:\nEXEC_NESTLOOPDEBUG\nEXEC_SORTDEBUG\nEXEC_MERGEJOINDEBUG\n\nReviewed-by: Tom Lane \nReviewed-by: Tatsuya Kawata \nDiscussion: https://postgr.es/m/aoFfAvp6xdk7Cdi6@paquier.xyz", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "e2c812f1475dc75ab6f8a39fb5696d8d32d05fa1", "date": "2026-08-22T16:23:21+09:00", "subject": "Track RI fast-path FK-check batches per subtransaction", "body": "Commit 4113873 confined RI fast-path batching to the top transaction\nlevel to avoid mishandling the batch cache on subtransaction abort.\nThat disabled batching for a foreign-key load wrapped in a savepoint,\nsuch as:\n\n BEGIN; SAVEPOINT s; COPY fk_table FROM ...\n\nThis was a surprising performance cliff and departed from the usual\nper-subtransaction resource handling.\n\nTrack cache entries per subtransaction instead. Add AtEOSubXact_RI(),\ncalled from CommitSubTransaction() and AbortSubTransaction() after\nResourceOwnerRelease(). On abort, it removes only entries opened by\nthe ending subtransaction, whose resources have just been released,\nwhile leaving entries opened by an outer level intact. Thus, an inner\nsubtransaction abort during outer-level trigger firing does not\ndiscard the outer statement's batch. On commit, no matching entry is\nexpected because its batch should already have been flushed at\nstatement end. Each entry records the subtransaction that opened its\nresources.\n\nAfter an abort, the remaining slot storage and per-entry flush contexts\nare reclaimed when TopTransactionContext is reset at top-level\ntransaction end.\n\nA fast-path batch is filled and flushed within a single trigger-firing\ncycle, so every row added to an entry must come from the subtransaction\nthat created it. AtEOSubXact_RI() relies on this invariant to identify\nan aborting subtransaction's entries by the subid stamped at entry\ncreation. Assert the invariant in ri_FastPathBatchAdd().\n\nAdd regression coverage for batching during nested firing inside a\nsubtransaction, both with different constraints and with the same\nconstraint at the inner and outer firing levels.\n\nReported-by: Noah Misch \nReported-by: Nikolay Samokhvalov \nDiscussion: https://postgr.es/m/20260705222115.be.noahmisch@microsoft.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "1d365db156acbd50a13048692c75a91db8dcd4ef", "date": "2026-08-22T00:23:59-07:00", "subject": "perf(cache): hash the environment-invariant part of the fs module cache key once", "body": "", "style": "conventional", "type": "perf", "scope": "cache", "breaking": false} {"repo": "vite", "sha": "a98c8d95d552c6fa98ab66fb810f5b39ad5abf49", "date": "2026-08-22T15:25:55+08:00", "subject": "chore: remove eslint comment", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "neovim", "sha": "f9ea56f95970a8c72573a798c6ff77b7ca77f1e7", "date": "2026-08-22T15:58:47+08:00", "subject": "test: fix $NVIM_LOG_FILE from environment being overwritten", "body": "Instead of setting it in the test runner (which runs in an Nvim instance\nthat already has a log file), make run_tests.zig set a $NVIM_LOG_FILE\nfallback path like what RunTests.cmake does.", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "neovim", "sha": "855ff5f78f342d325c4421a71e41e7b72c5157fa", "date": "2026-08-22T19:16:30+10:00", "subject": "fix(zip): detect encryption without letting unzip prompt #41422", "body": "", "style": "conventional", "type": "fix", "scope": "zip", "breaking": false} {"repo": "angular", "sha": "4cf8c55930feb5fd65f57d27f9b08f91a7719c0b", "date": "2026-08-22T12:49:42+02:00", "subject": "ci: pass required inputs for adev preview artifact upload", "body": "The `pack-and-upload-artifact` action from `dev-infra` recently added conditional steps that require `triggering-label` and `angular-robot-key` to be passed. Without these inputs, the internal action steps were silently skipped, resulting in no artifact being uploaded. This caused the subsequent deploy workflow to fail when it couldn't find the `adev-preview` artifact.\n\n(cherry picked from commit 3be5facf044d48ac572f7e07ad57846c2c361419)", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "angular", "sha": "3be5facf044d48ac572f7e07ad57846c2c361419", "date": "2026-08-22T12:49:42+02:00", "subject": "ci: pass required inputs for adev preview artifact upload", "body": "The `pack-and-upload-artifact` action from `dev-infra` recently added conditional steps that require `triggering-label` and `angular-robot-key` to be passed. Without these inputs, the internal action steps were silently skipped, resulting in no artifact being uploaded. This caused the subsequent deploy workflow to fail when it couldn't find the `adev-preview` artifact.", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "neovim", "sha": "2c5f37f6af962bc9ed65ca87b8272800badfb78f", "date": "2026-08-22T19:40:17+08:00", "subject": "fix(lsp): empty filterText/sortText/insertText not treated as unset #41421", "body": "Problem:\nEmpty string is falsy, so the spec's label fallback applies, but there only checks\nfor nil. An empty filterText drops the item, an empty sortText sorts it\nfirst, and an empty insertText leaves the word empty.\n\nSolution:\nTreat empty string as unset.", "style": "conventional", "type": "fix", "scope": "lsp", "breaking": false} {"repo": "curl", "sha": "5c61e168698a72b87437d58ed728bcdea6d5db42", "date": "2026-08-22T15:01:09+03:00", "subject": "docs: make 5 example snippets compile cleanly with clang", "body": "verify-examples.pl invokes 'gcc', which on macOS is clang. With\n-Wall -Wextra... -Werror -pedantic, clang rejects two patterns that\nGNU gcc accepts:\n\n- passing a char * to %p without a cast (-Wformat-pedantic): the C\n standard says %p expects void *, so cast explicitly in the\n CURLOPT_PROGRESSDATA, CURLOPT_PROGRESSFUNCTION, CURLOPT_XFERINFODATA\n and CURLOPT_XFERINFOFUNCTION examples\n\n- returning CURLcode result from main() when curl_easy_init() returned\n NULL leaves it uninitialized (-Wsometimes-uninitialized) in the\n CURLINFO_TLS_SSL_PTR example: initialize it to CURLE_OK like other\n man pages already do\n\nCloses #22638", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "tmux", "sha": "faba4112894fbb9bb35796fe67d4b0075e30fe4e", "date": "2026-08-22T13:51:03+01:00", "subject": "layout: only mark scrollbar for redraw when the pane actually changed", "body": "wp->flags |= PANE_REDRAWSCROLLBAR was set unconditionally whenever a pane\n reserved a scrollbar, even if layout_fix_panes() left its geometry\n completely unchanged - forcing a needless scrollbar redraw on every layout\n pass. Move it inside the existing \"did this pane's geometry actually\n change\" check.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "b53077d5f0c293b4d34a14dac0ae39e78255c2ce", "date": "2026-08-22T14:59:25+02:00", "subject": "fix(coverity): performance inefficiencies (PASS_BY_VALUE)", "body": "The perf cost is negligible, but it's easy to address so might as well.\n\n CID 653826: Performance inefficiencies (PASS_BY_VALUE)\n /src/nvim/input_cmdatom.c: 385 in atom_push_raw()\n 384 /// Takes ownership of the atom's allocated members. Caller sets `atom.changed`.\n >>> CID 653826: Performance inefficiencies (PASS_BY_VALUE)\n >>> Passing parameter atom of type \"CmdAtom\" (size 176 bytes) by value, which exceeds the low threshold of 128 bytes.\n 385 void atom_push_raw(bool cascade, CmdAtom atom)\n\n CID 653825: Performance inefficiencies (PASS_BY_VALUE)\n /src/nvim/input_cmdatom.c: 438 in atom_stage_set()\n 436 /// Stages an atom built before its command executes (do_pending_operator() prep-exempt, Visual\n 437 /// ops), in the command's frame; pushed at frame end, once `changed` is known.\n >>> CID 653825: Performance inefficiencies (PASS_BY_VALUE)\n >>> Passing parameter atom of type \"CmdAtom\" (size 176 bytes) by value, which exceeds the low threshold of 128 bytes.\n 438 static void atom_stage_set(CmdAtom atom)\n\n CID 653824: Performance inefficiencies (PASS_BY_VALUE)\n /src/nvim/input_cmdatom.c: 427 in atom_push()\n 425 /// Pushes an atom (emit + maybe cascade), or drops it if replay/Visual/internal-op already\n 426 /// in-progress.\n >>> CID 653824: Performance inefficiencies (PASS_BY_VALUE)\n >>> Passing parameter atom of type \"CmdAtom\" (size 176 bytes) by value, which exceeds the low threshold of 128 bytes.\n 427 static void atom_push(bool cascade, CmdAtom atom)", "style": "conventional", "type": "fix", "scope": "coverity", "breaking": false} {"repo": "angular", "sha": "8eeb7245359bb8c0c296e5170a926fe2ecbcecdd", "date": "2026-08-22T16:46:05+03:00", "subject": "fix(docs-infra): give duplicated headings their own anchor ids", "body": "Eight guides have headings that generate the same anchor id twice or more.\nBecause an id resolves to the first element that claims it, the table of\ncontents lists two entries pointing at the same place, the second section\ncannot be reached from the ToC at all, and the copy link button on the\nlater heading hands out a URL for the earlier one.\n\nOn `guide/forms/reactive-forms` the two \"Display the component\" steps sit\n5176px apart and both ToC entries scroll to the first.\n\nGive the later heading on each page an explicit id with the `{#custom-id}`\nsyntax the pipeline already supports. Only later duplicates are changed, so\nevery anchor that resolves today keeps pointing at the same heading.\n\n(cherry picked from commit 698a14d735b856a7a59004978fc1a84376087435)", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "698a14d735b856a7a59004978fc1a84376087435", "date": "2026-08-22T16:46:05+03:00", "subject": "fix(docs-infra): give duplicated headings their own anchor ids", "body": "Eight guides have headings that generate the same anchor id twice or more.\nBecause an id resolves to the first element that claims it, the table of\ncontents lists two entries pointing at the same place, the second section\ncannot be reached from the ToC at all, and the copy link button on the\nlater heading hands out a URL for the earlier one.\n\nOn `guide/forms/reactive-forms` the two \"Display the component\" steps sit\n5176px apart and both ToC entries scroll to the first.\n\nGive the later heading on each page an explicit id with the `{#custom-id}`\nsyntax the pipeline already supports. Only later duplicates are changed, so\nevery anchor that resolves today keeps pointing at the same heading.", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "b6e83c5e72172eb8e6021774ae413801e7939fde", "date": "2026-08-22T17:01:01+03:00", "subject": "fix(docs-infra): fail the build on duplicate heading anchors", "body": "Two headings on the same page can generate the same anchor id, and every\nlink to it then resolves to whichever comes first. Nothing caught this\nbecause the route manifest keeps anchors in a `Set`, so a repeat collapses\ninto one entry before any check runs, and the existing link validation only\nasks whether an anchor exists, which a duplicate satisfies.\n\nThe generator now inspects each page's headings while they are still an\nordered list and fails with the offending pages and anchors. The scanning\nand the duplicate check move to `heading.mts` so they are covered by tests\nnext to `getIdFromHeading`, which should keep this from coming back.\n\n(cherry picked from commit 5e8a70a010be0f0e1ef0df1ab658f20c72403837)", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "5e8a70a010be0f0e1ef0df1ab658f20c72403837", "date": "2026-08-22T17:01:01+03:00", "subject": "fix(docs-infra): fail the build on duplicate heading anchors", "body": "Two headings on the same page can generate the same anchor id, and every\nlink to it then resolves to whichever comes first. Nothing caught this\nbecause the route manifest keeps anchors in a `Set`, so a repeat collapses\ninto one entry before any check runs, and the existing link validation only\nasks whether an anchor exists, which a duplicate satisfies.\n\nThe generator now inspects each page's headings while they are still an\nordered list and fails with the offending pages and anchors. The scanning\nand the duplicate check move to `heading.mts` so they are covered by tests\nnext to `getIdFromHeading`, which should keep this from coming back.", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "vitest", "sha": "c3ba16b35847d27fb9f01f38cdcb047c77e121f8", "date": "2026-08-22T07:17:01-07:00", "subject": "fix(vm): fall back to compiling from source when a module's code cache is rejected", "body": "", "style": "conventional", "type": "fix", "scope": "vm", "breaking": false} {"repo": "neovim", "sha": "dadaa7fcade459ebd1fef2ee4eada1a389ea036d", "date": "2026-08-22T15:19:16+01:00", "subject": "fix(ui2): offset extmarks copied to pager by srow #41425", "body": "Problem: wrong end_row when copying extmarks to the pager buffer when expanding\na message. Results in incorrect highlights or errors such as \"invalid `end_col`\nout of range\".\n\nSolution: don't forget to offset end_row by srow, as is already done for the\ncopied lines and extmark rows.", "style": "conventional", "type": "fix", "scope": "ui2", "breaking": false} {"repo": "tmux", "sha": "d6074895bb40f4a4d5d5c7551d6cdcd8480f98ff", "date": "2026-08-22T15:45:29+01:00", "subject": "screen-redraw, screen-write: add damage-rectangle tracking and composition", "body": "Introduce a per-window list of damaged rectangles (struct redraw_damage)\nand redraw_damage_window() to record them, redraw_client_damage() to\nconsume them by composing exactly the damaged cells (via a new\nredraw_draw_damage_rect(), which also force-refreshes any pane-status\nspan it touches, since window_make_pane_status()'s content-diff check\nhas no way to know the physical cells were disturbed by something else).\n\nredraw_draw_span() now takes an explicit [clip_x, clip_x + clip_n) range\ninstead of always drawing a span's full width, so a damage rectangle can\nredraw just the portion of a span it actually covers.\n\nscreen_write_redraw_cb() - the fallback when a write can't be applied\ndirectly to the terminal - now reports damage for just the affected row\nvia this mechanism, instead of unconditionally flagging the whole pane\nfor a full redraw.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "f5839915537597ae047a0a9fd674bc377071f1b3", "date": "2026-08-22T20:19:55+05:30", "subject": "fix(url): escape cmd.exe special chars in vim.ui.open() URL #41394", "body": "Problem:\nOn Windows, vim.ui.open() passes URLs to `cmd.exe /c start`. cmd.exe\ntreats `&` as a command separator, so query strings are truncated.\n\nSolution:\nCaret-escape `&|<>^%!` in URIs when the open handler is cmd.exe.", "style": "conventional", "type": "fix", "scope": "url", "breaking": false} {"repo": "tmux", "sha": "f255f089fc8c7685e21c4abaca0a656b5234d227", "date": "2026-08-22T16:32:21+01:00", "subject": "window: add floating-pane damage/redraw helpers", "body": "window_pane_redraw_floating() reports damage for only a floating pane's\nold and new rectangle (via redraw_damage_window(), grown by one cell to\ncover its border frame - see the \"floating\" case in screen-redraw.c),\ninstead of the caller falling back to a full window redraw. Any other\npane whose *scrollbar strip* - not its whole body - intersects either\nrectangle still gets PANE_REDRAWSCROLLBAR directly, since scrollbars\naren't covered by the damage system.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "48cdd85886b30fed94f176857585c9a674f5413d", "date": "2026-08-22T16:36:25+01:00", "subject": "server-client: consume window damage during the normal redraw pass", "body": "redraw_client_damage() (added previously, unused until now) is called\nfrom server_client_check_redraw()'s normal redraw pass, and\nserver_client_any_pane_redraw() now also checks for pending window\ndamage so a client with only damage (no PANE_REDRAW/PANE_REDRAWSCROLLBAR\nflags) still gets its redraw pass run.\n\nserver_client_check_redraw() now returns whether the redraw was deferred\n(waiting for outstanding tty output to drain) rather than performed. A\ndeferred redraw no longer escalates to a full CLIENT_REDRAWWINDOW to\navoid losing what was pending - server_client_loop() now only clears\nPANE_REDRAW, PANE_REDRAWSCROLLBAR and window damage once every client\nviewing a window actually drew this pass (tracked via a new per-window\nredraw_deferred flag), otherwise they're left in place and retried in\ntheir normal, narrowly-scoped form.\n\nserver_client_set_session() now redraws only if the client's session or\ncurrent window actually changed, not on every call (e.g. switch-client\n-t= from clicking a pane name in the status line resolves here even\nwhen nothing besides the active pane changed).\n\nA drag callback's mouse_drag_update() now opens a sync region itself\n(tty_sync_start()) before its first write, so a fast-path write it makes\ndirectly and a later correction arriving via redraw_client_damage() end\nup in the same atomic terminal update instead of two visible frames.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "b26eeb5229964391e0b631114c863894f2b22f3e", "date": "2026-08-22T16:42:56+01:00", "subject": "window: redraw only borders/status on active-pane change", "body": "window_set_active_pane() unconditionally called server_redraw_window(w)\non every active-pane change, redrawing every pane's content even though\nonly the previous and new active pane's border/status appearance\nactually changed. Unzooming (which does change every pane's geometry)\nstill gets the full redraw; otherwise this now only redraws borders and\nstatus.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "b254f557e4f7fc5589ad877ef852f562ba2fea19", "date": "2026-08-22T16:53:08+01:00", "subject": "md-join-pane, cmd-split-window, cmd-resize-pane: use shared floating-pane redraw instead of full client redraw", "body": "The three interactive mouse-drag paths that move or resize a floating\npane (move-pane -M's Alt-drag, split-window/new-pane's interactive\nresize, and resize-pane's own border drag) each unconditionally called\nserver_redraw_window(w), redrawing every pane in the window for a change\nthat only ever disturbs the floating pane's own old and new rectangle.\nSwitch all three to window_pane_redraw_floating().", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "da4895d559cb794f3e976b58dddfd3ea0a720d4b", "date": "2026-08-22T16:54:25+01:00", "subject": "popup: damage-based redraw during drag/resize instead of full client redraw", "body": "popup_handle_drag()'s MOVE and SIZE branches each called\nserver_redraw_client(c) unconditionally, redrawing the client's entire\nwindow on every drag step. Report damage for just the popup's old and\nnew rectangle instead, via a new popup_damage() (translating from raw\nclient/tty coordinates into window coordinates), and set\nCLIENT_REDRAWOVERLAY so the popup itself still redraws.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "513869744f520c8068af81b3089a7fe4bdd40467", "date": "2026-08-22T18:20:05+02:00", "subject": "GHA/checkdocs: verify examples with picky warnings enables, also with clang", "body": "Verify example snippets with the picky warning options used when\ncompiling curl core code and standalone example code. Also test with\nclang. Extract picky warnings available via a new, internal, CMake\noption.\n\nAlso:\n- fix issues found.\n- tried reproducing #22638 in CI (specifically the casts for `%p`),\n turns out it needs apple clang 17, and no longer reproducing with\n apple clang 21.\n\nFollow-up to 5c61e168698a72b87437d58ed728bcdea6d5db42 #22638\n\nCloses #22639", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "axios", "sha": "9767b28e5f3f763ecee651fd9d3b044a8c7ba66a", "date": "2026-08-22T19:16:57+02:00", "subject": "docs: archive Node 20 runtime baseline plan", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "postgres", "sha": "f20c4278342f6afc44b856e98a0850f9dd2df0a8", "date": "2026-08-22T14:50:40-04:00", "subject": "Fix GIN posting tree page deletion with incomplete splits.", "body": "GIN posting tree page deletion failed to consider whether the target\npage's left sibling page, or the deletion target itself, was marked as\nincompletely split. Page deletion finds the target page's left sibling\nby walking the parent's downlinks, but an incompletely split page's new\nright half is part of the sibling chain despite having no downlink.\nDeletion could therefore overwrite the rightlink of the wrong page,\ndisconnecting the split's still-live right half from the sibling chain.\nScans would then silently miss tuples from that page.\n\nTo fix, teach the relevant page deletion path to avoid deleting a\nposting tree page whose left sibling is marked incompletely split (and\nto avoid doing so when the target page itself is so marked). This is\nessentially the same approach used by nbtree page deletion.\n\nClaude Code found this problem. The committed test case is a simplified\nversion of the one that it wrote to demonstrate this bug.\n\nAuthor: Peter Geoghegan \nReviewed-by: Andrey Borodin \nDiscussion: https://postgr.es/m/CAH2-Wz=sKJcn+OtfVN9rdg+Ps9e4cuQWNP-9t12UE2d8nEG90Q@mail.gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "7037e1effe9e3c98d43b36a287dd39e87eaf1bb2", "date": "2026-08-22T15:07:45-04:00", "subject": "fix(write): :write can target a garbage filename #41432", "body": "Problem:\n`buf_write()` captures the buffer's `fname`/`sfname`/`ffname`, then emits its\nprogress-message before opening the file. This may run user code\nsynchronously: the `Progress` autocmd, and the `msg_show` handler of an\nin-process UI (ui2). Either can change the CWD, and `shorten_fnames()`\nthen frees/reallocs every buffer's short name. The rest of `buf_write()`\nreads the freed name...\n\n \"foldtext()\" [New] 41L, 997B written\n E212: Can't open file for writing: illegal byte sequence\n\nASAN, with ui2 enabled and a `BufEnter` handler that runs `:lcd`:\n\n READ path_skip_sep <- path_tail <- match_file_list <- buf_write\n FREE shorten_buf_fname <- shorten_fnames <- update_cwd <- set_curbuf\n <- win_set_buf <- nvim_open_win <- ui2 msg_show handler\n <- ui_call_msg_show <- msg_ext_ui_flush <- buf_write\n\nSolution:\nCopy the names after the `*Pre` autocmds.\n\nNote: `readfile()` has the same shape, but its messages pass no\nprogress-id, so they skip `msg_progress()`. Safe, for now...", "style": "conventional", "type": "fix", "scope": "write", "breaking": false} {"repo": "neovim", "sha": "9b0bc7edef82200a7461332b75208a2160854226", "date": "2026-08-22T12:48:16-07:00", "subject": "fix(plines): 'linebreak' ignores inline virtual text width #41227", "body": "Problem:\nWith 'linebreak', a word is moved to the next screen line when it does\nnot fit on the current one. The check measures only the buffer text, so\ninline virtual text anchored inside the word is not counted. The word is\nkept on a line it cannot fit, and the virtual text is split across the\nscreen line boundary.\n\nSolution:\nCount the width of inline virtual text anchored within the word, so the\ncheck uses the width that is actually displayed.", "style": "conventional", "type": "fix", "scope": "plines", "breaking": false} {"repo": "tmux", "sha": "ebb44d47157363a113fa437e2c0e0eaf25de610d", "date": "2026-08-22T21:16:54+01:00", "subject": "image: report damage narrowly instead of flagging the whole pane for redraw", "body": "image_redraw_area() used to unconditionally flag the whole pane\n(PANE_REDRAW) whenever any part of it needed an image recomposited - a\npane can be much taller than the area actually disturbed (a small scroll\nregion, or a single line insert/delete far from where an image sits), so\nthis erased and retransmitted every image anywhere in the pane regardless\nof proximity to the affected area. Report damage for just the area via\nredraw_damage_window() instead.\n\nimage_redraw_scroll() similarly now scopes to the current scroll region\nrather than redrawing the whole pane via image_redraw_all().", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "e40ef98b1ea8b8bfd7152cb55b20cfffc539ec46", "date": "2026-08-22T21:19:35+01:00", "subject": "image: extend existing placements when a pane grows wider", "body": "image_write() only ever creates spans for as much of an image as fit in\nthe pane at the time it was placed - the rest of the image's pixels are\nstill retained (struct image is immutable and kept for as long as any\nplacement references it), but nothing revisited that clipping decision,\nso a pane that was too narrow when an image was displayed stayed clipped\nforever, even after growing wide enough to fit the rest. Unlike height,\nwhich recovers via ordinary scrollback, there is no equivalent \"scroll\nright\" - this is the only way the extra width is ever recovered.\n\nimage_grid_resize_width() finds every placement referenced in a grid's\nrows, works out how far its spans already reach and its origin column,\nand extends them (via a new image_extend_row(), factored out of\nimage_write()'s own span-building loop) up to whichever is smaller: the\nimage's own full width or the new pane width. window_pane_resize() calls\nit after a pane grows wider.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "b296666e419cdf7a673fdac7c348d0d95577f5ac", "date": "2026-08-22T16:24:05-04:00", "subject": "fix(cwd): stale buffer names after temp context-switch #41433", "body": "Problem:\n1. `ctx_dirs_restore()` is the only chdir site that doesn't re-shorten\n buffer names; `post_chdir()`, `update_cwd()` and `do_autochdir()` all\n call `shorten_fnames(true)`. So after a temp window-context switch\n that moved the CWD, every buffer's `b_fname` is still relative to the\n other directory. ui2 renders messages in a float, and entering it is\n such a switch, so with a `:bcd` in a `BufReadPost` handler 'statusline'\n \"%f\" shows \".config/nvim/init.lua\" while the CWD is already\n `~/.config/nvim`, and :write resolves the name against it:\n ```\n E212: Can't open file for writing: no such file or directory\n ```\n2. `msg_multihl()` leaves `msg_ext_id` pointing at the caller's storage\n when nothing was emitted (e.g. 'msg_silent'): the reset only ran on\n a flush that produced chunks. The next message then ships a dead\n stack frame as its msg_show id. After \":silent write\" the id is\n buf_write()'s `msg_id[MAXPATHL + 32]`:\n ```\n id = \"\\0\\0\\0\\0\\0\\0\\0\\0\\29\\0(k\\24R\\17p\\1\\0\\0\\0...\"\n ```\n\nSolution:\n1. `shorten_fnames(true)` after restoring the CWD. Drop `cs_save_sfname`,\n which was a partial workaround for the same bug.\n2. Release the id in `msg_multihl()`, where it is set and the caller's\n frame is still alive. Stop reading `.data.integer` out of a String\n union member.", "style": "conventional", "type": "fix", "scope": "cwd", "breaking": false} {"repo": "tmux", "sha": "4e2abff2d67d229c9ef4778924fed3f79b325274", "date": "2026-08-22T21:28:37+01:00", "subject": "window-copy: avoid erasing and needlessly re-flashing images during copy-mode redraws", "body": "window_copy_write_one() used to write text/highlight styling directly\nover image-covered cells via the normal path, which - since a character\nwrite typically clears whatever pixel content a terminal was showing\nthere - erased the image with nothing to redraw it back in. Image-covered\ncells now bypass styling and go straight into the grid instead (a\nnon-graphical client's ASCII fallback for the image still depends on this\nsame write, so it can't just be skipped).\n\nwindow_copy_write_line()'s call to image_redraw_area() used to fire\nunconditionally on every redraw, needlessly recompositing (and briefly\nflashing) the image on every single cursor step even after the erasure\nabove was fixed, even though nothing about the image had changed. It now\nonly fires when window_copy_update_image_refresh() (new) determines the\nview's underlying history position actually moved since the last redraw.\n\nAlso adds window_copy_visible_has_images(), used by a later commit's\nscroll-region fix as a bypass to a full redraw when the visible range\ncarries image data - line insert/delete only shifts character cells, not\nalready-sent image pixels, so scroll's normal fast path leaves stale or\nmissing image content behind.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "2a535f7d82bccc16121eb0b2df547dcb152ceee0", "date": "2026-08-22T21:30:01+01:00", "subject": "image: keep the scrolled-off portion of a placement instead of discarding it", "body": "image_write() handles an image taller than the pane by scrolling the\nscreen up to make room, but only ever created spans for the rows that\nended up on screen afterwards - the rows immediately scrolled off (up to\norigin_y of them) were pushed into history as blank, spanless rows and\nthat part of the image was permanently discarded. Unlike width, which has\nno \"scroll right\" to recover a permanent clip, height already has\nordinary scrollback, so this was pure waste: scrolling back up just\nshowed empty space where part of the image used to be.\n\nimage_write() now also calls image_extend_row() for those origin_y rows,\ncapped to however many real history rows screen_write_scrollup() actually\ncreated (it clamps its own request to one scroll-region height per call,\nso a single call may not push everywhere origin_y implies - looping it\nuntil the full amount is pushed, done here too, makes that gap rare, but\nthe cap keeps the fill-in safe even if it isn't).", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "9ca3dfba7989ab62e172cfd611d8acdfd56606ac", "date": "2026-08-22T17:45:54-04:00", "subject": "Provide a C-ctype variant expected file for test_regex_utf8", "body": "test_regex_utf8 decides whether to run by looking at the database\nencoding alone, but two of its cases, [[:graph:]] and [[:print:]] over\nE'xᔀሷ', depend on the ctype as well. In a database with encoding\nUTF8 and locale C they match just the x, because isgraph() and isprint()\nare false for anything outside ASCII, and the file fails.\n\nNo buildfarm animal builds such a cluster, which is why this went\nunnoticed. A pending buildfarm client change will let an\nanimal be configured that way.\n\nFix by providing a second expected file holding the C-ctype answers, in\nthe manner of json_encoding.sql, which carries expected files for UTF8\nand for SQL_ASCII. The .sql needs no change.\n\nThis applies to releases 15 and 16 only. A better solution is available\nfor release 17 and up.\n\nBackpatch to v15, v16 only.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "33b8a42dfba090cf140f8e5d01c4eddf2c08100c", "date": "2026-08-22T17:47:24-04:00", "subject": "Pin two ctype-dependent test_regex_utf8 cases to pg_c_utf8", "body": "test_regex_utf8 decides whether to run by looking at the database\nencoding alone, but two of its cases, [[:graph:]] and [[:print:]] over\nE'xᔀሷ', depend on the ctype as well. In a database with encoding\nUTF8 and locale C they match just the x, because isgraph() and isprint()\nare false for anything outside ASCII, and the file fails.\n\nNo buildfarm animal builds such a cluster, which is why this went\nunnoticed, and why the to_date() crash in 18.5 went undetected for want\nof exactly this coverage. A pending buildfarm client change will let an\nanimal be configured that way.\n\nFix by giving the two cases an explicit collation, so that they exercise\na fixed Unicode ctype instead of whatever the database happened to be\ninitialized with. test_regex() already passes its input collation down\nto the regex compiler. The expected results are unchanged; only the\nechoed queries differ.\n\nBackpatch-through: 17 (15 and 16 get a different fix)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "f534e06c70ccea81581a89a8e4edb87765dffd80", "date": "2026-08-23T09:22:52+08:00", "subject": "test(netrw): fix failure on Windows", "body": "", "style": "conventional", "type": "test", "scope": "netrw", "breaking": false} {"repo": "neovim", "sha": "9744c76af0f5da01b04b4a39a381621ffe94a4ea", "date": "2026-08-22T22:19:41-04:00", "subject": "build(vim-patch): v8.2.2331 is n/a", "body": "Vim9script variable declaration via const/final", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "vite", "sha": "ea5ac7c73bd1a1ac9f5363ea09f98566d5d3b483", "date": "2026-08-23T12:21:35+08:00", "subject": "ci: fix bot comment formatting", "body": "", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "openssl", "sha": "84ad0cc3fef8af6724b80803092cfa58acc168cc", "date": "2026-08-23T11:46:36+07:00", "subject": "Fix tick race in radix fault-injection scripts", "body": "Guard any `OP_SET_INJECT_WORD`+`OP_WRITE` with 2 new functions `OP_ENGINE_TICK_ENABLE`/`OP_ENGINE_TICK_DISABLE` that enable/disable ticking in the QUIC engine. Otherwise, the radix test interpreter can tick between the 2 ops and cause the peer to close the connection prematurely\n\nReviewed-by: Saša Nedvědický \nReviewed-by: Norbert Pocs \nMerge-date: Tue Sep 15 08:53:14 2026\nMerged-from: https://github.com/openssl/openssl/pull/32469", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "8770c49de4d992aa901157bc0d78de88eb12880e", "date": "2026-08-23T09:02:27+03:00", "subject": "docs: remove doubled word in SECURITY-ADVISORY.md", "body": "Closes #22642", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vite", "sha": "7bdad37a4a772936d24e0116ac32a1f09fd8a055", "date": "2026-08-23T08:30:22+02:00", "subject": "chore: add 97ad042 to .git-blame-ignore-revs", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "neovim", "sha": "805bf3c81cd886cced71090a9d6d93064398f790", "date": "2026-08-23T02:39:18-04:00", "subject": "build(vim-patch): v9.0.0627 is n/a", "body": "", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "openssl", "sha": "3dc109a97a9f793c040c130ce9d2de4db7d475c7", "date": "2026-08-23T10:39:26+03:00", "subject": "docs: fix doubled words in man pages and design docs", "body": "Fix 23 occurrences of accidentally doubled words like 'in in',\n'the the', 'been been' across man1/man3/man7 pages, internal\ndocumentation and design documents.\n\nCLA: trivial\nReviewed-by: Frederik Wedel-Heinen \nReviewed-by: Paul Dale \nMerge-date: Thu Aug 27 16:00:39 2026\nMerged-from: https://github.com/openssl/openssl/pull/32461\n(cherry picked from commit 6b736d031e9bc7df908292f7f0a09ed5fd562f5a)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "6b736d031e9bc7df908292f7f0a09ed5fd562f5a", "date": "2026-08-23T10:39:26+03:00", "subject": "docs: fix doubled words in man pages and design docs", "body": "Fix 23 occurrences of accidentally doubled words like 'in in',\n'the the', 'been been' across man1/man3/man7 pages, internal\ndocumentation and design documents.\n\nCLA: trivial\nReviewed-by: Frederik Wedel-Heinen \nReviewed-by: Paul Dale \nMerge-date: Thu Aug 27 16:00:39 2026\nMerged-from: https://github.com/openssl/openssl/pull/32461", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "neovim", "sha": "ebbf27928919403a898f27d781a52c4358016c39", "date": "2026-08-23T00:55:09-07:00", "subject": "fix(build): fix wasmtime version regex", "body": "Problem: Regex assumes a single-digit patch number, but LTS wasmtime releases (which we track) can run into double digits.\n\nSolution: Allow for multi-digit patch numbers.\n(cherry picked from commit 1deeff9c3898701255be467614d84180accd78b9)", "style": "conventional", "type": "fix", "scope": "build", "breaking": false} {"repo": "neovim", "sha": "1deeff9c3898701255be467614d84180accd78b9", "date": "2026-08-23T00:55:09-07:00", "subject": "fix(build): fix wasmtime version regex", "body": "Problem: Regex assumes a single-digit patch number, but LTS wasmtime releases (which we track) can run into double digits.\n\nSolution: Allow for multi-digit patch numbers.", "style": "conventional", "type": "fix", "scope": "build", "breaking": false} {"repo": "curl", "sha": "81ade541200869e07880c59183a578d7102517d2", "date": "2026-08-23T14:42:03+05:30", "subject": "altsvc: continue after unknown parameters", "body": "Parse extension names and values before deciding whether a parameter is\nknown. This prevents long names and token or quoted-string values from\nhiding later ma or persist parameters, as required by RFC 7838 section\n3.\n\nAdd unit coverage for long extension names and both value forms,\nincluding an escaped quote and semicolon inside a quoted string.\n\nFixes #22644\nCloses #22676", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "4f9d2f4dadc83191200de7d2154c957a711e8c3d", "date": "2026-08-23T15:37:51+05:30", "subject": "fix(css): keep newline-separated srcset candidates intact", "body": "", "style": "conventional", "type": "fix", "scope": "css", "breaking": false} {"repo": "angular", "sha": "787a8e97f1282f4e15c59b322b80962ddd480cf6", "date": "2026-08-23T13:18:42+03:00", "subject": "refactor(forms): type async validator subscription", "body": "Use the Subscription type returned by Observable.subscribe instead of storing async validator subscriptions as any.\n\n(cherry picked from commit a4a428ece246ca2fa11fe51e8d66ae6b697502b9)", "style": "conventional", "type": "refactor", "scope": "forms", "breaking": false} {"repo": "angular", "sha": "a4a428ece246ca2fa11fe51e8d66ae6b697502b9", "date": "2026-08-23T13:18:42+03:00", "subject": "refactor(forms): type async validator subscription", "body": "Use the Subscription type returned by Observable.subscribe instead of storing async validator subscriptions as any.", "style": "conventional", "type": "refactor", "scope": "forms", "breaking": false} {"repo": "angular", "sha": "5dfe753bfc3adf86da09a523e63478c63721daf4", "date": "2026-08-23T13:36:07+03:00", "subject": "refactor(router): replace internal any types", "body": "Infer URL parameter values as strings and type preloading work as Observable to match the values produced by these internal flows.\n\n(cherry picked from commit f275d289a0a999e91730c8e22553d771b5ab3fd5)", "style": "conventional", "type": "refactor", "scope": "router", "breaking": false} {"repo": "angular", "sha": "f275d289a0a999e91730c8e22553d771b5ab3fd5", "date": "2026-08-23T13:36:07+03:00", "subject": "refactor(router): replace internal any types", "body": "Infer URL parameter values as strings and type preloading work as Observable to match the values produced by these internal flows.", "style": "conventional", "type": "refactor", "scope": "router", "breaking": false} {"repo": "neovim", "sha": "8ce1aac477a33ab1fe2a0b590ff278518dc0e473", "date": "2026-08-23T06:13:41-05:00", "subject": "feat(pos): mutable vim.Pos, vim.Range fields #41318", "body": "Allow plugins to set the human-readable fields.", "style": "conventional", "type": "feat", "scope": "pos", "breaking": false} {"repo": "neovim", "sha": "042e33786130bef2da62ac8fc4678b76c2c038ad", "date": "2026-08-23T07:37:03-04:00", "subject": "fix(vim-patch): sync list_vimpatch_tokens(), list_vimpatch_numbers() #41440", "body": "Developer and CI may think a Vim patch is merged\nbut list_vimpatch_numbers() and the \"version.c\" update suggest\notherwise.", "style": "conventional", "type": "fix", "scope": "vim-patch", "breaking": false} {"repo": "angular", "sha": "dab363011c04806030243b189101774d47c061fb", "date": "2026-08-23T14:37:43+02:00", "subject": "build: enable stableTypeOrdering tsconfig flag", "body": "`stableTypeOrdering` was recently enabled in the internal builds. While there aren't any breakages in Angular, these changes enable it so we can catch potential issues earlier.\n\n(cherry picked from commit c18261f3fd87f70a0c8cce16c64c9bd38d16e6c7)", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "angular", "sha": "c18261f3fd87f70a0c8cce16c64c9bd38d16e6c7", "date": "2026-08-23T14:37:43+02:00", "subject": "build: enable stableTypeOrdering tsconfig flag", "body": "`stableTypeOrdering` was recently enabled in the internal builds. While there aren't any breakages in Angular, these changes enable it so we can catch potential issues earlier.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "babel", "sha": "c813803e5c055cde4843bd288bcb36a867b2cab4", "date": "2026-08-23T10:27:11-04:00", "subject": "Disallow new import prop access", "body": "* test: fix incorrect link in test262 allowlist.md\n\n* fix(parser): disallow new import() with prop\n\n* test: update test262 allowlist", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "4080ace060477c8a6b4d6a20f937a3c960c2f25c", "date": "2026-08-23T16:46:21+02:00", "subject": "hsts: only match the exact strings", "body": "Fixes #22643\nCloses #22677", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "a199cc54cd61c9af1ea7b4415fbc9a53f4e8d7f0", "date": "2026-08-23T16:46:21+02:00", "subject": "hsts: only match the exact strings", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "85c8f6f3a5b107ee605122019cea4fb4edcfe61d", "date": "2026-08-23T17:03:06+02:00", "subject": "cookie: allow loading cookies for localhost", "body": "The added code to avoid loading PSL cookies from files from PR #22500\n(c041895, not shipped in a release), calls psl_is_public_suffix(),\nreturns TRUE also for \"localhost\" (!).\n\nWe now allow cookies again for localhost. Added test 320 to verify.\n\nReported-by: Athos Ribeiro\nFixes #22645\nCloses #22646", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "bca9c4afa77fc8396d0ab63511bfedb9dcdd5c89", "date": "2026-08-23T20:58:22+05:30", "subject": "fix(man): allow :Man to open by absolute path #41401", "body": "Problem:\nman -w echoes any existent file back unchanged. :Man rejected all such\npaths, so :Man /usr/share/man/man1/bash.1 failed even for real man pages.\n\nSolution:\nAccept an echoed path when it looks like a man page (.../man1/foo.1).", "style": "conventional", "type": "fix", "scope": "man", "breaking": false} {"repo": "neovim", "sha": "c5cb0350a6dc578e6ca73eefdd58159e1e039cbf", "date": "2026-08-23T21:00:17+05:30", "subject": "fix(lsp): do not expand $VAR in tagfunc filenames #41398", "body": "Problem:\nCTRL-] via lsp tagfunc passes URI paths through :tag, which expands\n\\$VAR, so files like people_.\\$personId.tsx fail with E429.\n\nSolution:\nfnameescape() the filename in tag items so \\$ is treated as literal.", "style": "conventional", "type": "fix", "scope": "lsp", "breaking": false} {"repo": "neovim", "sha": "211bef231519be3e71adc176b078c0f981e861ae", "date": "2026-08-24T01:34:44+10:00", "subject": "docs(dir): how to decorate a listing #41315", "body": "", "style": "conventional", "type": "docs", "scope": "dir", "breaking": false} {"repo": "angular", "sha": "1123cb54f7fc13f032841a4d75c3bfaaaa7d2359", "date": "2026-08-23T17:45:12+02:00", "subject": "docs: add missing tap import and AppConfig interface to environment configuration reference", "body": "(cherry picked from commit d953d2eda90809f228ddf012c7e0f229f127e3ab)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "d953d2eda90809f228ddf012c7e0f229f127e3ab", "date": "2026-08-23T17:45:12+02:00", "subject": "docs: add missing tap import and AppConfig interface to environment configuration reference", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "528e3b97bcf1ab9d564adc1826f2af8d30125b7b", "date": "2026-08-23T17:52:37+02:00", "subject": "docs: modernize creating-services reference with signal state and asReadonly", "body": "(cherry picked from commit 15cf923f88f23f4c278c806c16196148b4baa88d)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "15cf923f88f23f4c278c806c16196148b4baa88d", "date": "2026-08-23T17:52:37+02:00", "subject": "docs: modernize creating-services reference with signal state and asReadonly", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "neovim", "sha": "354ea3bf2a3a11b69fb5eac78b6334e8131ce4c4", "date": "2026-08-23T12:38:07-04:00", "subject": "docs: misc #41452", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "postgres", "sha": "165aa5040a02e13011f07e3c6669ac0edf4f8258", "date": "2026-08-23T12:42:55-04:00", "subject": "Allow an aggregate's planner support function to be set via CREATE AGGREGATE.", "body": "Commit 42473b3b3 implemented SupportRequestSimplifyAggref, but failed\nto think about what infrastructure would be required for an extension\nto use that: there is no SQL-level mechanism for attaching a planner\nsupport function to an aggregate. That seems pretty critical for a\nfeature that's primarily intended to be used by extensions.\n\nTo fix, add a SUPPORT clause to CREATE AGGREGATE, and teach pg_dump\n(and thereby pg_upgrade) about dumping this aggregate property.\nWe don't need to touch ALTER AGGREGATE, because it's already the case\nthat you're supposed to use CREATE OR REPLACE AGGREGATE if you want\nto alter any aggregate-specific properties set by CREATE AGGREGATE.\n(Maybe at some point we'll think that that policy ought to change,\nbut I don't think this one feature moves the needle enough.)\n\nPer report from Andrei Lepikhov, who also provided some of the\nnew documentation text.\n\nReported-by: Andrei Lepikhov \nAuthor: Tom Lane \nReviewed-by: Andrei Lepikhov \nDiscussion: https://postgr.es/m/8f58c96d-d3c7-4c0f-9898-116f00eeaff6@gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "42ae9e43a802a702015db3d5016fe1aa9ed4ace1", "date": "2026-08-23T19:03:05+02:00", "subject": "build: version bump", "body": "", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "neovim", "sha": "bb82f9612cdbfb2ac6929167f041061e9e3d4a0a", "date": "2026-08-23T13:58:45-04:00", "subject": "fix(cmdatom): \"!\" operator hardcodes its range #41451", "body": "Problem:\nThe \"!\" operator stuffs its cmdline continuation (`:.,.+1!`), so its\nframe ends before capture (stuff pending) and the redo-prep disappears\nwith it.\n\n !ipsort => { type='excmd', lhs=':.,.+1!sort', keys=':.,.+1!sort' }\n\nCompare to builtin \".\" which works bc `do_bang()` completes the redo\n(`!ip` + `sort`).\n\nSolution:\nAppoint the stuffed continuation frame as the \"redo-prep\" frame.\n\n !ipsort => { type='operator', operator='!', lhs='!ipsort', keys='!ipsort' }\n\nNotes:\n- atom_cmd_end(): a frame ending with stuff pending re-points its\n redo-prep to the next frame.\n- atom_cmd_start(): a stuffed continuation frame (`KeyStuffed`) keeps\n the redo-prep; flushed stuff discards it.", "style": "conventional", "type": "fix", "scope": "cmdatom", "breaking": false} {"repo": "neovim", "sha": "69bf8c7792a45433515c6d409c5381ff7660d815", "date": "2026-08-23T15:04:11-04:00", "subject": "fix(cmdatom): terminal-mode keys leak into mapping lhs #41455", "body": "Problem:\nA mapping that enters terminal-mode (`:FzfLua files` via `:startinsert`)\nnever reaches a composite-end: terminal-mode runs no \"normal\" CmdFrames.\nThe composite collects the entire terminal session (and more) into `lhs`.\n\nE.g. if I have `` mapped to open `:FzfLua files`, then interact\nwith fzf-lua UI, the emitted CmdAtom looks like:\n\n … => { type='mapping', lhs='', keys=nil }\n\nSolution:\nEnd the composite when terminal-mode is entered, which emits a more\nmeaningful and repeatable atom:\n\n … => { type='mapping', lhs='', keys=nil }", "style": "conventional", "type": "fix", "scope": "cmdatom", "breaking": false} {"repo": "curl", "sha": "a334227e8ac74047c372c94bbb8950173f9a409c", "date": "2026-08-23T15:20:43-04:00", "subject": "tidy-up: fix typos in docs and comments", "body": "Correct grammar and spelling in documentation and source comments. Also\nfix the Aspell configuration key and one user-facing OpenSSL error\nmessage.\n\nCloses #22647", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "ba29223b387e470b9814b8dec730db1848fda1dd", "date": "2026-08-23T23:02:56+03:00", "subject": "refactor(docs-infra): drop exclude patterns for examples that no longer exist", "body": "The `embeddable` glob excludes nine example directories. Four of them,\n`testing`, `ssr`, `resolution-modifiers` and `dependency-injection`, were\nremoved by #66753 and #61686 without updating this file, so those patterns\nmatch nothing.\n\nThree of the four sit under a TODO about examples that do not compile. Those\nwere deleted rather than fixed, which leaves `reactive-forms` and\n`form-validation` as the only two the note still applies to.\n\n(cherry picked from commit c6a6ad0a44a9fa3d98fcf28adfe2c3912d69b267)", "style": "conventional", "type": "refactor", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "c6a6ad0a44a9fa3d98fcf28adfe2c3912d69b267", "date": "2026-08-23T23:02:56+03:00", "subject": "refactor(docs-infra): drop exclude patterns for examples that no longer exist", "body": "The `embeddable` glob excludes nine example directories. Four of them,\n`testing`, `ssr`, `resolution-modifiers` and `dependency-injection`, were\nremoved by #66753 and #61686 without updating this file, so those patterns\nmatch nothing.\n\nThree of the four sit under a TODO about examples that do not compile. Those\nwere deleted rather than fixed, which leaves `reactive-forms` and\n`form-validation` as the only two the note still applies to.", "style": "conventional", "type": "refactor", "scope": "docs-infra", "breaking": false} {"repo": "curl", "sha": "740d9520a9691b99be195f41408024f83abb1654", "date": "2026-08-23T22:57:36+02:00", "subject": "docs: mention possible auth option conflicts", "body": "Do not mix HTTP auth options with custom Authorization headers.\n\nCloses #22648", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "neovim", "sha": "716835c2325eee1e04e8c0694dbf61605cfa0eae", "date": "2026-08-24T04:39:26+05:30", "subject": "fix(ui2): dismiss msgs after mapped motions #41450", "body": "Problem:\nMapped keys can produce on_key callbacks with an empty `typed` value.\nso mapped motions such as `j -> gj` are not recognized as typed input\nand do not dismiss messages\n\nSolution:\nTrack whether an empty `typed` callback follows typed input, allowing\nmapped motions to dismiss messages like directly typed motions.", "style": "conventional", "type": "fix", "scope": "ui2", "breaking": false} {"repo": "neovim", "sha": "eeefe7ca65dfe6d32d17d2dd438bd8139c3dfa47", "date": "2026-08-23T19:32:07-04:00", "subject": "fix(cwd): E812 when a message opens a window during a file read #41458", "body": "Problem:\nWhen ui2 is enabled, opening an already-open file in another Neovim\ninstance results in the error `E812: Autocommands changed buffer…`.\n\nAnalysis:\nOn E812 the file is not loaded. The default SwapExists handler notifies\n`W325: Ignoring swapfile…`, and ui2 shows that message by opening\na window, which is a temp context switch. `ctx_dirs_restore()`\nre-shortens every buffer name on the way out, and `shorten_buf_fname()`\nalways frees and reallocates `b_sfname`. `readfile()` aliases `b_fname`\nacross `check_need_swap()` and compares the pointer to detect a rename.\n\nRegression by b296666e419c, which replaced the `cs_save_sfname` restore\n(that kept curbuf's pointer) with `shorten_fnames(true)`.\n\nSolution:\nKeep the allocation in `shorten_buf_fname()` when the short name is\nunchanged. Pointer stability is what the E200/E201/E812 guards actually\nassert.", "style": "conventional", "type": "fix", "scope": "cwd", "breaking": false} {"repo": "vite", "sha": "3e34883afff1f4823a406dafa3c9e449e80074a5", "date": "2026-08-24T08:00:39+08:00", "subject": "ci: narrow semantic-pull-request triggers", "body": "", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "vite", "sha": "5fdfd26ed10ffd08928a8d13b71650f3444221ca", "date": "2026-08-24T08:01:37+08:00", "subject": "chore: update contributing", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "290d7df6aae126f9fb5570bdce8aa355dbbbc600", "date": "2026-08-23T17:18:40-07:00", "subject": "refactor(runtime-vapor): insert hydration anchors synchronously", "body": "", "style": "conventional", "type": "refactor", "scope": "runtime-vapor", "breaking": false} {"repo": "ant-design", "sha": "dd159f456cede17980e8fa8d453a49980fb378f2", "date": "2026-08-24T08:29:48+08:00", "subject": "fix(Watermark): avoid drawing zero-sized canvas", "body": "", "style": "conventional", "type": "fix", "scope": "Watermark", "breaking": false} {"repo": "vue-core", "sha": "29ed4b0bcd774fc83a5a21587c1a26ff70a325f6", "date": "2026-08-23T17:54:18-07:00", "subject": "perf(runtime-vapor): cache template adopt target", "body": "v-for single-root +15.5%, multi-root +9.2%;", "style": "conventional", "type": "perf", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "e81ddcea27cda21139950f31a07e398a1c15b2c4", "date": "2026-08-23T18:22:37-07:00", "subject": "refactor(runtime-vapor): unify block hydration tail", "body": "", "style": "conventional", "type": "refactor", "scope": "runtime-vapor", "breaking": false} {"repo": "postgres", "sha": "89d091f602fa59326d7cdf51e134e5525ef80977", "date": "2026-08-24T10:48:36+09:00", "subject": "Move some node-read support functions from nodes.h to readfuncs.h", "body": "The functions moved in this commit have been exposed since c1ddd2361f6e,\nand there still seem to be a couple of projects out there that require\nthem, as they can be more efficient than nodeRead(). Note that all\nthese functions are only called in readfuncs.c in terms of the core\ncode.\n\nA follow-up patch plans to improve the in-core support of pg_strtok(),\nmaking it thread-safe. This change removes the need to expose internals\nspecific to node reads in nodes.h, which is a header with a more general\npurpose than readfuncs.h.\n\nBased on a suggestion by me.\n\nAuthor: Matthias van de Meent \nDiscussion: https://postgr.es/m/aofW6ntfWJjk_yBL@paquier.xyz", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "9905d7d6973b425fb808520e00ac4399e681f635", "date": "2026-08-23T19:09:49-07:00", "subject": "perf(runtime-vapor): avoid quadratic scan when leaving a hydration boundary", "body": "", "style": "conventional", "type": "perf", "scope": "runtime-vapor", "breaking": false} {"repo": "neovim", "sha": "15f6b7a7f545e4504c04fc4d93cf8bde356e49f7", "date": "2026-08-23T22:11:30-04:00", "subject": "build(vim-patch): vim9 uf_type_list is n/a", "body": "From v8.2.0149 .\nTarget v9.0.1605 .", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "vue-core", "sha": "0274379a7af01dd77fe8bbc1b2e87ecb0076bb32", "date": "2026-08-23T19:19:57-07:00", "subject": "refactor(runtime-vapor): name the dynamic anchor resolution rules", "body": "", "style": "conventional", "type": "refactor", "scope": "runtime-vapor", "breaking": false} {"repo": "neovim", "sha": "50094679655981ef480f23d5ab0d060b2e31190b", "date": "2026-08-23T22:26:38-04:00", "subject": "build(vim-patch): v9.0.1760 is n/a", "body": "", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "vue-core", "sha": "9454de7ec601f59b38126ea7bd000b005d454572", "date": "2026-08-23T19:35:13-07:00", "subject": "refactor(runtime-vapor): reify the slot hydration ledger as a session", "body": "", "style": "conventional", "type": "refactor", "scope": "runtime-vapor", "breaking": false} {"repo": "vite", "sha": "8b3b4894afd98a8807ef985b601410d40b795d79", "date": "2026-08-24T11:38:43+09:00", "subject": "test(css): fix css async order", "body": "", "style": "conventional", "type": "test", "scope": "css", "breaking": false} {"repo": "ant-design", "sha": "c8de04a82b5a90d7d1fd6245ecf7aaffa0aed290", "date": "2026-08-24T09:58:51+07:00", "subject": "fix(DirectoryTree): treat undefined expandedKeys as uncontrolled", "body": "", "style": "conventional", "type": "fix", "scope": "DirectoryTree", "breaking": false} {"repo": "vue-core", "sha": "aa4fa0016cc16760c593823b25d7506e73f0eacc", "date": "2026-08-23T20:06:16-07:00", "subject": "refactor(runtime-vapor): normalize the hydration cursor at its write funnel", "body": "", "style": "conventional", "type": "refactor", "scope": "runtime-vapor", "breaking": false} {"repo": "postgres", "sha": "9673a0aa92f4fd75c564e2c8e14d71500a7b55ae", "date": "2026-08-24T12:07:39+09:00", "subject": "Make stringToNode() infrastructure thread-safe", "body": "While stringToNode() was technically reentrant-safe, this was only for\nsingle-threaded workloads: the string decoded was stored in a global\nvariable. This may be unsafe depending on the execution model, and\ndefinitely unsafe in multi-threaded environments.\n\nThis patch moves the global variables into a struct (ReadNodeContext),\nwhose pointer is passed around on the stack. Support functions for node\nread, extensible nodes (note: these have never been tested, should be\ntested) and pg_strtok() are updated to handle the context data across\nthe call stack.\n\nCode simplifications are mostly in stringToNodeInternal(), that can rely\non the context structure rather than a save-and-reassign logic based on\nthe global variables previously available.\n\nAuthor: Matthias van de Meent \nReviewed-by: Peter Eisentraut \nReviewed-by: Michael Paquier \nReviewed-by: Yuhang Qiu \nDiscussion: https://postgr.es/m/CAEze2WjsGnOM4MB+n1LLdhhUg6r=hwWZRv+-heDE4UtzRdgjyg@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "8f8b07edaec13659031fd744220f13496c17e4f3", "date": "2026-08-23T20:32:29-07:00", "subject": "refactor(runtime-vapor): apply hydration series cleanup review", "body": "", "style": "conventional", "type": "refactor", "scope": "runtime-vapor", "breaking": false} {"repo": "vite", "sha": "c31c499ca04c7557d9093291b08b6e3087df0542", "date": "2026-08-23T20:42:27-07:00", "subject": "feat(create-vite): add nub package manager support", "body": "", "style": "conventional", "type": "feat", "scope": "create-vite", "breaking": false} {"repo": "vite", "sha": "21a3ee9575702536568149dcfe18f409ea447bfc", "date": "2026-08-24T11:43:17+08:00", "subject": "chore: remove outdated `minimumReleaseAgeExclude` config", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "angular", "sha": "5e6f3445dc00d5610a916f3b0d88b843ea7c44a2", "date": "2026-08-24T00:11:47-04:00", "subject": "refactor(devtools): update profiler instruction text for clearing recordings", "body": "I suspect this text used to be correct but became outdated after UI changes. These is no \"refresh\" button.\n\n(cherry picked from commit 434acc506e0fde854db409edc4b11445c370f35d)", "style": "conventional", "type": "refactor", "scope": "devtools", "breaking": false} {"repo": "angular", "sha": "434acc506e0fde854db409edc4b11445c370f35d", "date": "2026-08-24T00:11:47-04:00", "subject": "refactor(devtools): update profiler instruction text for clearing recordings", "body": "I suspect this text used to be correct but became outdated after UI changes. These is no \"refresh\" button.", "style": "conventional", "type": "refactor", "scope": "devtools", "breaking": false} {"repo": "vite", "sha": "2e8355d67c19c68e867ee633121fc91e007c846c", "date": "2026-08-24T12:13:10+08:00", "subject": "chore(docs): remove markdown-it-image-size", "body": "", "style": "conventional", "type": "chore", "scope": "docs", "breaking": false} {"repo": "angular", "sha": "950603097b6a94bf76f1ff018e4471b11ee5bcb7", "date": "2026-08-24T00:14:59-04:00", "subject": "perf(devtools): stop Map copying in handleFilter", "body": "Before: handleFilter cloned the entire matchedNodes Map on every match inside the iteration loop.\n\nAfter: handleFilter accumulates matches in a local Map and performs a single atomic update to matchedNodes at the end of the loop in O(n) time.", "style": "conventional", "type": "perf", "scope": "devtools", "breaking": false} {"repo": "openssl", "sha": "7bba68cafbfd9e96780fad8ae1f438d5cfb92388", "date": "2026-08-24T07:11:05+02:00", "subject": "CHANGES.md: restore removed \"OpenSSL 4.1\" header", "body": "For some inconceivable reason, the header got removed in commit\n2357b35ff7b8 \"Ensure TLS 1.3 ciphersuites are actually for TLS 1.3\"\nwithout any justification, or even mention. Restore it.\n\nFixes: 2357b35ff7b8 \"Ensure TLS 1.3 ciphersuites are actually for TLS 1.3\"\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Tomas Mraz \nReviewed-by: Andrew Dinh \nMerge-date: Fri Aug 28 11:16:06 2026\nMerged-from: https://github.com/openssl/openssl/pull/32484", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "f2f9bd7251db765d1a9c6f434e247f167912ce11", "date": "2026-08-24T07:11:05+02:00", "subject": "CHANGES.md: restore removed \"OpenSSL 4.0\" header", "body": "For some inconceivable reason, the header got removed in commit\n2357b35ff7b8 \"Ensure TLS 1.3 ciphersuites are actually for TLS 1.3\"\nwithout any justification, or even mention. Restore it.\n\nFixes: 2357b35ff7b8 \"Ensure TLS 1.3 ciphersuites are actually for TLS 1.3\"\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nReviewed-by: Bob Beck \nMerge-date: Tue Aug 25 11:33:46 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "19fbbd80b5cd4a503afab8004fcca8f4a76d230f", "date": "2026-08-24T07:17:05+02:00", "subject": "CHANGES.md: rephrase TLS 1.3 ticket lifetime cap change log record", "body": "Complements: 21c92ae589af \"Enforce RFC 8446 ticket lifetime limit for TLS 1.3 client\"\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Tomas Mraz \nReviewed-by: Andrew Dinh \nMerge-date: Fri Aug 28 11:16:07 2026\nMerged-from: https://github.com/openssl/openssl/pull/32484", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "898880d642310a92564acfde91cbf5a3f8a61e48", "date": "2026-08-24T07:17:05+02:00", "subject": "CHANGES.md: rephrase TLS 1.3 ticket lifetime cap change log record", "body": "Complements: 7e82c32134aa \"Enforce RFC 8446 ticket lifetime limit for TLS 1.3 client\"\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nReviewed-by: Bob Beck \nMerge-date: Tue Aug 25 11:38:30 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "46fbd9dc042d15c486f4302b356649a082283b67", "date": "2026-08-24T07:17:05+02:00", "subject": "CHANGES.md: rephrase TLS 1.3 ticket lifetime cap change log record", "body": "Complements: 601399c31b43 \"Enforce RFC 8446 ticket lifetime limit for TLS 1.3 client\"\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nReviewed-by: Bob Beck \nMerge-date: Tue Aug 25 11:37:40 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "9bde5cc62e3394e44d9d2160a9de5f0a16d1571d", "date": "2026-08-24T07:17:05+02:00", "subject": "CHANGES.md: rephrase TLS 1.3 ticket lifetime cap change log record", "body": "Complements: b0576dc5d422 \"Enforce RFC 8446 ticket lifetime limit for TLS 1.3 client\"\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nReviewed-by: Bob Beck \nMerge-date: Tue Aug 25 11:36:54 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "e0640cc881fafa1c7458c7104f497d7f9d79c35a", "date": "2026-08-24T07:17:05+02:00", "subject": "CHANGES.md: rephrase TLS 1.3 ticket lifetime cap change log record", "body": "Complements: 7aa9ffab4525 \"Enforce RFC 8446 ticket lifetime limit for TLS 1.3 client\"\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nReviewed-by: Bob Beck \nMerge-date: Tue Aug 25 11:35:47 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "7c89950df42713cbfe60460b92b5485dbd89a4ac", "date": "2026-08-24T07:17:05+02:00", "subject": "CHANGES.md: rephrase TLS 1.3 ticket lifetime cap change log record", "body": "Complements: 21c92ae589af \"Enforce RFC 8446 ticket lifetime limit for TLS 1.3 client\"\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nReviewed-by: Bob Beck \nMerge-date: Tue Aug 25 11:33:47 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "9e2dcd506b2a23e970d3ad20785b28b94500cbe5", "date": "2026-08-24T07:20:12+02:00", "subject": "CHANGES.md, NEWS.md: reorder HollowByte change log record, add missing period", "body": "Move HollowByte change log record after records with assigned CVE\nnumbers and severity ratings, add period at the end of the lead\nsentence, add a link to the PR in the comment.\n\nComplements: c4c5d2bbd7ab \"Update CHANGES/NEWS to mention the HollowByte fix\"\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Tomas Mraz \nReviewed-by: Andrew Dinh \nMerge-date: Fri Aug 28 11:16:08 2026\nMerged-from: https://github.com/openssl/openssl/pull/32484", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "c3357d35866ac3c21f84d5fce11aa2f3b6aa839b", "date": "2026-08-24T07:20:12+02:00", "subject": "CHANGES.md, NEWS.md: reorder HollowByte change log record, add missing period", "body": "Move HollowByte change log record after records with assigned CVE\nnumbers and severity ratings, add period at the end of the lead\nsentence, add a link to the PR in the comment.\n\nComplements: d01f86f21948 \"Update CHANGES/NEWS to mention the HollowByte fix\"\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nReviewed-by: Bob Beck \nMerge-date: Tue Aug 25 11:38:32 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "618369fa587c2224b2348906438081911505260f", "date": "2026-08-24T07:20:12+02:00", "subject": "CHANGES.md, NEWS.md: reorder HollowByte change log record, add missing period", "body": "Move HollowByte change log record after records with assigned CVE\nnumbers and severity ratings, add period at the end of the lead\nsentence, add a link to the PR in the comment.\n\nComplements: c500fbd8a446 \"Update CHANGES/NEWS to mention the HollowByte fix\"\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nReviewed-by: Bob Beck \nMerge-date: Tue Aug 25 11:37:41 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "ffdd18b969117556d20852288c7c2d2ca1493e89", "date": "2026-08-24T07:20:12+02:00", "subject": "CHANGES.md, NEWS.md: reorder HollowByte change log record, add missing period", "body": "Move HollowByte change log record after records with assigned CVE\nnumbers and severity ratings, add period at the end of the lead\nsentence, add a link to the PR in the comment.\n\nComplements: 4db90ebab383 \"Update CHANGES/NEWS to mention the HollowByte fix\"\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nReviewed-by: Bob Beck \nMerge-date: Tue Aug 25 11:36:55 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "a0a2ec74062a2a5bf9fa4e2564eaee473d7dae91", "date": "2026-08-24T07:20:12+02:00", "subject": "CHANGES.md, NEWS.md: reorder HollowByte change log record, add missing period", "body": "Move HollowByte change log record after records with assigned CVE\nnumbers and severity ratings, add period at the end of the lead\nsentence, add a link to the PR in the comment.\n\nComplements: 659de2429f07 \"Update CHANGES/NEWS to mention the HollowByte fix\"\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nReviewed-by: Bob Beck \nMerge-date: Tue Aug 25 11:35:49 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "c8dbd23e216cefe4bd92372b4f7b4ae5265dbe68", "date": "2026-08-24T07:20:12+02:00", "subject": "CHANGES.md, NEWS.md: reorder HollowByte change log record, add missing period", "body": "Move HollowByte change log record after records with assigned CVE\nnumbers and severity ratings, add period at the end of the lead\nsentence, add a link to the PR in the comment.\n\nComplements: c4c5d2bbd7ab \"Update CHANGES/NEWS to mention the HollowByte fix\"\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nReviewed-by: Bob Beck \nMerge-date: Tue Aug 25 11:33:48 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "b350fb26856456235503f95ec40e19de10660a0e", "date": "2026-08-24T07:24:24+02:00", "subject": "CHANGES.md, NEWS.md: update for 4.0.2", "body": "4.0.2 CHANGES.md includes the following:\n * CVE-2026-14456, CVE-2026-14457, CVE-2026-18798, CVE-2026-54874,\n CVE-2026-54876, CVE-2026-63072, CVE-2026-63073, CVE-2026-63074,\n CVE-2026-63075, CVE-2026-63076, CVE-2026-75803\n * https://github.com/openssl/openssl/pull/31174\n \"Enforce RFC 8446 ticket lifetime limit for TLS 1.3 client\"\n (already present)\n * https://github.com/openssl/openssl/pull/31572\n \"[4.0, 3.6, 3.5, 3.4] Add icx compiler version support in perl asm scripts\"\n * https://github.com/openssl/openssl/pull/31749\n \"Add documentation for OPENSSL_armcap\"\n * https://github.com/openssl/openssl/pull/31764\n \"x509: fix OCSP BasicResponse leak during verification\"\n (included as CVE-2026-54876)\n * https://github.com/openssl/openssl/pull/32052\n \"QUIC server: limit the number of pending connections\"\n (included as CVE-2026-14456)\n * https://github.com/openssl/openssl/pull/32300\n \"[4.0] Reject explicitly supplied invalid tags and generate tags for empty AEAD messages\"\n (included as CVE-2026-75803)\n * https://github.com/openssl/openssl/pull/32427\n \"Backport #32256 openssl 4.0 to 3.0\"\n\n4.0.2 NEWS.md includes the following:\n * CVE-2026-14456, CVE-2026-14457, CVE-2026-18798, CVE-2026-54874,\n CVE-2026-54876, CVE-2026-63072, CVE-2026-63073, CVE-2026-63074,\n CVE-2026-63075, CVE-2026-63076, CVE-2026-75803\n * https://github.com/openssl/openssl/pull/31764\n \"x509: fix OCSP BasicResponse leak during verification\"\n (included as CVE-2026-54876)\n * https://github.com/openssl/openssl/pull/32052\n \"QUIC server: limit the number of pending connections\"\n (included as CVE-2026-14456)\n * https://github.com/openssl/openssl/pull/32300\n \"[4.0] Reject explicitly supplied invalid tags and generate tags for empty AEAD messages\"\n (included as CVE-2026-75803)\n * https://github.com/openssl/openssl/pull/32427\n \"Backport #32256 openssl 4.0 to 3.0\"\n\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Tomas Mraz \nReviewed-", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "2f6e1a9f7036965b7a4e0bdf60847a542bf21213", "date": "2026-08-24T07:24:24+02:00", "subject": "CHANGES.md, NEWS.md: update for 3.0.22", "body": "3.0.22 CHANGES.md includes the following:\n * CVE-2026-54874, CVE-2026-63072, CVE-2026-63074, CVE-2026-63076,\n CVE-2026-75803\n * https://github.com/openssl/openssl/pull/31174\n \"Enforce RFC 8446 ticket lifetime limit for TLS 1.3 client\"\n (already present)\n * https://github.com/openssl/openssl/pull/31578\n \"Backport PR #30313 into 3.0 branch.\"\n * https://github.com/openssl/openssl/pull/32417\n \"[3.0] Reject explicitly supplied invalid tags and generate tags for empty AEAD messages\"\n (included as CVE-2026-75803)\n * https://github.com/openssl/openssl/pull/32427\n \"Backport #32256 openssl 4.0 to 3.0\"\n\n3.0.22 NEWS.md includes the following:\n * CVE-2026-54874, CVE-2026-63072, CVE-2026-63074, CVE-2026-63076,\n CVE-2026-75803\n * https://github.com/openssl/openssl/pull/32417\n \"[3.0] Reject explicitly supplied invalid tags and generate tags for empty AEAD messages\"\n (included as CVE-2026-75803)\n * https://github.com/openssl/openssl/pull/32427\n \"Backport #32256 openssl 4.0 to 3.0\"\n\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nReviewed-by: Bob Beck \nMerge-date: Tue Aug 25 11:38:33 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "3f4d9fa0a489aeff3e42245a10d9a33370144a91", "date": "2026-08-24T07:24:24+02:00", "subject": "CHANGES.md, NEWS.md: update for 3.4.7", "body": "3.4.7 CHANGES.md includes the following:\n * CVE-2026-14457, CVE-2026-54874, CVE-2026-63072, CVE-2026-63073,\n CVE-2026-63074, CVE-2026-63075, CVE-2026-63076, CVE-2026-75803\n * https://github.com/openssl/openssl/pull/31174\n \"Enforce RFC 8446 ticket lifetime limit for TLS 1.3 client\"\n (already present)\n * https://github.com/openssl/openssl/pull/31572\n \"[4.0, 3.6, 3.5, 3.4] Add icx compiler version support in perl asm scripts\"\n * https://github.com/openssl/openssl/pull/32416\n \"[3.5,3.4] Reject explicitly supplied invalid tags and generate tags for empty AEAD messages\"\n (included as CVE-2026-75803)\n * https://github.com/openssl/openssl/pull/32427\n \"Backport #32256 openssl 4.0 to 3.0\"\n\n3.4.7 NEWS.md includes the following:\n * CVE-2026-14457, CVE-2026-54874, CVE-2026-63072, CVE-2026-63073,\n CVE-2026-63074, CVE-2026-63075, CVE-2026-63076, CVE-2026-75803\n * https://github.com/openssl/openssl/pull/32416\n \"[3.5,3.4] Reject explicitly supplied invalid tags and generate tags for empty AEAD messages\"\n (included as CVE-2026-75803)\n * https://github.com/openssl/openssl/pull/32427\n \"Backport #32256 openssl 4.0 to 3.0\"\n\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nReviewed-by: Bob Beck \nMerge-date: Tue Aug 25 11:37:42 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "4d79d67312c6cf8c0d620be2653fab0ffedabede", "date": "2026-08-24T07:24:24+02:00", "subject": "CHANGES.md, NEWS.md: update for 3.5.8", "body": "3.5.8 CHANGES.md includes the following:\n * CVE-2026-14456, CVE-2026-14457, CVE-2026-18798, CVE-2026-54874,\n CVE-2026-63072, CVE-2026-63073, CVE-2026-63074, CVE-2026-63075,\n CVE-2026-63076, CVE-2026-75803\n * https://github.com/openssl/openssl/pull/31174\n \"Enforce RFC 8446 ticket lifetime limit for TLS 1.3 client\"\n (already present)\n * https://github.com/openssl/openssl/pull/31572\n \"[4.0, 3.6, 3.5, 3.4] Add icx compiler version support in perl asm scripts\"\n * https://github.com/openssl/openssl/pull/31749\n \"Add documentation for OPENSSL_armcap\"\n * https://github.com/openssl/openssl/pull/32052\n \"QUIC server: limit the number of pending connections\"\n (included as CVE-2026-14456)\n * https://github.com/openssl/openssl/pull/32416\n \"[3.5,3.4] Reject explicitly supplied invalid tags and generate tags for empty AEAD messages\"\n (included as CVE-2026-75803)\n * https://github.com/openssl/openssl/pull/32427\n \"Backport #32256 openssl 4.0 to 3.0\"\n\n3.5.8 NEWS.md includes the following:\n * CVE-2026-14456, CVE-2026-14457, CVE-2026-18798, CVE-2026-54874,\n CVE-2026-63072, CVE-2026-63073, CVE-2026-63074, CVE-2026-63075,\n CVE-2026-63076, CVE-2026-75803\n * https://github.com/openssl/openssl/pull/32052\n \"QUIC server: limit the number of pending connections\"\n (included as CVE-2026-14456)\n * https://github.com/openssl/openssl/pull/32416\n \"[3.5,3.4] Reject explicitly supplied invalid tags and generate tags for empty AEAD messages\"\n (included as CVE-2026-75803)\n * https://github.com/openssl/openssl/pull/32427\n \"Backport #32256 openssl 4.0 to 3.0\"\n\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nReviewed-by: Bob Beck \nMerge-date: Tue Aug 25 11:36:56 2026", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "e3004fd1e308fdc0792b3bee6988ebf7c5dad160", "date": "2026-08-24T07:24:24+02:00", "subject": "CHANGES.md, NEWS.md: update for 3.6.4", "body": "3.6.4 CHANGES.md includes the following:\n * CVE-2026-14456, CVE-2026-14457, CVE-2026-18798, CVE-2026-54874,\n CVE-2026-54876, CVE-2026-63072, CVE-2026-63073, CVE-2026-63074,\n CVE-2026-63075, CVE-2026-63076, CVE-2026-75803\n * https://github.com/openssl/openssl/pull/31174\n \"Enforce RFC 8446 ticket lifetime limit for TLS 1.3 client\"\n (already present)\n * https://github.com/openssl/openssl/pull/31572\n \"[4.0, 3.6, 3.5, 3.4] Add icx compiler version support in perl asm scripts\"\n * https://github.com/openssl/openssl/pull/31749\n \"Add documentation for OPENSSL_armcap\"\n * https://github.com/openssl/openssl/pull/31764\n \"x509: fix OCSP BasicResponse leak during verification\"\n (included as CVE-2026-54876)\n * https://github.com/openssl/openssl/pull/32052\n \"QUIC server: limit the number of pending connections\"\n (included as CVE-2026-14456)\n * https://github.com/openssl/openssl/pull/32259\n \"[3.6] Reject explicitly supplied invalid tags and generate tags for empty AEAD messages\"\n (included as CVE-2026-75803)\n * https://github.com/openssl/openssl/pull/32427\n \"Backport #32256 openssl 4.0 to 3.0\"\n\n3.6.4 NEWS.md includes the following:\n * CVE-2026-14456, CVE-2026-14457, CVE-2026-18798, CVE-2026-54874,\n CVE-2026-54876, CVE-2026-63072, CVE-2026-63073, CVE-2026-63074,\n CVE-2026-63075, CVE-2026-63076, CVE-2026-75803\n * https://github.com/openssl/openssl/pull/31764\n \"x509: fix OCSP BasicResponse leak during verification\"\n (included as CVE-2026-54876)\n * https://github.com/openssl/openssl/pull/32052\n \"QUIC server: limit the number of pending connections\"\n (included as CVE-2026-14456)\n * https://github.com/openssl/openssl/pull/32259\n \"[3.6] Reject explicitly supplied invalid tags and generate tags for empty AEAD messages\"\n (included as CVE-2026-75803)\n * https://github.com/openssl/openssl/pull/32427\n \"Backport #32256 openssl 4.0 to 3.0\"\n\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Milan Broz \nReviewed-by: Tom", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "05e9936df7f098b011209d046c71f0869c9de57d", "date": "2026-08-24T07:24:24+02:00", "subject": "CHANGES.md, NEWS.md: update for 4.0.2", "body": "4.0.2 CHANGES.md includes the following:\n * CVE-2026-14456, CVE-2026-14457, CVE-2026-18798, CVE-2026-54874,\n CVE-2026-54876, CVE-2026-63072, CVE-2026-63073, CVE-2026-63074,\n CVE-2026-63075, CVE-2026-63076, CVE-2026-75803\n * https://github.com/openssl/openssl/pull/31174\n \"Enforce RFC 8446 ticket lifetime limit for TLS 1.3 client\"\n (already present)\n * https://github.com/openssl/openssl/pull/31572\n \"[4.0, 3.6, 3.5, 3.4] Add icx compiler version support in perl asm scripts\"\n * https://github.com/openssl/openssl/pull/31749\n \"Add documentation for OPENSSL_armcap\"\n * https://github.com/openssl/openssl/pull/31764\n \"x509: fix OCSP BasicResponse leak during verification\"\n (included as CVE-2026-54876)\n * https://github.com/openssl/openssl/pull/32052\n \"QUIC server: limit the number of pending connections\"\n (included as CVE-2026-14456)\n * https://github.com/openssl/openssl/pull/32300\n \"[4.0] Reject explicitly supplied invalid tags and generate tags for empty AEAD messages\"\n (included as CVE-2026-75803)\n * https://github.com/openssl/openssl/pull/32427\n \"Backport #32256 openssl 4.0 to 3.0\"\n\n4.0.2 NEWS.md includes the following:\n * CVE-2026-14456, CVE-2026-14457, CVE-2026-18798, CVE-2026-54874,\n CVE-2026-54876, CVE-2026-63072, CVE-2026-63073, CVE-2026-63074,\n CVE-2026-63075, CVE-2026-63076, CVE-2026-75803\n * https://github.com/openssl/openssl/pull/31764\n \"x509: fix OCSP BasicResponse leak during verification\"\n (included as CVE-2026-54876)\n * https://github.com/openssl/openssl/pull/32052\n \"QUIC server: limit the number of pending connections\"\n (included as CVE-2026-14456)\n * https://github.com/openssl/openssl/pull/32300\n \"[4.0] Reject explicitly supplied invalid tags and generate tags for empty AEAD messages\"\n (included as CVE-2026-75803)\n * https://github.com/openssl/openssl/pull/32427\n \"Backport #32256 openssl 4.0 to 3.0\"\n\nSigned-off-by: Eugene Syromiatnikov \n\n14456\n\nReviewed-by: Milan Broz \nReviewed", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "3d44f75d0959313b1fc6a4e556d89927308a8cbe", "date": "2026-08-24T07:05:23Z", "subject": "Do not leak buffer name on failure, from Jeong, Heon.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "995b39579506049aadbe321fbc4fb9818c50e2a3", "date": "2026-08-24T07:06:58Z", "subject": "Do not leak format when drawing preview, from Jeong, Heon.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "4cabc2ae85eea7d80a4d7f7fc75c8191f4c72c86", "date": "2026-08-24T07:08:19Z", "subject": "Do not leak path when destroying client, from Jeong, Heon.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "7792a74c76c03ad0074880d9e6bbdc27cf3afb7f", "date": "2026-08-24T07:10:38Z", "subject": "Free pane fallback range when freeing pane, from Jeong, Heon.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "c9f5ecb4b55dcdb80af914eb2320f71f0f1234ec", "date": "2026-08-24T07:14:54Z", "subject": "Do not leak cached status line, from Jeong, Heon.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "22cdf6acac4fae0ff18c2d1453da80aa121072ee", "date": "2026-08-24T08:18:18+01:00", "subject": "3.7d bits.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "402d366bba94cf13eaddd793a889f56f91cfcb61", "date": "2026-08-24T07:26:43Z", "subject": "Add {} to shell escape characters (special in bash/ksh/csh), GitHub issue 5514.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "a1bca29bdd7c8ae2d7e195e47bca1f471f333e54", "date": "2026-08-24T09:27:42+02:00", "subject": "lib: silence gcc-16 compiler warnings `-Wmaybe-uninitialized`", "body": "Seen in the 'curl-for-win / Windows gcc zlib-classic (x64)' CI job,\nafter it got an upstream upgrade from gcc-15 to gcc-16:\n```\nlib/http.c:206:6: error: 'out.str' may be used uninitialized [-Werror=maybe-uninitialized]\n 206 | if(header_has_value(&header, &out)) {\n | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\nlib/api.h:188:3: error: 'guard.data' may be used uninitialized [-Werror=maybe-uninitialized]\n 188 | Curl_mapi_enter((g), (m), CURL_MAPI_FN_##fn, (r)) && (m)\n | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\nlib/multi.c:1695:6: note: in expansion of macro 'CURL_MAPI_ENTER'\n 1695 | if(CURL_MAPI_ENTER(&guard, m, multi_poll, &mresult)) {\n | ^~~~~~~~~~~~~~~\n```\nRef: https://github.com/curl/curl/actions/runs/32626356576/job/97162205296#step:3:4906\n\nCloses #22651", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "b7125da85c18eb903978c6b96a117ae021ffdcc1", "date": "2026-08-24T09:32:44+02:00", "subject": "feat: user-event API", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "postgres", "sha": "8646214ee20b3754d69f9d087baa7eeb1b7e1d0b", "date": "2026-08-24T09:46:08+02:00", "subject": "Use ArchiveHandle.lookaheadSize instead of a hardcoded 512", "body": "_discoverArchiveFormat allocates the lookahead buffer used to identify\nthe archive format and records its size in AH->lookaheadSize, but then\nused a hardcoded size in the two plades that need it. Use the field\ninstead, so that the buffer and its size cannot drift apart and risk\nout-of-bound reads.\n\nAuthor: Aleksander Alekseev \nReviewed-by: Michael Paquier \nReviewed-by: Daniel Gustafsson \nDiscussion: https://postgr.es/m/CAJ7c6TPqLZOMk3hPrgLQvrOg9BCY4SV6bUC_wx54fyquRkBckw@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "7710394c3ce6794752e1273218236a5477e3fb66", "date": "2026-08-24T00:50:06-07:00", "subject": "fix(runtime-vapor): count each hydration cursor exit once", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "ant-design", "sha": "44bd6dd334c316f86a291a3bcf3fa3a8278c0b1d", "date": "2026-08-24T15:53:01+08:00", "subject": "fix(Form): render numeric zero labels", "body": "", "style": "conventional", "type": "fix", "scope": "Form", "breaking": false} {"repo": "curl", "sha": "5a50313147292a8232b35457a52c7b65bc2517c3", "date": "2026-08-24T03:53:23-04:00", "subject": "appveyor: save an artifact for the vs2026 debug build", "body": "- Save an artifact (zip of curl.exe and dependencies) for the job\n 'CM VS2026, Debug, x64, OpenSSL 3.6 + Schannel, Static, Unicode,\n Build-tests & examples, clang-cl'.\n\nFor user testing.\n\nRef: https://github.com/curl/curl/issues/22466#issuecomment-5389439586\n\nCloses https://github.com/curl/curl/pull/22653", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "5849fd07a68bba529b751a878030980452eb6ca5", "date": "2026-08-24T01:00:28-07:00", "subject": "test(runtime-vapor): make the fallback bench ladder exercise fallback", "body": "", "style": "conventional", "type": "test", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "31391ba3d0a1f5b2245468e4d8090ef388269a56", "date": "2026-08-24T01:09:11-07:00", "subject": "fix(runtime-vapor): recover when the server rendered fewer logical children", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "curl", "sha": "541968e36153f8433ae12d2b9dc8f0c7b9449457", "date": "2026-08-24T10:18:22+02:00", "subject": "circleci: update URL in comment, include in future URL checks", "body": "Also:\n- mdlinkcheck: check two CI files for broken URLs.\n\nCloses #22654", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "0b9697f76a4507db97ffc9a5694eef0970288d3b", "date": "2026-08-24T01:31:13-07:00", "subject": "refactor(runtime-vapor): hold the anchor invariant in the constructor", "body": "", "style": "conventional", "type": "refactor", "scope": "runtime-vapor", "breaking": false} {"repo": "curl", "sha": "50bd20abb215c1fa3cc5d05c470be44fa7b4b42e", "date": "2026-08-24T10:54:36+02:00", "subject": "circleci: bump images to their latest revision", "body": "Ref: https://circleci.com/developer/images?imageType=machine\n\nCloses #22655", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "2687751471c04e07a9c5da78338569d97aff7fd4", "date": "2026-08-24T11:07:17+02:00", "subject": "protocol: simpler Curl_getn_scheme runs faster", "body": "Iterating unit test 1627 50,000 times show the new version to be 31%\nfaster on my machine.\n\n- unit1627: add more test strings, In particular three, five and six letter\n non-existing schemes.\n\n- remove scripts/schemetable.c, not used anymore\n\nCloses #22658", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "c091e25413b2e68630c87d83f899df3699b603a3", "date": "2026-08-24T02:30:10-07:00", "subject": "revert(runtime-vapor): recover when the server rendered fewer logical children", "body": "This reverts commit 491218b07.", "style": "conventional", "type": "revert", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "9e6d1e5634eca58c692a56d4aaa90b465a375c93", "date": "2026-08-24T02:38:59-07:00", "subject": "fix(runtime-vapor): skip the restore on a second cursor exit", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "angular", "sha": "87af7a46b4cdd443dd15bde5dcf986fc5714712f", "date": "2026-08-24T11:43:40+02:00", "subject": "docs: add self-closing tags guideline to components reference", "body": "(cherry picked from commit 719e1ada7413af3d17f2c68e9f6ccdd1abd283df)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "719e1ada7413af3d17f2c68e9f6ccdd1abd283df", "date": "2026-08-24T11:43:40+02:00", "subject": "docs: add self-closing tags guideline to components reference", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "curl", "sha": "c2676bf9e639e31e44b12a2cc82cf2046231b056", "date": "2026-08-24T11:55:29+02:00", "subject": "vssh: silence gcc-11 `-Wnull-dereference`, dedupe `CURL_EASY_STR()` calls", "body": "Silencing:\n```\nIn file included from libcurl_unity.c:179:\nvssh/vssh.c: In function ‘Curl_ssh_setup_pkey.part.0’:\nvssh/vssh.c:423:50: error: potential null pointer dereference [-Werror=null-dereference]\n 423 | CURL_EASY_STR(data, STRING_SSH_PUBLIC_KEY)[0]) {\nlibtool: compile: gcc -DHAVE_CONFIG_H -I../include -I../lib -I../lib -I. -DBUILDING_LIBCURL -DCURL_STATICLIB -DUNITTESTS -D_GNU_SOURCE -Werror-implicit-functi\n```\nRef: https://app.circleci.com/pipelines/gh/curl/curl/20825/workflows/c3eb591b-5a7d-40da-baa8-d2d9980b0cb2/jobs/181277\n\nFollow-up to c8df3defd9664cd276adc33754210ee07b5bbf71 #22628\n\nCloses #22656", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "2a382ffb6173b14919acf3a430b7504eb5793063", "date": "2026-08-24T18:14:49+08:00", "subject": "fix(lsp): completion overwrites text before the word boundary #41463", "body": "Problem:\nOne item with an edit range before the word boundary changes the start\ncolumn for all items. For tsserver `str.`, completing `str.char` can\nbecome `strcharAt` and break filtering.\n\nSolution:\nPrepend the missing text and filter on it.", "style": "conventional", "type": "fix", "scope": "lsp", "breaking": false} {"repo": "postgres", "sha": "53a981ed5bbaa8344da88dfa3550e6cfb6ecf3c9", "date": "2026-08-24T22:43:27+12:00", "subject": "Fix incorrect nmembers calculation in test_bitmapset", "body": "This calculation was failing to correctly account for the min_value when\ncalculating how many random members to add to the Bitmapset. What the\ncode was meant to do was add a random number of members between the\nminimum and maximum values. The code failed to account for the minimum\nvalue being non-zero, which could result in too many members being added\nto the set, which could reduce the effectiveness of the test. In reality,\nthe only caller to this function did pass 0 as the min_value, so there is\nno live bug being fixed here.\n\nFix this so that the Bitmapset is more sparsely populated with members\nbetween the minimum and maximum values. In passing, swap the order of the\nmin_value and max_value parameters. Having the minimum value appear\nbefore the maximum value makes more sense. Also align the names of both\nparameters, and adjust the 'offset' calculation to widen the offset range\nby 1 so that the possible range for that random value is -max_value to\nmax_value.\n\nAuthor: David Rowley \nReviewed-by: Greg Burd \nDiscussion: https://postgr.es/m/CAApHDvpfZqdMRmR-6W9UXb+Tanex5wK6PsL=XzYCV4P3pD36Ng@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "d1702a24b3629e25d972685d4c45d3c51b6beccb", "date": "2026-08-24T12:46:30+02:00", "subject": "ngtcp2+openssL: fix early data", "body": "Openssl needs an additional call into its API at just the right time to\n*really* enable early data use. Add checks after handshake if early data\nwas really successful, more stringent than we used before.\n\ntest_02_32: enable early data test wiht more backends and platforms to\nsee how CI now copes with it.\n\nFixes #22649\nCloses #22657", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "ad74a8dada2aaa17bfd0b8cc7b4119ff7a8ac04b", "date": "2026-08-24T13:47:53+02:00", "subject": "docs: add deprecation steps for EOL package versions", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "postgres", "sha": "250942b0d8212f8bdb10b9d63a911aeea5dd4de0", "date": "2026-08-24T14:20:47+02:00", "subject": "Fix untranslatable message that was pasted together at run time", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "f66eeef8a5f62e6465963f1ed1e9c223abde6e24", "date": "2026-08-24T13:36:34+01:00", "subject": "Added a fallback in server_client_check_redraw() as recommended by codex to catch an unserviced edge-case.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "143bbfbb1239cba40b9adc52559c358aaa69a277", "date": "2026-08-24T14:41:57+02:00", "subject": "fix(lsp): clamp inline completion range to the line on accept #41435", "body": "Problem:\n`Completor:accept()` hands the item's range straight to\n`nvim_buf_set_text()`. That range is resolved when the response arrives,\nso text deleted before accepting (backspacing within the 200 ms\ndebounce) leaves the range end past what is left of the line, and the\naccept aborts with `Invalid 'end_col': out of range`.\n\n18d6436 fixed the same staleness in `Completor:show()`, which drops an\nitem once its range *start* falls outside the buffer. That guard does\nnot cover the range end, and `accept()` was never given one, so this is\na hole left by that fix rather than a regression of it.\n\nSolution:\nClamp the end of the range to the end of the line before writing. The\nstart needs no clamp: `show()` dropped the item one event loop tick\nearlier if it had gone out of range.", "style": "conventional", "type": "fix", "scope": "lsp", "breaking": false} {"repo": "curl", "sha": "75e85f0e860d9a18ea895bdf10c744bb4d89c8f5", "date": "2026-08-24T09:06:15-04:00", "subject": "mprintf: stage output and emit it in runs", "body": "alloc_addbyter() appended one byte at a time through curlx_dyn_addn(),\npaying its bounds checks, growth logic, memcpy and NUL-termination once\nper output byte, and formatf() had no way to hand an output function\nmore than one byte at a time.\n\nStage dynbuf output in a 512-byte buffer and append a chunk at a time,\nand add an optional block-output callback alongside the per-byte stream\ncallback, so literal segments, %s bodies, digit runs and padding are\nemitted as runs. Each output function gets a memcpy-based variant:\naddrun(), alloc_addrun(), storerun() and fwrite_wrapper().\n\nThe dynbuf size limit is still reported at exactly the byte that exceeds\nit: alloc_addbyter() and alloc_addrun() check whether what they are\nabout to stage can still fit and, when it cannot, flush and append that\nbyte alone.\n\nA staged byte is not committed when it is accepted, so formatf() commits\npending output before %n and performs it only if that succeeds. Without\nthat, an allocation failure could leave %n executed where the\nbyte-at-a-time path would have stopped short of it.\n\nstream_zrun(), which stops at a NUL, scans with a byte loop rather than\nmemchr(), because the length it is given is the requested precision and\nis not clamped to the string length.\n\ncurl_maprintf() of a request line: -73% on x86-64, -67% on a Cortex-A55,\n-68% on a 250MHz e300c3. Of one header line: -81%, -77%, -71%. Of a\n300-byte all-%s body: -83%, -78%, -77%.\n\nCloses #22766", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "1e9f80ec18c06be06211c93536d90a81955e661a", "date": "2026-08-24T06:38:47-07:00", "subject": "perf(vm): don't prewarm modules the worker never requests", "body": "", "style": "conventional", "type": "perf", "scope": "vm", "breaking": false} {"repo": "angular", "sha": "745cc30139146d3bbbed0cd81d651e39380aec4a", "date": "2026-08-24T16:49:11+03:00", "subject": "refactor(docs-infra): remove example code that nothing references", "body": "The `built-in-directives` and `structural-directives` example apps are no\nlonger referenced. Two guide rewrites moved their pages to inline code fences\nand dropped the last references without removing the apps: #69134 for\nstructural directives, and #69822 for built-in directives, whose page is gone\nentirely. Both apps sat inside the `embeddable` glob, so they were still\ncompiled as preview components on every build.\n\nThe example e2e suites are excluded from every consumer by design: previews\nignore `*.e2e-spec.ts` and `*.po.ts`, and the zip and StackBlitz pipelines\nboth list the suffix in `EXCLUDE_FILES`. They could not run in any case,\nsince every spec declares a `driver` and never assigns it before calling\n`driver.get('')`, left over from an abandoned protractor to selenium\nmigration.\n\nThe eslint configuration in the same directory is also unused. Nothing in the\nrepository depends on eslint, and `tsconfig.eslint.json` extends a path that\nno longer exists.\n\n(cherry picked from commit 90fc9475f24d5839588978923ead5b6adbefbe54)", "style": "conventional", "type": "refactor", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "90fc9475f24d5839588978923ead5b6adbefbe54", "date": "2026-08-24T16:49:11+03:00", "subject": "refactor(docs-infra): remove example code that nothing references", "body": "The `built-in-directives` and `structural-directives` example apps are no\nlonger referenced. Two guide rewrites moved their pages to inline code fences\nand dropped the last references without removing the apps: #69134 for\nstructural directives, and #69822 for built-in directives, whose page is gone\nentirely. Both apps sat inside the `embeddable` glob, so they were still\ncompiled as preview components on every build.\n\nThe example e2e suites are excluded from every consumer by design: previews\nignore `*.e2e-spec.ts` and `*.po.ts`, and the zip and StackBlitz pipelines\nboth list the suffix in `EXCLUDE_FILES`. They could not run in any case,\nsince every spec declares a `driver` and never assigns it before calling\n`driver.get('')`, left over from an abandoned protractor to selenium\nmigration.\n\nThe eslint configuration in the same directory is also unused. Nothing in the\nrepository depends on eslint, and `tsconfig.eslint.json` extends a path that\nno longer exists.", "style": "conventional", "type": "refactor", "scope": "docs-infra", "breaking": false} {"repo": "tmux", "sha": "7d358208b9e2dce3b041b17d611bb285f1363673", "date": "2026-08-24T14:07:56Z", "subject": "Free any remaining items when destroying screen, from Jeong, Heon.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "8c440c469bb6f075fe26a1199ae6dcd2314394c9", "date": "2026-08-24T10:18:47-04:00", "subject": "fix(ui2): ui2 misinterprets getchar() input as pager_char (CR) #41465", "body": "Problem:\n`vim.on_key()` callbacks are invoked for keys that `getchar()` consumes,\nwhich never reach the main loop, and there is no way to tell such a key\napart. With ui2 the `on_key` handler activated after interactive\ncmdline, handles it as `pager_char` (thus enters the ui2 pager).\n\nSolution:\nSkip `vim.on_key()` callbacks during `getchar()`/`getcharstr()`.", "style": "conventional", "type": "fix", "scope": "ui2", "breaking": false} {"repo": "tmux", "sha": "41e701ad8408a3411bb59198a067c4ef471abba8", "date": "2026-08-24T15:27:25+01:00", "subject": "Return early if image is zero height when generating fallback, otherwise we can allocated too little space. Reported by Vivek Parikh.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "e3bf9680babe4ad22f0a73cc940065859bd9ac0c", "date": "2026-08-24T19:38:41+05:00", "subject": "docs: clarify onSameUrlNavigation reload behavior", "body": "fixes #70367\n\n(cherry picked from commit 3848b7f8bcc7c78d5650b31b9615ba3fcb9cfe38)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "3848b7f8bcc7c78d5650b31b9615ba3fcb9cfe38", "date": "2026-08-24T19:38:41+05:00", "subject": "docs: clarify onSameUrlNavigation reload behavior", "body": "fixes #70367", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "neovim", "sha": "d1b811b6c2d93fe13442d61331a9276fea417f7b", "date": "2026-08-24T16:51:39+02:00", "subject": "fix(lsp): advance inline completion range over matching input #41434", "body": "Problem:\n`Completor:accept()` applies the item against the range the server\nanswered with. Characters typed after the request but before accepting\nlie inside neither that range nor the replacement, so they are left\nbehind by the insertion: typing `a` after `foob` and accepting `foobar`\nwithin the 200 ms debounce yields `foobara`.\n\nSolution:\nGrow the item's range over the characters typed since the request, so\nthat `accept()` replaces them. Only while the candidate still matches,\ndecided by the longest common prefix `show()` already computes, so input\nthat contradicts it is left alone. Snippet items are unaffected, since\n`accept()` ignores the range for them.", "style": "conventional", "type": "fix", "scope": "lsp", "breaking": false} {"repo": "tmux", "sha": "5b54c7d41af85cd7f15f0c272d8980db945e2033", "date": "2026-08-24T15:05:26Z", "subject": "Always use synchronized updates in copy mode (like when not in the active pane), since copy mode will almost always moved the cursor. From Artem Goldenberg in GitHub issue 5525.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "0fd1609afe255b275c2560db6db584081a16e8a0", "date": "2026-08-24T18:50:02+03:00", "subject": "postgres_fdw: Fix flaky push down FUNCTION RTE test", "body": "The unnest() test introduced in 0ee83dd4a99 did not produce a\ndeterministic EXPLAIN plan output.\n\nThe two sides of this test's nested loop join cost about the same, so\nwhich one ended up as the outer wasn't stable and could flip between\nplatforms.\n\nUse an equality predicate on ft1.c3 instead of a range one. That\nrestricts ft1 to a single row, so it's clearly the outer of the nested\nloop and the Materialize on the inner side goes away, giving a stable\nplan.\n\nAuthor: Sami Imseih \nReviewed-by: Bharath Rupireddy \nDiscussion: https://postgr.es/m/CAA5RZ0uC%3DGOihOKcF7o3JVETF3e-Fkc4OUO5NbqMEKCTeSN-dQ%40mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "cd63e9e4332899976e4e852af24cb81f6bf7169b", "date": "2026-08-25T01:09:12+09:00", "subject": "feat(ui): toggle trace target highlight", "body": "", "style": "conventional", "type": "feat", "scope": "ui", "breaking": false} {"repo": "vitest", "sha": "c21865dfd0787e309aa5e974ac9ccd959a5bf0ea", "date": "2026-08-25T01:09:43+09:00", "subject": "fix: fix `recordArtifact` location with `vi.defineHelper`", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "openssl", "sha": "334af72270c87141ceb8553364d2b909cd353106", "date": "2026-08-24T19:47:42+02:00", "subject": "OSSL_FN: make bn_release() to be constant time.", "body": "When we perform some operations with BIGNUM using OSSL_FN functions,\nwe can (and normally should) implement them in a constant-time way.\nBut at the end we always call bn_release(), so it should be constant\ntime as well.\n\nAssisted-by: Claude:claude-opus-4-8\nReviewed-by: Matt Caswell \nReviewed-by: Richard Levitte \nMerge-date: Mon Sep 21 10:43:54 2026\nMerged-from: https://github.com/openssl/openssl/pull/32874", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "1e641c919fc1421493bf913feb607896982451a3", "date": "2026-08-24T21:02:36+03:00", "subject": "fix: no-loss-of-precision false positive with trailing decimal point", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "postgres", "sha": "ed37fefedb3f81cddca576bb256f162d5c036ab1", "date": "2026-08-24T11:51:52-07:00", "subject": "Add C test function for pg_locale.h APIs.", "body": "Test the API independently to account for fallback paths that aren't\nadequately tested from SQL.\n\nThe backport to 18 also tests the previously-supported behavior where\na size of -1 meant that the string was NUL-terminated. That behavior\nwas later removed in 19.\n\nReviewed-by: Andres Freund \nDiscussion: https://postgr.es/m/v36ssaygf7grb3qzfsjhtdzi7kqd45ds56nyuf7gi5qjml4qbb@ezmfqzmhlrs2\nBackpatch-through: 18", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "364fc8e72e97fcd24a1f272879af5a64ec851277", "date": "2026-08-24T11:51:59-07:00", "subject": "pg_locale.c: add explanatory comments.", "body": "Explain the purpose of case mapping functions, rather than just the\nAPI.\n\nSuggested-by: Andres Freund \nDiscussion: https://postgr.es/m/v36ssaygf7grb3qzfsjhtdzi7kqd45ds56nyuf7gi5qjml4qbb@ezmfqzmhlrs2\nBackpatch-through: 18", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "510f61555b9da5800ccd13de2336c90a64c0b9a5", "date": "2026-08-24T15:01:50-04:00", "subject": "fix(ui2): pager_char does not match keycode aliases #41468", "body": "Problem:\n`pager_char` compares the `keytrans()` result exactly. E.g. `pager_char\n= \"\"` does not work, it expects `` (uppercase).\n\nSolution:\nNormalize `pager_char` in enable().", "style": "conventional", "type": "fix", "scope": "ui2", "breaking": false} {"repo": "neovim", "sha": "1ef030f16226152fbba0b46c76bd0726aecd30b4", "date": "2026-08-24T22:07:53+02:00", "subject": "fix(help): :helptags regressions", "body": "Problem:\nParent commit regressed some behavior of the old C helptags-gen impl:\n- helpfiles in sub-directories are named by basename, so :help fails\n- \"help-tags\" always names \"tags\", never \"tags-nl\"\n- E150 E151 E152 E153 are never reported\n- existing tags file is not overwritten if no tags were found\n- a duplicate tag aborts the run, skipping the remaining directories\n- `*.TXT` and `*.FRX` (uppercase) are not recognized as helpfiles\n- tree-sitter-vimdoc accepts tags that the C parser rejected: `*a|b*`\n (breaks |links|) and unterminated `*tag`\n- PUC Lua \"<\" compares with (localized) `strcoll()`, so the tags file is\n not sorted by byte value (E432)\n- requiring `vim.treesitter` at load time breaks :help itself, not just\n :helptags, where the module is unavailable\n- `vim.pack` runs :helptags for plugins that have no \"doc/\" directory,\n so every install/update emits E150\n\nSolution:\n- Fail the build if generating helptags reports any `v:errmsg`.\n- Restore old behavior: tags generated for runtime/doc are now identical\n to those from the C implementation. Errors are non-fatal messages\n instead of exceptions, so all directories are still processed.\n- Load treesitter lazily, report a plain error if parser is missing.", "style": "conventional", "type": "fix", "scope": "help", "breaking": false} {"repo": "postgres", "sha": "8fa803dd695073a6fad49c49878b46783ac2f207", "date": "2026-08-24T13:33:31-07:00", "subject": "ltree/crc32.c: fix fragile code.", "body": "Explicitly make space for the NUL when casefolding.\n\nNo known bug in the previous code, because the previous buffer (size\n12) was more than large enough for folding any codepoint with enough\nroom left for a NUL. The builtin provider's limit is 7; ICU's limit\nseems to be 7 also; and libc always does 1:1 mappings so the real\nlimit is MAX_MULTIBYTE_CHAR_LEN + 1 (size 5).\n\nReviewed-by: Heikki Linnakangas \nReviewed-by: Rithvika Devisetti \nDiscussion: https://postgr.es/m/2cfbc37ae8deccd3825e36b7f31c391cbf8ff9b8.camel@j-davis.com\nBackpatch-through: 18", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "dc78d8f9ea78c30d31d622da37246d464d725bcc", "date": "2026-08-24T20:34:26Z", "subject": "Do not add messages to cfg_causes after config has finished, GitHub issue 5528.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "fb5d466a35213ef1a4719f94f57194a5fb05ec30", "date": "2026-08-24T22:43:47+02:00", "subject": "fix(help): :helptags on CRLF helpfiles", "body": "Problem:\nThe vimdoc parser does not treat \"\\r\" as whitespace, so in a CRLF\nhelpfile the codeblock rule fails and \"*\" pairs inside examples become\ntags. On Windows this generates duplicate \".\" and \"/\" tags:\n\n D:/a/neovim/neovim/build/bin/nvim.exe -u NONE -i NONE -e --headless -c \"helptags ++t doc\" -c \"exe 'cquit' !empty(v:errmsg)\"\"\n Error in command line:\n E154: Duplicate tag \".\" in gui.txt and repeat.txt\n E154: Duplicate tag \"/\" in pattern.txt and usr_08.txt\n E154: Duplicate tag \"/\" in usr_08.txt and pattern.txt\n\nNote: The old C parser was unaffected because it read helpfiles in text\nmode (`os_fopen(…, \"r\")`).\n\nSolution:\nStrip \"\\r\" before parsing.", "style": "conventional", "type": "fix", "scope": "help", "breaking": false} {"repo": "neovim", "sha": "8afc1bca27335ebd0703caa5759701d5ebc6e6b7", "date": "2026-08-24T22:45:49+02:00", "subject": "build: failing \"with-external-deps\" CI job", "body": "Problem:\nSince 20edb8fa9aac, generating helptags requires the\n`tree-sitter-vimdoc` parser, but the external_deps CI job builds with\nUSE_BUNDLED_TS_PARSERS disabled.\n\nSolution:\nSet USE_BUNDLED_TS_PARSERS.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "angular", "sha": "72559b471098f06a56fb1f71b6acc9a0b9b2ac17", "date": "2026-08-25T00:05:37+03:00", "subject": "fix(docs-infra): redirect three removed pages instead of 404ing", "body": "`guide/http/security` and `reference/concepts` both still exist as content but\nneither is routed, so each falls through to the 404 shell.\n\n`guide/http/security` was navigable from November 2023 until #54365 removed its\nentry in February 2024. #55029 then copied its XSRF sections into the security\nguide, which already covered XSSI, and #55060 repointed the remaining links, so\nit now redirects to `/best-practices/security`.\n\n`reference/concepts` was added by #54365 and removed by #58694 in November\n2024, leaving it navigable for nine months without a redirect. The same commit\nalso removed `guide/ngmodules`, the route its only card linked to, so both now\nredirect to `/guide/ngmodules/overview`, alongside the four `guide/ngmodules/*`\npaths already redirected there.\n\nBoth pages are removed, along with the Bazel package that existed only to build\nthe concepts page, and the stale `llms-list.md` entry for the HttpClient page.\nThe security guide takes its place in that list, so `llms-full.txt` keeps its\nXSSI coverage and picks up CSP, Trusted Types and sanitization with it.\n\n(cherry picked from commit 7da60d1920fee8a088615bbf32006c9bed17de72)", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "7da60d1920fee8a088615bbf32006c9bed17de72", "date": "2026-08-25T00:05:37+03:00", "subject": "fix(docs-infra): redirect three removed pages instead of 404ing", "body": "`guide/http/security` and `reference/concepts` both still exist as content but\nneither is routed, so each falls through to the 404 shell.\n\n`guide/http/security` was navigable from November 2023 until #54365 removed its\nentry in February 2024. #55029 then copied its XSRF sections into the security\nguide, which already covered XSSI, and #55060 repointed the remaining links, so\nit now redirects to `/best-practices/security`.\n\n`reference/concepts` was added by #54365 and removed by #58694 in November\n2024, leaving it navigable for nine months without a redirect. The same commit\nalso removed `guide/ngmodules`, the route its only card linked to, so both now\nredirect to `/guide/ngmodules/overview`, alongside the four `guide/ngmodules/*`\npaths already redirected there.\n\nBoth pages are removed, along with the Bazel package that existed only to build\nthe concepts page, and the stale `llms-list.md` entry for the HttpClient page.\nThe security guide takes its place in that list, so `llms-full.txt` keeps its\nXSSI coverage and picks up CSP, Trusted Types and sanitization with it.", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "tmux", "sha": "2d5328a8604b026d9b143c6c46d85283c0676ed5", "date": "2026-08-24T21:17:19Z", "subject": "If a pane has no scrollbar, don't bother to redraw it, and if redraw is deferred change scrollbar redraws into redraw-all-scrollbars not redraw entire window. GitHub issue 5529.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "50ea32162f8ebf2f6843c2f89e15abf9e85d828f", "date": "2026-08-24T21:19:40Z", "subject": "Correct a comparison ('\\0' should be 0).", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "fad70a09ff43724b521f7e450d894a93d54330a0", "date": "2026-08-24T16:28:19-05:00", "subject": "Fix autovacuum's handling of TOAST storage parameters.", "body": "The previous commit made VACUUM apply a main table's storage\nparameters to its TOAST table, as CREATE TABLE has long documented.\nAutovacuum still gets this wrong in two ways. It falls back to the\nmain table's autovacuum parameters only when the TOAST table has no\nreloptions at all, so setting a single toast.* parameter silently\nstops the rest from being inherited. And it never consults the\nmain table for the non-autovacuum parameters, such as\nvacuum_index_cleanup and vacuum_truncate.\n\nTo fix, combine the two sets with merge_toast_reloptions() rather\nthan choosing between them, and hand the main table's parameters\ndown to vacuum_rel() the way VACUUM now does.\npg_stat_autovacuum_scores uses the combined parameters for TOAST\ntables as well; it has to collect the main relations' parameters\nbefore it can do so, so it now makes a preliminary pass over\npg_class.\n\nAn existing shortcoming that this patch only makes worse is that\nautovacuum remains oblivious to concurrent storage parameter\nchanges on the main table. That is, the main table's parameters\nmay be captured long before its TOAST table is processed, and a\nuser may very well have altered the settings in the meantime.\nFixing that would likely require additional pg_class lookups, and\nit's not clear if it's worth the trouble.\n\nWhile this is a bug fix, it's too intrusive for back-patching, but\nthe issue seems to have gone unnoticed for a very long time,\nanyway.\n\nReviewed-by: Michael Paquier \nReviewed-by: Sami Imseih \nReviewed-by: Greg Burd \nTested-by: solai v \nDiscussion: https://postgr.es/m/aFRxC1W_kZU9OjJ9%40nathan", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "c244c62d8b9e7762be3d35090463c76042ab7f89", "date": "2026-08-24T16:28:19-05:00", "subject": "Fix VACUUM's handling of TOAST storage parameters.", "body": "Per the documentation for CREATE TABLE:\n\n If a table parameter value is set and the equivalent toast.\n parameter is not, the TOAST table will use the table's\n parameter value.\n\nPresently, VACUUM does no such thing. It reads the TOAST table's\nown reloptions, which hold only what was set through toast.*, so\nvacuum_index_cleanup, vacuum_max_eager_freeze_failure_rate, and\nvacuum_truncate settings on the main table have no effect on its\nTOAST table.\n\nTo fix, add merge_toast_reloptions(), which walks the parse table\nfor StdRdOptions and takes the main table's value for anything the\nTOAST table left at its default. vacuum_rel() hands the main\ntable's parameters down when recursing to a TOAST table, and it\nmerges them into a copy of the TOAST table's parameters before the\nvalues are used. This doesn't help VACUUM against a TOAST table\ndirectly (e.g., \"VACUUM pg_toast.pg_toast_5432\"), but that's\nprobably okay because it's not the main supported way to vacuum a\nTOAST table (see VACUUM's PROCESS_MAIN and PROCESS_TOAST options).\n\nA follow-up commit will do the same for autovacuum. While this is\na bug fix, it's too intrusive for back-patching, but the issue\nseems to have gone unnoticed for a very long time, anyway.\n\nReviewed-by: Michael Paquier \nReviewed-by: Sami Imseih \nReviewed-by: Greg Burd \nTested-by: solai v \nDiscussion: https://postgr.es/m/aFRxC1W_kZU9OjJ9%40nathan", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "732af45f431913752092151256a175de2197cb99", "date": "2026-08-24T23:37:49+02:00", "subject": "CVE-2026-80230: OpenSSL pinning bypass", "body": "openssl: no server cert is only okay if also not pinned", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "5267ed859d545534d0c21675a2b70af5a3b6e3ef", "date": "2026-08-24T23:37:49+02:00", "subject": "openssl: no server cert is only okay if also not pinned", "body": "Reported-by: Stanislav Fort\nCloses #22660", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "1e9b06a8e07c3bd394459e68902b79d7b55c2035", "date": "2026-08-24T14:39:51-07:00", "subject": "pg_locale.h: define case mapping limits.", "body": "Clearly define case mapping upper bounds that cover all providers,\nlocales, and encodings: PG_MAX_CASEMAP_MBCHARS,\nPG_MAX_CASEMAP_EXPANSION, and PG_CASEMAP_BUFSZ.\n\nAlso redefine UTF8_CASEMAP_BUFSZ to include room for a NUL for\nconsistency with PG_CASEMAP_BUFSZ.\n\nUse PG_CASEMAP_BUFSZ for the static buffer in ltree/crc32.c.\n\nRemove macros UNICODE_CASEMAP_LEN (which was redundant with\nUNICODE_MAX_CASEMAP_CODEPOINTS from unicode_limits.h), and\nUNICODE_CASEMAP_BUFSZ which was not as descriptive or helpful.\n\nSuggested-by: Heikki Linnakangas \nReviewed-by: Rithvika Devisetti \nDiscussion: https://postgr.es/m/2cfbc37ae8deccd3825e36b7f31c391cbf8ff9b8.camel@j-davis.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "d6e4e22e9143749a9b629f2149e78edf2b6e0ec0", "date": "2026-08-25T10:43:08+12:00", "subject": "Align random test function parameters in test_bitmapset", "body": "53a981ed5 adjusted test_random_offset_operations() so that the minimum\nvalue parameter came before the maximum value parameter. Here we adjust\ntest_random_operations() so that it accepts absolute numbers for the\nminimum and maximum values. Previously the maximum value to put in the\nrandom Bitmapset was the minimum value plus the range. Also align the\nparameter order with test_random_offset_operations() to make these\nfunctions consistent with each other.\n\nBackpatch to v19 as test_random_operations() is new there.\n\nAuthor: David Rowley \nReviewed-by: Greg Burd \nDiscussion: https://postgr.es/m/66d041b4-32f4-4097-bb2c-0e7147337a5e@app.fastmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "a447d806fb304a8af2265ee85d43280bc11b019a", "date": "2026-08-24T17:18:24-07:00", "subject": "fix(runtime-vapor): render a null dynamic component as an empty branch", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "23a2b2ca7c36c15cd774daa313fbad32e44e3dfb", "date": "2026-08-25T05:28:46+05:00", "subject": "fix(runtime-vapor): set value as attribute so form reset restores it", "body": "close #15338", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "2473b781032a82c02172470c600cbe38d0204fc6", "date": "2026-08-24T17:37:58-07:00", "subject": "fix(runtime-vapor): claim runtime structural anchors so block validity matches across builds", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "64671b5c54bae054f454bb35e20e680046d473af", "date": "2026-08-25T05:41:18+05:00", "subject": "fix(runtime-vapor): do not skip the initial DOM prop set", "body": "close #15339", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "openssl", "sha": "55aeaef7d55fa792fde39146311450064b608196", "date": "2026-08-25T11:19:35+10:00", "subject": "rsa: use generated tries in crypto helpers", "body": "Assisted-by: ChatGPT:gpt-5.6Sol\n\nReviewed-by: Shane Lontis \nReviewed-by: Mounir Idrassi \n(Merged from https://github.com/openssl/openssl/pull/32502)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "f82466c814030cb72b873662360b1dccfd21c6a7", "date": "2026-08-25T11:19:58+10:00", "subject": "providers: convert RSA keymgmt parameter parsing", "body": "Assisted-by: ChatGPT:gpt-5.6Sol\n\nReviewed-by: Shane Lontis \nReviewed-by: Mounir Idrassi \n(Merged from https://github.com/openssl/openssl/pull/32502)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "ac9b57ccfc42f3d306aac925c42a00b3900ea82d", "date": "2026-08-24T18:20:19-07:00", "subject": "refactor(runtime-vapor): drop the write-only OWNS_ANCHOR flag", "body": "", "style": "conventional", "type": "refactor", "scope": "runtime-vapor", "breaking": false} {"repo": "postgres", "sha": "7e29fcd1ae0083a7fd45f14ab7b9ad3ea0c97e9c", "date": "2026-08-25T10:22:12+09:00", "subject": "Don't assume DISTINCT ON implies uniqueness when the tlist has SRFs", "body": "query_is_distinct_for() treated a subquery's DISTINCT ON clause as\nproof that its output is unique over the DISTINCT ON columns, even if\nthe targetlist contains set-returning functions. That's not true:\nwhen the query has an ORDER BY, the planner postpones evaluation of\nSRFs that are not DISTINCT ON or ORDER BY columns until after the\nUnique step, so the subquery can produce duplicates of the DISTINCT ON\ncolumns. Relying on this bogus uniqueness proof allowed join removal\nand unique-inner joins to produce wrong results.\n\nPlain DISTINCT is not affected, since all tlist columns are DISTINCT\ncolumns there, and so any SRFs get expanded before the Unique step.\n\nTo fix, make query_supports_distinctness() and query_is_distinct_for()\nrefuse to prove distinctness via DISTINCT ON if the targetlist\ncontains any SRFs. This is more conservative than necessary, since\nthe SRFs are only postponed when there is an ORDER BY and none of them\nappear in a sort/group column, but it doesn't seem worth the trouble\nto check that precisely.\n\nAuthor: Richard Guo \nReviewed-by: Tom Lane \nDiscussion: https://postgr.es/m/CAMbWs4-hfd1Pyy_zBejsVUSy-3dx16rz2hgUakkKnAg3qg2q=Q@mail.gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "84833e2453e12bdd68b31e09512fc036990e4b60", "date": "2026-08-25T09:39:03+08:00", "subject": "fix(runtime-vapor): align DOM prop updates with VDOM", "body": "Compare DOM prop updates against the previous raw binding value instead of\nthe browser-normalized live property.\n\nThis ensures changed bindings are written even when the DOM property already\nnormalizes to the same value, while preserving initial writes and same-binding\nskips. Add regressions for reflected string and numeric properties.", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "postgres", "sha": "db2d99323f1882af87989e698d3678affc4a5557", "date": "2026-08-25T10:49:44+09:00", "subject": "Fix CPU cost of right-semi and right-anti hash joins", "body": "final_cost_hashjoin() assumed that the rows a hash join produces come\nfrom the outer side. That does not hold for JOIN_RIGHT_SEMI and\nJOIN_RIGHT_ANTI, which produce rows from the inner side: matched inner\nrows for the one, unmatched inner rows for the other. hashjointuples,\nwhich carries both the cpu_tuple_cost charge and the remaining qual\ncosts, was therefore counting the wrong rows, and could be far off in\neither direction.\n\nCompute hashjointuples from the inner side for these two join types,\nmirroring what JOIN_SEMI and JOIN_ANTI already do with the outer side.\nThe fraction we need is semifactors.outer_match_frac: it is derived\nfrom the SpecialJoinInfo, so despite its name it always describes the\nsemijoin's left-hand side, which is the inner side here. Compute the\nsemijoin factors for these join types in all cases; previously that\nhappened only when the inner side was provably unique.\n\nThe same mix-up affects outer_matched_rows when the inner side is\nknown unique, where the outer row count was multiplied by the inner\nside's match fraction. A unique inner side means each outer row has\nat most one match, so use the number of matching pairs instead.\n\nA right anti join evaluates the non-hashed joinquals once per tuple\npassing the hash clauses, without any short-circuit. Charge the\nremaining qual costs on that pair count, and only cpu_tuple_cost on\nthe emitted rows. A right semi join short-circuits already-matched\ninner tuples and keeps the emitted-row charge, as JOIN_SEMI does.\n\nNestloop and mergejoin need no equivalent fix: neither supports\nJOIN_RIGHT_SEMI, nestloop doesn't support JOIN_RIGHT_ANTI either, and\nfinal_cost_mergejoin() takes its count from approx_tuple_count(),\nwhich multiplies the two input sizes together and so gives the same\nanswer whichever side is on the outside.\n\nNo backpatch as this could result in plan changes.\n\nAuthor: Richard Guo \nReviewed-by: Ayush Tiwari \nReviewed-by: Haibo Yan \nReviewed-by: Mounir Idrassi \n(Merged from https://github.com/openssl/openssl/pull/32502)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "37843d65ea0f3b4eed7c37d1eb591791c0cc01d8", "date": "2026-08-25T12:01:49+10:00", "subject": "ec: use generated tries in crypto helpers", "body": "Assisted-by: ChatGPT:gpt-5.6Sol\n\nReviewed-by: Shane Lontis \nReviewed-by: Mounir Idrassi \n(Merged from https://github.com/openssl/openssl/pull/32502)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "a5b1f17776fc4109dc5e6289da6ceb06b3e3170d", "date": "2026-08-25T12:02:16+10:00", "subject": "providers: convert EC keymgmt parameter parsing", "body": "Assisted-by: ChatGPT:gpt-5.6Sol\n\nReviewed-by: Shane Lontis \nReviewed-by: Mounir Idrassi \n(Merged from https://github.com/openssl/openssl/pull/32502)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "7c8d34a13fa0b47bf22eefeb9c79db9374f10048", "date": "2026-08-24T19:04:58-07:00", "subject": "fix(runtime-vapor): key the null dynamic component branch by its resolved sentinel", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "neovim", "sha": "c6ba487b89479301f71b06007b82e37cf8329095", "date": "2026-08-24T23:16:02-04:00", "subject": "build(vim-patch): n/a ba43008a2d6788eeb8ff78ae39dedd28598ae658", "body": "", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "openssl", "sha": "19e8a1fc88f64e6b08de04f0e1a7e5fbd7e5d509", "date": "2026-08-25T15:09:04+10:00", "subject": "ec: name generated parameter groups", "body": "Assisted-by: ChatGPT:gpt-5.6Sol\n\nReviewed-by: Shane Lontis \nReviewed-by: Mounir Idrassi \n(Merged from https://github.com/openssl/openssl/pull/32502)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "c2ce945ac07ce51ee32779c7cf1650bb12179afe", "date": "2026-08-25T07:47:24+02:00", "subject": "vtls_scache: use case sensitive path match", "body": "Provided-by: 1rhino2 on hackerone", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "35c3c6e7d5f58e00ac05f15c1b297c962b9f6dfd", "date": "2026-08-25T13:56:38+08:00", "subject": "fix(Menu): keep horizontal submenu open while crossing hover path", "body": "The horizontal item's bottom `::after` underline intercepts the mouse\npath between the item and its popup, so crossing slowly along the bottom\nedge can trigger the 0.1s leave-delay and close the submenu unexpectedly.\n\nMake the decorative underline transparent to pointer events so the hover\ntarget stays the item itself. Dark theme was unaffected only because its\n`activeBarHeight` is 0 (no visible bar).\n\nRef #40337", "style": "conventional", "type": "fix", "scope": "Menu", "breaking": false} {"repo": "tmux", "sha": "1459c90a7fa6a70afd1e8438fa9985141e4002be", "date": "2026-08-25T06:04:33Z", "subject": "Tidy up args_make_commands reference counting to fix a memory leak, from Jeong, Heon.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "0851650c6be43fe23477fee380fc6672d0348e57", "date": "2026-08-25T06:14:16Z", "subject": "Do not leak exit message on client free, from Jeong, Heon.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "8991dafa37d2bd0960bdd0579819150ac4c57a80", "date": "2026-08-25T16:38:12+10:00", "subject": "rsa: use selection-specific import decoders", "body": "Assisted-by: ChatGPT:gpt-5.6Sol\n\nReviewed-by: Shane Lontis \nReviewed-by: Mounir Idrassi \n(Merged from https://github.com/openssl/openssl/pull/32502)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "e63961d7c16f7df5ba4b881f4bfa7a7920c628bd", "date": "2026-08-25T08:39:16+02:00", "subject": "doc: Document overflow handling of pg_class.relpages etc.", "body": "The pg_class fields relpages, relallvisible, and relallfrozen are\nstored as int32 in the catalog but block numbers are internally of\ntype uint32. This all works correctly because the values are\nimplicitly converted back and forth according to C conversion rules.\nBut this was apparently never explicitly documented, so add this.\n\nReviewed-by: Chao Li \nDiscussion: https://www.postgresql.org/message-id/flat/db993f34-e3e1-4e6e-85c3-4f248d2ff29b%40eisentraut.org", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "b345a554edf321e0ca53db0f4a69aac217805677", "date": "2026-08-25T08:39:16+02:00", "subject": "Fix signed/unsigned integer handling in pg_restore_relation_stats()", "body": "The pg_class fields relpages, relallvisible, and relallfrozen are of\ntype int32, but the statistics restoration code and some statistics\ncode in postgres_fdw internally dealt with them as uint32 for a little\nbit, after which they would turn back into int32. This doesn't make\nsense, so fix it to use int32 consistently throughout.\n\nIn passing, change a float to float4 to similarly make the C type\nconsistent with the SQL type.\n\nReviewed-by: Chao Li \nDiscussion: https://www.postgresql.org/message-id/flat/db993f34-e3e1-4e6e-85c3-4f248d2ff29b%40eisentraut.org", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "1dcbc259af03ecbf88806c91d869e1c97654433b", "date": "2026-08-25T08:48:27+02:00", "subject": "fix(core): instrument midldeware issue #17554", "body": "", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "nestjs", "sha": "35a29ccfe1b813c3c72ccb7e3920ad06e516a182", "date": "2026-08-25T08:53:37+02:00", "subject": "fix(core): instrument middleware fallback issue", "body": "", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "nestjs", "sha": "4593b0ad33a225cbcdbae518f5e0163cc0b65607", "date": "2026-08-25T09:12:20+02:00", "subject": "fix(core): preserve middleware arity #17554", "body": "", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "tmux", "sha": "6f4c9b32af0354c4afc5e901600c6b48aca25352", "date": "2026-08-25T07:23:30Z", "subject": "Do not pass NULL to qsort, and add checks for invalid session or window after swap.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "7ebaa889dc098c2adeb90254f7cf933f3609b790", "date": "2026-08-25T09:48:26+02:00", "subject": "chore: drop skip instrumentation - duplicated api", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "angular", "sha": "e8378dfeab39fca5bcdc47126b7859c8fdab7a28", "date": "2026-08-25T15:50:22+08:00", "subject": "fix(common): use locale NaN symbol in number formatting", "body": "Non-finite values all used NumberSymbol.Infinity, so formatNumber(NaN)\nrendered as infinity. Locale data already defines NumberSymbol.NaN.\n\n(cherry picked from commit 7596548e9b836a5d12e966bf296a4dbba044b78f)", "style": "conventional", "type": "fix", "scope": "common", "breaking": false} {"repo": "angular", "sha": "7596548e9b836a5d12e966bf296a4dbba044b78f", "date": "2026-08-25T15:50:22+08:00", "subject": "fix(common): use locale NaN symbol in number formatting", "body": "Non-finite values all used NumberSymbol.Infinity, so formatNumber(NaN)\nrendered as infinity. Locale data already defines NumberSymbol.NaN.", "style": "conventional", "type": "fix", "scope": "common", "breaking": false} {"repo": "eslint", "sha": "0430280e7cca9dc0fdbf0bc50464e98e84285c49", "date": "2026-08-25T10:50:27+03:00", "subject": "chore: improve ecosystem tests compatibility on Windows", "body": "* chore: improve ecosystem tests compatibility on Windows\n\n* run ecosystem tests in CI mode", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "postgres", "sha": "13db0dbad601916b6fb526f777c4949abc08c423", "date": "2026-08-25T09:55:16+02:00", "subject": "Fix accidentally casting away const", "body": "fixup for commit c83a133f2bb", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "a1de07418b89f1b30f9ca088306b2c1615f928c3", "date": "2026-08-25T03:58:17-04:00", "subject": "feat(ui2): lift ui2 options into 'messagesopt' (part 1) #41474", "body": "Problem:\nIn order for ui2 to graduate to the main \"messages ui\" its configuration\nneeds to graduate into actual options.\n\nSolution:\nMigrate some of its config to 'messagesopt':\n- \"maxheight\" (note: currently this is an integer treated as\n a \"percentage\"; if we want to support a row count we could allow\n values with units, like `\"42%\"`)\n- \"pager\"\n- \"timeout\"\n\nAlso improves error messages:\n\n messagesopt=hit-enter,history:500,bogus E474: Unknown item 'bogus'\n messagesopt=hit-enter,history:500,progress:x E474: 'progress' must be one of: , c\n messagesopt=hit-enter,history:abc E474: 'history' requires a number", "style": "conventional", "type": "feat", "scope": "ui2", "breaking": false} {"repo": "vite", "sha": "57f0bf51a08dec0ec507258434cead9b72718e3f", "date": "2026-08-25T16:04:23+08:00", "subject": "docs(guide): fix API documentation grammar and casing", "body": "", "style": "conventional", "type": "docs", "scope": "guide", "breaking": false} {"repo": "postgres", "sha": "5e1e32c2c6c61c411358ef101acb27db3969d4fb", "date": "2026-08-25T10:07:52+02:00", "subject": "Remove old readline API workaround", "body": "The casting away of const is no longer needed.\n\nrl_readline_name has been const since readline-4.2 (2001), libedit\nsince\n2018 (https://cvsweb.netbsd.org/bsdweb.cgi/src/lib/libedit/readline/readline.h?sortby=date,\nrev 1.44)\n\nSee also commit 8666cf65ea6 for a similar fix.\n\nReviewed-by: Heikki Linnakangas \nDiscussion: https://www.postgresql.org/message-id/flat/459217ca-3840-4452-8cd1-8f9b47e8d479%40eisentraut.org", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "8534b0b4c765b06de30d249c9bacecc2d10ac68b", "date": "2026-08-25T10:09:49+02:00", "subject": "Replace px_memset() with explicit_bzero()", "body": "The former was trying to do the same as the latter, but the latter has\nmore robust and portable implementations, and it's the standard across\nthe tree, so replace and remove px_memset().\n\nCo-authored-by: Dagfinn Ilmari Mannsåker \nDiscussion: https://www.postgresql.org/message-id/flat/a035c67f-4aa8-42e0-bb24-7b5abe927af4%40eisentraut.org\nDiscussion: https://www.postgresql.org/message-id/flat/87ldew2yqu.fsf%40wibble.ilmari.org", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "2c654c4ea1b6a960dd6cda1387dd1dd335e6e651", "date": "2026-08-25T20:16:52+12:00", "subject": "Fix incorrect cast in Assert", "body": "This was introduced in e3e26d04b, so backpatch to v19.\n\nReported-by: Lev Nikolaev \nDiscussion: https://postgr.es/m/3e991e23-2260-459f-b3d0-e84fb48d353a@tantorlabs.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "5767fece95ccb58c25cee0567dd83457b8e15b37", "date": "2026-08-25T10:26:50+02:00", "subject": "Change InputFunctionCall* to take a const char *str", "body": "It's reasonable to specify that the input string passed to these\nfunctions isn't modified.\n\nIn each case, the input string is passed down to CStringGetDatum(),\nwhich already takes a const char *, so this change just pushes that\nAPI specification one level up.\n\nA couple of callers benefit from this by no longer having to cast away\na const qualifier.\n\nA few more functions that call input functions internally are changed\nso that they can now offer a const qualified input string to their\ncallers.\n\nReviewed-by: Heikki Linnakangas \nDiscussion: https://www.postgresql.org/message-id/flat/459217ca-3840-4452-8cd1-8f9b47e8d479%40eisentraut.org", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "72dfda5f5ac3825fc26b82fa404225f3e7a0fc31", "date": "2026-08-25T10:38:11+02:00", "subject": "cmake: check libgsasl version at configure time", "body": "To match other dependencies with version requirements.\n\nFollow-up to 3da249e1f0716c06644ed3522a37a8bf81808012 #21609\nCherry-picked from #21404\n\nCloses #22667", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "cbc1e24d1be9d3ca8d0f98802c92e6233ab89532", "date": "2026-08-25T10:41:10+02:00", "subject": "pgcrypto: Consistent spelling of built-in", "body": "Consistently use \"built-in\" instead of \"built in\" and \"builtin\".\n\nAuthor: Erik Rijkers \nReviewed-by: Álvaro Herrera \nDiscussion: https://postgr.es/m/b6d69f32-253c-43ee-290e-f5f4c051a358@xs4all.nl", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "77027b347cf94b38aa93f86d213002cd608230a4", "date": "2026-08-25T17:50:01+09:00", "subject": "Further cleanup related to statistics import support in postgres_fdw.", "body": "* Reorder the conditions in an if-else block for efficiency.\n* Reorder struct definitions for readability.\n* Reorder arguments for some functions for consistency.\n* Move variables to the required scope.\n* Rename a variable to match an input argument.\n* Add/Tweak some comments/docs for clarity.\n\nAuthor: Etsuro Fujita \nReviewed-by: Bharath Rupireddy \nDiscussion: https://postgr.es/m/CAPmGK14otpa7XZyBO5GxDABUK9OPWFLsAwoQP3ugNdXq%3DK4J1g%40mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "81dcba3549b029e789d37c1cc45d44c947039ada", "date": "2026-08-25T10:51:10+02:00", "subject": "cmake/FindLibgsasl: fix to set `LIBGSASL_VERSION` with pkg-config detection", "body": "Follow-up to 16f073ef49f94412000218c9f6ad04e3fd7e4d01 #16973\n\nCloses #22662", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "3d6c4555a78a12899a731c2b37524c02d23b76e7", "date": "2026-08-25T10:52:03+02:00", "subject": "build: don't require the \"vimdoc\" parser to build Nvim", "body": "Problem:\nThe runs \":helptags\", which needs the tree-sitter-vimdoc parser.\nThis may cause problems for package maintainers / distros?\n\nSolution:\nUse running to generate the tags, like the zig build already does.\nIt also errors on duplicate tags, which is good.\n\nThe `:helptags` command still uses treesitter.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "curl", "sha": "b68876a969f569e1bb0812b957509d47cfe41258", "date": "2026-08-25T11:14:30+02:00", "subject": "CVE-2026-80229: OpenSSL provider use-after-free", "body": "openssl: avoid conn reuse if provider is used", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "abdd53533682433048ace23a67f1d964101a642c", "date": "2026-08-25T11:14:30+02:00", "subject": "CVE-2026-80229: OpenSSL provider use-after-free", "body": "Cherry-picked from 7ea37abc6ac0120ba5f6d9\n\nopenssl: avoid conn reuse if provider is used\n\nReported-by: Stanislav Fort\n\nCloses #22665", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "7ea37abc6ac0120ba5f6d94be8d196f7cf1506bb", "date": "2026-08-25T11:14:30+02:00", "subject": "openssl: avoid conn reuse if provider is used", "body": "Reported-by: Stanislav Fort\n\nCloses #22665", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "aed1209c02b81410d7536ed98f573767f1e7b877", "date": "2026-08-25T10:14:49+01:00", "subject": "regress: add Michael K. Darling's damage/redraw regression tests", "body": "From github.com/darlingm/tmux, branch pr5516-regression-fixes. Adds 9\nregression tests covering gaps found in the redraw-damage-rectangles\nbranch: screen-write full/region redraw fallback, same-session window\nswitches, wide-character clipping at damage edges, pane prompts and\nstatus lines surviving damage, floating-pane status format refresh,\nand multi-client damage delivery.\n\nredraw-multiclient.sh is adapted here to use ASCII pane borders\n(pane-border-lines simple) instead of darlingm's original UTF-8\nborders: the original reliably \"failed\" under this test's nested\ntmux-in-tmux harness (relaying through an outer tmux client) due to\nthat harness mis-rendering a cell that held a multi-byte UTF-8 border\ncharacter being overwritten by later plain content - confirmed to be\na nested-relay artifact, not a real bug, by replaying the identical\ndrag sequence against a real terminal (xterm), where it never\nreproduces. ASCII borders avoid the artifact; the test still reliably\ncatches the real \"damage consumed by only one client\" bug it targets\n(verified by reintroducing that bug and confirming the test fails).\n\nThe other 8 tests are added verbatim from darlingm's branch. Four of\nthem (popup-drag-status-line.sh, popup-drag-wide-character.sh,\npopup-drag-pane-prompt.sh, switch-client-redraw.sh) currently FAIL on\nthis branch, since the source fixes they test for have not been\nmerged yet - only the tests are being added here.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "48e33179d4f39896489d925e49d2fd0c2dcc594a", "date": "2026-08-25T10:25:29+01:00", "subject": "server-client, popup, screen-redraw: fix damage-system regressions", "body": "Fixes the four bugs caught by the regression tests added in aed1209c\n(popup-drag-status-line.sh, popup-drag-wide-character.sh,\npopup-drag-pane-prompt.sh, switch-client-redraw.sh), based on fixes\nfrom Michael K. Darling (github.com/darlingm/tmux, pr5516-regression-\nfixes), reviewed and adapted:\n\n- server-client.c: server_client_set_session()'s check for whether the\n client's window actually changed compared old->curw to s->curw, but\n when old == s these read the same, already-updated field, so a\n same-session window switch was never detected. Compare against the\n client's own cached redraw scene instead (redraw_client_has_window(),\n new in screen-redraw.c/tmux.h). Taken from darlingm as-is.\n\n- popup.c: popup_damage() only translated a popup's client-coordinate\n rectangle into window coordinates, so a popup dragged across the\n status line never triggered a status-line redraw once it moved away -\n status_redraw()'s own \"skip if content unchanged\" optimization\n suppressed it, since only the popup moved, not the status content.\n Now detects overlap with the status line and forces a redraw via the\n existing (previously unused) CLIENT_REDRAWSTATUSALWAYS flag, and\n properly clips the reported rectangle to the pane area for\n status-at-top/bottom/off. Taken from darlingm as-is.\n\n- screen-redraw.c: redraw_draw_damage_rect() clipped a span to a damage\n rectangle's raw geometric edges, which have no idea what's in the\n grid, so a clip edge could land mid-character and tear a wide\n character in half. Added redraw_damage_grow_span_clip(): widen the\n clip by one cell on each edge that isn't already at the span's own\n boundary. Reimplemented simpler than darlingm's version (which walked\n grid cells per span type via a switch and direct grid lookups) -\n since no grid cell is ever wider than two columns, an unconditional\n one-cell margin is always enough to pull a split character back in,\n with no need to inspect grid content at all.\n\n- screen-redraw.c: redraw_dr", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "60870831a825f627545836884466b0179701d23c", "date": "2026-08-25T11:25:41+02:00", "subject": "build: fold lists to use one source filename per line", "body": "To reduce merge conflicts.\n\nCloses #22666", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "f94e9eb15ba2a22f69aef234cb81333764d0b298", "date": "2026-08-25T11:34:27+02:00", "subject": "fix(core): preserve middleware arity and harden instance decorator", "body": "", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "tmux", "sha": "f41b983e042774453b9a62e76b60b9dca12c2237", "date": "2026-08-25T10:49:46+01:00", "subject": "screen-write, tty, popup, window: fix remaining untested damage gaps", "body": "Two more fixes based on Michael K. Darling's branch\n(github.com/darlingm/tmux, pr5516-regression-fixes), taken as-is -\nneither is caught by any test in regress/ yet, found by code review\nrather than a failing test:\n\n- screen_write_redraw_cb() (screen-write.c) reported damage for only a\n single row, using ttyctx->ocy as if every fallback redraw were a\n single-cell write. But it's also the callback for cases that can\n legitimately span many rows - a large scroll-region fallback\n (tty_redraw_region(), when tty_large_region() or the pane is\n obscured), a full reset, and entering/leaving the alternate screen.\n For those, only the top row of the affected area ever got marked as\n damaged, leaving the rest stale until an unrelated redraw happened to\n cover it. Changed the shared tty_ctx_redraw_cb typedef to carry\n (py, ny) - the actual row range - and updated every call site to pass\n the range it actually knows about, instead of hardcoding a single\n row.\n\n- window_pane_redraw_floating() (window.c) never refreshed the status\n line after moving/resizing a floating pane, so a status format\n depending on that pane's geometry (e.g. #{pane_width}) could go\n stale until an unrelated status refresh happened. Added a\n server_status_window(w) call.\n\nAlso confirmed the window_pane_scrollbar_intersects() parameter\nnaming cleanup (loop -> wp) discussed earlier was already done in an\nearlier \"Cleanup.\" commit - nothing left to do there.\n\nAll 9 tests in regress/ plus the two pre-existing floating-pane tests\nplus a further 19-test sweep of redraw/tty/input/sync-adjacent\nregress tests pass.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "0c993800e8722de854911020db95d0399206b619", "date": "2026-08-25T12:14:46+02:00", "subject": "chore: fix test types", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "45485b54210e06a517c1ebf86b42b1ea99fc3fe2", "date": "2026-08-25T12:48:22+02:00", "subject": "fix(core): circular durable providers issue #17562", "body": "", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "tmux", "sha": "84a2ccdbcb6bf95117eacd4b4343bbaf2a97d0dc", "date": "2026-08-25T12:02:20+01:00", "subject": "build: warn that --enable-sixel is deprecated", "body": "", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "postgres", "sha": "31ad154242e2dc9fa9bbc4ebbe31f559e2282d9c", "date": "2026-08-25T14:11:35+03:00", "subject": "Fix crash on trying to expand a shared memory hash table", "body": "When a hash table grows large enough we try to expand the directory,\nbut we shouldn't try to do that if the hash table is marked as\nHASH_FIXED_SIZE. Before commit 9fe9ecd516b that was harmless,\nalthough it was a little weird to expand the directory when allocating\nthe new element was doomed to fail later anyway. But the new\nallocator function added in that commit for shared memory hash tables\ndid not expect to be called after hash table creation at all and would\njust crash. Fix by ensuring that the allocator is not called after\nhash table creation for HASH_FIXED_SIZE tables.\n\nThe crash started with commit 9fe9ecd516b in v19, so backpatch to v19.\nTo trigger the crash, you needed a shared memory hash table of just\nthe right size so that we would attempt to resize it: power-of-2 and\nat least HASH_SEGSIZE (256) elements. With any other size, you would\nrun out of allocated elements first.\n\nAuthor: Konstantin Knizhnik \nReviewed-by: Rahila Syed \nDiscussion: https://www.postgresql.org/message-id/d59221f2-b3d2-41ad-8bf0-d581b42e4cba@garret.ru\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "cf06854c62a36744ff26878327df76c989f9fb79", "date": "2026-08-25T14:12:28+03:00", "subject": "refactor(docs-infra): remove the Windows tile icons and other unused icons", "body": "The `msapplication` tile configuration serves no purpose now, so the\n`browserconfig.xml` it points at, the five tile images and the two meta tags\nin `index.html` are removed together.\n\n`shared-docs/icons/twitter.svg` is also unused, left behind when the footer\nmoved to X and Bluesky. Its three siblings in that directory are all still\nreferenced.\n\n(cherry picked from commit a67d4027e239d924d8ae805261b89bc6eedb2c35)", "style": "conventional", "type": "refactor", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "a67d4027e239d924d8ae805261b89bc6eedb2c35", "date": "2026-08-25T14:12:28+03:00", "subject": "refactor(docs-infra): remove the Windows tile icons and other unused icons", "body": "The `msapplication` tile configuration serves no purpose now, so the\n`browserconfig.xml` it points at, the five tile images and the two meta tags\nin `index.html` are removed together.\n\n`shared-docs/icons/twitter.svg` is also unused, left behind when the footer\nmoved to X and Bluesky. Its three siblings in that directory are all still\nreferenced.", "style": "conventional", "type": "refactor", "scope": "docs-infra", "breaking": false} {"repo": "tmux", "sha": "aa6b52eff9b3950cfd2429345a1edf14bf029dff", "date": "2026-08-25T12:14:41+01:00", "subject": "window: redraw border status after moving floating panes", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "9339a7a2de437ed93f9cc3da7f32d0100412d599", "date": "2026-08-25T11:27:58Z", "subject": "fix(platform-server): avoid stripping unicode whitespace during url resolution", "body": "Avoid trimming urlStr with String.prototype.trim() in parseUrl to ensure URL parsing and resolution align with the WHATWG URL standard.", "style": "conventional", "type": "fix", "scope": "platform-server", "breaking": false} {"repo": "angular", "sha": "5aa6d97deb9ef1de14e23748b7fa74f97d183132", "date": "2026-08-25T11:27:58Z", "subject": "fix(platform-server): avoid stripping unicode whitespace during url resolution", "body": "Avoid trimming urlStr with String.prototype.trim() in resolveUrl to ensure URL parsing and resolution align with the WHATWG URL standard.", "style": "conventional", "type": "fix", "scope": "platform-server", "breaking": false} {"repo": "angular", "sha": "71e52d1396b9cef98652929b73e08c4cde645970", "date": "2026-08-25T11:27:58Z", "subject": "fix(platform-server): avoid stripping unicode whitespace during url resolution", "body": "Avoid trimming urlStr with String.prototype.trim() in resolveUrl to ensure URL parsing and resolution align with the WHATWG URL standard.\n\n(cherry picked from commit 3e924cc8dbbb57f23b262cb8f0d7e2bd0673034c)", "style": "conventional", "type": "fix", "scope": "platform-server", "breaking": false} {"repo": "curl", "sha": "7be1e70cb6bcd83e130ecfe8cb91b6a7dcdeff42", "date": "2026-08-25T13:47:11+02:00", "subject": "vtls: move 'native_ca_store' ssl_config_data => ssl_primary_config", "body": "And include it as key for connection reuse matching.\n\nReported-by: Stanislav Fort\n\nCloses #22668", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "74cba75dffbb1697f100767fe37c93e50ad9cda2", "date": "2026-08-25T19:59:24+08:00", "subject": "fix(Drawer): render header when only extra is provided", "body": "* fix(Drawer): render header when only extra is provided\n\n* fix(Drawer): support zero values for title and extra", "style": "conventional", "type": "fix", "scope": "Drawer", "breaking": false} {"repo": "nestjs", "sha": "c921a60ea114c0e1a6e048c0bd52d6c9d92f8cb6", "date": "2026-08-25T20:12:05+08:00", "subject": "fix(fastify): append headers instead of overwriting", "body": "", "style": "conventional", "type": "fix", "scope": "fastify", "breaking": false} {"repo": "babel", "sha": "b5366178f8bcd3f47213396cdc51c7a5e0ac529e", "date": "2026-08-25T08:20:30-04:00", "subject": "fix(generator): keep parens for ?. chain in tag", "body": "", "style": "conventional", "type": "fix", "scope": "generator", "breaking": false} {"repo": "curl", "sha": "f2c03b1c40a5ba819f90cda236d77b1511832223", "date": "2026-08-25T14:28:07+02:00", "subject": "quic: upload improvements", "body": "- set sockopt to receive ECN\n- set socktop to dontfrag\n- vquic: provide ECN on receive to callbacks\n- ngtcp2: manage sendbuf more carefully, do not add\n more than QUIC can keep in flight\n\nCloses #22669", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "babel", "sha": "4cea20d65b950fd10971e5946c2f69ee7e859f55", "date": "2026-08-25T20:38:56+08:00", "subject": "Use built-in glob in Node.js", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "babel", "sha": "2d762de8e58b81fcde6606ea9e9eeef151c48425", "date": "2026-08-25T08:39:39-04:00", "subject": "fix: assign closure body after handling return", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "babel", "sha": "a946b816dbaaad9de654229668bdbc296c359a8d", "date": "2026-08-25T08:57:57-04:00", "subject": "fix: improve for loop closure environment check", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "neovim", "sha": "8ca36b9783d84e29bf6cb085c38e717d832b5594", "date": "2026-08-25T15:01:28+02:00", "subject": "test(harness): support `{retries=…}`", "body": "Problem:\nCannot retry a test with its full `after_each`/`before_each` lifecycle.\n\nSolution:\n- Overload `it()` to accept an `opts` param:\n ```\n it('flaky', { retries = 2 }, function(ctx) end) -- 3 attempts.\n ```\n- Pass `ctx` to test functions.\n- Fix a bug in `t.read_file_list()`.", "style": "conventional", "type": "test", "scope": "harness", "breaking": false} {"repo": "postgres", "sha": "3c982c9bf1af994090a181c68f9407ff39d3ddf7", "date": "2026-08-25T15:25:03+02:00", "subject": "Add test for view referencing labels shared by vertex and edge tables", "body": "Add a test for view containing labels which are shared by both vertex\nand edge tables. When such a label is dropped from only vertex tables\nor only edge tables, the view may be rendered invalid because it does\nnot find any elements associated with the label. Right now, you get\nan error at run time. It would be correct per SQL standard to make\nthis a syntax error (parse time error), which would require some\nrestrict/cascade action when dropping labels. This is not a big\nproblem, because you still get a clean error, but it might be worth\nimproving sometime. Add a test case to record the current behavior.\n\nAuthor: Ashutosh Bapat \nDiscussion: https://www.postgresql.org/message-id/flat/CAExHW5toFV8b984ipuBSAar-W50K%3D%2BXmPOcG5ZUZNP4jrf4zSg%40mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "03dfbe333c589939539eead20d81abc9e2034735", "date": "2026-08-25T15:42:53+02:00", "subject": "test(channel): unreliable \"chansend sends lines…\"", "body": "", "style": "conventional", "type": "test", "scope": "channel", "breaking": false} {"repo": "openssl", "sha": "0d09e9187c4193653f75d3e49d692873124a788a", "date": "2026-08-25T16:12:29+02:00", "subject": "CHANGES.md, NEWS.md: tfix in CVE-2026-54876 description", "body": "Fix a typo in the description of CVE-2026-54876: \"memory\" instead\nof \"nemory\".\n\nFixes: 05e9936df7f0 \"CHANGES.md, NEWS.md: update for 4.0.2\"\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Tom Cosgrove \nReviewed-by: Andrew Dinh \nReviewed-by: Tomas Mraz \nMerge-date: Fri Aug 28 11:17:59 2026\nMerged-from: https://github.com/openssl/openssl/pull/32488", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "f5bb3c33ae300ffc4ad74de25e048926fa6ceae2", "date": "2026-08-25T16:12:29+02:00", "subject": "CHANGES.md, NEWS.md: tfix in CVE-2026-54876 description", "body": "Fix a typo in the description of CVE-2026-54876: \"memory\" instead\nof \"nemory\".\n\nFixes: 05e9936df7f0 \"CHANGES.md, NEWS.md: update for 4.0.2\"\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Tom Cosgrove \nReviewed-by: Andrew Dinh \nReviewed-by: Tomas Mraz \nMerge-date: Fri Aug 28 11:17:41 2026\nMerged-from: https://github.com/openssl/openssl/pull/32488", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "72766ecf8ab41a3384f8e1e1dc8b00c554322b78", "date": "2026-08-25T16:24:01+02:00", "subject": "connection reuse: age check", "body": "If a transfer has CURLOPT_MAXLIFETIME_CONN set, verify the connection\nage of a matched connection before reuse, independent if it is currently\nin use (for multiplexed connections).\n\nCloses #22671", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "37e52a2f463e500fbe3ccbb16b8ae175f5315f21", "date": "2026-08-25T23:34:42+09:00", "subject": "QUIC: fix deferred SSL_listen_ex() connection adoption", "body": "SSL_listen_ex() could attach a deferred incoming channel without\nbinding its QUIC TLS layer or fully adopting the caller-provided\nconnection into the listener hierarchy. The resulting connection could\nnot complete its handshake and retained stale port and ownership state.\n\nKeep deferred channels idle until TLS is attached, bind the handshake\nlayer during peeloff, reparent the connection, and drain unpeeled\ndeferred channels during listener teardown. Make the regression test\nrequire handshake completion and stream I/O.\n\nFixes #32490\nAssisted-by: Pi:Kimi-k3\nAssisted-by: Codex:gpt-5.6-sol\n\nReviewed-by: Neil Horman \nReviewed-by: Saša Nedvědický \nMerge-date: Tue Sep 1 14:19:16 2026\nMerged-from: https://github.com/openssl/openssl/pull/32491", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "51bfb8c606246e21556e37fa312bb636129637fa", "date": "2026-08-25T23:36:06+09:00", "subject": "QUIC: harden SSL_listen_ex() argument and failure handling", "body": "Require new_conn to be a fresh standalone, non-thread-assisted QUIC\nconnection before replacing its transport state. Reject started, shut\ndown, stream-bearing, BIO-configured, already parented, or\nthread-assisted objects instead of destructively adopting them.\n\nPerform allocation and listener refcount setup before removing the\nincoming channel, return -1 for invalid arguments and internal failures,\nand reserve 0 for no connection being available. Document the\nreleased-API behavior change and cover it with driver-managed\nallocation-failure and invalid-target tests.\n\nAssisted-by: Pi:Kimi-k3\nAssisted-by: Codex:gpt-5.6-sol\nReviewed-by: Neil Horman \nReviewed-by: Saša Nedvědický \nMerge-date: Tue Sep 1 14:19:17 2026\nMerged-from: https://github.com/openssl/openssl/pull/32491", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "3aa323584ef61ed16bbc7a84a7eb4f7467199d87", "date": "2026-08-25T16:37:05+02:00", "subject": "test(shell): unreliable \"throttles shell-command output…\"", "body": "", "style": "conventional", "type": "test", "scope": "shell", "breaking": false} {"repo": "openssl", "sha": "f5793a17fd9125e6699010b43ce76a5d8d9bc7b9", "date": "2026-08-25T16:39:40+02:00", "subject": "OSSL_FN: Define OSSL_FN_BITS in include/crypto/fn.h", "body": "OSSL_FN_BITS was defined locally in crypto/fn/fn_local.h and redefined\nin test/fn_api_test.c, while its sibling OSSL_FN_BYTES was defined in\ninclude/crypto/fn.h. Move the definition next to OSSL_FN_BYTES and\ndrop the local duplicates.\n\nAlso replace all 'OSSL_FN_BYTES * 8' spellings with OSSL_FN_BITS.\n\nAssisted-by: Pi:moonshotai/kimi-k3\nSigned-off-by: Richard Levitte \nReviewed-by: Jakub Zelenka \nReviewed-by: Matt Caswell \nMerge-date: Wed Sep 16 12:09:42 2026\nMerged-from: https://github.com/openssl/openssl/pull/32489", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "5634542be580750ffb1a5766470f9e9c72719696", "date": "2026-08-25T19:04:54+03:00", "subject": "docs: Clarify eqeqeq suggestion behavior", "body": "* docs: Clarify eqeqeq suggestion behavior\n\nFixes #21254\n\n* docs: Move eqeqeq link to further reading\n\n* docs: Add eqeqeq further reading metadata", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "curl", "sha": "bb8ec5eb650678ca75e23b23729f25fc10825923", "date": "2026-08-25T12:12:35-04:00", "subject": "strcase: inline the raw case conversions", "body": "Curl_raw_toupper() and Curl_raw_tolower() were out-of-line functions\ndoing a single table lookup, and casecompare() invokes one of them twice\nper byte compared, so every case-insensitive comparison in the library\npaid two calls per byte: header matching, scheme and host compares,\ncookie domains.\n\nMake them static CURL_INLINE functions in strcase.h and give the two\nmapping tables external linkage so the lookup inlines in every\ntranslation unit without relying on LTO or unity mode. The tables and\nthe mapping they perform are unchanged, so this is not a behavior\nchange.\n\ncurl_strequal() on identical header names: -49% on a Cortex-A55, -51%\non a 250MHz e300c3.\n\nCloses #22672", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "ce8a897f985b2cfd1194f6c03fe5388f9bc6d6bf", "date": "2026-08-25T17:37:27+01:00", "subject": "feat(lua): add vim.async", "body": "Problem: Nvim has many Lua APIs that start callback-driven work: timers,\njobs, libuv handles, and other event-loop tasks. Callers that need to\nsequence or cancel that work have to build their own coroutine wrappers,\ntask bookkeeping, and cleanup rules. This makes async control flow hard\nto share, test, and document.\n\nSolution: Add `vim.async`, a structured-concurrency module vendored from\nasync.nvim. It provides task handles, await/pawait helpers,\nsleep/timeout helpers, completion-order iteration, and semaphores on top\nof Nvim's event loop.\n\nThe API follows the same broad model as Trio: async work has an owner,\ntasks are awaited explicitly, and cancellation is cooperative. Include\ngenerated vimdoc with an introductory overview and examples, a news\nentry, and functional tests for the new module.\n\nAI-assisted", "style": "conventional", "type": "feat", "scope": "lua", "breaking": false} {"repo": "angular", "sha": "ea76f7242898235bc8ac7aa132f048c57dfeecef", "date": "2026-08-25T19:55:04+03:00", "subject": "refactor(devtools): use a custom logger and error", "body": "Use a custom logger that prefixes the messages with `[Angular DevTools]`;\nAdd ability for dev-only log messages;\nAdd Angular-DevTools-specific `Error`;\n\n(cherry picked from commit 1371c946aa9ac5e857059cfb6394d5c2a0d83e1a)", "style": "conventional", "type": "refactor", "scope": "devtools", "breaking": false} {"repo": "angular", "sha": "1371c946aa9ac5e857059cfb6394d5c2a0d83e1a", "date": "2026-08-25T19:55:04+03:00", "subject": "refactor(devtools): use a custom logger and error", "body": "Use a custom logger that prefixes the messages with `[Angular DevTools]`;\nAdd ability for dev-only log messages;\nAdd Angular-DevTools-specific `Error`;", "style": "conventional", "type": "refactor", "scope": "devtools", "breaking": false} {"repo": "openssl", "sha": "4a35f52ad31dac49dea226596ed0169d6fb85a1d", "date": "2026-08-25T19:12:35+02:00", "subject": "Fix CVE-2026-63072 test", "body": "Fixes the build/test failure when OpenSSL is configured with\nno-x963kdf.\n\nReviewed-by: Neil Horman \nReviewed-by: Eugene Syromiatnikov \nMerge-date: Wed Aug 26 09:33:00 2026\nMerged-from: https://github.com/openssl/openssl/pull/32496", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "bf69e855beb2c12fb5c8e5cbfdebec5583a533b9", "date": "2026-08-25T20:54:25+03:00", "subject": "refactor(docs-infra): remove example configs for tooling that no longer exists", "body": "An example-config.json marked a directory as a doc example and told aio's\nexample tooling which boilerplate to use and what to run against it. #56496\nremoved that tooling in June 2024. Eight of these fifteen files are empty and no\ncode has read any of them since; the name survives only in the zip and\nStackBlitz exclude lists, which skip it rather than open it.\n\n(cherry picked from commit ddfd8a8e9b8418613eaa1020b76260270f21b275)", "style": "conventional", "type": "refactor", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "ddfd8a8e9b8418613eaa1020b76260270f21b275", "date": "2026-08-25T20:54:25+03:00", "subject": "refactor(docs-infra): remove example configs for tooling that no longer exists", "body": "An example-config.json marked a directory as a doc example and told aio's\nexample tooling which boilerplate to use and what to run against it. #56496\nremoved that tooling in June 2024. Eight of these fifteen files are empty and no\ncode has read any of them since; the name survives only in the zip and\nStackBlitz exclude lists, which skip it rather than open it.", "style": "conventional", "type": "refactor", "scope": "docs-infra", "breaking": false} {"repo": "openssl", "sha": "df49dc92c656e2cd166554d568f76efcfd2e6330", "date": "2026-08-25T20:07:41+02:00", "subject": "Fix CVE-2026-63072 test", "body": "Fixes the build/test failure when OpenSSL is configured with no-x963kdf.\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Nikola Pajkovsky \nMerge-date: Wed Aug 26 09:33:58 2026\nMerged-from: https://github.com/openssl/openssl/pull/32499\n(cherry picked from commit 32e51c18283f4c9095478c7b5f06a44d32c297f3)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "32e51c18283f4c9095478c7b5f06a44d32c297f3", "date": "2026-08-25T20:07:41+02:00", "subject": "Fix CVE-2026-63072 test", "body": "Fixes the build/test failure when OpenSSL is configured with no-x963kdf.\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Nikola Pajkovsky \nMerge-date: Wed Aug 26 09:33:58 2026\nMerged-from: https://github.com/openssl/openssl/pull/32499", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "604bda445ac48af6871d9259e10038944ea90e06", "date": "2026-08-25T14:15:48-04:00", "subject": "test(harness): avoid overlong socket filename #41488", "body": "Problem:\nTest sockets live under `$TMPDIR`, which the harness points at the build\ndir. On macOS/BSD `sockaddr_un.sun_path` is 104 bytes, and a CI build\npath plus \"nvim..\" leaves little room:\n\n /Users/runner/work/neovim/neovim/build/Xtest_tmpdir_terminal/nvim.runner/aBcDeF/nvim.12345.0\n\nSolution:\nPoint XDG_RUNTIME_DIR (`stdpath('run')`) at \"/tmp/nvim_\". 28 bytes:\n\n /tmp/nvim_19916/nvim.19919.1\n\nTODO?:\n- `TEMP_DIR_NAMES` prefers `$TMPDIR` over `/tmp`, so on macOS\n `stdpath('run')` defaults to the long `/var/folders//<…>/T/` path\n instead of the short `/tmp` alias.\n\nNote:\n- The 104-byte limit applies to the `bind()` arg, not its \"realpath\",\n to, so `/tmp/…` symlinks can be used to workaround the limit.", "style": "conventional", "type": "test", "scope": "harness", "breaking": false} {"repo": "tmux", "sha": "8e48c889dd229f4bb3046c4efbe01459aacf7b04", "date": "2026-08-25T19:36:06+01:00", "subject": "Add a resize test.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "b26b240613da0ace7abe80f2c89f74ce11db8eb9", "date": "2026-08-25T18:38:05Z", "subject": "Fix tiled pane resizing with floating layout cells, reported by Kevin Hovsäter.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "9c52dbf216f3b84086a27f0a77d5e5cc3e828274", "date": "2026-08-25T21:18:59+02:00", "subject": "refactor(compiler-cli): ensure eager dependencies used inside defer blocks are not dropped", "body": "When `onlyExplicitDeferDependencyImports` is enabled, dependencies that are used exclusively inside a `@defer` block but are provided eagerly (i.e., in the `imports` array but omitted from `deferredImports`) were incorrectly dropped from the generated `dependencies` array.\n\nThis occurred because the template binder intentionally omits directives used only inside defer blocks from the eager directives list, and the defer block resolver only tracks dependencies explicitly listed in `deferredImports`.\n\nThis commit fixes the issue by ensuring that any template dependency used within a defer block that is not deferred is explicitly preserved as an eager dependency, ensuring it is available at runtime.", "style": "conventional", "type": "refactor", "scope": "compiler-cli", "breaking": false} {"repo": "neovim", "sha": "7ecab736f59c85eb2bacadc1ba67f5f705b9c954", "date": "2026-08-25T15:42:33-04:00", "subject": "ci: no AI advertisements #41492", "body": "Problem:\nClaude and friends love to sprinkle advertisements in commit messages.\n\nSolution:\nAllow a generic \"AI-assisted\" token only.\nReject useless mentions of AI services, sessions, and other cringe.\n\nDisclosure: This PR was authored by an sapient turnip 🫜, raised in\nmoist, fluffy soil, tilled by the blurry wings of a hummingbird.", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "angular", "sha": "60f8a785fd29b9c8a5338ccff3f86c1b54babb2b", "date": "2026-08-25T21:08:41Z", "subject": "refactor(core): defer foreign component rendering to post-update pass", "body": "Foreign components imported via foreignImports and created by the\nɵɵforeignComponent instruction were previously rendered eagerly in the\ncreation phase (rf & 1) of the template. This restricted which\nproperties could be passed to foreign component props, as parent-bound\ninputs (@Input(), input(), input.required()), properties initialized in\nngOnInit(), and pull-based view queries (viewChild()) were not yet\ninitialized at creation time.\n\nThis change defers foreign component rendering to run as a view effect\nduring the update pass:\n- Update ɵɵforeignComponent in core to schedule component rendering via\n createViewEffect (executed in runEffectsInView during refreshView),\n executed with setActiveConsumer(null) to prevent reactive context\n leakage and destroyed immediately upon first execution.\n- Update ɵɵforeignComponent to strictly accept props as a factory function\n (() => props) or null.\n- Update the compiler template pipeline to wrap foreign component props\n in an arrow function closure (() => ({ ... })).\n- Hoist creation-time foreign content projection instructions\n (ɵɵforeignContent, ɵɵforeignContentFn) into creation-phase variable\n declarations before ɵɵforeignComponent so creation-time context is\n captured safely.\n\n(cherry picked from commit 34817da7354f2a4e55f277a991d4345a4ca8a91d)", "style": "conventional", "type": "refactor", "scope": "core", "breaking": false} {"repo": "angular", "sha": "34817da7354f2a4e55f277a991d4345a4ca8a91d", "date": "2026-08-25T21:08:41Z", "subject": "refactor(core): defer foreign component rendering to post-update pass", "body": "Foreign components imported via foreignImports and created by the\nɵɵforeignComponent instruction were previously rendered eagerly in the\ncreation phase (rf & 1) of the template. This restricted which\nproperties could be passed to foreign component props, as parent-bound\ninputs (@Input(), input(), input.required()), properties initialized in\nngOnInit(), and pull-based view queries (viewChild()) were not yet\ninitialized at creation time.\n\nThis change defers foreign component rendering to run as a view effect\nduring the update pass:\n- Update ɵɵforeignComponent in core to schedule component rendering via\n createViewEffect (executed in runEffectsInView during refreshView),\n executed with setActiveConsumer(null) to prevent reactive context\n leakage and destroyed immediately upon first execution.\n- Update ɵɵforeignComponent to strictly accept props as a factory function\n (() => props) or null.\n- Update the compiler template pipeline to wrap foreign component props\n in an arrow function closure (() => ({ ... })).\n- Hoist creation-time foreign content projection instructions\n (ɵɵforeignContent, ɵɵforeignContentFn) into creation-phase variable\n declarations before ɵɵforeignComponent so creation-time context is\n captured safely.", "style": "conventional", "type": "refactor", "scope": "core", "breaking": false} {"repo": "angular", "sha": "b199bdfa2ad8b3fe189e7b4cef9dbea600cd4732", "date": "2026-08-25T21:22:04Z", "subject": "fix(core): explicitly reject foreign components in JIT mode", "body": "Foreign components are only supported in AOT mode. Using them in\nJIT mode previously resulted in silent failures or confusing runtime\nerrors (such as unknown element errors or crashed template ingestion).\n\nThis commit adds explicit validation in JIT compilation:\n- Throws an error during component compilation if `foreignImports` is\n specified on `@Component`.\n- Throws an error during standalone import verification if a foreign\n component is mistakenly passed to `@Component.imports`.\n\n(cherry picked from commit 915a03ae856a595bd5fa0f364928869db8daec19)", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "angular", "sha": "915a03ae856a595bd5fa0f364928869db8daec19", "date": "2026-08-25T21:22:04Z", "subject": "fix(core): explicitly reject foreign components in JIT mode", "body": "Foreign components are only supported in AOT mode. Using them in\nJIT mode previously resulted in silent failures or confusing runtime\nerrors (such as unknown element errors or crashed template ingestion).\n\nThis commit adds explicit validation in JIT compilation:\n- Throws an error during component compilation if `foreignImports` is\n specified on `@Component`.\n- Throws an error during standalone import verification if a foreign\n component is mistakenly passed to `@Component.imports`.", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "angular", "sha": "afe8499a14c1bf8ab42e6e30b35917a873745540", "date": "2026-08-25T22:02:28Z", "subject": "fix(compiler-cli): default template diagnostic related message source file to template", "body": "For external templates (using `templateUrl`), primary diagnostics are\nreported against the synthetic `ts.SourceFile` representing the HTML\ntemplate document. However, secondary related messages (such as those in\n`foreign_component.ts` and `oob.ts`) were explicitly passing the\ncomponent's TypeScript file as `sourceFile`.\n\nBecause the character offsets (`start` and `end`) originate from the\nHTML template AST, associating them with the TypeScript source file\ncaused IDEs and CLI diagnostics to map HTML offsets onto the `.ts` file,\nresulting in corrupt or out-of-bounds source locations.\n\nThis commit resolves the issue by:\n1. Making `sourceFile` optional in `makeTemplateDiagnostic` and related\n checker interfaces (`TemplateTypeChecker`, `TemplateContext`).\n2. Defaulting `relatedMessage.sourceFile` to the template's source file\n (`sf` for external/indirect templates, or the component `.ts` file\n for direct inline templates) when not explicitly provided.\n3. Removing explicit `sourceFile: this.sourceMapping.node.getSourceFile()`\n mappings from `foreign_component.ts` and DOM element checks in\n `oob.ts`, allowing them to automatically resolve to the template file.\n4. Adding unit test coverage for external templates encountering foreign\n component conflicts with related messages.\n\n(cherry picked from commit a46292af260f241da2f9f4e4cbb7d40b14a678f1)", "style": "conventional", "type": "fix", "scope": "compiler-cli", "breaking": false} {"repo": "angular", "sha": "a46292af260f241da2f9f4e4cbb7d40b14a678f1", "date": "2026-08-25T22:02:28Z", "subject": "fix(compiler-cli): default template diagnostic related message source file to template", "body": "For external templates (using `templateUrl`), primary diagnostics are\nreported against the synthetic `ts.SourceFile` representing the HTML\ntemplate document. However, secondary related messages (such as those in\n`foreign_component.ts` and `oob.ts`) were explicitly passing the\ncomponent's TypeScript file as `sourceFile`.\n\nBecause the character offsets (`start` and `end`) originate from the\nHTML template AST, associating them with the TypeScript source file\ncaused IDEs and CLI diagnostics to map HTML offsets onto the `.ts` file,\nresulting in corrupt or out-of-bounds source locations.\n\nThis commit resolves the issue by:\n1. Making `sourceFile` optional in `makeTemplateDiagnostic` and related\n checker interfaces (`TemplateTypeChecker`, `TemplateContext`).\n2. Defaulting `relatedMessage.sourceFile` to the template's source file\n (`sf` for external/indirect templates, or the component `.ts` file\n for direct inline templates) when not explicitly provided.\n3. Removing explicit `sourceFile: this.sourceMapping.node.getSourceFile()`\n mappings from `foreign_component.ts` and DOM element checks in\n `oob.ts`, allowing them to automatically resolve to the template file.\n4. Adding unit test coverage for external templates encountering foreign\n component conflicts with related messages.", "style": "conventional", "type": "fix", "scope": "compiler-cli", "breaking": false} {"repo": "curl", "sha": "ec0d5bb696788b675f05fbe7d7fc31b7206cdce5", "date": "2026-08-26T00:08:51+02:00", "subject": "urldata: sort the connectdata struct fields by size", "body": "To minimize holes\n\nCloses #22678", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "c84079ebf2098c07bcb7b5b942ef74d0f7b70eb0", "date": "2026-08-25T15:20:56-07:00", "subject": "refactor(router): add component input binding support for router resources", "body": "Integrates component input binding (via `withComponentInputBinding()`) with router resources.\n\n(cherry picked from commit df83a340880d690883579b5f886e775276b0ae68)", "style": "conventional", "type": "refactor", "scope": "router", "breaking": false} {"repo": "angular", "sha": "df83a340880d690883579b5f886e775276b0ae68", "date": "2026-08-25T15:20:56-07:00", "subject": "refactor(router): add component input binding support for router resources", "body": "Integrates component input binding (via `withComponentInputBinding()`) with router resources.", "style": "conventional", "type": "refactor", "scope": "router", "breaking": false} {"repo": "postgres", "sha": "dc8628e55d443120f4759a969ec329474a838ee3", "date": "2026-08-26T10:18:07+09:00", "subject": "doc: Update REPACK-related table rewrite documentation", "body": "Adding REPACK introduced table-rewrite behavior that overlaps with\nVACUUM FULL and CLUSTER, but some documentation still mentioned only one\nof these commands or described the progress and locking behavior\nimprecisely.\n\nClarify that table-rewriting operations, including REPACK, can change\nCTIDs. Also describe VACUUM FULL, CLUSTER, and REPACK consistently as\ntable-rewriting operations for disk-space recovery where appropriate,\ndocument the pg_stat_progress_repack and compatibility\npg_stat_progress_cluster views consistently, and clarify the ACCESS\nEXCLUSIVE locking behavior of REPACK, including CONCURRENTLY.\n\nBackpatch to v19, where REPACK was introduced.\n\nAuthor: Marcos Pegoraro \nAuthor: Fujii Masao \nDiscussion: https://postgr.es/m/CAB-JLwaLcHyf=OnLnBsJEbdd57WjWPpK5T7iGFJvAQphjTZj_A@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "f88655b527aed3a64172ab74ecb69a3def18cbfc", "date": "2026-08-25T22:59:44-04:00", "subject": "build(vim-patch): v8.2.0922 is n/a", "body": "", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "openssl", "sha": "25a7beec2bd3a8a63b1c03c79de603b139aaf4ed", "date": "2026-08-26T13:06:55+10:00", "subject": "crypto: remove unused multiprime parameter helper", "body": "Assisted-by: ChatGPT:gpt-5.6Sol\n\nReviewed-by: Shane Lontis \nReviewed-by: Mounir Idrassi \n(Merged from https://github.com/openssl/openssl/pull/32502)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "2f4df67f5d00836979bc72bf6113adb45aff48df", "date": "2026-08-26T10:49:31+05:30", "subject": "Remove the redundant remote_final_lsn variable from the apply worker.", "body": "The apply worker tracked the final LSN of the remote transaction being\napplied in a separate global variable remote_final_lsn, even though\napply_error_callback_arg already tracked the same transaction's xid and\nfinish LSN for error context reporting, and both were updated together\nat the same places.\n\nRename ApplyErrorCallbackArg to ApplyRemoteCtx and treat its\nremote_xid/finish_lsn pair as the generic descriptor of the remote\ntransaction currently being applied, replacing all uses of\nremote_final_lsn so the redundant global can be removed.\n\nThere is no behavior change.\n\nRecording the remote transaction being applied in a single place should\nalso help the proposed conflict log table patch, which needs the same\ninformation when logging a conflict.\n\nAuthor: Zhijie Hou \nAuthor: Amit Kapila \nReviewed-by: Hayato Kuroda \nReviewed-by: shveta malik \nReviewed-by: Dilip Kumar \nDiscussion: https://postgr.es/m/TY4PR01MB177186ED4D38C5356294CC51C94A42@TY4PR01MB17718.jpnprd01.prod.outlook.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "41f3c6fff88ade015669cac5c42db946e0b6f5c9", "date": "2026-08-26T12:37:38+07:00", "subject": "fix(utils): match timestamp query parameter with proper delimiters", "body": "", "style": "conventional", "type": "fix", "scope": "utils", "breaking": false} {"repo": "openssl", "sha": "cc8f5a0e14dfacb4d506d47220b3169f6344a56b", "date": "2026-08-25T23:45:24-07:00", "subject": "Keep GREASE key shares valid after HelloRetryRequest", "body": "SSL_OP_GREASE prepends a reserved key share to every ClientHello. When a\nserver requests a new group in a HelloRetryRequest, the second ClientHello\nmust contain only the requested key share. The extra GREASE entry therefore\ncauses the handshake to fail with a bad key share error.\n\nOmit the GREASE key share only when the retry requests a replacement. Retain\nit for cookie-only retries, where the original key share must be repeated.\n\nExtend the GREASE test to capture both ClientHellos and cover selected-group\nand stateless cookie retries.\n\nAssisted-by: Codex:gpt-5.6-sol\nReviewed-by: Richard Levitte \nReviewed-by: Mounir Idrassi \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:40:03 2026\nMerged-from: https://github.com/openssl/openssl/pull/32505", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "8be35a48bfe787b7d126cb4c74ee1779526df559", "date": "2026-08-26T08:48:10+02:00", "subject": "FAQ: correct an option typo", "body": "Closes #22679", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "12e823ba6d9e7bbc572605442652390252705a0d", "date": "2026-08-26T08:48:14+02:00", "subject": "25-test_verify.t and test/certs/: add mixed 3-level RSA/ECC chain and validation test for it", "body": "Reviewed-by: Mounir Idrassi \nReviewed-by: Andrew Dinh \nMergeDate: Thu Sep 10 13:24:41 2026\n(Merged from https://github.com/openssl/openssl/pull/31611)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "7370bc3cbe52ade78320e1dded028e9ca5b42b66", "date": "2026-08-26T15:10:23+08:00", "subject": "ci: skip all bot handling if skip label exist", "body": "", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "b1e6c185d88a821939514c636d4c50f3ae2b14e2", "date": "2026-08-26T15:20:37+08:00", "subject": "chore(build): migrate to Vite+ 0.3", "body": "", "style": "conventional", "type": "chore", "scope": "build", "breaking": false} {"repo": "curl", "sha": "b73a4f00b2a6748d6f36d737cc3869c9a2e96e1e", "date": "2026-08-26T09:33:27+02:00", "subject": "Makefile.am: improve etags", "body": "The Makefile generated with autotools offers a 'tags' target that\ninvokes etags to build a symbol database in 'TAGS'.\n\nFor some reason it never indexes the symbols in the public include files\nand this is a somewhat crude way to make it to so. The downside with\nthis take is that it repeats the header files names in the root\nMakefile. I still consider it valuable. The file set changes very rarely\nand this tags target is not critical.\n\nCloses #22680", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "fc80bbe1b81ee417a322f6ec91ea6fc701418fbc", "date": "2026-08-26T15:35:50+08:00", "subject": "refactor(runtime-vapor): replace vnode-hook hijacking with internal iu/ibu notification channel", "body": "", "style": "conventional", "type": "refactor", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "f09c102c2077b445873130ca598ae2f2d6e9e849", "date": "2026-08-26T15:35:51+08:00", "subject": "refactor(runtime-vapor): key interop vnode-hook suppression by driving vnode", "body": "", "style": "conventional", "type": "refactor", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "e3387324feb2319e2d3487b82f93fe340eadfc48", "date": "2026-08-26T15:35:51+08:00", "subject": "refactor(runtime-vapor): consolidate slot outlet resolution and hydration", "body": "* refactor(runtime-vapor): restructure the VDOM slot outlet into an explicit state machine\n\n* refactor(runtime-vapor): construct all slot boundary contexts through one factory\n\n* refactor(runtime-vapor): name the exposed-branch rule in one place\n\n* refactor(runtime-vapor): extract the shared slot hydration idioms\n\n* refactor(runtime-vapor): share the deferred slot anchor idioms across hosts\n\n* refactor(runtime-vapor): share the deferred slot attach skeleton\n\n* refactor(runtime-vapor): apply slot outlet cleanup review\n\n* fix(runtime-vapor): keep the hydration graph tree-shakable from the slot outlet\n\n* fix(runtime-vapor): keep the interop slot outlet tree-shakable and minifiable", "style": "conventional", "type": "refactor", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "e178a042dc58f60e449bdd17f01f4dd2ca4fa76e", "date": "2026-08-26T15:35:52+08:00", "subject": "refactor(runtime-vapor): protocolize foreign-content access on interop fragments", "body": "", "style": "conventional", "type": "refactor", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "01728bbddfcbce73be3c110f7304ad3e08dfebc3", "date": "2026-08-26T15:35:52+08:00", "subject": "refactor(runtime-vapor): replace string slot interop markers with symbol keys", "body": "* refactor(runtime-vapor): replace string slot interop markers with symbol keys\n\n* fix: dts tests", "style": "conventional", "type": "refactor", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "37c81f936cbe404f5ec00a6827020febda3c41ce", "date": "2026-08-26T15:35:53+08:00", "subject": "refactor(runtime-vapor): type the interop seam", "body": "* refactor(runtime-vapor): type the interop seam\n\n* chore: improve", "style": "conventional", "type": "refactor", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "c1296028cfaf640d22b88612de145b284a5f5024", "date": "2026-08-26T15:35:53+08:00", "subject": "refactor(runtime-vapor): consolidate the unmount-suspense ambient behind named operations", "body": "", "style": "conventional", "type": "refactor", "scope": "runtime-vapor", "breaking": false} {"repo": "curl", "sha": "c3c70440e8d9938c0bb732595751e8aebc614f24", "date": "2026-08-26T09:41:03+02:00", "subject": "setopt: make NULL `CURLOPT_AWS_SIGV4` disable aws-sigv4 auth", "body": "Disable `CURLAUTH_AWS_SIGV4` when `CURLOPT_AWS_SIGV4` is set to NULL.\nSyncing behavior with `CURLAUTH_HTTPSIG`.\n\nReported-by: Stanislav Fort\n\nCloses #22681", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "58e71306b2b07d584aa351007af714e65b857a82", "date": "2026-08-26T01:01:48-07:00", "subject": "perf(expect): avoid redefining matcher state", "body": "", "style": "conventional", "type": "perf", "scope": "expect", "breaking": false} {"repo": "vite", "sha": "a6a09f4f394c507ccb0a43ba9c66e5a5e36643c7", "date": "2026-08-26T16:05:57+08:00", "subject": "fix(bundled-dev): mark payload delivered on client report", "body": "Claude-Session: https://claude.ai/code/session_01XbpHn3CJVhLzV1iRk8TARS", "style": "conventional", "type": "fix", "scope": "bundled-dev", "breaking": false} {"repo": "vite", "sha": "bfdb22754f792e8f7e40b125659d79ab590d8a7b", "date": "2026-08-26T16:06:27+08:00", "subject": "docs: update development tools list", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vite", "sha": "90d08f0c068a024aa5051a40b1130fb9a46849e5", "date": "2026-08-26T16:11:41+08:00", "subject": "docs: remove outdated locale links", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vite", "sha": "493cc7d43269860fe499a30980d729b0adc93d2c", "date": "2026-08-26T16:13:26+08:00", "subject": "chore: remove useless files", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "curl", "sha": "c30335595736ea30168e8a8916aa9338cc3c6412", "date": "2026-08-26T04:19:08-04:00", "subject": "tool_doswin: fix stdin data truncation", "body": "- Don't do premature write shutdown (SHUT_WR) on the socket used to read\n the stdin data (aka socket_r).\n\n- Don't do premature read shutdown (SHUT_RD) on the socket used to write\n the stdin data (aka socket_w).\n\n- Remove broken setsockopt call for SO_DONTLINGER on socket_r.\n\n- Wait until all data sent from socket_w is acknowledged by socket_r.\n\nPrior to this change a partial shutdown of the write side of the read\nsocket that the main thread reads stdin data from (socket_r) happened\nbefore reading the data from the socket. The partial shutdown caused a\nFIN to be sent which lingered for ~2 minutes (typical OS configured\ntime), and after that time the connection was terminated.\n\nTheory of operation to read stdin on windows is a connection is\nestablished to relay stdin from a dedicated thread (write to socket_w)\nto the main thread (read from socket_r) so reading stdin is\nnon-blocking. Data truncation would occur if this process was not\ncompleted by the OS FIN reply wait time.\n\nNote this FIN wait time is not the same as SO_LINGER wait time, the\nlatter only used by closesocket.\n\nRef: https://github.com/curl/curl/pull/22383#issuecomment-5399969303\n\nCloses https://github.com/curl/curl/pull/22682", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "c27b1e3a9d23b8dc21a20fee6d47eba1ac9d13eb", "date": "2026-08-26T16:25:56+08:00", "subject": "docs: standardize CodeSandbox capitalization", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "735260aa6afa746078cd55a1eded84f19daf986c", "date": "2026-08-26T16:34:20+08:00", "subject": "docs(blog): fix broken size-limit workflow links", "body": "", "style": "conventional", "type": "docs", "scope": "blog", "breaking": false} {"repo": "angular", "sha": "98960556a6d09abae6c6a541bd378113377f8c0d", "date": "2026-08-26T08:40:34Z", "subject": "build: update mcp-b webmcp dependencies to v5", "body": "See associated pull request for more information.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "5df8ec5feb69cd775a1fec374f67d86124c979fe", "date": "2026-08-26T17:03:04+08:00", "subject": "docs: update npm and social links to canonical urls", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "curl", "sha": "f30d9ef341de9199bca97f3a3956676931d20e55", "date": "2026-08-26T11:07:43+02:00", "subject": "http digest: tie peer/credentials on input", "body": "Tie the peer/origin already on input handling and flush stale data when\nit changed.\n\nVerified by test 1985\n\nReported-by: Igor Morgenstern\n\nCloses #22683", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "05f37b2b5eac5feab774a990a64ad16e2d03f5f5", "date": "2026-08-26T14:46:49+05:30", "subject": "Skip relations dropped concurrently in GetSubscriptionRelations().", "body": "GetSubscriptionRelations() can see a pg_subscription_rel row for a\nrelation whose pg_class row was removed by a concurrent DROP. In this\ncase, get_rel_relkind() returns '\\0'. Skip such relations since they no\nlonger exist and do not need synchronization.\n\nReplace the assertion on relkind with an error. Once the dropped\nrelation case is handled, any other unexpected relkind indicates a\nrelation kind that cannot be part of a subscription and should be\nreported rather than ignored.\n\nAuthor: Vignesh C \nReviewed-by: Amit Kapila \nReviewed-by: Bharath Rupireddy \nReviewed-by: Hayato Kuroda \nReviewed-by: Kyotaro Horiguchi \nBackpatch-through: 19\nDiscussion: https://postgr.es/m/CALDaNm3eeKocRtQyUPdg2kaxJ-VAo5pwcNytDKNT1Yc0t2V_ug%40mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "ce165cda4f54e832b35d29be2ff86cfeaaea90a3", "date": "2026-08-26T11:30:12+02:00", "subject": "tests: support switching off the stdout redirect per-test", "body": "The `` tag now supports `option=\"no-stdout\"` for this purpose.\n\nCloses #22684", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "e5727a412badbd1721e91169df26e55b30065a8a", "date": "2026-08-26T11:37:35+02:00", "subject": "GHA: run on `curl-*` branches", "body": "For rock-solid curl.\n\nCloses #22736", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "6fafe470822f1b4b5db38d0d57206e10df96ad0b", "date": "2026-08-26T11:45:42+02:00", "subject": "tool_cb_hdr: do not truncate etags output to stdout", "body": "Verified by test 321\n\nReported-by: Stanislav Fort\nCloses #22685", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "f8992b6a177a2ccfb5dc04137088534bad68681c", "date": "2026-08-26T11:53:36+02:00", "subject": "dist: fix to drop test bundle .c files from the source tarball", "body": "Drop `libtests.c`, `servers.c`, `tunits.c`, `units.c` from source distro\ntarball.\n\nMinor downside: it needs the bundle name be hardcoded into the\n`nodist__SOURCES =` lines.\n\nAlso:\n- confirmed that `CLEANFILES` is still required after this.\n\nRef: #22722 (alternate solution)\nRef: #22723 (experiment)\nFollow-up to 71cf0d1fca9e1f53524e1545ef0c08d174458d80 #14772\n\nCloses #22686", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "b259ece92a951e799d8ace7a865d6269cc1cdd71", "date": "2026-08-26T12:59:18+03:00", "subject": "Fix registering shmem callbacks in single-user mode", "body": "RegisterShmemCallbacks() should not be called from the postmaster\nprocess after postmaster startup. Add an assertion for that, and fix\nthe check for whether it's being called for \"after startup\"\nallocations to take single-user mode into account.\n\nAuthor: Ayush Tiwari \nDiscussion: https://www.postgresql.org/message-id/CAJTYsWU7epjUL1-xrpYeOrmkmq20M2Q22Y_SRb8PWQG8KNQyDw@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "6017e6ffedff391c5e8ec6f0d3a2ca85b5fda26a", "date": "2026-08-26T12:03:11+02:00", "subject": "tool_paramhlp: --proto only supports one modifier", "body": "1 - clarify the documentation\n\n2 - make curl exit with an error for unknown protocols to make users\n better notice when they get it wrong\n\nVerified in test 322\n\nCloses #22687", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "9193fd38ac797a3179db6c2f75b75658cf80a985", "date": "2026-08-26T12:35:43+02:00", "subject": "fixup the parser more", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "ecb03bc7d41c2e85bd2876d308488ed9b9a400c8", "date": "2026-08-26T19:04:37+08:00", "subject": "test: pin rolldown to the pkg.pr.new build of rolldown#10783", "body": "`overrides.rolldown: $rolldown` sends the root version to every package, so the URL\nalso reaches tsdown's `rolldown` subdependency. `blockExoticSubdeps` rejects that,\nso it is turned off while this pin is in place.\n\nNot for merge.", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "neovim", "sha": "6addf6758d7deca2722046d41fbe100f099059ac", "date": "2026-08-26T13:06:17+02:00", "subject": "fix(ui): too many win_float_pos events for the popupmenu (partial fix)", "body": "I got annoyed about so many repeated `win_float_pos` events\nfor the popupmenu. This fixes some of them but not all of them.\n\nvibe-less explanation: `need_highlight_changed = true` is very expensive\nshould not be needed when using a window-local highlight namespace. This\nwas only necessary when overriding the global highlight namespace. This\ncan instead be handled by using the correct `hl_attr_active` instead of\n`highlight_attr`.\n\nAlso `pum_grid.pending_comp_index_update` can be cleared when using\nwin_float_pos prior to redraw.\n\nI wanted to add a test for no repeated win_float_pos event in the same\nredraw:flush cycle but that requires deeper cleanups, like\ngetting rid of crazy redraw panic intermingled deep into insexpand.c\n(if \"state\" is correct, a single update_screen() after the weird\nmultilayer recursive dance should be enough.)", "style": "conventional", "type": "fix", "scope": "ui", "breaking": false} {"repo": "nestjs", "sha": "c9d59f24d44b8d2dc4c602fe6dba47b22b991742", "date": "2026-08-26T13:15:22+02:00", "subject": "chore: expose missing internals", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "neovim", "sha": "d393abad77c54ea84b1474d30e70517ff00ad334", "date": "2026-08-26T19:18:22+08:00", "subject": "test: fix wrong order of args to eq(), neq() #41499", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "vite", "sha": "42bc61afff01ac5fa7bab213c2e2b96c3ca12a7e", "date": "2026-08-26T19:21:07+08:00", "subject": "test: regenerate lockfile with tarball integrity for the rolldown pin", "body": "pnpm 10.34.5 writes URL dependency entries without an `integrity` field, and then\nrefuses to install them (ERR_PNPM_MISSING_TARBALL_INTEGRITY). Regenerated with\n`pnpm@11.23.0 install --update-checksums`, which fills the checksums in, then\nconverted `patchedDependencies` back to the shape pnpm 10 expects.", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "openssl", "sha": "81d84030f432322afa28a858a6756b23d606d921", "date": "2026-08-26T13:24:20+02:00", "subject": "Fix build warnings in quic_srtm.c", "body": "ssl/quic/quic_srtm.c: the SRTM invariant-check helpers use several\nvariables only inside assert() calls, so they trigger unused-variable\nwarnings when assertions are compiled out. Mark them ossl_unused.\n\nReviewed-by: Norbert Pocs \nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nMerge-date: Wed Aug 26 12:38:35 2026\nMerged-from: https://github.com/openssl/openssl/pull/32510\n(cherry picked from commit 5b09bb7d1159a31a7ae0a8f654262dc8858b95c8)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "cda4558815fed8c30bc0d32dc280fbc3c8251971", "date": "2026-08-26T13:24:20+02:00", "subject": "Fix build warnings in fuzzing and ANSI C builds", "body": "ssl/quic/quic_srtm.c: the SRTM invariant-check helpers use several\nvariables only inside assert() calls, so they trigger unused-variable\nwarnings when assertions are compiled out. Mark them ossl_unused.\n\ntest/cmp_vfy_test.c: SETUP_TEST_FIXTURE declares a variable, so invoke\nit together with the other declarations instead of after the first\nstatement, fixing a declaration-after-statement error on ANSI C (C90)\ncompilers.\n\nReviewed-by: Norbert Pocs \nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nMerge-date: Wed Aug 26 12:38:35 2026\nMerged-from: https://github.com/openssl/openssl/pull/32510\n(cherry picked from commit 5b09bb7d1159a31a7ae0a8f654262dc8858b95c8)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "5b09bb7d1159a31a7ae0a8f654262dc8858b95c8", "date": "2026-08-26T13:24:20+02:00", "subject": "Fix build warnings in fuzzing and ANSI C builds", "body": "ssl/quic/quic_srtm.c: the SRTM invariant-check helpers use several\nvariables only inside assert() calls, so they trigger unused-variable\nwarnings when assertions are compiled out. Mark them ossl_unused.\n\ntest/cmp_vfy_test.c: SETUP_TEST_FIXTURE declares a variable, so invoke\nit together with the other declarations instead of after the first\nstatement, fixing a declaration-after-statement error on ANSI C (C90)\ncompilers.\n\nReviewed-by: Norbert Pocs \nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nMerge-date: Wed Aug 26 12:38:35 2026\nMerged-from: https://github.com/openssl/openssl/pull/32510", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "f117f36bb1ae1abbf5c92440c75972a7d0673dc8", "date": "2026-08-26T13:24:20+02:00", "subject": "Fix build warnings in fuzzing and ANSI C builds", "body": "ssl/quic/quic_srtm.c: the SRTM invariant-check helpers use several\nvariables only inside assert() calls, so they trigger unused-variable\nwarnings when assertions are compiled out. Mark them ossl_unused.\n\ntest/cmp_vfy_test.c: SETUP_TEST_FIXTURE declares a variable, so invoke\nit together with the other declarations instead of after the first\nstatement, fixing a declaration-after-statement error on ANSI C (C90)\ncompilers.\n\nReviewed-by: Norbert Pocs \nReviewed-by: Tomas Mraz \nMerge-date: Wed Aug 26 12:45:11 2026\nMerged-from: https://github.com/openssl/openssl/pull/32507", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "280ae314e0aa4a9eb979e0a6da728a634de0b8aa", "date": "2026-08-26T19:41:31+08:00", "subject": "docs(Statistic): correct milliseconds terminology", "body": "", "style": "conventional", "type": "docs", "scope": "Statistic", "breaking": false} {"repo": "openssl", "sha": "91ae3fe14c59bf10f68bd50f1aaee0bc9fd597aa", "date": "2026-08-26T20:43:32+09:00", "subject": "CMS/PKCS7: do not leave stale errors on the queue", "body": "EVP_{CIPHER,MD}_fetch() raise an error when the requested algorithm is\nnot available, unlike EVP_get_{cipher,digest}bynid().\n\nThese are used here only to probe whether particular algorithms are\navailable in the currently active providers. Remove the error queue\nentries so they are not exposed to the user.\n\nFixes: 02d0f65c9f85 \"CMS/PKCS7: use EVP_CIPHER_fetch() for SMIMECapabilities\"\nReviewed-by: Jakub Zelenka \nReviewed-by: Dmitry Belyavskiy \nMerge-date: Thu Sep 3 09:22:38 2026\nMerged-from: https://github.com/openssl/openssl/pull/32513", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "ed0338befd1d865a8ea1fbaa90013a096dedd07a", "date": "2026-08-26T13:46:13+02:00", "subject": "wolfssl: do not run Curl_wssl_setup_x509_store() twice", "body": "The flag preventing a rerun was not set when a cached_store\nwas used.\n\nCURL_UNCONST an ech string for my old wolfssl version\n\nCloses #22689", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "4433f71d6c189b58b2288efc976786e216994e66", "date": "2026-08-26T14:19:03+02:00", "subject": "macos sectrust: fail ocsp verify when not builtin", "body": "When compiling macOS SecTrust with a compiler that does not support\n__builtin_available(), OCSP verfication is not available. This means it\nis not Apple's clang.\n\nInstead of ignoring this, fail the handshake when ocsp status is\nrequested.\n\nCloses #22690", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "71cabc8f7cb3953ce97845a2ce11202e11983841", "date": "2026-08-26T14:28:14+02:00", "subject": "scripts/testnum: provide a few available test numbers", "body": "Closes #22691", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "66a487fae42d0f675f063749424b4755127d58b2", "date": "2026-08-26T14:29:20+02:00", "subject": "s_server: Do not override -Verify with -verify_return_error", "body": "Or the verification flags in instead of overriding\nthe previous value with -verify_return_error.\n\nFixes e5f01903ed7\n\nReviewed-by: Milan Broz \nReviewed-by: Dmitry Belyavskiy \nMerge-date: Thu Aug 27 16:15:34 2026\nMerged-from: https://github.com/openssl/openssl/pull/32512\n(cherry picked from commit d2307ad06d5510352ce2fc9ad2c87461fd60966a)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "d2307ad06d5510352ce2fc9ad2c87461fd60966a", "date": "2026-08-26T14:29:20+02:00", "subject": "s_server: Do not override -Verify with -verify_return_error", "body": "Or the verification flags in instead of overriding\nthe previous value with -verify_return_error.\n\nFixes e5f01903ed7\n\nReviewed-by: Milan Broz \nReviewed-by: Dmitry Belyavskiy \nMerge-date: Thu Aug 27 16:15:34 2026\nMerged-from: https://github.com/openssl/openssl/pull/32512", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "5f6a5e84a5e8cef301e15ac2e8f8e9837b002ec6", "date": "2026-08-26T15:32:43+03:00", "subject": "feat(coverage): switch to `@vitest/istanbuljs` packages", "body": "", "style": "conventional", "type": "feat", "scope": "coverage", "breaking": false} {"repo": "neovim", "sha": "e7ae1b3c10ad16713b0f0353eb2bb445b0e73c89", "date": "2026-08-26T08:32:58-04:00", "subject": "fix(ui2): cmdwin is not special #41508", "body": "Problem:\nUnreliable ui2 test:\n\n FAILED …/ui/messages2_spec.lua @ 277: messages2 multiline messages and pager\n …/ui/messages2_spec.lua:277: Row 1 did not match.\n Expected:\n ...\n |*{1::}echo \"foo\" | echo \"bar\\nbaz\\n\"->repeat(&lines) |\n |*{1::}messages |\n |*{1::}^ |\n ...\n Actual:\n ...\n |*{9:vim.schedule callback: ...ork/neovim/neovim/runt [+7]}|\n\nSolution:\nui2 is doing contortions to handle the old cmdwin behavior; stop doing\nthat, it's no longer necessary since b2bf7bcfb151.", "style": "conventional", "type": "fix", "scope": "ui2", "breaking": false} {"repo": "curl", "sha": "ee4cb4a6c500a1e16dc6a1312640b75bf6080aa1", "date": "2026-08-26T14:38:51+02:00", "subject": "schannel: refresh stream sizes after renegotiation", "body": "Move the SECPKG_ATTR_STREAM_SIZES query to schannel_connect()\ncompletion, where it runs after both initial negotiation and\nrenegotiation. This avoids relying on schannel_send() to refresh the\ncached values later.\n\nSuggested-by: Jay Satiro\n\nCloses https://github.com/curl/curl/pull/22692", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "dad084c3947df430bd0ee62bc96feb389d2bfe2b", "date": "2026-08-26T08:41:02-04:00", "subject": "ci: merge \"ai-assisted\" job into \"label\" #41509", "body": "Problem:\nToo many \"jobs\" listed in the CI report.\n\nSolution:\n- Do the \"ai-assisted\" step in the \"label\" job. No need for them to be\n separate.\n- Also drop the \"already labeled\" check; `gh pr edit --add-label` is\n idempotent.\n- Note: AI-assisted PRs are now labeled only on the \"opened\" event,\n which is fine.", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "curl", "sha": "f8df560ed9faa2ff3d6198b50f55eaf0315fcb39", "date": "2026-08-26T14:46:46+02:00", "subject": "test322: fix for disabled-manual builds", "body": "Follow-up to 6017e6ffedff391c5e8ec6f0d3a2ca\n\nCloses #22693", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "c2e0cbdd8e2467231c5c5216240435af9ba10c91", "date": "2026-08-26T14:55:57+02:00", "subject": "quic: move the RXE definition to a local header", "body": "The RXE structure carries the reference count of a received packet and\nembeds the OSSL_QRX_PKT handed out to the users of the QRX as its first\nmember. It is defined in quic_record_rx.c, so a test that wants to build\na packet without a QRX behind it cannot allocate one or look at its\nreference count, and would have to guess the size of the structure and\nthe offset of the fields it needs.\n\nMove the definition to a local header, following the other local headers\nin ssl/quic, so that a test can include it and construct a packet of its\nown rather than the library having to provide a constructor for it.\n\nAssisted-by: Claude:claude-opus-5\n\nReviewed-by: Tomas Mraz \nReviewed-by: Saša Nedvědický \nMergeDate: Fri Aug 28 14:03:45 2026\n(Merged from https://github.com/openssl/openssl/pull/32515)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "e9b031b1bcb3991b848446458093683adc261dbb", "date": "2026-08-26T14:56:03+02:00", "subject": "test: cover the packet pinning path of QUIC stream reassembly", "body": "Until now every rstream test queued data with a NULL packet, so the\nproduction path where received chunks pin their OSSL_QRX_PKT via\nreference counting was never exercised by any unit test. A NULL packet\nis never passed in production, so the tests now always queue data in a\nmock packet rather than toggling it.\n\nHave test_rstream_simple and test_rstream_random queue every frame in a\nmock packet, and add a new test_rstream_pkt which asserts the reference\ncounting behaviour directly: references held while chunks are buffered,\nshared packets referenced once per frame, references released when frames\nare consumed, dropped by overlapping frames, moved to the ring buffer or\nfreed with the stream, and cleansing of a packet backed chunk wiping\nexactly the chunk data.\n\nAssisted-by: Claude:claude-fable-5\n\nReviewed-by: Tomas Mraz \nReviewed-by: Saša Nedvědický \nMergeDate: Fri Aug 28 14:03:46 2026\n(Merged from https://github.com/openssl/openssl/pull/32515)", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "neovim", "sha": "8d9a798d6d99c60d9434584d0d5dc38f3f57b2bd", "date": "2026-08-26T09:09:17-04:00", "subject": "fix(man): don't grow a highlight into the previous line #41490", "body": "Problem:\nWith these two lines (`\\b` marking a backspace):\n\n f\\bfoo\n xb\\bb\n\nline 1's bold run ends at byte 1 and line 2's begins at byte 1, so line 1\nrenders \"fo\" in bold rather than \"f\", and line 2's \"b\" is not bold at all.\n\nSolution:\nOnly grow a highlight group that is on the current row.", "style": "conventional", "type": "fix", "scope": "man", "breaking": false} {"repo": "curl", "sha": "c878bb36f838d054adc0be27a219409842bb8365", "date": "2026-08-26T15:15:14+02:00", "subject": "openssl: aws-lc ocsp workaround", "body": "SSL_get_tlsext_status_ocsp_resp() is documented to return -1 on a\nmissing OCSP stapling from the server, however AWS-LC is breaking that\nby returning 0.\n\nSince it does not semantically make a difference if the data is reported\nmissing or has 0 length, treat 0 also as a missing stapling.\n\nReported-by: Stanislav Fort\n\nCloses #22694", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "0fe2f5bfc21f4010f86935aebb3defe6d1ceaaf3", "date": "2026-08-26T15:21:44+02:00", "subject": "TLS 1.3: Apply the cap to the lifetime hint in the session", "body": "Commit 5a85e41 added a 7 day cap on the TLSv1.3 ticket lifetime as required by\nRFC 9846, but applied it to a local variable after the value had already been\nstored in the session and after its last use, so it had no effect. The uncapped\nvalue was still returned by SSL_SESSION_get_ticket_lifetime_hint() and used by\nthe client when deciding whether to offer the ticket for resumption. Apply the\ncap to the lifetime hint in the session.\n\n- RFC 9846 4.7.1: Servers MUST NOT use any value greater than 604800 seconds\n (7 days). The value of zero indicates that the ticket should be discarded\n immediately. Clients MUST NOT use tickets for longer than 7 days after\n issuance, regardless of the ticket_lifetime, and MAY delete tickets earlier\n based on local policy.\n\nFixes #32244\n\nReviewed-by: Tom Cosgrove \nReviewed-by: Richard Levitte \nReviewed-by: Mounir Idrassi \nMerge-date: Tue Sep 1 13:47:48 2026\nMerged-from: https://github.com/openssl/openssl/pull/32240\n(cherry picked from commit 66985198d23337651f48c3affbdc7564bcdb903a)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "66985198d23337651f48c3affbdc7564bcdb903a", "date": "2026-08-26T15:21:44+02:00", "subject": "TLS 1.3: Apply the cap to the lifetime hint in the session", "body": "Commit 5a85e41 added a 7 day cap on the TLSv1.3 ticket lifetime as required by\nRFC 9846, but applied it to a local variable after the value had already been\nstored in the session and after its last use, so it had no effect. The uncapped\nvalue was still returned by SSL_SESSION_get_ticket_lifetime_hint() and used by\nthe client when deciding whether to offer the ticket for resumption. Apply the\ncap to the lifetime hint in the session.\n\n- RFC 9846 4.7.1: Servers MUST NOT use any value greater than 604800 seconds\n (7 days). The value of zero indicates that the ticket should be discarded\n immediately. Clients MUST NOT use tickets for longer than 7 days after\n issuance, regardless of the ticket_lifetime, and MAY delete tickets earlier\n based on local policy.\n\nFixes #32244\n\nReviewed-by: Tom Cosgrove \nReviewed-by: Richard Levitte \nReviewed-by: Mounir Idrassi \nMerge-date: Tue Sep 1 13:47:48 2026\nMerged-from: https://github.com/openssl/openssl/pull/32240", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "27f6123a70b9e3f5d39e5987a314769e58bf2921", "date": "2026-08-26T14:43:29+01:00", "subject": "Fix clang-format violation in tls_process_client_hello()", "body": "Fix a minor style violation in ssl/statem/statem_srvr.c.\n\nThe violation arrived with the DTLS 1.3 series and was never caught\nlocally, because git rebase does not run the pre-commit hooks over the\ncommits it replays. CI did not catch it either: the check-style job\nenumerates a pull request's changed files with `gh pr view --json\nfiles`, which returns at most 100 entries, and the DTLS 1.3 pull\nrequest changed 186 files -- ssl/statem/statem_srvr.c fell outside the\nchecked set. The following commit fixes that workflow bug.\n\nAssisted-by: Claude:claude-opus-5\nReviewed-by: Nikola Pajkovsky \nReviewed-by: Paul Dale \nReviewed-by: Tomas Mraz \nMerge-date: Mon Sep 7 14:09:39 2026\nMerged-from: https://github.com/openssl/openssl/pull/32516", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "b4f5e4908b3fb53f18ecc5180bcc6a3e9b9b54e3", "date": "2026-08-26T14:45:09+01:00", "subject": "Check every changed file in the coding style CI job", "body": "The check-style job built its file list with `gh pr view --json files`.\nThat returns only the first 100 entries - gh's GraphQL query asks for\n100 and does not follow further pages - so any pull request touching\nmore than 100 files had the remainder silently skipped, and the job\nreported a pass having never examined them.\n\nThis was not hypothetical. The DTLS 1.3 pull request changed 186 files,\nso 86 were dropped - 41 of them .c/.h files clang-format would have\nexamined - and the job passed while ssl/statem/statem_srvr.c carried a\nreal formatting violation.\n\nUse the REST files endpoint with --paginate instead, which follows every\npage. Two details worth noting: this endpoint names the field\n`filename` rather than the `path` exposed by gh's GraphQL mapping, and\nper_page is set only to reduce the request count, since the REST\ndefault of 30 would fetch the same 186 files over seven requests rather\nthan two.\n\nChecked against that same pull request: the old command yields 100\npaths, the new one yields all 186, matching `git diff --name-only`\nexactly. Running pre-commit over each list against the unfixed tree\nconfirms the difference - the 100-file list passes, the full list fails\non ssl/statem/statem_srvr.c.\n\nAssisted-by: Claude:claude-opus-5\nReviewed-by: Nikola Pajkovsky \nReviewed-by: Paul Dale \nReviewed-by: Tomas Mraz \nMerge-date: Mon Sep 7 14:09:40 2026\nMerged-from: https://github.com/openssl/openssl/pull/32516", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "4777d27a19a4f61c2eee8b6e40a650bcc0a44118", "date": "2026-08-26T23:10:57+09:00", "subject": "fix(terminal): truecolor SGR with colour space id #41491", "body": "libvterm reads the first three sub-parameters of an SGR 38:2 or 48:2 sequence\nas R:G:B, so a colour space id shifts the channels: the empty slot becomes red\nvia CSI_ARG_MISSING truncating to 255, and green and blue move over one.\n\nSkip the colour space id when the colon-separated group holds more than three\narguments. The group length comes from CSI_ARG_HAS_MORE rather than the raw\nargument count, so a following semicolon-separated parameter is not consumed.\n\n(cherry picked from commit 0346958153bb0d991ad684bd8ba30ec2c056f35c)", "style": "conventional", "type": "fix", "scope": "terminal", "breaking": false} {"repo": "neovim", "sha": "0346958153bb0d991ad684bd8ba30ec2c056f35c", "date": "2026-08-26T23:10:57+09:00", "subject": "fix(terminal): truecolor SGR with colour space id #41491", "body": "libvterm reads the first three sub-parameters of an SGR 38:2 or 48:2 sequence\nas R:G:B, so a colour space id shifts the channels: the empty slot becomes red\nvia CSI_ARG_MISSING truncating to 255, and green and blue move over one.\n\nSkip the colour space id when the colon-separated group holds more than three\narguments. The group length comes from CSI_ARG_HAS_MORE rather than the raw\nargument count, so a following semicolon-separated parameter is not consumed.", "style": "conventional", "type": "fix", "scope": "terminal", "breaking": false} {"repo": "openssl", "sha": "4ffef6491996fc4aab919ea8a6bc05f7aaa15b33", "date": "2026-08-26T16:36:28+02:00", "subject": "test: cover a long lagging read of packet backed stream data", "body": "Queue many small contiguous frames that each pin their own packet while\nthe reader stays behind, so a large number of packets are held at once and\nthen released as the data is finally consumed. This exercises the packet\nreference lifecycle at a scale the other tests do not reach, and gives a\nreassembly that copies data out of packets under memory pressure something\nto run against.\n\nAssisted-by: Claude:claude-opus-4-8\n\nReviewed-by: Tomas Mraz \nReviewed-by: Saša Nedvědický \nMergeDate: Fri Aug 28 14:03:47 2026\n(Merged from https://github.com/openssl/openssl/pull/32515)", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "postgres", "sha": "c949e6524c0f306513176e8f96bd8b4bdd49d667", "date": "2026-08-27T02:43:34+12:00", "subject": "Close relations opened specifically for AFTER triggers", "body": "39dcfda2d fixed an incorrect reuse of ResultRelInfos for AFTER triggers\nwhen the ResultRelInfo needed to have a different ri_RootResultRelInfo.\nThat caused an issue in logical replication apply workers as\nfinish_edata() neglects to call ExecCloseResultRelations() and instead\nrelies on ExecCleanupTupleRouting() to close relations opened during\npartitioning's tuple routing. Since 39dcfda2d, because we may have done\nsome additional table_opens() calls due to having to create an additional\nResultRelInfo because of requirements to have a different\nri_RootResultRelInfo, we should now be explicitly closing any relations\nopened on ResultRelInfos in EState's es_trig_target_relations.\n\nSince finish_edate() seems to want to avoid calling\nExecCloseResultRelations(), add a new external function named\nExecCloseTrigTargetRelations().\n\nReported-by: Hayato Kuroda (Fujitsu) \nAuthor: Hayato Kuroda (Fujitsu) \nAuthor: David Rowley \nReviewed-by: Zhijie Hou (Fujitsu) \nDiscussion: https://postgr.es/m/OS9PR01MB121491E7E05950D108AF9A6D8F5A72@OS9PR01MB12149.jpnprd01.prod.outlook.com\nBackpatch-through: 15", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "bb9a5087b8d9fb78f7a3ecfd29faefc37cff4f0d", "date": "2026-08-26T22:55:45+08:00", "subject": "fix(health): powershell \"echo\" fails without an arg #41477", "body": "", "style": "conventional", "type": "fix", "scope": "health", "breaking": false} {"repo": "curl", "sha": "e031bbf85108b4b18fca8ca3a86f7496a0a6fb0b", "date": "2026-08-26T17:01:38+02:00", "subject": "os400: port latest header files changes to ILE/RPG interface", "body": "Closes #22695", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "dfe7be4d2a700c5a2a22f8060358078518166b08", "date": "2026-08-26T15:09:31Z", "subject": "refactor(compiler): emit any as type argument for ɵɵInjectableDeclaration", "body": "The static `ɵprov` field emitted on `@Injectable()` classes uses `ɵɵInjectableDeclaration`.\nWhen a subclass extends a generic `@Injectable()` base class with contravariant parameters\n(such as callback/transformer methods depending on generic type parameters), TypeScript's\nstatic side inheritance check (`typeof Sub extends typeof Super`) fails with `TS2417` because\n`ɵɵInjectableDeclaration` is not assignable to `ɵɵInjectableDeclaration>`.\n\nUsing `any` (`o.DYNAMIC_TYPE`) in `createInjectableType` avoids strict variance checks on\nstatic inheritance for internal Ivy definitions and aligns with other Ivy declaration types.\n\n(cherry picked from commit caeab598c9828a4f9014232d91812addfd44b4b3)", "style": "conventional", "type": "refactor", "scope": "compiler", "breaking": false} {"repo": "angular", "sha": "caeab598c9828a4f9014232d91812addfd44b4b3", "date": "2026-08-26T15:09:31Z", "subject": "refactor(compiler): emit any as type argument for ɵɵInjectableDeclaration", "body": "The static `ɵprov` field emitted on `@Injectable()` classes uses `ɵɵInjectableDeclaration`.\nWhen a subclass extends a generic `@Injectable()` base class with contravariant parameters\n(such as callback/transformer methods depending on generic type parameters), TypeScript's\nstatic side inheritance check (`typeof Sub extends typeof Super`) fails with `TS2417` because\n`ɵɵInjectableDeclaration` is not assignable to `ɵɵInjectableDeclaration>`.\n\nUsing `any` (`o.DYNAMIC_TYPE`) in `createInjectableType` avoids strict variance checks on\nstatic inheritance for internal Ivy definitions and aligns with other Ivy declaration types.", "style": "conventional", "type": "refactor", "scope": "compiler", "breaking": false} {"repo": "openssl", "sha": "6b2ced985e6c6305431888aef532b5de625ea02e", "date": "2026-08-26T17:42:43+02:00", "subject": "test: reassemble small out of order frames and check the bytes", "body": "Deliver a buffer as small packet backed frames in a random order with\noverlapping retransmits, each carrying its own copy of its bytes, then\nread it back and compare. The random order drives insertion at the head,\nthe tail and the middle of the reassembly, and the frame size is swept\nacross the boundary where a design may change how it stores a chunk, so\nshort frames and their overlaps are exercised in every combination with\nand without cleanse. It passes on the current implementation and would\ncatch a reassembly that mishandles any of those.\n\nAssisted-by: Claude:claude-opus-4-8\n\nReviewed-by: Tomas Mraz \nReviewed-by: Saša Nedvědický \nMergeDate: Fri Aug 28 14:03:49 2026\n(Merged from https://github.com/openssl/openssl/pull/32515)", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "openssl", "sha": "3e6538233748eeddcfc19295d23d1dff77e26c74", "date": "2026-08-26T18:03:15+02:00", "subject": "CI: skip rejection-path test under fuzzing builds", "body": "Assisted-by: Claude:claude-fable-5\nReviewed-by: Jakub Zelenka \nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nMerge-date: Thu Aug 27 07:09:00 2026\nMerged-from: https://github.com/openssl/openssl/pull/32521\n(cherry picked from commit 3be64e3d8016161c19ebd86012936d121e208cc1)", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "openssl", "sha": "3be64e3d8016161c19ebd86012936d121e208cc1", "date": "2026-08-26T18:03:15+02:00", "subject": "CI: skip rejection-path test under fuzzing builds", "body": "Assisted-by: Claude:claude-fable-5\nReviewed-by: Jakub Zelenka \nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nMerge-date: Thu Aug 27 07:09:00 2026\nMerged-from: https://github.com/openssl/openssl/pull/32521", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "angular", "sha": "c1ac4b750ab1d0638b4c352214df1f4c4dbda8da", "date": "2026-08-26T19:22:19+03:00", "subject": "refactor(devtools): convert remaining console calls to the custom log", "body": "Convert the remaining calls introduced by the #70254.", "style": "conventional", "type": "refactor", "scope": "devtools", "breaking": false} {"repo": "ant-design", "sha": "554bf7719157f3aff12220aab80d868416b6544a", "date": "2026-08-27T00:56:43+08:00", "subject": "fix(Statistic): preserve semantic nodes for zero values", "body": "", "style": "conventional", "type": "fix", "scope": "Statistic", "breaking": false} {"repo": "ant-design", "sha": "9770453b04cf1b51e6a73ca06d0da45c42951fa0", "date": "2026-08-26T22:28:22+05:30", "subject": "fix(Alert, Empty): render numeric zero for content nodes", "body": "* fix(Alert): render numeric zero for title, description and action\n\n* fix(Empty): render numeric zero for description and children", "style": "conventional", "type": "fix", "scope": "Alert, Empty", "breaking": false} {"repo": "ant-design", "sha": "500177bbe47cd2b1ae088085b37333933baae214", "date": "2026-08-27T00:59:43+08:00", "subject": "fix(Menu): stop icon drift after inline collapse", "body": "* fix(Menu): stop icon drift after inline collapse\n\n* Update components/menu/style/vertical.ts\n\nCo-authored-by: lijianan <574980606@qq.com>\nSigned-off-by: 高艳兵 \n\n* fix(Menu): correct collapsed transition syntax\n\n---------", "style": "conventional", "type": "fix", "scope": "Menu", "breaking": false} {"repo": "openssl", "sha": "00ad61ce2ad576675549514003b850e10a6530be", "date": "2026-08-26T19:54:14+02:00", "subject": "add VS2013 compiler testing job to CI workflow", "body": "Reviewed-by: Tomas Mraz \nReviewed-by: Bob Beck \nReviewed-by: Neil Horman \nMergeDate: Fri Aug 28 16:18:03 2026\n(Merged from https://github.com/openssl/openssl/pull/32537)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "97fd91933df73cfdc73b53832a41b8b9d0d9cefd", "date": "2026-08-26T14:24:53-04:00", "subject": "Accept CRLs whose IDP names the issuer of a certificate without CDP", "body": "When a certificate has no CRL distribution points extension (or none of\nits distribution points matches the CRL), crl_crldp_check() accepted a\nCRL issued by the certificate issuer only if the CRL had no issuing\ndistribution point extension or the IDP carried no distribution point\nname. A CRL whose IDP names the certificate issuer was rejected with\nX509_V_ERR_DIFFERENT_CRL_SCOPE.\n\nRFC 5280, section 6.3.3, last paragraph, requires the opposite:\n\n If the revocation status has not been determined, repeat the process\n above with any available CRLs not specified in a distribution point\n but issued by the certificate issuer. For the processing of such a\n CRL, assume a DP with both the reasons and the cRLIssuer fields\n omitted and a distribution point name of the certificate issuer.\n That is, the sequence of names in fullName is generated from the\n certificate issuer field as well as the certificate issuerAltName\n extension.\n\nStep (b)(2)(i) of the same section then requires that \"one of the names\nin the IDP matches one of the names in the DP\", i.e. the IDP is in scope\nif one of its names is the certificate issuer name or one of the names\nin the certificate's issuerAltName extension.\n\nImplement that check in a new idp_check_issuer() and use it in the\nfallback at the end of crl_crldp_check(). An IDP that only carries a\nURI, or a nameRelativeToCRLIssuer, does not match the certificate issuer\nand is still rejected as out of scope, as before.\n\nExtend make_empty_crl() in test/recipes/25-test_verify.t to accept CRL\nextensions and add two cases: a CRL whose IDP fullName is the CA's\ndirectoryName is accepted for a certificate without CDP, and one whose\nIDP fullName is only a URI is still rejected with \"different CRL scope\".\n\nFixes #23325\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Jakub Zelenka \nReviewed-by: David von Oheimb \nMerge-date: Mon Sep 7 18:38:10 2026\nMerged-from: https://github.com/op", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "ec8aaf845f4b7606ccf3b97247ebddd3548de917", "date": "2026-08-26T12:04:03-07:00", "subject": "Fix crash in subscription refresh on concurrent relation drop.", "body": "Commit 46b4f5c11b0 made the logical replication origin checks quote\nthe schema and relation names they interpolate into the query sent to\nthe publisher. Those names can be NULL, which that commit overlooked.\n\nAlterSubscription_refresh() collects the OIDs of the relations already\npresent in pg_subscription_rel and hands them to\ncheck_publications_origin_tables() and\ncheck_publications_origin_sequences(), which append the\nschema-qualified name of each one to the query so that\nalready-subscribed relations are excluded from the check. The\nrelations are never locked, so one of them can be dropped concurrently\nbefore its name is read, and get_rel_name() and get_namespace_name()\nreturn NULL. quote_literal_cstr() dereferences it and crashes the\nbackend.\n\nThis commit fixes this by skipping a relation whose name is no longer\navailable. A dropped relation is not synchronized anyway, and the\nappended clauses only exclude relations from a check whose sole effect\nis a WARNING, so omitting one can at most produce a spurious WARNING.\n\nThe window is reachable from ALTER SUBSCRIPTION ... REFRESH\nPUBLICATION and from SET, ADD and DROP PUBLICATION, which refresh by\ndefault, but only when copy_data is true and origin is none. Backpatch\nto v16 as commit 46b4f5c11b0 was back-patched that far. The sequence\npath exists only in v19 and later.\n\nThe test is applied to v19 and later only. Adding it to v17 and v18\nwould require enabling injection point support in\nsrc/test/subscription there, and v16 predates injection points\nentirely. That is more test infrastructure churn on stable branches\nthan this fix warrants.\n\nReported-by: SATYANARAYANA NARLAPURAM \nAuthor: SATYANARAYANA NARLAPURAM \nCo-authored-by: Bharath Rupireddy \nReviewed-by: Ajin Cherian \nReviewed-by: Masahiko Sawada \nDiscussion: https://postgr.es/m/CAHg+QDcd_o3707Ey8c8b7HkE-t14g8c0tk8ME3ctywDsh3ut8g@m", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "79b101486c1d792600b79f90579b788385e85878", "date": "2026-08-26T15:38:17-04:00", "subject": "Export subxip[] for snapshots taken during recovery.", "body": "A snapshot taken during recovery stores all of its in-progress XIDs in\nsubxip, every running top-level XID included, leaving xip empty. Unlike\nwith other snapshots, its suboverflowed flag does not mean that subxip\nis redundant. We nevertheless treated it that way during snapshot\nexport, so an importing session could see in-progress transactions as\naborted. This misbehavior could also lead to hint bits being\nincorrectly set on the standby; affected tuples then wrongly appeared\nvisible or invisible to sessions that never imported the snapshot.\n\nTo fix, teach snapshot export to include the subxip[] array regardless\nof the overflow flag when the snapshot is taken during recovery. This\nis in line with how CopySnapshot() and SerializeSnapshot() already\nhandle the same issue.\n\nClaude Code diagnosed this problem. The committed TAP test is a\nsimplified version of the one that it wrote to demonstrate this bug.\n\nOversight in commit 6c2003f8a, which enabled snapshot export and import\nduring recovery.\n\nAuthor: Peter Geoghegan \nAuthor: Bertrand Drouvot \nBug: #17846\nDiscussion: https://postgr.es/m/CAH2-WzmHVeYY%3Dpjz9x8DhhxVjXHX0pvoQ-MdiB1Tt6%3Do2GTiKg%40mail.gmail.com\nDiscussion: https://postgr.es/m/17846-1a0e5ce976f4c01a@postgresql.org\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "594f5e09e733a07613d2cc88c70a160cdab3a6a9", "date": "2026-08-26T22:36:17+02:00", "subject": "dtls: BIO_snprintf() to snprintf() in the DTLS echo demos", "body": "The BIO_snprintf() deprecation series was not rebased over the DTLS\ndemos before merging, which breaks no-deprecated builds with an\nimplicit-function-declaration error. Convert the port formatting\ncalls to snprintf() as done for the other callers in that series.\n\nAssisted-by: Claude:claude-fable-5\n\nReviewed-by: Dmitry Belyavskiy \nReviewed-by: Saša Nedvědický \nMergeDate: Wed Aug 26 22:21:52 2026\n(Merged from https://github.com/openssl/openssl/pull/32533)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "0b5759caa299e2f86abc8b140c56e16fd03cc744", "date": "2026-08-26T22:54:00+02:00", "subject": "CURLOPT_SOCKOPTFUNCTION.md: ALREADY_CONNECTED does not work for HTTP/3", "body": "Reported-by: Stanislav Fort\n\nCloses #22696", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "28471c849c58cb77ea3fc032a8ca09d191d3fd63", "date": "2026-08-26T23:13:13+02:00", "subject": "CODE_STYLE.md: describe how we write typecasts", "body": "Closes #22697", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "4fc45a9b38bdcc38da07236a98116f9fc9e13636", "date": "2026-08-26T21:57:24Z", "subject": "refactor(compiler-cli): index bound directive inputs and outputs in templates", "body": "Update the template indexer to discover and record bound directive inputs (property bindings, static text attributes) and outputs (event bindings).\n\nThis associates template binding identifiers with their target directive or component class declarations, enabling indexers and language tooling to properly resolve and cross-reference bound directive inputs and outputs.", "style": "conventional", "type": "refactor", "scope": "compiler-cli", "breaking": false} {"repo": "openssl", "sha": "45e53df37e1b521c80c377cfee4c709ab832c605", "date": "2026-08-27T07:44:04+09:00", "subject": "QUIC: complete SSL_listen_ex() hierarchy adoption", "body": "Treat domain flags as a property of the QUIC object hierarchy so\nexplicitly configured domains remain effective for listeners, accepted\nconnections and their streams.\n\nKeep the queued channel in place until TLS attachment is committed,\nconfigure deferred qlog state before the channel runs, install the\nlistener message callbacks, and reset the adopted shell to\naccepted-connection mode, blocking and event-handling defaults.\n\nExtend the public API regression to cover explicit-domain ancestry,\ndiffering context flags and modes, callback identity, stream\ninheritance, and listener lifetime.\n\nAssisted-by: Codex:gpt-5.6-sol\nReviewed-by: Neil Horman \nReviewed-by: Saša Nedvědický \nMerge-date: Tue Sep 1 14:19:18 2026\nMerged-from: https://github.com/openssl/openssl/pull/32491", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "ba666ecef61355c51640676aa1b17c97ff4636c0", "date": "2026-08-27T01:53:04+02:00", "subject": "clang-tidy: merge nested `if()`s", "body": "Reported by clang-tidy 23.1.0+ via `readability-redundant-nested-if`.\n\nDo not enforce at this time, due to a handful of places where it 1)\napplies conditionally for non-debug builds (cf-h2-proxy.c), 2)\ndecreases readability (cookie.c, probably could be fixed by using temp\nvariables or code refactor), 3) breaks code structure (protocol.c).\n\nRef: https://clang.llvm.org/extra/clang-tidy/checks/readability/redundant-nested-if.html\n\nCloses #22793", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "5b06840fa25e824a1cc7a466789c35b932cdc315", "date": "2026-08-27T02:05:56+02:00", "subject": "tidy-up: drop redundant parentheses", "body": "Flagged by clang-tidy 23.1.0 (`readability-redundant-parentheses`).\n\nAlso:\n- examples/websocket: add missing parantheses.\n\nCloses #22749", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "6171ccb558abb132eca9f76e38984f0a3ab49ab5", "date": "2026-08-27T08:19:18+08:00", "subject": "fix(Skeleton): hide decorative image placeholder", "body": "* fix(Skeleton): hide decorative image placeholder\n\n* test(Skeleton): update image demo snapshots", "style": "conventional", "type": "fix", "scope": "Skeleton", "breaking": false} {"repo": "vue-core", "sha": "6e1814aa355aad8dd32a89873e95267297ae0609", "date": "2026-08-27T02:19:37+02:00", "subject": "fix(hydration): handle async component unmount before lazy hydration", "body": "", "style": "conventional", "type": "fix", "scope": "hydration", "breaking": false} {"repo": "vue-core", "sha": "a72036f66bcb9a4f3b3737cda19eeba1bc531e15", "date": "2026-08-27T08:20:07+08:00", "subject": "fix(hydration): handle moving unresolved async fragment", "body": "", "style": "conventional", "type": "fix", "scope": "hydration", "breaking": false} {"repo": "vue-core", "sha": "b535917d46d5c6e28c1d3b9185bfa1665127c73a", "date": "2026-08-27T08:20:41+08:00", "subject": "fix(runtime-core): avoid caching unmounted suspense children", "body": "close #15288", "style": "conventional", "type": "fix", "scope": "runtime-core", "breaking": false} {"repo": "vue-core", "sha": "6eaecc14d8151a7d258b80ce0519e5ecc17a2bef", "date": "2026-08-27T08:21:18+08:00", "subject": "fix(v-model): re-sync select when model is overridden in change handler", "body": "close #10505", "style": "conventional", "type": "fix", "scope": "v-model", "breaking": false} {"repo": "vue-core", "sha": "ef82a26775952a665545d3771788bd23b2efec91", "date": "2026-08-27T08:21:42+08:00", "subject": "fix(shared): correctly compare Map and Set values", "body": "close #15320", "style": "conventional", "type": "fix", "scope": "shared", "breaking": false} {"repo": "vue-core", "sha": "8654f35118a1165ac30fd398ceaff8849c44d9d5", "date": "2026-08-26T21:22:13-03:00", "subject": "fix(runtime-core): resolve $el for dev root comment fragment", "body": "close #12680", "style": "conventional", "type": "fix", "scope": "runtime-core", "breaking": false} {"repo": "vue-core", "sha": "cd1974562357dd73101d048f207a694a4e5310ec", "date": "2026-08-27T02:22:33+02:00", "subject": "fix(suspense): don't treat the leaving branch as the fallback while its mount is pending", "body": "close #15332", "style": "conventional", "type": "fix", "scope": "suspense", "breaking": false} {"repo": "vue-core", "sha": "f8d42e1cf2a0ccb43c50fad6e972cb130c53824e", "date": "2026-08-27T05:22:52+05:00", "subject": "fix(runtime-core): keep .trim result when combined with .number v-model modifier", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-core", "breaking": false} {"repo": "vue-core", "sha": "31da934fcbe73ab8ed187bf820c80b24ced474be", "date": "2026-08-27T05:24:27+05:00", "subject": "fix(runtime-dom): support !important on CSS custom properties in style binding", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-dom", "breaking": false} {"repo": "vue-core", "sha": "38577161c0cfcd4e1013eec0030b781a32fc3744", "date": "2026-08-27T05:24:53+05:00", "subject": "fix(compiler-core): handle invalid static arg in same-name v-bind shorthand", "body": "", "style": "conventional", "type": "fix", "scope": "compiler-core", "breaking": false} {"repo": "neovim", "sha": "8ae46cbd81ee80497fe2d74b623953e5a19764dd", "date": "2026-08-26T20:25:11-04:00", "subject": "build(vim-patch): v8.2.3589 is n/a", "body": "termopen() and term_start() are not 1-1 because of Windows.\nSee Vim's \":h job_start\".\n\n```\n{command} can be a String. This works best on MS-Windows. On\nUnix it is split up in white space separated parts to be\npassed to execvp(). Arguments in double quotes can contain\nwhite space.\n\n{command} can be a List, where the first item is the\nexecutable and further items are the arguments. All items are\nconverted to String. This works best on Unix.\n```", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "neovim", "sha": "38039e639b927376df0dbc0bdb779c9400d35a6c", "date": "2026-08-26T20:37:43-04:00", "subject": "build(vim-patch): v8.2.3761 is n/a", "body": "", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "ant-design", "sha": "2e07d06003cbff82780b97bae0b02693d73bfe06", "date": "2026-08-27T08:38:08+08:00", "subject": "docs(Tree): clarify useTree memo dependencies", "body": "", "style": "conventional", "type": "docs", "scope": "Tree", "breaking": false} {"repo": "neovim", "sha": "0b52504cd8dcaa3ce59a088f4ecc6852d17ad38e", "date": "2026-08-26T20:42:55-04:00", "subject": "build(vim-patch): v8.2.4157 is n/a", "body": "", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "openssl", "sha": "ee8da734a879a235b18ae4752d25e5ffec1c4bc4", "date": "2026-08-27T11:09:44+10:00", "subject": "providers: convert non-legacy cipher parameter parsing", "body": "Use generated trie decoders for common, CTS and TDES parameter handling.\n\nAssisted-by: ChatGPT:gpt-5.6Sol\nReviewed-by: Simo Sorce \nReviewed-by: Shane Lontis \nMerge-date: Thu Sep 3 02:16:47 2026\nMerged-from: https://github.com/openssl/openssl/pull/32536", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "b88dacdd3df3bb462e95d49f18d550e211032087", "date": "2026-08-27T11:10:03+10:00", "subject": "providers: correct advertised cipher tls-mac & multiblock AAD parameter typess", "body": "Match TLS MAC and multiblock AAD metadata to the parameter types actually processed.\n\nAssisted-by: ChatGPT:gpt-5.6Sol\nReviewed-by: Simo Sorce \nReviewed-by: Shane Lontis \nMerge-date: Thu Sep 3 02:16:50 2026\nMerged-from: https://github.com/openssl/openssl/pull/32536", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "28f4711b01926b60c2226a8b1ff48b1bc5ea4eb1", "date": "2026-08-27T11:10:57+10:00", "subject": "providers: convert legacy cipher parameter parsing", "body": "Use generated trie decoders for the legacy-provider DES, RC2 and RC5 implementations.\n\nAssisted-by: ChatGPT:gpt-5.6Sol\nReviewed-by: Simo Sorce \nReviewed-by: Shane Lontis \nMerge-date: Thu Sep 3 02:16:55 2026\nMerged-from: https://github.com/openssl/openssl/pull/32536", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "8bd7f02756d303ded5e3d3c9e12a63d6e9ee9775", "date": "2026-08-27T11:12:09+10:00", "subject": "test: cover non-legacy cipher parameter decoders", "body": "Check parameter metadata exposed by common, CTS, multiblock and TDES cipher implementations.\n\nAssisted-by: ChatGPT:gpt-5.6Sol\nReviewed-by: Simo Sorce \nReviewed-by: Shane Lontis \nMerge-date: Thu Sep 3 02:17:00 2026\nMerged-from: https://github.com/openssl/openssl/pull/32536", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "openssl", "sha": "cd9b0832ceb254848562bf8d39c32d2ead0434eb", "date": "2026-08-27T11:12:09+10:00", "subject": "test: cover non-legacy cipher parameter decoders", "body": "Check parameter metadata exposed by common, CTS, multiblock and TDES cipher implementations.\n\nAssisted-by: ChatGPT:gpt-5.6Sol\n\nReviewed-by: Richard Levitte \nReviewed-by: Nikola Pajkovsky \n(Merged from https://github.com/openssl/openssl/pull/32559)", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "openssl", "sha": "8c5cd40b493bec0bf66003779a41531a15418f1d", "date": "2026-08-27T11:12:09+10:00", "subject": "test: cover non-legacy cipher parameter decoders", "body": "Check parameter metadata exposed by common, CTS, multiblock and TDES cipher implementations.\n\nAssisted-by: ChatGPT:gpt-5.6Sol\n\nReviewed-by: Richard Levitte \nReviewed-by: Nikola Pajkovsky \n(Merged from https://github.com/openssl/openssl/pull/32561)\n\n(cherry picked from commit 5af1b7064c8dbadff7d6edd4f5340f6b9679874d)", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "openssl", "sha": "5af1b7064c8dbadff7d6edd4f5340f6b9679874d", "date": "2026-08-27T11:12:09+10:00", "subject": "test: cover non-legacy cipher parameter decoders", "body": "Check parameter metadata exposed by common, CTS, multiblock and TDES cipher implementations.\n\nAssisted-by: ChatGPT:gpt-5.6Sol\n\nReviewed-by: Richard Levitte \nReviewed-by: Nikola Pajkovsky \n(Merged from https://github.com/openssl/openssl/pull/32561)", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "openssl", "sha": "277556a093908c17dd95a5171b3969458fbcd039", "date": "2026-08-27T11:12:09+10:00", "subject": "test: cover non-legacy cipher parameter decoders", "body": "Check parameter metadata exposed by common, CTS, multiblock and TDES cipher implementations.\n\nAssisted-by: ChatGPT:gpt-5.6Sol\n\nReviewed-by: Richard Levitte \nReviewed-by: Nikola Pajkovsky \n(Merged from https://github.com/openssl/openssl/pull/32560)", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "openssl", "sha": "585842e512d244c85e17a0fdaf2114d40bd87028", "date": "2026-08-27T11:13:16+10:00", "subject": "test: cover legacy cipher parameter decoders", "body": "Check DES, RC2 and RC5 parameter metadata.\n\nAssisted-by: ChatGPT:gpt-5.6Sol\nReviewed-by: Simo Sorce \nReviewed-by: Shane Lontis \nMerge-date: Thu Sep 3 02:17:03 2026\nMerged-from: https://github.com/openssl/openssl/pull/32536", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "7840537b6e09712e0bbb6e80189a4c051d1fe031", "date": "2026-08-27T09:42:25+08:00", "subject": "fix: keep transfer pagination on a valid page", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "b6039507fceafac37b06d208119cec2a468286a7", "date": "2026-08-27T09:46:04+08:00", "subject": "fix(Steps): hide decorative panel arrows", "body": "", "style": "conventional", "type": "fix", "scope": "Steps", "breaking": false} {"repo": "vite", "sha": "ee644014aab61e546742b862a7d7b0d6c7d67a7b", "date": "2026-08-27T04:19:58+02:00", "subject": "chore: delete unused `PluginContainerOptions`", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "991a885bb13185020f4b9d7d77dc1b0c3d9b49d1", "date": "2026-08-26T19:48:58-07:00", "subject": "fix(hmr): cover vapor fast-path reload with the hmr updating flag and reset it on failed updates", "body": "", "style": "conventional", "type": "fix", "scope": "hmr", "breaking": false} {"repo": "neovim", "sha": "19516b78b9891392d28c8d32236e8bbf51d23db2", "date": "2026-08-26T23:30:46-04:00", "subject": "build(vim-patch): v9.2.0565 is n/a", "body": "", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "postgres", "sha": "ea9c9432c828c6ba9e38bdd0bccd3f63507594fc", "date": "2026-08-27T10:12:53+05:30", "subject": "Don't choose an invalid index for REPLICA IDENTITY FULL lookups.", "body": "For a REPLICA IDENTITY FULL remote relation whose local counterpart has\nno primary key or replica identity, FindUsableIndexForReplicaIdentityFull()\nchooses the first index of a suitable shape from RelationGetIndexList().\nThat list excludes only indexes that are not indislive, so an invalid\nindex left behind by a failed CREATE INDEX CONCURRENTLY can be selected.\n\nSuch an index need not contain every row. Consequently, changes for rows\nthat it fails to find can be silently dropped as missing-tuple conflicts.\nIf the index contains no rows at all, the scan can instead error out and\ncause the apply worker to exit.\n\nSkip invalid indexes, as the planner does.\n\nAuthor: Mikhail Nikalayeu \nReviewed-by: Miłosz Bieniek \nReviewed-by: Amit Kapila \nReviewed-by: Shlok Kyal \nReviewed-by: Vignesh C \nReviewed-by: Ajin Cherian \nDiscussion: https://postgr.es/m/CADzfLwWuubcbJBDRZ_J1SSqHDNjNmUYSAgf5y=17LxmP401xbw@mail.gmail.com\nBackpatch-through: 16, where it was introduced", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "37979c62fd731157ff2632c3f08f2ee74d86ad09", "date": "2026-08-27T15:40:30+10:00", "subject": "Validate AAD size correctness for MB aes-cbc-hmac cipher", "body": "This was missed by the fuzzer due to the type error in the advertised\nparameter lists. With the corrected list, the fuzzer can trigger crashes.\n\nAssisted-by: ChatGPT:gpt-5.6Sol\nReviewed-by: Simo Sorce \nReviewed-by: Shane Lontis \nMerge-date: Thu Sep 3 02:16:43 2026\nMerged-from: https://github.com/openssl/openssl/pull/32536", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "caff91bf3c8ca6cd0cbe7cfa6c1a155a94a2a0f3", "date": "2026-08-27T07:51:34+02:00", "subject": "ngtcp2: avoid NULL deref in cf_ngtcp2_send", "body": "In disable-verbose builds\n\nPointed out by Codex Security\n\nCloses #22698", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "9ab65a1e443c3a87f60fb628193ece792cbdccfc", "date": "2026-08-26T23:10:46-07:00", "subject": "fix(runtime-vapor): own each dev render generation with a render scope for HMR", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "ant-design", "sha": "c09c0e6647f7a17b3be57141201ca8cee47c4d78", "date": "2026-08-27T14:14:03+08:00", "subject": "fix(List): preserve numeric zero row keys", "body": "", "style": "conventional", "type": "fix", "scope": "List", "breaking": false} {"repo": "curl", "sha": "40e5dbf5808ccacaab9e68f4dfc769e8fd17ead5", "date": "2026-08-27T08:33:30+02:00", "subject": "CVE-2026-80255: secure cookie attribute bypass with tab", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "4f6aa41a0145e930e766775dbe860883d350aa0a", "date": "2026-08-27T08:33:30+02:00", "subject": "cookie: improve TAB handling", "body": "For entries with a leading tab. Verified in test 2885.\n\nReported-by: Stanislav Fort\nCloses #22699", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "a1710d420c3df32cfe81215befaa55e97038f0ae", "date": "2026-08-27T14:53:32+08:00", "subject": "feat: support aria-describedby on handles", "body": "* feat: support aria-describedby on handles\n\n* add types\n\n* patch", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "9e93add8ac96989d44e72fd5c9cd80001e9db885", "date": "2026-08-27T15:08:06+08:00", "subject": "fix(Checkbox): allow group to override global disabled state", "body": "* fix(Checkbox): allow group to override global disabled state\n\n* test(Checkbox): update group demo snapshots\n\n---------", "style": "conventional", "type": "fix", "scope": "Checkbox", "breaking": false} {"repo": "ant-design", "sha": "bcde005e9fe8be0037bcde5390e012086b768d78", "date": "2026-08-27T15:09:25+08:00", "subject": "fix(Alert): restrict global closable type", "body": "", "style": "conventional", "type": "fix", "scope": "Alert", "breaking": false} {"repo": "ant-design", "sha": "a26f19a740a494150ea79debdf1f13b30de73c24", "date": "2026-08-27T15:10:23+08:00", "subject": "fix(Tag): prioritize local CheckableTag styles", "body": "", "style": "conventional", "type": "fix", "scope": "Tag", "breaking": false} {"repo": "ant-design", "sha": "e8fbc23658a323812864d56d53b5d25b73784dd1", "date": "2026-08-27T15:10:47+08:00", "subject": "fix(Table): preserve custom header keydown events", "body": "", "style": "conventional", "type": "fix", "scope": "Table", "breaking": false} {"repo": "curl", "sha": "621e507300e6c83966567b6ae0352ca82544df13", "date": "2026-08-27T09:16:11+02:00", "subject": "libcurl.md: emphasize that the output needs checking", "body": "Closes #22700", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "3c25112b5824d568fe2f0c6f8a081ee322950554", "date": "2026-08-27T09:24:32+02:00", "subject": "chore: update peer deps", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "807f23735492a13eb4d42775184bdf3a67bd6cd6", "date": "2026-08-27T13:10:44+05:30", "subject": "fix(Card, Tag, Breadcrumb, Segmented, FloatButton): render numeric zero for content nodes", "body": "", "style": "conventional", "type": "fix", "scope": "Card, Tag, Breadcrumb, Segmented, FloatButton", "breaking": false} {"repo": "ant-design", "sha": "dc4490a27543b3f3215298dd474aa337ec75234b", "date": "2026-08-27T15:45:10+08:00", "subject": "fix(Grid): clear stale row alignment classes", "body": "", "style": "conventional", "type": "fix", "scope": "Grid", "breaking": false} {"repo": "vitest", "sha": "1c00c94686ff6f96ec756503ed765b31068692a8", "date": "2026-08-27T09:45:43+02:00", "subject": "chore: extend comment policy", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "curl", "sha": "3b547de301342b347fb3c6ea6ae7979070793a66", "date": "2026-08-27T09:47:53+02:00", "subject": "imap: APPEND CRLF fix", "body": "Deny an APPEND operation with CRLF conversion requested as we are unable\nto predict the final upload size and APPEND requires us to announce it\nbefore starting the upload.\n\nThe check for unknown sizes had already been there, but the crlf\nconversion that renders a known size unknown was not accounted for.\n\nCloses #22702", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "53509aa83b1edc8c4f37a7e9d0a7e952d082a4d8", "date": "2026-08-27T10:22:12+02:00", "subject": "Fix potential double-free in dsa_finish()", "body": "dsa_finish() (the default DSA method's finish slot) frees the cached\nMontgomery context, dsa->method_mont_p, without NULLing the pointer.\nOnce the cache has been warmed (e.g. by signing), a subsequent\nfinish call — e.g. via DSA_free() after DSA_set_method() — frees it\nagain.\n\nNULL the pointer after freeing.\n\nFixes: https://github.com/openssl/openssl/issues/32541\nAssisted-by: Pi:moonshotai/kimi-k3\nSigned-off-by: Richard Levitte \nReviewed-by: Jakub Zelenka \nReviewed-by: Matt Caswell \nMerge-date: Sun Aug 30 14:54:56 2026\nMerged-from: https://github.com/openssl/openssl/pull/32543\n(cherry picked from commit 73f3bd88714957382a0a9e6e9e761ddb76db82a4)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "73f3bd88714957382a0a9e6e9e761ddb76db82a4", "date": "2026-08-27T10:22:12+02:00", "subject": "Fix potential double-free in dsa_finish()", "body": "dsa_finish() (the default DSA method's finish slot) frees the cached\nMontgomery context, dsa->method_mont_p, without NULLing the pointer.\nOnce the cache has been warmed (e.g. by signing), a subsequent\nfinish call — e.g. via DSA_free() after DSA_set_method() — frees it\nagain.\n\nNULL the pointer after freeing.\n\nFixes: https://github.com/openssl/openssl/issues/32541\nAssisted-by: Pi:moonshotai/kimi-k3\nSigned-off-by: Richard Levitte \nReviewed-by: Jakub Zelenka \nReviewed-by: Matt Caswell \nMerge-date: Sun Aug 30 14:54:56 2026\nMerged-from: https://github.com/openssl/openssl/pull/32543", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "a628f6fa945cf76c39c2790d1b8c7b1e593ce55b", "date": "2026-08-27T10:22:20+02:00", "subject": "Fix potential double-free in dh_finish()", "body": "dh_finish() (the default DH method's finish slot) frees the cached\nMontgomery context, dh->method_mont_p, without NULLing the pointer.\nOnce the cache has been warmed, a subsequent finish call — e.g. via\nDH_free() after DH_set_method() — frees it again.\n\nNULL the pointer after freeing.\n\nFixes: https://github.com/openssl/openssl/issues/32541\nAssisted-by: Pi:moonshotai/kimi-k3\nSigned-off-by: Richard Levitte \nReviewed-by: Jakub Zelenka \nReviewed-by: Matt Caswell \nMerge-date: Sun Aug 30 14:54:59 2026\nMerged-from: https://github.com/openssl/openssl/pull/32543\n(cherry picked from commit a32df286644d478e08707c13aa2aa5632f596510)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "a32df286644d478e08707c13aa2aa5632f596510", "date": "2026-08-27T10:22:20+02:00", "subject": "Fix potential double-free in dh_finish()", "body": "dh_finish() (the default DH method's finish slot) frees the cached\nMontgomery context, dh->method_mont_p, without NULLing the pointer.\nOnce the cache has been warmed, a subsequent finish call — e.g. via\nDH_free() after DH_set_method() — frees it again.\n\nNULL the pointer after freeing.\n\nFixes: https://github.com/openssl/openssl/issues/32541\nAssisted-by: Pi:moonshotai/kimi-k3\nSigned-off-by: Richard Levitte \nReviewed-by: Jakub Zelenka \nReviewed-by: Matt Caswell \nMerge-date: Sun Aug 30 14:54:59 2026\nMerged-from: https://github.com/openssl/openssl/pull/32543", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "8b362cbc35b29b06223b57581c30a01c8cebf7ed", "date": "2026-08-27T08:41:28Z", "subject": "build: update cypress-io/github-action action to v7.4.3", "body": "See associated pull request for more information.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "postgres", "sha": "8646d1d5904503f7e8cbb7166e857fed63c6e49c", "date": "2026-08-27T10:42:14+02:00", "subject": "Fix inferred property graph keys with INCLUDE columns", "body": "Inferred property graph keys used all attributes stored in the primary\nkey index, causing non-key INCLUDE columns to become part of the graph\nkey. Use only the index's key attributes.\n\nAuthor: Muhammad Taha Naveed \nDiscussion: https://www.postgresql.org/message-id/flat/CAPTqav%2BVUjYgm1jZy0Scy%3D1-PfdgznVj2%3DPwH8disiXF76HEow%40mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "dcc2a2dcd4120543aa8a967e26dcd5dba20053d1", "date": "2026-08-27T10:52:07+02:00", "subject": "evp_extra: test DSA/DH set_method regression", "body": "In test_low_level_dsa_method() and test_low_level_dh_method(), warm the\nMontgomery cache (by signing / deriving) before the *_set_method() call\nso the default finish slot uses the cache pointer for free. After the\nfix on the corresponding issue, this should not crash any more.\n\nThe DSA warm uses DSA_do_sign(); the DH warm uses DH_compute_key()\nwith the key's own pub_key as peer (self-derive, purpose is to touch\nthe Montgomery path, not perform a real derivation).\n\nRefs: https://github.com/openssl/openssl/issues/32541\nAssisted-by: Pi:moonshotai/kimi-k3\nSigned-off-by: Richard Levitte \nReviewed-by: Jakub Zelenka \nReviewed-by: Matt Caswell \nMerge-date: Sun Aug 30 14:55:00 2026\nMerged-from: https://github.com/openssl/openssl/pull/32543\n(cherry picked from commit 47aae188a6bdc65675ea9c2a9e770bef48a5d781)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "47aae188a6bdc65675ea9c2a9e770bef48a5d781", "date": "2026-08-27T10:52:07+02:00", "subject": "evp_extra: test DSA/DH set_method regression", "body": "In test_low_level_dsa_method() and test_low_level_dh_method(), warm the\nMontgomery cache (by signing / deriving) before the *_set_method() call\nso the default finish slot uses the cache pointer for free. After the\nfix on the corresponding issue, this should not crash any more.\n\nThe DSA warm uses DSA_do_sign(); the DH warm uses DH_compute_key()\nwith the key's own pub_key as peer (self-derive, purpose is to touch\nthe Montgomery path, not perform a real derivation).\n\nRefs: https://github.com/openssl/openssl/issues/32541\nAssisted-by: Pi:moonshotai/kimi-k3\nSigned-off-by: Richard Levitte \nReviewed-by: Jakub Zelenka \nReviewed-by: Matt Caswell \nMerge-date: Sun Aug 30 14:55:00 2026\nMerged-from: https://github.com/openssl/openssl/pull/32543", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "6e2334c1e5b47af2e00af3ea9bdf0561eb107135", "date": "2026-08-27T10:05:22+01:00", "subject": "Strengthen the guidance on bulk submission of pull requests", "body": "The previous wording asked contributors planning \"many pull requests\"\nto start with a sample of three or four and open an issue explaining\ntheir process. It relied on the contributor recognising in advance\nthat they were about to submit many, and someone working through a\nbacklog one item at a time might never make that judgement.\n\nState the limit as a property of what is open at any one time rather\nthan of the contributor's intentions, and say what happens when it is\nexceeded. Ask for the three or four to be chosen by importance: the\nsubmitter is the only person who has seen the whole list, so an\nunranked pile moves the triage onto the reviewers.\n\nAssisted-by: Claude:claude-opus-5\nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:28:51 2026\nMerged-from: https://github.com/openssl/openssl/pull/32544", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "dc5422494490d5ac9385384ec2078d0f43504e1d", "date": "2026-08-27T11:05:54+02:00", "subject": "Fix the VS2013 build of crypto/defaults.c", "body": "TEXT(REGISTRY_KEY) widens only the first literal of the concatenation and\nleaves the rest narrow, which MSVC 12.0 rejects with C2308. Open the key\nwith RegOpenKeyExA instead; the value is still read wide.\n\nAssisted-by: Claude Code:claude-opus-5\nCo-Authored-By: Claude Opus 5 (1M context) \n\nReviewed-by: Tomas Mraz \nReviewed-by: Bob Beck \nReviewed-by: Neil Horman \nMergeDate: Fri Aug 28 16:18:05 2026\n(Merged from https://github.com/openssl/openssl/pull/32537)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "052c9eac52d7eb02a61cfe7c70289c13f06a50e1", "date": "2026-08-27T11:16:25+02:00", "subject": "Fix the VS2013 build of providers/fips/fipsprov.c", "body": "MSVC's C compiler has no C99 \"inline\"; use ossl_inline, which expands to\n__inline there. Only built with enable-fips.\n\nAssisted-by: Claude Code:claude-opus-5\nCo-Authored-By: Claude Opus 5 (1M context) \n\nReviewed-by: Tomas Mraz \nReviewed-by: Bob Beck \nReviewed-by: Neil Horman \nMergeDate: Fri Aug 28 16:18:07 2026\n(Merged from https://github.com/openssl/openssl/pull/32537)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "0417bf52969a1d57f6bc92b9abd5c8a01f9bc9fe", "date": "2026-08-27T11:32:55+02:00", "subject": "Fix pg_event_trigger_ddl_commands for GRANT ON PROPERTY GRAPH", "body": "stringify_grant_objtype() treated OBJECT_PROPGRAPH as unused, so\npg_event_trigger_ddl_commands() failed with \"unsupported object type\"\nwhen a ddl_command_end trigger inspected GRANT/REVOKE on a property\ngraph. Return \"PROPERTY GRAPH\" like the GRANT command syntax.\n\nBug: #19637\nReported-by: Alexander Lakhin \nAuthor: Andrey Rachitskiy \nReviewed-by: Fujii Masao \nDiscussion: https://www.postgresql.org/message-id/flat/19637-4446f72945492ed8%40postgresql.org", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "d013a19fe21902ec86ec6f52519ed81d596db32f", "date": "2026-08-27T11:34:36+02:00", "subject": "multi: xfer table initial size and growth", "body": "Fine tune the multi's transfer table initial size and\ngrowth increments:\n- change initial capacity from 512 to 128\n- lower the amount of unused entries to trigger growth\n from a 4th to an 8th of the capacity\n- grow capacity by 50% inside limits of [128, 2048]\n\nThis lets 100 transfers run without need to regrowth\nand performs less reallocations for larger parallelism.\n\nOn 64-bit architectures, the 2048 growth limits grows\nthe table by 16KB which seems a good chunk and prevents\nover-large growth.\n\nCloses #22705", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "02b0c80422bf8b0ae7a3e839d162bb5e28b81e65", "date": "2026-08-27T06:14:24-04:00", "subject": "fix(lua): blast radius of broken `_G.debug` #41507", "body": "Problem:\n`nlua_pcall()` references `_G.debug.traceback`. If user code deletes it\nor breaks it some other way, various Lua features are broken.\n\n _G.debug = nil\n vim.schedule(function() end)\n vim.wait(100)\n\n E5113: Lua chunk: attempt to index a nil value\n stack traceback:\n [C]: in function 'loop_poll'\n [string \"vim/_core/editor\"]:176: in function 'wait'\n crash.lua:3: in main chunk\n PANIC: unprotected error in call to Lua API (attempt to index a nil value)\n\nSolution:\nCheck `_G.debug.traceback` before using it as errfunc. If it's broken,\nomit the traceback and say so in the error message.\n\nNote: We could cache `_G.debug` in LUA_REGISTRYINDEX on startup, but\nthat would prevent plugins from providing custom functionality there\n(and we happen to do so in `tui_spec.lua` for example).", "style": "conventional", "type": "fix", "scope": "lua", "breaking": false} {"repo": "nestjs", "sha": "a333a9dae6169537da3954c5b1ac35202b057fcb", "date": "2026-08-27T12:22:31+02:00", "subject": "sample: upgrade to v12", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "6f51be6fb09e4ea563f749387680fa09c49e789e", "date": "2026-08-27T12:22:39+02:00", "subject": "build: pin `vscode-languageserver-textdocument` version", "body": "This will prevent mismatches with the parent directory", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "curl", "sha": "bdda3a268f0fd1a1bb4b16b9322aa6bc055b8e76", "date": "2026-08-27T12:50:01+02:00", "subject": "mprintf: fix long double output", "body": "curl does not use this, and we discourage users from using these\nfunctions but since we claim printf() compatibility this still seems\nworth fixing.\n\nThis handling downgrades all input long doubles into regular doubles in\nthe output because in our CI platform set they caused problems in\nmultiple places.\n\nVerified in test 557.\n\nReported-by: Stanislav Fort\nCloses #22706", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "dc6adf3cb7e9b166600419baedef1cb7e7328ac6", "date": "2026-08-27T13:18:38+02:00", "subject": "proxy: CONNECT trailers handling", "body": "Make sure trailers from a CONNECT response are marked as such, so\nfiltering them on CURLOPT_SUPPRESS_CONNECT_HEADERS works correctly.\n\nAdd test2349 and new DBG env var to verify.\n\nCloses #22708", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "ba2ea49b3cba05a948d38afe6191a4338e063dce", "date": "2026-08-27T14:31:44+03:00", "subject": "Don't create a shell type for function returning an array", "body": "Refactor the checks in the function to move all the conditions for\nwhen to attempt creating a shell type into one place. Add a check for\nthe array syntax.\n\nIn addition to rejecting array syntax, another user-visible effect is\nthat the error message is now different if the type specified a\ntypmod. You now get \"type does not exist\" instead of the more\nspecific \"type modifier cannot be specified for shell type\". That\nseems better; the implicit shell type creation exists only for\nbackwards compatibility, and it never worked with type modifiers, so\nif there's a type modifier it's most likely not because the user tried\nto create a shell type,\n\nAdd test for the array syntax, the type modifier, and some other cases\nfor which we don't create shell types.\n\nDiscussion: https://www.postgresql.org/message-id/de673feb-41b4-4685-b24b-6408b95e58ab@iki.fi\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "d02476899f22f851729b6c95290bc73a22a3383d", "date": "2026-08-27T13:40:57+02:00", "subject": "setopt: allow setting a referer from CURLINFO_REFERER", "body": "Reported-by: Stanislav Fort\n\nVerified by test 2397\n\nCloses #22709", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "5845bc89b0a5748947c1e7a004fcba93d024a2d9", "date": "2026-08-27T19:44:37+08:00", "subject": "fix(Notification): upgrade non-submitting close controls", "body": "", "style": "conventional", "type": "fix", "scope": "Notification", "breaking": false} {"repo": "nestjs", "sha": "fff48189fd9151d7526cca3ade298630c9f2e289", "date": "2026-08-27T15:53:10+04:00", "subject": "fix(core): load deferred forward-ref providers from their host module", "body": "When a forwardRef'd provider is requested under a request/durable context\nwhile its instance host carries a settled donePromise (inherited from the\nstatic host by cloneStaticInstance), resolveComponentHost defers the load\nwith the consumer's moduleRef instead of the provider's host module.\nloadInstance then misses the token in the wrong module's provider\ncollection and throws a RuntimeException before its try/catch, the\ndeferred catch swallows it, and the host stays pending forever with the\nnever-constructed Object.create placeholder - cached per durable context\nuntil the process restarts.\n\nThe sibling call sites (the plain path and the branch added by #17563)\nalready pass instanceWrapper.host ?? moduleRef; this applies the same\ncorrection to the deferred path. Regression test: a forwardRef cycle\nspanning two modules, made durable transitively through a request-scoped\ndurable factory, consumed from a third module through the sparse\nctor-metadata path.\n\nFollow-up to #17562 / #17563", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "ant-design", "sha": "d58fe298deafb588f83a90dd02db505614078f22", "date": "2026-08-27T19:56:07+08:00", "subject": "docs(react): capitalize JavaScript in introduce.en-US.md", "body": "Corrected the spelling of 'Javascript' to 'JavaScript'.", "style": "conventional", "type": "docs", "scope": "react", "breaking": false} {"repo": "curl", "sha": "9499d987b96ee5350f3e822258f45d2e287c0be0", "date": "2026-08-27T13:58:25+02:00", "subject": "os400: rewrite upper ebcdic wrappers using dynbuf", "body": "Function dyn_addn_CCSID() extends dynbuf strings to arbitrary encodings,\npossibly multi-byte. It converts the input data using iterative iconv()\ncalls and appends the converted characters to the dynbuf string. Most\nEBCDIC wrappers are now based on this function.\n\nAll memory allocations in these wrappers are now performed via the\nlibcurl malloc interface.\n\nCloses #22675", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "2ceeb27078d890514a1e87b0d99f366d9a4a796f", "date": "2026-08-27T15:06:54+03:00", "subject": "fix(core): cancel stale debounce timers to prevent timer leaks", "body": "The debounce() utility scheduled a setTimeout when using a numeric wait value, but the scheduled timer was never cleared when a new value arrived, the observable threw, or the injector was destroyed. This caused pending timers to fire after invalidation and leak beyond the owner's lifecycle.\n\nRefactor timer cancellation into a dedicated helper and track the pending timer id so any stale timer is cleared on new values or teardown. Adds tests covering cancellation on new values, errors, and injector destruction.\n\n(cherry picked from commit 0904f90b13ac2a1bad22a2b43fb232957dba44c3)", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "angular", "sha": "0904f90b13ac2a1bad22a2b43fb232957dba44c3", "date": "2026-08-27T15:06:54+03:00", "subject": "fix(core): cancel stale debounce timers to prevent timer leaks", "body": "The debounce() utility scheduled a setTimeout when using a numeric wait value, but the scheduled timer was never cleared when a new value arrived, the observable threw, or the injector was destroyed. This caused pending timers to fire after invalidation and leak beyond the owner's lifecycle.\n\nRefactor timer cancellation into a dedicated helper and track the pending timer id so any stale timer is cleared on new values or teardown. Adds tests covering cancellation on new values, errors, and injector destruction.", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "curl", "sha": "72352daba01ba8072b6c1c0254cb77860ed1ab06", "date": "2026-08-27T14:36:05+02:00", "subject": "connection reuse: check SSL configs when doing a scheme upgrade", "body": "When reusing a SSL variant of a scheme for a non-SSL transfer,\ndo also check if the SSL configurations match.\n\nCloses #22710", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "3a5cb6574f64fc4484538b0d1336720240856c8f", "date": "2026-08-27T14:52:25+02:00", "subject": "Dynamic compression: do not leave errors behind on success", "body": "When loading the compression libraries dynamically\non demand, do not leave errors in the error queue\nwhen the library is simply missing and DSO_load()\nfails. We return success so there should be no error\nleft behind.\nFixes https://github.com/openssl/openssl/issues/23558\n\nReviewed-by: Richard Levitte \nReviewed-by: Nikola Pajkovsky \nMerge-date: Wed Sep 16 09:14:53 2026\nMerged-from: https://github.com/openssl/openssl/pull/32546", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "b2f2f7fea96415a92d8481c8dc6b413dd3f63f8e", "date": "2026-08-27T14:52:25+02:00", "subject": "Dynamic compression: do not leave errors behind on success", "body": "When loading the compression libraries dynamically\non demand, do not leave errors in the error queue\nwhen the library is simply missing and DSO_load()\nfails. We return success so there should be no error\nleft behind.\nFixes https://github.com/openssl/openssl/issues/23558\n\nReviewed-by: Richard Levitte \nReviewed-by: Nikola Pajkovsky \nMerge-date: Wed Sep 16 09:14:42 2026\nMerged-from: https://github.com/openssl/openssl/pull/32546", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "ce5f88cb62b88ebbd3c2a1e0c2a7f39d580a74aa", "date": "2026-08-27T14:52:25+02:00", "subject": "Dynamic compression: do not leave errors behind on success", "body": "When loading the compression libraries dynamically\non demand, do not leave errors in the error queue\nwhen the library is simply missing and DSO_load()\nfails. We return success so there should be no error\nleft behind.\nFixes https://github.com/openssl/openssl/issues/23558\n\nReviewed-by: Richard Levitte \nReviewed-by: Nikola Pajkovsky \nMerge-date: Wed Sep 16 09:14:31 2026\nMerged-from: https://github.com/openssl/openssl/pull/32546", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "08b9f85fa45435b22084c9a708a04f415b510877", "date": "2026-08-27T14:52:25+02:00", "subject": "Dynamic compression: do not leave errors behind on success", "body": "When loading the compression libraries dynamically\non demand, do not leave errors in the error queue\nwhen the library is simply missing and DSO_load()\nfails. We return success so there should be no error\nleft behind.\nFixes https://github.com/openssl/openssl/issues/23558\n\nReviewed-by: Richard Levitte \nReviewed-by: Nikola Pajkovsky \nMerge-date: Wed Sep 16 09:14:20 2026\nMerged-from: https://github.com/openssl/openssl/pull/32546", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "8249c0206982e5d41f5005e001fdb9d199d6ec8f", "date": "2026-08-27T14:52:25+02:00", "subject": "Dynamic compression: do not leave errors behind on success", "body": "When loading the compression libraries dynamically\non demand, do not leave errors in the error queue\nwhen the library is simply missing and DSO_load()\nfails. We return success so there should be no error\nleft behind.\nFixes https://github.com/openssl/openssl/issues/23558\n\nReviewed-by: Richard Levitte \nReviewed-by: Nikola Pajkovsky \nMerge-date: Wed Sep 16 09:14:08 2026\nMerged-from: https://github.com/openssl/openssl/pull/32546", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "017d9aa5acb453e5066b3649320c53c739ef487d", "date": "2026-08-27T14:52:25+02:00", "subject": "Dynamic compression: do not leave errors behind on success", "body": "When loading the compression libraries dynamically\non demand, do not leave errors in the error queue\nwhen the library is simply missing and DSO_load()\nfails. We return success so there should be no error\nleft behind.\nFixes https://github.com/openssl/openssl/issues/23558\n\nReviewed-by: Richard Levitte \nReviewed-by: Nikola Pajkovsky \nMerge-date: Wed Sep 16 09:13:51 2026\nMerged-from: https://github.com/openssl/openssl/pull/32546", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "9b7e528426ab7b34836271525cd0de7f4fde3e0f", "date": "2026-08-27T10:04:33-04:00", "subject": "fix(scrollbind): window jumps back and forth with tall virt_lines block", "body": "Problem:\nScrolling a 'scrollbind' window whose peer has a virt_lines block\ntaller than the window makes the peer jump back and forth.\n\nSolution:\nAfter scrollup()/scrolldown(), sync w_scbind_pos to the actual\nget_vtopline() when the target was not clamped to buffer bounds.\n\nRef: esmuellert/codediff.nvim#519\nAI-assisted\n\n(cherry picked from commit b781e95a24bd2d201f4afbe4729635114893cdf5)", "style": "conventional", "type": "fix", "scope": "scrollbind", "breaking": false} {"repo": "neovim", "sha": "b781e95a24bd2d201f4afbe4729635114893cdf5", "date": "2026-08-27T10:04:33-04:00", "subject": "fix(scrollbind): window jumps back and forth with tall virt_lines block", "body": "Problem:\nScrolling a 'scrollbind' window whose peer has a virt_lines block\ntaller than the window makes the peer jump back and forth.\n\nSolution:\nAfter scrollup()/scrolldown(), sync w_scbind_pos to the actual\nget_vtopline() when the target was not clamped to buffer bounds.\n\nRef: esmuellert/codediff.nvim#519\nAI-assisted", "style": "conventional", "type": "fix", "scope": "scrollbind", "breaking": false} {"repo": "angular", "sha": "d457cc9d2e44d539e429bfad71829eac2dbb74e8", "date": "2026-08-27T16:14:10+02:00", "subject": "fix(zone.js): preserve accessor and non-enumerable event listener options", "body": "`copyEventListenerOptions` copied the caller's options with `{...options}`\nbefore forwarding to the native `addEventListener`. Object spread only\ncopies own enumerable data properties, whereas the native call reads\neach dictionary member via WebIDL — a plain `[[Get]]` per member, which\ninvokes accessors and ignores enumerability. The copy was therefore\nlossy in a way the native call is not:\n\n- `Object.defineProperty(opts, 'passive', { get })` (the shape used by\n MDN's passive-listener feature test) — the getter was never invoked,\n so libraries that use the feature test fall back to the legacy boolean\n and register every listener as non-passive.\n- `Object.defineProperty(opts, 'capture', { get: () => true })` — the\n listener was silently registered on the bubbling phase.\n- `Object.defineProperty(opts, 'once', { get: () => true })` — the\n listener fired on every dispatch.\n\n`signal` was already special-cased for `AbortController.prototype.signal`\nafter #54142; that patch generalises the workaround to every recognised\nmember.\n\nThe copy itself was the correct fix for #54142 (frozen/readonly options)\nand is preserved. The fix reads each recognised member from the source\nvia `[[Get]]` when the spread did not, which recovers accessors and\nnon-enumerable properties without double-invoking any getter. The list\nof recognised members is hoisted to module scope so it isn't allocated\non every `patchEventTarget` invocation.\n\nThe call site is reordered to `buildEventListenerOptions(\ncopyEventListenerOptions(...))` so the passive-events code path also\nspreads a normalised data object rather than the caller's raw input.\n\nFixes #70431\n\nCo-authored-by: Matthieu Riegler \n(cherry picked from commit 4c4a705ce38d08e087b599fc243a9d7dc40966a5)", "style": "conventional", "type": "fix", "scope": "zone.js", "breaking": false} {"repo": "angular", "sha": "4c4a705ce38d08e087b599fc243a9d7dc40966a5", "date": "2026-08-27T16:14:10+02:00", "subject": "fix(zone.js): preserve accessor and non-enumerable event listener options", "body": "`copyEventListenerOptions` copied the caller's options with `{...options}`\nbefore forwarding to the native `addEventListener`. Object spread only\ncopies own enumerable data properties, whereas the native call reads\neach dictionary member via WebIDL — a plain `[[Get]]` per member, which\ninvokes accessors and ignores enumerability. The copy was therefore\nlossy in a way the native call is not:\n\n- `Object.defineProperty(opts, 'passive', { get })` (the shape used by\n MDN's passive-listener feature test) — the getter was never invoked,\n so libraries that use the feature test fall back to the legacy boolean\n and register every listener as non-passive.\n- `Object.defineProperty(opts, 'capture', { get: () => true })` — the\n listener was silently registered on the bubbling phase.\n- `Object.defineProperty(opts, 'once', { get: () => true })` — the\n listener fired on every dispatch.\n\n`signal` was already special-cased for `AbortController.prototype.signal`\nafter #54142; that patch generalises the workaround to every recognised\nmember.\n\nThe copy itself was the correct fix for #54142 (frozen/readonly options)\nand is preserved. The fix reads each recognised member from the source\nvia `[[Get]]` when the spread did not, which recovers accessors and\nnon-enumerable properties without double-invoking any getter. The list\nof recognised members is hoisted to module scope so it isn't allocated\non every `patchEventTarget` invocation.\n\nThe call site is reordered to `buildEventListenerOptions(\ncopyEventListenerOptions(...))` so the passive-events code path also\nspreads a normalised data object rather than the caller's raw input.\n\nFixes #70431", "style": "conventional", "type": "fix", "scope": "zone.js", "breaking": false} {"repo": "openssl", "sha": "99e44fd5676c8363325373f2811bbdd9364aacef", "date": "2026-08-27T16:17:32+02:00", "subject": "Don't use BIO_snprintf as it is deprecated", "body": "Signed-off-by: Norbert Pocs \nReviewed-by: Milan Broz \nReviewed-by: Tomas Mraz \nMerge-date: Thu Aug 27 14:35:30 2026\nMerged-from: https://github.com/openssl/openssl/pull/32547", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "04689f128884f9456ba152aff823dd195f0e0059", "date": "2026-08-27T19:28:23+05:00", "subject": "fix(core): omit connection header on http2 sse responses", "body": "", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "postgres", "sha": "4c5291435d40a0cf02b131b680a71e42243fcd04", "date": "2026-08-27T23:40:20+09:00", "subject": "Report specific SQLSTATEs for stats restore errors", "body": "The pg_restore_*_stats() functions could report SQLSTATE XX000 for\ninvalid variadic arguments, such as an unmatched name/value pair, a NULL\nargument name, or a non-text argument name. Attribute and extended\nstatistics restores could also report XX000 when the supplied statistics\nexceed the number of slots PostgreSQL can store.\n\nThese are not internal errors. They result from invalid caller input or\na PostgreSQL implementation limit, but the lack of specific SQLSTATEs\nmade clients treat them as internal errors.\n\nAssign appropriate SQLSTATEs to these errors so that applications and\ntests can classify them correctly.\n\nBackpatch to v19, but no further; changing ERRCODE assignments in\nreleased stable branches doesn't seem like a good idea.\n\nBug: #19629\nReported-by: Zheng Wang \nReported-by: Yanjie Zhao\nReported-by: Yiyang Liu\nAuthor: Fujii Masao \nDiscussion: https://postgr.es/m/19629-76babc04b683594d@postgresql.org\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "61df463c5e144d7f98515917303838b9a1a2876b", "date": "2026-08-27T10:48:33-04:00", "subject": "fix(cmdatom): insert-session entered by feedkeys() ignores typed input #41518", "body": "Problem:\nAn insert-session entered by a scheduled `feedkeys('i','n')` is\nclassified on entry (as \"not typed\"), and not reevaluated after that,\nthus user input following it is not captured.\n\nRepro: an `:imap` that does `feedkeys('','n')` and schedules\nre-entering insert, emits one CmdAtom for the first press, then nothing\nelse.\n\nSolution:\nSample `maptick` (ticked by `gotchars()` on typed input, including\nmappings) at session start; if it advanced by session end, the session\nis user input.", "style": "conventional", "type": "fix", "scope": "cmdatom", "breaking": false} {"repo": "postgres", "sha": "321ec8ee1c016fde47aeebb9e0b3882bcc4c6b97", "date": "2026-08-27T23:57:47+09:00", "subject": "Stabilize 019_replslot_limit", "body": "The test assumed that advancing WAL would lead to a checkpoint that\ninvalidates the obsolete replication slot. If a checkpoint that started\nbefore the WAL switch completes first, the following checkpoint can be\nskipped as idle, so the expected walsender termination is not logged.\n\nForce a CHECKPOINT in a background psql session after advancing WAL, so\nthe slot invalidation is exercised deterministically.\n\nThis has been observed on buildfarm members alligator and partridge:\nhttps://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=alligator&dt=2024-12-13%2001%3A24%3A58\nhttps://buildfarm.postgresql.org/cgi-bin/show_log.pl?nm=partridge&dt=2026-08-06%2018%3A00%3A11\n\nBackpatch to all supported versions.\n\nReported-by: Alexander Lakhin \nAuthor: Hayato Kuroda \nReviewed-by: Alexander Lakhin \nReviewed-by: Fujii Masao \nDiscussion: https://postgr.es/m/0b07ead5-a5da-445e-9698-a7d340708bdf@gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "29c285923a1a9c7124fb78e371ef3bc82eca5989", "date": "2026-08-27T15:59:40+01:00", "subject": "CI: Add AArch64 GCS validation build job", "body": "Add an AArch64 CI job on ubuntu-26.04-arm runner which builds with\n-mbranch-protection=standard and links with -z gcs=always and\n-z gcs-report=error. This verifies the presence of AArch64 GCS\nannotations for all object files. The job will fail at link time if\nGCS annotations are either missing or incompatible. The CI job does\nnot run `make tests` because there is no GCS enabled environment in\nthe CI setup in which to run these tests currently.\n\nSigned-off-by: Gowtham Suresh Kumar \nReviewed-by: Mounir Idrassi \nReviewed-by: Tom Cosgrove \nMerge-date: Tue Sep 15 10:23:40 2026\nMerged-from: https://github.com/openssl/openssl/pull/32631", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "angular", "sha": "6f91982c0d8d9736d820dfc25c91485660374e63", "date": "2026-08-27T15:50:46Z", "subject": "build: update dependency vscode-languageserver-textdocument to v1.0.14", "body": "See associated pull request for more information.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "angular", "sha": "bee058eef83bc86d9a84c66e9ccea4abb8ebfe5f", "date": "2026-08-27T19:12:36+03:00", "subject": "fix(docs-infra): point the tutorial .gitignore at live GitHub docs", "body": "The common tutorial scaffold's .gitignore opens with a link to\nhelp.github.com/ignore-files/, which 404s. That directory is copied into every\ntutorial and playground, so the dead link ships to anyone who opens one.\n\n(cherry picked from commit 843a65cba0cc501bcd0ab9cf6383e1b8e84405a9)", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "843a65cba0cc501bcd0ab9cf6383e1b8e84405a9", "date": "2026-08-27T19:12:36+03:00", "subject": "fix(docs-infra): point the tutorial .gitignore at live GitHub docs", "body": "The common tutorial scaffold's .gitignore opens with a link to\nhelp.github.com/ignore-files/, which 404s. That directory is copied into every\ntutorial and playground, so the dead link ships to anyone who opens one.", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "postgres", "sha": "aec25d6107ddf03430947f67d662bdedffc064c2", "date": "2026-08-28T01:17:19+09:00", "subject": "Fix temporary WAL receiver slot handling on timeline switches", "body": "Previously, when wal_receiver_create_temp_slot was enabled, a timeline\nswitch could cause the walreceiver to try to create the same temporary\nreplication slot again on the same connection. The slot had already been\ncreated before the first streaming attempt and still existed, so the\nsecond creation attempt failed with a FATAL error such as\n\"could not create replication slot ...\".\n\nThe walreceiver would later be restarted and streaming replication could\ncontinue, so this did not permanently break replication. Nevertheless,\nthe unexpected failure is a bug and should be fixed.\n\nFix this by tracking whether the temporary replication slot has already\nbeen created for the lifetime of the walreceiver and skipping subsequent\ncreation attempts. Also copy the retained slot name to shared memory on\neach streaming attempt, since RequestXLogStreaming() clears it when\nstreaming is restarted without a configured primary slot. This also\nkeeps pg_stat_wal_receiver.slot_name populated after timeline switches.\n\nBackpatch to all supported versions.\n\nAuthor: ChangAo Chen \nReviewed-by: Quan Zongliang \nReviewed-by: Fujii Masao \nDiscussion: https://postgr.es/m/tencent_628FDAF814231923BC8E8357BBBC50F94207@qq.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "074410485014a0f8f5bc8cac61c1b419ca29caa2", "date": "2026-08-27T19:21:50+03:00", "subject": "docs: update dead links in the contributing docs", "body": "The fork guide link 404s since GitHub retired help.github.com, and the\nrules_nodejs debugging docs moved to the bazel-contrib org. Also points the\ngit setup link straight at its current home rather than through a redirect.\n\n(cherry picked from commit 519c52eea06a0c69a9a625687eab1d7835951e09)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "519c52eea06a0c69a9a625687eab1d7835951e09", "date": "2026-08-27T19:21:50+03:00", "subject": "docs: update dead links in the contributing docs", "body": "The fork guide link 404s since GitHub retired help.github.com, and the\nrules_nodejs debugging docs moved to the bazel-contrib org. Also points the\ngit setup link straight at its current home rather than through a redirect.", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "271222650b44fcb931866d1e9c2ff61cc9b6c677", "date": "2026-08-27T18:46:18+02:00", "subject": "dtls: fix flaky cookie exchange loop in listener test", "body": "drive_until_connection_queued() spun through its iterations with a\nzero-timeout poll and no waiting, so over real loopback sockets the loop\ncould run dry in microseconds if a datagram was not yet readable. On\nmacOS loopback delivery is asynchronous, which made the blocking mode\ntests fail intermittently on slow runners with \"cookie exchange loop did\nnot converge\". Sleep between iterations so a delayed datagram can arrive,\nand raise the iteration cap so the budget also covers the client's 1s\ninitial retransmit timer if a datagram is lost outright.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Richard Levitte \nReviewed-by: Saša Nedvědický \nMerge-date: Tue Sep 1 14:42:31 2026\nMerged-from: https://github.com/openssl/openssl/pull/32551", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "400c810ddbd7890f751913977e4442d71a0ebfc4", "date": "2026-08-27T14:14:15-04:00", "subject": "Fix empty FOREIGN_JOIN sublist validation", "body": "FOREIGN_JOIN target sublists must contain at least two relation\nidentifiers. However, the parser checked only for sublists with\nexactly one identifier, so FOREIGN_JOIN(()) was accepted.\n\nReject sublists with fewer than two relation identifiers, and add\nregression coverage.\n\nAuthor: Chao Li \nDiscussion: http://postgr.es/m/BEDC04E0-6732-4310-95BA-6EC34BC1442C@gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "ad22e89d4d849bb32748e69c876ef5c8ed97451c", "date": "2026-08-27T21:42:40+03:00", "subject": "refactor(devtools): point the .gitignore comment at live GitHub docs", "body": "help.github.com/ignore-files/ 404s since GitHub retired that domain.", "style": "conventional", "type": "refactor", "scope": "devtools", "breaking": false} {"repo": "openssl", "sha": "92c8bb86aa7b3ba98f634938f4cf04a364f30698", "date": "2026-08-27T20:48:08+02:00", "subject": "quic: collect rejected incoming streams and return their stream credit", "body": "A stream rejected by the incoming stream policy is never placed on the accept\nqueue, so it is never retired for the purposes of MAX_STREAMS RXFC and the peer\nis never granted credit for another stream. After the initial limit of 100\nstreams has been reached, the peer cannot open any further stream for the\nlifetime of the connection.\n\nSuch a stream is also never garbage collected. Rejecting a stream resets its\nsend part, which frees the send stream, and the acknowledgement handler skips a\nchunk whose send stream is already gone before it confirms the STOP_SENDING and\nRESET_STREAM frames. The stream therefore never records that its STOP_SENDING\nwas acknowledged, never becomes ready for garbage collection and remains in the\nstream map until the connection is torn down.\n\nConfirm both frames even when the send stream has already been freed, and retire\nthe stream for MAX_STREAMS RXFC when it is rejected. The two changes belong\ntogether: returning the stream credit on its own would replace a leak bounded by\nthe stream limit with an unbounded one.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Saša Nedvědický \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:30:03 2026\nMerged-from: https://github.com/openssl/openssl/pull/32557\n(cherry picked from commit 4f5b40c8fb1b12435c8a564dcbcd2abd393f425d)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "4f5b40c8fb1b12435c8a564dcbcd2abd393f425d", "date": "2026-08-27T20:48:08+02:00", "subject": "quic: collect rejected incoming streams and return their stream credit", "body": "A stream rejected by the incoming stream policy is never placed on the accept\nqueue, so it is never retired for the purposes of MAX_STREAMS RXFC and the peer\nis never granted credit for another stream. After the initial limit of 100\nstreams has been reached, the peer cannot open any further stream for the\nlifetime of the connection.\n\nSuch a stream is also never garbage collected. Rejecting a stream resets its\nsend part, which frees the send stream, and the acknowledgement handler skips a\nchunk whose send stream is already gone before it confirms the STOP_SENDING and\nRESET_STREAM frames. The stream therefore never records that its STOP_SENDING\nwas acknowledged, never becomes ready for garbage collection and remains in the\nstream map until the connection is torn down.\n\nConfirm both frames even when the send stream has already been freed, and retire\nthe stream for MAX_STREAMS RXFC when it is rejected. The two changes belong\ntogether: returning the stream credit on its own would replace a leak bounded by\nthe stream limit with an unbounded one.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Saša Nedvědický \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 14:30:03 2026\nMerged-from: https://github.com/openssl/openssl/pull/32557", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "e721ce48dc78b58827c35d6a13f9f72a591e68f2", "date": "2026-08-27T12:10:49-07:00", "subject": "pg_upgrade: Read nextMultiOffset as a 64-bit value.", "body": "Commit bd8d9c9bdfa widened MultiXactOffset to 64 bits and widened\nControlData.chkpnt_nxtmxoff accordingly, but get_control_data() still\nread the \"Latest checkpoint's NextMultiOffset\" line with str2uint(),\nwhich returns unsigned int.\n\nThis commit adds str2uint64(), mirroring the str2uint() helper used\nfor the other control file fields, and reads the offset with it.\n\nBackpatch to v19, where MultiXactOffset was widened.\n\nReviewed-by: Heikki Linnakangas \nReviewed-by: Chao Li \nDiscussion: https://postgr.es/m/CAD21AoCvzerscfU8o4ARQ793yAGHpQ72r2x5apeC_W2-k=SLCQ@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "3fcb7167198e1b2ecf8cc0d934c8ffe53755e0ae", "date": "2026-08-27T12:25:04-07:00", "subject": "Report next_multi_offset as bigint in pg_control_checkpoint().", "body": "Commit bd8d9c9bdfa widened MultiXactOffset to 64 bits, but\npg_control_checkpoint() still handle checkPointCopy.nextMultiOffset as\nxid type and declared next_multi_offset column as xid. Since xid is 32\nbits wide, an offset above 2^32 was reported truncated, while\npg_controldata printed the full value of the same field.\n\nThis commit reports the column as bigint instead. That matches\npg_get_multixact_stats(), which already reports num_members and\nmembers_size, both derived from these same offsets, as int8.\n\nBackpatch to v19, where MultiXactOffset was widened.\n\nBump catalog version.\n\nReviewed-by: Heikki Linnakangas \nReviewed-by: Chao Li \nDiscussion: https://postgr.es/m/CAD21AoCvzerscfU8o4ARQ793yAGHpQ72r2x5apeC_W2-k=SLCQ@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "6f1dc268ccfc6f01822b528886fe329ce3e65592", "date": "2026-08-27T12:45:27-07:00", "subject": "docs(router): Add router resource docs", "body": "Adds doc (but no navigation entry) for router resources. Feature not released yet\nbut getting documentation in place.", "style": "conventional", "type": "docs", "scope": "router", "breaking": false} {"repo": "neovim", "sha": "617f9e628c3d4b3945ae7d8422f0c91e167efbfe", "date": "2026-08-28T05:48:34+10:00", "subject": "fix(session): window sizes lost in tabpages after an excluded window #41478", "body": "Problem: `restore_size` is scoped to all of `makeopens()`, so a window\nexcluded by 'sessionoptions' latches it off for every tabpage written\nafterwards: with `sessionoptions-=terminal`, a terminal in tab 2 makes\ntab 3 restore with `wincmd =` instead of its stored sizes.\n\nSolution: Recompute it per tabpage.", "style": "conventional", "type": "fix", "scope": "session", "breaking": false} {"repo": "neovim", "sha": "7e2e3f8c255f7ae65bd9fdc3feba6b4efd21416f", "date": "2026-08-27T16:12:11-04:00", "subject": "feat(editor): undo restores cursor position #41520", "body": "Problem:\nUndo places the cursor wherever the cursor happened to sit at \"save\ntime\" (`uh_cursor` is sampled lazily on the first change).\nExamples:\n- `i` preserves, but `a` does not\n- `diw`, `atest`, `d^` abandon the original position\n- `D`, `o` restore it (by accident).\n\nSolution:\n`composite` tracks the pending atom (and its `origin`) across frames.\nA `stuffed` continuation frame inherits the `origin` + prepped redo.\nStore `origin` info in the undo header, so undo can restore it.\n\n- Not for a mid-command undo break (i_CTRL-G_u).\n- Undoing a mapping restores where the mapping started (which\n technically may be different than where the \"edit\" started).", "style": "conventional", "type": "feat", "scope": "editor", "breaking": false} {"repo": "curl", "sha": "e39ff5113b7c0d24e914d47bc719ae1eda601877", "date": "2026-08-28T04:13:41+08:00", "subject": "http: reject resumed uploads", "body": "HTTP has no interoperable mechanism for resuming POST or PUT uploads.\nReject non-zero resume offsets for all HTTP upload request types instead\nof sending partial or complete data with non-standard semantics.\n\nKeep zero offsets and explicit --range uploads working. Update the tool\nerrors, documentation, and tests.\n\nFixes #22703\nCloses #22712", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "dabfa45243288f9078af13c85541f9c1a7967eaf", "date": "2026-08-27T15:39:21-05:00", "subject": "docs: update link/file for JetBrains `AGENTS.md`", "body": "(cherry picked from commit 0017b87b5215627b0d9d7f1c466ec561a1b6ad40)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "0017b87b5215627b0d9d7f1c466ec561a1b6ad40", "date": "2026-08-27T15:39:21-05:00", "subject": "docs: update link/file for JetBrains `AGENTS.md`", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "curl", "sha": "740ffee03b600709932a18154ef3ade001069f2b", "date": "2026-08-27T23:14:13+02:00", "subject": "gopher: fix partial sends of CRLF", "body": "Extract the buffer-sending logic into its own function so it can also be\nused when sending CRLF. Previously, CRLF could be partially sent to the\nserver without reporting an error or sending the remaining bytes.\n\nThis commit preserves how the old code handled errors while writing CRLF\nto the client, which is why the extracted function has a bool parameter.\n\nThis also includes a test case that verifies that the header written to\nthe client contains exactly one CRLF. This test case also detects the\nold buggy behaviour of partial sends.\n\nCloses #22714", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "282057f32a3d7d3a9563c83d6a32196223411c8a", "date": "2026-08-27T23:20:42+02:00", "subject": "mprintf: avoid never-ending loop for positive-infinite", "body": "Floating point is odd but it features this concept of \"infinite\" (which\ncan happen with the new typecast from long double to double) and\napparently dividing infinite with 10 is still infinte.\n\nFollow-up to bdda3a268f0fd1a1bb4\n\nPointed out by: Codex Security\nCloses #22713", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "929a94a5fe9a32f7d3251719ae36bf831878f247", "date": "2026-08-28T07:47:50+10:00", "subject": "cts: remove unrequited .inc files", "body": "There is nothing special about these after the CTS refactoring, so they can\nbe merged inline.\n\nReviewed-by: Simo Sorce \nReviewed-by: Shane Lontis \nMerge-date: Thu Sep 3 02:17:07 2026\nMerged-from: https://github.com/openssl/openssl/pull/32536", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "fd2452b8240add00f95bf1cb83e7f6d503758f19", "date": "2026-08-28T08:32:15+10:00", "subject": "aes-ctr-hmac: add range checking on AAD", "body": "Also correct parameter type from unsigned integer to octet string.\nThis triggers fuzzing failures.\n\nAssisted-by: ChatGPT:gpt-5.6Sol\n\nReviewed-by: Richard Levitte \nReviewed-by: Nikola Pajkovsky \n(Merged from https://github.com/openssl/openssl/pull/32559)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "1013e4368aa6a51a16708df5775499604d3d644f", "date": "2026-08-28T08:32:15+10:00", "subject": "aes-ctr-hmac: add range checking on AAD", "body": "Also correct parameter type from unsigned integer to octet string.\nThis triggers fuzzing failures.\n\nAssisted-by: ChatGPT:gpt-5.6Sol\n\nReviewed-by: Richard Levitte \nReviewed-by: Nikola Pajkovsky \n(Merged from https://github.com/openssl/openssl/pull/32561)\n\n(cherry picked from commit 1740422a1ac4aafc92c8a68a41ef53d603915e35)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "1740422a1ac4aafc92c8a68a41ef53d603915e35", "date": "2026-08-28T08:32:15+10:00", "subject": "aes-ctr-hmac: add range checking on AAD", "body": "Also correct parameter type from unsigned integer to octet string.\nThis triggers fuzzing failures.\n\nAssisted-by: ChatGPT:gpt-5.6Sol\n\nReviewed-by: Richard Levitte \nReviewed-by: Nikola Pajkovsky \n(Merged from https://github.com/openssl/openssl/pull/32561)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "74546d50dcef8739f64f989d18478822491e9731", "date": "2026-08-28T08:32:15+10:00", "subject": "aes-ctr-hmac: add range checking on AAD", "body": "Also correct parameter type from unsigned integer to octet string.\nThis triggers fuzzing failures.\n\nAssisted-by: ChatGPT:gpt-5.6Sol\n\nReviewed-by: Richard Levitte \nReviewed-by: Nikola Pajkovsky \n(Merged from https://github.com/openssl/openssl/pull/32560)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "7137a41223079b4b172aeccb5031347fcc947b79", "date": "2026-08-27T22:44:03Z", "subject": "feat(router): stabilize auto cleanup injectors feature", "body": "Removes experimental tags and stabilizes withAutoCleanupInjectors, AutoCleanupInjectorsFeature,\nand RouteReuseStrategy cleanup methods, while re-exporting withExperimentalAutoCleanupInjectors as deprecated.", "style": "conventional", "type": "feat", "scope": "router", "breaking": false} {"repo": "angular", "sha": "f4b2e437d9b82d6e32911c7220604e32d26ee69f", "date": "2026-08-28T01:48:30+03:00", "subject": "build: point the integration .gitignore comments at live GitHub docs", "body": "help.github.com/ignore-files/ 404s since GitHub retired that domain. Five of the\nfiles still linked to it over plain http as well.\n\n(cherry picked from commit 1e8dd80458d5a1fc47243a96f90891edd85ef047)", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "angular", "sha": "1e8dd80458d5a1fc47243a96f90891edd85ef047", "date": "2026-08-28T01:48:30+03:00", "subject": "build: point the integration .gitignore comments at live GitHub docs", "body": "help.github.com/ignore-files/ 404s since GitHub retired that domain. Five of the\nfiles still linked to it over plain http as well.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "angular", "sha": "9340d69d314d36c4a200935b3a543d4edbf0b59c", "date": "2026-08-28T01:52:40+03:00", "subject": "docs(core): repoint or drop dead jsperf links in the render3 notes", "body": "jsperf.com now returns 410 for every benchmark. Two of them still exist on the\nsuccessor site and are repointed at jsperf.app; the other five are gone, so the\nlinks are removed and the performance claims they backed are kept.\n\n(cherry picked from commit 2a4903fec1cd77ba80090eb793e3b153434256fa)", "style": "conventional", "type": "docs", "scope": "core", "breaking": false} {"repo": "angular", "sha": "2a4903fec1cd77ba80090eb793e3b153434256fa", "date": "2026-08-28T01:52:40+03:00", "subject": "docs(core): repoint or drop dead jsperf links in the render3 notes", "body": "jsperf.com now returns 410 for every benchmark. Two of them still exist on the\nsuccessor site and are repointed at jsperf.app; the other five are gone, so the\nlinks are removed and the performance claims they backed are kept.", "style": "conventional", "type": "docs", "scope": "core", "breaking": false} {"repo": "ant-design", "sha": "e315024afad5ff69075c8ccf1ab34b99a9ef745a", "date": "2026-08-28T08:21:40+08:00", "subject": "fix(Upload): avoid leaving canvas after GIF preview", "body": "", "style": "conventional", "type": "fix", "scope": "Upload", "breaking": false} {"repo": "vue-core", "sha": "70bd789b9636e7fd74a136991a4ee78bef060a3c", "date": "2026-08-28T08:40:18+08:00", "subject": "fix(scheduler): restore scheduler state after flush errors", "body": "", "style": "conventional", "type": "fix", "scope": "scheduler", "breaking": false} {"repo": "vue-core", "sha": "f945b0edb587efc03a7e50c680906562c1da9fb1", "date": "2026-08-28T09:00:42+08:00", "subject": "fix(runtime-vapor): clear v-for index alias for non-object sources", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "postgres", "sha": "978caef1936fc06a6dcc75283115ede953ab6d34", "date": "2026-08-28T13:13:04+12:00", "subject": "Fix incorrect multi-column RANGE partition pruning", "body": "When performing partition pruning with a RANGE partitioned table where\nthe pruning quals are only present for a leading prefix of the partition\nkey, it was possible that partition pruning would accidentally prune away\nsome partitions which shouldn't be pruned and include some partitions that\nwere not needed.\n\nThis happened due to an incorrectly coded loop bound which was\nterminating the loop when the bound reached the first or last element in\nthe partition bound array. This resulted in those end elements not being\nchecked in cases where they should be checked. It appears that it might\nhave been coded this way to avoid stepping off the array, but that was\ndone incorrectly as it failed to take into account the direction of travel\nthrough the array (the loop can go forwards or backwards). I.e., it's\nvalid to loop when 'off' is the last element if we're going backwards\nthrough the array, and valid to loop if 'off' is 0 and we're looping\nforward through the array, but the code as it was didn't allow that.\n\nHere we fix this by moving the loop condition check to after we've\ncalculated the array element to process, and break from the loop if that\nelement is beyond either end of the array.\n\nExample of accidentally pruned partition:\n\np: partition by range (a, b);\np1: for values from (1, 4) to (1, 7);\np2: for values from (1, 7) to (3, 8);\np3: for values from (4, 8) to (6, 9);\ndef: default;\n\nselect * from p where a <= 1;\n\nHere p2 was pruned by mistake.\n\nExample of accidentally not pruning a partition:\n\np: partition by range (a, b);\np1: for values from (7, 2) to (7, 7);\ndef: default;\n\nselect * from p where a > 7;\n\nNo partitions would be pruned in this case, despite it being impossible\nfor matching rows to exist in p1.\n\nAuthor: David Rowley \nReviewed-by: Ayush Tiwari \nReviewed-by: Tender Wang \nDiscussion: https://postgr.es/m/CAApHDvp5ne9AWaH-tG1Lke-USLz3NwWLWTUdP5NT7ypKtcFqcg@mail.gmail.com\nBackpatch-throug", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "985f834cc72707ac68f233a24bd4d6e6438e303b", "date": "2026-08-27T22:38:22-04:00", "subject": "build(vim-patch): v8.2.4758 is n/a", "body": "", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "angular", "sha": "8a2e63c87784a92e701893ccf520ee5078ab3cbf", "date": "2026-08-28T08:43:19+06:00", "subject": "docs: correct highlight index for pipe examples in README", "body": "(cherry picked from commit 54f67e1318348f929326ffd3b40e8bc50cf92c38)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "54f67e1318348f929326ffd3b40e8bc50cf92c38", "date": "2026-08-28T08:43:19+06:00", "subject": "docs: correct highlight index for pipe examples in README", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "a06ff888336cb1f095fa854958f795749723f77f", "date": "2026-08-28T10:43:59+08:00", "subject": "fix(Select): exclude internal popup visibility callback", "body": "", "style": "conventional", "type": "fix", "scope": "Select", "breaking": false} {"repo": "ant-design", "sha": "fa0c3577d2f0d8d058f53d4a0cb36e8d24c71a05", "date": "2026-08-28T10:44:48+08:00", "subject": "fix(Transfer): disable select all for disabled search results", "body": "", "style": "conventional", "type": "fix", "scope": "Transfer", "breaking": false} {"repo": "neovim", "sha": "9e8f032f331ed61ad0213d27edc241aa5898aae3", "date": "2026-08-27T23:18:34-04:00", "subject": "build(vim-patch): v9.2.0420 is n/a", "body": "", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "neovim", "sha": "ee19522048c0b607a2b2b379cdfc020f13bff95b", "date": "2026-08-28T00:14:40-04:00", "subject": "build(vim-patch): v8.2.2872 is n/a", "body": "", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "postgres", "sha": "718f281672a91856f0d241aa5602d4d14f055a93", "date": "2026-08-28T14:51:36+09:00", "subject": "Propagate disabled_nodes to single-child Append paths", "body": "create_append_path() skips cost_append() when an Append has exactly\none child whose parallel awareness matches its own, since setrefs.c\nstrips such an Append out entirely. In that case it copies the\nchild's rowcount and costs directly, but it failed to copy\ndisabled_nodes. An Append over a disabled child therefore claimed to\ncontain no disabled nodes, letting a disabled path win over one that\nis not disabled.\n\nThis is a regression in v18; before e22253467, disable_cost was folded\ninto a path's startup and total costs, so it rode along in the fields\nthis shortcut already copies.\n\nBack-patch to v18. This can change plans in stable branches, but only\nfor installations that have explicitly disabled a node type, and only\nto stop using the node they asked us to avoid.\n\nReported-by: Man Zeng \nAuthor: Tender Wang \nReviewed-by: Richard Guo \nReviewed-by: David Rowley \nDiscussion: https://postgr.es/m/CAHewXNm_Zx5EDoaD7wo7bq6cfRroNznS+RBCzT_p2-CWQXpgSw@mail.gmail.com\nBackpatch-through: 18", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "f2164110f2d0aa233211a1986c4c55bebe0ddf3c", "date": "2026-08-28T15:03:34+09:00", "subject": "fix(common): reject null file input in parse-file pipe", "body": "ParseFilePipe.thereAreNoFilesIn() only checked for `undefined` via\nisUndefined(), so a `null` value was treated as \"a file is present\".\nThis let `null` bypass the fileIsRequired check, and skip validators\nthat already treat a falsy file as trivially valid (e.g.\nMaxFileSizeValidator returns true when `!file`), silently letting a\nrequired file argument resolve to `null` instead of throwing.\n\nSwitch to isNil() so `null` and `undefined` are handled the same way.", "style": "conventional", "type": "fix", "scope": "common", "breaking": false} {"repo": "angular", "sha": "2877c3b995189257ad5c4fcc5fa0ae00929881a9", "date": "2026-08-28T06:25:31Z", "subject": "build: update dependency node to v24.20.0", "body": "See associated pull request for more information.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "vitest", "sha": "91be1d71cde84c6e150fda352eb1d79a099f60e5", "date": "2026-08-28T08:42:06+02:00", "subject": "chore(ai): always close PRs classified as `automation`", "body": "", "style": "conventional", "type": "chore", "scope": "ai", "breaking": false} {"repo": "vitest", "sha": "a6d5ea2d81f459b9dbd42199cd2c35efa415f580", "date": "2026-08-28T08:44:53+02:00", "subject": "perf(vm): let environments opt out of the module graph prewarm", "body": "", "style": "conventional", "type": "perf", "scope": "vm", "breaking": false} {"repo": "angular", "sha": "43aaeaec29e15d7b2504e8930c92c94b898ca200", "date": "2026-08-28T08:57:22+02:00", "subject": "fix(forms): avoid writing to `name` input on ControlValueAccessor", "body": "Fixes that signal forms were overridding the `name` input of the CVA with an auto-generated one. This can break directives like `mat-radio` that assign their own `name`.\n\nFixes #69677.\n\n(cherry picked from commit 38861ac41bb373870e961f41626de6b6b31120f1)", "style": "conventional", "type": "fix", "scope": "forms", "breaking": false} {"repo": "angular", "sha": "38861ac41bb373870e961f41626de6b6b31120f1", "date": "2026-08-28T08:57:22+02:00", "subject": "fix(forms): avoid writing to `name` input on ControlValueAccessor", "body": "Fixes that signal forms were overridding the `name` input of the CVA with an auto-generated one. This can break directives like `mat-radio` that assign their own `name`.\n\nFixes #69677.", "style": "conventional", "type": "fix", "scope": "forms", "breaking": false} {"repo": "vitest", "sha": "3cd019a0a952f43b79d85b2750d189ce0a17066d", "date": "2026-08-28T16:01:32+09:00", "subject": "fix(ui): persist trace step-iframe split pane sizes", "body": "", "style": "conventional", "type": "fix", "scope": "ui", "breaking": false} {"repo": "ant-design", "sha": "e7676edf76173eee3af4cab89f05185eb094fed1", "date": "2026-08-28T15:12:47+08:00", "subject": "fix(Image): use RTL icons in PreviewGroup", "body": "", "style": "conventional", "type": "fix", "scope": "Image", "breaking": false} {"repo": "nestjs", "sha": "7f79233c7ff2cdbb7ae20a0e60fd190ceb14df57", "date": "2026-08-28T16:16:10+09:00", "subject": "fix(core): use `isNil` instead of `isUndefined` in provider classifiers", "body": "", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "tmux", "sha": "013ada58c864c68402aa28c7ad8eef542f346c9d", "date": "2026-08-28T07:36:01Z", "subject": "Ignore focus events earlier to avoid them interfering with prefix, GitHub issue 5539.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "2c613007a056229c39c1b25ee37e6792ab056934", "date": "2026-08-28T17:49:05+10:00", "subject": "remove redundant checks", "body": "Reviewed-by: Richard Levitte \nReviewed-by: Nikola Pajkovsky \n(Merged from https://github.com/openssl/openssl/pull/32559)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "b41b92a42472317335e56ee26ddda1d741e9bfca", "date": "2026-08-28T17:49:05+10:00", "subject": "remove redundant checks", "body": "Reviewed-by: Richard Levitte \nReviewed-by: Nikola Pajkovsky \n(Merged from https://github.com/openssl/openssl/pull/32561)\n\n(cherry picked from commit dd7199f67021e7fb01a70324103aa27fdfa1a1e0)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "dd7199f67021e7fb01a70324103aa27fdfa1a1e0", "date": "2026-08-28T17:49:05+10:00", "subject": "remove redundant checks", "body": "Reviewed-by: Richard Levitte \nReviewed-by: Nikola Pajkovsky \n(Merged from https://github.com/openssl/openssl/pull/32561)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "2b2170254f1ef3adae9f040a0a2fbdacb9331f7a", "date": "2026-08-28T17:49:05+10:00", "subject": "remove redundant checks", "body": "Reviewed-by: Richard Levitte \nReviewed-by: Nikola Pajkovsky \n(Merged from https://github.com/openssl/openssl/pull/32560)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "fad90b666df98c93ee165341b0c8101c7b10ad7b", "date": "2026-08-28T09:55:25+02:00", "subject": "h2 push: use squeaky clean easy handle", "body": "When a HTTP/2 push requires a new easy handle, create a clean easy and\ndo *not* duplicate the existing one. This avoids messy use of the API\n`curl_easy_duphandle()`, which has some caveats the applictions must\nobserve and cannot do so for a h2 push.\n\nUsing a clean easy handle is also better since the pushed stream has no\nreal relation to the easy handle that was in use when the HTTP/2 push\narrived (it could be any easy using the connection).\n\nInheriting settings/inputs from a somewhat \"random\" existing easy seems\nwrong.\n\nCloses #22715", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "52b452948fe8fb11062fa88bb41f3c56d7e0f025", "date": "2026-08-28T08:56:12+01:00", "subject": "Check for buffer write error.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "181e818b9378692d42ab3e2b4153a037942c2195", "date": "2026-08-28T08:02:16Z", "subject": "Do no free input context if it is NULL when popup create fails.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "fa1d46d1cc32dfe1b824d5617c48a2c92950db03", "date": "2026-08-28T10:04:06+02:00", "subject": "apps: test ciphers app output modes and cipher selection", "body": "Add a new test recipe covering the ciphers app output: the non-verbose\ncolon-separated list including rejection of an unknown cipher string,\n-v cipher descriptions, -V cipher code prefixes and -stdname standard\ncipher names, including their combination and a check that AES128-SHA\nmaps to the TLS_RSA_WITH_AES_128_CBC_SHA standard name. Also check\nthat -psk makes PSK ciphers supported and that -ciphersuites configures\nthe TLSv1.3 ciphersuites and rejects an unknown one.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Daniel Kubec \nReviewed-by: Richard Levitte \nMerge-date: Tue Sep 15 15:17:09 2026\nMerged-from: https://github.com/openssl/openssl/pull/32569", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "78dad45f3bd13727653633fa43ec62cc336fe6e6", "date": "2026-08-28T09:06:45+01:00", "subject": "Add timeout for lock and don't keep credentials for regress, from Andreas Fehlner.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "621b63dff5641cd96afa5bec26ca18a389961db3", "date": "2026-08-28T16:10:05+08:00", "subject": "demo(Modal): make draggable demo closable on touch devices", "body": "The draggable Modal demo gated dragging on a hover-driven `disabled`\nstate (`onMouseOver`/`onMouseOut` on the title). This works for mouse\nbut breaks on touch: after dragging, `disabled` stays `false`, so\n`react-draggable`'s `touchstart` listener calls `preventDefault()` and\nswallows the synthetic `click` on the close button — the modal can no\nlonger be closed (#44151). The first drag also often failed because\n`mouseover` lags `touchstart`.\n\nReplace the hover+`disabled` simulation with `react-draggable`'s native\n`handle`, pointing at the title element. Dragging is now driven by a\nCSS selector, so mouse and touch share one code path: the title drags,\nthe body/close button do not, and the close button is never\n`preventDefault`-ed. Cursor `move` is narrowed to the title only,\nmatching the original demo.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "1c98c3a2a34682a6d0fa0077b5bcf563e97683ee", "date": "2026-08-28T08:11:08Z", "subject": "Do not walk off end of grid lines if last line is wrapped, reported by Moe Khalilov.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "1c0936e33687e141278359743f916ad0f70298ba", "date": "2026-08-28T17:15:27+09:00", "subject": "ktls_send_ctrl_message(): Fix parameter name from 'lengthi' to 'length'", "body": "In include/internal/ktls.h, the variable/parameter lengthi is used\nin Kernel TLS (KTLS) helper declarations and macros, whereas all\nrelated functions across the codebase consistently use length.\n\nThis naming inconsistency triggers static analysis warnings\n(including potential CWE-467 flags regarding parameter/pointer\nevaluation in macro expansions) and creates confusion during maintenance.\nThe identifier lengthi appears to be an unintended typo introduced\nduring the original KTLS interface definition.\n\nCLA: trivial\nReviewed-by: Neil Horman \nReviewed-by: Norbert Pocs \nReviewed-by: Paul Dale \nMerge-date: Mon Sep 7 18:42:31 2026\nMerged-from: https://github.com/openssl/openssl/pull/32568\n(cherry picked from commit f289977675358dbd9d048e86f00d89b74c53ad7b)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "f289977675358dbd9d048e86f00d89b74c53ad7b", "date": "2026-08-28T17:15:27+09:00", "subject": "ktls_send_ctrl_message(): Fix parameter name from 'lengthi' to 'length'", "body": "In include/internal/ktls.h, the variable/parameter lengthi is used\nin Kernel TLS (KTLS) helper declarations and macros, whereas all\nrelated functions across the codebase consistently use length.\n\nThis naming inconsistency triggers static analysis warnings\n(including potential CWE-467 flags regarding parameter/pointer\nevaluation in macro expansions) and creates confusion during maintenance.\nThe identifier lengthi appears to be an unintended typo introduced\nduring the original KTLS interface definition.\n\nCLA: trivial\nReviewed-by: Neil Horman \nReviewed-by: Norbert Pocs \nReviewed-by: Paul Dale \nMerge-date: Mon Sep 7 18:42:31 2026\nMerged-from: https://github.com/openssl/openssl/pull/32568", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "0e11020a0a644ce640848b75d47b77e518b0344b", "date": "2026-08-28T10:23:47+02:00", "subject": "apps: test enc app opaque symmetric key options", "body": "Add a new test recipe covering the enc app skey handling: an opaque key\nimported via -skeymgmt/-skeyopt raw bytes produces the same AES\nciphertext as the equivalent raw -K key and decrypts back, the raw key\nand skeyopt options are mutually exclusive, and unknown skeymgmt names\nor malformed skeyopt values are rejected. With module support, the\nfake-cipher provider covers a full encrypt/decrypt roundtrip through a\nprovider-implemented cipher, including defaulting the skeymgmt name to\nthe cipher name when -skeymgmt is not given.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Daniel Kubec \nReviewed-by: Dmitry Belyavskiy \nReviewed-by: Tomas Mraz \nMerge-date: Mon Sep 7 13:28:07 2026\nMerged-from: https://github.com/openssl/openssl/pull/32570", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "3b8723ce3d88b2bf871185736bc33e3cf9a189f4", "date": "2026-08-28T08:36:46Z", "subject": "fix(platform-server): resolve HTTP(S) URLs without authority as relative during SSR", "body": "Under the WHATWG URL standard, HTTP and HTTPS URLs lacking an authority\n(e.g., `http:/path` or `http:path`) resolve as relative paths when resolved\nagainst an origin of the same scheme. Previously, `relativeUrlsTransformerInterceptorFn`\ntreated any URL with a scheme as an absolute URL, bypassing base resolution in SSR\nand allowing Node fetch to parse the path as a cross-origin host.\n\nThis commit updates SSR URL resolution and the HTTP interceptor to ensure\nHTTP(S) URLs without an authority are resolved against the current origin,\npreventing unexpected origin changes and aligning SSR with browser behavior.\n\nFixes #70447\n\n(cherry picked from commit b3bb36ad87334f7a86908c1b2b4353c1cc514c2e)", "style": "conventional", "type": "fix", "scope": "platform-server", "breaking": false} {"repo": "angular", "sha": "b3bb36ad87334f7a86908c1b2b4353c1cc514c2e", "date": "2026-08-28T08:36:46Z", "subject": "fix(platform-server): resolve HTTP(S) URLs without authority as relative during SSR", "body": "Under the WHATWG URL standard, HTTP and HTTPS URLs lacking an authority\n(e.g., `http:/path` or `http:path`) resolve as relative paths when resolved\nagainst an origin of the same scheme. Previously, `relativeUrlsTransformerInterceptorFn`\ntreated any URL with a scheme as an absolute URL, bypassing base resolution in SSR\nand allowing Node fetch to parse the path as a cross-origin host.\n\nThis commit updates SSR URL resolution and the HTTP interceptor to ensure\nHTTP(S) URLs without an authority are resolved against the current origin,\npreventing unexpected origin changes and aligning SSR with browser behavior.\n\nFixes #70447", "style": "conventional", "type": "fix", "scope": "platform-server", "breaking": false} {"repo": "openssl", "sha": "19c76c202c4d3bf76361d272364a5fb12d1766d8", "date": "2026-08-28T10:44:28+02:00", "subject": "apps: test enc -kfile option", "body": "Add a subtest covering -kfile: a passphrase read from a file with a\ntrailing CRLF encrypts data that decrypts with the same passphrase\ngiven via -k, an empty passphrase file is rejected and so is a file\ncontaining only a newline.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Daniel Kubec \nReviewed-by: Paul Dale \nMerge-date: Mon Sep 7 13:29:19 2026\nMerged-from: https://github.com/openssl/openssl/pull/32574", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "d28c9f613d0ae19abbf6a34fa0f105ef9e6d3a64", "date": "2026-08-28T10:25:42+01:00", "subject": "Do not loop forever if a pane ends up with borders the wrong way round.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "0ebe0e978535f10f6fe54536ca2e2e6b304a4d5a", "date": "2026-08-28T11:43:53+02:00", "subject": "tool: checkfds, open on null device", "body": "When stdin/out/err are closed on tool start, checkfds() tries to fill\nthem by invoking `pipe()` which is may allocate file descriptors in an\nunwanted order and may create a pipe between stdout and stdin which is\nnot what we want.\n\nUse '/dev/null' if HAVE_PIPE is defined to open the closed descriptors.\n\nCaveat: I am not 100% sure that HAVE_PIPE implies that /dev/null\nworks, but it seems likely?\n\nCloses #22716", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "5d2b8670c632fab72a60fc85529550e0accc5e71", "date": "2026-08-28T12:00:03+02:00", "subject": "ipv6 scope_id: set from first peer", "body": "When not explicitly set, give the connection a present scope_id from the\nfirst peer of a connection, not from its origin. This accounts for\nproxies.\n\nCloses #22717", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "9f440907e8c1ec695677d334f9141d65e4e24178", "date": "2026-08-28T12:09:32+02:00", "subject": "refactor(core): remove old `deferredImports` structure", "body": "This finishes the migration to the keyed object structure", "style": "conventional", "type": "refactor", "scope": "core", "breaking": false} {"repo": "openssl", "sha": "89688b473e1769d0874a565d467cef4bb2642431", "date": "2026-08-28T13:07:37+02:00", "subject": "fuzz/x509v3.c: drop a dead store of cert", "body": "Setting cert to NULL before assinging another X509_new() result\nto it is unnecessary, as was pointed out by Coverity. Drop\nthe dead store.\n\nResolves: https://scan5.scan.coverity.com/#/project-view/65248/10222?selectedIssue=1700563\nComplements: 9da88ee430a0 \"Add X509V3_EXT_add_nconf fuzzer\"\nSigned-off-by: Norbert Pocs \nReviewed-by: Paul Dale \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Neil Horman \nReviewed-by: Eugene Syromiatnikov \nMerge-date: Mon Aug 31 05:19:12 2026\nMerged-from: https://github.com/openssl/openssl/pull/32576", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "fe2d3591c33d8cc5d23a734f616b681df5e1147a", "date": "2026-08-28T13:20:11+02:00", "subject": "share unlink: forget connection", "body": "When removing a share from an easy handle and the share caches\nconnections, invalidate the easy handle's `lastconnect_id` as\nfuture lookups would go into another connection cache and\npossibly find a wrong connection.\n\nCloses #22718", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "3e07f37212d81bace9a8f0a93b99cd0fca7bf6e6", "date": "2026-08-28T21:31:57+10:00", "subject": "fips: update the sources list removing the AES CTS .inc file", "body": "The file's contents were merged into aes_cts.c and refactored into\ncipher_cts.\n\nReviewed-by: Simo Sorce \nReviewed-by: Shane Lontis \nMerge-date: Thu Sep 3 02:17:10 2026\nMerged-from: https://github.com/openssl/openssl/pull/32536", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "48322e9f09f2b51f77d37c5e04a3f370bb36bfc4", "date": "2026-08-28T13:39:14+02:00", "subject": "mprintf: acknowledge %F", "body": "The code previously just ignored it, causing variadic argument\ndesynchronization\n\nExtend test 557 to verify\n\nReported-by: Stanislav Fort\n\nCloses #22719", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "fb083f3850c242e14613de517811a3818bef98db", "date": "2026-08-28T04:43:46-07:00", "subject": "docs(gen_help_html): generate Typst format #40665", "body": "Problem:\ngen_help_html.lua only outputs HTML, but Typst is useful for producing PDFs.\n\nSolution:\nAdd gen_one_typ() and ts_node_to_typ(), which mirror the existing\ngen_one_html() and ts_node_to_html().", "style": "conventional", "type": "docs", "scope": "gen_help_html", "breaking": false} {"repo": "vitest", "sha": "488c04680e4fbb9e9bff796871ed665eef517e68", "date": "2026-08-28T13:44:32+02:00", "subject": "fix: invalidate resolved setup files (fix #11073)", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "curl", "sha": "8071d7adc2ce8b29a49b8888aed1c346eaea4a3f", "date": "2026-08-28T14:01:25+02:00", "subject": "schannel: clear borrowed sslContext on close", "body": "The current Negotiate and NTLM consumers were traced, and no reachable\npost-free dereference was found.\n\nThe pointer is still left dangling after Schannel teardown, so clear it\nwhen the context is destroyed.\n\nCloses https://github.com/curl/curl/pull/22720", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "a5081e01ec185a962001f265f29992a33a983ca8", "date": "2026-08-28T14:07:42+02:00", "subject": "HISTORY: September 1999: started using CVS", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "19ad8bf452bbbe159caa990d6344f8122eb607eb", "date": "2026-08-28T08:14:48-04:00", "subject": "fix(core): do not tune the shared socket for http2 sse responses", "body": "`SseStream` tunes the request socket for streaming. Under HTTP/2 that\nsocket is a proxy: Node routes `setTimeout` to the shared `Http2Session`\nand the remaining calls to the shared TCP socket.\n\nOne `@Sse()` request therefore set `session.timeout` to 0 for the whole\nconnection. Every other stream on that connection lost its idle timeout,\nincluding requests that never used SSE, and the setting was not restored\nafter the SSE stream closed.\n\nSkip the tuning when the request is HTTP/2. It is not needed there: a\ngraceful `session.close()` waits for active streams, and Node arms no\nper-stream timeout to preserve.\n\nCloses #17605", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "openssl", "sha": "cdf6816a7aaef6eadd70e29af71d60f1eb4ab827", "date": "2026-08-28T14:20:45+02:00", "subject": "apps: test dgst -hex and -binary options", "body": "Add coverage for the -binary digest output path (raw digest bytes\nwritten via -out) and for -hex overriding the default binary output\nwhen signing, checking both the output format and that the hex value\nmatches the binary signature.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Daniel Kubec \nReviewed-by: Richard Levitte \nMerge-date: Tue Sep 15 15:08:53 2026\nMerged-from: https://github.com/openssl/openssl/pull/32577", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "c10a129007cc1f57b04a1b62ca6fa7a3cb3a1d5a", "date": "2026-08-28T14:33:58+02:00", "subject": "fix: accept awaited value in vi.when thenResolve", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "curl", "sha": "bd81f15ec295a2d2719babc7677b343938c7120a", "date": "2026-08-28T14:38:47+02:00", "subject": "schannel: clear PFX password before free", "body": "Closes #22721", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "f5d4b5975d9f54426c258ddf6c3f802bc17bee1b", "date": "2026-08-28T13:39:16+01:00", "subject": "fix(tabpage): focusing a tab while closing it, fails assert #41475", "body": "Problem:\nIf an autocommand handler focuses the tab page we're closing during a\n`:tabonly` (with some conditions), we hit an assert failure in\n`win_close_othertab()`.\n\nFor this to occur, we need:\n- `nvim_buf_delete()` to trigger `close_windows()` (as is done in\n\t`cmdwin.lua`'s `_cleanup()`)\n- `close_windows()` then calls `win_close_othertab()`, removing the\n\twindow (`win_free_mem()`)\n- then in the caller, `tabpage_close_other()`, the loop continues (we\n\tdon't detect `tp_lastwin == wp` since we've unlinked the window)\n- the loop assumes that `curtab != tp`\n\nbut we've refocused `curtab` so the `ex_win_close()` call passes `tp` as\n`curtab`, causing the assert to fail\n\nSolution:\nDetect the focus of `curtab` and abort closing the tab.", "style": "conventional", "type": "fix", "scope": "tabpage", "breaking": false} {"repo": "angular", "sha": "05c4d5a8354228100b51176f295ed5dee4f3febc", "date": "2026-08-28T14:50:31+02:00", "subject": "feat(core): add utility for testing directives", "body": "Adds `TestBed.createDirective` to make testing directives easier.\n\nFixes #54164.", "style": "conventional", "type": "feat", "scope": "core", "breaking": false} {"repo": "curl", "sha": "2a4212ce518d1b1daee59155dbf10f7dc727165b", "date": "2026-08-28T14:55:23+02:00", "subject": "autotools: alpha-sort `ETAGS_ARGS` list", "body": "Aka `CURL_ETAGS_FILES`.\n\nFollow-up to b73a4f00b2a6748d6f36d737cc3869c9a2e96e1e #22680\n\nCloses #22773", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "7144265aa0ba8a1e5c8962a44033a4e85fe18198", "date": "2026-08-28T15:12:42+02:00", "subject": "apps: cover the kdf -binary option in the test recipe", "body": "Add a test deriving a key with -binary -out and comparing the raw\nbytes against the expected value, also covering -out to a file.\nUse diag() instead of print for TAP failure diagnostics.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Daniel Kubec \nReviewed-by: Paul Dale \nMerge-date: Mon Sep 7 14:11:18 2026\nMerged-from: https://github.com/openssl/openssl/pull/32578", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "05852c3cf102192e47e26c2a37d03abd021d5a91", "date": "2026-08-28T09:36:13-04:00", "subject": "Make platform guards in two regression tests match meson builds", "body": "collate.linux.utf8 skips itself unless version() matches \"linux-gnu\",\nand infinite_recurse skips itself when version() matches\n\"powerpc64[^,]*-linux-gnu\". configure substitutes the GNU host triplet\ninto that string, but the meson build composes it from\nhost_machine.cpu_family() and host_system, which never carries the ABI\nsuffix. So ever since meson support arrived in 16, collate.linux.utf8\nhas not run at all on a meson build, and infinite_recurse has been\nrunning on ppc64 Linux the very case it means to stay away from.\nMeson documentation says it reports 'ppc64' instead of 'powerpc64'.\n\nFix by matching \"-linux[-,]\" and \"p(ower)?pc64[^,]*-linux\", which match\nall spellings. Keeping the punctuation on either side confines the\nmatch to the platform field. Neither pattern excludes musl, but\ncollate.linux.utf8's other conditions already require a set of glibc\nlocales to be present.\n\nBackpatch to 16, where the meson build was introduced.\n\nDiscussion: https://postgr.es/m/a40b19da-9a02-47b4-8afd-2bbbde8db1e8@dunslane.net", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "da8da39102788aabe2d479354cb146651825b094", "date": "2026-08-28T10:04:58-04:00", "subject": "doc PG 19 relnotes: move replication items to logical replicat.", "body": "Reported-by: Masahiko Sawada\n\nAuthor: Masahiko Sawada\n\nDiscussion: https://postgr.es/m/CAD21AoC1dJGqngg_cdT_ayQjOdw6gmSBfVOTtAWOq-5C+XeLZQ@mail.gmail.com\n\nBackpatch-through: 19 only", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "52ebad23b0c8cef49ad3a3fa6be7fe391601594e", "date": "2026-08-28T17:06:03+03:00", "subject": "docs(zone.js): point the Bluebird link at the project repo", "body": "bluebirdjs.com no longer resolves. npm lists the GitHub repo as the package's\nhomepage, so the link now goes there.\n\n(cherry picked from commit 731c838dbb4160e82a1003996bca4dc1695fc237)", "style": "conventional", "type": "docs", "scope": "zone.js", "breaking": false} {"repo": "angular", "sha": "731c838dbb4160e82a1003996bca4dc1695fc237", "date": "2026-08-28T17:06:03+03:00", "subject": "docs(zone.js): point the Bluebird link at the project repo", "body": "bluebirdjs.com no longer resolves. npm lists the GitHub repo as the package's\nhomepage, so the link now goes there.", "style": "conventional", "type": "docs", "scope": "zone.js", "breaking": false} {"repo": "curl", "sha": "c2a04c080d79e1eb5d99bc0a73fd71710aa6d345", "date": "2026-08-28T16:19:15+02:00", "subject": "distfiles.sh: check tarball for unexpected files", "body": "Verify if the source tarball has any files that we did not expect.\n\nFollow-up to f8992b6a177a2ccfb5dc04137088534bad68681c #22686\n\nCloses #22724", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "aaab8cdac4d347fea513f5174ebf7b03bb5e61c2", "date": "2026-08-28T10:24:40-04:00", "subject": "fix(gen_help_html): cleanup #41525", "body": "", "style": "conventional", "type": "fix", "scope": "gen_help_html", "breaking": false} {"repo": "openssl", "sha": "518e7ede45cda59427f41af8d521380f4bc12ca2", "date": "2026-08-28T16:39:56+02:00", "subject": "apps: cover the prime -generate option in the test recipe", "body": "Add tests generating a prime in decimal and hex output, verifying the\nresult is reported prime by feeding it back for primality checking,\nchecking the -safe option produces p with (p-1)/2 also prime, and\ncovering the error cases of a missing -bits and a trailing number\nargument.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Daniel Kubec \nReviewed-by: Paul Dale \nMerge-date: Mon Sep 7 14:12:35 2026\nMerged-from: https://github.com/openssl/openssl/pull/32580", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "92d44b200e12c7ecbaee57d37fcd5e8803811358", "date": "2026-08-28T17:48:46+03:00", "subject": "Fix backend state after a failed after-startup shmem request", "body": "RegisterShmemCallbacks() left the backend in a bad state, if an error\noccurred in the callbacks or if an allocation failed. Firstly,\n'shmem_request_state' was left in wrong state, causing a subsequent\ncall to RegisterShmemCallbacks() to wrongly take the postmaster\nstartup codepath or assertion failures in some other functions.\nSecondly, the 'pending_shmem_requests' list was not properly cleaned\nup, causing a subsequent RegisterShmemCallbacks() to try to process\nthe stale, already-freed requests.\n\nTo fix, add a PG_TRY() block to clean those things up on error.\n\nAuthor: Ayush Tiwari \nReviewed-by: Ashutosh Bapat \nDiscussion: https://www.postgresql.org/message-id/CAJTYsWVRRWH48=PcuAo_2Y4Ap6M0QRmzxgUfFkNRtdWK74LjBQ@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "9d888cd3f34c4e9a8f7fbe7e78c70bbd84a481de", "date": "2026-08-28T17:49:38+03:00", "subject": "Track which shmem areas have been fully initialized", "body": "If SHMEM_CALLBACKS_ALLOW_AFTER_STARTUP is used to allocate shared\nmemory after startup, but the initialization fails half-way through,\nthe shmem area is left in an indeterminate state. Furthermore, if\nmultiple shmem areas are registered in one RegisterShmemCallbacks()\ncall, some might be allocated while others are not.\n\nThis commit adds an explicit 'initialized' flag to each shmem area.\nWe still leave behind an uninitialized area on error, but at least\nthey are now clearly marked, and you get a slightly nicer error\nmessage if you try to re-register them. It'd be nice to clean up more\nthoroughly and support actually retrying the allocations, but in\npractice, the most likely reason for a shmem allocation or\ninitialization to fail is that you are out of shared memory and\nretrying wouldn't help with that.\n\nThis isn't exactly a new problem, the old ShmemInitStruct() interface\nhad similar issues if the initialization code failed, or if you\nallocated multiple structs and some allocations failed. It was just\nleft to the calling code to deal with it.\n\nAuthor: Ayush Tiwari \nReviewed-by: Ashutosh Bapat \nDiscussion: https://www.postgresql.org/message-id/CAJTYsWVRRWH48=PcuAo_2Y4Ap6M0QRmzxgUfFkNRtdWK74LjBQ@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "130f79b8b7e9aa3c9af9ff7b81fc1c76cec01729", "date": "2026-08-28T16:54:03+02:00", "subject": "fix: ignore computed member calls in static collection", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "angular", "sha": "d470afa8b2f10d4bca728af77c35b4e2c887e03f", "date": "2026-08-28T16:59:17+02:00", "subject": "refactor(platform-server): ensure error boundaries work with ssr/hydration", "body": "Added some tests to ensure error boundaries work with SSR and hydration.\n\nPR Close #70463", "style": "conventional", "type": "refactor", "scope": "platform-server", "breaking": false} {"repo": "curl", "sha": "273534cdc71dcc3b844ef46da4d48dd6d9a627bf", "date": "2026-08-28T17:04:35+02:00", "subject": "sspi: enable channel-binding in mingw-w64 <9 builds", "body": "Drop `SECPKG_ATTR_ENDPOINT_BINDINGS` guards and declare the missing\nmacro and struct for mingw-w64 <9. To bring these builds on par with the\nrest of supported Windows toolchains.\n\nFollow-up to 09662337441c40c23da7b557c4cceacd7cc3b76e #3321\n\nCloses #22725", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "449ef4b7a57bfb22b38527e57b714de71cfde7dd", "date": "2026-08-29T00:09:39+09:00", "subject": "QUIC: clarify SSL_listen_ex() ex data retention", "body": "Refer directly to application ex data stored on new_conn, avoiding\nterminology tied to the internal QUIC object layout.\n\nRestore the original quic_local.h copyright line since the file is\nnot modified.\n\nReviewed-by: Neil Horman \nReviewed-by: Saša Nedvědický \nMerge-date: Tue Sep 1 14:19:19 2026\nMerged-from: https://github.com/openssl/openssl/pull/32491", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "1c4271b6e5b81f52f9d1af02cbe7f47826cc6fb5", "date": "2026-08-28T09:17:26-06:00", "subject": "Print the correct length for the private key", "body": "When the private key length is not what we expect don't\nprint the public key length.\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Norbert Pocs \nMerge-date: Tue Sep 1 14:32:19 2026\nMerged-from: https://github.com/openssl/openssl/pull/32582\n(cherry picked from commit b0a7d15aff32d796b0f98be170a3c4a92973a5da)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "b0a7d15aff32d796b0f98be170a3c4a92973a5da", "date": "2026-08-28T09:17:26-06:00", "subject": "Print the correct length for the private key", "body": "When the private key length is not what we expect don't\nprint the public key length.\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Norbert Pocs \nMerge-date: Tue Sep 1 14:32:19 2026\nMerged-from: https://github.com/openssl/openssl/pull/32582", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "e28d725e469d991fc80d324cdd39a2b4b6d7e5f7", "date": "2026-08-28T11:21:42-04:00", "subject": "refactor(scripts): drop download-unicode-files.sh #41528", "body": "Unused since 26be6446e5ea, which removed src/unicode/ in favor of\nutf8proc.", "style": "conventional", "type": "refactor", "scope": "scripts", "breaking": false} {"repo": "curl", "sha": "b6bf15bf5cadccb2f2e71b68d17792069d707df2", "date": "2026-08-28T08:32:52-07:00", "subject": "tests: enable and fix some new Python ruff warnings", "body": "Closes #22726", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "b6d5507e20dafb7df76e119d5087ba649a18f350", "date": "2026-08-28T17:38:13+02:00", "subject": "CVE-2026-82209: domain-scoped PSL domain cookie", "body": "cookie: cookies set for an exact PSL domain is host-only\n\nAdded test 2318 to verify.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "95c1e8915dce64606bd753fd47fc0bd236e31cd6", "date": "2026-08-28T17:38:13+02:00", "subject": "cookie: cookies set for an exact PSL domain is host-only", "body": "Added test 2318 to verify.\n\nReported-by: Stanislav Fort (Aisle Research)\n\nCloses #22730", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "00c7baea512685987f28f2cbc10bbcd3e7904d70", "date": "2026-08-28T09:05:02-07:00", "subject": "CI: add a test of labeler patterns", "body": "When source files are renamed or removed, labeler patterns can refer to\nnonexistent files. Verify that the globs in the labeler config are still\nvalid so they can be updated when corresponding source changes occur.\nThis check does nothing when new files are added; those must be manually\nchecked to see if they warrant new matching patterns in the config.\nRemove a stale config entry found in the process of integrating this\ncheck.\n\nCloses #22727", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "vitest", "sha": "b628cb68050729dd82fbe5998198f4347040ebf0", "date": "2026-08-28T18:23:39+02:00", "subject": "feat: promote parseSpecifications out of experimental", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "neovim", "sha": "2514256d95d7161ec55e8e54c01f667a1551809c", "date": "2026-08-28T18:46:03+02:00", "subject": "refactor(input): exec stuffed keys eagerly", "body": "Problem:\nThe Vim \"stuff\" concept breaks the ability to reason about the call\nstack and thus the boundaries of a `CmdAtom`: a stuffed translation (\"x\"\n=> \"dl\") defers to the main loop. This \"continuation\" must be modeled in\n`CmdAtom`, by checking global flags at undefined times, during undefined\ncircumstances.\n\nSolution:\n- After a stuffed \"translation\", eagerly execute the stuff buffer\n (`exec_stuffed()`).\n- Delete the CmdAtom \"continuation\" junk.\n\nNote:\n- op_colon runs its cmdline \"nested\", but that's fine because operators\n already nest interactive sessions there (op_change runs edit()), and\n the cmdline is frameless so the operator's frame can own/capture it.", "style": "conventional", "type": "refactor", "scope": "input", "breaking": false} {"repo": "vitest", "sha": "7db80dc27e5948010c00160ed0b86570baad6ce8", "date": "2026-08-28T18:46:45+02:00", "subject": "chore: release v5.0.0-rc.3", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "curl", "sha": "61aa5b0263c7f4ff4396c376537a0bf297bdf91a", "date": "2026-08-28T09:53:03-07:00", "subject": "CI: add a test of Python code using pyrefly", "body": "pytype is no longer supported and will stop working one day when a newer\nPython becomes the default. pyrefly is a bit too pedantic in its default\nconfiguration, so run it with only the basic set of checks.\n\nCloses #22728", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "vitest", "sha": "e08c45c3921ab1c517021836bc73a13f1310c538", "date": "2026-08-28T10:19:14-07:00", "subject": "perf(expect): lazily format spy matcher failures", "body": "", "style": "conventional", "type": "perf", "scope": "expect", "breaking": false} {"repo": "neovim", "sha": "61958f233584b0c75f6dbc650d6b1d3c01096404", "date": "2026-08-28T18:34:31+01:00", "subject": "fix(cmdwin): allow a user to switch to other buffers #41199", "body": "", "style": "conventional", "type": "fix", "scope": "cmdwin", "breaking": false} {"repo": "postgres", "sha": "2ebf25e7d70a8fce31ace78d723fa9271ab8af72", "date": "2026-08-28T15:12:26-04:00", "subject": "Perform join removal by editing the query's jointree.", "body": "analyzejoins.c decided which joins could be dropped by consulting the\nplanner's derived data structures, but then implemented the removal\nby updating those structures in-place. That is a lot of fiddly work,\nand nothing keeps it in step with the rest of the planner:\nremove_leftjoinrel_from_query only bothered to update \"parts of the\nplanner's data structures that will actually be consulted later\", with\nno good way to know what those are. Bug #19560 is one consequence.\nIn that report, removing a join leaves an EquivalenceClass that now\ngives rise to a base restriction clause, but base restriction clauses\nhave already been generated and nothing reconsiders them, so the WHERE\ncondition disappears from the plan and we return wrong answers.\n\nThe self-join elimination code has the same design and the same type\nof hazard. We have seen many related bugs over the years too, so it's\ntime to do something drastic.\n\nTo fix, do the removals by editing root->parse->jointree (which is a\nfar simpler and more stable representation than the derived data),\nand then have query_planner() discard everything it computed from the\njointree and derive it over again. This requires quite a bit less\ncode, and doesn't require touching analyzejoins.c every time we change\nthe data derived by query_planner(). For typical cases it can actually\nsave a bit of planning time, though in cases where we have to iterate\nthe derivation loop many times it does add some time.\n\nreduce_unique_semijoins() gets the same treatment: rather than deleting\nthe semijoin's SpecialJoinInfo and relying on the jointree not being\nconsulted again, it now changes the JoinExpr's jointype to JOIN_INNER\nand recalculates everything.\n\nSome plans change in the join regression test. Qual evaluation order\nshifts in a few cases, because the conditions now reach later planning\nin jointree order rather than in whatever order the removal code\nre-distributed them. A few plans improve, since the rebuilt relation\ntargetlists no longer c", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "axios", "sha": "d29be181f85f6fe93397a07b1f69606d9622637b", "date": "2026-08-28T21:18:17+02:00", "subject": "fix: harden request option resolution and no_proxy matching", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "postgres", "sha": "7ddb9c41a13534011bb64a856fb5a60e41b82b75", "date": "2026-08-28T15:11:06-05:00", "subject": "Fix pg_stat_autovacuum_scores for TOAST tables.", "body": "In v19, pg_stat_autovacuum_scores computes a TOAST table's scores\nfrom its own storage parameters alone. On the other hand,\nautovacuum falls back to the main table's parameters when the TOAST\ntable has none. This means that the view may report scores that\ndon't match what autovacuum would calculate. This contradicts the\ndocumented promise that the view generates its results the same way\nautovacuum workers do. To fix, teach the view to do the same\nfallback. As in do_autovacuum(), we must make a preliminary pass\nover pg_class to collect the main tables' parameters, since the\npg_class scan may see TOAST tables before their main tables.\n\nCommit fad70a09ff for v20 improved autovacuum's handling of TOAST\nstorage parameters and adjusted the view to match, but it was\ndeemed too intrusive to back-patch. This fix is for v19 only.\n\nOversight in commit 87f61f0c82.\n\nReported-by: Masahiko Sawada \nReviewed-by: Masahiko Sawada \nDiscussion: https://postgr.es/m/CAD21AoB1CJRVfCDh8qYuD3eueiygXxk7F3nybgjN0RZXSD-QUw%40mail.gmail.com\nBackpatch-through: 19 only", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "434c12601c59f3c1e484682eb0ef66538859679c", "date": "2026-08-28T16:37:11-04:00", "subject": "Update .abi-compliance-history for join removal fix.", "body": "Bug: #19560\nReported-by: Orestis Markou \nAuthor: Tom Lane \nReviewed-by: Richard Guo \nReviewed-by: Thom Brown \nReviewed-by: Jacob Brazeal \nDiscussion: https://postgr.es/m/1186816.1784573544@sss.pgh.pa.us\nBackpatch-through: 16-18", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "cdec5b423a2a323b6a286a416b33fe935bfd290f", "date": "2026-08-28T23:03:03+02:00", "subject": "HISTORY.md: PSL support in 2015", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "4af0528a0e49d06b997c443b044cb013503e1546", "date": "2026-08-28T23:24:47+02:00", "subject": "Propagate rebalanced cost limit to parallel vacuum workers", "body": "AutoVacuumUpdateCostLimit() runs after each nap in vacuum_delay_point()\nand follows av_nworkersForBalance, but the new limit never reached the\nshared cost params in the vacuum DSM: propagation only required config\nreload. Parallel workers computed their delays from the stale limit,\nso a parallel autovacuum could run at up to twice the configured budget\n(or half of it) until the next SIGHUP, contradicting the propagation\npromise in maintenance.sgml.\n\nCall parallel_vacuum_propagate_shared_delay_params() after rebalancing.\nGated on the leader: parallel workers take the same nap path and must\nnot overwrite the shared parameters.\n\nThis also adds a test to validate the behaviour: pause the leader at the\nexisting injection point, start a second autovacuum worker and hold it at\na new injection point placed after it joined the balance, then check the\nfirst parameter load of the parallel workers reports the balanced limit.\nThe hold is needed because a second worker left running can finish its own\nvacuum before the leader resumes, which puts the balance back where it\nstarted. Autovacuum is disabled for everything but the two test tables\nvia thresholds, as a worker spawned by catalog churn would get trapped at\nthe hold point and starve the test of its second worker slot.\n\nBackpatch to v19 where autovacuum gained the ability to use parallel\nvacuum workers.\n\nAuthor: Zsolt Parragi \nReviewed-by: Bharath Rupireddy \nReviewed-by: Masahiko Sawada \nReviewed-by: Daniel Gustafsson \nDiscussion: https://postgr.es/m/CAN4CZFOZtEPwGQ6oa9LvHvN522zEp8h_dW9hHExSh7pVXofoKQ@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "bdbd99ab109c47996038077c5dd1ba8a1e2783a2", "date": "2026-08-29T00:59:26+02:00", "subject": "test(tui): unreliable \"TUI :detach % detaches other UIs\"", "body": "", "style": "conventional", "type": "test", "scope": "tui", "breaking": false} {"repo": "neovim", "sha": "bc16be3cd91b1d7508f91da97f5c95881eea5fc7", "date": "2026-08-29T02:03:12+02:00", "subject": "fix(cmdatom): operator with Lua textobject is not type=operator", "body": "Problem:\nAn operator completed by a Lua `:omap` textobject emits\n`CmdAtom.type=\"mapping\"` (lhs-only, no keys) instead of `type=\"operator\"`.\n\nAnalysis:\n`atom_redo_set()` declined K_LUA, though the prepped redo\n(\"op\" + K_LUA + id + CR) is exactly what \".\" replays. A no-edit \"g@\"\nemits nothing at all.\n\nSolution:\n- `atom_redo_set`: don't decline K_LUA; the redo route now captures the\n operator atom.\n- `atom_capture_cmd`: don't early-return if the frame has prepped redo.\n- op_function(): save/restore redobuff when invoking 'operatorfunc',\n like `call_user_func()` does for Vimscript. (Else the Lua callback\n may clobber the prepped \"g@\" redo / dot-repeat.)\n\nfix #41482\n\nTODO:\n- async Lua (timer/vim.schedule) can still clobber the pending dot-repeat...", "style": "conventional", "type": "fix", "scope": "cmdatom", "breaking": false} {"repo": "nestjs", "sha": "86a7f16388eb71cd63a9e54d1e9f343a0927c7f1", "date": "2026-08-29T13:28:56+09:00", "subject": "fix(core): throw `UndefinedDependencyException` for nil inject entries", "body": "", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "nestjs", "sha": "24db9f733537903f373ba5e68260ce37ef259ce9", "date": "2026-08-29T13:33:01+09:00", "subject": "fix(core): fix type error in dependencies cast", "body": "", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "openssl", "sha": "f7d2d2acef273929cf75e17374f263b483e58662", "date": "2026-08-29T02:20:52-04:00", "subject": "AEADs: add an error to the queue on tag mismatch", "body": "Decrypt with a wrong tag for each AEAD and inspect the error reason left\non the queue. Repeat for ciphertext only, AAD only, and ciphertext + AAD.\nThe tag rejection must fail at EVP_DecryptFinal_ex(), and leave error\nPROV_R_BAD_DECRYPT on the queue.\n\nAssisted-by: Claude:claude-opus-4-8\nReviewed-by: Neil Horman \nReviewed-by: Daniel Kubec \nReviewed-by: Tomas Mraz \nMerge-date: Fri Sep 4 17:39:21 2026\nMerged-from: https://github.com/openssl/openssl/pull/32587", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "8a2bb9ca241bbd82a0da536f6f39dca9037dd046", "date": "2026-08-29T10:01:42+02:00", "subject": "conncache: lower limit for auto maxconnects", "body": "When the `maxconnets` for a multi is not set, we calculate\nan \"auto\" limit based on the number of running transfers.\n\nIntroduce a minimum of the transfer count, so that we do\nnot evict connctions from the cache when many transfers\nfinish in the same round, causing an \"unnatural\" low running\ncount.\n\nThis should address our low `h1-request` perf stat drop.\n\nCloses #22734", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "007e81ac0ad66bd0be4887d88a276df292ae0bed", "date": "2026-08-29T17:08:02+09:00", "subject": "ci: skip EMFILE test on Node.js 26", "body": "* ci: fix `EMFILE` test on Node.js 26.8+\n\n* wip", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "eslint", "sha": "5661fa65fde9fd4c14f0b730e3cee6a42fc657c1", "date": "2026-08-29T13:49:38+05:30", "subject": "feat: no-extra-bind false negatives with class fields and static blocks", "body": "* fix: no-extra-bind should flag unnecessary bind with class fields/static blocks\n\n* apply review suggestions", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "eslint", "sha": "8724829f69f8ed80c876e3a5a017da199ce78739", "date": "2026-08-29T10:43:12+02:00", "subject": "docs: update compat table links", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vitest", "sha": "01a82372c54ccab2d9df53b3bfe80e5cf195bcd1", "date": "2026-08-29T08:51:58Z", "subject": "fix: prevent oxc/esbuild target duplication when merging configs", "body": "When a user has oxc.target or esbuild.target as an array in their Vite\nconfig, Vitest's ViteConfigPlugin was creating a new object with the\nsame target array. When Vite's mergeConfig merged these together,\nit would concatenate the arrays, causing duplicates.\n\nFor example, if user config had:\n oxc: { target: ['chrome121', 'firefox118'] }\n\nVitest would create:\n oxc: { target: ['chrome121', 'firefox118'] }\n\nAnd mergeConfig would produce:\n oxc: { target: ['chrome121', 'firefox118', 'chrome121', 'firefox118'] }\n\nThis caused OXC to fail with duplicate entries.\n\nThe fix is to not override oxc/esbuild config if the user has already\nprovided one with a target. We only set defaults when the user hasn't\nprovided the config or when they've provided it without a target.\n\nCloses #11035", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "curl", "sha": "916ffb10a8d2be4c8ad80c2e0e38f3fe583c7ea9", "date": "2026-08-29T10:53:15+02:00", "subject": "spnego_sspi: drop redundant UNICODE branch", "body": "Follow-up to 8a8ff47b63aff1d72c1796dbdb4955f243beac23 #22559\nFollow-up to a8881e5e1d2e22d4b085d45ab6c8ce1df251d602 #21315 #22410\n\nCloses #22737", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "b82235540f57bedccf6059236f7dfa32337687da", "date": "2026-08-29T11:02:50+02:00", "subject": "sspi: add local helper macro to avoid UNICODE branching", "body": "Also drop redundant cast and comment.\n\nCloses #22738", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "83ba3cb69c3ac30c2d03d2c577c6fc53dbcca856", "date": "2026-08-29T11:32:13+02:00", "subject": "apps: cover the dhparam -dsaparam file input in the test recipe", "body": "Reading DSA parameters from a file with -dsaparam and converting them\nto X9.42 DH parameters was not covered, nor were the error paths when\nthe input file key type does not match the -dsaparam expectation.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Daniel Kubec \nReviewed-by: Paul Dale \nMerge-date: Mon Sep 7 14:14:08 2026\nMerged-from: https://github.com/openssl/openssl/pull/32588", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "a00995705779b63ed288ecee67f696809b7b1d56", "date": "2026-08-29T12:21:16+02:00", "subject": "schannel: add ALPN support for mingw-w64 <9 and \nReviewed-by: Daniel Kubec \nMerge-date: Wed Sep 16 08:48:56 2026\nMerged-from: https://github.com/openssl/openssl/pull/32590", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "b9c5f2bcd93487d737a57bd6ffeeb3aa6528b322", "date": "2026-08-29T12:32:20+02:00", "subject": "apps: cover the dsa error cases in the test recipe", "body": "Cover the bad option failures (invalid input and output formats,\nextra arguments, unknown cipher and invalid password argument) as\nwell as the failures when loading an invalid or non-DSA key file\nand when requesting an unsupported output format, verifying the\nerror messages printed to stderr with the shared app_fails helper.\n\nReviewed-by: Richard Levitte \nReviewed-by: Daniel Kubec \nMerge-date: Wed Sep 16 08:48:56 2026\nMerged-from: https://github.com/openssl/openssl/pull/32590", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "e2b597de0ef14598901703f5aa52f8c962007d02", "date": "2026-08-29T18:52:23+08:00", "subject": "docs(config): fix wording and terminology", "body": "", "style": "conventional", "type": "docs", "scope": "config", "breaking": false} {"repo": "openssl", "sha": "ab566c794e6564d0624a509c533b66d82b0ce99d", "date": "2026-08-29T13:36:14+02:00", "subject": "apps: cover the rsa error cases in the test recipe", "body": "Cover the bad option failures (invalid input and output formats,\nextra arguments, unknown cipher and invalid password argument) as\nwell as the failures when checking a public key, loading an invalid\nor non-RSA key file and when requesting an unsupported or impossible\noutput format, verifying the error messages printed to stderr with\nthe shared app_fails helper.\n\nReviewed-by: Richard Levitte \nReviewed-by: Daniel Kubec \nMerge-date: Wed Sep 16 08:47:39 2026\nMerged-from: https://github.com/openssl/openssl/pull/32591", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "c9713ddbd0da8ba15e5f2d3af354e24759998ac9", "date": "2026-08-29T13:52:08+02:00", "subject": "apps: print an error for a non-positive genrsa bits argument", "body": "Previously a negative or zero numbits argument made genrsa exit\nwith a failure status without printing any error message.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Daniel Kubec \nReviewed-by: Richard Levitte \nMerge-date: Tue Sep 15 15:20:10 2026\nMerged-from: https://github.com/openssl/openssl/pull/32593", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "d6d126cc6fd623ce7efe85a1de2e0bf60230aaab", "date": "2026-08-29T13:52:08+02:00", "subject": "apps: print an error for a non-positive genrsa bits argument", "body": "Previously a negative or zero numbits argument made genrsa exit\nwith a failure status without printing any error message.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Dmitry Belyavskiy \nReviewed-by: Tomas Mraz \nMerge-date: Mon Sep 7 14:15:50 2026\nMerged-from: https://github.com/openssl/openssl/pull/32592", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "a6b29d79c0a81397fb509f53a9fdeb9aabc42bb0", "date": "2026-08-29T14:57:01+03:00", "subject": "docs: correct out-of-range highlight indices in two code blocks", "body": "Both indices point past the end of their block, so the intended lines are\nnever highlighted. The reactive forms block renders with no highlight at all,\nand the first example in output-interop.md highlights nothing while its\nsibling block is unaffected.\n\n(cherry picked from commit cd3c8cdd3a145c25017c48dc2fa9a8795e5a523d)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "cd3c8cdd3a145c25017c48dc2fa9a8795e5a523d", "date": "2026-08-29T14:57:01+03:00", "subject": "docs: correct out-of-range highlight indices in two code blocks", "body": "Both indices point past the end of their block, so the intended lines are\nnever highlighted. The reactive forms block renders with no highlight at all,\nand the first example in output-interop.md highlights nothing while its\nsibling block is unaffected.", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "4d4b0d910c3bcf7b95d06af098c5714919c3998b", "date": "2026-08-29T13:58:26+02:00", "subject": "test framework: add app_fails and slurp_file to OpenSSL::Test", "body": "Several test recipes duplicate a helper that runs an openssl\napplication expecting a failure exit status and checks that the\nstderr output matches a regular expression. Add a shared app_fails\nfunction so new recipes do not need to repeat it, together with a\nslurp_file function for reading back a captured output file.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Dmitry Belyavskiy \nReviewed-by: Tomas Mraz \nMerge-date: Mon Sep 7 14:15:49 2026\nMerged-from: https://github.com/openssl/openssl/pull/32592", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "da7d7a35fda97cff05e57ab1c13a72010857cf4b", "date": "2026-08-29T13:59:19+02:00", "subject": "apps: cover the genrsa error cases and verbose mode in the test recipe", "body": "Cover the bad option failures (unknown cipher, invalid primes,\nbits and password arguments) as well as the key generation failure\nwith an invalid number of primes, verifying the error messages\nprinted to stderr. Also cover the -verbose progress output and\ncheck that -quiet suppresses it.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Dmitry Belyavskiy \nReviewed-by: Tomas Mraz \nMerge-date: Mon Sep 7 14:15:50 2026\nMerged-from: https://github.com/openssl/openssl/pull/32592", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "5f972fbf20a59d8bdd089995965b1669a6e8a87d", "date": "2026-08-29T14:00:47+02:00", "subject": "apps: cover the gendsa error cases and verbose mode in the test recipe", "body": "Cover the bad option failures (missing and extra arguments, unknown\ncipher and invalid password argument) as well as the failures when\nloading a nonexistent, invalid or non-DSA parameters file, verifying\nthe error messages printed to stderr. Also extend the -verbose test\nto check the reported progress output and cover the -quiet option.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Dmitry Belyavskiy \nReviewed-by: Tomas Mraz \nMerge-date: Mon Sep 7 14:15:51 2026\nMerged-from: https://github.com/openssl/openssl/pull/32592", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "b2bd6a885aef5be6c0e7ce548015a6cfb0d769ab", "date": "2026-08-29T14:01:25+02:00", "subject": "apps: cover the genpkey error cases and verbose mode in the test recipe", "body": "Cover the bad option failures (missing algorithm, extra arguments,\ninvalid output format, unknown algorithm, key option, cipher and\npassword arguments) as well as the failures when loading a\nnonexistent or invalid parameters file, verifying the error messages\nprinted to stderr. Also cover the -verbose progress output and the\n-quiet option.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Dmitry Belyavskiy \nReviewed-by: Tomas Mraz \nMerge-date: Mon Sep 7 14:15:52 2026\nMerged-from: https://github.com/openssl/openssl/pull/32592", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "098d172373cc3f85aa9c38d0aa99beb4fd2a3e73", "date": "2026-08-29T20:09:27+08:00", "subject": "feat(ColorPicker): support custom gradient angles", "body": "", "style": "conventional", "type": "feat", "scope": "ColorPicker", "breaking": false} {"repo": "eslint", "sha": "08a02be429e21fc93d86c8dd16cec6dc945ea2c1", "date": "2026-08-29T21:12:47+09:00", "subject": "test: add error locations to `no-extra-boolean-cast`", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "openssl", "sha": "1b0887947f1599f7a8ed486c6d36aa659cc18efc", "date": "2026-08-29T14:31:22+02:00", "subject": "apps: cover the genrsa negative bits argument in the test recipe", "body": "Assisted-by: Claude:claude-fable-5\nReviewed-by: Daniel Kubec \nReviewed-by: Richard Levitte \nMerge-date: Tue Sep 15 15:20:11 2026\nMerged-from: https://github.com/openssl/openssl/pull/32593", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "ab707262b1ca4d5b9891f43ca18a59b0dd4154fe", "date": "2026-08-29T06:13:04-07:00", "subject": "fix(lsp): blank line before Content-Length breaks header parsing #41524", "body": "Problem:\nA blank line before the Content-Length header makes header parsing\nfail with \"Content-Length not found in header\". An LF in the 'name'\nstate falls through to the 'invalid' state, which only return to\n'name' at the *next* LF. That LF terminates the next line, so the\nline with Content-Length is swallowed.\nThe same issue happens when a junk line is a partial match of the\nheader name (e.g. \"Cont\\nContent-Length: ...\").\n\nSolution:\nWhen seeing an LF in the 'name' state, reset the cursor and stay in\n'name' rather than entering 'invalid'.", "style": "conventional", "type": "fix", "scope": "lsp", "breaking": false} {"repo": "openssl", "sha": "6caadd9aa37acac83200d77284efafc754892438", "date": "2026-08-29T15:18:17+02:00", "subject": "apps: cover the pkeyutl asn1parse and hexdump options in the test recipe", "body": "Use RSA -verifyrecover to recover the signed DigestInfo blob and check\nthat -asn1parse pretty-prints its ASN.1 structure and that -hexdump\nprints its hex dump. Also cover -asn1parse on a raw RSA signature,\nwhere the encoding error is reported but the command still succeeds.\n\nAlso fix the stale skip count of the RSA SKIP block, which was not\nupdated when the -rev subtest was added.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Daniel Kubec \nReviewed-by: Richard Levitte \nMerge-date: Tue Sep 15 15:18:22 2026\nMerged-from: https://github.com/openssl/openssl/pull/32595", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "abce26ecc997b7021bbdf9bb78e10af598885121", "date": "2026-08-29T18:39:57+05:00", "subject": "fix(common): serialize errorCode in built-in exception responses", "body": "", "style": "conventional", "type": "fix", "scope": "common", "breaking": false} {"repo": "openssl", "sha": "39952ff9120b47cf424a60d4c972a69778c1d7f7", "date": "2026-08-29T15:40:42+02:00", "subject": "apps: cover the ec no_public option in the test recipe", "body": "Write the testec-p256.pem private key in DER form with and without\n-no_public and check that the stripped encoding is smaller and matches\na new checked-in reference file. Also check that the public key is\nrecomputed from the private scalar when loading such a key, by\ncomparing -pubout output against the existing ec-conv-unc.der\nreference.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Daniel Kubec \nReviewed-by: Paul Dale \nMerge-date: Mon Sep 7 14:16:44 2026\nMerged-from: https://github.com/openssl/openssl/pull/32596", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "af94588df6f5c14683f553577adb5235be8efe52", "date": "2026-08-29T21:49:24+08:00", "subject": "chore: pin nwsapi to avoid CI failures", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "neovim", "sha": "063c76f500883600b3a5be186a0bf98c25002063", "date": "2026-08-29T10:01:08-04:00", "subject": "build(vim-patch): v8.2.3068 is almost auto n/a", "body": "Nvim uses utf8proc or an older process involving \"src/unicode/\"\nand \"download-unicode-files.sh\" to decouple unicode characters\nfrom \"mbyte.c\".\nThese unicode patches should be auto N/A but the diff hunk xfuncname\nis wrong for utf_char2cells().\nIs it because of the `#ifdef`?\nNeed a volunteer to inspect if these patches still have relevant\nchanges\n\nNvim\n\n```c\n/// same as utf_composinglike but operating on UCS-4 values\nbool utf_iscomposing(int c1, int c2, GraphemeState *state)\n{\n return (!utf8proc_grapheme_break_stateful(c1, c2, state)\n || arabic_combine(c1, c2));\n}\n```\n\nVim\n\n```c\n/*\n * Return TRUE if \"c\" is a composing UTF-8 character. This means it will be\n * drawn on top of the preceding character.\n * Based on code from Markus Kuhn.\n */\n int\nutf_iscomposing(int c)\n{\n // Sorted list of non-overlapping intervals.\n // Generated by ../runtime/tools/unicode.vim.\n static struct interval combining[] =\n {\n\t{0x0300, 0x036f},\n\t{0x0483, 0x0489},\n\t{0x0591, 0x05bd},\n // ...\n\t{0x0300, 0x036f},\n\t{0x0483, 0x0489},\n\t{0x0591, 0x05bd},\n };\n\n return intable(combining, sizeof(combining), c);\n}\n```", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "openssl", "sha": "da09078c973e5f4e44b7e5c5714cbe0d6e4f09ce", "date": "2026-08-29T10:11:41-04:00", "subject": "Build for the right targets on msvc2013 in compiler zoo CI", "body": "We need to build for the MSVC2013 targets when building with MSVC2013,\nso that we pull in our home rolled version of snprintf\n\nReviewed-by: Saša Nedvědický \nReviewed-by: Norbert Pocs \nMerge-date: Tue Sep 1 11:51:13 2026\nMerged-from: https://github.com/openssl/openssl/pull/32597", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "536eba5ef618f63512a237548d96fbe14332523a", "date": "2026-08-29T23:29:40+09:00", "subject": "refactor(websockets): drop the redundant pattern append in the ws context creator", "body": "`WsProxy` already appends the message pattern to the argument list before it invokes\nthe target callback, so the `args.push(targetPattern)` inside that callback added the\nsame value a second time. Guards, interceptors and pipes received the pattern twice.\n\nBehaviour is unchanged: `getPattern()` reads the last element and both entries held the\nsame value. Added a regression test that asserts the pattern appears exactly once and is\nstill reachable through `getPattern()`.", "style": "conventional", "type": "refactor", "scope": "websockets", "breaking": false} {"repo": "nestjs", "sha": "75958d7ab15a9b17f7aba7b486f7b8c4cae5fb78", "date": "2026-08-29T23:30:44+09:00", "subject": "fix(scripts): stop hiding build failures in the setup scripts", "body": "`prepare.sh` and `run-integration.sh` sent the build output to /dev/null and had no\nshebang, no `set -e` and no exit status check. A failed build was invisible, the script\ncarried on into the Docker steps and still exited 0.\n\nThe `&>` redirect was also not portable. Under dash, which is /bin/sh on most Debian and\nUbuntu systems, it parses as \"run in the background\" plus a separate no-op redirect, so\nthe build raced the steps that followed instead of being muted.\n\nAdds `#!/usr/bin/env bash` and `set -eu` to both scripts and drops the redirect.\n`pipefail` is left out on purpose so the scripts still work when invoked with `sh` on a\ndash system; neither script uses a pipeline and neither reads a shell variable.\n\nAlso corrects CONTRIBUTING.md. `npm run build` runs `move:node_modules`, which copies the\ncompiled output into the root `node_modules/@nestjs`; `move:samples` is a separate task\nthat only `build:samples` invokes. Two places claimed the build moves packages into the\n`sample` directories. `prepare.sh` also starts Docker containers and waits for RabbitMQ,\nwhich its description did not mention, and `npm run test` needs neither a build nor\nDocker, so it now comes before the integration setup.", "style": "conventional", "type": "fix", "scope": "scripts", "breaking": false} {"repo": "neovim", "sha": "84e429a0b2c3b0a17485d2f4484d840297524832", "date": "2026-08-29T16:35:41+02:00", "subject": "build(zig): use system library for libiconv on macOS", "body": "Problem: Unlike CMake, `build.zig` downloads and builds GNU libiconv\non macOS instead of linking to the system framework.\n\nSolution: Only pull in the libiconv dependency on Windows and link\nagainst system framework on macOS.", "style": "conventional", "type": "build", "scope": "zig", "breaking": false} {"repo": "neovim", "sha": "fbc5a769aa44c3a66f1ddd59cc9930c0dd7f365b", "date": "2026-08-29T20:52:37+05:30", "subject": "fix(ui2): don't draw MsgSeparator if cmdheight=0 #41531", "body": "Problem:\nWhen 'cmdheight' is 0, the MsgSeparator can obscure the statusline when\na message is displayed, making its contents disappear until\nthe next redraw.\n\nSolution:\nSkip the separator when 'cmdheight=0' available.", "style": "conventional", "type": "fix", "scope": "ui2", "breaking": false} {"repo": "postgres", "sha": "f23de46e15ba3faf38a97915df29ae98a472a16b", "date": "2026-08-29T10:23:23-05:00", "subject": "Disallow ONLY in REPACK commands.", "body": "The REPACK grammar accepts ONLY before the table name and * after\nthe table name, but that's neither documented nor handled in the\ncode. Perhaps REPACK should support that syntax, but for now let's\njust bring it in line with its documentation.\n\nOversight in commit ac58465e06.\n\nReviewed-by: Antonin Houska \nDiscussion: https://postgr.es/m/apBTsWGwkLXVh8Ow%40nathan\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "5b278a4dd6e826271498b4553bf2010904786542", "date": "2026-08-29T17:34:43+02:00", "subject": "test: use fs.watch for the async leak fixture", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "9b71cad975504156529a80b90a9b25083ce4cd94", "date": "2026-08-29T21:52:26+06:00", "subject": "fix(core): correctly discover value providers in discovery service", "body": "Providers registered via useValue with a custom discovery decorator\nwere not being discovered by DiscoveryService.getProviders because a\nternary condition in DiscoverableMetaHostCollection was missing a\nnegation operator.\n\nFixes #17617", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "vitest", "sha": "ccd3059efeb9e2c6b3a01c2fa289c23e6e83a865", "date": "2026-08-29T18:06:18+02:00", "subject": "docs: document TZ limitation in thread pools", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vitest", "sha": "848d7243e608efcfb6c31baaa722f17a094a841d", "date": "2026-08-29T18:06:45+02:00", "subject": "fix: don't duplicte `oxc.target` if user provides a custom array", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "vitest", "sha": "3f2ddc21a83f3ebf16be50a1b31982ee9060266f", "date": "2026-08-29T18:08:13+02:00", "subject": "fix(vm): link module graphs one at a time", "body": "", "style": "conventional", "type": "fix", "scope": "vm", "breaking": false} {"repo": "curl", "sha": "8ec97d9fff05445c1a72d12d1cf36518b5fc28e8", "date": "2026-08-29T18:26:46+02:00", "subject": "tests/perf: add a subdir for libcurl performance test programs", "body": "These programs does not build by default since they are a bit niche and\nmaybe not interesting to most.\n\nThe idea is to have a set of tests here that can be used to verify\nvarious libcurl function's performance when we have no other good means\nof doing so. Performance is best tested relatively. Does it run slower\nor faster than before?\n\n'make -C tests/perf' builds the performance test tools in an autotools\nbuild.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "d4fe1986fe334d22940039e815ccdbcc907baa6a", "date": "2026-08-29T18:43:32+02:00", "subject": "feat: promote clearCache out of experimental", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "babel", "sha": "d061c26365c42272f3c7666451fbe25743c22a20", "date": "2026-08-29T16:28:51-04:00", "subject": "fix(remove): unwrap single sequence expression", "body": "* fix(remove): unwrap single sequence expression\n\n* Apply suggestions from code review\n\nCo-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>\n\n---------", "style": "conventional", "type": "fix", "scope": "remove", "breaking": false} {"repo": "neovim", "sha": "c47a2fbf7b1a04f2959b8dcc13a40598b3a374a5", "date": "2026-08-29T19:37:46-04:00", "subject": "build(vim-patch): v8.2.3597 is n/a", "body": "Nvim already has 'scrollback' option.", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "postgres", "sha": "b4c8afdfaf52ce4a5839be6b4d205e696b00a543", "date": "2026-08-29T19:56:49-04:00", "subject": "Harden spell.c against out-of-order FLAG lines in Hunspell files.", "body": "The compound flags collected from COMPOUNDFLAG and friends are stored in\neither the string or the integer member of a union, according to the\nflag mode that the affix file's FLAG line declares. NIImportOOAffixes()\nconverted each flag as soon as it read it, using the mode in effect at\nthat point, and recorded that mode in the entry. Since FLAG may appear\nanywhere in the file, including after the compound flags, entries written\nbefore and after it could disagree about which member of the union holds\nthe flag.\n\nIn assert-enabled builds, this would result in an assertion failure.\nOtherwise, cmpcmdflag() takes the mode from its first argument and\napplies it to both, so it can read an integer as a char pointer and pass\nthat to strcmp(). Depending on which way the mismatch goes, the result\nis a segfault while sorting the array, a segfault in the bsearch() that\nlater looks flags up (the lookup key is built with the final mode, so\nthis happens even when the array itself is consistent), or, when both\nmembers happen to be readable, no crash at all and a compound flag that\nis never found, which silently disables compound word splitting. This\nisn't a security bug because we consider dictionary files to be trusted\ndata, but it's still worth fixing. (In practice, dictionary files\nusually put the FLAG line first, which is why this went unreported for\nso long.)\n\nFix by keeping the flags as strings while the file is read and converting\nthem once it has been read in full, when the mode is final. This also\nmakes the position of the FLAG line irrelevant, which is how the flags on\nAF, SFX and PFX lines are already treated: those are parsed in a second\npass and so always use the final mode. That precedent is reason for\nbehaving this way rather than throwing an error.\n\nThe old ispell file format reaches addCompoundAffixFlagValue() too,\nfrom NIImportAffixes(), and returns without entering NIImportOOAffixes(),\nso it needs the conversion step as well.\n\nWhile we're here, also fix some i", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "2fb8da5a245661287833b05a1b2e275ddf83bbd7", "date": "2026-08-29T20:11:56-04:00", "subject": "Drop the per-entry copy of the flag mode in CompoundAffixFlags.", "body": "Each CompoundAffixFlag carried its own copy of the dictionary's flag\nmode, which decides whether the union holds a string or an integer.\nThe comment on the field explained why: cmpcmdflag() needs the mode,\nand at the time there was no bsearch() variant that would allow\npassing that context through. But that problem is long gone, so let's\nremove the redundant storage. No flexibility is lost because things\nwould not work anyway if different flags within a dictionary have\ndifferent representations; and the preceding commit has seen to it\nthat they can't. At present, this saves no storage on 64-bit machines\nbecause of alignment issues, but it still seems like good\nsimplification.\n\nNo back-patch: there's little value in this except cleanliness, and\nalthough probably no outside code is looking at this data structure,\nthis'd be an API/ABI break if any does.\n\nAuthor: Ewan Young \nReviewed-by: Tom Lane \nDiscussion: https://postgr.es/m/CAON2xHN3QmsaySM6DGWa1gttcbJoFh0wjAE-_ZpSPo=LKN1hYw@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "9b6635965fda49fd4566612df7293730c30f930e", "date": "2026-08-30T08:24:03+08:00", "subject": "types(runtime-vapor): export DefineVaporSetupFnComponent", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "97f06b936e8a758f920d107b2f7ae14edb595437", "date": "2026-08-29T20:24:27-04:00", "subject": "build(vim-patch): v8.2.4535 is n/a", "body": "Nvim removed support for DOS 8.3 filename.", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "babel", "sha": "2b42cacdf5b5605c654612011a5f9cb0dee73513", "date": "2026-08-30T04:10:04+03:00", "subject": "fix: import built helper instead of src/ in tsRewriteRelativeImportExtensions test", "body": "* fix: allow babel-helpers unit tests to import from src/\n\nTest files must not import from src/ per the no-restricted-imports\nrule, but this one legitimately needs to: the helpers/ tsconfig only\ntype-checks (noEmit), so there's no compiled lib/helpers/*.js to\nimport instead. Scope an exception to packages/babel-helpers/test/\nunittests/ rather than testing against src directly.\n\n* review\n\n---------", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "curl", "sha": "d623b84832a88fe7a859a81c9d64fe0708335e29", "date": "2026-08-30T00:51:14-04:00", "subject": "schannel: fix ALPN erroneously disabled", "body": "- In easy.c global_init, initialize win32 before ssl.\n\nSchannel SSL init depends on win32 initialization.\n\nPrior to this change applications with no compatibility manifest using a\nlibcurl with Schannel had ALPN erroneously disabled. During Schannel\ninit there is a version info check so that ALPN is only enabled for\nWindows 8.1 or later (>= NT 6.3), but since win32 was not yet\ninitialized the version check would fail even if the OS was more recent.\n\nThis change ensures that the pointer to function RtlVerifyVersionInfo is\ninitialized before SSL initialization so that version checks during SSL\ninitialization do not use VerifyVersionInfo as a fallback.\n\nOn Windows RtlVerifyVersionInfo compares against the actual OS version\ninfo whereas VerifyVersionInfo has behavior that varies depending on\nwhether the application (eg curl tool) has a compatibility manifest. If\nthere is no manifest, VerifyVersionInfo treats the OS version to compare\nagainst as Windows 8 (NT 6.2) even on later versions.\n\nSo basically NT 6.2 < NT 6.3 caused ALPN to be erroneously disabled.\n\nReported-by: Jay Satiro\n\nRef: https://learn.microsoft.com/en-us/windows/win32/api/winbase/nf-winbase-verifyversioninfoa\n\nFixes https://github.com/curl/curl/pull/22739#issuecomment-5464013950\n\nCloses #22746", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "d73f32ace70a5a99747fbc8b637217f07b174b2a", "date": "2026-08-30T13:17:15+08:00", "subject": "type: restore jest-puppeteer global types", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "fb31e73c99815ad23bd0b75b2f08fca9a1d759f7", "date": "2026-08-30T13:17:59+08:00", "subject": "docs(blog): fix preposition", "body": "", "style": "conventional", "type": "docs", "scope": "blog", "breaking": false} {"repo": "ant-design", "sha": "df3386b019b2b64a2daed3e3470360e61b21a3d4", "date": "2026-08-30T13:19:38+08:00", "subject": "fix(locale): correct Traditional Chinese picker terms", "body": "", "style": "conventional", "type": "fix", "scope": "locale", "breaking": false} {"repo": "openssl", "sha": "398553b5a2f65f5c8e34d72a6f6b2ee7ea086b10", "date": "2026-08-30T18:28:36+09:00", "subject": "DTLS 1.3: truncate the unified header sequence number and reconstruct it", "body": "The DTLS 1.3 unified header carries only the low 8 or 16 bits of the\nrecord sequence number. Passing the full uint64_t value to\nWPACKET_put_bytes_u16() made SSL_write() fail when the counter reached\n65536.\n\nMask only the value encoded on the wire and in the corresponding AEAD\nadditional data, leaving the full counter intact for the nonce. On\nreceipt, reconstruct the full value relative to the right edge of the\ncurrent epoch's replay window, as recommended by RFC 9147 section 4.2.2.\n\nAdd boundary vectors for both wire widths and an end-to-end test across\nthe 16-bit wrap.\n\nFixes #32584\n\nAssisted-by: pi:Hy4-preview\nAssisted-by: Codex:gpt-5.6-sol\nReviewed-by: Ryan Hooper \nReviewed-by: Andrew Dinh \nMerge-date: Mon Sep 7 14:20:03 2026\nMerged-from: https://github.com/openssl/openssl/pull/32601", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "5e0b1464b2313075c5fa94632fe66cc370fa7d9e", "date": "2026-08-30T12:19:38+02:00", "subject": "docs(cmake): remove misleading comment", "body": "We still require this file since the upstream version can't find the\nglibc version on Linux.", "style": "conventional", "type": "docs", "scope": "cmake", "breaking": false} {"repo": "neovim", "sha": "eb18f29748b3ab232cdd0649a9d87f294cab5ad6", "date": "2026-08-30T07:13:02-04:00", "subject": "test(tui): unreliable \"exits immediately when stdin is closed\" #41547", "body": "Problem:\n0c091cedc291 fixed the \"immediate exit\" but the 2s `nvim_get_proc()`\ncheck still fails on slow (ASAN) CI.\n\nSolution:\nCheck the pid after the `screen:expect`. Anyway, `assert_nolog` is the\n\"meaningful\" part of the test since 0c091cedc291.", "style": "conventional", "type": "test", "scope": "tui", "breaking": false} {"repo": "ant-design", "sha": "bd10fa56d4ac8955c15feb475db4f2130aaa4773", "date": "2026-08-30T19:32:59+08:00", "subject": "fix(Radio): treat undefined group value as uncontrolled", "body": "", "style": "conventional", "type": "fix", "scope": "Radio", "breaking": false} {"repo": "nestjs", "sha": "a9bf20b6936f53a136bf845c480f19ae534a2d8b", "date": "2026-08-30T12:12:23Z", "subject": "fix(core): Serialize errorCode for plain HttpException responses", "body": "A plain HttpException constructed with a string body and an errorCode\noption sets exception.errorCode but BaseExceptionFilter rebuilds the\nresponse body from getResponse() and drops the errorCode, so the\nserialized body is {statusCode:403,message:Forbidden} instead of\nincluding errorCode.\n\nThe built-in exceptions path was fixed by #17615 (forwards\nhttpExceptionOptions.errorCode into createBody); this closes the same\ngap for the bare-string path in @nestjs/core, as offered as a follow-up\nin that PR. Object responses are preserved verbatim and the no-errorCode\ncase is unchanged.\n\nTests: two new cases in exceptions-handler.spec.ts; the errorCode one\nfails without this change.", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "neovim", "sha": "10d7cca784e7914a7acb6b4a513e617565dddc84", "date": "2026-08-30T05:32:55-07:00", "subject": "fix(snippet): cancel session on ESC in insert-mode #41555", "body": "Problem:\nIf a snippet does not have a placeholder, we use insert mode instead of\nselect mode. From here leaves the session and highlight active.\n\nSolution:\nCancel the session on .\n\n(cherry picked from commit 39862231b2889db55568ad805f1885fd8cfc5248)", "style": "conventional", "type": "fix", "scope": "snippet", "breaking": false} {"repo": "neovim", "sha": "39862231b2889db55568ad805f1885fd8cfc5248", "date": "2026-08-30T05:32:55-07:00", "subject": "fix(snippet): cancel session on ESC in insert-mode #41555", "body": "Problem:\nIf a snippet does not have a placeholder, we use insert mode instead of\nselect mode. From here leaves the session and highlight active.\n\nSolution:\nCancel the session on .", "style": "conventional", "type": "fix", "scope": "snippet", "breaking": false} {"repo": "curl", "sha": "b4242c52c17fcf5c2fcd7e5d40a53c1a1b197d9a", "date": "2026-08-30T14:42:00+02:00", "subject": "fixup the linux CI", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "9f05091b596e3015bf214ea15f6c56f6eb7647f6", "date": "2026-08-30T15:20:25+02:00", "subject": "GHA/non-native: bump cross-platform-actions to 1.5.0", "body": "Also:\n- tried `microvm` with NetBSD 11, but ended up not using it because\n tests failed to fork, then the process either hung or took too long to\n fit in 15 minutes.\n Ref: https://github.com/curl/curl/actions/runs/33314257008/job/99268019716?pr=22748\n\nRefs:\nhttps://github.com/cross-platform-actions/action/releases/tag/v1.5.0\nhttps://github.com/cross-platform-actions/action/blob/v1.5.0/readme.md#variants-variant\n\nCloses #22748", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "ffbb7e6d528f798d2dc82feb059006d32f519cc9", "date": "2026-08-30T16:33:14+03:00", "subject": "docs: correct stale visibleLines ranges in the first-app tutorial", "body": "The tutorial's example files shifted without the README line references\nfollowing, so several steps point at the wrong code. In two cases every\nrequested line is past the end of the file; on the inputs step this renders\nas a code block with no lines visible at all.\n\n(cherry picked from commit 46b02cc18122ee4221e41f518ed65d53ea21c4d1)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "46b02cc18122ee4221e41f518ed65d53ea21c4d1", "date": "2026-08-30T16:33:14+03:00", "subject": "docs: correct stale visibleLines ranges in the first-app tutorial", "body": "The tutorial's example files shifted without the README line references\nfollowing, so several steps point at the wrong code. In two cases every\nrequested line is past the end of the file; on the inputs step this renders\nas a code block with no lines visible at all.", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "neovim", "sha": "25f7c87a702fa246004dad85c93d9df421c22ad1", "date": "2026-08-30T15:50:09+02:00", "subject": "fix(mappings): replaying a deleted Lua mapping is UB", "body": "Problem:\nReplaying a deleted Lua mapping, may call an arbitrary function.\n\nRHS of a Lua mapping embeds its LuaRef (``). The raw\nkeys may outlive the ref (redobuff \".\", CmdAtom `keys`). If the mapping\nis deleted, replaying it either (1) dereferences a freed registry slot,\nor (2) calls whatever callback reused the slot (autocmd, timer, other\nmapping).\n\nSolution:\nAssign a monotonic (never recycled) id to Lua mappings and encode the\nmapping keys as ``.\n\nNote: in the case of Vimscript, a deleted function raises E117, but if\nthe function is redefined with the same name, the mapping will find it.\n\nAlternatives?:\n- Globally ensure `LuaRef` ids are not recycled.\n - Problem: could exhaust `int` in a long-lived Nvim session? Also,\n difficult to impl bc the \"recycling\" is done by `luaL_ref` itself.\n\nref: 5ac2e47acc999472042df4f10f8f7b5ffa72ba3e", "style": "conventional", "type": "fix", "scope": "mappings", "breaking": false} {"repo": "postgres", "sha": "92819e57945d106317eaaa4273180dd8663f91bd", "date": "2026-08-30T09:13:09-05:00", "subject": "doc: Mark up \"option\" in the CHECKPOINT synopsis.", "body": "Oversight in commit a4f126516e.\n\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "59335b5e71d552229f74ac41f3ef6b5a4d966535", "date": "2026-08-30T09:40:40-05:00", "subject": "doc: Fix the data type named in the regdatabase entry.", "body": "Oversight in commit b45137f315.\n\nBackpatch-through: 19 only", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "nestjs", "sha": "52b5e10a90cbfa11de46de68e5270b9f26612c73", "date": "2026-08-30T21:48:13+06:00", "subject": "fix(core): restore logger after silent lazy module load", "body": "LazyModuleLoader.load() with { logger: false } replaced the shared\nInstanceLoader logger with a SilentLogger but never restored it,\npermanently silencing logs for all subsequent load() calls.\n\nFixes #17628", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "nestjs", "sha": "9847b54434d46a5d9f9a85667c1639ca5cc85178", "date": "2026-08-30T11:51:22-04:00", "subject": "fix(ws): remove the upgrade listener on dispose", "body": "`ensureHttpServerExists` attaches an anonymous `upgrade` listener and\nkeeps no reference to it. `dispose()` closes the servers the adapter\ncreated, which takes their listeners with them, but it deliberately\nskips a server passed in by the caller. Nothing removes the listener\nthere, and no reference is left to remove it with.\n\nThe listener outlives `app.close()` and still answers upgrades. Once a\nlater app registers a gateway on a different path, the stale listener\nfinds no match and destroys the socket, so the live gateway never sees\nthe connection.\n\nKeep the handler and take it off in `dispose()`.\n\nCloses #17629", "style": "conventional", "type": "fix", "scope": "ws", "breaking": false} {"repo": "curl", "sha": "589885149f50b28eae5c7217bc4fd81e51cbc216", "date": "2026-08-30T18:33:33+02:00", "subject": "easy_lock: silence `portability-no-assembler` with clang-tidy 23.1.0+", "body": "```\nlib/easy_lock.h:74:7: warning: do not use inline assembler in safety-critical code [portability-no-assembler]\n 74 | __asm__ volatile(\"yield\" ::: \"memory\");\n | ^\n```\n\nCloses #22751", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "2a395a0420516ac193ba7068e6449a7cc9d808d2", "date": "2026-08-30T18:35:13+02:00", "subject": "easy: fix unused global on non-Windows", "body": "Detected by llvm/clang 23.1.0:\n```\nlib/easy.c:83:13: warning: variable 'easy_init_flags' set but not used [clang-diagnostic-unused-but-set-global]\n 83 | static long easy_init_flags;\n | ^\nlib/easy.c:83:13: warning: variable 'easy_init_flags' set but not used [-Wunused-but-set-global]\n 83 | static long easy_init_flags;\n | ^\n```\n\nCloses #22750", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "babel", "sha": "c095ce19f49a727cc73fb28dbb511986d2062c49", "date": "2026-08-30T14:06:50-04:00", "subject": "fix(bugfix): bound check before array access", "body": "", "style": "conventional", "type": "fix", "scope": "bugfix", "breaking": false} {"repo": "ant-design", "sha": "c99d71d06cf35e73747ac8aca1eab2a1a5de2a7b", "date": "2026-08-31T02:07:27+08:00", "subject": "fix(demo): disable upload button in disabled demo", "body": "", "style": "conventional", "type": "fix", "scope": "demo", "breaking": false} {"repo": "angular", "sha": "ddc8d40a83e9e0cde442c2ac5717c258dc9f9a08", "date": "2026-08-30T22:54:00+03:00", "subject": "fix(docs-infra): restore the edit link on decorative header pages", "body": "The 34 pages using `` render their title\nthrough `getPageTitle()` without passing the markdown file path, so\nthe \"Edit this page\" link is silently dropped. Every other page keeps\nit. Compare https://v19.angular.dev/guide/components, which still has\nthe pencil, against https://angular.dev/guide/components, which does\nnot.\n\n`filePath` was required until #63536 made it optional, so API\ndescriptions with no editable source could render a title without a\nlink. That removed the compile error forcing the decorative header to\nsupply it, and the argument was lost with nothing to catch it.\n\nPass the path again and cover both header variants with tests, since\nthe edit link had no coverage at all.\n\n(cherry picked from commit acdac1cb89df320b43cd20f1163ad17480d06b2f)", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "acdac1cb89df320b43cd20f1163ad17480d06b2f", "date": "2026-08-30T22:54:00+03:00", "subject": "fix(docs-infra): restore the edit link on decorative header pages", "body": "The 34 pages using `` render their title\nthrough `getPageTitle()` without passing the markdown file path, so\nthe \"Edit this page\" link is silently dropped. Every other page keeps\nit. Compare https://v19.angular.dev/guide/components, which still has\nthe pencil, against https://angular.dev/guide/components, which does\nnot.\n\n`filePath` was required until #63536 made it optional, so API\ndescriptions with no editable source could render a title without a\nlink. That removed the compile error forcing the decorative header to\nsupply it, and the argument was lost with nothing to catch it.\n\nPass the path again and cover both header variants with tests, since\nthe edit link had no coverage at all.", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "d90698dae759f21f29052e521ffaec6c3a79f6d1", "date": "2026-08-30T20:01:13Z", "subject": "fix(compiler-cli): check uninvoked signal aliases in extended diagnostic", "body": "Extended template diagnostic interpolated_signal_not_invoked previously\nonly checked symbols with kind === SymbolKind.Expression. When a signal\nis aliased via @let (or a template variable), getSymbolOfNode returns\na LetDeclarationSymbol (or VariableSymbol), causing the diagnostic to\nskip checking uninvoked usages of signal aliases in interpolations and\nbindings.\n\nThis commit updates interpolated_signal_not_invoked to also check\nLetDeclarationSymbol and VariableSymbol, using the usage site's AST\nname span for reporting the diagnostic.\n\nCloses #70476\n\n(cherry picked from commit faafd18a4c0c5db161e5d7056e9abfb8675904cb)", "style": "conventional", "type": "fix", "scope": "compiler-cli", "breaking": false} {"repo": "angular", "sha": "faafd18a4c0c5db161e5d7056e9abfb8675904cb", "date": "2026-08-30T20:01:13Z", "subject": "fix(compiler-cli): check uninvoked signal aliases in extended diagnostic", "body": "Extended template diagnostic interpolated_signal_not_invoked previously\nonly checked symbols with kind === SymbolKind.Expression. When a signal\nis aliased via @let (or a template variable), getSymbolOfNode returns\na LetDeclarationSymbol (or VariableSymbol), causing the diagnostic to\nskip checking uninvoked usages of signal aliases in interpolations and\nbindings.\n\nThis commit updates interpolated_signal_not_invoked to also check\nLetDeclarationSymbol and VariableSymbol, using the usage site's AST\nname span for reporting the diagnostic.\n\nCloses #70476", "style": "conventional", "type": "fix", "scope": "compiler-cli", "breaking": false} {"repo": "curl", "sha": "5bb0bea0845042109f3f1d201861d249f817cf12", "date": "2026-08-30T22:03:45+02:00", "subject": "runtests: Added runnernums in order to have O(1) lookup", "body": "This pull request adds runnernums the reverse of runnerids. In this way\nthe iteration over runnerids is no longer needed. The check can happen\ndirectly in runnernums.\n\nSigned-off-by: Mihai-Gabriel Marin \n\nCloses #22753", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "6a3ebdb9f05185e0e73be9a2bbd6992896fdc0f8", "date": "2026-08-30T22:15:44+02:00", "subject": "winsock: drop redundant version checks at initialization", "body": "All supported Windows versions (Vista+) support Winsock 2.2, therefore\nthese checks always succeed for a binary that loaded and started up\nsuccessfully.\n\nRef: https://learn.microsoft.com/windows/win32/api/winsock/nf-winsock-wsastartup\n\nCloses #22754", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "15818f49e109a10c12b56132eaeedc3d31fcaf54", "date": "2026-08-30T18:18:10-03:00", "subject": "cf-socket: avoid broken NetBSD SOCK_NONBLOCK", "body": "NetBSD releases before 6.1 expose SOCK_NONBLOCK without setting the\nsocket-layer nonblocking state. Use curlx_nonblock() on those releases\nfor connection and wakeup sockets so transfers cannot block while\nwaiting for a response. Exercise repeated short MIME reads in test1662.\n\nFixes #22586\nReported-by: John Verne\nCloses #22755", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "fcde6334507993cc1edb39922aae34dcc95d0add", "date": "2026-08-31T00:01:17+02:00", "subject": "global_init: tidy up, simplify, fix issues", "body": "- ssh: deinit before tearing down Winsock. (also to sync with reverse\n init order.)\n\n- winsock: fix to deinitialize if any of the subsequent initialization\n fails.\n\n- win32: init version info and clock before anything else. To make\n both usable by init code and also because these inits cannot fail.\n\n- win32: move init logic from system_win32.c back into easy.c, drop\n the umbrella functions to allow more granular control over init order,\n to simplify, and expose the context.\n\n- sspi: init before Winsock, to avoid a cycling the latter in case of\n an init failure.\n\n- trc: init after win32 version info and clock but before others.\n To allow tracing during initialization.\n\n- async: sync error message with internal function name.\n\nBug: https://github.com/curl/curl/pull/22746/files#discussion_r3889939149\n\nCloses #22756", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "7dc0592fca3190a9f6ce4d83e8af33d88c5d33e7", "date": "2026-08-31T04:03:59+05:30", "subject": "fix(ui2): cursor flicker in cmdline with matchit #41245", "body": "Problem:\nThe cmdline_hide callback causes an unnecessary cursor move to the cmdline,\nbefore the normal redraw updates the cursor back to the current window. This\nappears as \"flicker\" when using plugins such as matchit (legacy \":\" mappings\ninstead of \"\" mappings).\n\nSolution:\nSkip the immediate redraw for cmdline_hide events. The normal redraw\nstill updates the cursor after the cmdline window is hidden.", "style": "conventional", "type": "fix", "scope": "ui2", "breaking": false} {"repo": "nestjs", "sha": "1637abbe9b0ef4e7a6b4af8b224ca036b6961f15", "date": "2026-08-30T19:36:54-03:00", "subject": "test(ws): cover the message handler binding failure paths", "body": "The three failures that the ws adapter currently swallows had no unit\ncoverage: a throwing custom message parser, a malformed frame, and an event\nwith no registered handler.\n\nAdds a case for each, plus the void-parser and happy paths, and one that pins\nthe current behaviour of a throwing handler so the connection teardown risk\nstays visible.\n\nRefs #17624", "style": "conventional", "type": "test", "scope": "ws", "breaking": false} {"repo": "openssl", "sha": "6d2d67f95498842976e63938a44985d2adb0e3a1", "date": "2026-08-31T08:37:03+10:00", "subject": "Move SSL3_RT_MAX_PLAIN_LENGTH to prov_ssl.h", "body": "Reviewed-by: Simo Sorce \nReviewed-by: Shane Lontis \nMerge-date: Thu Sep 3 02:17:16 2026\nMerged-from: https://github.com/openssl/openssl/pull/32536", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "512becd155f18d347184690a95a82f3df1645b72", "date": "2026-08-31T08:40:27+10:00", "subject": "aes-cbc-hmac-sha: don't include ssl3.h since the define is in prov_ssl.h now", "body": "Reviewed-by: Simo Sorce \nReviewed-by: Shane Lontis \nMerge-date: Thu Sep 3 02:17:20 2026\nMerged-from: https://github.com/openssl/openssl/pull/32536", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "84f66fa687c24f691802de39a35a5fb0ed0a4e6e", "date": "2026-08-30T18:42:59-04:00", "subject": "build(vim-patch): v9.0.1557 is n/a", "body": "Global variables with \"_for_testing\" suffix are modified\nin \"f_test_override()\", N/A to Nvim.\nIf a test needs them, then port the relevant code outside\n\"f_test_override()\" via LuaJIT C FFI.", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "nestjs", "sha": "b763af173694d620e3f233dc0fc8f39e2e7b7326", "date": "2026-08-30T19:47:06-03:00", "subject": "refactor(ws): extract the default message parser to a module constant", "body": "Moving it out of the field initialiser gives the adapter a stable reference to\ncompare against, so it can tell whether the parser in use is still the built-in\none. Comparing the reference rather than tracking a flag covers every way the\nparser can be replaced: the constructor option, the setter, and a subclass\nassigning the protected field directly.\n\nPure refactor, no behaviour change.\n\nRefs #17624", "style": "conventional", "type": "refactor", "scope": "ws", "breaking": false} {"repo": "nestjs", "sha": "39fbddae51281ca56bf2fed9123d98e61509066c", "date": "2026-08-30T19:47:29-03:00", "subject": "fix(ws): stop discarding parser and handler lookup failures silently", "body": "The message handler binding wrapped its whole body in a single try with an\nempty catch, so three unrelated failures ended up indistinguishable from a\ndropped packet: a custom message parser that throws, a malformed frame, and an\nevent with no registered handler.\n\nThe handler lookup used a non-null assertion followed by a destructure, so an\nunrecognised event reached that catch as a TypeError. It is now an explicit\nguard that keeps the single map lookup introduced in 8c7718e38.\n\nA custom parser that throws is now reported, since nothing else in the process\nwould reveal it. The default parser stays silent because it throws on\nclient-controlled input, and reporting that would turn a public socket into a\nlog flood target.\n\nThe catch around the handler invocation is kept on purpose: without it a\nsynchronous throw propagates through the mergeMap that binds the client and\ntears down that client's source stream, silencing every later message from it.\n\nCloses #17624", "style": "conventional", "type": "fix", "scope": "ws", "breaking": false} {"repo": "openssl", "sha": "c3fa1250b5cff20554e5be9aa6346e665edfc2bc", "date": "2026-08-31T08:50:27+10:00", "subject": "changes: add note about finishing cipher OSSL_PARAM parsers", "body": "Reviewed-by: Simo Sorce \nReviewed-by: Shane Lontis \nMerge-date: Thu Sep 3 02:17:25 2026\nMerged-from: https://github.com/openssl/openssl/pull/32536", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "f190aa8ecea728e6ce7fb7a6250e03df4d4eaca5", "date": "2026-08-31T01:00:33+02:00", "subject": "lib: drop unused `system_win32.h` includes", "body": "Closes #22757", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "bb3adbe52e04e07362a6c0bf4375c16d14e12144", "date": "2026-08-30T19:14:21-04:00", "subject": "build(vim-patch): v9.0.1994 is n/a", "body": "", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "neovim", "sha": "b542f3a8f8689a9a033411456fb1c6576916c66b", "date": "2026-08-30T19:38:34-04:00", "subject": "fix(cmdatom): edit-repeat mapping #41564", "body": "Problem:\nThe \".\" example mapping at `:h edit-repeat` doesn't work well with\n`nvim_feedkeys(…, 'mt', false)`.\n\nSolution:\nUse `vim.b[ev.buf].maxseq` instead of `undotree()`.", "style": "conventional", "type": "fix", "scope": "cmdatom", "breaking": false} {"repo": "neovim", "sha": "9d51351218393e4ba0c81673a0a6dec70d19fbf0", "date": "2026-08-30T19:46:50-04:00", "subject": "build(vim-patch): 2ca7d5f4831ce3d86cdc3f0c844fc50ea0deb7e5 is n/a", "body": "v9.1.1029 removes runtime/doc/uganda.rux.", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "openssl", "sha": "66e691b18629396285d6f78a8b1b1eae36b92ebb", "date": "2026-08-31T09:28:19+09:00", "subject": "test: use the portable null device in Perl recipes", "body": "The symbol presence recipe hardcodes /dev/null in commands that run\nthrough the shell. On native Windows, cmd.exe treats that as\n:\\dev\\null. If the parent directory is absent, redirection fails\nbefore nm starts and the test can pass with empty symbol lists.\n\nUse File::Spec's null device for those commands. GNU nm still doesn't\ndistinguish mergeable COFF definitions and can't read dynamic symbols\nfrom Microsoft DLLs. Skip this developer test under native Windows\nPerl instead of treating its output as reliable.\n\nThe X.509 recipe has a related negative test. Failure to open /dev/null\ncan satisfy its assertion before invalid extension handling is tested.\nUse File::Spec's null device so the command fails for the intended\nreason.\n\nAssisted-by: codex:gpt-5.6-sol\nReviewed-by: Richard Levitte \nReviewed-by: Bob Beck \nReviewed-by: Tomas Mraz \nMerge-date: Fri Sep 4 08:11:49 2026\nMerged-from: https://github.com/openssl/openssl/pull/32604\n(cherry picked from commit cf206ac07a6f4d31fadd017b7d2e90af6074acae)", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "openssl", "sha": "cf206ac07a6f4d31fadd017b7d2e90af6074acae", "date": "2026-08-31T09:28:19+09:00", "subject": "test: use the portable null device in Perl recipes", "body": "The symbol presence recipe hardcodes /dev/null in commands that run\nthrough the shell. On native Windows, cmd.exe treats that as\n:\\dev\\null. If the parent directory is absent, redirection fails\nbefore nm starts and the test can pass with empty symbol lists.\n\nUse File::Spec's null device for those commands. GNU nm still doesn't\ndistinguish mergeable COFF definitions and can't read dynamic symbols\nfrom Microsoft DLLs. Skip this developer test under native Windows\nPerl instead of treating its output as reliable.\n\nThe X.509 recipe has a related negative test. Failure to open /dev/null\ncan satisfy its assertion before invalid extension handling is tested.\nUse File::Spec's null device so the command fails for the intended\nreason.\n\nAssisted-by: codex:gpt-5.6-sol\nReviewed-by: Richard Levitte \nReviewed-by: Bob Beck \nReviewed-by: Tomas Mraz \nMerge-date: Fri Sep 4 08:11:49 2026\nMerged-from: https://github.com/openssl/openssl/pull/32604", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "c9b32d17657fa257f2564dec3da7059d0cf96e34", "date": "2026-08-31T09:20:11+08:00", "subject": "types(runtime-vapor): support async setup in defineVaporComponent", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "169af351ca91aaf53b7126bed93d26076a90cea1", "date": "2026-08-31T09:20:47+08:00", "subject": "types(runtime-vapor): infer exposed type from options setup", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "f5a92eef63c0b59df3b901029f83377ea4e8e355", "date": "2026-08-31T09:23:53+08:00", "subject": "fix(runtime-vapor): reuse SSR v-if anchors in dynamic elements", "body": "close #15371", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "3495d0bdc098a7e8c6269c967b12dcba5ffea7c2", "date": "2026-08-31T09:25:09+08:00", "subject": "fix(compiler-vapor): drop comment children of Transition/TransitionGroup", "body": "close #15372", "style": "conventional", "type": "fix", "scope": "compiler-vapor", "breaking": false} {"repo": "openssl", "sha": "45368a61f74842c36bd2499e921b2f49163b6763", "date": "2026-08-31T12:18:41+10:00", "subject": "ec_params: add include guard to crypto/ec_params.h.in", "body": "Assisted-by: ChatGPT:gpt-5.6Sol\n\nReviewed-by: Shane Lontis \nReviewed-by: Mounir Idrassi \n(Merged from https://github.com/openssl/openssl/pull/32502)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "c24270bc1cbf88062263c7a067469bdb905534af", "date": "2026-08-30T19:59:18-07:00", "subject": "wcurl: import v2026.08.30", "body": "Closes #22758", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "da0854b984ee3bf6827dd332843622f98406f118", "date": "2026-08-31T13:09:17+10:00", "subject": "ec: apply group options to explicit parameters", "body": "Apply encoding and point-conversion options after constructing an EC group from explicit parameters, and add regression coverage for accepted and invalid option values.\n\nAssisted-by: ChatGPT:gpt-5.6Sol\n\nReviewed-by: Shane Lontis \nReviewed-by: Mounir Idrassi \n(Merged from https://github.com/openssl/openssl/pull/32502)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "173480905e8556188884d1ca938545cabe798952", "date": "2026-08-31T08:39:24+05:30", "subject": "aes-ctr-avx512: add VAES-512 AES-CTR encryption", "body": "Add a VAES-512 accelerated AES-CTR encryption path built on the shared\n512-bit primitives. Counter blocks are generated and encrypted 4/8/16 at\na time; the low 64-bit counter word is clamped so that a 64-bit carry is\nhandled by a scalar step before VAES resumes, and payloads below the\nprovider threshold fall back to the existing aesni CTR path.\n\nThe provider gains a CTR hook that routes eligible payloads to the VAES\nimplementation, and aes_platform.h exposes VAES_CTR_ELIGIBLE (aliased to\nVAES512_ELIGIBLE) with the ossl_aes_ctr_vaes prototypes.\n\nThe existing CTR known-answer vectors were all below the provider threshold\nthat dispatches to the VAES path, so that code was never exercised. Add\nNIST SP800-38A 64-byte CTR KATs (AES-128/192/256) and a parametrized C test\nspanning 64B-5000B across all key sizes and counter-carry boundaries,\nvalidated against an independent ECB-based CTR reference.\n\nReviewed-by: Simo Sorce \nReviewed-by: Nikola Pajkovsky \nMerge-date: Thu Sep 17 16:26:00 2026\nMerged-from: https://github.com/openssl/openssl/pull/30755", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "138bea88eb25e88da2e2fbadc62d7a221ed31acf", "date": "2026-08-31T08:39:24+05:30", "subject": "aes: factor out shared VAES-512 primitives", "body": "The VAES-512 AES implementation file will host more than one mode. Rename\nit to a mode-neutral name ahead of adding the AES-CTR implementation and\nfactoring out the shared VAES-512 primitives. This commit is a pure rename\nwith no content change.\n\nRefactor the VAES-512 CBC decryption code to sit on top of a small set of\nshared 512-bit primitives so that additional AES modes can reuse them:\n\n - portable compiler abstractions for ISA targeting and inlining,\n - OSSL_VAES512_DEFINE_ROUNDS, a macro emitting the 1x/2x/4x parallel\n 512-bit AES round functions for any round count and direction,\n - OSSL_VAES512_LOAD_ROUNDKEYS / OSSL_VAES512_CLEAR_ROUNDKEYS helpers,\n - ossl_vaes512_cpu_capable(), a shared runtime capability check.\n\nThe generated machine code for CBC decryption is byte-identical to the\nprevious version. aes_platform.h grows a VAES512_ELIGIBLE base guard with\nVAES_CBC_ELIGIBLE kept as an alias.\n\nRename the assembly cleanup helper and perlasm module from CBC-specific\nnames to shared aes-vaes symbols so additional modes can reuse the same\nstack scrub routine.\n\nReviewed-by: Simo Sorce \nReviewed-by: Nikola Pajkovsky \nMerge-date: Thu Sep 17 16:25:59 2026\nMerged-from: https://github.com/openssl/openssl/pull/30755", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "e034dbadecfd6ce141d616c6e98d477a99653ffd", "date": "2026-08-30T23:28:02-04:00", "subject": "build(vim-patch): n/a 5400a5d4269874fe4f1c35dfdd3c039ea17dfd62", "body": "comment,editorconfig plugins in Vim's runtime/pack/, are N/A.\nEditorconfig is unrelated but Vim did not fix the bundled version\nfrom editorconfig-vim and did not port the upstream fix.\n\nhttps://github.com/editorconfig/editorconfig-vim/commit/0d54ea863089fb13be423b4aed6cca35f3a5d778", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "openssl", "sha": "862229aa7928a7ddd8404e38f0131a865c28c151", "date": "2026-08-31T13:37:54+10:00", "subject": "fips: removal of ec_kmgmt_imexport.inc requires metadata updates", "body": "Remove the obsolete include and its entries from the module sources and checksum files.\n\nAssisted-by: ChatGPT:gpt-5.6Sol\n\nReviewed-by: Shane Lontis \nReviewed-by: Mounir Idrassi \n(Merged from https://github.com/openssl/openssl/pull/32502)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "0e944fe3579ecd525e5ffe7b7db4a7e1c34de49e", "date": "2026-08-31T12:51:14+09:00", "subject": "test_extensible: Test module for extensible nodes and custom scans", "body": "This commit adds a new test module that implements a CustomScan based on\nan ExtensibleNode. These APIs exist in the core code for ages but had\nzero coverage. This now covers the read, copy and out routines for\nextensible nodes, with SQL function wrappers for each. CustomScan nodes\nare tested with callbacks to manipulate the executor, based on the idea\nof repeating the number of times each row is returned in a single scan,\nusing a GUC (default at 2 tuples for each row, can be tweaked).\n\nThe code of this module is primitive to keep its logic simpler. For\nexample, custom scans happen only when matching one specific, hardcoded\nrelation name, and work only for base relations (this can be extended at\nwill, as required).\n\nThis module can be used as a template for developers willing to develop\ntheir own extensions. \"Writing a Custom Scan Provider\" covers the\ndocumentation for custom scans.\n\nAuthor: Aleksander Alekseev \nReviewed-by: Jan Nidzwetzki \nReviewed-by: Michael Paquier \nReviewed-by: Rafia Sabih \nDiscussion: https://postgr.es/m/CAJ7c6TNfn9Fv_Je1etA6rrgq1onVvXbjwBTkbkd4kVQhcu11gg%40mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "0892db24f5e8fd1004fadf851f836b0e56f28dff", "date": "2026-08-31T07:10:16+02:00", "subject": "providers: check OSSL_DECODER_CTX_set_construct_data() return value", "body": "It may return a failure if the decoder is frozen, and its return value\nis checked elsewhere, as was pointed out by Coverity. Handle the return\nvalue.\n\n[ Only changes to file_store.c are applied, as winstore_store.c is not\n present in openssl-3.0, by virtue of winstore having been added during\n openssl-3.2 development window. ]\n\nResolves: https://scan5.scan.coverity.com/#/project-view/65248/10222?selectedIssue=1700562\nComplements: 63f187cfedd2 \"STORE: Add a built-in 'file:' storemgmt implementation (loader)\"\nSigned-off-by: Norbert Pocs \nReviewed-by: Paul Dale \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Neil Horman \nReviewed-by: Eugene Syromiatnikov \nMerge-date: Mon Aug 31 05:19:13 2026\nMerged-from: https://github.com/openssl/openssl/pull/32576\n(cherry picked from commit acde7bc05a9979abce12c4f92d2a0aa0a3d61175)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "c5453a0ae9e1e440039c886d8f81ee932aed083e", "date": "2026-08-31T07:10:16+02:00", "subject": "providers: check OSSL_DECODER_CTX_set_construct_data() return value", "body": "It may return a failure if the decoder is frozen, and its return value\nis checked elsewhere, as was pointed out by Coverity. Handle the return\nvalue.\n\nResolves: https://scan5.scan.coverity.com/#/project-view/65248/10222?selectedIssue=1700562\nComplements: 63f187cfedd2 \"STORE: Add a built-in 'file:' storemgmt implementation (loader)\"\nComplements: 606e0426a148 \"Add support for loading root CAs from Windows crypto API\"\nSigned-off-by: Norbert Pocs \nReviewed-by: Paul Dale \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Neil Horman \nReviewed-by: Eugene Syromiatnikov \nMerge-date: Mon Aug 31 05:19:13 2026\nMerged-from: https://github.com/openssl/openssl/pull/32576\n(cherry picked from commit acde7bc05a9979abce12c4f92d2a0aa0a3d61175)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "acde7bc05a9979abce12c4f92d2a0aa0a3d61175", "date": "2026-08-31T07:10:16+02:00", "subject": "providers: check OSSL_DECODER_CTX_set_construct_data() return value", "body": "It may return a failure if the decoder is frozen, and its return value\nis checked elsewhere, as was pointed out by Coverity. Handle the return\nvalue.\n\nResolves: https://scan5.scan.coverity.com/#/project-view/65248/10222?selectedIssue=1700562\nComplements: 63f187cfedd2 \"STORE: Add a built-in 'file:' storemgmt implementation (loader)\"\nComplements: 606e0426a148 \"Add support for loading root CAs from Windows crypto API\"\nSigned-off-by: Norbert Pocs \nReviewed-by: Paul Dale \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Neil Horman \nReviewed-by: Eugene Syromiatnikov \nMerge-date: Mon Aug 31 05:19:13 2026\nMerged-from: https://github.com/openssl/openssl/pull/32576", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "eb0e147e0798ea61ccd125004dabf37ffe750aa1", "date": "2026-08-31T09:02:31+03:00", "subject": "refactor(docs-infra): remove two more unreferenced guide images", "body": "`input.svg` and `output.svg` are leftovers that #70335 missed. Their\nlast reference was deleted by #54829 when the AIO guides were removed,\nand nothing in the repo mentions them or their directory. Removing both\nempties `adev/src/assets/images/guide/inputs-outputs`.\n\n(cherry picked from commit a92f6057d3e02436010e246f9e76bcc8a227039a)", "style": "conventional", "type": "refactor", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "a92f6057d3e02436010e246f9e76bcc8a227039a", "date": "2026-08-31T09:02:31+03:00", "subject": "refactor(docs-infra): remove two more unreferenced guide images", "body": "`input.svg` and `output.svg` are leftovers that #70335 missed. Their\nlast reference was deleted by #54829 when the AIO guides were removed,\nand nothing in the repo mentions them or their directory. Removing both\nempties `adev/src/assets/images/guide/inputs-outputs`.", "style": "conventional", "type": "refactor", "scope": "docs-infra", "breaking": false} {"repo": "ant-design", "sha": "6ffc83184c2621cf140cb59a1754816de9277c66", "date": "2026-08-31T14:05:11+08:00", "subject": "fix(Tooltip): correct alignment near viewport boundaries", "body": "", "style": "conventional", "type": "fix", "scope": "Tooltip", "breaking": false} {"repo": "vue-core", "sha": "8ac7525e609a73833892babec424c6def7695d47", "date": "2026-08-31T14:12:45+08:00", "subject": "fix(runtime-vapor): avoid rebinding fired once handlers in dynamic props", "body": "Closes #15378", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "tmux", "sha": "39357c829c62c1e0c0af6e7a8715b0073c65e8e9", "date": "2026-08-31T07:41:48Z", "subject": "Clear stale marks when a search regex fails to compile, GitHub issue 5547 from Uzair Aftab.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "6a1ecdcce34cd2682b3bf06b14a10ce4f24c8165", "date": "2026-08-31T07:42:56Z", "subject": "Do not try to append zero size cells, from Uzair Aftab.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "602d215ae50d8b8f93ddbaacac1f9f139cfe4f20", "date": "2026-08-31T10:42:56+03:00", "subject": "fix(ui): add suite level report view", "body": "", "style": "conventional", "type": "fix", "scope": "ui", "breaking": false} {"repo": "tmux", "sha": "c552d1bf167a240d28f82b4d78a604a12da3326a", "date": "2026-08-31T07:44:39Z", "subject": "Layout cell can be replaced by unzoom, fixes use after free. From Uzair Aftab.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "af3e4d2e5b63fa2db14834b17b20b40771744b91", "date": "2026-08-31T07:46:55Z", "subject": "Do not allow popups for control clients, from Noam Stolero.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "8b3834c34fd54df0be8798ad6412aab7d58d118d", "date": "2026-08-31T07:51:56Z", "subject": "Do not leak temporary commands when expanding aliases, from Jeong, Heon.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "0f48cf3d7f17b11a131cda4ba870612d4ac96b66", "date": "2026-08-31T10:02:57+02:00", "subject": "conncache: guess maxconnects different", "body": "Take the number of attached transfers instead of the running\nones as guidance how many connections the multi should cache.\n\nIn addition, when guessing, only evict a connection if it has\nnot been in use for a second or longer.\n\nFollow-up to 8a2bb9ca241bbd82a0da536f6\n\nCloses #22759", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "092b42879ae8cdbb672c92896b6295af02bc0e16", "date": "2026-08-31T16:35:37+08:00", "subject": "fix(runtime-vapor): avoid re-registering fired once listeners", "body": "Revert #15383 and re-fix #15378", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "71207256955b4f7c4147c2a28ab22664e5e0748e", "date": "2026-08-31T17:11:32+08:00", "subject": "fix(runtime-vapor): scope each teleport children render so slot re-runs stop stale effects", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "fb78bb0388760c884c58ba69f24909be561416d6", "date": "2026-08-31T17:11:32+08:00", "subject": "fix(runtime-vapor): buffer deferred teleport target mounts on the suspense boundary", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "curl", "sha": "73f18365ab8e47912afcd5ee9404c7493d9d0fb2", "date": "2026-08-31T11:24:39+02:00", "subject": "GHA/non-native: disable riscv64 FreeBSD job due to upstream issue", "body": "CI jobs fail early with these errors:\n```\npkg: https://cross-platform-actions.github.io/freebsd-pkg-repo/FreeBSD:15:riscv64/meta.txz: Not Found\npkg: https://cross-platform-actions.github.io/freebsd-pkg-repo/FreeBSD:15:riscv64/data.pkg: Not Found\npkg: https://cross-platform-actions.github.io/freebsd-pkg-repo/FreeBSD:15:riscv64/data.tzst: Not Found\npkg: https://cross-platform-actions.github.io/freebsd-pkg-repo/FreeBSD:15:riscv64/packagesite.pkg: Not Found\npkg: https://cross-platform-actions.github.io/freebsd-pkg-repo/FreeBSD:15:riscv64/packagesite.tzst: Not Found\n```\nRef: https://github.com/curl/curl/actions/runs/33252993091/job/99101785035#step:4:35 (first seen)\n\nRef: https://github.com/cross-platform-actions/action/issues/166\nFollow-up to 9f05091b596e3015bf214ea15f6c56f6eb7647f6 #22748\n\nCloses #22760", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "c5dbf3f09b406586d5ce6ce0a3d634d1a07b4f04", "date": "2026-08-31T17:57:30+08:00", "subject": "docs: fix i18n contribution guide link", "body": "* docs: fix i18n contribution guide link\n\n* chore: rerun ci\n\n---------", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vitest", "sha": "dc10f5f8fb69b026afcf56f8a7a9dd36bd964d73", "date": "2026-08-31T12:22:33+02:00", "subject": "fix(browser): report the action error when a task times out", "body": "", "style": "conventional", "type": "fix", "scope": "browser", "breaking": false} {"repo": "tmux", "sha": "e9634d40749a5ae330aabf5aa46a81505b094a6b", "date": "2026-08-31T11:25:09+01:00", "subject": "Set used for popup ranges.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "ab00fc4a45add3600d693f329bc6c68ac4c25def", "date": "2026-08-31T12:26:26+02:00", "subject": "clock: save one call", "body": "When multi is processing a transfer, it assesses its timeleft and does a\nprogress update repeatedly. Both got a fresh timestamp, but can use the\nsame one. Change implementation for that.\n\nCloses #22761", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "58f02ae3663435854902ca8c2de049e75c4610f9", "date": "2026-08-31T19:27:17+09:00", "subject": "feat(ui): add link to open playwright trace", "body": "", "style": "conventional", "type": "feat", "scope": "ui", "breaking": false} {"repo": "vitest", "sha": "2122ffdfb42d86c9a2f1238100d6a031312cae32", "date": "2026-08-31T15:59:46+05:30", "subject": "fix: propagate --maxWorkers to projects", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "b42812e3c1875456b18c077ec620e3de9b1d088f", "date": "2026-08-31T18:33:59+08:00", "subject": "fix(Slider): keep edge tooltips within viewport", "body": "", "style": "conventional", "type": "fix", "scope": "Slider", "breaking": false} {"repo": "vitest", "sha": "51e9494166d8d0108a621fce80c41b6dba760bae", "date": "2026-08-31T12:36:10+02:00", "subject": "feat!: parse files statically in vitest list by default", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": true} {"repo": "ant-design", "sha": "3c356d47c07a1067ccde0ee6773590e1aafd98d2", "date": "2026-08-31T18:37:03+08:00", "subject": "test(Slider): update tooltip arrow snapshots", "body": "", "style": "conventional", "type": "test", "scope": "Slider", "breaking": false} {"repo": "neovim", "sha": "ad42ee1c41eb3fb2b828eba0e52752446b5901ed", "date": "2026-08-31T16:22:20+05:30", "subject": "fix(cmdwin): handle UTF-8 characters containing 0x80 #41566", "body": "Problem:\nConfirming cmdwin with a UTF-8 character containing 0x80 does not complete the\ncommand.\n\nSolution:\nEscape K_SPECIAL bytes while feeding the cmdwin input after confirmation.", "style": "conventional", "type": "fix", "scope": "cmdwin", "breaking": false} {"repo": "curl", "sha": "4ab4e3f846a93ec58621d05cc6fee024e66642f3", "date": "2026-08-31T12:53:16+02:00", "subject": "strings: add length to CURL_EASY_STR_SET()", "body": "Since we check the input lengths of strings before setting, pass the\nknown length to the hash set.\n\nCloses #22762", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "897f51fd2493046c52ec9539b7d02fe3763bd63e", "date": "2026-08-31T14:05:15+03:00", "subject": "chore: release v5.0.0-rc.4", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "vitest", "sha": "c4093fdeb87fe6a3066f6c6d637129a031802e94", "date": "2026-08-31T13:14:48+02:00", "subject": "ci: cancel ci when pr is merged or closed", "body": "", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "angular", "sha": "5cf7764e1b1479f7854563609d87cb1196d576ea", "date": "2026-08-31T14:37:11+03:00", "subject": "docs: close the unclosed docs-step tags in two guides", "body": "`reactive-forms.md` and `app-shell.md` each open a `` that is\nnever closed. The tokenizers match up to the *next* closing tag instead\nof failing, so the unclosed tag pulls in the content that follows and\nleaves the block after it unparsed in the output.\n\nOn https://angular.dev/guide/forms/reactive-forms the \"Creating nested\nform groups\" section shows raw source: \"To create a nested group in\n`profileForm`, add a nested `address` element\", backticks and all. Its\nheading is swallowed into the unclosed step instead of being an h3, and\nthe page renders 20 of its 21 steps. On\nhttps://angular.dev/ecosystem/service-workers/app-shell the third step\nis missing entirely.\n\n`reactive-forms.md` has been broken since #52536, `app-shell.md` since\n#55550.\n\n(cherry picked from commit 0752282c8f1fe50a7f71bc0c9a2a649343227a59)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "0752282c8f1fe50a7f71bc0c9a2a649343227a59", "date": "2026-08-31T14:37:11+03:00", "subject": "docs: close the unclosed docs-step tags in two guides", "body": "`reactive-forms.md` and `app-shell.md` each open a `` that is\nnever closed. The tokenizers match up to the *next* closing tag instead\nof failing, so the unclosed tag pulls in the content that follows and\nleaves the block after it unparsed in the output.\n\nOn https://angular.dev/guide/forms/reactive-forms the \"Creating nested\nform groups\" section shows raw source: \"To create a nested group in\n`profileForm`, add a nested `address` element\", backticks and all. Its\nheading is swallowed into the unclosed step instead of being an h3, and\nthe page renders 20 of its 21 steps. On\nhttps://angular.dev/ecosystem/service-workers/app-shell the third step\nis missing entirely.\n\n`reactive-forms.md` has been broken since #52536, `app-shell.md` since\n#55550.", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "341bc0deb7fc5d88bf955dc8680c415999e734d7", "date": "2026-08-31T14:37:16+03:00", "subject": "fix(docs-infra): reject unclosed paired docs elements", "body": "An unclosed `` or `` produces no error, just wrong\noutput, so the two cases fixed in the previous commit went unnoticed for\nyears.\n\nValidate that the paired custom elements balance before parsing. A\nmismatch now names the file and the counts instead of silently\nswallowing a section.\n\n(cherry picked from commit d9afca095b42f03779bdf1ed687d90908eae8a81)", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "d9afca095b42f03779bdf1ed687d90908eae8a81", "date": "2026-08-31T14:37:16+03:00", "subject": "fix(docs-infra): reject unclosed paired docs elements", "body": "An unclosed `` or `` produces no error, just wrong\noutput, so the two cases fixed in the previous commit went unnoticed for\nyears.\n\nValidate that the paired custom elements balance before parsing. A\nmismatch now names the file and the counts instead of silently\nswallowing a section.", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "curl", "sha": "779b84936b293d5983fd4d24b1472d59346fc6d5", "date": "2026-08-31T13:42:44+02:00", "subject": "build: drop `HTTP_ONLY` option", "body": "Drop `HTTP_ONLY` C macro for raw builds, and `HTTP_ONLY=ON` for CMake.\n\nIt's a niche option that's been historically incomplete throughout its\nexistence, also a maintenance pain to get it right and synced with\nprotocols and with each other. Note there was no matching `./configure`\noption either. At the time of this commit, it failed to disable SCP/SFTP\nif CMake found libssh/libssh2 (the default), and C flags missed to\ndisable IPFS in addition.\n\nAlso:\n- drop from AppVeyor CI build mix.\n- GHA/macos: replace with individual disable options.\n\nFollow-up to 4c5307b45655ba75ab066564afdc0c111a8b9291\nFollow-up to 08ef208fb78fb2eabc5cec08c23e74e251eac898\nCloses #22744\nFixes #22740\n\nCloses #22764", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "openssl", "sha": "689fb0a3847076199289b9099f9032e5045e23dd", "date": "2026-08-31T20:45:05+09:00", "subject": "Fix DTLS 1.3 accepting unauthenticated plaintext alerts", "body": "DTLS 1.3 accepted forged DTLSPlaintext records after traffic keys\nwere installed. Injected alerts could terminate or desynchronise an\nassociation, while plaintext handshake and ACK records could invoke\nWPACKET_cleanup() on an uninitialised WPACKET.\n\nSilently discard DTLSPlaintext in nonzero epochs, as required by RFC\n9147, after consuming the complete fragment. Harden tls13_cipher() by\nhonouring allow_plain_alerts, rejecting non-unified DTLS headers, and\nensuring WPACKET is initialised before cleanup.\n\nKeep validate_record_header unset because invalid DTLS records must be\nsilently discarded rather than converted into fatal alerts.\n\nAssisted-by: pi:kimi-k3\nReviewed-by: Ryan Hooper \nReviewed-by: Frederik Wedel-Heinen \nMerge-date: Mon Sep 7 14:41:16 2026\nMerged-from: https://github.com/openssl/openssl/pull/32622", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "9b2caf34f54f5eb2f26d271c2b2cdea71fa1cf40", "date": "2026-08-31T13:46:46+02:00", "subject": "transfer: DID handling", "body": "When a transfer leaves its DID state and goes to PERFORM, it should\nnot hog the multi processing when there are other transfers alive.\nMake multistate_did() return CURLM_OK then which triggers the\nprocessing of the next easy handle.\n\nThis gives more fairness among easy handles:\n- on a fast server, responses may arrive very quickly and processing\n a response is costly. All this time is then lost to other transfers\n that have not even sent their request\n- on a slow server, directly trying to receive the response will not\n work and is an unnecessary network call.\n\nLocal scorecard tests with h1 requests showed that the former handling\nwas only using 10 connections for 40 parallel requests after an\ninitial peak of 40 connections. The change makes libcurl use all 40\nconnections continuously.\n\nThis also explains the h1-requests memory use development. When we\nclosed connections more aggressively, the 40 initial connections\nshrank and released their memory. But the parallelism was not good.\n\nCloses #22763", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "3f2a5bb05fd4c67814214b308cb1444b500aec10", "date": "2026-08-31T13:57:45+02:00", "subject": "curl_setup.h: merge two `#if` blocks", "body": "Also: alpha-sort disable options within the block.\n\nCloses #22765", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "f2bfd611e25dbc3ab39164b2ac93c0aeae9f5f75", "date": "2026-08-31T12:41:03Z", "subject": "tmux: add utf8 terminal-features", "body": "OK: nicm@", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "c275b5de5a835276d3e25b7491fd0f0664295ee9", "date": "2026-08-31T20:41:19+08:00", "subject": "fix(lsp): separate adjacent nested folding ranges #41428", "body": "Problem:\nFolding range markers are overwritten while ranges are evaluated. A range ending on a row can hide another range starting there, and multiple nested ranges ending together emit only the innermost ending level.\n\nSolution:\nTrack starts and the number of ends per row before emitting markers. Prefer starts on shared boundary rows and use the outermost level when nested ranges end together.\n\nAI-assisted", "style": "conventional", "type": "fix", "scope": "lsp", "breaking": false} {"repo": "openssl", "sha": "e8d7ba81af8bfc0212be496aa21f0e3f4b8e266b", "date": "2026-08-31T15:09:25+02:00", "subject": "CMS_verify.pod,openssl-cms.pod.in: fix doc mistakes w.r.t. CMS_VERIFY_PARTIAL introduced by PR #27604", "body": "Reviewed-by: Mounir Idrassi \nReviewed-by: Dmitry Belyavskiy \nReviewed-by: Eugene Syromiatnikov \nMerge-date: Thu Sep 3 09:26:57 2026\nMerged-from: https://github.com/openssl/openssl/pull/22285", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "3d3a147a814ec8fd62c3080247faf17320714e6a", "date": "2026-08-31T22:34:51+09:00", "subject": "test: DTLS 1.3 forged DTLSPlaintext alert regression tests", "body": "Add regression coverage for post-handshake and buffered mid-handshake\nforged DTLSPlaintext alerts, including shutdown, replay-window,\nframing, malformed, and overlong cases. Also cover plaintext handshake\nand ACK records. Verify keyed epochs silently discard every record\nwhile application traffic continues.\n\nKeep an epoch-0 control proving legitimate plaintext alerts are still\ndelivered. Change TLSProxy's DTLS 1.3 teardown close_notify to\nDTLSCiphertext using p_ossltest's passthrough AEAD; its former\nDTLSPlaintext record is now correctly discarded and caused proxy\nrecipes to hang.\n\nAll forged-record cases fail without the record-layer fix.\n\nAssisted-by: pi:kimi-k3\nReviewed-by: Ryan Hooper \nReviewed-by: Frederik Wedel-Heinen \nMerge-date: Mon Sep 7 14:41:17 2026\nMerged-from: https://github.com/openssl/openssl/pull/32622", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "angular", "sha": "120aac91089f8f2665013e8f3ecd9b31ce6d1828", "date": "2026-08-31T16:12:52+02:00", "subject": "fix(vscode-extension): handle escaped delimiters in inline template and styles highlighting", "body": "The TextMate grammars for inline templates and inline styles ended the\nstring at the first occurrence of the delimiter, so an escaped delimiter\n(e.g. \\`) terminated the highlighting prematurely and the rest of the\ntemplate was no longer highlighted as HTML.\n\nEscape sequences are now consumed before delegating to the HTML/CSS\ngrammars, so escaped delimiters no longer end the string. Also removes\nthe stray pipes from the string delimiter character class, which\nunintentionally matched a literal '|'.\n\nFixes #65493\n\n(cherry picked from commit 1d9e445d1a9200e58c22e04ac2bc586b8607d1f4)", "style": "conventional", "type": "fix", "scope": "vscode-extension", "breaking": false} {"repo": "angular", "sha": "1d9e445d1a9200e58c22e04ac2bc586b8607d1f4", "date": "2026-08-31T16:12:52+02:00", "subject": "fix(vscode-extension): handle escaped delimiters in inline template and styles highlighting", "body": "The TextMate grammars for inline templates and inline styles ended the\nstring at the first occurrence of the delimiter, so an escaped delimiter\n(e.g. \\`) terminated the highlighting prematurely and the rest of the\ntemplate was no longer highlighted as HTML.\n\nEscape sequences are now consumed before delegating to the HTML/CSS\ngrammars, so escaped delimiters no longer end the string. Also removes\nthe stray pipes from the string delimiter character class, which\nunintentionally matched a literal '|'.\n\nFixes #65493", "style": "conventional", "type": "fix", "scope": "vscode-extension", "breaking": false} {"repo": "angular", "sha": "c3e51011155cda9ab15467c16d40257bd95e4378", "date": "2026-08-31T17:14:03+03:00", "subject": "fix(docs-infra): parse docs-callout attributes correctly", "body": "Three ways a callout could be misparsed:\n\n- A title quoted with `'` or a backtick was dropped, leaving an empty\n heading. Two callouts lose their title on angular.dev today, on\n guide/forms/template-driven-forms and guide/i18n/prepare. The first has\n to use single quotes because its title contains `\"pristine\"`.\n- A title containing `>` was dropped, because the attribute capture\n stopped at the first `>` even inside a quoted value.\n- The severity was matched anywhere in the tag, so a title such as\n \"Why this is important\" silently rendered an important callout.\n\nScan attributes with quoting in mind, accept all three quote characters\nas #69268 did for docs-code-block, and match the severity flags against\nthe tag with attribute values removed. The i18n callout also spelled the\nattribute `header`, which the extension has never read.\n\n(cherry picked from commit b70edd2768e587b83af09717f4501d1e52d2a950)", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "b70edd2768e587b83af09717f4501d1e52d2a950", "date": "2026-08-31T17:14:03+03:00", "subject": "fix(docs-infra): parse docs-callout attributes correctly", "body": "Three ways a callout could be misparsed:\n\n- A title quoted with `'` or a backtick was dropped, leaving an empty\n heading. Two callouts lose their title on angular.dev today, on\n guide/forms/template-driven-forms and guide/i18n/prepare. The first has\n to use single quotes because its title contains `\"pristine\"`.\n- A title containing `>` was dropped, because the attribute capture\n stopped at the first `>` even inside a quoted value.\n- The severity was matched anywhere in the tag, so a title such as\n \"Why this is important\" silently rendered an important callout.\n\nScan attributes with quoting in mind, accept all three quote characters\nas #69268 did for docs-code-block, and match the severity flags against\nthe tag with attribute values removed. The i18n callout also spelled the\nattribute `header`, which the extension has never read.", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "nestjs", "sha": "783d539898cb3f3c45c6662add7813c02a0f67f1", "date": "2026-08-31T23:50:53+09:00", "subject": "refactor(core): extract duplicate debug-mode check into shared helper", "body": "", "style": "conventional", "type": "refactor", "scope": "core", "breaking": false} {"repo": "openssl", "sha": "1edb85330e6f7add4a2c3615d0a2a1d999980646", "date": "2026-08-31T22:13:34+07:00", "subject": "Migrate script_55 to script_59", "body": "Assisted-by: Claude:claude-sonnet-5\nReviewed-by: Neil Horman \nReviewed-by: Norbert Pocs \nMerge-date: Tue Sep 15 08:55:48 2026\nMerged-from: https://github.com/openssl/openssl/pull/32619", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "8cfead542087aa19f5c18bae87d4a6e6103038cd", "date": "2026-08-31T17:21:07+02:00", "subject": "doc/man3/EVP_PKEY_get_size.pod: remove stray \">\" from HTML table formatting", "body": "There is a stray closing angle bracket in HTML-specific formatting\nof the security levels table, remove it.\n\nFixes: 73188a01bd99 \"doc: document EVP_PKEY_get_security_category function\"\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Kurt Roeckx \nReviewed-by: Dmitry Belyavskiy \nReviewed-by: Paul Dale \nMerge-date: Thu Sep 3 06:40:43 2026\nMerged-from: https://github.com/openssl/openssl/pull/32623", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "63eb836949f904b7efb76565a046bb8b5bd6ad08", "date": "2026-08-31T17:21:07+02:00", "subject": "doc/man3/EVP_PKEY_get_size.pod: remove stray \">\" from HTML table formatting", "body": "There is a stray closing angle bracket in HTML-specific formatting\nof the security levels table, remove it.\n\nFixes: 73188a01bd99 \"doc: document EVP_PKEY_get_security_category function\"\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Kurt Roeckx \nReviewed-by: Dmitry Belyavskiy \nReviewed-by: Paul Dale \nMerge-date: Thu Sep 3 06:40:33 2026\nMerged-from: https://github.com/openssl/openssl/pull/32623", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "159cbf3cc120deb2f2167ed318029879d22e2794", "date": "2026-08-31T17:21:07+02:00", "subject": "doc/man3/EVP_PKEY_get_size.pod: remove stray \">\" from HTML table formatting", "body": "There is a stray closing angle bracket in HTML-specific formatting\nof the security levels table, remove it.\n\nFixes: 73188a01bd99 \"doc: document EVP_PKEY_get_security_category function\"\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Kurt Roeckx \nReviewed-by: Dmitry Belyavskiy \nReviewed-by: Paul Dale \nMerge-date: Thu Sep 3 06:40:16 2026\nMerged-from: https://github.com/openssl/openssl/pull/32623", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "710640d8faebcbd4ef1b58c6d91b59aa7374700b", "date": "2026-08-31T16:49:32+01:00", "subject": "Fix pane-border-status not being redrawn properly.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "503162ea7a1088bf7b5660c373311c42f6e29c4c", "date": "2026-08-31T18:03:55+02:00", "subject": "ossl_store_register_loader_int(): make sure to check that scheme is non-empty", "body": "Reviewed-by: Eugene Syromiatnikov \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Simo Sorce \nMergeDate: Thu Sep 10 13:32:56 2026\n(Merged from https://github.com/openssl/openssl/pull/27507)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "f9fbb82f1e252a8e769d326f4b6880e43493ca19", "date": "2026-08-31T18:26:24+02:00", "subject": "We don't need linking s_lib into FIPS provider", "body": "Reviewed-by: Igor Ustinov \nReviewed-by: Paul Dale \nReviewed-by: Simo Sorce \nMergeDate: Wed Sep 2 15:04:35 2026\n(Merged from https://github.com/openssl/openssl/pull/32626)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "6a0e3cc0645305336b482a63079889e2d286dee3", "date": "2026-08-31T10:44:19-06:00", "subject": "Don't let stderr block when filling up.", "body": "Always discard stderr if we are not keeping it.\n\nFixes: #32128\nReviewed-by: Saša Nedvědický \nReviewed-by: Norbert Pocs \nMerge-date: Thu Sep 3 06:39:06 2026\nMerged-from: https://github.com/openssl/openssl/pull/32615\n(cherry picked from commit 0fff0090507a205e6d37072d6ea151b68d0f74b6)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "0fff0090507a205e6d37072d6ea151b68d0f74b6", "date": "2026-08-31T10:44:19-06:00", "subject": "Don't let stderr block when filling up.", "body": "Always discard stderr if we are not keeping it.\n\nFixes: #32128\nReviewed-by: Saša Nedvědický \nReviewed-by: Norbert Pocs \nMerge-date: Thu Sep 3 06:39:06 2026\nMerged-from: https://github.com/openssl/openssl/pull/32615", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "7188c20c7141924aeddfe7b5767a9b8ae64c71db", "date": "2026-08-31T17:46:19+01:00", "subject": "Avoid redrawing kitty images when terminal scrolling moves them.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "4186d7169b068a3f248e9795702d41ea9842a954", "date": "2026-08-31T12:07:07-05:00", "subject": "doc: Drop stale note about extension LWLock names.", "body": "Oversight in commit 38b602b028.\n\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "347486775812573f09c79e1f4d26d1f7506d261f", "date": "2026-08-31T12:34:24-05:00", "subject": "doc: Fix link on pg_dsm_registry_allocations page.", "body": "Oversight in commit 283e823f9d.\n\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "e2050759a43ff033f172b52ae0f98f3ce21d6481", "date": "2026-08-31T19:48:44+02:00", "subject": "docs: add error boundaries entries", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "neovim", "sha": "4af10a8cd825e8dedf49f965627c533dcc17b1ef", "date": "2026-08-31T14:21:04-04:00", "subject": "backport: feat(pos): mutable vim.Pos, vim.Range fields", "body": "Allow plugins to set the human-readable fields.\n\n(cherry picked from commit 8ce1aac477a33ab1fe2a0b590ff278518dc0e473)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "c65d378bb4fded436f197aa7bc52c03f507b8bd4", "date": "2026-08-31T14:08:39-05:00", "subject": "fix(core): validate SVG animation attributes outside the SVG namespace", "body": "Make the SVG animation security context depend on the tag name instead of the namespace the element was created in. An animation element declared outside an `` is created in the HTML namespace, but still animates once it ends up inside an SVG subtree, so to and `attributeName` bindings were reaching the DOM unvalidated.\n\nFixes #70490\n\n(cherry picked from commit 7168bed663ba2a53f087a68a251e8ccfd938eb97)", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "angular", "sha": "7168bed663ba2a53f087a68a251e8ccfd938eb97", "date": "2026-08-31T14:08:39-05:00", "subject": "fix(core): validate SVG animation attributes outside the SVG namespace", "body": "Make the SVG animation security context depend on the tag name instead of the namespace the element was created in. An animation element declared outside an `` is created in the HTML namespace, but still animates once it ends up inside an SVG subtree, so to and `attributeName` bindings were reaching the DOM unvalidated.\n\nFixes #70490", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "tmux", "sha": "0530a7d89a2c4eb375c2b0ca1f87bd3e8978e109", "date": "2026-08-31T19:27:46Z", "subject": "Add format variables for history change counters, from Michael K Darling.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "b402687fcd9f99d82b5c47b41e082f8032df1530", "date": "2026-08-31T19:34:09Z", "subject": "Add a generation counter to pane output, from Michael K Darling.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "415d800e7bb0c3a1c468f2d6538e3772df13ebd2", "date": "2026-08-31T21:48:22+02:00", "subject": "build: add performance tests", "body": "Add a performance test bundle with the necessary build logic, and 3\ntests to start with.\n\nThe new test binary is built by default as part of the existing test\ntargets and builds cleanly on all supported platforms. It's not run in\nCI. Build time overhead at this point is negligible. If it becomes\nhigher, the perf target may receive separate build controls.\n\nAlso apply small tidy-ups to existing build files.\n\nC-code-authored-by: Daniel Stenberg\nBuild-and-integration-by: Viktor Szakats\n\nCloses #22741\nCloses #22747", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "tmux", "sha": "9c1d5a080be8c1c1943fbfd0eee3677ed79cfe47", "date": "2026-08-31T20:52:54+01:00", "subject": "Use hash for actions, from Andreas Fehlner.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "543adc4de024dda4f3b8f368893c7d01a3cf0ea8", "date": "2026-08-31T22:00:40+02:00", "subject": "openssl: avoid strlen() on the data from OpenSSL", "body": "It is not guaranteed to be null terminated. Use memchr() instead.\n\nReported-by: Max Dymond\nCloses #22767", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "9ee317d3212efe6dd5fe72af801e1d8e5735fcd8", "date": "2026-08-31T13:20:26-07:00", "subject": "ci: remove shared primitives pullapprove group", "body": "This group is no longer necessary as responsibilities for that are now spread amongst teams more equally.\n\n(cherry picked from commit f3ed75eb6bb7dc81a60b8c8ae5b434aabe63397b)", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "angular", "sha": "f3ed75eb6bb7dc81a60b8c8ae5b434aabe63397b", "date": "2026-08-31T13:20:26-07:00", "subject": "ci: remove shared primitives pullapprove group", "body": "This group is no longer necessary as responsibilities for that are now spread amongst teams more equally.", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "curl", "sha": "c23bc84be4c4330b3af065e5e83350686832ee73", "date": "2026-08-31T22:23:48+02:00", "subject": "CURLOPT_USERNAME.md: ambient username caveats", "body": "Closes #22768", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "bf416baf7d7c69b5ca64ebac206b6f129a0708f2", "date": "2026-08-31T16:31:46-04:00", "subject": "DTLS 1.3 Enforce the RFC 9147 Section 8 sending epoch limit", "body": "dtls1_increment_epoch() capped the read/write epoch at UINT16_MAX for\nDTLS 1.2, but skipped that check entirely for DTLS 1.3, relying only\non the full 64-bit wrap as a backstop. RFC 9147 Section 8 requires a\nmuch tighter bound for senders: \"sending implementations MUST NOT\nallow the epoch to exceed 2^48-1\", motivated by AES-128's 128-bit key\ngiving a non-negligible key-reuse probability at anything near 2^64\nrekeys. Receivers are held to the looser Section 6.1 ceiling instead,\nso this is intentionally a write-side-only, DTLS-1.3-only check.\n\nAdd DTLS1_3_MAX_EPOCH (2^48-1) and enforce it in the write branch of\ndtls1_increment_epoch() before the increment, mirroring the existing\nUINT16_MAX guard's pre-increment style. With both the DTLS 1.2 and\nDTLS 1.3 write-side ceilings now always intercepting w_conn_epoch\nwell below UINT64_MAX, the post-increment wrap-to-zero check on that\npath can no longer be reached, so it is removed.\n\nAdd test_dtls13_increment_epoch_max() to test/dtls13_internal_test.c,\nwhich negotiates a real DTLS 1.3 connection (SSL_CONNECTION_IS_DTLS13()\ndepends on the negotiated method and can't be forced directly), then\nwrites w_conn_epoch to one below the limit and calls the real\nincrement function at and past the boundary.\n\nFixes: https://github.com/openssl/openssl/issues/32511\nAssisted-by: Claude:claude-sonnet-5\nReviewed-by: Frederik Wedel-Heinen \nReviewed-by: Jakub Zelenka \nMerge-date: Mon Sep 7 14:28:46 2026\nMerged-from: https://github.com/openssl/openssl/pull/32617", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "1d7ea23fa1c23e0be156a6886228d98509ecff8d", "date": "2026-08-31T22:49:29+02:00", "subject": "perf: restore the -DDEBUGBUILD for debug builds", "body": "Otherwise the base64 test segfaults\n\nCloses #22769", "style": "conventional", "type": "perf", "scope": "", "breaking": false} {"repo": "curl", "sha": "bf87b94d2d585989dd31309744b922d18bf5a4ac", "date": "2026-08-31T22:49:29+02:00", "subject": "perf: restore the -DDEBUGBUILD for debug builds", "body": "Otherwise the base64 test segfaults", "style": "conventional", "type": "perf", "scope": "", "breaking": false} {"repo": "curl", "sha": "fd1f5ec1d1227f89337a6a532e1a0ea259e768b3", "date": "2026-08-31T23:02:53+02:00", "subject": "perf/urlparser: provide a built-in set of URLs", "body": "Instead of loading a list. Iterate more loops instead by default.\n\nCloses #22770", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "fc1a44a1917f354b292b246e2956cde18f0e5b1f", "date": "2026-09-01T00:10:30+03:00", "subject": "test(core): add failing test for jsaction action map resolving Object.prototype properties", "body": "Demonstrates that a plain {} used as the parsed jsaction action map resolves an inherited Object.prototype member (e.g. toString) for a key that was never explicitly parsed into the map, instead of undefined. Callers only check `!== undefined`, so this gets treated as a real action.\n\nThis commit adds the test only, without the fix, so the failure is visible on its own.\n\n(cherry picked from commit 5cb3ed2958a4358cfc2a712950c56e41fe3a7413)", "style": "conventional", "type": "test", "scope": "core", "breaking": false} {"repo": "angular", "sha": "5cb3ed2958a4358cfc2a712950c56e41fe3a7413", "date": "2026-09-01T00:10:30+03:00", "subject": "test(core): add failing test for jsaction action map resolving Object.prototype properties", "body": "Demonstrates that a plain {} used as the parsed jsaction action map resolves an inherited Object.prototype member (e.g. toString) for a key that was never explicitly parsed into the map, instead of undefined. Callers only check `!== undefined`, so this gets treated as a real action.\n\nThis commit adds the test only, without the fix, so the failure is visible on its own.", "style": "conventional", "type": "test", "scope": "core", "breaking": false} {"repo": "angular", "sha": "d7d5401cd5e8ce23d8bb3f0e7ae08323b307d0d5", "date": "2026-09-01T00:10:37+03:00", "subject": "fix(core): don't use plain objects as maps for jsaction data", "body": "The parsed jsaction action map, the parser cache keyed by the raw jsaction attribute value, and the fallback in getDefaulted were all plain {} objects. Looking up a key that was never explicitly set but happens to match an inherited Object.prototype member (toString, constructor, hasOwnProperty, ...) resolves to that inherited value instead of undefined, and callers only check `!== undefined` to decide whether there's a real action or cache entry.\n\nUse Object.create(null) instead, which has no prototype to accidentally collide with.\n\n(cherry picked from commit 8d8859cb2113b90db3da1315877a2d7fd4ceb392)", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "angular", "sha": "8d8859cb2113b90db3da1315877a2d7fd4ceb392", "date": "2026-09-01T00:10:37+03:00", "subject": "fix(core): don't use plain objects as maps for jsaction data", "body": "The parsed jsaction action map, the parser cache keyed by the raw jsaction attribute value, and the fallback in getDefaulted were all plain {} objects. Looking up a key that was never explicitly set but happens to match an inherited Object.prototype member (toString, constructor, hasOwnProperty, ...) resolves to that inherited value instead of undefined, and callers only check `!== undefined` to decide whether there's a real action or cache entry.\n\nUse Object.create(null) instead, which has no prototype to accidentally collide with.", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "curl", "sha": "ea656bc8565d033aef716c395cd700581e2a61bd", "date": "2026-08-31T23:11:57+02:00", "subject": "fixup the same for CMake", "body": "from Viktor", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "7fe13265aca2706b5cc6b55f6f7df042a7b87b72", "date": "2026-08-31T23:23:57+02:00", "subject": "perf/percent: basic percent encode/decode performance test", "body": "Closes #22771", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "c6cc6c592f30901345d94ef75e0d42c1894fae6c", "date": "2026-09-01T00:27:17+03:00", "subject": "feat: check `Object.prototype` property names in `new-cap`", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "curl", "sha": "f9c45e9b241ccd6165c88811698cdf9e9d896fa7", "date": "2026-09-01T00:06:51+02:00", "subject": "perf: 0-pad the nanosecond decimals correctly", "body": "Closes #22772", "style": "conventional", "type": "perf", "scope": "", "breaking": false} {"repo": "postgres", "sha": "dcf14ef31cd2d867004ede87d8060716ea62619d", "date": "2026-09-01T08:12:40+09:00", "subject": "Fix comment in attribute_stats.c", "body": "Oversight in ce207d2a7901.\n\nAuthor: Corey Huinker \nDiscussion: https://postgr.es/m/CADkLM=eo7MtuCE=YjovW+=ASw1=q39qQ3qarrsw+EKfU901ztA@mail.gmail.com\nBackpatch-through: 18", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "36fc3f01346727bedecbc1805f26dc245dafd5aa", "date": "2026-08-31T19:09:28-05:00", "subject": "docs(docs-infra): improve update checkbox accessibility", "body": "Associate each recommendation with its checkbox so assistive technology has a descriptive label and non-interactive text toggles the control. Keep embedded links independently operable.\n\n(cherry picked from commit d8e74e7d39f020655b2bef91b955bd9cace14c6d)", "style": "conventional", "type": "docs", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "d8e74e7d39f020655b2bef91b955bd9cace14c6d", "date": "2026-08-31T19:09:28-05:00", "subject": "docs(docs-infra): improve update checkbox accessibility", "body": "Associate each recommendation with its checkbox so assistive technology has a descriptive label and non-interactive text toggles the control. Keep embedded links independently operable.", "style": "conventional", "type": "docs", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "6e299da8cfc7df49029d5d8bb7d5739560e1aaa4", "date": "2026-08-31T17:14:38-07:00", "subject": "feat(forms): hard-code `readOnlyHint` and `untrustedContentHint` for WebMCP implicit signal forms", "body": "Signal forms can safely assume `{readOnlyHint: false, untrustedContentHint: false}` given their context. A form _must_ alter page DOM (or else how would a user interact with it) and is therefore definitionally side-effectful. We also assume returned content is trusted, given it is currently hard-coded or derived from application errors.\n\nIn the future, we might want to consider cases where application errors are explicitly untrusted or where application developers choose to customize the response text with something which might be untrusted. For now, that's out of scope and we'll worry about it when a compelling use case arises.", "style": "conventional", "type": "feat", "scope": "forms", "breaking": false} {"repo": "angular", "sha": "91a2bf84250f9b0a28a8a68e1eded6e9506bc5e9", "date": "2026-08-31T17:16:06-07:00", "subject": "feat(core): support annotations in WebMCP tool declarations", "body": "This updates `declareExperimentalWebMcpTool` to allow annotations to be provided to opt-in the tool explicitly into being read only or returning untrusted content.", "style": "conventional", "type": "feat", "scope": "core", "breaking": false} {"repo": "ant-design", "sha": "e6edee0dbc8bc9712bb58c135ad57bd53fe8f74e", "date": "2026-09-01T08:19:52+08:00", "subject": "docs: fix invalid links", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "78e3a0a68663a0b9419cb20045d862bfe181c557", "date": "2026-09-01T08:30:32+08:00", "subject": "fix(Layout): handle undefined collapsed as uncontrolled", "body": "", "style": "conventional", "type": "fix", "scope": "Layout", "breaking": false} {"repo": "neovim", "sha": "9d2a31b05e018749f14c7c6688379b059ba3ea75", "date": "2026-08-31T20:49:50-04:00", "subject": "fix(cwd): flicker after temp context-switch #41561 #41582", "body": "Problem:\n`shorten_fnames()` always redraws the statusline/tabline, even if no\nbuffer name changed. Since b296666e a temp context-switch\n(`win_execute()`, `vim._with{win=}`) restores the CWD, so every such\nswitch flickers the message area.\n\nSolution:\nRedraw only if `shorten_buf_fname()` actually changed a name.", "style": "conventional", "type": "fix", "scope": "cwd", "breaking": false} {"repo": "vue-core", "sha": "051630e0e57dd610235539afc91212f7b35898b8", "date": "2026-09-01T08:53:21+08:00", "subject": "refactor(runtime-vapor): align fragment lifecycle hook names", "body": "", "style": "conventional", "type": "refactor", "scope": "runtime-vapor", "breaking": false} {"repo": "neovim", "sha": "ca992e82d40cb8ed31331ba75cf0a58c6d693ac7", "date": "2026-09-01T10:04:57+09:00", "subject": "fix(buffer): keep trailing slash significant in URI names #41577", "body": "Problem:\nLiteral path comparison ignored one trailing slash for every buffer name,\nincluding URIs. Generic URI syntax does not make a non-empty path\nequivalent to the same path with a trailing slash, so distinct URI\nbuffers collapsed into one.\n\nSolution:\nRequire equal lengths when comparing URI buffer names, while retaining\ntrailing-separator normalization for filesystem paths.\n\nAI-assisted", "style": "conventional", "type": "fix", "scope": "buffer", "breaking": false} {"repo": "vue-core", "sha": "e376b1de26105b02ccacd14bb77728bf09c78ee9", "date": "2026-08-31T18:41:14-07:00", "subject": "refactor(vapor): split NON_STABLE slot marker out of VaporSlotFlags", "body": "", "style": "conventional", "type": "refactor", "scope": "vapor", "breaking": false} {"repo": "vue-core", "sha": "eb9f523f2a41dd964ffdbeac2cd5ddc06ffdb029", "date": "2026-08-31T18:51:23-07:00", "subject": "refactor(vapor): collapse slot outlet forwarding flags to a 2-bit mode", "body": "", "style": "conventional", "type": "refactor", "scope": "vapor", "breaking": false} {"repo": "postgres", "sha": "f1c6f7ec95dc0434726738a5733491ef00a5a640", "date": "2026-09-01T11:42:13+09:00", "subject": "Fix integer to_char() overflow with V format", "body": "When to_char() formatted an integer value with a V pattern, it could\nreturn an incorrect result instead of reporting an overflow. V shifts\nthe decimal point by multiplying the input value by a power of ten before\nformatting it, so, for example,\n\n to_char(3, '9V999999999')\n\nrequires computing 3 * 10^9. This result does not fit in int4, but\nthe integer variant of to_char() performed the multiplication using a\nplain int32 expression. The intermediate result could therefore\noverflow, causing the function to output incorrect digits instead of\nraising \"integer out of range\".\n\nUse dtoi4() and int4mul() for this calculation so that both an\nout-of-range multiplier and an out-of-range product are detected, as\nwith ordinary integer arithmetic. This also matches the existing int8\nimplementation, which uses dtoi8() and int8mul() for the same\noperation.\n\nAfter this change, to_char() with V format either returns the\ncorrectly formatted result when the scaled value fits in int4, or\nraises \"integer out of range\" when it does not.\n\nBackpatch to all supported versions.\n\nReported-by: Andrey Rachitskiy \nAuthor: Andrey Rachitskiy \nReviewed-by: Miłosz Bieniek \nReviewed-by: Fujii Masao \nDiscussion: https://postgr.es/m/CAB8bMivEfqZxOVdzc3kZDN++XshmkEz2t7dfGBU8+oUm864EZg@mail.gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "735915e81cb7e5fac4c3f7ba748a294b6f4dc34c", "date": "2026-09-01T11:31:54+08:00", "subject": "fix(Tooltip): scope arrow clamping to unrestricted shifts", "body": "", "style": "conventional", "type": "fix", "scope": "Tooltip", "breaking": false} {"repo": "ant-design", "sha": "f8e2216c969ed3445c8c8b7a9463617582669631", "date": "2026-09-01T11:39:32+08:00", "subject": "fix(Input.Password): ignore repeated key activation", "body": "", "style": "conventional", "type": "fix", "scope": "Input.Password", "breaking": false} {"repo": "vite", "sha": "472385e6ec4b21e3167c7abf9769883d1c9675f8", "date": "2026-09-01T11:48:39+08:00", "subject": "feat(config): warn on named imports from JSON modules", "body": "", "style": "conventional", "type": "feat", "scope": "config", "breaking": false} {"repo": "postgres", "sha": "07df2b253277321d68d515615ec6c0a3d70ec591", "date": "2026-09-01T12:54:31+09:00", "subject": "doc: reformat GRAPH_TABLE examples", "body": "The GRAPH_TABLE examples in ddl.sgml and queries.sgml were written\nas single long lines, making their structure harder to read in the\ngenerated documentation.\n\nReformat these examples so that the graph name, MATCH clause, and\nCOLUMNS clause appear on separate lines, making them easier to read.\n\nAuthor: Koshino Taiki \nReviewed-by: Ashutosh Bapat \nReviewed-by: Fujii Masao \nDiscussion: https://postgr.es/m/OS9P286MB64860BD6CD6D4B4B0E1CEDE894A32@OS9P286MB6486.JPNP286.PROD.OUTLOOK.COM\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "fdef04f112aadfea40ad3c448d96a49a04c168bd", "date": "2026-09-01T13:55:00+09:00", "subject": "feat: add warning for unsupported hooks in plugin returned from `applyToEnvironment` hook", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "vite", "sha": "b50e1b4a3d66128a4076e19769b2e29657985516", "date": "2026-09-01T12:10:09+07:00", "subject": "fix(utils): handle dot in srcset density descriptor", "body": "", "style": "conventional", "type": "fix", "scope": "utils", "breaking": false} {"repo": "vue-core", "sha": "c9736498f4794c93fa2083ca6a31a894d6c3c888", "date": "2026-09-01T13:30:28+08:00", "subject": "fix(runtime-vapor): de-structured rest does not include inherited properties", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "postgres", "sha": "8e483af5515ec4ee90d17a3864f5bb764e4e9c47", "date": "2026-09-01T14:32:31+09:00", "subject": "Use pg_neg_s{32,64}_overflow() for some overflow checks", "body": "The files updated in this commit open-coded checks based on a minimal\nvalue (as of \"x == PG_INTnn_MIN\") combined with negative calculations.\nMore efficient equivalents are available in int.h with\npg_neg_s32_overflow() and pg_neg_s64_overflow().\n\nExtracted from a larger patch by the same author. These spots have\nproved to reduce the number of instructions generated after testing\nacross some versions of clang and gcc, with and without the\n__builtin_*_overflow() functions. Some unlikely() can be removed, but\nnot all, cash.c standing as an exception. There should be no behavioral\nchanges.\n\nSuggested-by: Dagfinn Ilmari Mannsåker \nAuthor: Ewan Young \nReviewed-by: Michael Paquier \nDiscussion: https://postgr.es/m/CAON2xHO4tTFiow2KKLbpiOEJNL4Th5p1QVTm4mSdEU+KW1rrfQ@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "57961337e43b3efabf2640e549529a0b11f3ccfc", "date": "2026-09-01T13:40:03+08:00", "subject": "fix(Tag): ignore repeated key activation", "body": "* fix(Tag): ignore repeated key activation\n\n* fix(Tag): suppress repeated key defaults\n\n---------", "style": "conventional", "type": "fix", "scope": "Tag", "breaking": false} {"repo": "vue-core", "sha": "2774aab7eda52fb9336757e70b57cf8553a338fa", "date": "2026-08-31T22:46:44-07:00", "subject": "perf(runtime-vapor): take the slot fast path for unmarked outlets under a boundary", "body": "", "style": "conventional", "type": "perf", "scope": "runtime-vapor", "breaking": false} {"repo": "curl", "sha": "2b4d88aa5e435d4e1ee1a230ea6f8923f78a8f47", "date": "2026-09-01T07:54:11+02:00", "subject": "THANKS: contributors from the 8.22.0 release", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "cc4ac6cff73c153de2b9016fb58f01118e2d0fc5", "date": "2026-08-31T22:54:43-07:00", "subject": "fix(runtime-vapor): harden slot fast-path gates", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "curl", "sha": "cfce7bdaf819628c0351dc9c5bd8687d3cdd6ac3", "date": "2026-09-01T07:55:01+02:00", "subject": "VERSIONS: add 8.22.0 and the pending next release", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "ec890236c683d65010387d6a83af69658134cdf0", "date": "2026-08-31T23:01:02-07:00", "subject": "perf(runtime-vapor): mirror the slot fast path in the interop outlet", "body": "", "style": "conventional", "type": "perf", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "4b3adcb1429ab6c36cab2b145345d7b818b6a10f", "date": "2026-08-31T23:07:43-07:00", "subject": "perf(runtime-vapor): apply slot stability analysis under an enclosing boundary", "body": "", "style": "conventional", "type": "perf", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "5a06a6357ebdec83bbebe555a0781cd07e29d5db", "date": "2026-08-31T23:13:09-07:00", "subject": "perf(runtime-vapor): wrap once slot fallbacks once on the interop path", "body": "", "style": "conventional", "type": "perf", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "bc6d040ae966d004d4fcfb68fef9d042ca81551a", "date": "2026-08-31T23:16:35-07:00", "subject": "refactor(compiler-vapor): share the dev flag-annotation emission", "body": "", "style": "conventional", "type": "refactor", "scope": "compiler-vapor", "breaking": false} {"repo": "vue-core", "sha": "ae12889a2bf3585bf015fed0cc09b61aed5fc8eb", "date": "2026-08-31T23:20:01-07:00", "subject": "refactor(runtime-vapor): align interop slot decode names with the shared helpers", "body": "", "style": "conventional", "type": "refactor", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "947b0fc0454be59c3c8920a338d8fc0551dc31c6", "date": "2026-08-31T23:25:46-07:00", "subject": "perf(runtime-vapor): resolve raw slots in the fast-path decision", "body": "", "style": "conventional", "type": "perf", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "2964d2a3ceddf34a7ebed48682d120843b39f34d", "date": "2026-08-31T23:27:33-07:00", "subject": "test(runtime-vapor): use the forwarded-slot helper for hand-built outlets", "body": "", "style": "conventional", "type": "test", "scope": "runtime-vapor", "breaking": false} {"repo": "ant-design", "sha": "a1c0958081c3f598483443ef4f53656d860a6352", "date": "2026-09-01T14:35:47+08:00", "subject": "refactor: reuse @rc-component/util type guards", "body": "", "style": "conventional", "type": "refactor", "scope": "", "breaking": false} {"repo": "eslint", "sha": "9cecb8a0a2348070abf72321965d41919c7cc626", "date": "2026-09-01T15:45:04+09:00", "subject": "docs: document `\\c` control letter escapes in no-control-regex", "body": "* fix: no-control-regex false negative for \\c control letter escapes\n\nThe rule detected \\x, \\u, and raw control characters in regex\npatterns but missed \\cA-\\cZ (and \\ca-\\cz) control letter escapes,\nwhich also produce control characters in the U+0000-U+001F range.\n\nFixes #21281\n\n* docs: document \\c control letter escapes in no-control-regex\n\nClarify that \\cA-\\cZ (and \\ca-\\cz) control letter escapes are\nintentionally allowed by this rule, since they explicitly signal\nintent to match a control character, unlike the hex/unicode/raw\nforms the rule disallows. Add tests covering this behavior.\n\nFixes #21281", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "8f5e9e9e1dea3839d61a8437afde0129b234f1f2", "date": "2026-08-31T23:47:28-07:00", "subject": "fix(runtime-vapor): gate interop fallback delegation with a SLOT_RESOLVER bit", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "4352822084edab122e215731393c8648a9ad13e8", "date": "2026-09-01T00:20:16-07:00", "subject": "refactor(compiler-vapor): compute v-if flag names only in dev", "body": "", "style": "conventional", "type": "refactor", "scope": "compiler-vapor", "breaking": false} {"repo": "vue-core", "sha": "9fde9acd3360f2280c46e81005e179d7ed06e09b", "date": "2026-09-01T00:21:42-07:00", "subject": "refactor(compiler-vapor): emit dev flag annotations behind __DEV__ guards", "body": "", "style": "conventional", "type": "refactor", "scope": "compiler-vapor", "breaking": false} {"repo": "curl", "sha": "1c58877705b172da892cb9c18c5edb572f2d1d73", "date": "2026-09-01T09:22:24+02:00", "subject": "sspi ambient user checks: remove again", "body": "With the clarification that the application has to take extra care when\nchanging an \"ambient user\" setting, remove the special check in\nconnection matching that had been imperfect from the start.\n\nDocumented here: https://curl.se/libcurl/c/CURLOPT_USERNAME.html\n\nCloses #22775", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "520e08c7f0a5892fbc05f3af79da408c4132e53a", "date": "2026-09-01T16:30:55+09:00", "subject": "docs: clarify parameter description in `getText`", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "eslint", "sha": "9f4a364ab0ade048dfce1f37792b1d461d866e55", "date": "2026-09-01T10:32:13+03:00", "subject": "fix: Ignore static imports in no-unreachable", "body": "* fix: Ignore static imports in no-unreachable\n\nFixes #21273\n\n* test: add split import unreachable coverage\n\n---------", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "curl", "sha": "72986f2aa68b273572add2546b79bb509a7216ec", "date": "2026-09-01T09:40:10+02:00", "subject": "http2: remove assert in ingress processing", "body": "Remove the assert that every `data` triggering ingress processing\nhas an opened HTTP/2 stream. This is not true for connection\nshutdown processing, at the least.\n\nReported-By: Max Dymond\nCloses #22776", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "94b4cb02e90fa01bae6c61caa7023bdc83a53cd6", "date": "2026-09-01T15:49:46+08:00", "subject": "docs: remove stale create-react-app guide references", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "3eb4c6eb1d1ecec7e303d753265788e48140fff2", "date": "2026-09-01T16:00:49+08:00", "subject": "fix(Input.Search): preserve custom enterButton onMouseDown", "body": "", "style": "conventional", "type": "fix", "scope": "Input.Search", "breaking": false} {"repo": "postgres", "sha": "3103bdafde6b0dfa8edbcefb48ac51097c6a9c1e", "date": "2026-09-01T10:13:30+02:00", "subject": "Handle PG_INT32_MIN negation overflow in right()", "body": "A negative n means \"return all but the first |n| characters\", so\ntext_right() negates n before clipping. Negating PG_INT32_MIN\noverflows; with -fwrapv the result is PG_INT32_MIN again, still\nnegative, and pg_mbcharcliplen() then returns an offset of zero,\nso the whole string is returned where the correct answer is an\nempty string:\n\n SELECT right('abcdef', (-2147483648)::int4); -- 'abcdef', want ''\n SELECT right('abcdef', -2147483647); -- '', correct\n\nClamp to PG_INT32_MAX instead. Any n whose absolute value is at\nleast the string's length skips all of it, and a text value cannot\nbe longer than PG_INT32_MAX, so this gives the same answer for every\nother input. Erroring out, as text_format_string_conversion() does\nfor a width of INT_MIN, would not be correct here: unlike a format\nwidth, an out-of-range skip count has a well-defined result.\n\nUsing pg_neg_s32_overflow() would be a slightly more optimal fix but\nas it's only available in PostgreSQL 18 and later the decision was\ntaken to apply the same fix to all backbranches.\n\nBackpatch to all supported versions.\n\nAuthor: Ewan Young \nReviewed-by: Daniel Gustafsson \nReviewed-by: Dagfinn Ilmari Mannsåker \nReviewed-by: David Rowley \nReviewed-by: Chao Li \nDiscussion: https://postgr.es/m/CAON2xHNnBz-AcPJgDmd5_39+8qR5AUKEZk4X3ZM-0zdsATn8kQ@mail.gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "729bac9eb1042a722f21c6f6dfb5d0f6b0e6c67a", "date": "2026-09-01T10:13:30+02:00", "subject": "Fix right() with the most negative integer", "body": "A negative n means \"return all but the first |n| characters\", so\ntext_right() negates n before clipping. Negating PG_INT32_MIN overflows;\nwith -fwrapv the result is PG_INT32_MIN again, still negative, and\npg_mbcharcliplen() then returns an offset of zero, so the whole string is\nreturned where the correct answer is an empty string:\n\n SELECT right('abcdef', (-2147483648)::int4); -- 'abcdef', want ''\n SELECT right('abcdef', -2147483647); -- '', correct\n\nClamp to PG_INT32_MAX instead. Any n whose absolute value is at least the\nstring's length skips all of it, and a text value cannot be longer than\nPG_INT32_MAX, so this gives the same answer for every other input. Note\nthat erroring out, as text_format_string_conversion() does for a width of\nINT_MIN a few hundred lines away, would not be right here: unlike a format\nwidth, an out-of-range skip count has a well-defined result.\n\ntext_left() is not affected. Its negative case computes the character\nlength plus n rather than negating n, and since the length is non-negative\nand bounded by the varlena size limit that sum cannot overflow.\n\nBackpatch to all supported versions.\n\nAuthor: Ewan Young \nReviewed-by: Daniel Gustafsson \nReviewed-by: Dagfinn Ilmari Mannsåker \nReviewed-by: David Rowley \nReviewed-by: Chao Li \nDiscussion: https://postgr.es/m/CAON2xHNnBz-AcPJgDmd5_39+8qR5AUKEZk4X3ZM-0zdsATn8kQ@mail.gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "33d0e986f23ae4d923436511817a489a65d122da", "date": "2026-09-01T17:38:41+09:00", "subject": "ci: make backport releases possible", "body": "", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "vite", "sha": "f40efefbb3630cdb7235286bc2b51673d9fbfc27", "date": "2026-09-01T10:39:44+02:00", "subject": "refactor: delete unused `esbuildPlugin`", "body": "", "style": "conventional", "type": "refactor", "scope": "", "breaking": false} {"repo": "postgres", "sha": "135b867a530cac2e3796d87c852b53bef40f0077", "date": "2026-09-01T11:14:25+02:00", "subject": "Wait for checksum state transition in test", "body": "The test for disabling data checksums after a failure didn't wait\nfor the state transition to complete before testing, which could\nlead to false negatives in testing. Fix by passing the wait param\nwhen disabling. Identified on buildfarm member culicidae.\n\nAuthor: Daniel Gustafsson \nReported-by: Alexander Lakhin \nReviewed-by: Alexander Lakhin \nDiscussion: https://postgr.es/m/b2f34fd3-8e98-450e-9e8c-ac08238e9116@gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "c046a8f10eb0610c125af6c00a34c021308b8bec", "date": "2026-09-01T17:15:11+08:00", "subject": "chore: remove obsolete nwsapi override", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "angular", "sha": "2a199ce68e9aaed2ed0e07822594becc9410d6ad", "date": "2026-09-01T11:31:12+02:00", "subject": "docs: Update request handler to use handle method", "body": "there is no request method the correct method is hanlde\n\n(cherry picked from commit e1bf5393ad0bc3ed6a6380626873b7971b154313)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "e1bf5393ad0bc3ed6a6380626873b7971b154313", "date": "2026-09-01T11:31:12+02:00", "subject": "docs: Update request handler to use handle method", "body": "there is no request method the correct method is hanlde", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vitest", "sha": "a44c6c17b030178185dca34b92c065ef719b6558", "date": "2026-09-01T11:34:33+02:00", "subject": "test: works in combination with `keyboard`", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "vitest", "sha": "1fcf81dc96bbea212eb158ae99d5d4e48eec8af6", "date": "2026-09-01T11:35:01+02:00", "subject": "fix: tracks previous target/coords", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "angular", "sha": "0669975ee7c89683b721618c26b72e702a73ff07", "date": "2026-09-01T11:38:51+02:00", "subject": "docs: fix global target event listener example", "body": "(cherry picked from commit 9fd77852fb96feef242337319812c32326f4276b)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "9fd77852fb96feef242337319812c32326f4276b", "date": "2026-09-01T11:38:51+02:00", "subject": "docs: fix global target event listener example", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "37809a1e52ab74e867f69430c3837968888ad270", "date": "2026-09-01T11:48:32+02:00", "subject": "store: add a fake store provider for API tests", "body": "Introduce a fake store provider with two loader schemes: \"fake\" with\na plain open() entry point and delete support, and \"fake-ex\" with an\nopen_ex() entry point and no delete support. The loader emits NAME\nobjects so it needs no decoders or other providers, and it records\nthe ctx params it receives so tests can assert what actually reached\nthe loader.\n\nUse it to cover previously unexercised parts of the OSSL_STORE API:\nthe legacy open function, open failure handling, all four search\ncriteria with their accessors, the post process callback, various\nerror conditions and deleting through a loader with and without\ndelete support.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Richard Levitte \nReviewed-by: Daniel Kubec \nMerge-date: Wed Sep 16 08:46:32 2026\nMerged-from: https://github.com/openssl/openssl/pull/32624", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "bce9bc86a99df5ede2249b69794a732f92964d87", "date": "2026-09-01T11:51:12+02:00", "subject": "fix: provide return type to `BrowserCommand` from `UserEventCommand`", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "postgres", "sha": "2a2dab67dc40a41a003545308423c1d3def02dea", "date": "2026-09-01T11:58:20+02:00", "subject": "Fix memset invocation", "body": "It was setting sha_buf_tmp but sizing it by sizeof sha_buf. These\nhappen to be the same size, so it's not an actual problem, but for\ncorrectness make the sizeof match.\n\nReported-by: Sehrope Sarkuni \nDiscussion: https://www.postgresql.org/message-id/flat/a035c67f-4aa8-42e0-bb24-7b5abe927af4%40eisentraut.org", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "3f5b46c1536edd85c56b00333721b8b63bb10edf", "date": "2026-09-01T12:04:15+02:00", "subject": "h2proxy: remember goaway \"last_stream_id\"", "body": "In the HTTP/2 proxy filter, remember the server announced\n\"last_stream_id\" form a GOAWAY frame, so that error handling does not\nreport a perfectly fine stream as broken.\n\nReported-By: Max Dymond\nCloses #22778", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "ec40430ed5d9d738a087544e6ad49aa1199308c8", "date": "2026-09-01T12:25:46+02:00", "subject": "OSSL_FN: Add OSSL_FN_to_bytes_be()", "body": "Export the value of OSSL_FN to a big-endian byte sequence in constant time.\n\nAssisted-by: Claude:claude-opus-4-8\nReviewed-by: Jakub Zelenka \nReviewed-by: Richard Levitte \nMerge-date: Mon Sep 21 10:07:02 2026\nMerged-from: https://github.com/openssl/openssl/pull/32866", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "610bf015996bfb35ffdcb0a3396efb381ba744ba", "date": "2026-09-01T13:46:40+02:00", "subject": "lib: expire/progress handling, do less clock calls", "body": "Save clock system calls by passing obtained timestamps to expire and\nprogress calls.\n\ntransfer: remove duplicate progress call cf-ip-happy: remove timestamps\nnever used cf-https-connect: remove timestamps never used\n\nCloses #22779", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "a62fa616113e68735db3e1dc353e09e675d6df6c", "date": "2026-09-01T13:50:51+02:00", "subject": "pgcrypto: remove unused BF_ASM macro", "body": "This has never been defined to anything but 0, and the _BF_body_r\nfunction it would call has never existed.\n\nAuthor: Dagfinn Ilmari Mannsåker \nReviewed-by: Daniel Gustafsson \nDiscussion: https://www.postgresql.org/message-id/flat/87ldew2yqu.fsf%40wibble.ilmari.org", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "8f3101d8208707b90e2cc93354d37a81cc2ce1ea", "date": "2026-09-01T12:49:49Z", "subject": "Extend word commands to use any Unicode whitespace character, GitHub issue 5562 from Hongyi Zhao.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "77309c07b081223841dd27ed89159dec93438d20", "date": "2026-09-01T13:50:56+01:00", "subject": "Whitespace regress test from Hongyi Zhao.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "717ebe4d0af0959ef26fbdf0b6d8c54dcdc32509", "date": "2026-09-01T09:03:23-04:00", "subject": "vtls: fix Curl_cert_hostcheck when the pattern is not a C string", "body": "https://github.com/curl/curl/pull/22767 removed some of the strlen\ncalls, but it wasn't quite sufficient. Curl_cert_hostcheck has a number\nof places where it implicitly assumes the buffer is zero-terminated:\n\n- The check for empty strings assume that, even if the buffer is\n zero-length, it can be dereferenced. This only works for C strings\n where there's always a zero byte in there. Replace this with a length\n check.\n\n- The strncmp(\"*.\") call will read past the buffer if the pattern equals\n \"*\". This works for C strings because it will stop at the zero byte\n first. For general buffers, this needs a length check first.\n\nThe various checks for the first and last character are actually OK,\nbecause the caller rejects empty strings first, but I similarly guarded\nthem for completeness. Likewise, I guarded the accesses of hostname in\nhopes it can eventually be de-C-string-ified, but the call to\nCurl_host_is_ipnum requires it to be a C string.\n\nUpdate unit1397 to cover these cases. Run each test with a\nseparate-allocated buffer. The two fixed invalid accesses are then\ncaught by ASan and valgrind.\n\nCloses #22795", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "bd8218e85d9305cda3abbe9a54a4288e538632e9", "date": "2026-09-01T13:04:29Z", "subject": "When changing selection-mode to line, set up the selection start and end correctly, GitHub issue 5545.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "9ebf9b1017c2b7d63e5bb938dfd46c7be1117f98", "date": "2026-09-01T16:23:23+03:00", "subject": "fix(ui2): respect options set during startup after enabling ui2 #41591", "body": "Problem: options related to ui2 (like `fillchars` with `msgsep`) do not\n take effect if set during startup after enabling ui2.\n\nSolution: explicitly check just after startup if relevant options were\n changed during startup.", "style": "conventional", "type": "fix", "scope": "ui2", "breaking": false} {"repo": "openssl", "sha": "6e31e19bf226b49bd13fcd425200fbbe4d005667", "date": "2026-09-01T22:48:35+09:00", "subject": "Fix double free in server client-certificate parse error path", "body": "Commit 95dcb1b719a1 (\"Allow SSL_set_retry_verify() on server\nsession\") split the server client-certificate processing path so that\napplications can retry verification on the server side. During that\nreshuffle, tls_process_client_certificate() lost the ownership-clearing\nassignment after sk_X509_push(sk, x).\n\nWhen parsing an X.509 client Certificate message, if at least one\ncertificate is parsed successfully and malformed trailing data then makes\nthe next loop iteration fail its length checks, x still aliases the\nprevious stack-owned certificate. The err path frees x and then frees the\nstack, causing a second X509_free() on the same object.\n\nThat second free performs an atomic reference count update on freed heap\nstorage. The observable runtime effect is allocator- and build-dependent:\na normal functional handshake test may still see only the expected\nparse-time decode_error, while ASan, valgrind, or enabled refcount\nassertions detect the invalid access.\n\nRestore the ownership-clearing assignment after the stack push. Add a\nTLS 1.2 memory-BIO regression test which sends one valid DER-encoded\nclient certificate followed by malformed trailing data and checks that\nthe server rejects it with decode_error / SSL_R_CERT_LENGTH_MISMATCH.\n\nFixes: 95dcb1b719a1 (\"Allow SSL_set_retry_verify() on server session\")\n\nAssisted-by: pi:kimi-k3\nReviewed-by: Jakub Zelenka \nReviewed-by: Viktor Dukhovni \nMerge-date: Wed Sep 16 08:44:01 2026\nMerged-from: https://github.com/openssl/openssl/pull/32627", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "aee3523489be43eecdd94c5cf89cf3939a237364", "date": "2026-09-01T22:48:35+09:00", "subject": "Fix double free in server client-certificate parse error path", "body": "Commit 95dcb1b719a1 (\"Allow SSL_set_retry_verify() on server\nsession\") split the server client-certificate processing path so that\napplications can retry verification on the server side. During that\nreshuffle, tls_process_client_certificate() lost the ownership-clearing\nassignment after sk_X509_push(sk, x).\n\nWhen parsing an X.509 client Certificate message, if at least one\ncertificate is parsed successfully and malformed trailing data then makes\nthe next loop iteration fail its length checks, x still aliases the\nprevious stack-owned certificate. The err path frees x and then frees the\nstack, causing a second X509_free() on the same object.\n\nThat second free performs an atomic reference count update on freed heap\nstorage. The observable runtime effect is allocator- and build-dependent:\na normal functional handshake test may still see only the expected\nparse-time decode_error, while ASan, valgrind, or enabled refcount\nassertions detect the invalid access.\n\nRestore the ownership-clearing assignment after the stack push. Add a\nTLS 1.2 memory-BIO regression test which sends one valid DER-encoded\nclient certificate followed by malformed trailing data and checks that\nthe server rejects it with decode_error / SSL_R_CERT_LENGTH_MISMATCH.\n\nFixes: 95dcb1b719a1 (\"Allow SSL_set_retry_verify() on server session\")\n\nAssisted-by: pi:kimi-k3\nReviewed-by: Jakub Zelenka \nReviewed-by: Viktor Dukhovni \nMerge-date: Wed Sep 16 08:43:45 2026\nMerged-from: https://github.com/openssl/openssl/pull/32627", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "03851df1105da45350e6d8db20d099299b27428c", "date": "2026-09-01T22:50:32+09:00", "subject": "doc: clarify aliasing of VALUES in FROM clauses", "body": "The VALUES reference page said that an AS clause was required when\nVALUES is used in a FROM clause. This was imprecise because AS is\noptional when specifying an alias.\n\nIn v14 and v15, a table alias is still required in this case, so state\nthat explicitly instead. In v16 and later, commit bcedd8f5fce made table\naliases for subqueries in FROM clauses optional, so state that the\ntable alias is optional there. In all branches, continue to recommend\nexplicit column aliases as good practice.\n\nBackpatch to all supported versions.\n\nAuthor: Ian Barwick \nReviewed-by: Laurenz Albe \nReviewed-by: Fujii Masao \nDiscussion: https://postgr.es/m/CAB8KJ=jrvZdNLLhYSiCgbjLTz2LKEjYw+-HVQkaF+kgk61KtHA@mail.gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "8e2cb142217f2efee1d10dcc02bfb75145ae775d", "date": "2026-09-01T17:00:51+03:00", "subject": "fix: `new-cap` false positive for `UTC` calls with `properties: false`", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "curl", "sha": "48e196426b148e6683914ce6a9206143175cb688", "date": "2026-09-01T16:31:57+02:00", "subject": "escape: polish encoding and decoding for speed", "body": "Primarily thanks to a new loopup table and taking bigger chunks to work\non instead of byte-by-byte, this PR does decoding twice as fast as\nbefore and it encodes five times faster (on my x86-64).\n\nCloses #22780", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "61e64cb19a08c23dca1df46f1ae09df514059fc6", "date": "2026-09-01T10:32:41-04:00", "subject": "Make after-startup shmem failure test reliable", "body": "The test expects a 128 kB request to exceed the shared memory available after\nstartup. Huge pages can round up the main shared memory segment enough for\nthat request to succeed, defeating the test.\n\nDisable huge pages for the allocation-failure phase so that it exercises the\nintended error path independently of shared memory segment rounding.\n\nAuthor: Ayush Tiwari \nReported-by: Andres Freund \nDiscussion: https://postgr.es/m/rpkhzrrmo7qiz7rr3qlufksb2oci33wcg3ao75le7dqindgoif@6yozkjwb6bdf\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "a6d1882cacca5e991dadbeb4e53f0693ff5289a7", "date": "2026-09-01T16:33:54+02:00", "subject": "fix(highlight): cterm default for Cursor hlgroup", "body": "Problem:\nThe `Cursor` highlight is invisible if 'termguicolors' is disabled.\n\nSolution:\nChange its default.", "style": "conventional", "type": "fix", "scope": "highlight", "breaking": false} {"repo": "angular", "sha": "00ec65546b93ee685fbf8f51989abeed720eceea", "date": "2026-09-01T17:58:21+03:00", "subject": "refactor(devtools): fix signal details UI", "body": "Fix different action buttons and value preview container size.", "style": "conventional", "type": "refactor", "scope": "devtools", "breaking": false} {"repo": "neovim", "sha": "7168e0d12f3474fdd1b7fc8db94739da6344e3c1", "date": "2026-09-01T17:15:51+02:00", "subject": "fix(mcursor): drop q= \"follow motion\" message", "body": "", "style": "conventional", "type": "fix", "scope": "mcursor", "breaking": false} {"repo": "neovim", "sha": "9a29622b545fc76c8b44d0e2b90f318a599eef39", "date": "2026-09-01T11:17:22-04:00", "subject": "feat(multicursor): MC HAMMER #41587", "body": "Other (squashed) commits:\n\nfix(tui): emit ui_send output atomically with the frame\n\nProblem:\ntui_ui_send() writes directly to the TTY, bypassing the output buffer.\nSequences sent via nvim_ui_send() (e.g. kitty multiple-cursors, or\nvisual-dot-repeat) always arrive in a separate TTY write from the frame\nthey were computed for. This manifests as \"tearing\", or e.g. in the case\nof multicursor the terminal renders text with stale cursor overlays.\n\nSolution:\n- tui_ui_send(): while a frame is being assembled (pending invalid\n regions or buffered output), buffer instead of writing directly.\n - Out-of-frame sends (tty queries, clear-on-disable) still write\n immediately.\n- mcursor.lua: emit the terminal-cursor update at the end of the redraw\n cycle (`on_end`, when screen positions are final) instead of\n vim.schedule().", "style": "conventional", "type": "feat", "scope": "multicursor", "breaking": false} {"repo": "curl", "sha": "30107badd5aa8569cf6005ff6a8adb81c25a4a17", "date": "2026-09-01T17:36:28+02:00", "subject": "tool_cb_wrt: try up to .9999 when no-clobbering", "body": "Previously it gave up after 100.\n\nSome users use this feature when downloading large number of files in a\nglob range and then 100 is on the small side.\n\nTweaked test 1683 to verify this even with the new limit.\n\nCloses #22784", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "753db54f8d50e49ff12ae7a35cc6f1acb5fa168c", "date": "2026-09-01T23:42:16+08:00", "subject": "Update package.json", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "5e38120ddc9896f33e63bd75d91a9c470c2a576a", "date": "2026-09-01T17:42:53+02:00", "subject": "doh: drop assert from doh_resp_decode_httpsrr()", "body": "When the incoming DNS packet is broken, there might be bytes left at the\nend that the parser ignores so let's not assert on that.\n\nReported-by: Max Dymond\nCloses #22785", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "9e5292a00323c0a80e0f02c629e06f738884ee3e", "date": "2026-09-01T17:50:05+02:00", "subject": "ngtcp2: shrink max sendbuf", "body": "The sendbuf for a HTTP/3 stream in ngtcp2 could grow up to 10MB which\ncan lead to large memory consumption without any real gains.\n\nShrink the max size to 640KB, because that is enough for everyone.\n\nCloses #22786", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "777f56a34e0add4a73fa2f9feabd7c4c4b84c63f", "date": "2026-09-01T19:10:38+03:00", "subject": "docs: explain that errors during early app startup miss ErrorHandler", "body": "There's a short window where Angular can't send errors to ErrorHandler\nyet: while it's still building the root module or root component. It\nneeds that root instance to look up the ErrorHandler you provided, so\nan error thrown before it exists just becomes a normal uncaught error\ninstead.\n\nThis mostly bites people using Angular elements, where a custom element\ntag already sitting in the page gets upgraded (and its component built)\nthe moment you call customElements.define, which can happen very early.\n\nAdded a short section to the error handling guide explaining why this\nhappens and the usual ways around it: setTimeout, APP_BOOTSTRAP_LISTENER,\nor moving element registration to ngDoBootstrap.\n\nFixes #29211\n\n(cherry picked from commit ca383059761b5ffe326d00f2835ab97037dd4fbf)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "ca383059761b5ffe326d00f2835ab97037dd4fbf", "date": "2026-09-01T19:10:38+03:00", "subject": "docs: explain that errors during early app startup miss ErrorHandler", "body": "There's a short window where Angular can't send errors to ErrorHandler\nyet: while it's still building the root module or root component. It\nneeds that root instance to look up the ErrorHandler you provided, so\nan error thrown before it exists just becomes a normal uncaught error\ninstead.\n\nThis mostly bites people using Angular elements, where a custom element\ntag already sitting in the page gets upgraded (and its component built)\nthe moment you call customElements.define, which can happen very early.\n\nAdded a short section to the error handling guide explaining why this\nhappens and the usual ways around it: setTimeout, APP_BOOTSTRAP_LISTENER,\nor moving element registration to ngDoBootstrap.\n\nFixes #29211", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "postgres", "sha": "f3ac68ffe38692582c24880ef10cc365ec683fa6", "date": "2026-09-01T12:16:45-04:00", "subject": "Handle XmlExpr more honestly in clauses.c.", "body": "eval_const_expressions() didn't constant-fold XmlExpr. In some cases\nit cannot, but often it can. Failure to do so can cause visible\nregressions in SQL-language function behavior compared to pre-v18,\nthough only in a rather narrow set of contexts: basically, if you're\ntrying to use a CASE to prevent evaluation of a failure-prone XML\nfunction.\n\nWhile we're at it, make contain_mutable_functions() handle XmlExpr\nmore precisely, and adjust some comments that justify not treating\nXmlExpr explicitly.\n\nThis seems worth sneaking into v19, but I'm hesitant to put it into\nv18. It's not really a bug fix, because we disclaim the safety of\nusing CASE this way, and it carries some risk of de-optimizing queries\nthat worked satisfactorily before.\n\nBug: #19487\nReported-by: Ilya Portnov \nAuthor: Andrey Rachitskiy \nCo-authored-by: Tom Lane \nDiscussion: https://postgr.es/m/19487-367258bc497b923a@postgresql.org\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "f89eca9cd6e9a751345d50200572f00e1ce02c78", "date": "2026-09-01T18:21:37+02:00", "subject": "tool_cb_wrt: inject the no-clobber number before the extension", "body": "Starting in curl 8.23.0, if the target file name uses an extension, curl\ninstead creates the new file name by inserting `-[num]` before the\nextenstion. Downloading to `image.jpg` with --no-clobber, curl first\ntries to save it as `image-1.jpg` then `image-2.jpg` etc.\n\nCloses #22787", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "6317e5b7e53e1faaaaafdbb2f35517e2e54bbe19", "date": "2026-09-01T18:24:49+02:00", "subject": "apps: cover request attributes in req test recipe", "body": "Generate certificate requests with attributes taken from the\nconfiguration, using both the prompting style entries and the\nprompt = no style, and check that the attributes appear in the\nprinted request text.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Paul Dale \nReviewed-by: Daniel Kubec \nMerge-date: Mon Sep 7 18:09:32 2026\nMerged-from: https://github.com/openssl/openssl/pull/32628", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "07749290fd9bd96503f05683ec5323a24d26dc0b", "date": "2026-09-01T11:26:19-05:00", "subject": "pg_dump: Remove some dead code.", "body": "Two members of TableInfo were set but not used. Remove them.\n\nAuthor: Aleksander Alekseev \nReviewed-by: Japin Li \nDiscussion: https://postgr.es/m/CAJ7c6TOET3UVUqXcXMpcoDAiZJvMFmHjnavj%2BFGdfXEKwe_56g%40mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "6ded7791946e8670cd16a30a5593aa6eebd81ade", "date": "2026-09-01T10:38:23-06:00", "subject": "Let TLSProxy tests set the expected NewSessionTicket count", "body": "The DTLS 1.3 handshake-complete check waited for two ticket acks, but\na server sends at most one ticket after a resumption, so the resuming\nconnection in 70-test_dtls13sessionid.t stalled until the proxy killed\ns_server. Add a per-connection expected_tickets setting (default two)\nand set it to one there.\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Tomas Mraz \nMerge-date: Tue Sep 1 17:04:41 2026\nMerged-from: https://github.com/openssl/openssl/pull/32629", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "6e3f29a049a3ff9f6737669db7b7b469c66c9e02", "date": "2026-09-01T19:39:07+03:00", "subject": "docs: highlight the search tutorial @for block as Angular", "body": "The block quotes a `.ts` file but declares `language=\"html\"`, and shiki's\nHTML grammar does not know Angular control flow, so the whole `@for` line\nrenders unstyled. Use `angular-ts`, matching the other blocks in the\ntutorial.\n\n(cherry picked from commit c03e872ed98e5e76c30646b8560b2949465d3800)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "c03e872ed98e5e76c30646b8560b2949465d3800", "date": "2026-09-01T19:39:07+03:00", "subject": "docs: highlight the search tutorial @for block as Angular", "body": "The block quotes a `.ts` file but declares `language=\"html\"`, and shiki's\nHTML grammar does not know Angular control flow, so the whole `@for` line\nrenders unstyled. Use `angular-ts`, matching the other blocks in the\ntutorial.", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "82733d90b5bc58b8d064ed49c282aa028664a1ed", "date": "2026-09-01T19:16:02+02:00", "subject": "test: tolerate DTLS peer change with fragment data", "body": "In the DTLS variant of the TLS 1.3 HRR tests a large key share can\nsplit the server's flight across two datagrams, leaving the proxy\nwith an incomplete message fragment when the client rejects the\nServerHello and its alert arrives first. The proxy treated any peer\nchange with pending fragment data as fatal and died, aborting the\nwhole test recipe. This made the no-ec run-checker build fail\nintermittently, since only the ffdhe key shares are large enough to\nforce the fragmentation.\n\nA DTLS peer may legitimately abort mid-flight, so discard the stale\nfragment data and carry on instead. TLS keeps the strict behaviour.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Nikola Pajkovsky \nReviewed-by: Tomas Mraz \nMerge-date: Fri Sep 4 17:40:52 2026\nMerged-from: https://github.com/openssl/openssl/pull/32630", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "angular", "sha": "cc05deab7204bb2146cae2c90b12e79a2b5e1846", "date": "2026-09-01T12:17:43-05:00", "subject": "test(forms): Add utility functions and update test files to use them", "body": "Moves common helper functions into a shared test utility to reduce duplication.\n\n(cherry picked from commit 318fefad303921a5452eea91c52ac09f8981f39c)", "style": "conventional", "type": "test", "scope": "forms", "breaking": false} {"repo": "angular", "sha": "318fefad303921a5452eea91c52ac09f8981f39c", "date": "2026-09-01T12:17:43-05:00", "subject": "test(forms): Add utility functions and update test files to use them", "body": "Moves common helper functions into a shared test utility to reduce duplication.", "style": "conventional", "type": "test", "scope": "forms", "breaking": false} {"repo": "angular", "sha": "ef48630a14f0bc8ba0a46d3fc7555c2a29f26a41", "date": "2026-09-01T12:18:06-05:00", "subject": "test(core): replace act with actAsync", "body": "Replaces the custom `act` helper function with the standardized `actAsync` utility\n\n(cherry picked from commit 063eda2db2366fab88935416902bff4d81828eea)", "style": "conventional", "type": "test", "scope": "core", "breaking": false} {"repo": "angular", "sha": "063eda2db2366fab88935416902bff4d81828eea", "date": "2026-09-01T12:18:06-05:00", "subject": "test(core): replace act with actAsync", "body": "Replaces the custom `act` helper function with the standardized `actAsync` utility", "style": "conventional", "type": "test", "scope": "core", "breaking": false} {"repo": "angular", "sha": "b2903de6813e649c85a1c850abca605807e9e2be", "date": "2026-09-01T19:24:09+02:00", "subject": "ci: move `packages/private` under `fw-general`", "body": "This doesn't need dev-infra approval as its mostly testing utils.\n\n(cherry picked from commit f44bf0fe221288d03107cc78acfbe333a9a9b7b5)", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "angular", "sha": "f44bf0fe221288d03107cc78acfbe333a9a9b7b5", "date": "2026-09-01T19:24:09+02:00", "subject": "ci: move `packages/private` under `fw-general`", "body": "This doesn't need dev-infra approval as its mostly testing utils.", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "angular", "sha": "ed2e401d2da9a6dd5c87006573a8e42934f58f62", "date": "2026-09-01T20:25:33+03:00", "subject": "fix(core): don't fail NgModule checks for a pipe that extends a base class", "body": "When an abstract base class has something like a lifecycle hook, Angular\ncompiles it as a directive with no selector. If a pipe extends that base\nclass, the pipe picks up the base class's directive definition through\nnormal class inheritance.\n\nThe NgModule dev-mode checks then got confused by that inherited\ndefinition and reported the pipe as a broken directive:\n\n - \"Directive SomePipe has no selector, please add it!\"\n - or, if the base class was a default (standalone) abstract directive,\n \"SomePipe is marked as standalone and can't be declared...\"\n\nBoth only happened in tests (TestBed), not when running the app.\n\nNow the selector check and the standalone check both look at the pipe's\nown definition and ignore a directive definition that only came from a\nbase class.\n\nFixes #36427\n\n(cherry picked from commit 58d536bc8362f3942c929e58be551d871b00cea7)", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "angular", "sha": "58d536bc8362f3942c929e58be551d871b00cea7", "date": "2026-09-01T20:25:33+03:00", "subject": "fix(core): don't fail NgModule checks for a pipe that extends a base class", "body": "When an abstract base class has something like a lifecycle hook, Angular\ncompiles it as a directive with no selector. If a pipe extends that base\nclass, the pipe picks up the base class's directive definition through\nnormal class inheritance.\n\nThe NgModule dev-mode checks then got confused by that inherited\ndefinition and reported the pipe as a broken directive:\n\n - \"Directive SomePipe has no selector, please add it!\"\n - or, if the base class was a default (standalone) abstract directive,\n \"SomePipe is marked as standalone and can't be declared...\"\n\nBoth only happened in tests (TestBed), not when running the app.\n\nNow the selector check and the standalone check both look at the pipe's\nown definition and ignore a directive definition that only came from a\nbase class.\n\nFixes #36427", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "postgres", "sha": "ef67d599eff86203371e38f2d08ec6fa6236b5b2", "date": "2026-09-01T10:26:10-07:00", "subject": "libpq: Stop greasing protocol connections", "body": "This reverts commit 4966bd3ed95e0d02bf220c4bad2b292963827dd4, which\nintroduced the 19beta grease campaign, now that Beta 3 has been stamped.\n(The reservation and documentation for the grease version code, done in\nd8d7c5dc8f, remains.)\n\nDiscussion: https://postgr.es/m/CAOYmi%2Bkqt_gn-wvUjJG1j3MM4LDDsRMURSPz2eMxfABu0iGtpw%40mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "59471367e33daa080025280dce833b7e4a2a3154", "date": "2026-09-01T19:47:40+02:00", "subject": "apps: cover the -extfile option in the ca recipe", "body": "Sign a request taking the X509v3 extensions from a separate\nextension file, both via its default section and with a section\nexplicitly selected using the -extensions option, and check that\nthe extensions from the chosen section end up in the certificate.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Paul Dale \nReviewed-by: Nikola Pajkovsky \nMerge-date: Mon Sep 7 18:11:16 2026\nMerged-from: https://github.com/openssl/openssl/pull/32632", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "f576890160af67ddf739920238838da647fcf2c0", "date": "2026-09-01T19:56:45+02:00", "subject": "apps: cover the -crlexts option in the ca recipe", "body": "Generate a CRL taking its extensions from a dedicated section of\nthe configuration file and check that the CRL is version 2 and\ncarries the requested extension.\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Paul Dale \nReviewed-by: Nikola Pajkovsky \nMerge-date: Mon Sep 7 18:11:17 2026\nMerged-from: https://github.com/openssl/openssl/pull/32632", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "37661c9e656ff2acb77e5ff948733b912574e354", "date": "2026-09-01T23:33:27+05:30", "subject": "fix: address comments", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "curl", "sha": "3a2c261293c204589ec2410dafb001425389f03a", "date": "2026-09-01T14:16:23-04:00", "subject": "CURLOPT_SHARE.md: remove bad explanation for cookie sharing", "body": "- Remove it because the correct explanation is already documented in\n CURL_LOCK_DATA_COOKIE in CURLSHOPT_SHARE.md.\n\nPrior to this change the doc said if you share cookies that the cookie\nengine is enabled. That is no longer true since d0a7ee3f (precedes curl\n7.67.0).\n\nReported-by: Artem Prilutskiy\n\nRef: https://curl.se/libcurl/c/CURLSHOPT_SHARE.html#CURLLOCKDATACOOKIE\n\nFixes https://github.com/curl/curl/issues/22788\nCloses https://github.com/curl/curl/pull/22790", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "610976420a331ab007621ae30f664d17c8d2eebe", "date": "2026-09-01T20:45:13+02:00", "subject": "Fix assertion failures in DELETE FOR PORTION OF tuple routing", "body": "We must handle cases where an inserted temporal leftover gets routed to\na different partition than the original tuple's. This happens if the\npartition key depends on the application-time column used in the FOR\nPORTION OF clause. Everything works for UPDATE, but for DELETE we can\nhit some Asserts, since that operation never needed to be covered\nbefore. One Assert is for the RETURNING list; the other, for WITH CHECK\nOPTION. This commit updates both to include CMD_DELETE.\n\nAuthor: Paul A Jungwirth \nReported-by: Peter Eisentraut \nDiscussion: https://postgr.es/m/03276924-6d55-470d-8192-a49c053c940f@eisentraut.org", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "f26ab219e28ceb88b29f9165800879548dcc6646", "date": "2026-09-01T14:20:16-05:00", "subject": "Fix spurious errors in COPY FROM (FORMAT {text,csv}).", "body": "The SIMD path for these commands reads ahead into the input buffer\nvia CopyLoadInputBuf(), which reports encoding and conversion\nerrors. The problem is that the errant data might lie beyond the\nend-of-copy marker, which we aren't copying and thus needn't error\nfor. To fix, add a \"speculative\" parameter to CopyLoadInputBuf()\nthat defers these errors, and use it in the SIMD path. The errors\nare instead reported only if we actually reach the errant data.\nTesting indicates this does not meaningfully impact performance.\n\nOversight in commit e0a3a3fd53.\n\nReviewed-by: Greg Burd \nReviewed-by: Nazir Bilal Yavuz \nDiscussion: https://postgr.es/m/apWOr1kgVSqweBnU%40nathan\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "e44904ddda9b25fe3d44dcb39747ee4aa0502db3", "date": "2026-09-01T22:35:52+03:00", "subject": "fix(docs-infra): keep the collapsed code state when a code block is reshown", "body": "The code block lives inside `@if (showCode())`, so hiding it destroys the DOM and\nshowing it builds a fresh copy with no hidden lines. Nothing reapplied\n`setCodeLinesVisibility()`, so a collapsed block came back showing the whole file.\n\nReapply it once the block is rendered again. It already branches on `expanded()`,\nso a block that was expanded stays expanded.\n\n(cherry picked from commit 90b9e081f551b722635204fc095ae3e7ce5f89e0)", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "90b9e081f551b722635204fc095ae3e7ce5f89e0", "date": "2026-09-01T22:35:52+03:00", "subject": "fix(docs-infra): keep the collapsed code state when a code block is reshown", "body": "The code block lives inside `@if (showCode())`, so hiding it destroys the DOM and\nshowing it builds a fresh copy with no hidden lines. Nothing reapplied\n`setCodeLinesVisibility()`, so a collapsed block came back showing the whole file.\n\nReapply it once the block is rendered again. It already branches on `expanded()`,\nso a block that was expanded stays expanded.", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "tmux", "sha": "57a13664cc2cf0db1c6b4f575c4934bf4ec1c4ee", "date": "2026-09-01T19:50:58Z", "subject": "Do not allow cursor on/off to escape synchronized updates.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "c49dfdf496af9a0d835c37b2e9404b4284644c7e", "date": "2026-09-01T20:51:20+01:00", "subject": "Regress test for cursor updates.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "60e871658fbac9f823bee16811e15bf51ea6c210", "date": "2026-09-01T20:52:44+01:00", "subject": "Remove +x.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "5bbff5c2e9f78db0247a9897fae873bdff6a4be8", "date": "2026-09-01T21:02:57+01:00", "subject": "doh: reject undersized HTTPS RDATA", "body": "HTTPS RDATA requires at least two bytes for SvcPriority and one byte\nfor TargetName. Reject shorter records before storing them to avoid\nzero-sized allocations in debug builds.\n\nAdd unit coverage for RDATA lengths zero through three.\n\nCloses #22791", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "1bda0b23bce47ab19d518ba51b26b9caed967782", "date": "2026-09-01T15:20:01-05:00", "subject": "Fix scaling of autovacuum freeze scores.", "body": "Sufficiently large settings of the *_freeze_score_weight parameters\ncan lower the scaling thresholds below the *_freeze_max_age\nsettings. This can lead the scoring code to raise a value < 1.0 to\na power > 1.0, thus lowering the score, which is the opposite of\nwhat is intended. To fix, only exponentiate the scores when doing\nso won't lower their values.\n\nOversight in commit d7965d65fc.\n\nReviewed-by: Bharath Rupireddy \nReviewed-by: Greg Burd \nDiscussion: https://postgr.es/m/apWcbq91BAfhmsKU%40nathan\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "2d6aed1d92d22864472b7958c68f3013c81ea050", "date": "2026-09-01T21:52:47+01:00", "subject": "Add some more tests for recent stuff.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "6bb6c286a972ffe1e4d2e98d8e61cc66beb87822", "date": "2026-09-01T21:03:55Z", "subject": "Create screen before zoom may need to use it.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "59dc0e75c439aa88bd303fd0e3e02dae677e78b1", "date": "2026-09-01T22:07:08+01:00", "subject": "Add a couple more tests.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "87e0a082438264ad90b87fd74165ab4fd90f63ef", "date": "2026-09-02T06:17:48+09:00", "subject": "fix: prefer-object-has-own autofix breaks when Object is shadowed", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "postgres", "sha": "7da24355a2993721bf9bc54224b3bb3e601530ce", "date": "2026-09-01T16:27:18-05:00", "subject": "pg_dump: Fix performance regression in attribute statistics query.", "body": "After adding pg_stats.tableid, we taught pg_dump to use it in the\nattribute statistics query for servers running v19 or later.\nNotably, we did not give this new form of the query a \"redundant\nfilter clause\" like we did the pre-v19 form. Testing indicates\nthat produces a significant performance regression, and that adding\na similar filter clause to the v19-or-later form of the query fixes\nit.\n\nOversight in commit 4b5ba0c4ca.\n\nReviewed-by: Sami Imseih \nReviewed-by: Michael Paquier \nDiscussion: https://postgr.es/m/apQ9nvLW_zFdarYd%40nathan\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "d4485336057af536df877f06d189886948680dc9", "date": "2026-09-01T18:32:21-04:00", "subject": "X509: complete ASN.1 stream failure handling", "body": "Use buf directly when validating canonical output pointer movement.\n\nValidate NDEF prefix and suffix output passes and return failure when the stream flush fails. Add a two-digest PKCS7 allocation-failure regression.\n\nAssisted-by: OpenAI Codex:gpt-5.6-sol\nReviewed-by: Jakub Zelenka \nReviewed-by: Mounir Idrassi \nMerge-date: Mon Sep 14 14:13:34 2026\nMerged-from: https://github.com/openssl/openssl/pull/32047\n(cherry picked from commit 5fd4eae78531f9d73aebb3550eb3bdbcb4d09e8f)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "5fd4eae78531f9d73aebb3550eb3bdbcb4d09e8f", "date": "2026-09-01T18:32:21-04:00", "subject": "X509: complete ASN.1 stream failure handling", "body": "Use buf directly when validating canonical output pointer movement.\n\nValidate NDEF prefix and suffix output passes and return failure when the stream flush fails. Add a two-digest PKCS7 allocation-failure regression.\n\nAssisted-by: OpenAI Codex:gpt-5.6-sol\nReviewed-by: Jakub Zelenka \nReviewed-by: Mounir Idrassi \nMerge-date: Mon Sep 14 14:13:34 2026\nMerged-from: https://github.com/openssl/openssl/pull/32047", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "d7385ed015343355a01105ac9e67836a0fd23912", "date": "2026-09-02T01:20:40+02:00", "subject": "clang-tidy: enable check `bugprone-assignment-in-selection-statement`", "body": "Supported by clang-tidy 23.1.0+.\n\nRef: https://clang.llvm.org/extra/clang-tidy/checks/bugprone/assignment-in-selection-statement.html\n\nCloses #22792", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "c586a5336a5bd0fef2136424542f4738ae96bf94", "date": "2026-09-01T17:03:27-07:00", "subject": "Fix nonstandard SQL command doc id= values introduced in the v19 cycle.", "body": "While some instances of supported PostgreSQL versions use the\nsql-alter-table style, sql-altertable is the dominant style. Change\nthese before they're baked into the doc URLs of a supported version.\n\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "6dbc931a39976dbbffca99e4e9c77d4d2a8cc240", "date": "2026-09-01T17:03:27-07:00", "subject": "Fix user-facing English grammar introduced in the v19 cycle.", "body": "Backpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "a4a8573a75d2031ca531b57ff59b0327cdcd953b", "date": "2026-09-02T00:19:17Z", "subject": "fix(devtools): safely handle foreign root elements in feature detection", "body": "Prevent Angular DevTools from crashing when elements matching [ng-version] in the DOM do not belong to the host application (such as third-party custom elements or browser extensions). ng.getComponent returns null for these elements, which previously caused ng.getDirectiveMetadata to throw.\n\n(cherry picked from commit ef09d108f23adc1aa29ec260fcf04b7271ea53c1)", "style": "conventional", "type": "fix", "scope": "devtools", "breaking": false} {"repo": "angular", "sha": "ef09d108f23adc1aa29ec260fcf04b7271ea53c1", "date": "2026-09-02T00:19:17Z", "subject": "fix(devtools): safely handle foreign root elements in feature detection", "body": "Prevent Angular DevTools from crashing when elements matching [ng-version] in the DOM do not belong to the host application (such as third-party custom elements or browser extensions). ng.getComponent returns null for these elements, which previously caused ng.getDirectiveMetadata to throw.", "style": "conventional", "type": "fix", "scope": "devtools", "breaking": false} {"repo": "angular", "sha": "ec48c8305c40fbafbab51801f93bb4e4b6a25ea0", "date": "2026-09-02T00:19:21Z", "subject": "fix(core): return null when getDirectiveMetadata is called with null or undefined", "body": "Safely return null instead of throwing an unhandled TypeError when getDirectiveMetadata is called with a falsy directive or component instance.\n\n(cherry picked from commit cea6896a0fef4fa4f37a4c7a43c7a880900eb21d)", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "angular", "sha": "cea6896a0fef4fa4f37a4c7a43c7a880900eb21d", "date": "2026-09-02T00:19:21Z", "subject": "fix(core): return null when getDirectiveMetadata is called with null or undefined", "body": "Safely return null instead of throwing an unhandled TypeError when getDirectiveMetadata is called with a falsy directive or component instance.", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "postgres", "sha": "680f707fb2e4e3a954137e027e89afbf2d6f7944", "date": "2026-09-02T09:19:37+09:00", "subject": "Speed up pg_waldump TAP test", "body": "Several invocations decoded and printed the full WAL range even though\nthe test only needs to confirm that a command form works or that\ndecoding reaches a specific error. Limit those checks to one record,\nand start the fall-off-the-end checks near the end of the generated WAL.\n\nThis change reduces the IPC overhead overall, particularly on Windows,\nwithout reducing coverage.\n\nI am usually hesitant to backpatch such changes as this is only an\nimprovement, but the gains are too good in terms of IO and runtime, for\nboth the CI and the buildfarm. Based on the numbers provided, the CI\ntakes 30% less time to run the test with this change on Windows (worst\ncase shown on the lists). These tests have been introduced in\n96063e28366b.\n\nAuthor: Sehrope Sarkuni \nReviewed-by: Michael Paquier \nReviewed-by: Nazir Bilal Yavuz \nDiscussion: https://postgr.es/m/CAH7T-araSEdsNpxiKaMOW8kr_ZHhekCLHx5yzm2ksq4FdgULbA@mail.gmail.com\nBackpatch-through: 17", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "67abeb6253e2e0280733d70ffe763df00e227a29", "date": "2026-09-02T08:40:15+08:00", "subject": "fix(compiler-vapor): preserve empty text nodes in component slots", "body": "", "style": "conventional", "type": "fix", "scope": "compiler-vapor", "breaking": false} {"repo": "vue-core", "sha": "e8e40445536489461848a01c27a7dc2be089f5ba", "date": "2026-09-02T08:43:27+08:00", "subject": "fix(runtime-vapor): keep fallthrough live for custom element root instances", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "ant-design", "sha": "26f5d0af66c3ac629ba7d9aa3c927264b62363b6", "date": "2026-09-02T09:25:55+08:00", "subject": "demo: avoid third-party utility in Table ajax example", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "8f6fd3f5b5d700d6b6e63e14e8e0bbc7a68cc04e", "date": "2026-09-01T22:18:58-04:00", "subject": "Pass parameters to evp_keymgmt_newdata", "body": "Update EVP keymgmt utility and export functions to pass provided parameters,\nsuch as property queries, to evp_keymgmt_newdata() instead of NULL.\n\nImplement OSSL_FUNC_KEYMGMT_NEW_EX in the ML-DSA key management implementation\nto parse and apply property query parameters during key allocation, and add\ncorresponding test coverage.\n\nAssisted-by: Gemini:Gemini 3.7 Flash\nSigned-off-by: Simo Sorce \nReviewed-by: Dmitry Belyavskiy \nReviewed-by: Paul Dale \nMerge-date: Mon Sep 7 18:53:10 2026\nMerged-from: https://github.com/openssl/openssl/pull/32636", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "ccb121e83350a06428565a5541e2ee03626e4e2e", "date": "2026-09-02T10:56:57+08:00", "subject": "fix(runtime-vapor): update custom element slot outlet name in place", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "d2a80d535ac096816c654b1b25e75d6300f367a5", "date": "2026-09-02T10:56:58+08:00", "subject": "refactor(runtime-vapor): render native custom element slot outlets on the fast path", "body": "", "style": "conventional", "type": "refactor", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "ddf5e1866853e625c3d63c99d7266229838cb124", "date": "2026-09-02T10:56:58+08:00", "subject": "refactor(runtime-vapor): render custom element light DOM slots in place", "body": "* refactor(runtime-vapor): render custom element light DOM slots in place\n\nFor shadowRoot: false, the host parsed its light DOM children and the\ncomponent still rendered native outlets that were swapped for the\nparsed nodes after mount, then the block tree was patched to point at the\nmoved DOM. That replacement only ran on the root instance's updated hook,\nso an outlet revealed by a child component update was never replaced and\nthe light DOM content was lost. Dynamic slot names had the same blind spot.\n\nHand the parsed nodes to the component as static slots instead. Outlets\nresolve them through the regular slot pipeline, which removes the\npost-mount replacement, the block tree fixup, the customElementFallback\nfield and the per-flush outlet rescan. Native outlets are now only\nrendered for hosts with a shadow root.\n\nNested shadowRoot: false elements now mount before the parent's mounted\nhook, like a regular child component.\n\n* fix(runtime-dom): derive custom element shadow root from the created root", "style": "conventional", "type": "refactor", "scope": "runtime-vapor", "breaking": false} {"repo": "vite", "sha": "93164c3530a7b4fc7bbedfb986d6afa9546cdef3", "date": "2026-09-02T12:03:54+09:00", "subject": "feat: add top-level `tsconfig` option", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "vite", "sha": "db45bb113d4c9ad894f5238c0d1343a4277d957f", "date": "2026-09-02T11:32:25+08:00", "subject": "chore: enable `minimumReleaseAgeExcludePrune`", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "vite", "sha": "b120589f052fe8e0d5ea75bade1a0278c0bdfa7f", "date": "2026-09-02T11:35:38+08:00", "subject": "chore: update debugging playwright tests section in contributing.md", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "7793cab03924d29db45d64dadc1c81d0a8cf59e0", "date": "2026-09-02T11:43:37+08:00", "subject": "docs: explain renderable content behavior in FAQ", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "postgres", "sha": "cc9a8eb112fa9cf5eb868306955198adb98e497e", "date": "2026-09-02T13:17:39+09:00", "subject": "Remove fake FunctionCallInfos from the stats restore and import code", "body": "The statistics import and restore paths built fake FunctionCallInfos\nwith LOCAL_FCINFO() and InitFunctionCallInfoData(), never calling\nanything through them. The only data relevantly used was the values\ncarried around as NullableDatum arrays, read back with PG_GETARG_*()\nmacros.\n\nimport_relation_statistics() and import_attribute_statistics() were\npassing a set of NullablePointer pointers, leading to more round-trip\nmanipulations with the fake FunctionCallInfos. These are reworked so as\nthe values pushed to pg_class (for relation stats) and pg_statistic (for\nattribute stats) are passed around in dedicated structures with all the\nvalues assigned based on positional indexes (import code cares about\nless values than the restore code), simplifying their signatures and\nreducing the number of code lines required to pass the data around.\n\nThe stats_check_*() functions now take the values and the attribute\nnames rather than a StatsArgInfo with positional indexes, as this code\nis shared between the stats restore and import. This removes the\ndependency between the StatsArgInfo bits and the import code.\nStatsArgInfo is used only for the restore code, for argument mapping.\n\nNo behavioral change should be introduced in this commit; this is\n(should be) pure refactoring.\n\nReported-by: Robert Haas \nAuthor: Corey Huinker \nAuthor: Michael Paquier \nDiscussion: https://postgr.es/m/CADkLM=eo7MtuCE=YjovW+=ASw1=q39qQ3qarrsw+EKfU901ztA@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "fc8ed6362404c2abfb87f9f291aa9d19b8323d8c", "date": "2026-09-02T13:40:14+09:00", "subject": "Remove quals made redundant by reducing outer joins to antijoins", "body": "When reduce_outer_joins reduces an outer join to an antijoin, any IS\nNULL qual on a Var from the antijoin's nullable side is necessarily\ntrue. Previously, such quals were discarded later in\ndistribute_qual_to_rels, mainly to avoid bogus selectivity estimates.\nBut that discard was incomplete: the qual remained in the jointree,\nwhile its Vars were not counted in attr_needed. Since commit\n2ebf25e7d, join removal edits the jointree and expects it to contain\nno other references to a removed rel, so it could remove a rel that\nsuch a discarded qual still references, and then trip an assertion on\nthe qual's stale Var.\n\nTo fix, move this processing to an earlier phase: such quals are now\nremoved from the jointree by reduce_outer_joins itself. This way\nlater phases see a consistent query tree, and\ncheck_redundant_nullability_qual is no longer needed, so remove it.\n\nBack-patch to v16, as with commit 2ebf25e7d.\n\nReported-by: Tender Wang \nAuthor: Richard Guo \nReviewed-by: Tom Lane \nDiscussion: https://postgr.es/m/CAHewXNk8b0TsSy4dL=CO7FXL2W3WBm0BcdP-zwNJePa-Qj4HzA@mail.gmail.com\nBackpatch-through: 16", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "94021cc30533e5d8be447571d473e9982f684525", "date": "2026-09-02T00:46:23-04:00", "subject": "tests: Test certificates with embedded zero bytes", "body": "The code removed from https://github.com/curl/curl/pull/22767 was part\nof a mess from when cURL once that strings in certificates were C\nstrings and did not have embedded zeros. See these historical commits:\n\n* c0e8bed5bf7a7e56897e492a4dcc399621939995\n* 0b66efac9cb0b606d943870237f689d71ba798f7\n* 781b82baf5e481964d67971848e4f363915ef789\n\nSince it seems there was never a test for any of this, fill in some of\nthese missing tests now.\n\n(I was working on a fix for the same strlen issue but never got around\nto uploading it. Since #22767 has since landed, I figure I may as well\ncontribute the tests.)\n\nCloses #22794", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "6544d0a39d9c0c0e89dbcf38eb4bfb9201923816", "date": "2026-09-02T14:03:50+09:00", "subject": "Fix OID-counter-sensitive row ordering in foreign_key test", "body": "Several queries in the foreign_key test sort pg_constraint rows with\n\"ORDER BY oid::regclass::text\". These OIDs are constraint OIDs, not\nrelation OIDs, so the regclass cast renders them as numeric strings,\nand text comparison of numeric strings depends on digit count. If the\nOID counter crosses a power of ten while the test runs, the output\norder flips, causing spurious test failures. This cannot happen in a\nfresh cluster, but installcheck against a long-lived cluster can hit\nit.\n\nTo fix, sort by constraint name, which is unique within each of these\nqueries, and use COLLATE \"C\" so that the order does not depend on the\ndatabase's default collation either.\n\nAuthor: Richard Guo \nReviewed-by: Tom Lane \nReviewed-by: jian he \nDiscussion: https://postgr.es/m/CAMbWs49KaJFGs4rfBWk+YmHPTXZ+pJTMJ12iaz7ruZdSK+sC5w@mail.gmail.com\nBackpatch-through: 18", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "7635a3206797a9c793bcf2e43d107c9fe08dc17d", "date": "2026-09-02T14:14:22+09:00", "subject": "Apply some query jumbling to WAIT FOR LSN", "body": "The query is now stored as normalized in pg_stat_statements, with its\nLSN position ignored in the jumbling, resulting in:\nWAIT FOR LSN $1;\n\nThe option values in the WITH clause are left untouched.\n\nReported-by: Peter Eisentraut \nAuthor: Sirisha Chamarthi \nReviewed-by: Sami Imseih \nReviewed-by: Rithvika Devisetti \nDiscussion: https://postgr.es/m/1f04f78c-2cfa-47a0-997f-f02572cd10eb@eisentraut.org\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "d3c9ccdf1edec4f486a860e279d2ef2e22f5e4f1", "date": "2026-09-02T14:35:10+09:00", "subject": "Fix checkpointer restartpoint assertion failure", "body": "When recovery starts from a backup without a signal file, pg_subtrans\nis not started at the beginning of recovery and remains unstarted\nthroughout recovery. However, previously, a restartpoint run by\nthe checkpointer during recovery nevertheless tried to truncate\npg_subtrans, triggering the assertion failure:\n\n TRAP: failed Assert(\"TransactionIdIsValid(initial)\")\n\nThis commit fixes this by tracking whether pg_subtrans has been\nstarted during recovery, and have the checkpointer check this flag\nbefore truncating pg_subtrans at restartpoints.\n\nBackpatch to all supported versions.\n\nReported-by: Imran Zaheer \nAuthor: Imran Zaheer \nAuthor: Fujii Masao \nReviewed-by: Michael Paquier \nDiscussion: https://postgr.es/m/CA+UBfamzfEReT0VOGRUW_=_AecCYQ-CNKLR_T4Q+YEmJAH3fdg@mail.gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "843434b486066f106cff787bc77d42caa2785dfe", "date": "2026-09-02T08:11:52+02:00", "subject": "warnless: dedupe Intel compiler warning suppression pragmas", "body": "Closes #22796", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "d7af4b5acab1813cce6b85d9f1284e5b48c02c32", "date": "2026-09-02T06:15:07Z", "subject": "docs: decouple selection from aria toolbar examples and guide", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "e6343c87bf1fe99fcf236e65c7e13ef1946a278e", "date": "2026-09-02T08:25:48+02:00", "subject": "Dereference variable after NULL check", "body": "Fixes coverity issue 1700802\n\nSigned-off-by: Norbert Pocs \nReviewed-by: Paul Dale \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Tomas Mraz \nMerge-date: Fri Sep 4 13:32:42 2026\nMerged-from: https://github.com/openssl/openssl/pull/32637", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "dd7c135bb97ec7a5dad09fa9e1e2dd9c044ceee8", "date": "2026-09-02T15:32:54+09:00", "subject": "Fix qual pushdown past grouping through simple CASE", "body": "Commit 44fb59fc6 taught the grouping-conflict walker to treat the arg\nof a simple CASE as a direct operand of each WHEN comparison, but it\nonly checked the collation, on the assumption that the WHEN operator\nis always the type-default \"=\" and thus matches the grouping eqop.\nThat assumption fails once the arg is relabeled to another type: the\nWHEN then compares under that type's \"=\", which need not agree with\nthe grouping equality. For instance, with a DISTINCT over a citext\ncolumn, a qual such as \"CASE t::text WHEN 'A' THEN ...\" was pushed\nbelow the Unique, although the equivalent \"t::text = 'A'\" is correctly\nkept above it.\n\nInstead of special-casing the arg, have the walker bind a Var arg\nwhile walking the WHEN conditions and resolve each CaseTestExpr to it,\nso that the arg is checked exactly as each WHEN uses it: with the\nopfamily and collation checks of a direct operand when the WHEN is a\ncomparison, and as a non-operand reference otherwise. A non-Var arg\nis walked once as a non-operand, as before. The CaseTestExpr in an\nArrayCoerceExpr's elemexpr and a JsonConstructorExpr's coercion stand\nfor something else and are left alone.\n\nReported-by: Tender Wang \nAuthor: Richard Guo \nReviewed-by: Tender Wang \nReviewed-by: Ewan Young \nDiscussion: https://postgr.es/m/CAHewXNkvGTOgijRLjmudpg=wz0d-J8ChY2o7ksh3Be+Q_Bxwog@mail.gmail.com\nBackpatch-through: 18", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "5dbdacaa04c0d09077829dd252c5ee79b6185ff6", "date": "2026-09-02T02:49:25-04:00", "subject": "vquic: fix preprocessor check for old Darwin versions", "body": "Closes #22797", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "6fdb7f061bcff114fce12b8175eb056436b6eecf", "date": "2026-09-02T06:54:43Z", "subject": "build: update dependency mocha to v12", "body": "See associated pull request for more information.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "angular", "sha": "2bc91d085aa4c2757d355ec58e699c4fdcc95fac", "date": "2026-09-02T06:57:04Z", "subject": "build: update dependency bazel to v8.8.0", "body": "See associated pull request for more information.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "angular", "sha": "d880b77967230acd9ce1c035e982c1eccee7bdce", "date": "2026-09-02T10:15:12+03:00", "subject": "build: narrow the preview exclusions for two example apps", "body": "`reactive-forms` and `form-validation` were excluded from `embeddable` in\nfull, with a TODO to fix them. The examples are fine; four partial snapshots\nquoted by the guides are not, since each shares the final template of the\ncomponent it precedes, which references members it does not have yet.\nExcluding those four files clears all 30 errors and puts 23 app files back in\nthe preview pool, so both guides' final components can be previewed for the\nfirst time.\n\n(cherry picked from commit 431b170fd562198b9bbffb079bb13c2dfd3210b1)", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "angular", "sha": "431b170fd562198b9bbffb079bb13c2dfd3210b1", "date": "2026-09-02T10:15:12+03:00", "subject": "build: narrow the preview exclusions for two example apps", "body": "`reactive-forms` and `form-validation` were excluded from `embeddable` in\nfull, with a TODO to fix them. The examples are fine; four partial snapshots\nquoted by the guides are not, since each shares the final template of the\ncomponent it precedes, which references members it does not have yet.\nExcluding those four files clears all 30 errors and puts 23 app files back in\nthe preview pool, so both guides' final components can be previewed for the\nfirst time.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "postgres", "sha": "66f85615d5d650c47b6e8ea75e32a6817cb23dac", "date": "2026-09-02T16:22:32+09:00", "subject": "Add more test coverage for pg_clear_attribute_stats()", "body": "pg_clear_attribute_stats() was called exactly once in the whole test\nsuite, only for a correct cleanup. None of its failure scenarios were\ncovered, in terms of:\n- NULL input values\n- Incorrect object names (schema, attribute, relation)\n- Incorrect relkinds\n- Inherited pg_statistic entry cleanup\n\nAll these cases are now covered. This would have caught the failure\nfixed by 11ed011ae22, as one example.\n\nAuthor: Zhong ShiHao \nDiscussion: https://postgr.es/m/CAGRkXqQEFzVh_bQSoLGDqaRTq7Xta6OUoda8PDrWDPWcAuwk=w@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "978772cf5dd642b093382e9ff4b3a353e8a5d1a2", "date": "2026-09-02T09:27:56+02:00", "subject": "pgindent fix for a63df2664225d06d5be", "body": "Oversight in commit a63df2664225d06d5be per buildfarm member koel.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "7d3247ccd15be860b098679e39eff7b6c8040355", "date": "2026-09-02T09:32:50+02:00", "subject": "Add previous commit to .git-blame-ignore-revs", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "1dd17174622dde26585b8de5e1d9e4b6ef31541d", "date": "2026-09-02T08:43:21+01:00", "subject": "Fix tests for macOS.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "d082122f4ef0542b9aeabb28a876cf24b7340c8b", "date": "2026-09-02T16:09:16+08:00", "subject": "fix(Input.Password): respect controlled visibility state", "body": "", "style": "conventional", "type": "fix", "scope": "Input.Password", "breaking": false} {"repo": "angular", "sha": "94b1d3d1b7b87d21f16263563a3f71a2d3f7b546", "date": "2026-09-02T11:13:19+03:00", "subject": "docs: correct openFiles entries that point at missing files", "body": "Two tutorial steps list a file in `openFiles` that does not exist, and a\nmissing entry is dropped without complaint. `first-app/05-inputs` asks for\n`housinglocations.ts` when the file is `housinglocation.ts`; since\n`hiddenFiles` is everything not in `openFiles`, the interface that step\nteaches was marked hidden rather than opened.\n`signals/5-component-communication-with-signals` asks for\n`quantity-selector.ts`, which exists in neither `src` nor `answer`. Every\nopenFiles entry in the tutorials now resolves.\n\n(cherry picked from commit 45148dae444dfb7c631a7dbc36fc832409639ff8)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "45148dae444dfb7c631a7dbc36fc832409639ff8", "date": "2026-09-02T11:13:19+03:00", "subject": "docs: correct openFiles entries that point at missing files", "body": "Two tutorial steps list a file in `openFiles` that does not exist, and a\nmissing entry is dropped without complaint. `first-app/05-inputs` asks for\n`housinglocations.ts` when the file is `housinglocation.ts`; since\n`hiddenFiles` is everything not in `openFiles`, the interface that step\nteaches was marked hidden rather than opened.\n`signals/5-component-communication-with-signals` asks for\n`quantity-selector.ts`, which exists in neither `src` nor `answer`. Every\nopenFiles entry in the tutorials now resolves.", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "77d831751b3452a8291256fa187553a258e2cf1e", "date": "2026-09-02T16:35:56+08:00", "subject": "fix(Table): support empty string values in tree filters", "body": "", "style": "conventional", "type": "fix", "scope": "Table", "breaking": false} {"repo": "vue-core", "sha": "ee99fe5262fbeb44b330f644ac75cf58fc9535de", "date": "2026-09-02T16:36:03+08:00", "subject": "fix(runtime-vapor): key async component branches by state", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "c3fb7fffe4aed1a85d92915c2dc8024558b9fadb", "date": "2026-09-02T16:36:04+08:00", "subject": "refactor(runtime-vapor): create resolved async components without a wrapper", "body": "* refactor(runtime-vapor): create resolved async components without a wrapper\n\n* chore: minor tweaks\n\n* test: update tests\n\n* chore: update", "style": "conventional", "type": "refactor", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "00657e37669df8e818bb2ce1f5c16c7d81643908", "date": "2026-09-02T16:36:04+08:00", "subject": "fix(runtime-vapor): apply hydration strategies to resolved async components", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "29fa76acdea3d8dabe899bc35551bbce54baff9d", "date": "2026-09-02T16:36:05+08:00", "subject": "fix(runtime-vapor): keep async wrapper hooks across deferred hydration", "body": "* fix(runtime-vapor): keep async wrapper hooks across deferred hydration\n\n* refactor(runtime-vapor): tidy the async component paths\n\n* chore: minor tweaks", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "eb3f096c4d500d6d4875f4f94d9c3aa904e985f3", "date": "2026-09-02T16:36:05+08:00", "subject": "fix(runtime-vapor): skip rendering an async wrapper unmounted while loading under Suspense", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "783cc7a318b067685e6a73374a6b8482a6c288e6", "date": "2026-09-02T16:36:06+08:00", "subject": "perf(runtime-vapor): gate async wrapper branches for tree-shaking", "body": "* perf(runtime-vapor): gate async wrapper branches for tree-shaking\n\n* chore: update", "style": "conventional", "type": "perf", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "f6479b277a624b10967599846789b95b6cd34b35", "date": "2026-09-02T16:36:06+08:00", "subject": "refactor(runtime-core): make the lazy hydration \"patched\" guard opt-in", "body": "", "style": "conventional", "type": "refactor", "scope": "runtime-core", "breaking": false} {"repo": "neovim", "sha": "cccf3598514403a96c14aff6aba180efdbe471a3", "date": "2026-09-02T16:54:00+08:00", "subject": "build: disable some clang-tidy readability checks #41556", "body": "Problem:\nNew clang-tidy checks readability-trailing-comma and\nreadability-redundant-parentheses fire across the codebase.\n\nSolution:\nDisable readability-trailing-comma and readability-redundant-parentheses\nchecks in the clang-tidy configuration.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "neovim", "sha": "47ce0d1f87e0a80bf4920d83be824aec6a9d3f50", "date": "2026-09-02T04:54:35-04:00", "subject": "build(vim-patch): v9.1.0578 is n/a #41585", "body": "Nvim implementation of \":TOhtml\" fails on test_tohtml.vim\nbecause they compare with the sample test files.\nCustomizing the sample files to pass Vim's test is maintenance burden.\nNvim has lua functional tests to skip Vim's tests.\n\nNvim's test Makefile computes NEW_TESTS, NEW_TESTS_RES, unlike Vim.\nLikely N/A.\n\nFollowing runtime/ files are N/A after Lua rewrite.\n\n- runtime/autoload/tohtml.vim\n- runtime/syntax/2html.vim\n- runtime/plugin/tohtml.vim", "style": "conventional", "type": "build", "scope": "vim-patch", "breaking": false} {"repo": "tmux", "sha": "ad632726bda806d382e075c17e0161734894d1fa", "date": "2026-09-02T11:16:37+01:00", "subject": "Some more regress tests.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "ed131747c1a3dc7f816d26724c44537ac8acba44", "date": "2026-09-02T12:25:50+02:00", "subject": "x509: stricter parsing", "body": "Check that the \"validity\" ASN.1 field has a length matching the\ncontained date elements and fail the parsing otherwise.\n\nThis is a precaution against malformed cert data that should have been\ncaught by the TLS backend already. A bit added defense.\n\nCloses #22799", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "fe67952d34b9582df3cdf95a7bf62ca9ced5ef09", "date": "2026-09-02T12:50:08+02:00", "subject": "test1399: improve reliability", "body": "Disconnect test1399 from the system clock and provide\nfixed timestamps for recording progress and checking\nrecorded timer values.\n\nFixes #22798\nReported-by: Christian Hesse\nCloses #22800", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "9fa390aee696f7c3886bc9146e4051cbbe9593cb", "date": "2026-09-02T12:32:29+01:00", "subject": "Fix error in test.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "1c31526a90d342253d07cb4e672361baf867be9e", "date": "2026-09-02T13:45:52+02:00", "subject": "fix(vim.hl): convert finish column independently #41618", "body": "Problem:\nCoordinates-to-position conversion for the finish column in\n`vim.hl.range()` checks the start column. This results in an incorrect\nfinish column when only one of the column coordinates is `vim.v.maxcol`.\n\nSolution:\nCheck the finish column when converting the finish position. Update the\nexisting screen test so that a `vim.v.maxcol` finish highlights the\nend-of-line marker, matching the existing `-1` behavior.", "style": "conventional", "type": "fix", "scope": "vim.hl", "breaking": false} {"repo": "neovim", "sha": "a7101efcafafc76cb540a1b988e751ca39c20845", "date": "2026-09-02T14:08:05+02:00", "subject": "fix(health): guard against nil result from vim.system:wait()", "body": "vim.system():wait() can return nil when timeout is exceeded during\nsigkill (see #37922). check_stable_version() and check_head_hash()\ndirectly indexed result.code without a nil guard, causing:\n\n .../runtime/lua/vim/health/health.lua:560: attempt to index local\n 'result' (a nil value)\n\nwhen :checkhealth runs and the `git ls-remote` call times out (e.g.\noffline / firewalled). Fix by adding `not result or` to both guards,\nmatching the workaround already on master:\n\n if not result then return false, 'command failed' end\n\nin runtime/lua/vim/health/health.lua:system().\n\nRepro: `vim.system({\"git\",\"ls-remote\",...}, {timeout=1}):wait()` returns\nnil; patched code returns early instead of throwing.", "style": "conventional", "type": "fix", "scope": "health", "breaking": false} {"repo": "curl", "sha": "e470fcf1ec525cbaf0c455218ae774bd68e640fd", "date": "2026-09-02T14:09:30+02:00", "subject": "CVE-2026-82208: wolfSSL CA-cache hit overrides callback", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "fe4cd2babab4359313cef9d76dcc2eda37a69fdc", "date": "2026-09-02T14:12:07+02:00", "subject": "fix(health): restore pynvim version lookup #41619", "body": "Problem:\nFallback pynvim version lookup is broken since Vimscript-to-Lua rewrite.\n- `vim.fs.basename()` returns the module filename instead of the\n directory needed to discover adjacent metadata. This produces an\n empty metadata list, hiding the remaining issues.\n- `table.sort()` requires a Boolean comparator, while\n `vim.version.cmp()` returns a number.\n- `table.sort()` sorts in place and returns no value, so assigning its\n result discards the metadata list.\n\nSolution:\n- Use `vim.fs.dirname()` to discover adjacent package metadata.\n- Use `vim.version.gt()` as the descending Boolean comparator.\n- Sort the metadata list in place.\nThis restores fallback version detection when `neovim.VERSION` is\nunavailable.\n\n(cherry picked from commit 3ea7bc3f607bdc08940698e7d049036e1f788c66)", "style": "conventional", "type": "fix", "scope": "health", "breaking": false} {"repo": "neovim", "sha": "3ea7bc3f607bdc08940698e7d049036e1f788c66", "date": "2026-09-02T14:12:07+02:00", "subject": "fix(health): restore pynvim version lookup #41619", "body": "Problem:\nFallback pynvim version lookup is broken since Vimscript-to-Lua rewrite.\n- `vim.fs.basename()` returns the module filename instead of the\n directory needed to discover adjacent metadata. This produces an\n empty metadata list, hiding the remaining issues.\n- `table.sort()` requires a Boolean comparator, while\n `vim.version.cmp()` returns a number.\n- `table.sort()` sorts in place and returns no value, so assigning its\n result discards the metadata list.\n\nSolution:\n- Use `vim.fs.dirname()` to discover adjacent package metadata.\n- Use `vim.version.gt()` as the descending Boolean comparator.\n- Sort the metadata list in place.\nThis restores fallback version detection when `neovim.VERSION` is\nunavailable.", "style": "conventional", "type": "fix", "scope": "health", "breaking": false} {"repo": "curl", "sha": "38703fe2610bf6e79390bfc93a8558ffc0e2b02c", "date": "2026-09-02T14:22:17+02:00", "subject": "multihandle: some easy shrinks", "body": "- new `struct Curl_fixed_buf` keeping data and length in one alloc\n for upload/download/socket buffers held in the multi\n- max_host_connections changed to uint32_t\n- max_total_connections changed to uint32_t\n- maxconnects changed to uint32_t\n- last_timeout_ms changed to timediff_t\n- now_access_count removed, no one used it\n\nCloses #22801", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "1935d215b5c6bc8619342e6271cfd52837517112", "date": "2026-09-02T14:38:46+02:00", "subject": "Add metadata support to EVP_SKEY objects", "body": "Assisted-by: Claude Opus 4.6 \n\nReviewed-by: Mounir Idrassi \nReviewed-by: Simo Sorce \nMergeDate: Sat Sep 5 17:06:55 2026\n(Merged from https://github.com/openssl/openssl/pull/32644)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "0fb258d1f2ca65b7299ba7c47045634df05efbff", "date": "2026-09-02T15:46:47+03:00", "subject": "postgres_fdw: Fix \"may be used uninitialized\" warning in foreign_join_ok()", "body": "Commit 0ee83dd4a99 detected the mixed foreign x function-RTE cases in one\nif/else chain that only set a pair of bool flags, then acted on those flags in\na second chain. fpinfo_o and fpinfo_i are each assigned in one chain and read\nin the other, so proving them initialized requires correlating the flags with\nthe assignments. gcc 13 with -Og does not manage that and reports fpinfo_i as\npossibly uninitialized; clang's -Wconditional-uninitialized likewise reports\nboth variables.\n\nThe flags served no purpose beyond deferring the work, so merge the two chains\nand read both fdw_private pointers up front, unconditionally. A function rel\nnever has an fdw_private of its own, so the pointer is simply NULL until the\nstub replaces it, and each branch now assigns and uses the two variables in one\nplace.\n\nReported-by: Karina Litskevich \nReported-by: Marina Polyakova \nReviewed-by: Marina Polyakova \nDiscussion: https://postgr.es/m/CACiT8iamoL-%3D792e6JK2uCVvwDhfCg7DXHhvqN0P33jhGg2-8A%40mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "cccdee60bf95aff4addd12a43d9fe2a87f350e1b", "date": "2026-09-02T14:54:28+02:00", "subject": "Do not advertise OPENSSL_INIT_LOAD_CONFIG in optsdone too soon", "body": "If there was no explicit call to OPENSSL_init_crypto before starting\nthreading, it would result in a race between loading the config in\none thread, and attempting to use features dependent on the config,\ne.g., providers in other threads.\n\nThe first thread that gets into OPENSSL_init_crypto would start loading\nthe config and set `loading` to 1 The same thread may re-enter the function\nrecursively and should do nothing, but it would actually update `optsdone`\nwith OPENSSL_INIT_LOAD_CONFIG, even though the loading has not finished yet.\nThis commit fixes the issue by masking the opts bit out if we re-enter\nOPENSSL_init_crypto.\n\nAssisted-by: Claude:claude-opus-4.8\n(cherry picked from commit d115fd05e3986e90db18776ad49103fadccbe52f)\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Dmitry Belyavskiy \nMergeDate: Mon Sep 7 08:37:37 2026\n(Merged from https://github.com/openssl/openssl/pull/32660)\n\n(cherry picked from commit 95f41f0ee0eafe1a694ae4607a18679f63b8299e)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "95f41f0ee0eafe1a694ae4607a18679f63b8299e", "date": "2026-09-02T14:54:28+02:00", "subject": "Do not advertise OPENSSL_INIT_LOAD_CONFIG in optsdone too soon", "body": "If there was no explicit call to OPENSSL_init_crypto before starting\nthreading, it would result in a race between loading the config in\none thread, and attempting to use features dependent on the config,\ne.g., providers in other threads.\n\nThe first thread that gets into OPENSSL_init_crypto would start loading\nthe config and set `loading` to 1 The same thread may re-enter the function\nrecursively and should do nothing, but it would actually update `optsdone`\nwith OPENSSL_INIT_LOAD_CONFIG, even though the loading has not finished yet.\nThis commit fixes the issue by masking the opts bit out if we re-enter\nOPENSSL_init_crypto.\n\nAssisted-by: Claude:claude-opus-4.8\n(cherry picked from commit d115fd05e3986e90db18776ad49103fadccbe52f)\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Dmitry Belyavskiy \nMergeDate: Mon Sep 7 08:37:37 2026\n(Merged from https://github.com/openssl/openssl/pull/32660)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "457c55324724f385e968eeee00955b8b1bec8910", "date": "2026-09-02T14:54:28+02:00", "subject": "Do not advertise OPENSSL_INIT_LOAD_CONFIG in optsdone too soon", "body": "If there was no explicit call to OPENSSL_init_crypto before starting\nthreading, it would result in a race between loading the config in\none thread, and attempting to use features dependent on the config,\ne.g., providers in other threads.\n\nThe first thread that gets into OPENSSL_init_crypto would start loading\nthe config and set `loading` to 1 The same thread may re-enter the function\nrecursively and should do nothing, but it would actually update `optsdone`\nwith OPENSSL_INIT_LOAD_CONFIG, even though the loading has not finished yet.\nThis commit fixes the issue by masking the opts bit out if we re-enter\nOPENSSL_init_crypto.\n\nAssisted-by: Claude:claude-opus-4.8\n\nReviewed-by: Neil Horman \nReviewed-by: Dmitry Belyavskiy \nMergeDate: Fri Sep 4 07:15:24 2026\n(Merged from https://github.com/openssl/openssl/pull/32646)\n\n(cherry picked from commit 7122f8141674140654a1b8a9cc7cd790b9fbb9dd)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "7122f8141674140654a1b8a9cc7cd790b9fbb9dd", "date": "2026-09-02T14:54:28+02:00", "subject": "Do not advertise OPENSSL_INIT_LOAD_CONFIG in optsdone too soon", "body": "If there was no explicit call to OPENSSL_init_crypto before starting\nthreading, it would result in a race between loading the config in\none thread, and attempting to use features dependent on the config,\ne.g., providers in other threads.\n\nThe first thread that gets into OPENSSL_init_crypto would start loading\nthe config and set `loading` to 1 The same thread may re-enter the function\nrecursively and should do nothing, but it would actually update `optsdone`\nwith OPENSSL_INIT_LOAD_CONFIG, even though the loading has not finished yet.\nThis commit fixes the issue by masking the opts bit out if we re-enter\nOPENSSL_init_crypto.\n\nAssisted-by: Claude:claude-opus-4.8\n\nReviewed-by: Neil Horman \nReviewed-by: Dmitry Belyavskiy \nMergeDate: Fri Sep 4 07:15:24 2026\n(Merged from https://github.com/openssl/openssl/pull/32646)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "c2921730a753577832502f89223edfa67420f5cf", "date": "2026-09-02T14:58:59+02:00", "subject": "perf/base64: split into separate encoding plus decoding tests", "body": "Closes #22802", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "ea3c0e962aee86f0d45bd01edde7b41663b811e9", "date": "2026-09-02T15:06:44+02:00", "subject": "tests/perf: split \"percent\" into urlencode + urldecode", "body": "Closes #22803", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "9d50010a551a7bd3eea0c289b1e8e675e303a560", "date": "2026-09-02T15:26:51+02:00", "subject": "Check return code of BIO_ADDR_rawmake", "body": "Resolves coverity issue 1700561.\n\nSigned-off-by: Norbert Pocs \nReviewed-by: Neil Horman \nReviewed-by: Jakub Zelenka \nMerge-date: Wed Sep 16 09:12:45 2026\nMerged-from: https://github.com/openssl/openssl/pull/32661", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "aed91e349b20f170d4da031d9b5d08b609e7bba6", "date": "2026-09-02T15:26:51+02:00", "subject": "Check return code of BIO_ADDR_rawmake", "body": "Resolves coverity issue 1700561.\n\nSigned-off-by: Norbert Pocs \nReviewed-by: Neil Horman \nReviewed-by: Jakub Zelenka \nMerge-date: Wed Sep 16 09:12:34 2026\nMerged-from: https://github.com/openssl/openssl/pull/32661", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "fff887cdee2bd1ec78591d1dc14fc1e38f8583a6", "date": "2026-09-02T15:26:51+02:00", "subject": "Check return code of BIO_ADDR_rawmake", "body": "Resolves coverity issue 1700561.\n\nSigned-off-by: Norbert Pocs \nReviewed-by: Neil Horman \nReviewed-by: Jakub Zelenka \nMerge-date: Wed Sep 16 09:12:23 2026\nMerged-from: https://github.com/openssl/openssl/pull/32661", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "66616ef7a9e31adc38207f44ff609ca3ca47f76d", "date": "2026-09-02T15:26:51+02:00", "subject": "Check return code of BIO_ADDR_rawmake", "body": "Resolves coverity issue 1700561.\n\nSigned-off-by: Norbert Pocs \nReviewed-by: Neil Horman \nReviewed-by: Jakub Zelenka \nMerge-date: Wed Sep 16 09:12:11 2026\nMerged-from: https://github.com/openssl/openssl/pull/32661", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "c7439731f32818fff6b548b8080eeb53fb78af7e", "date": "2026-09-02T15:26:51+02:00", "subject": "Check return code of BIO_ADDR_rawmake", "body": "Resolves coverity issue 1700561.\n\nSigned-off-by: Norbert Pocs \nReviewed-by: Neil Horman \nReviewed-by: Jakub Zelenka \nMerge-date: Wed Sep 16 09:11:30 2026\nMerged-from: https://github.com/openssl/openssl/pull/32661", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "dd6c536ba96147082c18c0ce6e1bb69961e80e3a", "date": "2026-09-02T15:26:51+02:00", "subject": "Check return code of BIO_ADDR_rawmake", "body": "Resolves coverity issue 1700561.\n\nSigned-off-by: Norbert Pocs \nReviewed-by: Neil Horman \nReviewed-by: Jakub Zelenka \nMerge-date: Wed Sep 16 09:11:08 2026\nMerged-from: https://github.com/openssl/openssl/pull/32661", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "d67c45b532d0ec08c2f15591d8ac7fb27ae4cc63", "date": "2026-09-02T15:41:32+02:00", "subject": "Free WPACKETs on error paths", "body": "Fixes coverity issue 1700560.\n\nSigned-off-by: Norbert Pocs \nReviewed-by: Neil Horman \nReviewed-by: Jakub Zelenka \nMerge-date: Tue Sep 15 10:16:30 2026\nMerged-from: https://github.com/openssl/openssl/pull/32663", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "c4b4be0f6c737ab27c59aff61b682d7c29572bda", "date": "2026-09-02T15:41:32+02:00", "subject": "Free WPACKETs on error paths", "body": "Signed-off-by: Norbert Pocs \nReviewed-by: Neil Horman \nReviewed-by: Jakub Zelenka \nMerge-date: Tue Sep 15 10:13:37 2026\nMerged-from: https://github.com/openssl/openssl/pull/32665", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "daa0931f0389ee876a2ec7a82c137a25d2a325a9", "date": "2026-09-02T15:41:32+02:00", "subject": "Free WPACKETs on error paths", "body": "Signed-off-by: Norbert Pocs \nReviewed-by: Neil Horman \nReviewed-by: Jakub Zelenka \nMerge-date: Tue Sep 15 10:12:48 2026\nMerged-from: https://github.com/openssl/openssl/pull/32664", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "a5f8ec43ecac0591120ad0892755a1c19c2e9be4", "date": "2026-09-02T15:41:32+02:00", "subject": "Free WPACKETs on error paths", "body": "Signed-off-by: Norbert Pocs \nReviewed-by: Neil Horman \nReviewed-by: Jakub Zelenka \nMerge-date: Tue Sep 15 10:12:34 2026\nMerged-from: https://github.com/openssl/openssl/pull/32664", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "b4a5eda669d7dd1c7fbc5cb3af23449807f9218c", "date": "2026-09-02T15:41:32+02:00", "subject": "Free WPACKETs on error paths", "body": "Signed-off-by: Norbert Pocs \nReviewed-by: Neil Horman \nReviewed-by: Jakub Zelenka \nMerge-date: Tue Sep 15 10:12:18 2026\nMerged-from: https://github.com/openssl/openssl/pull/32664", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "e4aeafa81d19f5d0bfd12d1875ae9eac151332c4", "date": "2026-09-02T15:41:32+02:00", "subject": "Free WPACKETs on error paths", "body": "Fixes coverity issue 1700560.\n\nSigned-off-by: Norbert Pocs \nReviewed-by: Neil Horman \nReviewed-by: Jakub Zelenka \nMerge-date: Tue Sep 15 10:11:12 2026\nMerged-from: https://github.com/openssl/openssl/pull/32663", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "350fa5ad7ccc014bcdd0eea2bd76ba09a265452e", "date": "2026-09-02T09:42:40-04:00", "subject": "fix(multicursor): act on fold contents, not the fold itself #41622", "body": "Problem:\nIf an edit cascades to cursor(s) that happen to be within a closed fold,\nthe edit acts on the fold itself (`:h fold-behavior`). E.g. \"dd\" deletes\nthe entire fold, not the cursor's line within the fold.\n\nSolution:\nCheck `mc_replaying()` in `hasAnyFolding()`, so folds are ignored during\na multicursor cascade. This means multicursor always replays relative to\ntext *within* the fold.", "style": "conventional", "type": "fix", "scope": "multicursor", "breaking": false} {"repo": "vitest", "sha": "eed5bd5ce120acd223c23a113ebbbd9696617833", "date": "2026-09-02T16:02:31+02:00", "subject": "docs: write 'em", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "neovim", "sha": "5bc7dbb13ecb644e9cff2468503f826cfca01483", "date": "2026-09-02T16:08:37+02:00", "subject": "fix(loader): remove all cached loaders on disable #41613", "body": "Problem:\nWhen disabling, the loader list is traversed forwards while entries are\nremoved. As a result, some entries are skipped and remain active. Thus,\ndisabling `vim.loader` restores `_G.loadfile`, but does not fully\nrestore the original package loader chain.\n\nSolution:\nTraverse `package.loaders` in reverse when removing cached loaders.\nCapture the original `package.loaders` list in the test, and assert that\nthe list is restored properly.", "style": "conventional", "type": "fix", "scope": "loader", "breaking": false} {"repo": "neovim", "sha": "a41d00866676a70358d5568ef3c6e6b21f1d0c43", "date": "2026-09-02T16:09:20+02:00", "subject": "fix(trust): hash fileformat=mac buffers correctly #41616", "body": "Problem:\n`compute_hash()` reconstructs buffer contents with `fileformat` set to\n`mac` using CRLF (`dos`) line endings. The resulting hash does not match\nthe file's bytes.\n\nSolution:\nDefine the line ending for every supported file format. Use that mapping\nwhen reconstructing buffer contents for hashing.", "style": "conventional", "type": "fix", "scope": "trust", "breaking": false} {"repo": "neovim", "sha": "8321941ff20e5583e2190cdfc532352ff4441914", "date": "2026-09-02T16:10:18+02:00", "subject": "fix(vim.version): fix metatable for intersections #41615", "body": "Problem:\n`vim.version.intersect()` returns tables with `VersionRange` (method\ntable) as their metatable. Although the calculated bounds are correct,\nthe results do not expose `VersionRange` methods.\n\nSolution:\nConstruct intersection tables with `range_mt`, matching\n`vim.version.range()`. Add assertions covering method access and\nrepeated intersection.", "style": "conventional", "type": "fix", "scope": "vim.version", "breaking": false} {"repo": "postgres", "sha": "d870be195b3b7ee224a58ff562c1f1c252a96814", "date": "2026-09-02T09:22:40-05:00", "subject": "initdb: Pad rewritten GUC lines with spaces instead of tabs.", "body": "Commit f63ae72bbc converted postgresql.conf.sample to spaces so\nthat it lines up at any tab width, but replace_guc_value() still\npads with tabs when it re-aligns the trailing comment of a\nrewritten setting. Fix that.\n\nOversight in commit f63ae72bbc.\n\nReviewed-by: Chao Li \nDiscussion: https://postgr.es/m/apRezMywHRuR5h-6%40nathan\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "80e9708afa2a4e46967e5961dfc6a49fea84a118", "date": "2026-09-02T10:38:27-04:00", "subject": "fix(multicursor): atom may cascade in another buffer #41624", "body": "Problem:\nAn atom queued in buffer A cascades on B's cursors if the mapping ends\nin B (\"nnoremap X x:bnext\").\n\nSolution:\nCheck the atom's origin buffer (`CmdAtom.origin.buf`).\n\nNote: This does not preclude mappings etc from doing work in temporary\nthrowaway buffers, as long as they return to the origin buffer.", "style": "conventional", "type": "fix", "scope": "multicursor", "breaking": false} {"repo": "postgres", "sha": "61636c17b391bb5b93dbeeed24553198187752fc", "date": "2026-09-02T10:53:35-04:00", "subject": "Stabilize xid_wraparound/t/002_limits.pl test.", "body": "Not many buildfarm members run this test, since it's gated behind\nPG_TEST_EXTRA=xid_wraparound, but of those that do, the slower ones\nnot infrequently fail. The reason seems to be that the test expects\nan INSERT command to either succeed or fail, but there's a window\nwhere it can succeed while issuing a warning about impending\nwraparound. poll_query_until treats nonempty stderr as a failure,\nso it loops an extra time until the INSERT succeeds with no warning.\nThere seems no reason to treat this behavior as wrong, so adjust\nthe test to accept it.\n\nReported-by: Alexander Lakhin \nDiagnosed-by: Alexander Lakhin \nAuthor: Tom Lane \nDiscussion: https://postgr.es/m/7a68bee2-91bd-481d-be44-160e6beebc83@gmail.com\nBackpatch-through: 17", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "a8c5b98889019dfb7d76b6a020d36c51c28b18bf", "date": "2026-09-02T16:58:40+02:00", "subject": "chore: rename type", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "neovim", "sha": "c75aae2bfbfa06e119300c295ca5081557538dd6", "date": "2026-09-02T11:20:55-04:00", "subject": "fix(multicursor): 'autocomplete' during insert-cascade raises E565 #41625", "body": "Problem:\nWith mcursors and 'autocomplete', a non-literal key (``) during an\nactive completion attempts to insert-cascade, which attempts a nested\n`edit()`, which refuses with E565.\n\nSolution:\nDefer the flush while completion is active; the pending keys will be\nhandled later.", "style": "conventional", "type": "fix", "scope": "multicursor", "breaking": false} {"repo": "curl", "sha": "19d0f4acfa9df373e8e1f987c11b7165e1edac34", "date": "2026-09-02T18:51:40+03:30", "subject": "tool: (re)set SSL verification in --libcurl output", "body": "So that it shows it correctly even if toggled on/off between --next.\n\nFixes #22701\nCloses #22745", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "9e52bfd8d09b9adf10cb5e5810f2a35ae0f3ee7d", "date": "2026-09-02T09:55:32-06:00", "subject": "Convert the paths for generator scripts on windows.", "body": "jom as used in CI apparenty handles this for us, but\nnmake (at least in older versions) does not.\n\nwe passed CI because it appears to only exercise the jom\npath on windows.\n\nIf we convert the slashes in the paths ourselves it works for\nboth.\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nMerge-date: Mon Sep 7 18:21:29 2026\nMerged-from: https://github.com/openssl/openssl/pull/32648", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "b01224a3774c17959cbaf805fc4f79ffe90398c9", "date": "2026-09-02T16:56:24+01:00", "subject": "multi_ntfy: replace enabled bitset with inline flags", "body": "Store enabled notification types in the low bits of a uint32_t and use\nthe high bit for the pending-entry state.\n\nThis reduces curl_multi_ntfy to 40 bytes and removes an allocation from\nmulti-handle initialization.\n\nCloses #22804", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "f1d89e874a00c9286b9c7382b30817d61a185b27", "date": "2026-09-02T12:00:54-04:00", "subject": "fix(multicursor): 'operatorfunc' motion not followed #41627", "body": "Problem:\nWith follow-mode (\"q=\"), an 'operatorfunc' that only moves the cursor\ndoes not cascade.\n\nSolution:\nUpdate the `effect` condition to include cursor-moves if follow-mode is\nactive.", "style": "conventional", "type": "fix", "scope": "multicursor", "breaking": false} {"repo": "openssl", "sha": "3a9d737c152c59fddb2d04e3eb7a343121f963ec", "date": "2026-09-02T12:07:01-04:00", "subject": "Fix AES-GCM tag length query and the test meant to cover it", "body": "Querying OSSL_CIPHER_PARAM_AEAD_TAG with a NULL buffer returned a length\nof 0 instead of the tag length. Report ctx->taglen in that case, matching\nthe IV queries.\n\nThe test had a missing \"goto err\", meaning the test passed only when the\nassertion failed. Add the goto.\n\nBoth were introduced in #28232\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Neil Horman \nReviewed-by: Bob Beck \nReviewed-by: Paul Dale \nMerge-date: Mon Sep 7 18:55:12 2026\nMerged-from: https://github.com/openssl/openssl/pull/32649\n(cherry picked from commit 223e04f993f0df5c81358715eb7ed243ac31b1a6)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "223e04f993f0df5c81358715eb7ed243ac31b1a6", "date": "2026-09-02T12:07:01-04:00", "subject": "Fix AES-GCM tag length query and the test meant to cover it", "body": "Querying OSSL_CIPHER_PARAM_AEAD_TAG with a NULL buffer returned a length\nof 0 instead of the tag length. Report ctx->taglen in that case, matching\nthe IV queries.\n\nThe test had a missing \"goto err\", meaning the test passed only when the\nassertion failed. Add the goto.\n\nBoth were introduced in #28232\n\nAssisted-by: Claude:claude-fable-5\nReviewed-by: Neil Horman \nReviewed-by: Bob Beck \nReviewed-by: Paul Dale \nMerge-date: Mon Sep 7 18:55:12 2026\nMerged-from: https://github.com/openssl/openssl/pull/32649", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "de25b07fe444aa9abd5992cacc72fc36c74448d6", "date": "2026-09-02T18:07:16+02:00", "subject": "fix: use `keyDef.repeat` as starting point for tlast keydef", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "openssl", "sha": "e9e85dd09d0eabc9e5437007c6bd033d4ef8b6b1", "date": "2026-09-02T18:12:56+02:00", "subject": "Provider compat CI: Force copy of fips.so and fipsmodule.cnf", "body": "If the target branch is >4.0 the destination files\ncan be read-only.\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Norbert Pocs \nMerge-date: Mon Sep 7 18:16:30 2026\nMerged-from: https://github.com/openssl/openssl/pull/32662\n(cherry picked from commit e80b14876a69a758d2eb2a916937d846cba3967b)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "e80b14876a69a758d2eb2a916937d846cba3967b", "date": "2026-09-02T18:12:56+02:00", "subject": "Provider compat CI: Force copy of fips.so and fipsmodule.cnf", "body": "If the target branch is >4.0 the destination files\ncan be read-only.\n\nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Norbert Pocs \nMerge-date: Mon Sep 7 18:16:30 2026\nMerged-from: https://github.com/openssl/openssl/pull/32662", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "acfedc444cd02ea745df1feaa1309f19eeccea62", "date": "2026-09-02T18:12:56+02:00", "subject": "Provider compat CI: Force copy of fips.so and fipsmodule.cnf", "body": "If the target branch is >4.0 the destination files\ncan be read-only.\n\nReviewed-by: Viktor Dukhovni \nReviewed-by: Eugene Syromiatnikov \nMerge-date: Mon Sep 7 18:14:53 2026\nMerged-from: https://github.com/openssl/openssl/pull/32675", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "73c0180cb6823e01518a33e74ec0a6ce7f80d4d4", "date": "2026-09-02T17:16:59+01:00", "subject": "multi: remove redundant completion message list", "body": "Use the msgsent bitset to track unread application messages and scan it\nfrom curl_multi_info_read(). Remove entries when read or when their easy\nhandle is removed.\n\nThis removes the linked-list node from each easy handle. Completion order\nis not API-defined.\n\nAdd regression coverage for CURLMINFO_XFERS_DONE.\n\nCloses #22805", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "2f897694d8526000bc0283b3660476d21171fa93", "date": "2026-09-02T17:38:55+01:00", "subject": "multi: store timeout epoch as whole seconds", "body": "The timeout splay tree stores its keys as offsets from a fixed epoch.\nDropping the epoch's microseconds shifts every offset by the same\nconstant, which cancels when calculating timeout deltas.\n\nStore only the epoch's seconds and reconstruct a zero-microsecond\ncurltime for offset calculations. This reduces Curl_timeouts and\nCurl_multi by eight bytes on 64-bit builds without changing timeout\nordering or durations.\n\nCloses #22806", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "5ff30209d5fb2df85f1896d039b58551980e0c45", "date": "2026-09-02T10:51:57-06:00", "subject": "copyright.pm: spell the null device the platform's way", "body": "year_of() runs git through the platform shell with \"2>/dev/null\".\nUnder cmd.exe that is a redirect into a nonexistent \\dev directory,\nso every generator run printed \"The system cannot find the path\nspecified.\" twice and the copyright year silently fell back to the\ncurrent year. Use File::Spec->devnull() instead.\n\nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nMerge-date: Mon Sep 7 18:21:29 2026\nMerged-from: https://github.com/openssl/openssl/pull/32648", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "eb861a9d7782a78155b533856e956ccff1cc2dbc", "date": "2026-09-02T11:16:13-06:00", "subject": "Add a CI job to build once with Microsoft's nmake, not only jom", "body": "Every MSVC job builds with jom, which spawns commands differently from\nnmake, so makefile rules that nmake handles differently were never\nexercised; the new GENERATE rules for crypto/objects broke under nmake\nwhile CI stayed green. Add a build-only job that configures in-tree,\nas NOTES-WINDOWS.md describes, and builds with nmake, with assembly\nenabled so the perlasm/nasm generator rules are covered too.\n\nReviewed-by: Andrew Dinh \nReviewed-by: Neil Horman \nReviewed-by: Richard Levitte \nMerge-date: Tue Sep 15 10:18:53 2026\nMerged-from: https://github.com/openssl/openssl/pull/32651", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "75d20e6d0ffc4dbbd6c024f791ed24c5dbc8c3f0", "date": "2026-09-02T11:16:13-06:00", "subject": "Add a CI job to build once with Microsoft's nmake, not only jom", "body": "Every MSVC job builds with jom, which spawns commands differently from\nnmake, so makefile rules that nmake handles differently were never\nexercised; the new GENERATE rules for crypto/objects broke under nmake\nwhile CI stayed green. Add a build-only job that configures in-tree,\nas NOTES-WINDOWS.md describes, and builds with nmake, with assembly\nenabled so the perlasm/nasm generator rules are covered too.\n\nReviewed-by: Andrew Dinh \nReviewed-by: Neil Horman \nReviewed-by: Richard Levitte \nMerge-date: Tue Sep 15 10:01:16 2026\nMerged-from: https://github.com/openssl/openssl/pull/32651", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "eab7d91ade03c58e612b95055899aa12c2d87443", "date": "2026-09-02T18:24:18+01:00", "subject": "multi: reduce struct size with small cleanups", "body": "Use byte-sized multi API call-stack fields, remove the unused uint32\ntable destructor and replace the last timeout value with a bit.\n\nThis reduces Curl_multi by 32 bytes on 64-bit builds.\n\nCloses #22807", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "5e621df24bc2d14567dae5de7602c748eb67597d", "date": "2026-09-02T13:37:20-05:00", "subject": "fix(platform-server): preserve Unicode whitespace in ServerXhr URLs", "body": "Avoid String.prototype.trim() when classifying server request URLs because it\nstrips Unicode whitespace that WHATWG URL parsing preserves as path data.\nThis mismatch can cause xhr2 to send credentials to a different origin than\nthe application approved.\n\nClassify the unmodified request URL so whitespace-prefixed input falls through\nURL resolution, and cover NBSP- and BOM-prefixed HTTP URLs.\n\nFixes #70540\n\n(cherry picked from commit 44355181e25905c79045e6344e73c5648ed1d2a4)", "style": "conventional", "type": "fix", "scope": "platform-server", "breaking": false} {"repo": "angular", "sha": "44355181e25905c79045e6344e73c5648ed1d2a4", "date": "2026-09-02T13:37:20-05:00", "subject": "fix(platform-server): preserve Unicode whitespace in ServerXhr URLs", "body": "Avoid String.prototype.trim() when classifying server request URLs because it\nstrips Unicode whitespace that WHATWG URL parsing preserves as path data.\nThis mismatch can cause xhr2 to send credentials to a different origin than\nthe application approved.\n\nClassify the unmodified request URL so whitespace-prefixed input falls through\nURL resolution, and cover NBSP- and BOM-prefixed HTTP URLs.\n\nFixes #70540", "style": "conventional", "type": "fix", "scope": "platform-server", "breaking": false} {"repo": "postgres", "sha": "792094a5ce89f9953d913c835e4530e56b63026a", "date": "2026-09-02T20:43:32+02:00", "subject": "Coerce GRAPH_TABLE pattern WHERE clauses to boolean", "body": "transformGraphElementPattern() and transformGraphPattern() ran their\nWHERE clauses through transformExpr() without ever applying\ncoerce_to_boolean(), so a WHERE clause of any type was accepted and\nits bare datum was used as the qual, which resulted in wrong results\nwithout a diagnostic. Route both sites through\ntransformWhereClause(), like every other WHERE clause, and add\nregression tests for the element-level and pattern-level cases.\n\nAuthor: Ewan Young \nReviewed-by: Chao Li \nDiscussion: https://www.postgresql.org/message-id/flat/CAON2xHNZUS8ZwzVEJFeRirsC_-7EJvdbwABSzdGaJ_DF9DvHoQ%40mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "2451e34c404aba1038f5e749c85bb268b2f4525e", "date": "2026-09-02T12:22:12-07:00", "subject": "perf(spy): avoid scanning cleared mocks", "body": "", "style": "conventional", "type": "perf", "scope": "spy", "breaking": false} {"repo": "neovim", "sha": "40258d5d05446d8b048d4830db617f76b8d35b28", "date": "2026-09-02T22:04:32+02:00", "subject": "fix(health): correct reporting messages #41632", "body": "Problems:\n- Slow shell check measures time in nanoseconds, but reports seconds.\n- `kdch1` check incorrectly tests `kbs_entry` instead.\n- curl version is passed as an advice, so it is never reported.\n\nSolutions:\n- Scale the elapsed time to seconds before reporting.\n- Test `kdch1_entry` for `kdch1` check.\n- Format curl warning with `string.format()`.\n\nAI-assisted", "style": "conventional", "type": "fix", "scope": "health", "breaking": false} {"repo": "angular", "sha": "74c1716cefa3268528011b2a459c41ca3f75845c", "date": "2026-09-02T20:22:13Z", "subject": "refactor(compiler-cli): index pipes in template expressions", "body": "Update the template indexer to discover and record pipes used in template expressions.\n\nThis associates template pipe identifiers with their target pipe class declarations, enabling indexers and language tooling to properly resolve and cross-reference pipes.", "style": "conventional", "type": "refactor", "scope": "compiler-cli", "breaking": false} {"repo": "neovim", "sha": "45b645718c804b66394231f85f52413218fdafbd", "date": "2026-09-02T22:27:20+02:00", "subject": "fix(keymap): report the documented parameter name #41633", "body": "Problem:\nInvalid `modes` parameter for `vim.keymap.del()` reports wrong parameter\nname `\"mode\"`.\n\nSolution:\nReport correct parameter name `\"mode\"` instead.", "style": "conventional", "type": "fix", "scope": "keymap", "breaking": false} {"repo": "openssl", "sha": "862853b97484857b7d94865fcc3b862a5ed765df", "date": "2026-09-02T22:56:36+02:00", "subject": "Add tests for bio_dgram_pair", "body": "Add tests to ensure various race conditions are closed in bio_dgram_pair\nusage\n\nCo-authored-by: Neil Horman \nAssisted-by: Claude:claude-fable-5-1\n\nReviewed-by: Andrew Dinh \nReviewed-by: Neil Horman \nMergeDate: Fri Sep 11 20:50:59 2026\n(Merged from https://github.com/openssl/openssl/pull/32498)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "10fa98bb6e086ff498d4bfecd4b0cdbb802d079b", "date": "2026-09-02T22:58:26+02:00", "subject": "fix(pack): read document links from the request buffer #41638", "body": "Problem:\nThe `textDocument/documentLink` handler resolves the confirmation buffer\nfrom the request URI and checks it for `nil`, then reads lines from\nbuffer `0`. When the confirmation buffer is not current, links are\ncomputed from unrelated text and returned against the confirmation\nbuffer's line numbers, producing missing or misplaced links.\n\nSolution:\nRead lines from the resolved `bufnr`.\n\nAI-assisted", "style": "conventional", "type": "fix", "scope": "pack", "breaking": false} {"repo": "neovim", "sha": "a58cee4512a12688099a569c2e1c11ec6c1c608b", "date": "2026-09-02T23:01:32+02:00", "subject": "fix(lsp): respect range in linewise visual mode #41636", "body": "Problem:\nAn explicitly passed `range` to `vim.lsp.buf.format()` in linewise\nvisual mode is silently replaced by the selection.\n\nSolution:\nParenthesise the mode check.", "style": "conventional", "type": "fix", "scope": "lsp", "breaking": false} {"repo": "neovim", "sha": "0ecca23ee2449304f138090eee5e15d512d1cceb", "date": "2026-09-02T23:01:54+02:00", "subject": "fix(treesitter): accept a string title in inspect_tree() #41639", "body": "Problem:\n`inspect_tree()` documents `title` as\n`string|fun(bufnr:integer):string|nil`, but the implementation handles\nonly `nil` and function values. A string title leaves `title` unset and\nfails the assertion below.\n\nSolution:\nUse the string as the title.", "style": "conventional", "type": "fix", "scope": "treesitter", "breaking": false} {"repo": "openssl", "sha": "3af7a00956aabd0afc8b54932e8882091849f30a", "date": "2026-09-02T17:03:37-04:00", "subject": "fix MSVC 2013 builds in shared mode", "body": "MSVC is running ok with static linking at the moment, but shared builds\nstill fail, because all of our dynamically linked tests include\nlibtestutil.a, which references vsnprintf.\n\nFix it by including the msvc2013 printf code in libstestutil.a\n\nWhile we're at it, mix up the 2013 test a bit, so that one of the builds\ncompiles a static library, and the other does a shared library\n\nReviewed-by: Mounir Idrassi \nReviewed-by: Bob Beck \nMergeDate: Fri Sep 4 16:47:25 2026\n(Merged from https://github.com/openssl/openssl/pull/32652)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "6cc5a420b3f5e500ffde3ab7517a6624030191f0", "date": "2026-09-02T23:03:39+02:00", "subject": "clang-tidy: disable readability-redundant-parentheses", "body": "It is an opinion and a judgement call I think is better left for humans.\n\nRef: #22803 where this suddenly showed up for code not added in the PR.\n\nCloses #22809", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "a7195a4f0834127572cdb8119547dfe73dee9f8e", "date": "2026-09-02T21:20:18Z", "subject": "build(zone.js): fix release PR URL and filter changelog to zone.js scope", "body": "In recent conventional-changelog version updates, the underlying\n@conventional-changelog/git-client dropped support for the raw `grep` and\n`extendedRegexp` options, causing all monorepo commits to be included in the\nzone.js changelog. Additionally, without `tagPrefix: 'zone.js-'`, conventional-changelog\ntreated monorepo release tags as release boundaries, generating changelog sections\nfor every intermediate monorepo release.\n\nThis change:\n- Configures `tagPrefix: 'zone.js-'` and filters commits by `scope === 'zone.js'`\n via `writerOpts.transform` in the gulp changelog task.\n- Fixes the PR creation link in the release script to use the cross-fork\n compare URL format on GitHub.\n- Fixes the release commit SHA lookup in `cutReleaseWorkflow`.\n- Updates zone.js release documentation.\n\n(cherry picked from commit cd771174b91b83dc591d7c496dc4c14467ce1a57)", "style": "conventional", "type": "build", "scope": "zone.js", "breaking": false} {"repo": "angular", "sha": "cd771174b91b83dc591d7c496dc4c14467ce1a57", "date": "2026-09-02T21:20:18Z", "subject": "build(zone.js): fix release PR URL and filter changelog to zone.js scope", "body": "In recent conventional-changelog version updates, the underlying\n@conventional-changelog/git-client dropped support for the raw `grep` and\n`extendedRegexp` options, causing all monorepo commits to be included in the\nzone.js changelog. Additionally, without `tagPrefix: 'zone.js-'`, conventional-changelog\ntreated monorepo release tags as release boundaries, generating changelog sections\nfor every intermediate monorepo release.\n\nThis change:\n- Configures `tagPrefix: 'zone.js-'` and filters commits by `scope === 'zone.js'`\n via `writerOpts.transform` in the gulp changelog task.\n- Fixes the PR creation link in the release script to use the cross-fork\n compare URL format on GitHub.\n- Fixes the release commit SHA lookup in `cutReleaseWorkflow`.\n- Updates zone.js release documentation.", "style": "conventional", "type": "build", "scope": "zone.js", "breaking": false} {"repo": "curl", "sha": "0d68a05cc3dcf750cf36499eb0f0c9856f203a42", "date": "2026-09-02T23:39:10+02:00", "subject": "curl_addrinfo: drop redundant parentheses", "body": "Closes #22810", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "6f5406c678d5b8b4553bff4d7bc42573b5d6871f", "date": "2026-09-03T01:43:18+03:00", "subject": "fix(docs-infra): correct example viewer tab state and line numbering", "body": "Five more defects in the example viewer, following #70508.\n\nThe DOM was queried before Angular rendered it. `setCodeLinesVisibility()`\nwalks the rendered lines but ran synchronously on tab change, so it measured\nthe outgoing tab and the incoming file showed in full. The selected tab was\nalso lost when the code block was hidden and reshown, because the recreated\ntab group had no `[selectedIndex]` while `snippetCode` survived, leaving the\nstrip and the code disagreeing.\n\n`expandable` was computed once at startup by counting hidden DOM nodes, so a\ncollapsed tab offered no way to expand it, and recomputing that count on tab\nchange would drop the control whenever the block was expanded, since nothing\nis hidden then. Both paths now share one rule: a file is expandable when it\nhas a `visibleLinesRange` and either the block is expanded or the range\nactually hides lines, so a range that covers its whole file still gets no\ninert control.\n\nArray indices were also mixed with 1-based line numbers: the gap check tested\n`index - 1` for the preceding line, drawing a `...` separator inside\ncontiguous ranges, and the gutter tested `index` while the code tested\n`index + 1`, shifting every line number by one.\n\nFive new specs cover these, using the comma-separated range format the\npipeline emits; each fails with its fix reverted. The tab label also moves\nfrom 0.8125rem to 0.875rem to match the code beside it.\n\n(cherry picked from commit 26afa313f3f9871ed0172ed1b18ef019459716a7)", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "26afa313f3f9871ed0172ed1b18ef019459716a7", "date": "2026-09-03T01:43:18+03:00", "subject": "fix(docs-infra): correct example viewer tab state and line numbering", "body": "Five more defects in the example viewer, following #70508.\n\nThe DOM was queried before Angular rendered it. `setCodeLinesVisibility()`\nwalks the rendered lines but ran synchronously on tab change, so it measured\nthe outgoing tab and the incoming file showed in full. The selected tab was\nalso lost when the code block was hidden and reshown, because the recreated\ntab group had no `[selectedIndex]` while `snippetCode` survived, leaving the\nstrip and the code disagreeing.\n\n`expandable` was computed once at startup by counting hidden DOM nodes, so a\ncollapsed tab offered no way to expand it, and recomputing that count on tab\nchange would drop the control whenever the block was expanded, since nothing\nis hidden then. Both paths now share one rule: a file is expandable when it\nhas a `visibleLinesRange` and either the block is expanded or the range\nactually hides lines, so a range that covers its whole file still gets no\ninert control.\n\nArray indices were also mixed with 1-based line numbers: the gap check tested\n`index - 1` for the preceding line, drawing a `...` separator inside\ncontiguous ranges, and the gutter tested `index` while the code tested\n`index + 1`, shifting every line number by one.\n\nFive new specs cover these, using the comma-separated range format the\npipeline emits; each fails with its fix reverted. The tab label also moves\nfrom 0.8125rem to 0.875rem to match the code beside it.", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "4340a63c52fca4d08ef258a3c053f50932eab27e", "date": "2026-09-02T18:04:41-05:00", "subject": "fix(router): avoid view transitions when the user agent provides one", "body": "Preserve hasUAVisualTransition through Location and the Router navigation pipeline. This prevents withViewTransitions from starting an author transition after the browser has already performed one, including across redirects and when using experimental platform navigation.\n\n(cherry picked from commit af26a8c52178a0e972c2841051db44f9d04ac371)", "style": "conventional", "type": "fix", "scope": "router", "breaking": false} {"repo": "angular", "sha": "af26a8c52178a0e972c2841051db44f9d04ac371", "date": "2026-09-02T18:04:41-05:00", "subject": "fix(router): avoid view transitions when the user agent provides one", "body": "Preserve hasUAVisualTransition through Location and the Router navigation pipeline. This prevents withViewTransitions from starting an author transition after the browser has already performed one, including across redirects and when using experimental platform navigation.", "style": "conventional", "type": "fix", "scope": "router", "breaking": false} {"repo": "vue-core", "sha": "29f2410b01519bdbba925c7d63a23e9746da5847", "date": "2026-09-03T08:37:24+08:00", "subject": "fix(runtime-vapor): should preserve style merging order from static and v-bind props", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "d105e19e914bca6e1140bfec074e055ed0679e21", "date": "2026-09-03T06:37:58+06:00", "subject": "chore: fix typos in comments", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "angular", "sha": "e03f20a3f727f443ad31b36c769022813ca5efe9", "date": "2026-09-02T18:20:50-07:00", "subject": "fix(docs-infra): ignore external links when mapping navigation items to routes", "body": "When navigation items contain external URLs (e.g. https:// links for documentation or third-party resources), mapNavigationItemsToRoutes previously registered them as Angular Router route definitions. In recent versions of @angular/build, the static prerender worker asserts that discovered routes do not return empty content during SSG prerendering, causing production builds and CI adev-deploy to fail on these routes.\n\nThis change filters out external links in mapNavigationItemsToRoutes so only valid internal application paths are registered as Angular routes.\n\n(cherry picked from commit 502fa130aafa29fc390c47df4e13ce9e19eb3187)", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "502fa130aafa29fc390c47df4e13ce9e19eb3187", "date": "2026-09-02T18:20:50-07:00", "subject": "fix(docs-infra): ignore external links when mapping navigation items to routes", "body": "When navigation items contain external URLs (e.g. https:// links for documentation or third-party resources), mapNavigationItemsToRoutes previously registered them as Angular Router route definitions. In recent versions of @angular/build, the static prerender worker asserts that discovered routes do not return empty content during SSG prerendering, causing production builds and CI adev-deploy to fail on these routes.\n\nThis change filters out external links in mapNavigationItemsToRoutes so only valid internal application paths are registered as Angular routes.", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "5a1acd8e08d746a03d2a467c922b12d8439b8018", "date": "2026-09-02T18:51:27-07:00", "subject": "test(docs-infra): add route discovery and prerendering smoke test target", "body": "Add a dedicated smoke test target (//adev/scripts/routes:test) and routing unit smoke test (adev/src/app/routing/routes.spec.ts) to validate route discovery, representative content paths, and route mappings.\n\nThis verifies that external links are never mapped into client router routes (which causes static site prerendering builds to fail), ensures all navigation items resolve to existing markdown content files, and covers representative routes across all content sections.", "style": "conventional", "type": "test", "scope": "docs-infra", "breaking": false} {"repo": "babel", "sha": "289acef1c17672f8f845ed34a772026f66e43420", "date": "2026-09-03T04:05:48+02:00", "subject": "Fix import path for babel-node.js", "body": "* Fix import path for babel-node.js\n\n* Fix import path for babel-node.js\n\n* Update packages/babel-node/test/helpers/runner.js\n\nCo-authored-by: Huáng Jùnliàng \n\n---------", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "a830fa886cee8b5c366787c50aadb1d5f17b2348", "date": "2026-09-03T12:09:49+09:00", "subject": "Stabilize recovery conflict stats checks in 031_recovery_conflict.pl", "body": "The buildfarm member akepa reported a failure in the\n031_recovery_conflict.pl test.\n\nThe test checked pg_stat_database_conflicts immediately after detecting\na recovery conflict in the standby log. However, the conflict counter is\nflushed by the canceled backend during backend exit, so WAL replay\ncompletion and the log message did not guarantee that the updated\nstatistics are visible yet. So, previously, the test could see a conflict\ncounter of 0 even though the conflict had already occurred, triggering\nthe test failure.\n\nFix this by polling for the expected conflict counter instead of reading\nit only once, handling the asynchronous pgstats update.\n\nPer buildfarm member akepa.\n\nBackpatch to v17, where this test is enabled and has the same race.\n\nAuthor: Fujii Masao \nReviewed-by: Ayush Tiwari \nReviewed-by: Nazir Bilal Yavuz \nDiscussion: https://postgr.es/m/CAHGQGwHmiLNRfvJDAR=PmxQgf7DbzPSO1M-1RoqO8oy=t2G5KA@mail.gmail.com\nBackpatch-through: 17", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "57c1072380536656df87ebc59270f966b7a0f695", "date": "2026-09-03T12:16:54+09:00", "subject": "Avoid backend hang during temp table cleanup in deferrable transactions", "body": "Previously, a session that had created temporary objects could hang during\nexit if its default transaction mode was SERIALIZABLE READ ONLY DEFERRABLE\nand a concurrent serializable transaction was prepared. During backend\nexit, temporary-relation cleanup pushed a regular transaction snapshot,\nwhich honored the session's default settings and could wait for a safe\nserializable snapshot. So, if the conflicting transaction was prepared,\nthis wait could last indefinitely.\n\nThe wait occurred while the backend was already exiting and interrupts\nwere held off, so even pg_terminate_backend() could not cancel it. The\nbackend therefore remained until the prepared transaction was resolved.\n\nTemporary-relation cleanup only needs an active MVCC snapshot for fetching\nTOAST data from catalog tuples while dropping temporary objects. It does\nnot need a transaction snapshot affected by the user's default isolation\nsettings. So, use a catalog snapshot instead, which provides the needed\nprotection without entering the deferrable safe-snapshot wait.\n\nBackpatch to all supported versions.\n\nBug: #19441\nReported-by: Alexander Lakhin \nAuthor: Andrey Rachitskiy \nReviewed-by: Andrey Borodin \nReviewed-by: Fujii Masao \nDiscussion: https://postgr.es/m/19441-ec29f3b1363b4a68@postgresql.org\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "8b59e8098e97974e56ad68c92c24d178b59d90c4", "date": "2026-09-03T14:37:06+10:00", "subject": "tdes: restore query parameters", "body": "TDES doesn't do anything with these parameters but they do default via the\ncommon cipher code and return zero (false).\n\nReviewed-by: Shane Lontis \nReviewed-by: Mounir Idrassi \nMerge-date: Tue Sep 15 05:19:42 2026\nMerged-from: https://github.com/openssl/openssl/pull/32656", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "31fc1e5ee8322f11ad95b03e0f07194abad92384", "date": "2026-09-03T13:38:54+08:00", "subject": "feat(BorderBeam): support individual beam configuration", "body": "* feat(BorderBeam): support per-beam configuration\n\n* test(BorderBeam): update site table snapshots\n\n* refactor(BorderBeam): use getItemConfig for per-beam configuration\n\n* refactor(BorderBeam): use items for per-beam configuration\n\n* refactor(BorderBeam): simplify item config handling\n\n* chore: bump antd.min.js size-limit to 445 KiB", "style": "conventional", "type": "feat", "scope": "BorderBeam", "breaking": false} {"repo": "curl", "sha": "959cb7df1ca224e30855d29b3b7d44dff87df01a", "date": "2026-09-03T11:14:06+05:30", "subject": "docs: add the missing .html to the url-encode book link", "body": "The everything.curl.dev page is served at url-encode.html; without the\nsuffix it 404s. It is the only everything.curl.dev link in docs.\n\nCloses #22811", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "be6760bc99ad5e195ae75c524731bea83e7674d7", "date": "2026-09-03T13:45:59+08:00", "subject": "feat: Listy support scrollWidth", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "01b41a82daa15b1057afd53f4a980af67dcd8fbb", "date": "2026-09-03T13:52:20+08:00", "subject": "fix(runtime-vapor): derive transition persisted deterministically and let the renderer own v-show enter", "body": "Slot-propagated v-show roots derived `persisted` only under `appear` and\ncarried it across branch swaps via an element marker, so structural\nenter/leave flipped with `appear` and `mode`. v-show also drove enter/leave\non its first touch of a fresh branch element, firing hooks on hidden nodes\nand double-entering in slots.", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "fcae0e90264f3491df4c8eb0cfe2869a38181847", "date": "2026-09-03T13:52:21+08:00", "subject": "fix(runtime-vapor): stop resuming an out-in branch after unmount", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "d7d5f7d7002843e0f01d92626efc1b00bd3510fd", "date": "2026-09-03T13:52:22+08:00", "subject": "fix(runtime-vapor): rebind transition hooks on prop change for dynamic slot roots", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "17811d87cacbb99ae735f690c1543c002549ee91", "date": "2026-09-03T13:52:22+08:00", "subject": "fix(runtime-core): relocate persisted transition roots on forced leave", "body": "* fix(runtime-core): relocate persisted transition roots on forced leave\n\n* test(vapor-e2e): switch kept-alive page while the v-show leave is still in flight\n\n* chore: update", "style": "conventional", "type": "fix", "scope": "runtime-core", "breaking": false} {"repo": "vue-core", "sha": "78ff63b8f91d6b0460217fdb3f6b8c7050f7dd5b", "date": "2026-09-02T23:15:22-07:00", "subject": "fix(runtime-vapor): relocate v-for rows as reorders instead of enters", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "bb64d37c728da292bc608c124fccc276aa10164b", "date": "2026-09-02T23:34:54-07:00", "subject": "refactor(runtime-vapor): keep composed transition-group keys off $key", "body": "", "style": "conventional", "type": "refactor", "scope": "runtime-vapor", "breaking": false} {"repo": "openssl", "sha": "6e8ec7c3e1db835f7e395430404f331f5d09ccce", "date": "2026-09-03T08:51:07+02:00", "subject": "dtls: Add missing pqueue free", "body": "Fixes coverity issue 1700559.\n\nSigned-off-by: Norbert Pocs \nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nMerge-date: Thu Sep 17 16:50:59 2026\nMerged-from: https://github.com/openssl/openssl/pull/32666\n(cherry picked from commit 9bd26a3a3b4179e05178fb26974bee66203358ca)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "9bd26a3a3b4179e05178fb26974bee66203358ca", "date": "2026-09-03T08:51:07+02:00", "subject": "dtls: Add missing pqueue free", "body": "Fixes coverity issue 1700559.\n\nSigned-off-by: Norbert Pocs \nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nMerge-date: Thu Sep 17 16:50:59 2026\nMerged-from: https://github.com/openssl/openssl/pull/32666", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "ff0adb8079546e3b256c1ed496a02c92d205d442", "date": "2026-09-02T23:52:22-07:00", "subject": "fix(runtime-vapor): land v-for rows after leaving rows at the list tail", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "vue-core", "sha": "938df056304bf45c3bf6e6c5d4af8e28f0bb15d6", "date": "2026-09-02T23:54:22-07:00", "subject": "test(vapor-e2e): element child update triggers transition-group move", "body": "", "style": "conventional", "type": "test", "scope": "vapor-e2e", "breaking": false} {"repo": "openssl", "sha": "ded4af3921919251e2574c2b46e3d8951522c0c9", "date": "2026-09-03T09:06:47+02:00", "subject": "dtls: Don't check for NULL when it can't be", "body": "In dtls context SSL_CONNECTION->d1 is always set by dtls1_new(),\ntherefore checking for it only makes the static analysis to think it can\nbe otherwise and the variable can be dereferenced later in the function.\n\nFixes coverity issue 1700558 and 1700557.\n\nSigned-off-by: Norbert Pocs \nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nMerge-date: Thu Sep 17 16:51:00 2026\nMerged-from: https://github.com/openssl/openssl/pull/32666\n(cherry picked from commit 36e15462b1fd2af9d5189e07829b864c88a6925a)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "36e15462b1fd2af9d5189e07829b864c88a6925a", "date": "2026-09-03T09:06:47+02:00", "subject": "dtls: Don't check for NULL when it can't be", "body": "In dtls context SSL_CONNECTION->d1 is always set by dtls1_new(),\ntherefore checking for it only makes the static analysis to think it can\nbe otherwise and the variable can be dereferenced later in the function.\n\nFixes coverity issue 1700558 and 1700557.\n\nSigned-off-by: Norbert Pocs \nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nMerge-date: Thu Sep 17 16:51:00 2026\nMerged-from: https://github.com/openssl/openssl/pull/32666", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "9a8966879a051aad5e870f30d1c6c77a1a588186", "date": "2026-09-03T09:23:45+02:00", "subject": "fix(iter): keep take() predicate within bounds #41635", "body": "Problem:\n`IterArray:take()` iterates up to `self._tail`, but `_tail` is\nexclusive. When no element fails the predicate, the loop reads one index\npast the last element and calls the predicate with `nil`.\n\nSolution:\nStop at `self._tail - inc`, which is the last in-range index for both\niteration directions. Add tests using a predicate which dereferences its\nargument and matches every element, forward and reversed.\n\nAI-assisted", "style": "conventional", "type": "fix", "scope": "iter", "breaking": false} {"repo": "openssl", "sha": "a3ff226af9fb1568ce74f799784fa876fc8cc949", "date": "2026-09-03T09:35:23+02:00", "subject": "Fix wrong condition evaluation order", "body": "The OR has precedence therefore the code can fall over when dtls\nhas a 0 lenght record.\n\nThe condition evaluated as `(a || b) ? c : d`.\n\nFixes coverity issue 1700556.\n\nSigned-off-by: Norbert Pocs \nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nMerge-date: Thu Sep 17 16:51:01 2026\nMerged-from: https://github.com/openssl/openssl/pull/32666\n(cherry picked from commit 7ff33de472e0550159e330753558cb77666bc57b)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "7ff33de472e0550159e330753558cb77666bc57b", "date": "2026-09-03T09:35:23+02:00", "subject": "Fix wrong condition evaluation order", "body": "The OR has precedence therefore the code can fall over when dtls\nhas a 0 lenght record.\n\nThe condition evaluated as `(a || b) ? c : d`.\n\nFixes coverity issue 1700556.\n\nSigned-off-by: Norbert Pocs \nReviewed-by: Neil Horman \nReviewed-by: Andrew Dinh \nMerge-date: Thu Sep 17 16:51:01 2026\nMerged-from: https://github.com/openssl/openssl/pull/32666", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "b97cef6c0871310a2ebd9a7c12951a16411ac397", "date": "2026-09-03T09:48:17+02:00", "subject": "appveyor: fix description, drop redundant env", "body": "- drop redundant Unicode setting.\n Follow-up to 41198e09b66bd5f99fcec9fe9f0d562f70559309 #20390\n\n- sync OpenSSL version in description with actual.\n Follow-up to 81cdf4d8e5be60bf3fe498c823c34a42f2b08294 #21939\n\nReported-by: Jay Satiro\nRef: #22653\n\nCloses #22812", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "7f3dbfa4c3c444e4f2ff04d188c50f6778cfb020", "date": "2026-09-03T15:56:06+08:00", "subject": "refactor(runtime-vapor): resolve markerless v-for hydration anchors locally", "body": "", "style": "conventional", "type": "refactor", "scope": "runtime-vapor", "breaking": false} {"repo": "vite", "sha": "a6d43bc9e3464faa4d49f090e75e1ab334ffb7b0", "date": "2026-09-03T16:03:27+08:00", "subject": "fix(bundled-dev): mark payload delivered on client report", "body": "", "style": "conventional", "type": "fix", "scope": "bundled-dev", "breaking": false} {"repo": "neovim", "sha": "9222ed3a4a5a2da9b97ce9f95f7c25f09c52d479", "date": "2026-09-03T10:10:05+02:00", "subject": "fix(filetype): report documented parameter name #41649", "body": "Problem:\nInvalid `args` parameter for `vim.filetype.match()` reports wrong\nparameter name `\"arg\"`.\n\nSolution:\nReport correct parameter name `\"args\"` instead.", "style": "conventional", "type": "fix", "scope": "filetype", "breaking": false} {"repo": "vue-core", "sha": "c1de889dfc44986acac3a4e23116bf8aade4510c", "date": "2026-09-03T16:18:34+08:00", "subject": "refactor(runtime-vapor): track transition-group child props with a group-owned render effect", "body": "", "style": "conventional", "type": "refactor", "scope": "runtime-vapor", "breaking": false} {"repo": "postgres", "sha": "88aa6ca138a1eaa63618b2e6ef85ca6c7422dfed", "date": "2026-09-03T17:28:53+09:00", "subject": "pg_upgrade: Test --check with a running source server", "body": "4fff78f009 has added coverage for pg_upgrade --check, but only after\nstopping the source server. This adds coverage for the case where a\nsource server is still running, checking for port patterns and a success\ncase where the old server still runs.\n\nThe chdir() required for VPATH builds needs to be moved slightly\nearlier, to adapt to the generation of the logs.\n\nAuthor: Bertrand Drouvot \nReviewed-by: Ewan Young \nDiscussion: https://postgr.es/m/apU4/hmRv/4gv20W@bdtpg", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "5f3a49abf7ff7bb4a9eba7101eb8e8021960074f", "date": "2026-09-03T14:12:53+05:30", "subject": "Use WAL insert position for conflict detection.", "body": "The retain_dead_tuples feature waits for concurrent transactions to be\napplied before advancing conflict_detection_slot.xmin. It gets the WAL\nwrite position from the publisher and waits until the apply remote\nposition passes it. This ensures that dead tuples, commit timestamps, and\norigins needed for conflict detection (update_deleted,\nupdate_origin_differs, and delete_origin_differs) are retained while\nthose transactions are applied.\n\nHowever, with asynchronous commit enabled on the publisher, a committed\ntransaction may not have updated the WAL write position yet. As a result,\nthe reported WAL position can lag behind the latest committed transaction.\nThis can cause conflict_detection_slot.xmin to advance prematurely,\nallowing dead tuples needed for conflict detection to be removed before\nsubsequent asynchronously committed transactions are applied.\n\nReport the end of the last inserted WAL record instead. A transaction\ninserts its commit record before it is marked committed, so the insert\nposition cannot be behind a transaction that has already committed.\nTransactions that have entered the commit phase but have not yet inserted\ntheir commit record are still accounted for by the in-commit transaction\nID reported in the same message.\n\nIn addition to above, add missing trailing period to update_deleted\nconflict detail messages which are added for the same feature.\n\nReported-by: Nisha Moond \nAuthor: Zhijie Hou \nReviewed-by: Amit Kapila \nDiscussion: https://postgr.es/m/TY4PR01MB177182F547A62FC2666EC04EC94B72@TY4PR01MB17718.jpnprd01.prod.outlook.com\nBackpatch-through: 19, where it was introduced", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "315e10aae2b6a19dc4aaa6810283785824f8a6bb", "date": "2026-09-03T10:47:01+02:00", "subject": "RELEASE-NOTES: 8.14.2 details", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "c19b948ef8d34e5a15e41cd5a88cdd4bda848f8f", "date": "2026-09-03T13:49:38+05:00", "subject": "fix(core): apply SkipSelf to only the starting node in embedded views", "body": "The node-by-node walk used to resolve a token through an embedded view\ninjector forces the `Self` flag on every node injector lookup. Combined\nwith `SkipSelf`, that lookup can never match, so every node between the\ninjection point and the view boundary was skipped and the token was\nresolved from the custom injector instead of the nearest parent node.\n\nClear `SkipSelf` once the starting node has been checked, so only that\nnode is skipped. This also makes the strip at the embedded view injector\nredundant.\n\nFixes #70547\n\n(cherry picked from commit 70756b506c508faa908fe8917e1f6937bcd6df77)", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "angular", "sha": "70756b506c508faa908fe8917e1f6937bcd6df77", "date": "2026-09-03T13:49:38+05:00", "subject": "fix(core): apply SkipSelf to only the starting node in embedded views", "body": "The node-by-node walk used to resolve a token through an embedded view\ninjector forces the `Self` flag on every node injector lookup. Combined\nwith `SkipSelf`, that lookup can never match, so every node between the\ninjection point and the view boundary was skipped and the token was\nresolved from the custom injector instead of the nearest parent node.\n\nClear `SkipSelf` once the starting node has been checked, so only that\nnode is skipped. This also makes the strip at the embedded view injector\nredundant.\n\nFixes #70547", "style": "conventional", "type": "fix", "scope": "core", "breaking": false} {"repo": "curl", "sha": "6a18e7772807940f59ea281104a40e16e8b04d20", "date": "2026-09-03T10:51:22+02:00", "subject": "RELEASE-NOTES: provide 8.16.1 details", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "04197decd6b3674d9f827ebdf880fb4a0cf083b1", "date": "2026-09-03T16:56:41+08:00", "subject": "fix(statuscolumn): heap buffer overflow with sign/number column", "body": "Problem: Drawing 'statuscolumn' leads to a heap-buffer-overflow if a\n sign/number column comes after many items.\nSolution: Avoid curitem >= stl_items_len when writing to stl_items.", "style": "conventional", "type": "fix", "scope": "statuscolumn", "breaking": false} {"repo": "curl", "sha": "ce56d392bece39580d43f13649035a55505afcf7", "date": "2026-09-03T10:59:20+02:00", "subject": "CVE-2026-80231: native CA store conn reuse", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "0f2e0af0da0dd31652cd73a7d5a52c0246416eec", "date": "2026-09-03T11:00:08+02:00", "subject": "GHA/non-native: re-enable FreeBSD riscv64", "body": "It seems to be working again, with no change on the curl end.\n\nAlso: log the list of custom port sources.\n\nRefs:\nhttps://docs.freebsd.org/en/books/handbook/ports/\nhttps://github.com/cross-platform-actions/freebsd-builder/commit/5d49278777bb3825aac3395d2dffc087f55ddf6d\nhttps://github.com/cross-platform-actions/action/issues/166\n\nFollow-up to 73f18365ab8e47912afcd5ee9404c7493d9d0fb2 #22760\n\nCloses #22814", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "67f1586d29333fa994ef38168f1730f849783a36", "date": "2026-09-03T11:06:04+02:00", "subject": "RELEASE-NOTES: fixed for 8.20.1", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "52ec4a95bfb88c5d3a96b2ddc60cecfd74049ff6", "date": "2026-09-03T11:15:36+02:00", "subject": "connectdata: use less curltime", "body": "Replace several curltime members with timediff_t in connectdata.\n\nEveryone needs a shrink!\n\nCloses #22815", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "f419fad428639ce515a53ab7450c78ef60105960", "date": "2026-09-03T11:49:09+02:00", "subject": "Fix lack of message pluralization", "body": "Fixups for commits b99b74144f9, 79b101486c1, bf7d19be9b1.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "6555c4ee1ec2aea7eb541d558981b55eb751497c", "date": "2026-09-03T14:52:18+05:00", "subject": "fix(forms): mark control as dirty before setting its value in FVC interop", "body": "When a `FormValueControl` writes a value through the reactive forms interop,\nthe control is marked dirty only after `setValue()` has already emitted on\n`valueChanges`, so a subscriber that reads `dirty` still sees `false`.\n\nMark the control dirty first, matching the order the `ControlValueAccessor`\npath already uses in `updateControl()`.\n\nFixes #70554\n\n(cherry picked from commit 15bfc1a4befdef22188c23a4d0313140595012fd)", "style": "conventional", "type": "fix", "scope": "forms", "breaking": false} {"repo": "angular", "sha": "15bfc1a4befdef22188c23a4d0313140595012fd", "date": "2026-09-03T14:52:18+05:00", "subject": "fix(forms): mark control as dirty before setting its value in FVC interop", "body": "When a `FormValueControl` writes a value through the reactive forms interop,\nthe control is marked dirty only after `setValue()` has already emitted on\n`valueChanges`, so a subscriber that reads `dirty` still sees `false`.\n\nMark the control dirty first, matching the order the `ControlValueAccessor`\npath already uses in `updateControl()`.\n\nFixes #70554", "style": "conventional", "type": "fix", "scope": "forms", "breaking": false} {"repo": "ant-design", "sha": "ee8970114446ae3a646f1b13fa3c166022c347fa", "date": "2026-09-03T17:54:35+08:00", "subject": "docs: fix brand icon variants falling into general category on Icons page", "body": "* docs: fix brand icon variants falling into general category on Icons page\n\n* docs: also sort Sketch brand icon variants into brand category", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vitest", "sha": "c4473e48ee046938a9ab6784fe2735257c2c0f72", "date": "2026-09-03T13:31:02+03:00", "subject": "fix(coverage): prevent crash on `/@fs/` prepended virtual files", "body": "", "style": "conventional", "type": "fix", "scope": "coverage", "breaking": false} {"repo": "vitest", "sha": "54bfad25d1da4866a2c565b6e9e655a1c3de5183", "date": "2026-09-03T12:31:35+02:00", "subject": "chore: state cleanup", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "vitest", "sha": "68013e90ec528fd72106668d4dcbc56c6136e401", "date": "2026-09-03T12:32:03+02:00", "subject": "feat: implement `userEvent.pointer` for preview", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "vitest", "sha": "f08ce4b7144542af128dcb884150c42074223653", "date": "2026-09-03T12:34:50+02:00", "subject": "fix: apply queued mocks from doMock() in queue order (fixes #10706)", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "vitest", "sha": "73d7294840a09c6c46d90a602a3a8acb3f8ddc41", "date": "2026-09-03T19:35:08+09:00", "subject": "fix(ui): hide running count in static reports", "body": "", "style": "conventional", "type": "fix", "scope": "ui", "breaking": false} {"repo": "vitest", "sha": "584cf308437069385b0fb905fa3ff7d7b7f65a90", "date": "2026-09-03T12:38:28+02:00", "subject": "fix: add a warning if inline project has duplicate plugins due to unexpected `extends: true`", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "curl", "sha": "e26f6e408a80dc0adcb704b0e2c408e240f482cc", "date": "2026-09-03T11:39:47+01:00", "subject": "uint-bset: skip unused slots when counting", "body": "Start counting at first_slot_used to avoid scanning the known-empty\nprefix of the bitset.", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "73923b0dd85bb936ba2f63ee916dabaa0603340d", "date": "2026-09-03T12:40:58+02:00", "subject": "fix(ssh): compare table length in the SSH config #41637", "body": "Problem:\n`parse_ssh_config()` compares tables against a freshly allocated empty\ntable.\n- In `parse_multiple_values()`, the guard which avoids flushing an empty\n accumulator never applies. Runs of separators and trailing whitespace\n push empty strings into the results, and `is_valid()` does not filter\n them. `Host alpha beta ` parses as `{ 'alpha', '', 'beta' }`.\n- In `parse_value()`, the condition reduces to `chr == '\"' and quoted`.\n `quoted` starts false and only that branch sets it, so it can never\n become true: quotes are never recognised and are inserted literally,\n and the unterminated-quote check is unreachable.\n\nSolution:\nCompare `#val` instead. Add a test for repeated and trailing separators.\n\nAI-assisted", "style": "conventional", "type": "fix", "scope": "ssh", "breaking": false} {"repo": "vitest", "sha": "1bcb6c2e75cac1db58b159c2191b7d24859f60ad", "date": "2026-09-03T12:43:04+02:00", "subject": "test: modifiers are released", "body": "", "style": "conventional", "type": "test", "scope": "", "breaking": false} {"repo": "vitest", "sha": "dc19a988e0f3fcc69e7a4478881b822a918f3018", "date": "2026-09-03T12:50:42+02:00", "subject": "docs: explain the `pointer` API is for chains", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vitest", "sha": "d46a7472266d5bb43595ea51dcdd64ab0f560f12", "date": "2026-09-03T13:06:04+02:00", "subject": "fix: treat test.describe as a suite during static collection", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "vitest", "sha": "b848a609d7f7b23cf552de8e265cdfdd89aa66d8", "date": "2026-09-03T13:59:53+02:00", "subject": "chore: trigger push event", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "vitest", "sha": "d7bf1a7444e15db23c9bbb921e968296b0e3cdf6", "date": "2026-09-03T14:03:48+02:00", "subject": "docs: split migration into multiple pages, add a button to copy the prompt", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vitest", "sha": "8a11ab46db8acc91fcf1247112374673854bb10e", "date": "2026-09-03T14:24:51+02:00", "subject": "docs: add \"old version\" banner to v4 docs", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "neovim", "sha": "e1cde28ba28544a28ce18cfaf272e09204010460", "date": "2026-09-03T08:44:22-04:00", "subject": "fix(multicursor): live-mirror Visual-mode operator mapping #41655", "body": "Problem:\nInsert entered by a Visual-mode operator mapping (`xnoremap c c`) does\nnot live-mirror at cursors.\n\nSolution:\nDecide `typed` from the Visual session's own provenance (`vatom`, which\nspans the selection and knows whether it was user input) instead of only\nnegating it.", "style": "conventional", "type": "fix", "scope": "multicursor", "breaking": false} {"repo": "vitest", "sha": "d927e2e43a45c5112c609a60b49268c502ccf22d", "date": "2026-09-03T14:44:52+02:00", "subject": "docs: voidzero theme doesn't have a slot in production, only in dev", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vitest", "sha": "d05ab8efbdb742e7c3a3976aa9862cda234ed524", "date": "2026-09-03T14:47:26+02:00", "subject": "docs: add blog post for Vitest 5", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vitest", "sha": "5fd9ce33cb737a290456423fb700d07cf0971567", "date": "2026-09-03T14:58:30+02:00", "subject": "docs: fix migration links", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vitest", "sha": "1051116c6364092656ee2d73315b03b3cc4aa9a3", "date": "2026-09-03T15:12:04+02:00", "subject": "docs: mention vitest doctor in the blog post", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "vitest", "sha": "decfeb61c71a93372f84b6d43893df86a1756308", "date": "2026-09-03T15:15:14+02:00", "subject": "docs: mention default diagnostics", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "f74915e6a3b2b342446a050efb12719e95df8ea5", "date": "2026-09-03T15:52:01+02:00", "subject": "test/recipes/95-test_external_pyca_data/cryptography.sh: use python3 command", "body": "Some distributions do not provide \"python\" command. Change it\nto python3 (which is hopefully more universally available) and provide\nan ability to supply a different python command name, just in case.\n\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Jakub Zelenka \nReviewed-by: Nikola Pajkovsky \nMerge-date: Mon Sep 7 18:23:53 2026\nMerged-from: https://github.com/openssl/openssl/pull/32668\n(cherry picked from commit 55d5f10e065a3d09f76b47afaf46c3d56faf986c)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "55d5f10e065a3d09f76b47afaf46c3d56faf986c", "date": "2026-09-03T15:52:01+02:00", "subject": "test/recipes/95-test_external_pyca_data/cryptography.sh: use python3 command", "body": "Some distributions do not provide \"python\" command. Change it\nto python3 (which is hopefully more universally available) and provide\nan ability to supply a different python command name, just in case.\n\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Jakub Zelenka \nReviewed-by: Nikola Pajkovsky \nMerge-date: Mon Sep 7 18:23:53 2026\nMerged-from: https://github.com/openssl/openssl/pull/32668", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "493431986b24448d439609bc6cb3246e2a74e509", "date": "2026-09-03T16:02:27+02:00", "subject": "test/recipes/95-test_external_pyca_data/cryptography.sh: use BLDTOP, not SRCTOP", "body": "Create temporary directory under $BLDTOP and not $SRCTOP,\nas it should be.\n\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Jakub Zelenka \nReviewed-by: Nikola Pajkovsky \nMerge-date: Mon Sep 7 18:23:54 2026\nMerged-from: https://github.com/openssl/openssl/pull/32668\n(cherry picked from commit c5c6f692b753d7803db08ab97915bebd544a5422)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "c5c6f692b753d7803db08ab97915bebd544a5422", "date": "2026-09-03T16:02:27+02:00", "subject": "test/recipes/95-test_external_pyca_data/cryptography.sh: use BLDTOP, not SRCTOP", "body": "Create temporary directory under $BLDTOP and not $SRCTOP,\nas it should be.\n\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Jakub Zelenka \nReviewed-by: Nikola Pajkovsky \nMerge-date: Mon Sep 7 18:23:54 2026\nMerged-from: https://github.com/openssl/openssl/pull/32668", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "4ec2e04b9c89daf6f841281f484042892da572e3", "date": "2026-09-03T16:04:35+02:00", "subject": "test/recipes/95-test_external_pyca_data/cryptography.sh: update dep installation", "body": "* pyca has switched to dependency groups in commit 47.0.0~82 \"Convert\n development extras to dependency groups (#14638)\" (ea1fc9f37eb6),\n use pip --groups instead of the square bracket syntax.\n * pyca 40.0+ does not use CFLAGS/LDFLAGS for passing flags directly\n anymore, but rather needs OPENSSL{,_INCLUDE,_LIB}_DIR[1]. Synthesise\n an include dir (similarly to the way it is done\n in test/recipes/95-test_external_oqsprovider.t) and pass it along\n the library dir to the pip install command.\n\n[1] https://docs.rs/openssl/latest/openssl/#manual\n\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Jakub Zelenka \nReviewed-by: Nikola Pajkovsky \nMerge-date: Mon Sep 7 18:23:54 2026\nMerged-from: https://github.com/openssl/openssl/pull/32668\n(cherry picked from commit ead81eac94d6d039301a49f57fa891c9c6c47dc1)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "ead81eac94d6d039301a49f57fa891c9c6c47dc1", "date": "2026-09-03T16:04:35+02:00", "subject": "test/recipes/95-test_external_pyca_data/cryptography.sh: update dep installation", "body": "* pyca has switched to dependency groups in commit 47.0.0~82 \"Convert\n development extras to dependency groups (#14638)\" (ea1fc9f37eb6),\n use pip --groups instead of the square bracket syntax.\n * pyca 40.0+ does not use CFLAGS/LDFLAGS for passing flags directly\n anymore, but rather needs OPENSSL{,_INCLUDE,_LIB}_DIR[1]. Synthesise\n an include dir (similarly to the way it is done\n in test/recipes/95-test_external_oqsprovider.t) and pass it along\n the library dir to the pip install command.\n\n[1] https://docs.rs/openssl/latest/openssl/#manual\n\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Jakub Zelenka \nReviewed-by: Nikola Pajkovsky \nMerge-date: Mon Sep 7 18:23:54 2026\nMerged-from: https://github.com/openssl/openssl/pull/32668", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "b3861819962a32da5c56bca3d9f47b4754f49a57", "date": "2026-09-03T16:12:42+02:00", "subject": "Update pyca-cryptography and wycheproof submodules", "body": "Update pyca-cryptography to version 50.0 and wycheproof to the latest\nmain.\n\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Jakub Zelenka \nReviewed-by: Nikola Pajkovsky \nMerge-date: Mon Sep 7 18:23:55 2026\nMerged-from: https://github.com/openssl/openssl/pull/32668\n(cherry picked from commit 12060ef22bf1e5021a0cbfe9977b340ff8aa6e25)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "12060ef22bf1e5021a0cbfe9977b340ff8aa6e25", "date": "2026-09-03T16:12:42+02:00", "subject": "Update pyca-cryptography and wycheproof submodules", "body": "Update pyca-cryptography to version 50.0 and wycheproof to the latest\nmain.\n\nSigned-off-by: Eugene Syromiatnikov \nReviewed-by: Jakub Zelenka \nReviewed-by: Nikola Pajkovsky \nMerge-date: Mon Sep 7 18:23:55 2026\nMerged-from: https://github.com/openssl/openssl/pull/32668", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "c07b32149ab8e0aa92196c0938306498226f79aa", "date": "2026-09-03T22:17:38+08:00", "subject": "docs: add Listy to ConfigProvider component config", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "neovim", "sha": "3beb15da5704cef4243d1cc03ad32a826b6778e1", "date": "2026-09-03T10:28:17-04:00", "subject": "fix(multicursor): live-mirror 'autocomplete' #41660", "body": "Problem:\nDuring 'autocomplete', a completion-ending `` restarts completion\nimmediately. Multicursor insert-cascade skips replay (`edit()` refuses\nto nest if compl/pum is active), so the mcursors do not update until\nESC.\n\nSolution:\nFlush pending keys in `TRIGGER_AUTOCOMPLETE`, just before autocomplete\nrestarts again.", "style": "conventional", "type": "fix", "scope": "multicursor", "breaking": false} {"repo": "vite", "sha": "07d65709f3802c4196ac7074b5b723c1dabf2e34", "date": "2026-09-03T22:31:09+08:00", "subject": "docs(api-javascript): remove example file name", "body": "", "style": "conventional", "type": "docs", "scope": "api-javascript", "breaking": false} {"repo": "eslint", "sha": "1696682791661c13167eb905da2f38d1b8f4a3bf", "date": "2026-09-03T23:41:39+09:00", "subject": "ci: restore EMFILE test on Node.js 26", "body": "", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "neovim", "sha": "e1a28e52b2b9391889826cbfaec473d8b615e22f", "date": "2026-09-03T11:32:20-04:00", "subject": "fix(editor): shada \"force\" load clears v:oldfiles #41662", "body": "Problem:\n`shada_read()` rebuilds `v:oldfiles` for any \"forced\" read, which first\n*clears* the list. For an in-memory caller such as multicursor, the list\nis never repopulated. So `v:oldfiles` is empty after `Q` + edit.\n\nSolution:\nRebuild `v:oldfiles` only when `kShaDaGetOldfiles` is requested.\n`shada_read_everything()` (`:rshada[!]`, startup) always requests it.\n\nold bug from 411a06c8b6e9", "style": "conventional", "type": "fix", "scope": "editor", "breaking": false} {"repo": "curl", "sha": "2956af6c7451029fb12c12f8166dff211f930d52", "date": "2026-09-03T17:54:45+02:00", "subject": "HTTP-CONNECT: do not react to 401 responses", "body": "When a proxy answers with a 401 on a CONNECT, ignore Authentication\nheaders and fail the connect.\n\nAdd test2173 to verify.\n\nCloses #22817", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "ba0ee081c8746a148bfad1420919fa04dfb0b67a", "date": "2026-09-03T12:41:40-04:00", "subject": "fix(coverity): DEADCODE, FORWARD_NULL, UNINIT #41667", "body": "Problem:\n\n CID 655469: (DEADCODE)\n /build/src/nvim/auto/lua_api_c_bindings.generated.h: 7696 in nlua_api_nvim__mcursor_cascading()\n 7690\n 7691 exit_0:\n 7692 arena_mem_free(arena_finish(&arena));\n 7693 if (ERROR_SET(&err)) {\n 7694 luaL_where(lstate, 1);\n 7695 if (err_param) {\n >>> CID 655469: (DEADCODE)\n >>> Execution cannot reach this statement: \"lua_pushstring(lstate, \"Inv...\".\n 7696 lua_pushstring(lstate, \"Invalid '\");\n 7697 lua_pushstring(lstate, err_param);\n 7698 lua_pushstring(lstate, \"': \");\n 7699 }\n 7700 lua_pushstring(lstate, err.msg);\n 7701 api_clear_error(&err);\n /build/src/nvim/auto/lua_api_c_bindings.generated.h: 7702 in nlua_api_nvim__mcursor_cascading()\n 7696 lua_pushstring(lstate, \"Invalid '\");\n 7697 lua_pushstring(lstate, err_param);\n 7698 lua_pushstring(lstate, \"': \");\n 7699 }\n 7700 lua_pushstring(lstate, err.msg);\n 7701 api_clear_error(&err);\n >>> CID 655469: (DEADCODE)\n >>> Execution cannot reach the expression \"5\" inside this statement: \"lua_concat(lstate, (err_par...\".\n 7702 lua_concat(lstate, err_param ? 5 : 2);\n 7703 return lua_error(lstate);\n 7704 }\n 7705\n 7706 return 1;\n 7707 }\n\n CID 655468: Null pointer dereferences (FORWARD_NULL)\n /src/nvim/mcursor.c: 833 in mc_ins_cascade()\n 827 mc_ins_span_push(keys.items, NULL);\n 828 mc_ins_preview_rebase();\n 829 }\n 830 } else if (ins.data != NULL && ins.size < mc_ins_span.done_len) {\n 831 // Capture shrank without a restart signal, e.g. completion surgery rewrote the pending keys.\n 832 mc_ins_cascade_restart();\n >>> CID 655468: Null pointer dereferences (FORWARD_NU", "style": "conventional", "type": "fix", "scope": "coverity", "breaking": false} {"repo": "neovim", "sha": "3808c00fc86be0c8bb92fbc65d48a12cc4da21f5", "date": "2026-09-03T18:44:30+02:00", "subject": "fix(statusline): default 'statusline' improvements #41597", "body": "Problem: there should be exactly one cell of padding between sections,\nand exactly one cell of minimum padding between the left and the right.\n- Spaces between sections waste space when a section is empty.\n According to a comment in #33036, this was the reason to avoid `%k`\n and implement the keymap section with a vim expression, but other\n sections still have this problem.\n- The diagnostics section wastes space because it is not entirely empty\n when there are diagnostics in another buffer.\n- The terminal exit code section can touch the right side, e.g. the\n ruler, even though it belongs to the left side.\nSolution:\n- Use auto-hiding item groups (`%(` without width fields) to get rid of\n unneeded spaces when a section shows no information.\n This simplifies the 'showcmd' and 'keymap' sections in particular.\n- As a slight simplification, `term_exitcode` is moved into the flags\n section since it is formatted with square brackets like a flag.\n- Count the diagnostics for the current buffer specifically.\n- Ensure at least one cell of padding between the left and the right\n side by adding a space next to the separator `%=`.\n\nProblem: some sections are implemented with `%{%`, even though\nreevaluation of the expression result is not needed.\nThis leads to otherwise needless %-escaping in `progress_status`.\nSolution: use `%{` instead.\n\nProblem: `%{` and `%{%` (without items) replace spaces with fillchars.\nThis looks out-of-place inside the terminal exit code section, and in\ncontrast to all other sections, the 'busy' section is surrounded by\nfillchars, which looks inconsistent, and with some terminal-font\ncombinations, ◐ overlaps the fillchar, e.g. Alacritty & JetBrains Mono.\nSolution: use non-breaking spaces U+202F to avoid fillchar substitution.\n\nProblem: sections that appear/disappear frequently can make otherwise\nmore stable sections jump around a lot.\nSolution: sort the sections on the right roughly by volatility:\n'showcmd' in first place, 'keymap' next to the rule", "style": "conventional", "type": "fix", "scope": "statusline", "breaking": false} {"repo": "neovim", "sha": "da4355ab8f68c4e0290134e691ab7b9dedc2a5df", "date": "2026-09-03T13:22:50-04:00", "subject": "fix(multicursor): 'guicursor' uses Normal shape in Visual mode #41669", "body": "Problem:\n'guicursor' is not respected when extra cursors exist.\n\nPer-cursor selection toggles Visual off, and the sandbox restores it\nonly after restore_current_state(), whose ui_cursor_shape() caches the\nNormal shape.\n\nSolution:\nRestore Visual before restore_current_state().", "style": "conventional", "type": "fix", "scope": "multicursor", "breaking": false} {"repo": "angular", "sha": "bd9b45b5cc1dd904cc4a5de45f6de8e1564b70b6", "date": "2026-09-03T20:46:08+03:00", "subject": "feat(core): allow reading `Injector` from a view or content query", "body": "Queries could already read `ElementRef`, `TemplateRef` and\n`ViewContainerRef` from a matched node via the `read` option, but not the\nnode injector. Getting it required a helper directive on the element.\n\n`{read: Injector}` now returns the node injector of the matched element,\nso a component can resolve tokens as they are seen from that element. This\nis useful when wrapping third-party components that project templates and\nexpect directives inside those templates to inject the host component.\n\nWorks for `@ViewChild`/`@ContentChild` and the signal-based\n`viewChild`/`contentChild`.\n\nFixes #47760", "style": "conventional", "type": "feat", "scope": "core", "breaking": false} {"repo": "babel", "sha": "8e5565e427e9f285e644f6b221a1f64ccacd8757", "date": "2026-09-03T14:43:44-04:00", "subject": "fix(parser): do not form html entity for invalid codepoint", "body": "", "style": "conventional", "type": "fix", "scope": "parser", "breaking": false} {"repo": "babel", "sha": "a2c32a1385c64368c1b59a94c992fe5c15005f75", "date": "2026-09-03T20:44:52+02:00", "subject": "ci: update agentscan", "body": "", "style": "conventional", "type": "ci", "scope": "", "breaking": false} {"repo": "curl", "sha": "a935481d4e04690ec2eef266528c8f5aacdc3608", "date": "2026-09-03T14:51:33-04:00", "subject": "lib: CURL_GLOBAL_WIN32 -> CURL_GLOBAL_WINSOCK", "body": "- Change curl_global_init flag name from CURL_GLOBAL_WIN32 to\n CURL_GLOBAL_WINSOCK.\n\n- Retain CURL_GLOBAL_WIN32 as an alias for legacy reasons.\n\n- Remove lwIP initialization from CURL_GLOBAL_WINSOCK.\n\nPrior to this change lwIP init was controlled by the flag which did\nnot make sense. It may have some historical reason but if it does then\nthat should be represented by a different flag. AFAICT lwIP init is\nnot our responsibility, since it may need different init depending on\nhow it is used.\n\nThe name change WIN32 -> WINSOCK is to eliminate user confusion. The name\nWIN32 may imply a broader Windows initialization, which libcurl always\ndoes and is not controlled by this flag.\n\nReported-by: Viktor Szakats\n\nRef: https://github.com/curl/curl/pull/22813#issuecomment-5529065833\n\nCloses https://github.com/curl/curl/pull/22819", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "7b666e62be89e66aceef9f091a3c922ac5adf951", "date": "2026-09-03T19:12:36Z", "subject": "Wait for pane exit status in all cases not just remain-on-exit since events need it.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "4b1586fa81f2cca49b4195ceba7b18d516ece106", "date": "2026-09-03T20:12:44+01:00", "subject": "Test for exit status.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "5209695703db4096923c203f235d78aec0cbdec8", "date": "2026-09-03T15:20:30-04:00", "subject": "fix(multicursor): abandoned Visual selection does not move cursors #41670", "body": "Problem:\n\"v{motion}\" moves the primary cursor to the selection end, but the\nother cursors snap back to their anchors unless follow-mode (\"q=\") is\non. The selection preview showed them at the selection end.\n\nSolution:\nReplay the abandoned selection at every cursor, not only in follow-mode.\nAn abandoned typed selection now always emits a \"visual\" CmdAtom.", "style": "conventional", "type": "fix", "scope": "multicursor", "breaking": false} {"repo": "postgres", "sha": "903d9d12cc0adcd9a7f92c7b35333e9e40756fbf", "date": "2026-09-03T15:36:25-04:00", "subject": "Move visibility map memory ordering comments.", "body": "nodeIndexonlyscan.c has long carried a detailed explanation of why it is\nsafe for index-only scans to test a visibility map bit without locking\nthe visibility map buffer, despite the possibility of reading a stale\nvalue. Move that explanation into visibilitymap.c.\n\nThis is preparation for an upcoming commit that will move VM checks\nneeded by index-only scans out of the executor and into heapam.\n\nAuthor: Peter Geoghegan \nSuggested-by: Andres Freund \nReviewed-by: Andres Freund \nReviewed-by: Tomas Vondra \nDiscussion: https://postgr.es/m/t6mtqbv2mbfhjni4bvwdgoecppjmxvbyfwl6utovzv76xc2672%40k3o5ryevaeqv", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "8ca9cfb77838b3a17b7c9b5ef26cb2a26b5e1bb2", "date": "2026-09-03T15:16:16-05:00", "subject": "Convert PROC_HDR->startupBufferPinWaitBufId to a Buffer.", "body": "Presently, this field stores the buffer's index, i.e., the Buffer\nminus one, and it uses -1 to mean \"not waiting.\" The callers\nsubtract one before storing the value and add it back after reading\nit. This commit changes the field to store the Buffer itself and\nto use InvalidBuffer (0) as the sentinel, which gets rid of that\narithmetic.\n\nThis is preparatory work for a follow-up commit that will convert\nthe variable to an unsigned atomic, for which a negative sentinel\nmakes little sense.\n\nReviewed-by: Peter Eisentraut \nReviewed-by: Andres Freund \nReviewed-by: Zsolt Parragi \nDiscussion: https://postgr.es/m/alAJeRRzehDjLaF1%40nathan", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "83032dc42cab67dfdd33152d08434785977f069f", "date": "2026-09-03T16:27:09-04:00", "subject": "Add index scan pgstat test coverage.", "body": "Add tests that verify the exact values of the idx_tup_read and\nidx_tup_fetch counters in pg_stat_all_indexes. There was no existing\ncoverage of pg_stat_all_indexes, beyond some counter reset tests;\nidx_tup_read and idx_tup_fetch had no tests at all.\n\nThis is preparation for an upcoming commit that will move index scan\nheap fetches behind a new slot-based table AM interface, relocating the\ncode that maintains these counters.\n\nAuthor: Peter Geoghegan \nReviewed-by: Tomas Vondra \nDiscussion: https://postgr.es/m/CAH2-Wz%3DV0V5WGuo85D-OF%2BTM_CxU3q%3DMMkY-ebY0K8xc-9Xykw%40mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "386bd8408274351d3ea763214c3df2b58b351f86", "date": "2026-09-03T22:31:22+02:00", "subject": "Fix NULL dereference and memory leak in ASN1_item_dup()", "body": "If ASN1_item_d2i_ex() fails asn1_cb() can dereference return\nvalue of ASN1_item_d2i_ex() which is NULL causing NULL dereference.\nAlso if ASN1_item_d2i_ex() succeeds and asn1_cb() fails it causes\na memory leak.\n\nFound by: x509_req fuzzer (MFAIL test).\nAssisted-by: ChatGPT:gpt-5.6\n\nReviewed-by: Mounir Idrassi \nReviewed-by: Bob Beck \nMerge-date: Tue Sep 15 10:49:35 2026\nMerged-from: https://github.com/openssl/openssl/pull/32691", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "618a1dcc12af4b2d1e3a7c0637d88bfc0fc441c9", "date": "2026-09-03T22:31:22+02:00", "subject": "Fix NULL dereference and memory leak in ASN1_item_dup()", "body": "If ASN1_item_d2i_ex() fails asn1_cb() can dereference return\nvalue of ASN1_item_d2i_ex() which is NULL causing NULL dereference.\nAlso if ASN1_item_d2i_ex() succeeds and asn1_cb() fails it causes\na memory leak.\n\nFound by: x509_req fuzzer (MFAIL test).\nAssisted-by: ChatGPT:gpt-5.6\n\nReviewed-by: Mounir Idrassi \nReviewed-by: Bob Beck \nMerge-date: Tue Sep 15 10:48:44 2026\nMerged-from: https://github.com/openssl/openssl/pull/32691", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "be137374de048c9c2ac7746ce16d85f84280b8e1", "date": "2026-09-03T22:31:22+02:00", "subject": "Fix NULL dereference and memory leak in ASN1_item_dup()", "body": "If ASN1_item_d2i_ex() fails asn1_cb() can dereference return\nvalue of ASN1_item_d2i_ex() which is NULL causing NULL dereference.\nAlso if ASN1_item_d2i_ex() succeeds and asn1_cb() fails it causes\na memory leak.\n\nFound by: x509_req fuzzer (MFAIL test).\nAssisted-by: ChatGPT:gpt-5.6\n\nReviewed-by: Mounir Idrassi \nReviewed-by: Bob Beck \nMerge-date: Tue Sep 15 10:47:53 2026\nMerged-from: https://github.com/openssl/openssl/pull/32691", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "e5ab49eb37314acc29ff1208513a13b3bf69b996", "date": "2026-09-03T22:31:22+02:00", "subject": "Fix NULL dereference and memory leak in ASN1_item_dup()", "body": "If ASN1_item_d2i_ex() fails asn1_cb() can dereference return\nvalue of ASN1_item_d2i_ex() which is NULL causing NULL dereference.\nAlso if ASN1_item_d2i_ex() succeeds and asn1_cb() fails it causes\na memory leak.\n\nFound by: x509_req fuzzer (MFAIL test).\nAssisted-by: ChatGPT:gpt-5.6\n\nReviewed-by: Mounir Idrassi \nReviewed-by: Bob Beck \nMerge-date: Tue Sep 15 10:44:49 2026\nMerged-from: https://github.com/openssl/openssl/pull/32691", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "025394badc0421c46b3f55f3cfe8965013b78100", "date": "2026-09-03T22:31:22+02:00", "subject": "Fix NULL dereference and memory leak in ASN1_item_dup()", "body": "If ASN1_item_d2i_ex() fails asn1_cb() can dereference return\nvalue of ASN1_item_d2i_ex() which is NULL causing NULL dereference.\nAlso if ASN1_item_d2i_ex() succeeds and asn1_cb() fails it causes\na memory leak.\n\nFound by: x509_req fuzzer (MFAIL test).\nAssisted-by: ChatGPT:gpt-5.6\n\nReviewed-by: Mounir Idrassi \nReviewed-by: Bob Beck \nMerge-date: Tue Sep 15 10:32:30 2026\nMerged-from: https://github.com/openssl/openssl/pull/32691", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "ec24702d399ace69848b9346b082466a7dd324c9", "date": "2026-09-03T22:31:22+02:00", "subject": "Fix NULL dereference and memory leak in ASN1_item_dup()", "body": "If ASN1_item_d2i_ex() fails asn1_cb() can dereference return\nvalue of ASN1_item_d2i_ex() which is NULL causing NULL dereference.\nAlso if ASN1_item_d2i_ex() succeeds and asn1_cb() fails it causes\na memory leak.\n\nFound by: x509_req fuzzer (MFAIL test).\nAssisted-by: ChatGPT:gpt-5.6\n\nReviewed-by: Mounir Idrassi \nReviewed-by: Bob Beck \nMerge-date: Tue Sep 15 10:30:31 2026\nMerged-from: https://github.com/openssl/openssl/pull/32691", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "1fd8e7e85f4723105c458834c64df77c39f18dad", "date": "2026-09-03T23:39:06+03:00", "subject": "docs: use a supported alert level for the cross-field validation note", "body": "`WARNING:` is not an `AlertSeverityLevel`, so the paragraph rendered as plain\nbody text on https://angular.dev/guide/forms/signals/cross-field-logic while\nthe three other alerts on the page rendered normally. `CRITICAL` is the level\nkitchen-sink.md defines for warning the reader off a footgun.\n\nThe five `@Component({/* ... */})` collapses are the formatter's. The file has\ndrifted out of prettier since it landed, and `ng-dev format changed --check`\nchecks whole changed files.\n\n(cherry picked from commit 0029c3ba5adcdf09c3331ffa38393f0acacb4d3a)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "0029c3ba5adcdf09c3331ffa38393f0acacb4d3a", "date": "2026-09-03T23:39:06+03:00", "subject": "docs: use a supported alert level for the cross-field validation note", "body": "`WARNING:` is not an `AlertSeverityLevel`, so the paragraph rendered as plain\nbody text on https://angular.dev/guide/forms/signals/cross-field-logic while\nthe three other alerts on the page rendered normally. `CRITICAL` is the level\nkitchen-sink.md defines for warning the reader off a footgun.\n\nThe five `@Component({/* ... */})` collapses are the formatter's. The file has\ndrifted out of prettier since it landed, and `ng-dev format changed --check`\nchecks whole changed files.", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "openssl", "sha": "f99bbcd5afff69840f2bfed92458cd5ac092cf91", "date": "2026-09-03T16:50:20-04:00", "subject": "Dont have dgram_pairs track bios, track bio_dgram_pair_st instead", "body": "@ifanzki reported an asan issue with the quic_radix_test in:\nhttps://github.com/openssl/openssl/issues/32420\n\nAnd we've tried fixing it a few different ways, with varying amounts of\nsuccess\n\nThis commit represents an alternative approach which I think is more\ngeneral to the bio_dgram_pair bio type.\n\nThe problem:\nWe have a race condition as reported above in which freeing one side of\na BIO dgram pair results in a use after free\n\nThe root cause:\nBIO_new_bio_dgram_pair creates 2 dgram BIOS, each with a reference count\nof one (which is sensible, as each bio has a single owner). However,\nafter creating the bios, it calls BIO_make_bio_pair(), which links the\ntwo together. Under the covers in the bio_dgram_pair implementation,\neach bio stores the pointer of its peer bio _without_ taking an\nadditional reference on the bio. This means that, once the first bio is\nfreed, the second bio may be doing subsequent io, which requires\ndereferencing of the peer bio , which is already release,\nleading to the use after free case observed.\n\nNominally this isn't an issue, as most code that uses BIO_dgram_pairs\nare single threaded, and freeing of the component bios only occurs after\nany i/o is already terminated, avoiding the observed UAF. However, the\nQUIC radix test exercises the thread assisted mode of the quic stack,\nand in so doing passes each half of a bio dgram pair to a separate\nthread.\n\nThe fix:\n\nNominally we could resolve this by increasing the reference count on\neach of the peer bios, and dropping that ref count when each is freed,\nso that only the last to free actually deallocates both instances, but\nthis creates a chicken and egg issue. Since the the stored reference to\neach bio lives in the dgram pair implementation, upping the refcount of\nthe bios prevents either bio from ever reaching the implementation free\ncode (dgram_pair_free)\n\nSo instead I'm proposing the following in this commit:\n1) Instead of having the internal bio data track the peer bio, instead\n have the in", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "a6f9a206b7f31be24adc82d59bf352aa992a8111", "date": "2026-09-03T21:04:11Z", "subject": "Extend neww -S to work with -t as well as -t, GitHub issue 5571 from Mazunki Hoksaas.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "5d5f66239a60e5bb1784685418bf8c19f0e3bbb0", "date": "2026-09-03T22:04:34+01:00", "subject": "Test for -S from Mazunki Hoksaas.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "9a40e39b1e8780ba75e014cb87860249c405c37e", "date": "2026-09-03T16:10:37-05:00", "subject": "Auto-vectorize the varbit bitwise operators.", "body": "Compile varbit.c with -ftree-vectorize where available, and adjust\nthe loops in bit_and(), bit_or(), bitxor(), and bitnot() so that\nthey are amenable to being auto-vectorized. (Mainly, that involves\nreading the byte count into a local variable before the loop. As\nwritten, the bound is recomputed from the varlena header on every\niteration, because the compiler must assume that the loop might\nmodify the header.)\n\nReviewed-by: Heikki Linnakangas \nReviewed-by: Chao Li \nDiscussion: https://postgr.es/m/aphxrw6NkYkLMtZN%40nathan", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "93169e2e6231853c61c33432bceebf4ffc91015b", "date": "2026-09-03T21:35:38Z", "subject": "Do not attempt to write to control clients which have been destroyed, reported by Artur Penttinen in GitHub issue 5570.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "f7a563ccd1233f99b3b202c6c208ee37694c8cef", "date": "2026-09-03T17:37:03-04:00", "subject": "Add non-MVCC index scan test module.", "body": "Add test_indexscan, a new test module that provides an SQL-callable\nfunction that performs a full index scan using whatever snapshot type\nthe caller asks for. Also add a set of tests that verify and document a\nvariety of behaviors associated with some specific snapshot type.\n\nPreparation for an upcoming commit that will move index scan heap\nfetches behind a new slot-based table AM interface. Aside from testing\nnon-MVCC snapshot types, these new tests can catch bugs where an index\nscan incorrectly maintains xs_heap_continue during HOT chain traversal.\nSuch bugs are unlikely to affect MVCC scans, which makes it hard to\nwrite tests that detect them without relying on a custom test module.\n\nAuthor: Peter Geoghegan \nReviewed-by: Tomas Vondra \nDiscussion: https://postgr.es/m/CAH2-Wzn%2BC%3DmAMv9aW3Skfh80JPpHKT3yM%3DDYwkrrhYyG2f%2B_vg%40mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "tmux", "sha": "ce253864f88c08c97bc7ec32fda1fe50b2220345", "date": "2026-09-03T22:37:09+01:00", "subject": "Test for late run-shell on a client that has exited.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "647a96676a01b223b500fc14d7dcd93489576ee8", "date": "2026-09-03T16:48:44-06:00", "subject": "Make the cached SHA-1 fingerprint internal-only", "body": "X509::sha1_hash exists only so X509_cmp() can compare certificates by\nidentity, yet X509_digest() handed it to callers when asked for SHA-1,\nand it was computed under the certificate's libctx/propq, tying an\ninternal value to the context the certificate was parsed in.\n\nCompute it with the built-in SHA-1 over the DER encoding instead, via\nthe new ossl_x509_internal_fingerprint(), so it is the same in every\nlibrary context; the only remaining failure mode is an allocation\nfailure. Drop the X509_digest() shortcut so the cached value is never\nvisible to callers, and rename the field to fingerprint in X509 and\nX509_CRL. The CRL side gets the same treatment in a following commit.\n\nReviewed-by: Mounir Idrassi \nReviewed-by: Nikola Pajkovsky \nMerge-date: Thu Sep 17 16:44:33 2026\nMerged-from: https://github.com/openssl/openssl/pull/32686", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "2b3d11aaed891958f24aed05bf454524d603b6ea", "date": "2026-09-04T07:50:46+09:00", "subject": "Minimize footprint of TOAST_MAX_CHUNK_SIZE in heap and amcheck code", "body": "This refactoring eases a little bit a follow-up change to add support\nfor 8-byte TOAST value IDs, as the maximum chunk size allowed for a\nsingle chunk of TOASTed data depends on the size of varatt_external, its\nvalue ID being part of it.\n\nIt is worth noting that TOAST_MAX_CHUNK_SIZE resolves as a uintptr_t, so\nthe previous calculations were in 64-bit unsigned arithmetic. These are\nnow done in int32, which still works the same way due to an external\nTOAST value being bounded to 1GB.\n\nAuthor: Michael Paquier \nReviewed-by: Yugo Nagata \nReviewed-by: Bharath Rupireddy \nDiscussion: https://postgr.es/m/CALj2ACVnO6jjs5aRx3dSOBy9LV7=iakPBR1eOvoeB8rL8Q47zg@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "e12ed4ebaaf7a56b206bdd39e90f39a531250fa1", "date": "2026-09-03T17:11:33-06:00", "subject": "Make the cached CRL fingerprint internal-only", "body": "Give X509_CRL::fingerprint the same treatment as the certificate one:\ncompute it with the built-in SHA-1 via ossl_x509_internal_fingerprint()\nso it does not depend on the CRL's libctx/propq, and drop the\nX509_CRL_digest() shortcut that handed the cached value to callers.\nEXFLAG_NO_FINGERPRINT now only signals an allocation failure; update\nits documentation accordingly.\n\nReviewed-by: Mounir Idrassi \nReviewed-by: Nikola Pajkovsky \nMerge-date: Thu Sep 17 16:44:34 2026\nMerged-from: https://github.com/openssl/openssl/pull/32686", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "403dbe58caa08b30ed2c63dca42db5d952a6ae32", "date": "2026-09-03T17:28:44-06:00", "subject": "Compute X509_get_signature_info() on demand, not in the cache", "body": "The signature info was computed once by ossl_x509v3_cache_extensions()\nand stored in X509::siginf. Its security bits depend on a digest fetch\nunder the certificate's libctx/propq, so the cached answer was frozen\nto the context the certificate was parsed in.\n\nCompute it in X509_get_signature_info() instead. The siginf field and\nossl_x509_init_sig_info() are removed.\n\nReviewed-by: Mounir Idrassi \nReviewed-by: Nikola Pajkovsky \nMerge-date: Thu Sep 17 16:44:34 2026\nMerged-from: https://github.com/openssl/openssl/pull/32686", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "90c3f5110fc66dce374512738f2667c29eca2c5e", "date": "2026-09-03T17:51:43-06:00", "subject": "Use SipHash for the internal fingerprint", "body": "Replace SHA-1 with SipHash-2-4 (fixed key, 64-bit output) for the\ninternal X509 and X509_CRL fingerprint.\n\nAlways build crypto/siphash; no-siphash now only removes the provider\nMAC.\n\nReviewed-by: Mounir Idrassi \nReviewed-by: Nikola Pajkovsky \nMerge-date: Thu Sep 17 16:44:35 2026\nMerged-from: https://github.com/openssl/openssl/pull/32686", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "9d036bdd8fb794363feb2f3e02dacd675b497935", "date": "2026-09-04T10:22:41+10:00", "subject": "params: add directive for dealing with duplicated parameters", "body": "The generated param name decoders will, by default, return an error\nif a duplicate param name is encountered. Although passing a repeated\nparameter is a bug in the caller, this isn't always desirable behaviour.\n\nAdd the ability to accept the first occupance and to ignore later ones.\n\nReviewed-by: Viktor Dukhovni \nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Tomas Mraz \nMerge-date: Mon Sep 7 18:14:54 2026\nMerged-from: https://github.com/openssl/openssl/pull/32675", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "68b90853eef734f39b3157a5fdf7b1db0a9e268a", "date": "2026-09-04T10:24:56+10:00", "subject": "ec: accept duplicate instances of the point format parameter", "body": "The first value is kept. This is done for compatibility with\nold FIPS providers from <=4.0 branches. They export two point\nformat parameters with different values.\n\nReviewed-by: Viktor Dukhovni \nReviewed-by: Eugene Syromiatnikov \nReviewed-by: Tomas Mraz \nMerge-date: Mon Sep 7 18:14:55 2026\nMerged-from: https://github.com/openssl/openssl/pull/32675", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "ant-design", "sha": "4e86b57c2d2fd6c9ca6532c37e22da9c536d5b62", "date": "2026-09-04T08:41:37+08:00", "subject": "docs: mark mask object fields optional in Modal zh-CN doc", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "postgres", "sha": "cbec572918fd1dd197e13e21952d2b5f0be2e942", "date": "2026-09-04T10:33:39+09:00", "subject": "Disallow joins whose lateral references need an unformable outer join", "body": "If a LATERAL subquery references a PlaceHolderVar that must be\nevaluated at an outer join, the subquery's lateral_relids include that\nouter join's relid, since the lateral parameter is the outer join's\noutput. join_is_legal() verified that the rels named by a proposed\njoin's minimum parameterization could still be joined to from outside,\nbut it did not consider such outer-join relids. As a result, when\nidentity 3 permitted a commuted join order, we could approve a join\nthat includes part of that outer join's required input, even though\nthe outer join could then be completed only above the proposed join,\nleaving the lateral parameter forever unsatisfiable.\n\nIn assert-enabled builds this tripped the backstop Assert in\ntry_nestloop_path(). In production builds the bogus join generated\nonly paths whose parameterization can never be satisfied, so they\ncould never appear in a complete plan. This not only wasted planning\neffort, but also could mislead the clauseless-join heuristics into\nthinking that legality of this join means that some other join rel\nneed not be formed, and that could lead to failure to find any plan at\nall.\n\nSuch joins used to be rejected by the have_dangerous_phv() check in\njoin_is_legal(), which refused any join where a PlaceHolderVar's eval\nset overlapped the other side of the join without being contained in\nit. Commit a16ef313f removed that restriction, exposing this\noversight.\n\nTo fix, teach join_is_legal() to reject a proposed join if its minimum\nparameterization includes an outer-join relid whose outer join cannot\nbe formed strictly outside the join, that is, when any rel needed to\nform that outer join must become part of the proposed join's own join\ntree.\n\nBack-patch to v18, where commit a16ef313f removed the\nhave_dangerous_phv() restriction.\n\nAuthor: Richard Guo \nReviewed-by: Tom Lane \nReviewed-by: Tender Wang \nDiscussion: https://postgr.es/m/CAMbWs49ZveAYx9T5b0=gxn0pnvqq=MbN-Z", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "e8d6a4d3399c739772080d70c7f3c4d548a637c9", "date": "2026-09-04T12:30:05+09:00", "subject": "fix(build): keep hash placeholders as-is in `resolveFileUrl` hook", "body": "", "style": "conventional", "type": "fix", "scope": "build", "breaking": false} {"repo": "angular", "sha": "cf24431aa20f9e0bc348d0283ca49fd2b1e4ddfc", "date": "2026-09-04T11:29:01+06:00", "subject": "fix(docs-infra): render signal inputs, outputs and models using initializer syntax", "body": "The API reference rendered signal-based bindings with a hybrid line mixing decorator\nsyntax and the signal type, e.g. `readonly @Input('formField') field: InputSignal>;`\nfor a member declared as `readonly field = input.required>({alias: 'formField'});`.\nSignal bindings are never declared with `@Input()` / `@Output()`, so the rendered code did\nnot match how the source is authored.\n\nSignal inputs, outputs and models are now rendered in the initializer form the source uses:\n\n readonly field = input.required>({alias: 'formField'});\n readonly nameChange = output({alias: 'name'});\n readonly value = model();\n\n`PropertyEntry` carries no `isSignal` flag, so the discriminator is the resolved property\ntype: `InputSignal` / `InputSignalWithTransform` for inputs, `OutputEmitterRef` for outputs\nand `ModelSignal` for models. A model is tagged as both an input and an output, so both tags\nare stripped and only its input alias is passed to `model()` — the output alias is always\nderived from it. `outputFromObservable()` is deliberately excluded: it is typed as the\nbroader `OutputRef` and is not declared with an `output()` initializer. Decorator-based\ninputs and outputs continue to render with `@Input(...)` / `@Output(...)`.\n\n(cherry picked from commit 459aafb2fc05fe1e9dfda11672f8e7940d71106e)", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "459aafb2fc05fe1e9dfda11672f8e7940d71106e", "date": "2026-09-04T11:29:01+06:00", "subject": "fix(docs-infra): render signal inputs, outputs and models using initializer syntax", "body": "The API reference rendered signal-based bindings with a hybrid line mixing decorator\nsyntax and the signal type, e.g. `readonly @Input('formField') field: InputSignal>;`\nfor a member declared as `readonly field = input.required>({alias: 'formField'});`.\nSignal bindings are never declared with `@Input()` / `@Output()`, so the rendered code did\nnot match how the source is authored.\n\nSignal inputs, outputs and models are now rendered in the initializer form the source uses:\n\n readonly field = input.required>({alias: 'formField'});\n readonly nameChange = output({alias: 'name'});\n readonly value = model();\n\n`PropertyEntry` carries no `isSignal` flag, so the discriminator is the resolved property\ntype: `InputSignal` / `InputSignalWithTransform` for inputs, `OutputEmitterRef` for outputs\nand `ModelSignal` for models. A model is tagged as both an input and an output, so both tags\nare stripped and only its input alias is passed to `model()` — the output alias is always\nderived from it. `outputFromObservable()` is deliberately excluded: it is typed as the\nbroader `OutputRef` and is not declared with an `output()` initializer. Decorator-based\ninputs and outputs continue to render with `@Input(...)` / `@Output(...)`.", "style": "conventional", "type": "fix", "scope": "docs-infra", "breaking": false} {"repo": "angular", "sha": "635ef1bd80ee93e94babe7126e298ca2557eb646", "date": "2026-09-04T09:05:47+03:00", "subject": "refactor(devtools): contain hydration overlays functionality on the be", "body": "Move the refresh mechanism of the hydration overlays to the backend,\ndelegating only the feature toggling to the frontend.", "style": "conventional", "type": "refactor", "scope": "devtools", "breaking": false} {"repo": "openssl", "sha": "1e2c0ca19053db4e7ffe53eea3c994d5656ca40f", "date": "2026-09-04T08:43:05+02:00", "subject": "Don't cleanup uninitialised variable hdr", "body": "Fixes coverity issues 1700551.\n\nSigned-off-by: Norbert Pocs \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Kurt Roeckx \nMerge-date: Tue Sep 15 10:15:31 2026\nMerged-from: https://github.com/openssl/openssl/pull/32677", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "d6d07e77cf485628ddfcdc916d71499fdba80be5", "date": "2026-09-04T08:43:05+02:00", "subject": "Don't cleanup uninitialised variable hdr", "body": "Fixes coverity issues 1700551.\n\nSigned-off-by: Norbert Pocs \nReviewed-by: Nikola Pajkovsky \nReviewed-by: Kurt Roeckx \nMerge-date: Tue Sep 15 10:14:43 2026\nMerged-from: https://github.com/openssl/openssl/pull/32677", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "25317c38370c3d5f1cee54793fa776053b26ed79", "date": "2026-09-04T06:51:02Z", "subject": "build: update dependency cypress to v16", "body": "See associated pull request for more information.", "style": "conventional", "type": "build", "scope": "", "breaking": false} {"repo": "postgres", "sha": "ac1ccbea98f166548890f14c9ca9d4ab055ca291", "date": "2026-09-04T15:51:56+09:00", "subject": "pg_upgrade: Skip tests for live server checks on Windows", "body": "The tests added by 88aa6ca138a1 could fail on Windows, due to the fact\nthat pg_ctl is not able to compare the PID stored in postmaster.pid with\nthe postmaster that has been spawned. The postmaster may not be seen as\nrunning, and could fail the live check.\n\nNote that pg_ctl falls back on checking the start time recorded in\npostmaster.pid, with a lag of 2 seconds. So, outside the possibility of\nfixing pg_ctl so as the PID file checks become more reliable, an\nalternative solution would be to use the same trick as the pg_ctl test\n001_start_stop.pl, with a hardcoded sleep of *3* seconds, at least. I\ncannot get excited about spreading this pattern, though, so skip it is\nas we do not lose any coverage.\n\nPer report from buildfarm members fairywren.\n\nDiscussion: https://postgr.es/m/apphYnTYS1WtLkKU@paquier.xyz", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "9a012386b818da9530953da1967e6d01cc39e479", "date": "2026-09-04T09:07:37+02:00", "subject": "connect: fix for first ms connection shutdown", "body": "When a connection needed to start shutdown in the first millisecond\nsince its creation, the changed `start_ms` indicator did not change.\n\nInitialize `start_ms` with -1 so that 0ms are treated as a shutdown\nstart.\n\nCloses #22821", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "56f6dca25772ef5c674744640a265d63fc48af97", "date": "2026-09-04T12:57:49+05:30", "subject": "chore: rm stale code", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "neovim", "sha": "034fca6c539ce33760cad3d01db5e8473f61c4ac", "date": "2026-09-04T09:27:53+02:00", "subject": "fix(move): zz does not center the cursor with wrapped lines #41656", "body": "Problem:\n`scroll_cursor_halfway()` resets its `above` and `below` row counters on\nevery iteration of the `topline` scan, because they are declared inside\nthe loop. Each iteration then takes one line from each side regardless\nof how many screen rows those lines occupy, so tall wrapped lines below\nthe cursor leave it near the top of the window instead of centered.\n\nAffects the default `nosmoothscroll` path, reached from \"zz\" and from\n`update_topline()`. Regression from 9b9ccac6256 (vim-patch:9.0.1121),\nwhich moved the declarations to their first use while porting; the\nupstream patch did not touch them.\n\nSolution:\nDeclare the counters at function scope, as Vim does, so they accumulate\nacross iterations.", "style": "conventional", "type": "fix", "scope": "move", "breaking": false} {"repo": "eslint", "sha": "e9163a4245b0850d00ce2c1b0808bc5ed38fedf2", "date": "2026-09-04T13:00:21+05:30", "subject": "refactor: simplify logic", "body": "", "style": "conventional", "type": "refactor", "scope": "", "breaking": false} {"repo": "openssl", "sha": "9b8037e9457e93b4712d31d0bb090fbe05462be7", "date": "2026-09-04T09:32:27+02:00", "subject": "Make the extended tests CI jobs run on ready_for_review", "body": "This allows avoiding full CI re-run by the otherwise necessary\nclose & reopen of the PR to re-run these jobs.\n\nReviewed-by: Paul Dale \nReviewed-by: Nikola Pajkovsky \nMerge-date: Wed Sep 16 09:04:35 2026\nMerged-from: https://github.com/openssl/openssl/pull/32679", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "93afea22da01b5afeb1f98e1f52cfeef577f147b", "date": "2026-09-04T09:32:27+02:00", "subject": "Make the extended tests CI jobs run on ready_for_review", "body": "This allows avoiding full CI re-run by the otherwise necessary\nclose & reopen of the PR to re-run these jobs.\n\nReviewed-by: Paul Dale \nReviewed-by: Nikola Pajkovsky \nMerge-date: Wed Sep 16 09:04:15 2026\nMerged-from: https://github.com/openssl/openssl/pull/32679", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "5931ff628144262c0989c15137b999aa97cff17d", "date": "2026-09-04T09:32:27+02:00", "subject": "Make the extended tests CI jobs run on ready_for_review", "body": "This allows avoiding full CI re-run by the otherwise necessary\nclose & reopen of the PR to re-run these jobs.\n\nReviewed-by: Paul Dale \nReviewed-by: Nikola Pajkovsky \nMerge-date: Wed Sep 16 09:01:06 2026\nMerged-from: https://github.com/openssl/openssl/pull/32679", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "577b47ca5f2dc54fc90e3357dcaf7c2c01baaf08", "date": "2026-09-04T09:32:27+02:00", "subject": "Make the extended tests CI jobs run on ready_for_review", "body": "This allows avoiding full CI re-run by the otherwise necessary\nclose & reopen of the PR to re-run these jobs.\n\nReviewed-by: Paul Dale \nReviewed-by: Nikola Pajkovsky \nMerge-date: Wed Sep 16 09:00:16 2026\nMerged-from: https://github.com/openssl/openssl/pull/32679", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "97995c4185702a43e6553a037b4273a343f551fa", "date": "2026-09-04T09:32:27+02:00", "subject": "Make the extended tests CI jobs run on ready_for_review", "body": "This allows avoiding full CI re-run by the otherwise necessary\nclose & reopen of the PR to re-run these jobs.\n\nReviewed-by: Paul Dale \nReviewed-by: Nikola Pajkovsky \nMerge-date: Wed Sep 16 08:55:44 2026\nMerged-from: https://github.com/openssl/openssl/pull/32679", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "openssl", "sha": "1cdca2624e65accddcc6e6f46c3cc65098991929", "date": "2026-09-04T09:32:27+02:00", "subject": "Make the extended tests CI jobs run on ready_for_review", "body": "This allows avoiding full CI re-run by the otherwise necessary\nclose & reopen of the PR to re-run these jobs.\n\nReviewed-by: Paul Dale \nReviewed-by: Nikola Pajkovsky \nMerge-date: Wed Sep 16 08:55:22 2026\nMerged-from: https://github.com/openssl/openssl/pull/32679", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "03a317355fadd5297265f4767def019585eaa80d", "date": "2026-09-04T13:06:46+05:30", "subject": "refactor: improve logic as per coderabbit", "body": "", "style": "conventional", "type": "refactor", "scope": "", "breaking": false} {"repo": "vue-core", "sha": "3cff5931343cf542e59e251714016df408d89a59", "date": "2026-09-04T15:37:14+08:00", "subject": "fix(runtime-vapor): resolve hydration anchors by SSR fragment range ownership", "body": "", "style": "conventional", "type": "fix", "scope": "runtime-vapor", "breaking": false} {"repo": "neovim", "sha": "18f5d47543dfa21c915f2f7d42b0d6b1da9fdd8f", "date": "2026-09-04T03:37:48-04:00", "subject": "fix(multicursor): follow cursor-move by API #41675", "body": "Problem:\nA non-edit mapping that moves the cursor by API (Lua,\n`:call nvim_win_set_cursor()`, …) is not cascaded.\n\nSolution:\nFallback to LHS-replay if the cursor moved and follow-mode is enabled.", "style": "conventional", "type": "fix", "scope": "multicursor", "breaking": false} {"repo": "eslint", "sha": "264b4346d1963701df0c398b4aeb2f6e8b2af93e", "date": "2026-09-04T16:43:19+09:00", "subject": "feat: add `d` and `v` flags to `no-unexpected-multiline`", "body": "", "style": "conventional", "type": "feat", "scope": "", "breaking": false} {"repo": "curl", "sha": "c53420e42742a68004854a7fa354f9500b4b1516", "date": "2026-09-04T00:48:12-07:00", "subject": "ossl_verifyhost: remove assumption of null termination of ASN1_STRING.", "body": "This only affects hosts where the deprecated CN is used for matching DNS\nnames, instead of the modern way, which is by using a dNSName-typed SAN.\n\nA similar change but for certs that store the hostname in the SAN (as\nthey should nowadays) was in 543adc4de024dda4f3b8f368893c7d01a3cf0ea8.\n\nFixes #22822\nCloses #22824", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "ac00b0f8bdc59152e0d5b5f97c979595e9a5e2ad", "date": "2026-09-04T16:55:39+09:00", "subject": "Add pgstat test for reset of dead/live counters with table truncation", "body": "A patch has been proposed to refactor the table-level statistics to\ntreat the non-transactional and transactional parts as two separate\nentities, in the scope of a larger piece of work to introduce\ntransactional flushes.\n\nThis commit adds a test for a pattern that has proved to be problematic\nwhile splitting the relation stats data, and where we had no coverage\nyet: an entry marked as truncdropped resets the live/dead counters in\nthe flush callback, leaving any other data intact. The code path\nresetting the live/dead counters was covered, but nothing in the\nregression tests checked the stats views for it.\n\nExtracted from a larger patch by the same author.\n\nAuthor: Sami Imseih \nReviewed-by: Bertrand Drouvot \nReviewed-by: Michael Paquier \nDiscussion: https://postgr.es/m/aoU32VTkUtpdUewB@paquier.xyz", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "13f3455a6034d40871d805eb17eb7b27b22785f1", "date": "2026-09-04T10:58:45+03:00", "subject": "docs: correct filename labels on code examples", "body": "Four labels name a different file than the block loads: three tabs on the\nanimations guide say `leave.*` but load `leave-parent.*` (`leave.*` is a\nseparate example shown earlier on the page), and one block in the reactive\nforms section says `actor-form-template.component.html` but loads\n`actor-form-reactive.component.html`.\n\nThe other three leak the `.1`/`.2` variant suffix, which every other\nnumbered-variant header in adev strips.\n\nhttps://angular.dev/guide/animations\nhttps://angular.dev/guide/legacy-animations/reusable-animations\nhttps://angular.dev/guide/forms/form-validation\n\n(cherry picked from commit 3481b151676d0e8fc2721553f3219c56cbbda886)", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "angular", "sha": "3481b151676d0e8fc2721553f3219c56cbbda886", "date": "2026-09-04T10:58:45+03:00", "subject": "docs: correct filename labels on code examples", "body": "Four labels name a different file than the block loads: three tabs on the\nanimations guide say `leave.*` but load `leave-parent.*` (`leave.*` is a\nseparate example shown earlier on the page), and one block in the reactive\nforms section says `actor-form-template.component.html` but loads\n`actor-form-reactive.component.html`.\n\nThe other three leak the `.1`/`.2` variant suffix, which every other\nnumbered-variant header in adev strips.\n\nhttps://angular.dev/guide/animations\nhttps://angular.dev/guide/legacy-animations/reusable-animations\nhttps://angular.dev/guide/forms/form-validation", "style": "conventional", "type": "docs", "scope": "", "breaking": false} {"repo": "curl", "sha": "c5f9b8b40527ead840d9bbe19220e0b8515951eb", "date": "2026-09-04T10:00:17+02:00", "subject": "ftp: tighten use_ssl check for conn reuse", "body": "CURLUSESSL_* needs a bit more special check on connection reuse. Except\nCURLUSESSL_TRY, we expect the reused connection to have exactly the same\nsetting. On TRY, we are happy with anything that is no NONE.\n\nCloses #22823", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "9d8153223dbf47b9aecdc1474202aaee4845f146", "date": "2026-09-04T11:07:16+03:00", "subject": "fix: support `__proto__` in `/* exported */` comments", "body": "* fix: support `__proto__` in `/* exported */` comments\n\n* bump @eslint/plugin-kit", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "curl", "sha": "a226b5ea042831409b9d5899b170e835eb3dd960", "date": "2026-09-04T10:08:36+02:00", "subject": "CURLOPT*ISSUERCERT*: not verified cryptographically", "body": "Reported-by: Stanislav Fort\n\nCloses #22825", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "f042c25fb876d0da5ff5f35332bcc3a8b5a7fe66", "date": "2026-09-04T10:16:41+02:00", "subject": "openssl/gtls: reject CRLfile with native CA store, matching rustls", "body": "Closes #22826", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "649015d10abd6dcc2040589fab2d63b19fbe783a", "date": "2026-09-04T10:20:11+02:00", "subject": "ci(macos): skip brew upgrade", "body": "Problem: Support for Intel macOS is increasingly being dropped since\nApple itself retired the architecture (with macOS 27 already being\nARM-only). In particular, homebrew is already dropping support, with\nsome 3rd-party taps like `hashicorp/tap/vagrant` (that are enabled on GH\nrunners by default) no longer being available. This makes `brew upgrade`\nfail on Intel.\n\nSolution: Skip `brew upgrade` in `install_deps.sh`, as we don't rely on\nthe bleeding edge version of the installed dependencies. This skips\nchecking the tap (which we don't use) and hence lets the workflows pass.\n(It also shaves of some seconds from the CI time.)\n\nNote: This is strictly speaking a GHA issue and should be handled by\nadapting the runner images; but since GHA has already announced that\nthey will drop all Intel macOS images in 2027, this is unlikely to\nhappen.\n\nIt also means that this is a stopgap solution to allow 0.13 to keep Tier\n1 support for Intel macOS, but next year we'll have to drop support as\nwell.", "style": "conventional", "type": "ci", "scope": "macos", "breaking": false} {"repo": "angular", "sha": "3dfc855381a29b359d8d79da22551eaa69edafe7", "date": "2026-09-04T11:34:28+03:00", "subject": "test(devtools): run the orphaned supported-apis spec", "body": "`supported-apis.spec.ts` was added in #60585, in a commit that also edited the\n`ts_test_library` three lines below the `srcs` it was left out of. No revision\nof that BUILD file has ever listed it, so it has not run since March 2025. The\ntarget goes from 12 specs to 14.\n\nWiring it up alone would not have worked. Every `*IsSupported` helper calls\n`ngDebugClient()`, which throws when `window.ng` is undefined, and the old\n`expect(supported).toBeTruthy()` set no `ng` at all. It now stubs `ng` and\nchecks the flag set and that each flag tracks its own debug API.\n\n`ng-debug-api.spec.ts` adds an `[ng-version]` root and did not remove it, which\n`getAppRoots()` then picks up in the other file under jasmine's random\nordering, so it now clears the DOM in its own `afterEach`.\n\n`glob` matches the sibling `directive-forest/component-tree` target and keeps\nthe next spec in this directory from being dropped the same way.", "style": "conventional", "type": "test", "scope": "devtools", "breaking": false} {"repo": "curl", "sha": "00271cfcb95b6481a1f91ce0a0dad5d64d2e89a9", "date": "2026-09-04T10:47:32+02:00", "subject": "schannel: auto_client_cert conn matching", "body": "Move flag `auto_client_cert` into the connection SSL config struct so it\nis part of connection matching.\n\nCloses #22827", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "d97de71dcf3c94dd5a70eff630ab790beb07b97b", "date": "2026-09-04T10:55:40+02:00", "subject": "ci(macos): untap untrusted aws/tap on macOS", "body": "Problem: The 3rd-party default homebrew tap `aws/tap` is untrusted,\nwhich now triggers a warning in our macOS workflows.\n\nSolution: \"Untap\" (disable) the tap (which we don't use).", "style": "conventional", "type": "ci", "scope": "macos", "breaking": false} {"repo": "curl", "sha": "ea90a974463196953b1863c850fd2f3329169abc", "date": "2026-09-04T11:17:30+02:00", "subject": "digest: check peer equality for staleness", "body": "Check that current peer is *equal* from last peer's state information.\nThis accomodates also scheme changes, which the previous check did not.\n\nCloses #22828", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vitest", "sha": "9e1166959e14bd32298d8a0e85352431c769ec7a", "date": "2026-09-04T11:23:19+02:00", "subject": "chore: add more rules to AGENTS.md regarding ai contributions", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "postgres", "sha": "f4012672bb2e5f1b0a4641826e14e9d209049dea", "date": "2026-09-04T18:40:01+09:00", "subject": "postgres_fdw: Fix issues with statistics import.", "body": "When analyzing a foreign table pointing to a partitioned or inherited\ntable on the remote server, this method just imported relpages and\nreltuples for the parent table from the remote pg_class, causing a\ndiscrepancy or an error in results compared to the sampling method.\n\n* In the partitioning case, while the sampling method sets relpages=0,\n this method set relpages=-1. (For reltuples, the results of both\n methods match.)\n* In the inheritance case, while the sampling method sets\n reltuples=(# of rows in the parent and child tables), this method set\n reltuples=(# of rows in the parent table). (For relpages, the results\n of both methods match.)\n\nAs relpages and reltuples are used for costing foreign paths, these\nwould lead to incorrect plans or plan changes depending on the method.\nTo fix, do the following:\n\n* For the partitioning case, modify postgres_fdw to set relpages=0 to\n match the sampling method.\n* For the inheritance case, we could also modify it to set the correct\n reltuples value by fetching reltuples for the child tables as well,\n but 1) the change isn't small, and 2) table inheritance hasn't been\n used that much these days, so fix by disabling this method.\n\nWhile at it, adjust comments and docs slightly.\n\nOversight in commit 28972b6fc.\n\nAuthor: Etsuro Fujita \nDiscussion: https://postgr.es/m/CAPmGK173b2Xvyoeq%3DY7L09H%2BajgMgxC2fYe%3D7CCunZz8tGTfFA%40mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "vite", "sha": "3b8c45b0a07a29917b40d325912fbc55677b014b", "date": "2026-09-04T17:53:37+08:00", "subject": "chore: enable `minimumReleaseAgeStrict`", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "vite", "sha": "12e709ca4df1059747db1cb7c5d1cd71aba79a24", "date": "2026-09-04T18:12:06+08:00", "subject": "fix(html): don't inline preload link targets (fix #13355)", "body": "", "style": "conventional", "type": "fix", "scope": "html", "breaking": false} {"repo": "vite", "sha": "8492422b8f110625a90c702f42f30784e8cf19dc", "date": "2026-09-04T18:13:10+08:00", "subject": "fix: resolve the actual package root in findNearestMainPackageData for nested package.json", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "curl", "sha": "020bc8dfbaefa9fa0b61e0993a6463b011598ab3", "date": "2026-09-04T12:26:36+02:00", "subject": "digest: fix maximum length quoted value parsing", "body": "Separate the scan limit from the content buffer limit so maximum length\nDigest quoted values can be parsed correctly.\n\nCloses #22830", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "2124f968a696c540fb328c21babb433895878dc7", "date": "2026-09-04T06:28:42-04:00", "subject": "feat(multicursor): \"]C\"/\"[C\" leave a cursor at old position #41681", "body": "Problem:\n`]C` jumps the primary onto another cursor, which dedupes at the next\nedit. Every `]C` consumes a cursor.\n\nSolution:\n`]C` adds a cursor at the current position before jumping. This\neffectively \"rotates\" the primary cursor.", "style": "conventional", "type": "feat", "scope": "multicursor", "breaking": false} {"repo": "curl", "sha": "ca6d0a8ee88b5d778c7d2a327928d7f05da2de7f", "date": "2026-09-04T12:37:17+02:00", "subject": "http proxy: some code cleanups", "body": "- dedup code for CONNECT/CONNECT-UDP\n- pass proxy peer to h1/h2 filters for use in\n CONNECT reuqests (UDP needs it)\n\nCloses #22829", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "eslint", "sha": "b3d876b46083d67899eb1d9613118c1c583632a2", "date": "2026-09-04T12:47:31+02:00", "subject": "chore: disable npm audit in ecosystem tests", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false} {"repo": "curl", "sha": "7c0c7ca57b9774bff5e17f8ecf00db0cf5d7b67b", "date": "2026-09-04T12:14:34+01:00", "subject": "hash: select key operations by type", "body": "Store a key type in each Curl_hash instead of hash and comparison\nfunction pointers, resolving the type to shared callbacks internally.\n\nUse byte keys for general hashes and native curl_socket_t keys for the\nmulti event socket hash.\n\nCloses #22808", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "4c0a82810fbb21fea6fabd24597e8e1e9115e98e", "date": "2026-09-04T13:42:42+02:00", "subject": "ssl: Fix expected output for OpenSSL 4", "body": "When adding support for OpenSSL 4, one occurrence of OpenSSL error\nmessage matching was missed in the backpatches to PostgreSQL 17 and\n16.\n\nReported-by: Christoph Berg \nDiscussion: https://postgr.es/m/apky-gqiLVtFkR-w@msg.df7cb.de\nBackpatch-through: 16,17", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "curl", "sha": "9be5c6849225f2dbb3bf60249998c33575dd0910", "date": "2026-09-04T14:04:53+02:00", "subject": "digest: parse without storing 'value' in local buffer", "body": "Use of curlx_str_ functions avoids having a copy.\n\nCloses #22831", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "d6d51a0ef8234cc564fda74d244fa95e6deb30dc", "date": "2026-09-04T15:13:23+03:00", "subject": "test(devtools): run the orphaned app component spec", "body": "`app.component.spec.ts` has been in the tree since the initial DevTools commit\nin 2020, but no revision of `shell-browser/src/app/BUILD.bazel` has ever listed\nit, and that file has no globs. Two of the three specs in the directory are\nwired up; this one has never run.\n\nIt could not run as written. The setup declared `AppComponent` and imported\n`RouterModule.forRoot([])`, neither of which applies now that the component is\nstandalone and injects `MessageBus` and `DEEP_LINK_INSTANCE_ID`.\n\nIt now stubs the `chrome.devtools` surface that `ngOnInit` reads and covers\nboth lifecycle hooks: the listeners registered on init, the two performance\ntrack emits, the three guards on the deep link listener, and the listeners\nremoved on destroy.\n\nFollow-up to #70570, which fixed the same problem in ng-devtools-backend.", "style": "conventional", "type": "test", "scope": "devtools", "breaking": false} {"repo": "eslint", "sha": "427ac0a014066c36aa57fa8fa9af20fd9fb591e1", "date": "2026-09-04T14:58:59+02:00", "subject": "fix: use format strings in debug calls", "body": "* fix: use format strings in debug calls\n\n* tweaks\n\n* remove debug output for `ConfigLoader` class name\n\n* no suggestion if the template literal contains a format specifier\n\n* clarify comments\n\n* move the check for multiple arguments to the fixer function\n\n* remove sperfluous parameter", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "neovim", "sha": "6112d4a6741e10401fce0d7728cd89398878a77a", "date": "2026-09-04T15:42:09+02:00", "subject": "fix(fuzzy): fuzmatch_str_free() frees the wrong element #41691", "body": "Problem:\nThe loop frees `fuzmatch[count].str` on every iteration instead of\n`fuzmatch[i].str`.\n\nThe function has no callers in Nvim, so there is no impact today. The\nindexing is a slip from the port; Vim's implementation is correct.\n\nSolution:\nIndex with the loop variable, as Vim does.\n\nAI-assisted", "style": "conventional", "type": "fix", "scope": "fuzzy", "breaking": false} {"repo": "eslint", "sha": "bb47dc6da2399a8f76c0c0c3273e6bc314c480e5", "date": "2026-09-04T16:04:45+02:00", "subject": "fix: update dependency file-entry-cache to v11", "body": "* chore: update dependency file-entry-cache to v11\n\n* update lint-result-cache tests\n\n* update eslint tests\n\n* update eslint tests\n\n* fix content strategy\n\n* handle invalid JSON cache files\n\n* add test that fails with file-entry-cache v11\n\n* add another test that fails with file-entry-cache v11\n\n* add another test that fails with file-entry-cache v11\n\n* add more tests that fails with file-entry-cache v11\n\n* use file-entry-cache 11.1.5\n\n* fileEntryCache.create is no longer a getter\n\n* fix flaky test\n\n* remove handling invalid JSON\n\n* delete legacy properties\n\n* update range to exclude v11.1.6\n\n* change variable name in tests\n\n* use format string in debug call\n\nCo-authored-by: fnx <966276+DMartens@users.noreply.github.com>\n\n---------", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "postgres", "sha": "2227afac0171550b0e213173cadcefd165434b23", "date": "2026-09-04T23:04:59+09:00", "subject": "Use pg_parse_lsn() for LSNs in backup manifests", "body": "A user-supplied or modified backup manifest could contain an invalid WAL\nrange LSN that was nevertheless accepted by the manifest parser. For\nexample, an LSN component wider than 32 bits could be silently truncated,\nand trailing characters after an otherwise valid LSN could be ignored.\n\nThis could cause pg_verifybackup, pg_combinebackup, and incremental\nbackup validation to check WAL availability against a different location\nfrom the one specified in the manifest. PostgreSQL-generated manifests\nshould not contain such values, but tools consuming manifests should\nreject invalid input rather than reinterpret it.\n\nFix this by using the common pg_parse_lsn() helper to parse LSNs in\nbackup manifests. It requires the same syntax as the pg_lsn input\nfunction.\n\nAuthor: Ayush Tiwari \nReviewed-by: Yuefei Shi \nReviewed-by: Fujii Masao \nDiscussion: https://postgr.es/m/CAJTYsWXieRHb-ooV2XfjHAtBsS7+P5La_-o8-Cqi15CNDhh9hQ@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "angular", "sha": "3064f3f1dccd78177bf3b86f8ea231102884f0d7", "date": "2026-09-04T07:10:49-07:00", "subject": "feat(router): expose router resources in public API", "body": "Router resources integrate the Angular router with the Signals Resource API, allowing route-level data fetching during navigation transitions.", "style": "conventional", "type": "feat", "scope": "router", "breaking": false} {"repo": "vitest", "sha": "6108b819785bceff03edb2cc7b60e17779f21191", "date": "2026-09-04T18:12:48+03:30", "subject": "fix: exit 1 when vitest list fails collection (fix #11145)", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false} {"repo": "vitest", "sha": "455466c168a9d5cfecd98824860fbdc236f1f9ba", "date": "2026-09-04T22:43:33+08:00", "subject": "fix(types): make public declarations self-contained", "body": "", "style": "conventional", "type": "fix", "scope": "types", "breaking": false} {"repo": "openssl", "sha": "84348f44b4350e6fad355ffadd73e3a4f312231f", "date": "2026-09-04T16:45:27+02:00", "subject": "OSSL_FN: Clear OSSL_FN_CTX memory when freeing it", "body": "This was missed when OSSL_FN_CTX was first implemented, and mirrors\nhow BN_CTX_free() clears all numbers in its pool with BN_clear_free().\nThe arena is a single allocation, so one cleanse of the block covers\neverything carved out of it, in secure and non-secure memory alike.\n\nAssisted-by: Pi:moonshotai/kimi-k3\nSigned-off-by: Richard Levitte \nReviewed-by: Jakub Zelenka \nReviewed-by: Matt Caswell \nReviewed-by: Daniel Kubec \nMerge-date: Tue Sep 22 05:18:34 2026\nMerged-from: https://github.com/openssl/openssl/pull/32682", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "67b593861de13226d5bb23271aeaa16ce67024de", "date": "2026-09-04T09:48:31-05:00", "subject": "Convert FixedParallelState->last_xlog_end to an atomic variable.", "body": "Currently, this variable is an XLogRecPtr protected by a spinlock.\nBy converting it to an atomic variable, we can remove the spinlock.\n\nReviewed-by: Zsolt Parragi \nTested-by: solai v \nDiscussion: https://postgr.es/m/alAJeRRzehDjLaF1%40nathan", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "postgres", "sha": "168fa9dd424e3beabeb0b4c079a032b7b6f5297b", "date": "2026-09-04T16:49:59+02:00", "subject": "doc: Pluralize MXID when referring to Multixact IDs", "body": "Reported-by: Yugo Nagata \nReviewed-by: Pierre Forstmann \nDiscussion: https://postgr.es/m/TY4PR01MB17374866B8EBD3F507AFB1C7C94D12@TY4PR01MB17374.jpnprd01.prod.outlook.com", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "7d29bb79c779351c672cc5e4e7c879370b8ef6f0", "date": "2026-09-04T11:06:00-04:00", "subject": "backport: fix(fuzzy): fuzmatch_str_free() frees the wrong element", "body": "fix(fuzzy): fuzmatch_str_free() frees the wrong element #41691\n\nProblem:\nThe loop frees `fuzmatch[count].str` on every iteration instead of\n`fuzmatch[i].str`.\n\nThe function has no callers in Nvim, so there is no impact today. The\nindexing is a slip from the port; Vim's implementation is correct.\n\nSolution:\nIndex with the loop variable, as Vim does.\n\nAI-assisted\n\n(cherry picked from commit 6112d4a6741e10401fce0d7728cd89398878a77a)", "style": "normal", "type": "", "scope": "", "breaking": false} {"repo": "neovim", "sha": "17b908fc8ea56f4d57bb33981de690203dd072a2", "date": "2026-09-04T17:18:22+02:00", "subject": "fix(menu): menu_get() cannot tell :noremenu from :menu 
` block for those two instead of a\nhighlighted one.\n\nThe result is that on the compatForm API page the first overload's example is\nsyntax highlighted while the other two, which contain identical code, render as\nflat unstyled text.\n\nRebuilding //adev/src/assets:content with and without this change alters 1 of\n1565 pages. That page goes from 6 highlighted and 2 plain code blocks to 8\nhighlighted and none plain.\n\n(cherry picked from commit b9fb45697173d0ce433337822ebeb3bf0087ec97)", "style": "conventional", "type": "docs", "scope": "forms", "breaking": false}
{"repo": "angular", "sha": "b9fb45697173d0ce433337822ebeb3bf0087ec97", "date": "2026-09-09T12:16:18+03:00", "subject": "docs(forms): close the unterminated code fences in the compatForm docs", "body": "Two of the three `compatForm` overloads open a fenced code block in their\n`@example` and never close it. The fence then runs to the end of the tag, so the\nAPI renderer emits a plain `
` block for those two instead of a\nhighlighted one.\n\nThe result is that on the compatForm API page the first overload's example is\nsyntax highlighted while the other two, which contain identical code, render as\nflat unstyled text.\n\nRebuilding //adev/src/assets:content with and without this change alters 1 of\n1565 pages. That page goes from 6 highlighted and 2 plain code blocks to 8\nhighlighted and none plain.", "style": "conventional", "type": "docs", "scope": "forms", "breaking": false}
{"repo": "ant-design", "sha": "8d3c287113763071816b2348fac29059bfe39b8f", "date": "2026-09-09T17:40:25+08:00", "subject": "fix: align picker components with suffix API", "body": "* fix(DatePicker): migrate suffixIcon API to suffix\n\n* test(DatePicker): update renamed suffix demo snapshots\n\n---------", "style": "conventional", "type": "fix", "scope": "", "breaking": false}
{"repo": "curl", "sha": "66aaba7fb33f8dd9867d911b5a24ad3c11239b01", "date": "2026-09-09T11:53:58+02:00", "subject": "hash: change key arguments to const", "body": "hash.h functions had `void *key` as argument in several functions and\nthose should be `const void *key` instead as they are not modified.\n\nSaves a ton of `CURL_UNCONST` use, often on string constants where they\nare definitely not wanted.\n\nCloses #22886", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "eslint", "sha": "26d11bce3e0e21f223613d4bb4be3423839b229e", "date": "2026-09-09T13:38:20+03:00", "subject": "fix: don't report `__proto__` properties in `object-shorthand`", "body": "* fix: don't report `__proto__` properties in `object-shorthand`\n\n* explain __proto__ semantics in the isProtoKey JSDoc", "style": "conventional", "type": "fix", "scope": "", "breaking": false}
{"repo": "vitest", "sha": "23dda738ccacb11682426426cd7b876afd107621", "date": "2026-09-09T12:44:42+02:00", "subject": "fix: share the server on self-referencing `extends`", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false}
{"repo": "postgres", "sha": "40d811540a589cbddeb0736966f2541088134163", "date": "2026-09-09T12:58:00+02:00", "subject": "Fix error message for concurrent repack on materialized views", "body": "Previously, running REPACK (CONCURRENTLY) on a materialized view failed\nwith a confusing \"has no identity index\" error.\n\nREPACK (CONCURRENTLY) replays changes decoded from WAL, but a\nmaterialized view produces none, since it cannot be modified by DML and\nREFRESH rewrites it through a transient heap that is not logically\ndecoded.\n\nFix this by reporting a clear error for materialized views upfront,\nalongside the existing checks for system catalogs and TOAST tables, and\ndocument the limitation.\n\nAuthor: Bharath Rupireddy \nReviewed-by: Kiran Kaki \nBackpatch-through: 19\nDiscussion: https://postgr.es/m/CALj2ACXPvy4_LkQ0nHKgbZjAGK9fkDqRYUvXA3hgwxPm_X7Eng@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "vitest", "sha": "07670364b45e61f2c542d332a232fe81e4f88d82", "date": "2026-09-09T20:06:49+09:00", "subject": "refactor(ui): consolidate unocss style to use class", "body": "", "style": "conventional", "type": "refactor", "scope": "ui", "breaking": false}
{"repo": "neovim", "sha": "cb2e9cdfd3e81e644324ff42aaff6f360610d056", "date": "2026-09-09T07:25:32-04:00", "subject": "fix(multicursor): clipboard=unnamed, explicit \"+yy #41804", "body": "Problem:\n- With `clipboard=unnamed[plus]`, per-cursor yanks are joined into '\"'\n  on exit, but not the clipboard.\n- Multicursor `\"+yy` calls the clipboard provider per-cursor, behaving\n  as \"last wins\".\n\nSolution:\n- Never cascade to/from the clipboard provider.\n- Implicit clipboard (`clipboard=unnamed[plus]`) uses the cursor-local\n  unnamed register, so the cascade needs no `start_batch_changes()`. On\n  exit, the clipboard gets the same joined result as the `\"` register.\n- Operations that reference the explicit clipboard registers `\"+` / `\"*`\n  while multicursor is active, use the cached (primary) clipboard value.", "style": "conventional", "type": "fix", "scope": "multicursor", "breaking": false}
{"repo": "axios", "sha": "f9e8415617290b215dfc02f836ca11f3f5d9ef46", "date": "2026-09-09T13:30:21+02:00", "subject": "build!: migrate release bundles to Rolldown", "body": "", "style": "conventional", "type": "build", "scope": "", "breaking": true}
{"repo": "vitest", "sha": "91ab1588c72c9c5f7c638368a541cd20e0e8a934", "date": "2026-09-09T13:35:58+02:00", "subject": "fix(doctor): measure vm pools for custom environments", "body": "", "style": "conventional", "type": "fix", "scope": "doctor", "breaking": false}
{"repo": "vitest", "sha": "4f8ef0216f637ed2d3e66684982d930f20892347", "date": "2026-09-09T13:36:17+02:00", "subject": "docs: remove v0 link, update vitest API usage", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false}
{"repo": "vitest", "sha": "c119be016295b45a005e2a36367ea7d133b4f385", "date": "2026-09-09T13:37:09+02:00", "subject": "fix(expect): correct return value in `toMatchAriaSnapshot`", "body": "", "style": "conventional", "type": "fix", "scope": "expect", "breaking": false}
{"repo": "curl", "sha": "24c86d189a73ae2eb67448a616a90d393c826780", "date": "2026-09-09T13:56:22+02:00", "subject": "os400: document curl_formadd_CCSID multi-chunk peculiarity", "body": "Closes #22887", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "neovim", "sha": "0a844adc133f732b236b1669c705d9aac888a5a0", "date": "2026-09-09T20:04:23+08:00", "subject": "fix(lsp): newlines in the completion word from insertText #41801", "body": "Problem:\nA textEdit is cut to its first line but an insertText is not, so its\nnewlines reach the |complete-items| word.\n\nSolution:\nCut it the same way.", "style": "conventional", "type": "fix", "scope": "lsp", "breaking": false}
{"repo": "neovim", "sha": "d039f19af54e73e1f076c29f1cb412c14612da07", "date": "2026-09-09T13:04:46+01:00", "subject": "build: enable more EmmyLua checks", "body": "Enable checks for unused code, deprecated calls, return types, and\nannotations.\n\nFix the warnings and replace the deprecated highlight function. Keep\nexceptions where old APIs are still needed or the checker gets it wrong.\n\nAI-assisted", "style": "conventional", "type": "build", "scope": "", "breaking": false}
{"repo": "neovim", "sha": "8e0045f43ade397bbcf62c0d33aaf45e72e63610", "date": "2026-09-09T14:45:10+02:00", "subject": "feat(lsp): support `workspace/foldingRange/refresh` #41780", "body": "Problem:\nServers cannot request refreshing stale LSP folding ranges after a\nproject-wide change.\n\nSolution:\nAdvertise `workspace.foldingRange.refreshSupport` and re-request folding\nranges for active buffers when handling `workspace/foldingRange/refresh`.\n\nAI-assisted", "style": "conventional", "type": "feat", "scope": "lsp", "breaking": false}
{"repo": "curl", "sha": "6ee61c3fe1479a6fbe85238fa83873b87b87f747", "date": "2026-09-09T14:48:29+02:00", "subject": "lib: connection shutdown, cleanup code", "body": "Move all shutdown related functions to cshutdn.c and give them\n`Curl_cshutdn_*` prefix.\n\nAdd connection age handling methods to conncache.h, so that other code\ndoes not need to mess with details of how connections keep time.\n\nCloses #22888", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "tmux", "sha": "483912351048bf60937bb8df950bcea12606a3a2", "date": "2026-09-09T13:50:14+01:00", "subject": "Tweak changes.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "angular", "sha": "312e1d808902116fb8cd4e02d936260113453999", "date": "2026-09-09T15:09:10+02:00", "subject": "feat(compiler-cli): add strictUnclaimedEventNames option to catch misspelled output bindings", "body": "Adds an opt-in type-checking option that reports event bindings on\nelements with matched directives whose name matches neither an output\nof the matched directives nor a known native DOM event.\n\nSince the compiler cannot know the set of custom events that may bubble\nup from descendant elements, the check is heuristic: only\nsingle-identifier camelCase names are reported. Dash-separated names,\nwhich bubbling custom events conventionally use, native DOM events\n(compared ignoring case, to cover vendor-prefixed names such as\nwebkitAnimationEnd), key pseudo-events and targeted events are exempt.\nWith CUSTOM_ELEMENTS_SCHEMA, elements without a matched component are\nexempt as well; NO_ERRORS_SCHEMA disables the check entirely.\n\nThe option is opt-in everywhere for now; enabling it by default under\nstrictTemplates inside google3 (via the sync markers) will follow once\nthe google3 cleanup is complete.\n\nRelated to #43603", "style": "conventional", "type": "feat", "scope": "compiler-cli", "breaking": false}
{"repo": "tmux", "sha": "7049aa826ad3ebb9ded4017b90a88d113c1c67fa", "date": "2026-09-09T14:39:14+01:00", "subject": "Do not allow 0x0 panes and remove a dead member.", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "curl", "sha": "1143bc9ac6912743a0fef78d5ac81426b172976d", "date": "2026-09-09T15:56:00+02:00", "subject": "netrc: survive comment-only .netrc file", "body": "A .netrc file with only a comment and no newline caused an NULL\ndereference\n\nVerify with test 2429\n\nReported-by: Max Dymond\nCloses #22889", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "openssl", "sha": "4da7d5496632142cfd0c857208ad9705ba7fec1f", "date": "2026-09-09T16:11:00+02:00", "subject": "CHANGES.md, NEWS.md: Add a new 4.2 section", "body": "Reviewed-by: Eugene Syromiatnikov \nReviewed-by: Nikola Pajkovsky \nMerge-date: Fri Sep 11 13:05:05 2026\nMerged-from: https://github.com/openssl/openssl/pull/32768", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "openssl", "sha": "85eeff79b822c1923322b07625e492ef67fb8d61", "date": "2026-09-09T16:11:24+02:00", "subject": "Add 4.1 branch to various CI jobs", "body": "Reviewed-by: Eugene Syromiatnikov \nReviewed-by: Nikola Pajkovsky \nMerge-date: Fri Sep 11 13:05:06 2026\nMerged-from: https://github.com/openssl/openssl/pull/32768\n(cherry picked from commit b2335251a885ab5b48302ff59347e7faee2f1672)", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "openssl", "sha": "b2335251a885ab5b48302ff59347e7faee2f1672", "date": "2026-09-09T16:11:24+02:00", "subject": "Add 4.1 branch to various CI jobs", "body": "Reviewed-by: Eugene Syromiatnikov \nReviewed-by: Nikola Pajkovsky \nMerge-date: Fri Sep 11 13:05:06 2026\nMerged-from: https://github.com/openssl/openssl/pull/32768", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "neovim", "sha": "0c0b78c79e8b6eadf585e79974753b8f6e0f156d", "date": "2026-09-09T16:21:46+02:00", "subject": "fix(ui2): :quit on unsaved buffer shows hit-enter prompt #41575", "body": "Problem:\n\"Press any key\" modal prompt is shown when `:quit`\nfrom unsaved buffer, even if ui2 is enabled (which \nshould never show hit-enter prompt).\n\nSolution:\nSkip `wait_return` for `ext_messages` UIs.", "style": "conventional", "type": "fix", "scope": "ui2", "breaking": false}
{"repo": "openssl", "sha": "5e141884fb62368e57c31b9ffbc5fcb4e2fb679d", "date": "2026-09-09T16:23:35+02:00", "subject": "Remove 3.0 branch from various CI jobs as it is EOL", "body": "Reviewed-by: Eugene Syromiatnikov \nReviewed-by: Nikola Pajkovsky \nMerge-date: Fri Sep 11 13:05:07 2026\nMerged-from: https://github.com/openssl/openssl/pull/32768\n(cherry picked from commit 2238a94895d90cf221ae804d385a063280b81d0e)", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "openssl", "sha": "2238a94895d90cf221ae804d385a063280b81d0e", "date": "2026-09-09T16:23:35+02:00", "subject": "Remove 3.0 branch from various CI jobs as it is EOL", "body": "Reviewed-by: Eugene Syromiatnikov \nReviewed-by: Nikola Pajkovsky \nMerge-date: Fri Sep 11 13:05:07 2026\nMerged-from: https://github.com/openssl/openssl/pull/32768", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "angular", "sha": "5dabf81bdfc01e8d0ae7edc326b693f86c53dfc7", "date": "2026-09-09T09:24:37-05:00", "subject": "test(forms): migrate integration tests to zoneless scheduling", "body": "Use whenStable instead of manual change detection so integration tests exercise scheduled rendering. Await asynchronous callbacks and remove redundant timer waits to keep assertions within the test lifecycle.\n\n(cherry picked from commit 8afd85c6d752207f975718dbc6306c5ae079f2ca)", "style": "conventional", "type": "test", "scope": "forms", "breaking": false}
{"repo": "angular", "sha": "8afd85c6d752207f975718dbc6306c5ae079f2ca", "date": "2026-09-09T09:24:37-05:00", "subject": "test(forms): migrate integration tests to zoneless scheduling", "body": "Use whenStable instead of manual change detection so integration tests exercise scheduled rendering. Await asynchronous callbacks and remove redundant timer waits to keep assertions within the test lifecycle.", "style": "conventional", "type": "test", "scope": "forms", "breaking": false}
{"repo": "curl", "sha": "89ca8fe80ad4ca6434724a4ed287850deabe9766", "date": "2026-09-09T17:04:38+02:00", "subject": "cw-out: avoid O(n^2) work while buffering paused output", "body": "Improves paused-output buffer performance without changing protocol\nbehavior.\n\nThe buffer chain now keeps a tail pointer, so appending does not\nrepeatedly walk the list, and tracks the total buffered length instead\nof rescanning every buffer on each append. Flushing proceeds from the\noldest buffer iteratively.\n\nThis reduces buffering many paused writes from quadratic list/replay\nbookkeeping to linear work.\n\nCloses #22896", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "curl", "sha": "3341da7dc27256373de366b163e6b70a6606ba38", "date": "2026-09-09T17:04:42+02:00", "subject": "happy eyeballs: bound work from many duplicate/failing addresses", "body": "A DNS response with many duplicate addresses that fail to connect\nsynchronously could stall a transfer: failed attempts were never\npruned from the running list (cf_ip_ballers_prune only removes\nstill-ongoing ones), so every new address triggered a full rescan and\nan O(n) tail-walk to append, and the connection timeout was only\nchecked after the loop, not during it.\n\nFree hard-failed attempts immediately instead of accumulating them,\ntrack a tail pointer for O(1) appends, and check the timeout on each\niteration. Also cap and deduplicate the addresses converted from a\nc-ares response, closing off the unbounded input side.\n\nCloses #22897", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "curl", "sha": "a15c4a7f8c1bd0586012b532f51ac5aee00bfe0b", "date": "2026-09-09T17:04:45+02:00", "subject": "headers: avoid O(n^2) rescans in curl_easy_nextheader()", "body": "Every call rescanned the entire header list from the head and did a full\nname comparison against each entry just to recompute amount/ index for\nthe one header being returned, making a sequential enumeration of N\nheaders O(n^2).\n\nCache the previous pick's name/origin/request/amount/index; when the\nnext call continues from exactly that node with the same filters, reuse\nthe cached amount and increment the index instead of rescanning. Any\nother pattern (restart, filter change, non-sequential prev) falls back\nto the original full scan, so behavior is unchanged.\n\nCloses #22898", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "angular", "sha": "25228d0cc3c730266e2f7ea981f1c0f23dd49384", "date": "2026-09-09T17:22:08+02:00", "subject": "test: fix failing boundary test", "body": "Fixes a test that's breaking CI.", "style": "conventional", "type": "test", "scope": "", "breaking": false}
{"repo": "postgres", "sha": "539b7e8290e71a253ac4b70faed75e38a7125d97", "date": "2026-09-09T11:23:25-04:00", "subject": "Fix concurrency issues with DROP TABLESPACE", "body": "DROP TABLESPACE checked pg_shdepend for dependent objects without\nfirst locking the tablespace.  A concurrent command that recorded a\nshared dependency on the tablespace right after that check could still\ncommit, leaving an object whose pg_shdepend entry (or, for a relation,\npg_class.reltablespace) pointed to a tablespace that no longer existed.\n\nClose the race by having DropTableSpace() take an AccessExclusiveLock\non the tablespace before calling checkSharedDependencies().  That\nconflicts with the AccessShareLock shdepLockAndCheckObject() takes when\nrecording a new dependency, so the loser of the race blocks and\nrechecks once the winner commits.\n\nThat AccessExclusiveLock creates a new deadlock: ALTER TABLESPACE\nRENAME/SET and the internal ACL/owner updates in DROP OWNED and\nREASSIGN OWNED touched the catalog tuple without locking the\ntablespace, risking a lock-order cycle with DROP.  Fix by taking an\nAccessShareLock first in all four paths, rechecking pg_shdepend after\nany wait in the DROP OWNED and REASSIGN OWNED cases.  GRANT, REVOKE,\nand ALTER TABLESPACE ... OWNER TO already lock the object.\n\nAdd isolation tests covering both orderings of the original race\n(dependency-first and drop-first) and all four previously-unlocked\nupdate paths.\n\nAuthor: Ayush Tiwari \nReviewed-by: Andrew Dunstan \nDiscussion: https://postgr.es/m/CAJTYsWXjAQFnGKzXsht3XK8KHhyhontwFJw4JAHsUSfs_ptR4g@mail.gmail.com\nBackpatch-through: 14", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "angular", "sha": "91c2d6341413cc1bb2535abf0f8d7b8ec7dd7309", "date": "2026-09-09T17:48:11+02:00", "subject": "build: switch integration tests away from protractor", "body": "Reworks the integration tests to use Selenium Webdriver instead of Protractor.", "style": "conventional", "type": "build", "scope": "", "breaking": false}
{"repo": "angular", "sha": "679c50dcbe8cf7e5d0009e2b5fbf7289f305856b", "date": "2026-09-09T08:52:02-07:00", "subject": "refactor(router): Remove ActivatedRouteSnapshot from resourceContext", "body": "providing the snapshot is a bit incompatible with how things are meant to work here.\nThe snapshot is only accurate during the setup and would be out of date\non followup navigations", "style": "conventional", "type": "refactor", "scope": "router", "breaking": false}
{"repo": "neovim", "sha": "d24cafc06a78f6357674c2429706303538be48da", "date": "2026-09-09T12:10:01-04:00", "subject": "fix(multicursor): failed motion at primary is still cascaded #41812", "body": "Problem:\nIn follow-mode, a motion that fails at the primary cursor (e.g. \"j\" at\nEOB) still cascades to other cursors, so they move while the primary\ndoes not, which causes them to be deduplicated...\n\nAlso affects Visual sessions: \"Vjd\" at EOB deletes 1 line at the primary\nbut (potentially) 2 lines at every multicursor.\n\nSolution:\nIf a motion \"beeps\", don't replay it.\n\n\"3w\" near the end still replays (it moved), as does \"0\" at column\n0 (does not beep). This feels like a usable and intuitive compromise.", "style": "conventional", "type": "fix", "scope": "multicursor", "breaking": false}
{"repo": "openssl", "sha": "4ea4a7136112c135229e6a73fc301f54c617313b", "date": "2026-09-09T18:10:06+02:00", "subject": "Regression test for PKCS12_parse", "body": "Reviewed-by: Mounir Idrassi \nReviewed-by: Simo Sorce \nReviewed-by: Todd Short \nMerge-date: Tue Sep 15 15:23:48 2026\nMerged-from: https://github.com/openssl/openssl/pull/32773", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "neovim", "sha": "a55f134ee305ebf3cb75149025c5e75f5e2090e0", "date": "2026-09-09T12:31:26-05:00", "subject": "fix(lsp): stale client when rpc.connect fails #41792", "body": "Problem:\nWhen vim.lsp.rpc.connect is used for an LSP and the connection fails\n(due to ECONNREFUSED, for example), the LSP client is left\nuninitialized. Subsequent calls to vim.lsp.start will reuse the\nuninitialized client, but no new connection will be attempted.\n\nSolution:\nWhen vim.lsp.rpc.connect encounters a connection error, terminate the\nclient. This allows the user to do something in response to the failed\nconnection in the on_exit callback, and also makes it so subsequent\nattempts to start the LSP will try to establish a new connection rather\nthan using the previous uninitialized client.", "style": "conventional", "type": "fix", "scope": "lsp", "breaking": false}
{"repo": "neovim", "sha": "d663d8f4b3116a9cad4ade936218ccb4ed723c30", "date": "2026-09-09T18:43:20+01:00", "subject": "perf(help): generate helptags with LPeg", "body": "Problem: Generating helptags builds and queries a full vimdoc syntax\ntree. This is much slower than the previous scanner and also happens\nbefore discovering that the tags file cannot be written.\n\nSolution: Extract tags with an LPeg grammar, reading each file once and\nskipping example blocks and ordinary text. Open the output first so\nunwritable directories can return immediately.\n\nThis restores legacy handling of tags inside inline code spans. On an\narm64 Mac, the 5.36 MB runtime corpus takes 24 ms instead of 604 ms,\nwith identical generated tags.\n\nAI-assisted", "style": "conventional", "type": "perf", "scope": "help", "breaking": false}
{"repo": "eslint", "sha": "ac74e37322ebf122ada676f153dc55c81f3caab0", "date": "2026-09-09T14:08:38-04:00", "subject": "chore: Add AGENTS.md with AI disclosure requirements", "body": "* chore: Add AGENTS.md with AI disclosure requirements\n\nAdds an AGENTS.md file documenting commands, architecture, rule\nconventions, and code conventions for AI agents working in this repo.\nIncludes an AI disclosure requirement section aligned with the ESLint\nAI Usage Policy, requiring AI-authored issues, pull requests, and\ncomments to state the model that produced them.\n\nCLAUDE.md points at AGENTS.md so Claude Code picks up the same\ninstructions.\n\n* chore: Fold copilot-instructions.md into AGENTS.md\n\nMoves the project knowledge from .github/copilot-instructions.md into\nAGENTS.md so there is a single set of instructions for AI agents, and\nremoves the Copilot-specific file.\n\nContent brought over that AGENTS.md did not already cover: the\ntop-level directory map, test file layout and what requires tests,\nthe rule documentation outline including the \"When Not To Use It\"\nsection, and the fixer function guidance for fixable rules.\n\n* Update AGENTS.md\n\nCo-authored-by: Francesco Trotta \n\n* Update AGENTS.md\n\nCo-authored-by: Francesco Trotta \n\n* Update AGENTS.md\n\nCo-authored-by: Francesco Trotta \n\n---------", "style": "conventional", "type": "chore", "scope": "", "breaking": false}
{"repo": "openssl", "sha": "e6c03cbcc9aacce38e283541016a9179bf54e965", "date": "2026-09-09T15:00:21-04:00", "subject": "use BUF_MEM_free instead of OPENSSL_free to free BUF_MEM in str_copy", "body": "While reviewing a PR, I noticed that str_copy in conf_def.c allocates a\nBUF_MEM with BUF_MEM_free, but frees it with OPENSSL_free.  While this\nis fine here since we take ownership of the buf mem data pointer, it was\na bit hard to parse.\n\nChange it to follow the usual pattern of orphaning the BUF_MEM data by\nsetting buf->data to NULL prior to using the standard BUF_MEM_free call\n\nReviewed-by: Saša Nedvědický \nReviewed-by: Todd Short \nReviewed-by: Andrew Dinh \nMergeDate: Fri Sep 11 20:55:00 2026\n(Merged from https://github.com/openssl/openssl/pull/32774)", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "neovim", "sha": "ac167f943a905669b29f337f18d22a1efefeca92", "date": "2026-09-09T15:14:26-04:00", "subject": "fix(terminal): check file timestamps on terminal-job exit #41814", "body": "Problem:\nWhen a `:!` command runs, file timestamps are checked and W12 or\n`FileChangedShell` is triggered. But not when a `:terminal` job exits.\n\nOnly `enter_buffer()` checks, so whether you get the W12 warning depends\non how you happen to return to the file: `:term` works, `:tabnew | term`\ndoes not.\n\n    echo foo > testfile\n    nvim --clean testfile\n    \" Edit the buffer, don't save it.\n    :!mv testfile testfile2\n    :tabnew | term sh\n    touch testfile2 && mv testfile2 testfile\n    exit\n\n'autoread' is defeated because `:!mv` deletes the file, so\n`should_watch()` fails and the watcher stops. `ensure_watcher()` only\nre-runs on `BufReadPost`/`BufWritePost`/`OptionSet`, so nothing watches\nthe file when the terminal moves it back.\n\nSolution:\nSet `need_check_timestamps` on terminal-job exit, same as `:!cmd`\n(`os_shell`).", "style": "conventional", "type": "fix", "scope": "terminal", "breaking": false}
{"repo": "neovim", "sha": "08dd3adffb60ae6c031f9b8310febd65bb14604e", "date": "2026-09-10T03:42:20+08:00", "subject": "feat(api): add ev.win to autocmd callback args #39826", "body": "Problem: Autocmd callbacks currently have to rely on curwin, which may\nalready have changed by the time they run. This is especially noticeable\nfor WinNew, WinClosed, WinResized and WinScrolled.\n\nSolution: Track the window in the autocmd context and expose it as ev.win.\nPass it explicitly for the window events that need it.", "style": "conventional", "type": "feat", "scope": "api", "breaking": false}
{"repo": "angular", "sha": "b6517a6672708c6d4320a0c57c4f2ea8b6583793", "date": "2026-09-09T21:44:12+02:00", "subject": "build: resolve pnpm 12 breakages", "body": "Accounts for breaking changes in pnpm 12.", "style": "conventional", "type": "build", "scope": "", "breaking": false}
{"repo": "postgres", "sha": "94cc8c443a3115531be7abe414dad5034ee3bb57", "date": "2026-09-09T13:45:46-07:00", "subject": "Unify REPACK (CONCURRENTLY) error messages.", "body": "Commit 86f7c82cf102 reported the user catalog table restriction as a\ndetail, but the other checks in check_concurrent_repack_requirements()\nwere still a mix: some were hints, and the details repeated\n\"REPACK (CONCURRENTLY)\", which the primary message already names.\n\nReport them as details with the same \"This operation ...\" wording. The\none actionable hint, ALTER TABLE ... REPLICA IDENTITY USING INDEX,\nstays.\n\nBackpatch to v19, where REPACK (CONCURRENTLY) was introduced.\n\nAuthor: Bharath Rupireddy \nSuggested-by: Alvaro Herrera \nDiscussion: https://postgr.es/m/apBIFWzHYOaG0auN%40nathan\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "openssl", "sha": "774015ad0bd6bdd665253d0cb5e51427edef7aea", "date": "2026-09-09T22:50:53+02:00", "subject": "Fix memory leak on error on OSSL_PROVIDER_available()", "body": "`ossl_provider_free` must be called, but the `CRYPTO_THREAD_read_lock`\nerror path does not do this. By inverting the condition and\nrestructuring the code we can solve this cleanly without introducing a\ngoto.\n\nReviewed-by: Frederik Wedel-Heinen \nReviewed-by: Todd Short \nMerge-date: Wed Sep 16 13:41:24 2026\nMerged-from: https://github.com/openssl/openssl/pull/32775", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "openssl", "sha": "1e85928237e28369fdcfd44344028a74ae8c0215", "date": "2026-09-09T22:50:53+02:00", "subject": "Fix memory leak on error on OSSL_PROVIDER_available()", "body": "`ossl_provider_free` must be called, but the `CRYPTO_THREAD_read_lock`\nerror path does not do this. By inverting the condition and\nrestructuring the code we can solve this cleanly without introducing a\ngoto.\n\nReviewed-by: Frederik Wedel-Heinen \nReviewed-by: Todd Short \nMerge-date: Wed Sep 16 13:41:14 2026\nMerged-from: https://github.com/openssl/openssl/pull/32775", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "openssl", "sha": "4652075a19186130541310b0329a504a53e6eea9", "date": "2026-09-09T22:50:53+02:00", "subject": "Fix memory leak on error on OSSL_PROVIDER_available()", "body": "`ossl_provider_free` must be called, but the `CRYPTO_THREAD_read_lock`\nerror path does not do this. By inverting the condition and\nrestructuring the code we can solve this cleanly without introducing a\ngoto.\n\nReviewed-by: Frederik Wedel-Heinen \nReviewed-by: Todd Short \nMerge-date: Wed Sep 16 13:41:04 2026\nMerged-from: https://github.com/openssl/openssl/pull/32775", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "openssl", "sha": "d9cf714cf3863660da1a9f3a6362179ae3424ef6", "date": "2026-09-09T22:50:53+02:00", "subject": "Fix memory leak on error on OSSL_PROVIDER_available()", "body": "`ossl_provider_free` must be called, but the `CRYPTO_THREAD_read_lock`\nerror path does not do this. By inverting the condition and\nrestructuring the code we can solve this cleanly without introducing a\ngoto.\n\nReviewed-by: Frederik Wedel-Heinen \nReviewed-by: Todd Short \nMerge-date: Wed Sep 16 13:40:54 2026\nMerged-from: https://github.com/openssl/openssl/pull/32775", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "openssl", "sha": "515bc3d37c3b374d940b07a5ce9bbcef3ec9764a", "date": "2026-09-09T22:50:53+02:00", "subject": "Fix memory leak on error on OSSL_PROVIDER_available()", "body": "`ossl_provider_free` must be called, but the `CRYPTO_THREAD_read_lock`\nerror path does not do this. By inverting the condition and\nrestructuring the code we can solve this cleanly without introducing a\ngoto.\n\nReviewed-by: Frederik Wedel-Heinen \nReviewed-by: Todd Short \nMerge-date: Wed Sep 16 13:40:44 2026\nMerged-from: https://github.com/openssl/openssl/pull/32775", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "openssl", "sha": "c3cb4f142bacab224c3a72489fc27ca608aa1ef7", "date": "2026-09-09T22:50:53+02:00", "subject": "Fix memory leak on error on OSSL_PROVIDER_available()", "body": "`ossl_provider_free` must be called, but the `CRYPTO_THREAD_read_lock`\nerror path does not do this. By inverting the condition and\nrestructuring the code we can solve this cleanly without introducing a\ngoto.\n\nReviewed-by: Frederik Wedel-Heinen \nReviewed-by: Todd Short \nMerge-date: Wed Sep 16 13:40:31 2026\nMerged-from: https://github.com/openssl/openssl/pull/32775", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "postgres", "sha": "50ba390a4f81536aac13c1b8cf75c646a1156527", "date": "2026-09-09T16:02:24-05:00", "subject": "Fix hangs in COPY FROM (FORMAT text).", "body": "The SIMD path for this command reads ahead via CopyLoadInputBuf()\nwhenever fewer than sizeof(Vector8) bytes remain in the input\nbuffer, even if those bytes hold a complete end-of-copy marker.\nIf the input is a pipe whose writer has sent the marker but not\nclosed the pipe, that read blocks, and COPY waits for data it will\nnever use.  The scalar loop asks only for the bytes it needs, so\nit stops at the marker without reading any further.\n\nTo fix, teach CopyLoadInputBuf() to decline a speculative load\nwhen the caller's remaining bytes contain a backslash, which in\ntext mode might begin such a marker.  (CSV mode doesn't treat\n\\. as special, so it is unaffected.)  The SIMD path then hands\nthose bytes to the scalar loop, as it already does when a load\ncomes up short.  Checking for the marker in the SIMD helper itself\nwould be more direct, but a call there costs the compiler\nregisters on every line, which measurably slowed COPY of short\nlines in testing.\n\nOversight in commit e0a3a3fd53.\n\nAuthor: Nazir Bilal Yavuz \nReviewed-by: Greg Burd \nDiscussion: https://postgr.es/m/CAN55FZ1qFb4Yo3-MWeQfGQnu1_Ksx4xcFC2m3hyw8a--%2BeHQYQ%40mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "postgres", "sha": "d2b65863f04f99eab15eca01a54708d2f8ad6099", "date": "2026-09-09T16:30:04-05:00", "subject": "Add pg_nodiscard decorations to some Bitmapset functions.", "body": "There's no conceivable use-case for ignoring the result of the\nfunctions that return a pointer to a Bitmapset.  Add pg_nodiscard\ndecorations to those like commit a378ba49a5 did for some functions\nin pg_list.h.\n\nAuthor: Jan Nidzwetzki \nDiscussion: https://postgr.es/m/2020a70e-76f3-4661-afa6-82c2639251c2%40planetscale.com", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "babel", "sha": "acc19f92476347c1888381f4c9f0277b8e5a5745", "date": "2026-09-10T05:37:01+08:00", "subject": "perf: Improve `@babel/parser`", "body": "", "style": "conventional", "type": "perf", "scope": "", "breaking": false}
{"repo": "babel", "sha": "97caaba50bc8e56a6a01d781989d577b57532123", "date": "2026-09-09T23:37:47+02:00", "subject": "[ts] Allow merging `enum` into a `namespace`", "body": "", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "babel", "sha": "b7eef30765eaa5bdca36ba4f851c871316171af9", "date": "2026-09-09T17:40:23-04:00", "subject": "fix(ts): improve variable handling in namespace", "body": "", "style": "conventional", "type": "fix", "scope": "ts", "breaking": false}
{"repo": "postgres", "sha": "1129570f654e96d45bbc63397079171af29710b9", "date": "2026-09-09T17:48:54-04:00", "subject": "pg_plan_advice: Add CHECK_FOR_INTERRUPTS() and check_stack_depth()", "body": "Without CHECK_FOR_INTERRUPTS(), backends might not respond to query\ncancellations in a timely fashion, especially for JOIN_ORDER()\nspecifications, where matching against each join can take O(n^2) time\nin the length of the join order list in degenerate cases.\n\nRecursive functions should check_stack_depth().\n\nReported-by: Noah Misch \nDiscussion: https://postgr.es/m/CA+TgmoYmXy-jiP5qDhqNEiYFEBzQsArO6O2d9E8szNZqi1bePQ@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "angular", "sha": "96dbae53bbf818ad3ec8cb1796f614b5db3376b1", "date": "2026-09-09T14:53:38-07:00", "subject": "refactor(core): expose `consequentialHint` annotation", "body": "This allows `consequentialHint` to be passed to `ModelContext.registerTool`. From Angular's perspective, this is simply a pass-through option.\n\nSee: https://groups.google.com/a/chromium.org/g/chrome-ai-dev-preview/c/uHu5kfclbKw/m/dgorjfHPDgAJ", "style": "conventional", "type": "refactor", "scope": "core", "breaking": false}
{"repo": "angular", "sha": "8ca879e2d86c077e0eaeee043fc337fbc0040345", "date": "2026-09-09T14:53:44-07:00", "subject": "refactor(forms): expose `consequentialHint` annotation", "body": "Unlike `readOnlyHint` and `untrustedContentHint`, Angular cannot infer any reasonable default value for `consequentialHint` and simply exposes it to users to specify as appropriate.\n\nSee: https://groups.google.com/a/chromium.org/g/chrome-ai-dev-preview/c/uHu5kfclbKw/m/dgorjfHPDgAJ", "style": "conventional", "type": "refactor", "scope": "forms", "breaking": false}
{"repo": "angular", "sha": "0ee0a16c4ae92458757f79afbbfc1e3aa000eae6", "date": "2026-09-09T21:56:22Z", "subject": "fix(compiler): stop suppressing non-iterable values in `@for` expression", "body": "When generating type check blocks for `@for` loops, a non-null assertion is appended to the collection expression. If the collection expression is a compound expression (e.g. binary or logical operations like a && b), the lack of outer parentheses caused the ! to bind only to the rightmost operand (a && b!), leading to typecheck errors such as TS2532.\n\nThis wraps the expression via .wrapForTypeChecker() before appending !.\n\nTemplate type checking for compound expressions in `@for` loops may surface new type errors\n\nTemplate type checking for `@for` loops now more accurately checks the entire collection expression when compound expressions (such as `(x | async) || []` or logical/ternary operations) are used.\n\nPreviously, certain invalid expressions in `@for` loops could slip through template type checking unnoticed. As a result, you may see new TypeScript diagnostics (such as `TS2488: Type '{}' must have a '[Symbol.iterator]()' method that returns an iterator`) if the expression does not evaluate to an iterable type. For example, when a component property `x` is typed as `any`, `(x | async)` produces `unknown | null`, which causes `(x | async) || []` to type check as `{}` rather than an iterable array.\n\nTo fix these errors, ensure that values referenced in the template are strongly typed on the component class (for example, declaring `x: Observable` instead of `any`), or use `$any(...)` in the template (for example, `$any(x | async) || []`).", "style": "conventional", "type": "fix", "scope": "compiler", "breaking": false}
{"repo": "angular", "sha": "5e632b639f736de876823a596dd9e3a094b03e1b", "date": "2026-09-09T21:56:22Z", "subject": "fix(compiler): wrap `@for` collection expression before appending non-null assertion", "body": "When generating type check blocks for `@for` loops, a non-null assertion is appended to the collection expression. If the collection expression is a compound expression (e.g. binary or logical operations like a && b), the lack of outer parentheses caused the ! to bind only to the rightmost operand (a && b!), leading to typecheck errors such as TS2532.\n\nThis wraps the expression via .wrapForTypeChecker() before appending !.\n\n(cherry picked from commit 14dbbf9b68474a442a295ae165543c8923eae4b8)", "style": "conventional", "type": "fix", "scope": "compiler", "breaking": false}
{"repo": "angular", "sha": "14dbbf9b68474a442a295ae165543c8923eae4b8", "date": "2026-09-09T21:56:22Z", "subject": "fix(compiler): wrap `@for` collection expression before appending non-null assertion", "body": "When generating type check blocks for `@for` loops, a non-null assertion is appended to the collection expression. If the collection expression is a compound expression (e.g. binary or logical operations like a && b), the lack of outer parentheses caused the ! to bind only to the rightmost operand (a && b!), leading to typecheck errors such as TS2532.\n\nThis wraps the expression via .wrapForTypeChecker() before appending !.", "style": "conventional", "type": "fix", "scope": "compiler", "breaking": false}
{"repo": "babel", "sha": "6768aad8c742e12e1f3f4ef5f5babeb04b303959", "date": "2026-09-10T00:10:58+02:00", "subject": "fix: only collapse the empty-object check when it targets the declared binding", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false}
{"repo": "babel", "sha": "ed7fd27c3af01e36bae5aa668b727276be5ff42f", "date": "2026-09-09T18:30:17-04:00", "subject": "fix(parser): remove non-decimal prefix from bigint", "body": "* test(parser): serialize bigInt in snapshot\n\nThis change ensures that when the AST contains BigInt/RegExp, the test runner can still write to output.extended.json. Previously it will throw a TypeError because the builtin JSON.stringify does not know how to serialize bigint\n\n* fix(parser): remove non-decimal prefix from bigint\n\n* test: add test for invalid bigint literal\n\n* fix: support undefined in test serialize\n\n* cleanup: inline toBigInt\n\nFor Babel 8, the minimal supported Node.js version already supports BigInt.", "style": "conventional", "type": "fix", "scope": "parser", "breaking": false}
{"repo": "ant-design", "sha": "bdbbdd5179e440c2269c27355f1187bf2bf1b7f9", "date": "2026-09-10T06:46:54+08:00", "subject": "docs: update third-party library links in recommendation", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false}
{"repo": "angular", "sha": "66de56beb96a93699045c655fe8ef98d98f212df", "date": "2026-09-10T01:34:36+02:00", "subject": "refactor(compiler-cli): relax nullish coalescing non nullable diagnostics on indexed access", "body": "When noUncheckedIndexedAccess is not enabled, indexed accesses do not include undefined in the type. This relaxes the check for nullish coalescing similarly to optional chaining. Fixes #70655\n\nfixes #70655\n\n(cherry picked from commit 9090a758be395df1ba414f3cc10b938105578f63)", "style": "conventional", "type": "refactor", "scope": "compiler-cli", "breaking": false}
{"repo": "angular", "sha": "9090a758be395df1ba414f3cc10b938105578f63", "date": "2026-09-10T01:34:36+02:00", "subject": "refactor(compiler-cli): relax nullish coalescing non nullable diagnostics on indexed access", "body": "When noUncheckedIndexedAccess is not enabled, indexed accesses do not include undefined in the type. This relaxes the check for nullish coalescing similarly to optional chaining. Fixes #70655\n\nfixes #70655", "style": "conventional", "type": "refactor", "scope": "compiler-cli", "breaking": false}
{"repo": "postgres", "sha": "cb298616463d6e4f2c360e8d07b83b526fad375f", "date": "2026-09-10T09:39:18+09:00", "subject": "Widen Object ID counter to 8 bytes", "body": "This commit adds a new routine called GetNewObjectId8() in varsup.c,\nwhich is able to retrieve a 8-byte OID, based on the next OID retrieved\nfrom the control file.  GetNewObjectId() is kept compatible with its\norigin, where we still check that the lower 32 bits of the counter do\nnot wraparound, while handling the FirstNormalObjectId case.\n\nThis change is combined with an update of the control file, with\nCheckPoint.nextOid enlarged from Oid to Oid8.  pg_control_checkpoint(),\npg_resetwal and pg_upgrade are updated to adapt to this change, to make\nsure that the value is not truncated due to 32-bit limitations when\nused, making an upgrade to a newer version where 8-byte is supported a\ntransparent operation.\n\nA couple of comments in the xlog redo logic referred to wraparound\nissues, which is not the case anymore with this change in place, so\nrefresh these.\n\nThis change is required to be able to support external TOAST pointers\nbased on Oid8, under discussion, and can be used for other purposes.\n\nTests are added for pg_upgrade (2^32 value carried across upgrades),\npg_resetwal and pg_control_checkpoint(), hence across all the tools\nthat need to be covered by this change.\n\nBump catalog version.\nBump PG_CONTROL_VERSION.\nBump XLOG_PAGE_MAGIC, due to XLOG_NEXTOID.\n\nAuthor: Michael Paquier \nReviewed-by: Greg Burd \nReviewed-by: Bharath Rupireddy \nReviewed-by: Yugo Nagata \nDiscussion: https://postgr.es/m/af19kUjwjhaoUTLn@paquier.xyz", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "postgres", "sha": "3430fba3b0f0655e203a4b03dd531d7bcad86ff3", "date": "2026-09-10T09:46:44+09:00", "subject": "postgres_fdw: Fix crash when estimating joins with functions", "body": "Commit 0ee83dd4a99 allowed postgres_fdw to push down inner joins between\nforeign tables and functions in FROM clauses to the remote server.\nHowever, planning such a join with use_remote_estimate enabled could\ncause a segmentation fault. For example, the following query crashed\nwhen ft had use_remote_estimate set to true:\n\n    EXPLAIN SELECT * FROM generate_series(1, 3) g JOIN ft ON ft.i = g;\n\nThe crash occurred because init_func_stub_fpinfo() copied the connection\ninformation and options from the foreign side to the function-side stub,\nbut omitted the user mapping, leaving it NULL. When the function was the\nouter side of the join, the join inherited this NULL mapping and passed it\nto GetConnection() when requesting a remote estimate, causing a\nsegmentation fault.\n\nFix this by having init_func_stub_fpinfo() also copy the user mapping\nfrom the foreign side.\n\nAuthor: Fujii Masao \nReviewed-by: Alexander Pyhalov \nDiscussion: https://postgr.es/m/CAHGQGwGE9xPzCke0C0r4omTkh-hU08aBd2j9oSA0FYmJLCf9jA@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "curl", "sha": "110936726e518ff843e0a2063db8143be07dd286", "date": "2026-09-09T20:54:00-04:00", "subject": "tool_easysrc: reset the handle between --next operations", "body": "The curl tool creates a fresh easy handle for every operation, but the\n--libcurl output reuses a single handle for all of them and never resets\nit, so options set for one operation silently apply to the following\nones and the generated program does not do what the command line did.\n\n19d0f4acfa9d fixed this for SSL verification by re-emitting those two\noptions after --next. Everything else still leaks. For\n\n  curl -H \"X-First: yes\" --compressed http://host/one \\\n       --next http://host/two\n\nthe generated program sends X-First and Accept-Encoding on the second\ntransfer as well, while the command line sends them only on the first.\nConfirmed by compiling the generated program and comparing what both\nactually put on the wire against a local server:\n\n                     operation 1              operation 2\n  curl               X-First, Accept-Encoding  neither\n  generated program  X-First, Accept-Encoding  X-First, Accept-Encoding\n\nEmit curl_easy_reset() between operations instead, which covers the\ngeneral case: every option the next operation needs is already re-emitted\nafter the reset, because the tool writes out its options per operation.\n\nThe call is deferred rather than written straight after\ncurl_easy_perform() so that it only appears when another operation\nactually follows: a single-operation command generates no reset at all,\nand none is emitted before the trailing curl_easy_cleanup().\n\nExisting --libcurl coverage (tests 1400-1407, 1420, 1465, 1481) is all\nsingle-operation, and 1679 checks only the SSL options, so add test 1687\nfor the general case. It fails without this change.\n\nTest 1679's expected output gains the same curl_easy_reset() line, since\nit is a --next test too.\n\nRef: #22701\nCloses #22890", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "postgres", "sha": "be2602cfcc1578480ccefca66eda93975d2c10de", "date": "2026-09-10T10:03:13+09:00", "subject": "Fix more duplicate qual clauses in parameterized paths", "body": "Commit f836b688f ensured that at most one clone of a join clause is\nchosen for a parameterized path, so long as the clone variants are\nparse-tree distinct.  Two ways remained to end up enforcing the same\ncondition more than once, which wastes execution effort and applies\nthe clause's selectivity multiple times, underestimating the row\ncount.\n\nIf a commuting outer join nulls no Var referenced by a particular\nqual, the clone variants of that qual are parse-tree identical, and\nthe incompatible_relids test cannot tell them apart.  To fix, enforce\njust the first surviving variant, identified by rinfo_serial.\n\nThe other case is in get_joinrel_parampathinfo, where the\nclause-recovery pass for dropped EquivalenceClasses queries the EC\nmachinery with a context that overlaps the preceding loop's, so the\nderived-clause cache can hand back a clause that that loop already\naccepted.  To fix, skip any clause that is already present in the\ncollected list.\n\nAlso add assertions verifying that the clauses to be enforced at a\njoin or within a parameterized path contain no duplicate rinfo_serial,\nto catch any remaining or future violations.  These assertions are\nadded in master only, since hard prevention of duplicate clauses is a\nnew goal and other unfixed cases may remain.\n\nBack-patch to v19, as with commit f836b688f.\n\nAuthor: Richard Guo \nAuthor: Tom Lane \nReviewed-by: Tom Lane \nDiscussion: https://postgr.es/m/CAMbWs4_Ezb1Lnj7BqcTQSysC6B5UFZxfhib8xr6rqpu5uqr8dA@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "postgres", "sha": "7c620d2bf87bdb9508dd8cb40377f84388244dca", "date": "2026-09-10T10:03:53+09:00", "subject": "Fix duplicate enforcement of EC-derived conditions", "body": "A clause connecting the same two EC members can be requested from the\nEC machinery in different contexts.  When building parameterized index\npaths for a relation, generate_implied_equalities_for_column derives\nan indexable clause with parent_ec set, to mark it as redundant with\nother join clauses derived from the same EC.  Meanwhile, at a join\nwhere the relations of a multi-relation EC member first come together,\ngenerate_join_implied_equalities emits a clause equating the same two\nmembers with parent_ec unset, since that clause is the sole enforcer\nof its member there and must not be dropped as redundant.  Because\ncreate_join_clause treats parent_ec as part of its cache-lookup key,\nsuch requests can yield two different RestrictInfos for the same\ncondition, carrying different rinfo_serials.\n\nThis breaks the assumption that one condition has exactly one serial\nnumber, which create_nestloop_path relies on to drop join clauses that\nare already enforced within a parameterized inner path.  As a result,\nthe same condition could be enforced twice at different plan levels,\nwasting execution effort and applying the clause's selectivity twice,\nunderestimating the join's row count.\n\nTo fix, make create_join_clause copy the rinfo_serial from an existing\nclause that connects the same two members with the opposite parent_ec\nmarking, so that the two clauses keep their distinct markings but are\nrecognized as the same condition.\n\nAs a result, a duplicate join filter disappears from one of the\nexisting regression plans (the \"cardinal_number\" query in join.sql);\nthe other plan changes are join-order differences between plans of\nessentially equal cost.\n\nAuthor: Richard Guo \nReviewed-by: Tom Lane \nDiscussion: https://postgr.es/m/CAMbWs4-3yZeBMGQDpmYRPdV-sL_=6H2AN=MoeneOT+VUk28PPg@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "postgres", "sha": "7ac0c2d058adb0aa47b7deb444c1949b84b7d781", "date": "2026-09-10T11:02:21+09:00", "subject": "test_extensible: Improve handling of incorrect inputs", "body": "The SQL wrappers of the module relied on stringToNode(), that trusts\nits input.  When specifying node strings that to not match with the\nexpected ExtensibleNode \"TestExtNode\", the module would crash.\n\nThis adds a simple set of sanity checks to check the input, ensuring\nthat only an ExtensibleNode \"TestExtNode\" goes through.  Though far\nfrom critical, it's always nice to make these test modules not crash\nwhen used.\n\nOversight in 0e944fe3579.\n\nAuthor: Zsolt Parragi \nReviewed-by: Greg Burd \nDiscussion: https://postgr.es/m/CAN4CZFOTQXMZvuPxweO4Fm7+sjrPJUqU-JvYKETLBxuhqgbPjw@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "vite", "sha": "42a4ea7cd01dee615891c2e5675456d4b8a2550c", "date": "2026-09-10T10:19:08+08:00", "subject": "chore: avoid recompiling regex in mergeChangelog script", "body": "", "style": "conventional", "type": "chore", "scope": "", "breaking": false}
{"repo": "ant-design", "sha": "6e517e3dede347a01f033cbcb8c9247b8f10e42e", "date": "2026-09-10T10:24:26+08:00", "subject": "chore: upgrade rc Dropdown and Tabs dependencies", "body": "* chore: upgrade rc dropdown and tabs\n\n* chore: up size-limit", "style": "conventional", "type": "chore", "scope": "", "breaking": false}
{"repo": "postgres", "sha": "34c11958b6f12a90c0ec9fca3ce502a32280c30d", "date": "2026-09-10T11:25:24+09:00", "subject": "test_bitmapset: Improve handling of incorrect inputs", "body": "The SQL wrappers of the module relied on stringToNode(), that entirely\ntrusts its input.  This commit applies an extra round of validation, to\nprevent the test module from crashing on incorrect inputs, provided\nthrough the SQL functions.\n\nNo backpatch, as this is no critical issue.\n\nAuthor: Zsolt Parragi \nReviewed-by: Greg Burd \nDiscussion: https://postgr.es/m/CAN4CZFOTQXMZvuPxweO4Fm7+sjrPJUqU-JvYKETLBxuhqgbPjw@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "vite", "sha": "ef0dc17ada53d1169ae5a89cb8f6482831466755", "date": "2026-09-10T10:50:00+08:00", "subject": "fix: only treat whole `node_modules` path segments as dependencies (fix #17467)", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false}
{"repo": "postgres", "sha": "d1f742d5d12654173b022fd7723ab206a3e3a3de", "date": "2026-09-10T11:51:06+09:00", "subject": "Fix pg_stat_progress_cluster command reporting for unprivileged users", "body": "Previously, when another session ran REPACK, CLUSTER, or VACUUM FULL and\nthe user lacked permission to view its progress details,\npg_stat_progress_cluster incorrectly reported the command as CLUSTER\ninstead of NULL.\n\npg_stat_progress_cluster is based on pg_stat_progress_repack and\ntranslates REPACK into CLUSTER or VACUUM FULL depending on whether index\nordering is requested. However, this translation defaulted to CLUSTER,\nincorrectly converting NULL commands returned for insufficient privileges\ninto CLUSTER.\n\nFix this translation to preserve a NULL command when the user lacks\npermission to view the progress details.\n\nBump catalog version.\n\nAuthor: Zsolt Parragi \nReviewed-by: Fujii Masao \nReviewed-by: Chao Li \nDiscussion: https://postgr.es/m/CAN4CZFMzy2V_wvRCBW5K8=wVCk1-C7nq=8otXO=u+s_1KZHqyA@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "ant-design", "sha": "a6ec655d6b78bcf3a4cb384180e1c02e2f1ff80e", "date": "2026-09-10T09:52:04+07:00", "subject": "fix(Tabs): apply styles.popup.root to the more dropdown", "body": "* fix(Tabs): apply styles.popup.root to the more dropdown\n\n* test(Tabs): open the more dropdown via more.visible\n\n* test(Tabs): fold popup assertions into the main semantic case", "style": "conventional", "type": "fix", "scope": "Tabs", "breaking": false}
{"repo": "vite", "sha": "8abf700eeb2411d8402d08f8e2696effafdbe774", "date": "2026-09-10T10:52:45+08:00", "subject": "perf(proxy): pre-compile context matchers at server creation", "body": "", "style": "conventional", "type": "perf", "scope": "proxy", "breaking": false}
{"repo": "postgres", "sha": "e418e8eb92aa1dccca18b297a85c7ff38a7ee2a0", "date": "2026-09-10T11:54:08+09:00", "subject": "Fix REPACK command reporting in pg_stat_progress_cluster", "body": "pg_stat_progress_cluster translates REPACK into VACUUM FULL when\npg_stat_progress_repack.repack_index_relid is zero, and into CLUSTER\notherwise.\n\nPreviously, repack_index_relid was set only when an index scan was used.\nWhen REPACK used a sequential scan and sort to order the table by an index,\nthe field remained zero even though index ordering had been requested. As a\nresult, pg_stat_progress_cluster incorrectly reported the command as\nVACUUM FULL instead of CLUSTER.\n\nFix this by setting repack_index_relid to the OID of the requested ordering\nindex when progress reporting starts, regardless of the scan method. This\nallows pg_stat_progress_cluster to correctly report REPACK as CLUSTER when\nindex ordering is requested.\n\nAlso update the documentation to clarify the meaning of the index columns\nand how REPACK is translated into CLUSTER or VACUUM FULL.\n\nSuggested-by: Fujii Masao \nAuthor: Chao Li \nReviewed-by: Fujii Masao \nReviewed-by: Zsolt Parragi \nDiscussion: https://postgr.es/m/CAN4CZFMzy2V_wvRCBW5K8=wVCk1-C7nq=8otXO=u+s_1KZHqyA@mail.gmail.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "vite", "sha": "c8f7b280a2669733b84231c128c4926822b131d7", "date": "2026-09-09T22:58:27-04:00", "subject": "docs(ssr): explain server-only module reloads (fix #19114)", "body": "", "style": "conventional", "type": "docs", "scope": "ssr", "breaking": false}
{"repo": "vite", "sha": "9913672bee9c34a2df7fff4c2538783cd4f43b4e", "date": "2026-09-10T00:01:44-03:00", "subject": "fix: handle CRLF line endings in code frame positions", "body": "", "style": "conventional", "type": "fix", "scope": "", "breaking": false}
{"repo": "neovim", "sha": "4397aad1d3055928a4ac51a55464b82edba629d8", "date": "2026-09-10T10:30:07+07:00", "subject": "fix(mbyte): spacing marks need their own display cell", "body": "Problem:\nstrdisplaywidth('ทำ') is 1, but that cluster needs two cells. A SpacingMark has\npositive advance width (Unicode core spec D55), so it cannot share its base's\ncell: no Thai font fits a consonant plus SARA AM (U+0E33) in one cell, the\nnarrowest needing 119% of one and most 180-210%, measured against a 0.600 em\ncell. UAX#29 GB9a keeps the mark in its base's cluster, which is correct for\nsegmentation and says nothing about width.\n\nsetcellwidths() cannot work around it, because the collapse is per cluster, not\nper codepoint.\n\nVim reports 2 here. Devanagari and Bengali matras and Khmer and Tamil vowel\nsigns are the same case.\n\nSolution:\nAdd a cell for each remaining codepoint in the cluster whose boundclass is\nUTF8PROC_BOUNDCLASS_SPACINGMARK, clamped to the maximum cells per cluster.\nSegmentation is untouched, so the cursor and selection still treat the cluster\nas one unit.\n\nU+FF9E and U+FF9F, the halfwidth katakana sound marks, are Grapheme_Extend\nrather than SpacingMark, so they are matched separately. This is not a general\nsolution for grapheme extenders that carry width -- that needs a queryable\ntable -- but it does fix the reported regressions for Thai and Japanese\n(refs #34733).\n\nAI-assisted\n\nfixes #34733", "style": "conventional", "type": "fix", "scope": "mbyte", "breaking": false}
{"repo": "postgres", "sha": "b264391a846e8b5917b00dde2399c38843c75bb8", "date": "2026-09-09T20:39:53-07:00", "subject": "Fix pg_iswcased().", "body": "Commits 630706ced0 and 9c8de15969 lost some encoding-specific nuances:\nin libc with EUC encodings, or in ICU with any non-UTF8 encoding, we\ncannot easily classify characters, so we must assume they are\ncase-varying. Restore these behaviors.\n\nPreviously, ILIKE using an index scan could cause wrong results.\n\nReported-by: Noah Misch \nDiscussion: https://postgr.es/m/20260826221526.89.noahmisch@microsoft.com\nBackpatch-through: 19", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "postgres", "sha": "d8304bc661c29600f82ba62719457837edf47b08", "date": "2026-09-10T13:10:54+09:00", "subject": "pgbench: Improve handling of NULL in \\gset and \\aset", "body": "Previously, readCommandResponse() stored every column with PQgetvalue(),\nwhich returns an empty string for the SQL NULL value.  An empty string\nis not a valid value in the pgbench expression language, and when\nreturned by the server it was also indistinguishable from a NULL.\n\nWhen receiving a value, pgbench now checks for PQgetisnull() and sets a\nNULL value with setNullValue() before saving it.  This behavior is more\nconsistent with \\set, making the result saved testable with IS NULL.\n\nThis is a quality-of-life improvement, so no backpatch is done.\n\nAuthor: Shinya Kato \nReviewed-by: Kyotaro Horiguchi \nReviewed-by: Shihao Zhong \nDiscussion: https://postgr.es/m/CAOzEurT9J9QMmpJt_8pWZCziNYCUHOD=c-hoccOVYJLc_VxVRg@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "postgres", "sha": "c8985a09b7b52519f794c433e5f8f7311a77a2fc", "date": "2026-09-10T13:19:11+09:00", "subject": "doc: Mention pg_stat_force_next_flush()", "body": "pg_stat_force_next_flush() has primarily been used in in-core tests to\nforce cumulative statistics to be flushed at the end of a transaction\nwithout waiting for the PGSTAT_MIN_INTERVAL throttle, to speed up\nthings.\n\nIt is also useful in tests outside the core tree.  Spoiler: I have used\nit myself for some out-of-core work for the same reason as the in-core\ntests, and would have unlikely known about it without my work on the\ncore code.\n\nAuthor: Sami Imseih \nDiscussion: https://postgr.es/m/aqEONVuLhgBiqHW6@paquier.xyz", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "postgres", "sha": "3f660d86ba65d9747d5e2da418e24bd023271241", "date": "2026-09-10T13:31:09+09:00", "subject": "Take RI fast-path snapshot after locking referenced relation", "body": "ri_FastPathCheck() acquired its scan snapshot before opening the\nreferenced relation.  If it then waited for the relation lock in READ\nCOMMITTED mode, a referenced row committed during the wait would not be\nvisible to the old snapshot.  The check could consequently report a\nforeign key violation even though the referenced row existed.\n\nThe batched path acquires its scan snapshot after opening the relations,\nso it does not have the lock-wait problem.  The per-row path remains\nreachable during constraint validation and checks re-entered from a\nbatch flush.\n\nTake the snapshot after opening the referenced relation and reloading\nthe constraint information.  This also agrees with the SPI path, which\nacquires the referenced-relation lock before selecting the snapshot\nused for the check.\n\nAlso make the scan snapshot active for the duration of the check, so\nSTABLE cast and equality functions use the same snapshot as the index\nscan.  Otherwise, a STABLE cast can miss rows inserted by earlier AFTER\ntriggers and cause a valid foreign key to be rejected.  Add a\nregression test comparing this case with the SPI path.\n\nFinally, also advance the command counter after acquiring the\nreferenced-relation lock and before reloading the constraint\ninformation, so local invalidations are processed under the lock, as in\nthe SPI path.\n\nAdd an injection-point test that pauses a per-row check before locking\nthe referenced relation, commits the missing referenced row in another\nsession, and verifies that validation succeeds.  Force per-row\nvalidation by using a role without SELECT privilege on the referenced\ntable, so the test exercises this path even with batching enabled.\n\nAdd isolation-test coverage for the visibility of a referenced row\ncommitted after the referencing transaction has executed an earlier\ncommand.  A later command can see such a row in READ COMMITTED, but not\nin REPEATABLE READ or SERIALIZABLE.\n\nReviewed-by: Chao Li \nReviewed-by: Amit Kapila \nReviewed-by: Bertrand Drouvot \nReviewed-by: Michael Paquier \nDiscussion: https://postgr.es/m/aoU32VTkUtpdUewB@paquier.xyz", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "neovim", "sha": "5ebcb073b3b7434aaec492faadaaa0cf65457e2f", "date": "2026-09-10T00:55:59-04:00", "subject": "revert(scrollbind): revert #41519 due to regressions", "body": "Problem:\nResetting w_scbind_pos to the displayed position discards pending\nscroll distance. This breaks scrollbind with asymmetric folds and\nleaves persistent offsets after tall virtual-line blocks. Wrapped\nlines with smoothscroll also show increased follower lag.\n\nSolution:\nRevert the original change and its test. This restores the previous\nbehavior, including the original tall-virt_lines oscillation, while\na replacement is reviewed separately.\n\nThis reverts commit b781e95a24bd2d201f4afbe4729635114893cdf5.\n\nAI-assisted\n\n(cherry picked from commit 0c9012f295eedd7283bbf7f9a0fbef3781ba50f4)", "style": "conventional", "type": "revert", "scope": "scrollbind", "breaking": false}
{"repo": "neovim", "sha": "0c9012f295eedd7283bbf7f9a0fbef3781ba50f4", "date": "2026-09-10T00:55:59-04:00", "subject": "revert(scrollbind): revert #41519 due to regressions", "body": "Problem:\nResetting w_scbind_pos to the displayed position discards pending\nscroll distance. This breaks scrollbind with asymmetric folds and\nleaves persistent offsets after tall virtual-line blocks. Wrapped\nlines with smoothscroll also show increased follower lag.\n\nSolution:\nRevert the original change and its test. This restores the previous\nbehavior, including the original tall-virt_lines oscillation, while\na replacement is reviewed separately.\n\nThis reverts commit b781e95a24bd2d201f4afbe4729635114893cdf5.\n\nAI-assisted", "style": "conventional", "type": "revert", "scope": "scrollbind", "breaking": false}
{"repo": "vitest", "sha": "829b6d395efdb1545b6c640a2dffd1c8fabc96e5", "date": "2026-09-10T14:17:40+09:00", "subject": "test(runner): update fails lifecycle expectations", "body": "", "style": "conventional", "type": "test", "scope": "runner", "breaking": false}
{"repo": "postgres", "sha": "25649d6e791c2d803fff0ce8b69ed968028a5edc", "date": "2026-09-10T14:41:54+09:00", "subject": "Remove batching from RI fast-path checks", "body": "Commit b7b27eb41a5 added batching to the direct-index fast path for\nforeign key checks introduced by 2da86c1ef9b. Instead of probing the\nreferenced index once per row, it accumulated referencing rows and\nchecked them in groups, using SK_SEARCHARRAY for single-column foreign\nkeys.\n\nThe batching requires state to survive across trigger invocations and\nto be flushed at the end of each trigger-firing cycle. Follow-up work\nhas had to define how that state interacts with nested trigger firing,\nsubtransactions, deferred constraints, and SET CONSTRAINTS.\n\nFailure to handle one of those cases can leave a buffered check\nunperformed, allowing a transaction to commit a permanent foreign key\nviolation without reporting an error. With PostgreSQL 19 close to\nrelease, there is not enough time to gain confidence that all relevant\ntrigger and transaction states have been covered.\n\nRemove the batching and its after-trigger callback infrastructure,\nincluding the per-batch RI cache and associated subtransaction cleanup.\nRestore the early exit in AfterTriggerFireDeferred(), since batch\ncallbacks can no longer queue additional deferred triggers after\nafterTriggerInvokeEvents() returns.\n\nRemove tests that exercise only the batching implementation and its\ncallback and cache lifetime machinery. Retain tests that continue to\nexercise the underlying per-row fast path, including validation,\nscan-key construction, cross-type rechecks, deferred checks, and\nmetadata invalidation.\n\nKeep the underlying per-row fast path. It performs each check\nsynchronously and retains no pending checks across trigger invocations.\nAlso retain the fixes made to the per-row probe path, including support\nfor domain-typed referencing columns, restriction to btree referenced\nindexes, concurrent index replacement, metadata invalidation, and\nnullable referenced keys.\n\nThis removal applies only to REL_19_STABLE.\n\nDiscussion: https://postgr.es/m/CA+HiwqEDcCSDoYTtoXH4Ta_BpGN+revMZSLLj6GRz2_5ybwE1g@mail.gmail.com", "style": "normal", "type": "", "scope": "", "breaking": false}
{"repo": "vitest", "sha": "4944cf4988052950af174be8dfd688761d0e0588", "date": "2026-09-10T14:51:38+09:00", "subject": "fix(utils): fix `deepMerge` to handle prototype", "body": "", "style": "conventional", "type": "fix", "scope": "utils", "breaking": false}
{"repo": "ant-design", "sha": "636164932bb51898fb61be1b8a4b46077ae1c7d9", "date": "2026-09-10T13:59:52+08:00", "subject": "docs: keep only dnd-kit in drag-and-drop recommendations", "body": "", "style": "conventional", "type": "docs", "scope": "", "breaking": false}
{"repo": "vue-core", "sha": "65819bf4ca4e270e3080feab97bc5bf7bef0d6e9", "date": "2026-09-09T23:08:01-07:00", "subject": "refactor(runtime-vapor): render async setups from a settled hydration pass", "body": "An async setup cannot render inside its continuation: the hydration cursor\nand mode only exist during a synchronous pass. Compile the template of an\nawaiting