threads listlengths 1 2.99k |
|---|
[
{
"msg_contents": "Log Message:\n-----------\nRemove some of the most blatant brain-fade in the recent guc patch\n(it's so nice to have a buildfarm member that actively rejects naked\nuses of strcasecmp). This coding is still pretty awful, though, since\nit's going to be O(N^2) in the number of guc variables. ... |
[
{
"msg_contents": "Hello\n I am confused about the ideas of \"session\"&\"backend\"&\"process\" when I\nread the lmgr part of source code. What's the difference among them?\n\nThanks,\nRanc.\n\nHello I am confused about the ideas of \"session\"&\"backend\"&\"process\" when I read the lmgr part of source c... |
[
{
"msg_contents": "Hi all,\n\nThis patch adds a sentence on monitoring.sgml explaining that \nstats_row_level needs to be enabled if user wants to get last \nvacuum/analyze execution time.\n\nThis patch can be applied on 8.2 branch and HEAD.\n\nRegards.\n\n\n-- \nGuillaume.\nhttp://www.postgresqlfr.org\nhttp://... |
[
{
"msg_contents": "The UUID generation functions are tagged IMMUTABLE,\nshouldn't they be VOLATILE?\n\n-- \nmarko\n",
"msg_date": "Sun, 22 Apr 2007 22:33:15 +0300",
"msg_from": "\"Marko Kreen\" <markokr@gmail.com>",
"msg_from_op": true,
"msg_subject": "contrib/uuid-ossp: immutable vs. volatile"
... |
[
{
"msg_contents": "One of the few situations where I experience poor performance under PostgreSQL, compared to other, commercial databases, is when an EXISTS predicate is used. Actually, these often do perform quite well, but there are some situations where there are optimizations available which other product... |
[
{
"msg_contents": "Hello I'm trying to implement a new SortMerge algorithm, and see how it compares with the current algorithm. In doing so, rather than using select, I'd like to introduce a new verb so that Select is not modified. \n \n I looked at the source code, but could find the place where Postgres SQ... |
[
{
"msg_contents": "Hi,\n\nI was wondering if there is any possibility to add an enhancement in\npgAdmin Query.\nNow, when we do \"select * from smth\", results are listed in DataOutput\ntab and there we have columns with the same width (something about\n260px currently). It is realy annoying that in most cases ... |
[
{
"msg_contents": "How much effort would it be to add EXPLAIN/EXPLAIN ANALYSE capability to\npl/pgsql functions?\n\nwhat I mean, is either a special mode, where \"SELECT my_plpgsql_func()\"\nwould print all query plans instead or in addition to executing them, or\nsome way for EXPLAIN to pass some flags to func... |
[
{
"msg_contents": "-----Ursprüngliche Nachricht-----\n> Von: Tom Lane [mailto:tgl@sss.pgh.pa.us] \n> > \"Dorochevsky,Michel\" <michel.dorochevsky@softcon.de> writes:\n> > Here are two panic runs with the Heikki's LOCK_DEBUG patched postgres:\n> > www.dorochevsky.de/infos/PostgresPanicProblem-2007-04-23.zip\n... |
[
{
"msg_contents": "If you have a table with a bunch of children, and these children all\nhave a primary key which is generated from the same sequence, assuming\nthat you're partitioning based on date (ie, this is a transaction\nrecord table), it would be nice if the planner could spot that all\ntables have a pr... |
[
{
"msg_contents": "Hackers,\n\nI was thinking about the upcoming release on my 32-hour epic airplane ordeal, \nand realizing that it changes PostgreSQL in a lot of ways. Between major \nimprovements to performance, major changes to the file format, and changes to \nimplicit conversions breaking backwards compa... |
[
{
"msg_contents": "All,\n\nYou'll notice that the Hyena testing system for Solaris builds is down, due to \nan attack and some required system upgrading. Since we'll have the new \nSolaris machines up within a few days, I don't plan to revive Hyena's \nBuildfarm instance. We're going to put some benchmark cod... |
[
{
"msg_contents": "While poking at Michel Dorochevsky's issue, I noticed that if a DELETE\nis done in a table that is referenced by many different foreign keys,\nwe repeat the ri_Check_Pk_Match test over again for each FK. Seems like\nit would be nice to avoid this duplicated work. But right now I don't\nsee ... |
[
{
"msg_contents": "So I was informed today that UNION types integer and text cannot be \nmatched. Alright, but it failed to tell which particular expressions \nin this 3-branch, 30-columns-each UNION clause in a 100-line statement \nit was talking about. So I made the attached patch to give some better \npoin... |
[
{
"msg_contents": "In a PL/PgSQL set-returning function, returning the result set of a\nquery requires a FOR loop and repeated invocations of the RETURN NEXT\nstatement:\n\n FOR x in SELECT ... LOOP\n RETURN NEXT x;\n END LOOP;\n\nThis works, but it seems overly verbose. It occurred to me that we c... |
[
{
"msg_contents": "The values of pg_enc enumeration are changed between 8.2 and 8.3.\nPG_JOHAB was removed and PG_EUC_JIS_2004 was added. \nAre there any incompatibility issues here?\n\nFor example, if we initialize a server(8.3) using initdb(8.3) and libpq(8.2)\nwith UTF-8 encoding, the server is actually init... |
[
{
"msg_contents": "Hello\n\nIt is RETURN TABLE(SQL) via ANSI SQL 2003\n\nTable function support is in patch queue:\nhttp://archives.postgresql.org/pgsql-patches/2007-02/msg00216.php\nhttp://momjian.us/mhonarc/patches/msg00001.html\n\nRegards\nPavel Stehule\n\n\n\n>In a PL/PgSQL set-returning function, returning... |
[
{
"msg_contents": "A few of us on IRC were wondering what the status of tsearch2 is in 8.3 ?\nWas it decided to include it in core or did we decide to keep FTS as a \nplugin?\nSome brief comments from anyone on the inside of the whole FTS issue \nwould be greatly appreciated by us mere end users.\nRegards,\n- N... |
[
{
"msg_contents": "Hi all,\n\n I'm one of the Google SoC's students for PostgreSQL. My project is\nto implement column-level privilege in PG. Here is a description of my\nproject. Any and all help and/or comment is appreciated.\n\n Table-level privilege subsystem in PG is now used like this:\nGRANT { { SELECT |... |
[
{
"msg_contents": "I want to apologize for having my signature at the bottom of the 8.2.4,\netc. release announcement. There is no reason to mention my name,\ncompany, and web site at the bottom of something that is supposed to\ncome from the community. And I didn't even write that email --- I was\ndoing it f... |
[
{
"msg_contents": "The resolution to my problem with the select_common_type() error message \nturned out to be that this doesn't work:\n\npostgres=# select null union select null union select 1;\nERROR: UNION types text and integer cannot be matched\n\nThat's because it resolves the first two branches independ... |
[
{
"msg_contents": "The JDBC driver does not connect successfully to a CVS HEAD server\n(updated as of today) when using a database set to UTF8 encoding;\nenabling the debug messages in the driver shows that it is sending\nclient_encoding=UNICODE in the startup packet, but the server is\nreporting client_encodin... |
[
{
"msg_contents": "two of my buildfarm members had different but pretty weird looking\nfailures lately:\n\nhttp://www.pgbuildfarm.org/cgi-bin/show_log.pl?nm=quagga&dt=2007-04-25%2002:03:03\n\nand\n\nhttp://www.pgbuildfarm.org/cgi-bin/show_log.pl?nm=emu&dt=2007-04-24%2014:35:02\n\n\nany ideas on what might causi... |
[
{
"msg_contents": "Hi, I wanto joinin the developer group of postgresql。\n But, I just donot know how to put the first step, as I installed the\npostgresql, and also get the postgresql code. after that, I also installed\nthe cygwin on my computer( as my os is windows xp). but now I wonder what's\nmy next step. ... |
[
{
"msg_contents": "I just noticed that the stage logs aren't displayed against MSVC build\nhosts as they are for regular hosts, eg:\n\nhttp://www.pgbuildfarm.org/cgi-bin/show_log.pl?nm=mastodon&dt=2007-04-25%2001:00:02\n\nvs.\n\nhttp://www.pgbuildfarm.org/cgi-bin/show_log.pl?nm=narwhal&dt=2007-04-25%2002:00:03\... |
[
{
"msg_contents": "I found that autovacuum launcher does not launch any workers in HEAD.\n\nAFAICS, we track the time to be vaccumed of each database in the following way:\n\n1. In rebuild_database_list(), we initialize avl_dbase->adl_next_worker\n with (current_time + autovacuum_naptime / nDBs).\n2. In do_st... |
[
{
"msg_contents": "I'm seeing an ECPG-Check failure on Windows Vista - any ideas what might\nbe causing this?\n\nhttp://www.pgbuildfarm.org/cgi-bin/show_log.pl?nm=vaquita&dt=2007-04-24%2020:00:05\n\nThe only other Vista buildfarm member (baiji, on the same physical box)\nis running MSVC builds which don't yet t... |
[
{
"msg_contents": "In recovery, with full_pages_writes=on, we read in each page only to \noverwrite the contents with a full page image. That's a waste of time, \nand can have a surprisingly large effect on recovery time.\n\nAs a quick test on my laptop, I initialized a DBT-2 test with 5 \nwarehouses, and let i... |
[
{
"msg_contents": "Hi,\n\nI am trying to estimate the size of a table composed of 51754000 rows.\nEach row has 31 attributes: 16 x bit(24) and 15 x bit(8)\n\nSo, the payload should be:\n51754000 x ( 16 x 24 + 15 x 24 ) bits = 3115 MB\n\nNow, from what I understand from postgresql manual is that the overhead \ni... |
[
{
"msg_contents": "Hi ,\nAny body tell me how to kill a long running query in postgresql, is \nthere any statement to kill a query, and also tell me how to log slow \nqueries to a log file.\n\nRegards\nJ Mageshwaran\n\n********** DISCLAIMER **********\nInformation contained and transmitted by this E-MAIL is pro... |
[
{
"msg_contents": "I'm in the process of archiving data on one of my PG machines. After\nbacking up the data, I delete the old records and then run a \"vacuum\nfull\" on each table.\n\nI'm vacuuming the first table now and it is taking much longer than I\nexpected (I'm now past the 2-hour mark). Some info:\n\nV... |
[
{
"msg_contents": "I am leaving Friday for a two week trip to Sydney and Melbourne,\nAustralia, Mumbai and Pune, India, and London, England.\n\nI will be attending Open Cebit in Sydney, and a Sydney PostgreSQL Users\nGroup meeting. London is planning to put together a user group meeting.\nI expect both events ... |
[
{
"msg_contents": "This was another occurance of the strange create index failure on\nNarwhal - unfortunately, despite having 'keep_error_builds' => 1 in my\nBF config it seems to have removed the tree so I can't get the dump that\nTom wanted.\n\nDoes anyone know why the keep_error_builds option didn't work in ... |
[
{
"msg_contents": "\nFor a long time one of the big gripes we get is that when using psql\ninteractively if you turn autocommit off and you typo on the nth command\nyou've suddenly lost all your work. The response was always that we needed a\ngeneric subtransaction facility to handle it.\n\nWe've had such a fac... |
[
{
"msg_contents": "Hi,\nI try to update a tuple in pg_attribute table by using the function\nsimple_heap_update while somequery processing is going on. But when I use to\nsee the value of that tuple(just updated) once again by SearchSysCache\nfunction in the same query processing, I am not able to see the updat... |
[
{
"msg_contents": "Now that we are half-way though the scheduled feature freeze, I wanted\nto share my thoughts about this period.\n\nHaving just pushed all open items into the patches queue or 8.4 hold\nqueue, I am seeing that we have many more in-process patches than we\nnormally do at this stage of the proce... |
[
{
"msg_contents": "Hi,\n When the lock manager's data structures were split into \"partitions\",\nhow many such data structures can one partition control? Since we use\nLOCKTAG's hash value to decide the partition which the lock should in, can\nall locks be split into ONE partition?\n\n Regards,\n ranc... |
[
{
"msg_contents": "This email applies ONLY to the native Windows port of PostgreSQL, not to\nany other ports.\n\nVersion 8.0 was the first version of PostgreSQL that ran on Windows\nnatively. As most of you know, the porting work was a large and complex\ntask that spanned two development cycles and took a great... |
[
{
"msg_contents": "Hi,\nI have dropped a column (say column name is 'A') from the relation R. By\nsetting the attisdropped as true in the pg_catalog.pg_attribute table. But\nthe column is dropped locially not the physically. Can you please tell me\nwhen this column will be physically also dropped. Is this colum... |
[
{
"msg_contents": "\n\nNow that we seem to have MSVC building working tolerably well, I think \nwe need a bit of cleanup. In particular, I think the config setup needs \nto be more like the arguments we pass to the standard configure script. \nThis shouldn't be hard, but I think it should be done before we go t... |
[
{
"msg_contents": "\nIn summary, for a patch to be applied, someone has to understand the\npatch and the subsystem it modifies. In the past, most complex patches\ncame from experienced developers, so even if no one but the author fully\nunderstood the patch, we could rely on the author to some extent. With\nn... |
[
{
"msg_contents": "I've uploaded a preview release of the Windows installer for PostgreSQL\n8.3 to http://pgfoundry.org/frs/?group_id=1000007.\n\nThis preview is intended to get some initial testing and feedback on the\nnew architecture that we intend to use for future installers. In a\nnutshell, most of the bu... |
[
{
"msg_contents": "few days before, I said I wanto joinin the postgresql develope group. and\nrecently, I make some progress, such as get throught the base steps. but\nwhen i add some functions to the spatial data, such as box_add1() to box\ndata, but when I compile it, and install it to my computer, it turns o... |
[
{
"msg_contents": "Hey ya,\n\n Found a behaviour that caught my attention. While it's not a real\nproblem, and definetely not serious, I realize that it may be good to\nwrite about it, cause I couldn't find previous posts on the matter.\n\n Basically, I am preparing a presentation for an install party, and\n... |
[
{
"msg_contents": "my postgresql source code is at c:/mingw/postgresql and instal to\nC:/msys/1.0/local/pgsql/\nI add a function to src\\backend\\utils\\adt\\geo_ops.c as the following:\n*Datum\nbox_add2(PG_FUNCTION_ARGS)\n{\n BOX *box = PG_GETARG_BOX_P(0);\n Point *p = PG_GETARG_POINT_P(1);*\n\n* PG_RET... |
[
{
"msg_contents": "Arjen van der Meijden told me that according to the tweakers.net\nbenchmark, HEAD is noticeably slower than 8.2.4, and I soon confirmed\nhere that for small SELECT queries issued as separate transactions,\nthere's a significant difference. I think much of the difference stems\nfrom the fact ... |
[
{
"msg_contents": ">\n> my postgresql source code is at c:/mingw/postgresql and instal to\n> C:/msys/1.0/local/pgsql/\n> I add a function to src\\backend\\utils\\adt\\geo_ops.c as the following:\n> *Datum\n> box_add2(PG_FUNCTION_ARGS)\n> {\n> BOX *box = PG_GETARG_BOX_P(0);\n> Point *p = PG_GETARG_POINT... |
[
{
"msg_contents": "Need help. I want to create new system catalog and extend existing commands\n- CREATE ROLE and another commands connected with roles. I can't find how to\nadd new features into parser. Can you help me?\n\nThanks.\n\n Akmal.\n\nNeed help. I want to create new system catalog and extend existing... |
[
{
"msg_contents": "Hello!\n\nI wanna know where can I find the source code that writes tuples on disk,\nright now I'm reading about object persistence and I wanna see how it is\ndone in pgsql\n\nI hope you can help me\nregards\njorge\n\nHello!\n \nI wanna know where can I find the source code that writes tuples... |
[
{
"msg_contents": "I'm taking over Simon's heap page diagnostic functions to get them into \nshape for committing.\n\nI'm thinking of having a new contrib-module for them: pgforensics. All \nthe new functions go there, and I'm also going to move bt_page_items, \nbt_page_stats, and bt_metap functions from pgstat... |
[
{
"msg_contents": "On Mon, 2007-30-04 at 08:56 -0300, Heikki Linnakangaspgsql wrote:\n> Date: Mon, 30 Apr 2007 09:18:36 +0100\n> From: Heikki Linnakangas <heikki@enterprisedb.com>\n> To: Tom Lane <tgl@sss.pgh.pa.us>\n> Cc: Dave Page <dpage@postgresql.org>, Simon Riggs\n> <simon@2ndquadrant.com>, \n> Bruce Momj... |
[
{
"msg_contents": "I have developed a small patch to optimizer/util/plancat.c that \neliminates one of hte caveats associated with constraint exclusions, \nnamely the inability to avoid searching tables based on the results of \nstable functions. The new code expands non-volatile functions into \nconstant val... |
[
{
"msg_contents": "OK, so posted. ;-)\n\nTo clarify for the larger audience: without the plain \"gss\" \nmechanism, the \"gss-np\" mechanism provides exactly the same \nfunctionality as the existing krb5 mechanism. It will properly \nsecure the initial connection, but will not do anything once the \nconn... |
[
{
"msg_contents": "Hi,\nI want to add a column in the table and also want to hidde that column from\nthe user. So that user's normal insertion and update goes on (means user\nwill not supply the value for my hidden column). Is there any way to hidde\nthe column from the user. I tried to make my column (attisdro... |
[
{
"msg_contents": "I just realized that there's a problem with the patch I applied here\nhttp://archives.postgresql.org/pgsql-committers/2007-03/msg00057.php\nto ensure that the result type of an inline'd SQL function is correctly\nmarked in the resulting expression tree. To wit, it doesn't work for\nfunctions... |
[
{
"msg_contents": "\n\n> ------- Original Message -------\n> From: Andrew Dunstan <andrew@dunslane.net>\n> To: Josh Berkus <josh@agliodbs.com>\n> Sent: 01/05/07, 21:10:07\n> Subject: Re: [HACKERS] Feature freeze progress report\n> \n> So if the commercial \n> backers of PostgreSQL want better management of the ... |
[
{
"msg_contents": "Simon,\n\nI am forwarding this to pgsql-hackers. I can send the requirements once\nI get the right contact.\n\nThanks.\n\nAle Raza.\n\n\n-----Original Message-----\nFrom: Simon Riggs [mailto:simon@2ndquadrant.com] \nSent: Thursday, April 26, 2007 1:02 AM\nTo: Ale Raza\nCc: Tom Lane; postgresq... |
[
{
"msg_contents": "I notice that we have two versions of not INHERITing:\n\nALTER ROLE meek NOINHERIT earth;\n\nALTER TABLE meek NO INHERIT earth;\n\nIs there some merit in deciding on just one of these syntaxes? It seems\nlike we will have to support both the above, but we should encourage\njust one common way... |
[
{
"msg_contents": "Bruce Momjian <bruce@momjian.us> writes:\n> FYI, Tom, Heikki, I need one of you to post the list of patches and\n> where we think we are on each one, even if the list is imperfect.\n\nThis message is an attempt to sort out which patch queue entries have\nno hope of getting into 8.3 (and so we... |
[
{
"msg_contents": "Hello Tom,\n\nAll what you wrote is true. plpgpsm copies-and-pastes about 30% of code and \nIt is terrible for long run. But when I can change it? Writing differnt \nruntime is nonsense, better way is refactoring plpgsql and then sharing code \nwith its. It's not propable in 8.4 .. there wil... |
[
{
"msg_contents": "Hi,\n \n I am getting the parse error while i try to execute a simple sql query in postgres.\n \n java.sql.SQLException: ERROR: parser: parse error at or near \"and\" at character 58\n \n The Query has been changed and it is very much like the below one\n \n select * from emp where empName... |
[
{
"msg_contents": "Hi,\n\nI'm starting to review the \"synchronized scans\" and \"scan-resistant \nbuffer cache\" patches. The patches have complex interactions so I'm \ntaking a holistic approach.\n\nThere's four outstanding issues with the sync scans in particular:\n\n1. The simplistic hash approach. While it... |
[
{
"msg_contents": "Hi,\n\nWhat would cause or what should I do with a table that hangs during its \nreindexing? I can see in the messages the last table to reindex, so \nsequentially the next table must be the problem.\nI have statement_timeout set for 6 hours, then I get \"canceling statement \ndue to stateme... |
[
{
"msg_contents": "Hello.\n\nI wanted to create a new system catalog in Postgres. So I changed the source\ncode. Everything seems to be OK. I compiled it, but now after initdb I\nreceive:\n\n$ initdb -D /usr/home/postgres/post1\nThe files belonging to this database system will be owned by user\n\"postgres\".\nT... |
[
{
"msg_contents": "Hi,\n\nThe recently discovered autovacuum bug made me notice something that is\npossibly critical. The current autovacuum code makes an effort not to\nleave workers in a \"starting\" state for too long, lest there be failure\nto timely tend all databases needing vacuum.\n\nThis is how the la... |
[
{
"msg_contents": "I noticed that conversion_procs is sadly single-tasked to build. I am\nwondering if it would be acceptable to rework the Makefile.shlib to have\nan option to allow building multiple libs, by creating a rule to collect\nlibraries to build, and have each conversion_proc Makefile add a target\n... |
[
{
"msg_contents": "Here's a shorter version:\n\nOn the date variant, I wasn't sure how to handle intervals with parts \nsmaller than days:\nfloor, ceiling, round or error out\nTo get round, the last parameters of generate_series would be\nextract('epoch' FROM '1 day'::interval)::bigint * round(extract('epoch' F... |
[
{
"msg_contents": "Here's a shorter version:\n\nOn the date variant, I wasn't sure how to handle intervals with parts \nsmaller than days:\nfloor, ceiling, round or error out\nTo get round, the last parameters of generate_series would be\nextract('epoch' FROM '1 day'::interval)::bigint * round(extract('epoch' F... |
[
{
"msg_contents": "I'm reviewing the strxfrm patch, and while comparing that code to the\ncode in varstr_cmp (which uses the same UTF8/UTF16 workaround but for\nstrcoll instead), and I noticed that in varstr_cmp we have an\noptimization to use a stack based buffer instead of palloc if the string\nis short enoug... |
[
{
"msg_contents": "Hi all,\n\nAttached is an updated bitmap index patch. It contains bug fixes, API\nchanges, binary changes (page identifier to distinguish it from other\nindexes) and has been brought up to HEAD.\n\nI worked on a few approaches to VACUUM, none very satisfactory. The\nproblem is, breaking a com... |
[
{
"msg_contents": "For info, the buildfarm script failed to leave the broken tree behind \nagain so I was unable to get a dump from the affected index.\n\nAndrew; My run logs show that the script did think it was leaving the \ntree behind (it included the 'leaving error trees' text that you asked \nme to add to... |
[
{
"msg_contents": "We recently experienced a hard media failure. It turns out our admin guys\nhave been doing backups by backing up the pgsql data directory, but not\nprepping the db first.\n\nAfter a restore, when I try to start the database I get \"LOG: logger\nshutting down\" and nothing else. This is Postg... |
[
{
"msg_contents": "I often need in command line to get the code of function, so I make a patch\nfor pg_dump, thanks this patch pg_dump is able to dump only one functions or\nall the functions.\n\nThe argument is --object or -B\n\nExample:\n./pg_dump -Bfunction:test_it -Bfunction:dblink_open\nTo dump the functi... |
[
{
"msg_contents": "I just noticed that my recent change to prevent PG_RE_THROW from dying\nif there's noplace to longjmp to has provoked a whole lot of warnings\nthat were not there before. Apparently this is because gcc understands\nthat siglongjmp() never returns, but is not equally clueful about\npg_re_thro... |
[
{
"msg_contents": "I have done the following test and I am unable to understand the results. I have tried debugging the code and I have reached down to the Storage Layer. I am playing with the optimizer etc.. I no very little about the internals of the Executor.\n\nIf you could point out to me what possible e... |
[
{
"msg_contents": "Log Message:\n-----------\nTeach tuplesort.c about \"top N\" sorting, in which only the first N tuples\nneed be returned. We keep a heap of the current best N tuples and sift-up\nnew tuples into it as we scan the input. For M input tuples this means\nonly about M*log(N) comparisons instead ... |
[
{
"msg_contents": "Log Message:\n-----------\nPut Autotools-generated files into subdirectory config/; add macro files\nused from PostgreSQL there so you don't need a PostgreSQL source tree to\nbootstrap the code.\n\nModified Files:\n--------------\n psqlodbc:\n Makefile.am (r1.43 -> r1.44)\n (... |
[
{
"msg_contents": "Tom,\nDid you restart Postgres and drop file system caches?\n\nWhat I am suspecting is that some sort of prefetching is happening. \nI know that Postgres does not do prefetching. \nI also understand very little about OS/FileSystem level prefetching.\n\n----- Original Message ----\nFrom: Tom ... |
[
{
"msg_contents": "Hello,\nHopefully this hasn't been answered already. I attempted to do full\ndue-diligence on it, and could find no answer anywhere.\n\nBasically my problem is that my server has both python 2.3 and python\n2.4 installed (strange but unavoidable reasons behind this), and it\nappears that my ... |
[
{
"msg_contents": "Hello!\n\nI have a potential situation where I will have a lot of sensor data coming\nin very often. (every second or so) The sensor data is from physics type\nmeasurements, and will normally follow a slowly changing pattern with\nsinusoidal disturbances. The overall shape of the data is more... |
[
{
"msg_contents": "What is the reasoning behind having two different implementations of the\ndatetime types, with slightly different behavior? Do we intend to keep\nsupporting both FP- and integer-based datetimes indefinitely?\n\nClearly, there are some costs associated with maintaining two different\nimplement... |
[
{
"msg_contents": "\n\n> ------- Original Message -------\n> From: Bruce Momjian <bruce@momjian.us>\n> To: PostgreSQL-development <pgsql-hackers@postgresql.org>\n> Sent: 05/05/07, 03:00:25\n> Subject: [HACKERS] New idea for patch tracking\n> \n> As for #3, again, I don't want us to take on a burdensome patch tr... |
[
{
"msg_contents": "The current TODO list has:\n\n\tDependency Checking\n\t===================\n\t\n\t* Flush cached query plans when the dependent objects change,\n\t when the cardinality of parameters changes dramatically, or\n\t when new ANALYZE statistics are available\n\t\n\t A more complex solution woul... |
[
{
"msg_contents": "As promised I added a Patch Status site to the wiki:\n\nhttp://developer.postgresql.org/index.php/Todo:PatchStatus\n\nIt contains all the patches tom mentioned in his recent patch queue\ntriage mail with references to the patches (either on the queue or in\nthe archives) as well as the author... |
[
{
"msg_contents": "\n\nIn connection with completing David Fetter's array of composites patch, \nI am looking at doing some better name mangling for array types as \nrecently discussed. What I'm thinking of is prepending one or more \nunderscores to the type name up to some limit (NAMEDATALEN / 2 ?) and if \nne... |
[
{
"msg_contents": "\n\n> ------- Original Message -------\n> From: Bruce Momjian <bruce@momjian.us>\n> To: Dave Page <dpage@postgresql.org>\n> Sent: 05/05/07, 11:06:37\n> Subject: Re: [HACKERS] New idea for patch tracking\n> \n> > <snip tracker outline>\n> > \n> > Barring a few trivial details, that sounds almo... |
[
{
"msg_contents": "\nWhat is the approved way to iterate over a relation's attributes? I see \nthat lsyscache.c::get_relnatts() is marked NOT_USED and has been for \nnearly seven years. Maybe it's time to remove that code ;-)\n\ncheers\n\nandrew\n",
"msg_date": "Sat, 05 May 2007 13:09:31 -0400",
"msg_fr... |
[
{
"msg_contents": "\nI have been talking with Theo some more about his recent problem with \nbytea arguments and results (see recent discussion on -bugs and also \nrecent docs patch), what he needs is a way to have bytea (and possibly \nother unknown types) passed as binary data to and from plperl. The \nconv... |
[
{
"msg_contents": "Let me give you my approach to tracking. It might help set the stage\nfor moving forward. My goal has always been to foster discussion and\npull as many TODO items and patches from the discussion as possible (and\nothers do that as well by saying \"Please add to TODO\" or applying\npatches)... |
[
{
"msg_contents": "Hello, I have a system where there are mostly COPYs,\nwhich insert data into a table. Ocasionally a COPY will fail (and thus,\ndead rows appear), but as far as I can tell ROLLBACK is not reflected\nanywhere in the pg_stats_user_tables. And since there are no rows\nn_tup_upd or n_tup_del, th... |
[
{
"msg_contents": "As suggested in the TODO list (and as I need the functionality \nmyself), I have implemented the current_query interface to \ndebug_query_string.\n\nI'm not sure the best place to put this, suggestions welcome..\n\nPlease review the patch attached.\n\nCheers\nTom",
"msg_date": "Mon, 7 M... |
[
{
"msg_contents": "Just noticed buffile.c:292 doesn't look quite right where\nBufFileDumpBuffer calls FileWrite:\n bytestowrite = FileWrite(thisfile, file->buffer, bytestowrite);\nIt looks as though it would write the same data each time the\nloop is iterated. Would this be better?\n bytestowrite = Fil... |
[
{
"msg_contents": "I've been studying the SQL spec in a bit more detail and I'm suddenly\nthinking that we've got the behavior all wrong in the current\nGENERATED/IDENTITY patch. In particular, it looks to me like we've\nbeen implementing GENERATED ALWAYS AS (expr) according to the rules\nthat the spec in fact... |
[
{
"msg_contents": "\nI added a field to pg_type, updated all the bootstrap catalog entries, \nchanged Natts_pg_type, and (I think) fixed the two places in pg_type.c \nthat construct pg_type entries. However, initdb fails at the sysviews \ncreation stage - the core dump shows it failing as shown below. Anyone \n... |
[
{
"msg_contents": "Here's my roadmap for the \"scan-resistant buffer cache\" and \n\"synchronized scans\" patches.\n\n1. Fix the current vacuum behavior of throwing dirty buffers to the \nfreelist, forcing a lot of WAL flushes. Instead, use a backend-private \nring of shared buffers that are recycled. This is w... |
[
{
"msg_contents": "Hi,\n\nI am migrating from Sybase to Postgres.\n\ntrying to get a hold on the issue of future releases feature list.\n\n1. Where can I find the feature list for 8.3 ? When is it expected?\n2. And for 8.4?\n3. Who is the guy to ask about a feature like \"startup migrate\" in ORACLE?\n\nThanks... |
[
{
"msg_contents": "Hi,\n\nWe are using the following piece of code for updating a list\nsynchronously i.e. no two threads would update the list at a time. But\nthe postgres hangs for some reason after a few calls to UpdateList\nfunction. The program (i.e. postgres) does not even react to Ctrl+C or\nany other in... |
[
{
"msg_contents": "\nThis thread here became a TODO item:\n\nhttp://archives.postgresql.org/pgsql-hackers/2006-11/msg01011.php\n\nDuring that discussion a patch was produced that nobody seemed to have\nobjections to. The main problem seemed to be that it wouldn't always\ndowngrade the lock on a ROLLBACK TO (as ... |
[
{
"msg_contents": "Hi,\n\nI asked about this before, but the thread got hijacked to discuss \nanother buildfarm failure :-(. Currently our only Windows Vista \nbuildfarm member (Vaquita) fails every time (assuming it gets that far) \non ECPG's dt_test and update tests.\n\nI've checked the FS permissions, and se... |
[
{
"msg_contents": "Hello all,\n\nI am trying to move from GiST intarray index to GIN intarray index, but my\nGIN index is not being used by the planner.\n\nThe normal query is like that\n\nselect *\n from sourcetablewith_int4\n where ARRAY[myint] <@ myint_array\n and some_other_filters\n\n(with GiST index ev... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.