id stringlengths 4 10 | text stringlengths 4 2.14M | source stringclasses 2
values | created timestamp[s]date 2001-05-16 21:05:09 2025-01-01 03:38:30 | added stringdate 2025-04-01 04:05:38 2025-04-01 07:14:06 | metadata dict |
|---|---|---|---|---|---|
1742102057 | Install Sui failed on M1 Mac
Command
cargo install --locked --git https://github.com/MystenLabs/sui.git --branch devnet sui
Error
error: linking with cc failed: exit status: 1
······
ld: symbol(s) not found for architecture arm64
clang: error: linker command failed with exit code 1 (use -v to see invocation)
error: could not compile sui-framework due to previous error
System Information
rustup: stable-aarch64-apple-darwin unchanged - rustc 1.70.0 (90c541806 2023-05-31)
cmake: cmake version 3.26.4
OS: macOS Ventura 13.4, Apple M1 Pro
So how can I fix this and install Sui?
Command
cargo install --locked --git https://github.com/MystenLabs/sui.git --branch devnet sui
Error
error: linking with cc failed: exit status: 1 ······ ld: symbol(s) not found for architecture arm64 clang: error: linker command failed with exit code 1 (use -v to see invocation) error: could not compile sui-framework due to previous error
System Information
rustup: stable-aarch64-apple-darwin unchanged - rustc 1.70.0 (90c541806 2023-05-31)
cmake: cmake version 3.26.4
OS: macOS Ventura 13.4, Apple M1 Pro
So how can I fix this and install Sui?
It looks like a rust environment issue, not a sui issue.
Try to add the following sections to your ~/.cargo/config (if you don't have this file feel free to create):
[target.x86_64-apple-darwin]
rustflags = [
"-C", "link-arg=-undefined",
"-C", "link-arg=dynamic_lookup",
]
[target.aarch64-apple-darwin]
rustflags = [
"-C", "link-arg=-undefined",
"-C", "link-arg=dynamic_lookup",
]
Works for me! I'll let you know if I have any other issues.
@alphafitz11 Feel free to create a new issue if it still does not work for you.
| gharchive/issue | 2023-06-05T16:07:46 | 2025-04-01T04:32:49.107873 | {
"authors": [
"MarcoDotIO",
"alphafitz11",
"dustin6341",
"hello2jie",
"stefan-mysten"
],
"repo": "MystenLabs/sui",
"url": "https://github.com/MystenLabs/sui/issues/12344",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
1518161787 | narwhal does not shut down reliably during reconfiguration
To reproduce:
RUST_LOG=sui=debug,msim=error,narwhal=debug,info cargo simtest --test reconfiguration_tests test_passive_reconfig --run-ignored all --no-capture > /tmp/log
The exact output will depend on what commit you run this at (I was at e48c8b2d7), but here is what I observe:
; grep narwhal_manager /tmp/log
2022-01-03T02:04:56.000004Z INFO node{id=4 name="k#91a96f6b.."}: sui_core::narwhal_manager: crates/sui-core/src/narwhal_manager/mod.rs:155: Starting up narwhal
2022-01-03T02:04:56.000004Z INFO node{id=4 name="k#91a96f6b.."}: sui_core::narwhal_manager: crates/sui-core/src/narwhal_manager/mod.rs:188: Starting up narwhal is complete
2022-01-03T02:04:56.000011Z INFO node{id=5 name="k#8cbb4f31.."}: sui_core::narwhal_manager: crates/sui-core/src/narwhal_manager/mod.rs:155: Starting up narwhal
2022-01-03T02:04:56.000011Z INFO node{id=5 name="k#8cbb4f31.."}: sui_core::narwhal_manager: crates/sui-core/src/narwhal_manager/mod.rs:188: Starting up narwhal is complete
2022-01-03T02:04:56.000021Z INFO node{id=2 name="k#8ae0118f.."}: sui_core::narwhal_manager: crates/sui-core/src/narwhal_manager/mod.rs:155: Starting up narwhal
2022-01-03T02:04:56.000021Z INFO node{id=2 name="k#8ae0118f.."}: sui_core::narwhal_manager: crates/sui-core/src/narwhal_manager/mod.rs:188: Starting up narwhal is complete
2022-01-03T02:04:56.000097Z INFO node{id=3 name="k#a46530e7.."}: sui_core::narwhal_manager: crates/sui-core/src/narwhal_manager/mod.rs:155: Starting up narwhal
2022-01-03T02:04:56.000097Z INFO node{id=3 name="k#a46530e7.."}: sui_core::narwhal_manager: crates/sui-core/src/narwhal_manager/mod.rs:188: Starting up narwhal is complete
2022-01-03T02:05:02.789219Z INFO node{id=2 name="k#8ae0118f.."}: sui_core::narwhal_manager: crates/sui-core/src/narwhal_manager/mod.rs:120: Sending shutdown message to narwhal
2022-01-03T02:05:02.792368Z INFO node{id=3 name="k#a46530e7.."}: sui_core::narwhal_manager: crates/sui-core/src/narwhal_manager/mod.rs:120: Sending shutdown message to narwhal
2022-01-03T02:05:02.818062Z INFO node{id=5 name="k#8cbb4f31.."}: sui_core::narwhal_manager: crates/sui-core/src/narwhal_manager/mod.rs:120: Sending shutdown message to narwhal
2022-01-03T02:05:02.822548Z INFO node{id=4 name="k#91a96f6b.."}: sui_core::narwhal_manager: crates/sui-core/src/narwhal_manager/mod.rs:120: Sending shutdown message to narwhal
2022-01-03T02:05:02.964260Z INFO node{id=2 name="k#8ae0118f.."}: sui_core::narwhal_manager: crates/sui-core/src/narwhal_manager/mod.rs:134: Narwhal shutdown is complete
2022-01-03T02:05:02.964260Z INFO node{id=2 name="k#8ae0118f.."}: sui_core::narwhal_manager: crates/sui-core/src/narwhal_manager/mod.rs:155: Starting up narwhal
2022-01-03T02:05:02.964260Z INFO node{id=2 name="k#8ae0118f.."}: sui_core::narwhal_manager: crates/sui-core/src/narwhal_manager/mod.rs:188: Starting up narwhal is complete
2022-01-03T02:05:06.042422Z INFO node{id=4 name="k#91a96f6b.."}: sui_core::narwhal_manager: crates/sui-core/src/narwhal_manager/mod.rs:134: Narwhal shutdown is complete
2022-01-03T02:05:06.042422Z INFO node{id=4 name="k#91a96f6b.."}: sui_core::narwhal_manager: crates/sui-core/src/narwhal_manager/mod.rs:155: Starting up narwhal
2022-01-03T02:05:06.042422Z INFO node{id=4 name="k#91a96f6b.."}: sui_core::narwhal_manager: crates/sui-core/src/narwhal_manager/mod.rs:188: Starting up narwhal is complete
Notice that all 4 validators log the Sending shutdown message to narwhal message, but only 2 of them finish shutting down and starting up again.
I looked into what is causing this:
The immediate problem is that https://github.com/MystenLabs/sui/blob/main/crates/sui-core/src/narwhal_manager/mod.rs#L132 has no timeout, and so can hang forever. I tried adding a timeout and abort()ing the tasks forcibly if the timeout was reached, but this caused the ports not to be released, so narwhal wasn't able to start back up in the new epoch.
Then I started tracing the ReconfigureNotification::Shutdown message through the system, and became convinced that it will be almost impossible to make the shutdown process reliable given the way it is currently written. I don't have any specific ideas on how to fix this, since I don't know the code very well, but it seems to be that the reconfiguration path in narwhal_manager needs to have a way to force shutdown, e.g. by dropping owned values. The approach of making a request to the admin server (which can simply hang or fail) seems untenable.
There could be multiple issues. In private testnet I see these logs on a node
2023-01-03 15:33:11 | 2023-01-03T23:33:11.345255Z INFO sui_core::narwhal_manager: Sending shutdown message to narwhal
2023-01-03 15:33:27 | 2023-01-03T23:33:27.120944Z INFO sui_core::narwhal_manager: Narwhal shutdown is complete
2023-01-03 15:33:27 | 2023-01-03T23:33:27.153696Z INFO sui_core::narwhal_manager: Starting Narwhal with store path /opt/sui/db/consensus_db/epoch1
2023-01-03 15:33:27 | 2023-01-03T23:33:27.193061Z INFO sui_core::narwhal_manager: Starting up narwhal is complete
But the node did not seem to have unregistered the PrimaryReceiverHandler from the previous epoch either,
2023-01-03 15:33:29 | 2023-01-03T23:33:29.196447Z WARN request{route=/narwhal.PrimaryToPrimary/RequestVote remote_peer_id=e9cd3bcd direction=outbound}: anemo_tower::trace::on_failure: response failed error=Status code: 520 Unknown InvalidEpoch("Header verify", 0, 1) latency=0 ms
......
I agree reliable and fast shutdown via the reconfig message is hard. I believe we have discussed using mechanisms like JoinSet / TaskGroup to forcefully shutdown Narwhal. @akichidis @laura-makdah will have better ideas on why we choose the current approach.
its easy enough to forcibly abort tasks, but right now the tasks don't own the ports, so the port is still unavailable after aborting the tasks.
It would be easier if we don't have to shut down the network and reuse it
My main confusion is that Narwhal@epoch0 seems to have shutdown successfully, and Narwhal@epoch1 seems to have started successfully. Yet primary handler from epoch 0 still seems to be active. Might be an issue in how Narwhal interfaces with network. Agree that it would be better if Narwhal does not have to unregister and re-register network services.
https://github.com/MystenLabs/sui/pull/6630 can be relevant.
I don't understand the claim that this will be simpler without unregistering/reregistering network services. The problem is not with shutting down servers and starting them - the problem is that that step never happens because the signaling used to trigger that is unreliable.
(However, If not shutting down the network for some reason meant we didn't need all this complicated signaling then that would be different.)
Thanks for the above info all! I'll have a closer look on this once back from PTO on Monday, but just to better understand the issue flagged here - we are talking about instances of Narwhal where nodes never manage shutdown and process just hangs there, right?
Then I started tracing the ReconfigureNotification::Shutdown message through the system, and became convinced that it will be almost impossible to make the shutdown process reliable given the way it is currently written
The problem is not with shutting down servers and starting them - the problem is that that step never happens because the signaling used to trigger that is unreliable.
@mystenmark could you please elaborate more on your concerns here? - as I see challenges my self especially around networking but I would like to see your thoughts given your current investigation
Also there is a refactor PR here https://github.com/MystenLabs/sui/pull/7064 as an attempt to make the shutdown process more reliable by removing the need to talk to the network admin interfaces and eliminate the primary <-> worker communication during it.
#6630 can be relevant.
@mwtian that PR was an attempt to showcase that if we need to shutdown everything except networking is doable with some measured effort. @mystenmark so far the biggest source volatility in terms of shutting down a node has been the network of various reason (@bmwill made heroic efforts to give us in the anemo library a way to explicitly shutdown network) but there are challenges that still might exist. As network is holding references to other objects, not having the network shutdown can make things hang.
The problem is not with shutting down servers and starting them - the problem is that that step never happens because the signaling used to trigger that is unreliable.
@mystenmark could you please elaborate more on your concerns here? - as I see challenges my self especially around networking but I would like to see your thoughts given your current investigation
The argument I'm making is roughly this:
As a matter of first principles, if you have a file descriptor corresponding to a bound socket, you can close the file descriptor and then immediately rebind to the port. If we were using TCP, we'd probably have a tokio TcpListener, which we could drop and recreate, and that would just work. I understand anemo is different, but there shouldn't be any fundamental difficulties with closing a socket and re-opening it.
Therefore if we have a way of reliably calling code that will cause the sockets to be closed, there should not be any difficulties. The problem is that we do not seem to have a reliable way of doing that, because the reconfiguration path goes through many layers of network calls and channels.
Obviously we are not operating in a world of first principles, but in the world of making expedient changes to existing code. To that end, I attempted to fix this issue by putting a timeout on the join_all call, and then calling abort() on the JoinHandles to forcibly terminate the tasks. Unfortunately this did not resolve the issue because the tasks themselves don't own the network sockets, so nothing ever got closed, so they could not be rebound when starting the new tasks.
I don't know this code very well and don't want to strongly argue for any specific approaches to fix this problem. But it does seem to me like we are dealing with something like this:
Reliable Signaling
Unreliable Signaling
restart networking
Should work
Won't work reliably
don't restart networking
Should work
Won't work reliably
All that being said, I recognize that there may be contingent factors about the code that make a "don't restart networking" approach much simpler in practice.
even anemo doesn't own the network sockets, the underlying library quinn does and is running a number of background tasks which we do not explicitly have handles to. I've tried as best as I can to add the ability to explicitly shut down the network, and in my limited testing it has been successful although there may be cases where it is slow.
I have found the source of this hanging shutdown to be due to the failure to join on the task of the Narwhal to worker network call to shutdown. This means that Tasos's PR will potentially fix the issue.
even anemo doesn't own the network sockets, the underlying library quinn does and is running a number of background tasks which we do not explicitly have handles to.
ah. well, that's unfortunate. (and surprising! you can't shut down a listening socket in quinn?)
Thinking some more: Regardless of what approach we take for reconfiguration, it seems that we do want to be able to reliable shut down narwhal in the case where a validator leaves the committee and is demoted to a fullnode. Otherwise it may continue to try to talk to the current committee and spew all kinds of error logs. It shouldn't degrade the behavior of the committee since we need to handle the case where someone does this maliciously.
Regardless of what approach we take for reconfiguration, it seems that we do want to be able to reliable shut down narwhal in the case where a validator leaves the committee and is demoted to a fullnode. Otherwise it may continue to try to talk to the current committee and spew all kinds of error logs.
This is nice to have but not a must. A validator leaving the committee could really always shut down the entire process first. It's not very much a priority to handle this well.
| gharchive/issue | 2023-01-04T01:37:13 | 2025-04-01T04:32:49.126112 | {
"authors": [
"akichidis",
"bmwill",
"laura-makdah",
"lxfind",
"mwtian",
"mystenmark"
],
"repo": "MystenLabs/sui",
"url": "https://github.com/MystenLabs/sui/issues/7108",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
2605995703 | 🛑 Mysteryexe is down
In fcf37a7, Mysteryexe (https://mysteryexe.xyz) was down:
HTTP code: 403
Response time: 935 ms
Resolved: Mysteryexe is back up in 278bee6 after 14 minutes.
| gharchive/issue | 2024-10-22T16:55:38 | 2025-04-01T04:32:49.131639 | {
"authors": [
"Mysteryexe"
],
"repo": "Mysteryexe/uptime-checker",
"url": "https://github.com/Mysteryexe/uptime-checker/issues/758",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
1987989217 | "Anchor offset" in selected activity and "start offset" should use same time format
Checked for duplicates
No - I haven't checked
Is this a regression?
No - This is a new bug
Version
1.14.0
Describe the bug
The offset field in the selected activity panel uses a different time format than in the "offset" column in the activity directives table.. This makes it hard for the user to compare offsets in the different parts of the UI.
Reproduction
Add an activity to a plan. Activity directives table table display offset in hours (e.g., 02:03:25.77) but selected activity panel displays offset as "0d 2h 3m 25s 770ms 0us".
Logs
No response
System Info
Chrome
Severity
Minor
Any suggestion on which time format would be preferable?
Probably "0d 2h 3m 25s 770ms 0us" is preferred. Our plans are usually weeks to years long.
| gharchive/issue | 2023-11-10T16:52:47 | 2025-04-01T04:32:49.144421 | {
"authors": [
"AaronPlave",
"parkerabercrombie"
],
"repo": "NASA-AMMOS/aerie-ui",
"url": "https://github.com/NASA-AMMOS/aerie-ui/issues/995",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
2466800522 | Use gatsby browser layout
Wraps all pages in the Layout component from gatsby-browser.js instead of on the page component level. This eliminates the page flicker (coming from the delay in getting the page dimensions from the header container ref) that was occurring on navigation between each page.
@willemarcel I was able to resolve most of the failed tests by correctly mirroring the gatbsy-browser functions in gatsby-ssr. Unfortunately there are a couple tests that still fail. I believe this is a hydration issue. Strangely, while I have two tests failing locally too, they are two different tests.
| gharchive/pull-request | 2024-08-14T20:48:33 | 2025-04-01T04:32:49.146003 | {
"authors": [
"LanesGood"
],
"repo": "NASA-IMPACT/admg-casei",
"url": "https://github.com/NASA-IMPACT/admg-casei/pull/678",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
1558859058 | IPDA Logo
https://app.zenhub.com/files/267908922/49b0cf12-fccb-4600-8f46-7e14cb388bca/download
https://app.zenhub.com/files/267908922/310d6bfd-3adb-4672-98d0-d2000d653d38/download
@jordanpadams @nutjob4life are you validating this logo ?
@tloubrieu-jpl yes. this logo is good to go. @nutjob4life logo for the website! Thanks!
| gharchive/issue | 2023-01-26T22:12:21 | 2025-04-01T04:32:49.148793 | {
"authors": [
"edwardbarraza",
"jordanpadams",
"tloubrieu-jpl"
],
"repo": "NASA-PDS/design-team",
"url": "https://github.com/NASA-PDS/design-team/issues/169",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
2485214137 | Feature: 支持模糊匹配
>>> import fontra
>>> fontra.get_font_styles("Cascadia")
KeyError: "Font 'Cascadia' not found. Did you mean 'Cascadia Mono' ?"
>>> fontra.get_font("Cascadia", "Roman")
KeyError: "Font 'Cascadia' not found. Did you mean 'Cascadia Mono' ?"
>>> fontra.get_font("Cascadia Mono", "Roma")
KeyError: "Font style 'Roma' of font 'Cascadia Mono' not found. Did you mean 'Roman' ?"
感觉可以做一个类似 fc-match (fontconfig) 的东西，单独做一个用于获得最佳匹配的函数，然后给剩下的函数加个类似 fuzzy=True 或者 strict=False 的参数之类的
| gharchive/pull-request | 2024-08-25T12:12:07 | 2025-04-01T04:32:49.183680 | {
"authors": [
"KomoriDev",
"NCBM"
],
"repo": "NCBM/fontra",
"url": "https://github.com/NCBM/fontra/pull/2",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
328141372 | NCI-Agency/anet#382: Fix jsx-a11y/href-no-hash warning
By upgrading to the latest version of eslint-config-react-app (currently 3.0.0-next.66cc7a90).
@maradragan I've fixed all new issues (errors preventing the build and new warnings) that were introduced by the updated plugin.
| gharchive/pull-request | 2018-05-31T13:35:51 | 2025-04-01T04:32:49.217017 | {
"authors": [
"gjvoosten"
],
"repo": "NCI-Agency/anet",
"url": "https://github.com/NCI-Agency/anet/pull/636",
"license": "mit",
"license_type": "permissive",
"license_source": "bigquery"
} |
1341709634 | Automatically set/restrict language
Germany and Chile have raised concerns about language settings. Should we have it set so that language is a two stage process. When selecting a country the language defaults to local language with the user having to go through region then language if they wish to change language. Should we also restrict it so that a count can only be done in the local language or the english related to that country?
I'm not really sure what happens if a user chooses a language from a country with a different set of target species, presumably the options will not be correct?
@CCarvell
@JimChiazzese1 I am guessing the concerns are around your last point ie. a count being done in English but for the wrong country, in which case the options would not be correct as you suggest. Yes, selecting country first with a default to local language then proactively having to change language back to English would help this. @kazlauskis is this feasible in technical terms?
I can't remember if we imposed a country boundary on GPS locations to also restrict someone in Chile using the UK version.
I'm not really sure what happens if a user chooses a language from a country with a different set of target species, presumably the options will not be correct?
The options are tied to countries rather than languages. If a user selects Brazil, the options will be for Brazil but shown in the currently selected language. If that language doesn't have a translation, then it will default to English. Some countries will have multiple languages so we might want to keep this flexibility.
@JimChiazzese1 The app currently automates the language selection. At the point when the app is installed, it picks the language set from the device settings. So that if your primary device language is Spanish then it will use that by default and list the app countries' options in Spanish. Can you give me a scenario where this flow doesn't work?
| gharchive/issue | 2022-08-17T12:44:49 | 2025-04-01T04:32:49.257943 | {
"authors": [
"CCarvell",
"JimChiazzese1",
"kazlauskis"
],
"repo": "NERC-CEH/fit-count-app",
"url": "https://github.com/NERC-CEH/fit-count-app/issues/145",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
2353240325 | HTML find and replace fix
We have a potential solution to the problem:
@calnfynn from HsH who is on a Praktikum with the Open Science Lab TIB - this code fix appears to solve the problem. Please add this to your Notebook code and let me know if its works.
https://gist.github.com/calnfynn/360c5f5bdcff96001336c946f6b13b59
Hi @calnfynn - I need you to do two write-ups on this fix. These are needed to firstly, explain what the changes are and secondly, to show how to implement the changes in full.
The coder can figure these things out when they look at ending the Notebook, but these things need to be made explicit to the user before they get into things.
As an example: I only got that the markdownify library was converting the HTML to Markdown and we weren't replacing HTML characters any more. Also the user needs to know how to add the new import and have the library loaded into Codespace.
The above is good issue fixing practice.
It only needs to be brief.
This is a fix for the HTML bug where some HTML characters like umlats and punctuations marks are not displaying properly and causing Quarto rendering to fail on PDF output.
This is the code with the fix: https://gist.github.com/calnfynn/360c5f5bdcff96001336c946f6b13b59
Below are the instructions about how to use the bug fix.
First you need to add a new Python library:
Edit requirements.txt and a line markdownify
in the terminal run: pip install -r requirements.txt this will install the library
Next we edit your Jupyter Notebook:
At the top of cell 2: Add the instruction to import the new library. Paste in from markdownify import markdownify after import html.
In cell 2: Next is to replace the whole of the def get_text section with the code in the following Gist: Copy all from line 3 to line 19 - https://gist.github.com/calnfynn/360c5f5bdcff96001336c946f6b13b59
You can no run your Notebook and the text should now be cleanly rendered - now as a conversion from HTML to Markdown. Before you run the Notebook use the Clear all outputs button.
| gharchive/issue | 2024-06-14T12:12:33 | 2025-04-01T04:32:49.275664 | {
"authors": [
"mrchristian"
],
"repo": "NFDI4Culture/CPS-Demo",
"url": "https://github.com/NFDI4Culture/CPS-Demo/issues/12",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
310018640 | 0.7.0 beta版本 NzModalService.open办法是否被移除
What problem does this feature solve?
项目中使用的NzModalService.open办法一直都很不错,更新后无法使用
What does the proposed API look like?
希望可以更正
rename to create
| gharchive/issue | 2018-03-30T09:18:40 | 2025-04-01T04:32:49.277209 | {
"authors": [
"vthinkxie",
"xiaoreign"
],
"repo": "NG-ZORRO/ng-zorro-antd",
"url": "https://github.com/NG-ZORRO/ng-zorro-antd/issues/1250",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
337195882 | adding css "overflow-y: auto" to .antd-modal-body
What problem does this feature solve?
if there is too much content to show in the modal, some content will be outside of the modal window.
i can set css of internal div like this: overflow-y: auto, but if modal content dynamically changes, this fails again.
i manually modified .antd-modal-body, and it worked really well.
What does the proposed API look like?
adding css "overflow-y: auto" to .antd-modal-body
Since user can add it yourself, all our styles are synced with ant design react version.
We won't change it ourselves.
Since i'm new to web development, the only way i can think of changing it myself is modifying node_modules/ng-zorro-antd/src/modal/style/modal.less.
Is there another way please?
you can rewrite any style in your own style.less or style.css file.
| gharchive/issue | 2018-06-30T07:48:57 | 2025-04-01T04:32:49.280131 | {
"authors": [
"humblepride",
"vthinkxie"
],
"repo": "NG-ZORRO/ng-zorro-antd",
"url": "https://github.com/NG-ZORRO/ng-zorro-antd/issues/1759",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
739543798 | Cascader should restore its value after cancel search
Reproduction link
https://stackblitz.com/edit/angular-zz1hq5?file=src/app/app.component.ts
Steps to reproduce
select something from the cascader
type random string into the search bar
click outside the menu to close the popover
What is expected?
cascader reverted to original value
What is actually happening?
cascader remains with the search string
Environment
Info
ng-zorro-antd
10.1.1
Browser
Chrome latest
You can compare with react version https://vulov4--run.stackblitz.io
I would like to fix this issue
| gharchive/issue | 2020-11-10T03:41:01 | 2025-04-01T04:32:49.284210 | {
"authors": [
"rppig42",
"stygian-desolator"
],
"repo": "NG-ZORRO/ng-zorro-antd",
"url": "https://github.com/NG-ZORRO/ng-zorro-antd/issues/6036",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
2580206180 | 页面切换的时候select控件的宽度有个变宽的动画
Reproduction link
https://github.com/zhaopeiym/ng-test-app
Steps to reproduce
页面切换
What is expected?
宽度没有动态变化
What is actually happening?
宽度发生动态变化
Environment
Info
ng-zorro-antd
18.1.1
Browser
edge
页面切换，nz-date-picker控件的边框也会有颜色变化，灰色->黑色->灰色的变化效果。
| gharchive/issue | 2024-10-11T01:41:56 | 2025-04-01T04:32:49.287387 | {
"authors": [
"zhaopeiym"
],
"repo": "NG-ZORRO/ng-zorro-antd",
"url": "https://github.com/NG-ZORRO/ng-zorro-antd/issues/8834",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
367339892 | radix_sort.cpp: Add Radix Sort Algorithm
This adds Radix sort which sorts array by sorting from lowest significant digit to highest.
Closes https://github.com/NITSkmOS/Algorithms/issues/222
By submitting this pull request I confirm I've read and complied with the
below declarations.
[x] I have read the Contribution guidelines and I am confident that my PR reflects them.
[x] I have followed the commit guidelines for this project.
[x] My code follows the standard code structure.
[x] This pull request has a descriptive title. For example, {Tag}: Add {Algorithm/DS name} [{Language}], not Update README.md or Added new code.
[x] This pull request will be closed if I fail to update it even once in a continuous time span of 7 days.
[x] This pull request shall only be reviewed and merged once all the checks passes. No maintainer or supporter shall be obliged to review it before this condition is met.
[x] I have mentioned the issue number correctly (with hyperlink) in this pull request description.
After you submit your pull request, DO NOT click the 'Update Branch' button.
@sangamcse rebase
Hey! I'm GitMate.io! This pull request is being rebased automatically. Please DO NOT push while rebase is in progress or your changes would be lost permanently :warning:
Automated rebase with GitMate.io was successful! :tada:
ack 864eeaa
| gharchive/pull-request | 2018-10-05T19:31:15 | 2025-04-01T04:32:49.297627 | {
"authors": [
"paawan01",
"sangamcse"
],
"repo": "NITSkmOS/Algorithms",
"url": "https://github.com/NITSkmOS/Algorithms/pull/298",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
369938733 | binary_search_tree.cpp: Add Binary search tree
This adds C++ implementation of binary search tree with search, insert and delete functions.
Closes https://github.com/NITSkmOS/Algorithms/issues/54
By submitting this pull request I confirm I've read and complied with the
below declarations.
[x] I have read the Contribution guidelines and I am confident that my PR reflects them.
[x] I have followed the commit guidelines for this project.
[x] My code follows the standard code structure.
[x] This pull request has a descriptive title. For example, {Tag}: Add {Algorithm/DS name} [{Language}], not Update README.md or Added new code.
[x] This pull request will be closed if I fail to update it even once in a continuous time span of 7 days.
[x] This pull request shall only be reviewed and merged once all the checks passes. No maintainer or supporter shall be obliged to review it before this condition is met.
[x] I have mentioned the issue number correctly (with hyperlink) in this pull request description.
ack b9fe5a7
| gharchive/pull-request | 2018-10-14T20:09:27 | 2025-04-01T04:32:49.301895 | {
"authors": [
"paawan01",
"sangamcse"
],
"repo": "NITSkmOS/Algorithms",
"url": "https://github.com/NITSkmOS/Algorithms/pull/434",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
1282150334 | Off by one in Unbound 1.16.0 verbosity setting.
Describe the bug
The verbosity option in unbound.conf for controlling the log level information is being incremented plus one of what is defined in the config. .e.g. verbosity: -1 will result in a verbosity: 0 at runtime.
To reproduce
Steps to reproduce the behavior:
Follow build from source instructions here https://unbound.docs.nlnetlabs.nl/en/latest/getting-started/installation.html#building-from-source-compiling on Ubuntu 20.04.1 LTS
Use the unbound.conf as suggested here https://unbound.docs.nlnetlabs.nl/en/latest/getting-started/configuration.html
Append a verbosity: -1 option to the unbound.conf
sudo useradd unbound
sudo chown -R unbound /usr/local/etc/unbound
sudo ldconfig
sudo -u unbound unbound-anchor
sudo -u unbound unbound-control-setup
Enable remote-control in unbound.conf
sudo systemctl stop systemd-resolved
sudo unbound -v
unbound-control status
Expected behavior
Unbound fails to run due to attempting to start with an invalid verbosity level configuration of -1.
System:
Unbound version: 1.16.0
OS: Ubuntu 20.04.1 LTS
unbound -V output:
root@ubuntu-s-1vcpu-1gb-ams3-01:~# unbound -V
Version 1.16.0
Configure line:
Linked libs: mini-event internal (it uses select), OpenSSL 1.1.1f 31 Mar 2020
Linked modules: dns64 respip validator iterator
BSD licensed, see LICENSE in source package for details.
Report bugs to unbound-bugs@nlnetlabs.nl or https://github.com/NLnetLabs/unbound/issues
root@ubuntu-s-1vcpu-1gb-ams3-01:~#
Additional information
You can see that unbound-checkconf correctly identifies the invalid option of verbosity: -1 which if it is invalid should cause Unbound to not start when you attempt to execute Unbound but something is incrementing the verbosity statement at runtime.
root@ubuntu-s-1vcpu-1gb-ams3-01:~# sudo unbound -v
[1655978380] unbound[28024:0] notice: Start of unbound 1.16.0.
root@ubuntu-s-1vcpu-1gb-ams3-01:~# unbound-control status
version: 1.16.0
verbosity: 0
threads: 1
modules: 2 [ validator iterator ]
uptime: 9 seconds
options: reuseport control(ssl)
unbound (pid 28026) is running...
root@ubuntu-s-1vcpu-1gb-ams3-01:~# unbound-checkconf
[1655978446] unbound-checkconf[28028:0] fatal error: verbosity value < 0
root@ubuntu-s-1vcpu-1gb-ams3-01:~#
I haven't dug much more into this issue re: where in the code is causing this, but it is clearly a bug.
Hi @buggysolid, thanks for the in depth bug report. You're correct that the verbosity is incremented when unbound is started. This is due to the command you use to start unbound: unbound -v, which increments the config verbosity with 1. Unbound currently accepts verbosity levels under 0 and treats these as 0. Unless this causes any big errors, we don't see a reason to change this behaviour. I'll see if i can make the manpage a little clearer though!
Unbound currently accepts verbosity levels under 0 and treats these as 0. Unless this causes any big errors, we don't see a reason to change this behaviour.
It hasn't caused me any errors but I can think of a case in which someone runs unbound-checkconf before starting Unbound during a provisioning process to validate there config. If that utility borks AFAIK it will set a non-zero exit code.
I'll see if i can make the manpage a little clearer though!
That would be good, I wouldn't exactly describe it as a bug either but it wasn't clear from the documentation.
Closing this issue in the hope that the clarification in the manpage is enough
| gharchive/issue | 2022-06-23T10:02:50 | 2025-04-01T04:32:49.320991 | {
"authors": [
"TCY16",
"buggysolid"
],
"repo": "NLnetLabs/unbound",
"url": "https://github.com/NLnetLabs/unbound/issues/705",
"license": "bsd-3-clause",
"license_type": "permissive",
"license_source": "bigquery"
} |
1365529360 | [Q] Is there a way to resolve a list of domains on launch?
I'm looking to have unbound automatically resolve (and therefore cache) a list of domains (preferably via a file) whenever it launches. This way a list of frequently accessed {sub,}domains can be built and prefetched which would be great for whenever the server (or device hosting the server) restarts.
Is this possible to do at the moment?
There is the script, in the source tarball and code repository, contrib/warmup.sh that performs this task. There is also warmup.cmd for use on Windows.
Excellent! Just what I needed.
| gharchive/issue | 2022-09-08T04:55:22 | 2025-04-01T04:32:49.322893 | {
"authors": [
"txtsd",
"wcawijngaards"
],
"repo": "NLnetLabs/unbound",
"url": "https://github.com/NLnetLabs/unbound/issues/750",
"license": "bsd-3-clause",
"license_type": "permissive",
"license_source": "bigquery"
} |
1305075426 | Avoid treat normal hosts as unresponsive servers
This is a fix for issue #713
When infra-keep-probing is on, all hosts with expired entries were treated as
unresponsive servers and thus causing problems (see #713).
This commit change that, so that normal hosts with expired entries are treated
as unknown servers.
The locks fixed in commit https://github.com/NLnetLabs/unbound/commit/33bd49af8140e6744064a74eaf1c369ebed86610 . Thanks for the patch!
Thanks for merging it. Our internal patch is exact same as the final version.
I thought the lock changes just made the code nicer and they were identical. Do you mind explain a little bit about the reason the lock should be hold a bit longer? This is my first time look into Unbound code, so would like to learn more about that.
The host->rtt.rto that is used in the if statement in line 731 needs the lock on e. At line 699 it can seen that it is a part of e, e->data.
Ah, that make sense, I missed that. Thanks again!
| gharchive/pull-request | 2022-07-14T17:17:01 | 2025-04-01T04:32:49.325781 | {
"authors": [
"hunts",
"wcawijngaards"
],
"repo": "NLnetLabs/unbound",
"url": "https://github.com/NLnetLabs/unbound/pull/714",
"license": "bsd-3-clause",
"license_type": "permissive",
"license_source": "bigquery"
} |
276326141 | scat2gridgauss_zt always complains of irregular T axes
Reported by ansley b manke on 29 Mar 2012 18:24 UTC
Report from the Ferret Users List:
define axis/t=300:400:1 tax
define axis/z=0:300:1 zax
let chla_reg = scat2gridgauss_zt(dbar[i=1:5],date_new[i=1:5],chla[i=1:5],z[gz=zax],t[gt=tax],1.,1.,1.,0)
shade chla_reg
Bailing out of external function "scat2gridgauss_zt":
output axes must be REGULARLY spaced
**ERROR: error in external function
It turns out that the check for irregular destination axis is incorrect. There is no test for this function in the benchmark suite; tests for all directions of the gridding functions should be added.
Migrated-From: http://dunkel.pmel.noaa.gov/trac/ferret/ticket/1929
Comment by ansley.b.manke on 27 Apr 2012 19:33 UTC
The bug is fixed, and a test for all the scat2grid* functions in 2D combinations of xyzt is added to the benchmark suite.
| gharchive/issue | 2017-11-23T10:31:08 | 2025-04-01T04:32:49.405026 | {
"authors": [
"karlmsmith"
],
"repo": "NOAA-PMEL/Ferret",
"url": "https://github.com/NOAA-PMEL/Ferret/issues/1201",
"license": "Unlicense",
"license_type": "permissive",
"license_source": "github-api"
} |
276333734 | axis label sizes no longer controllable after "go magnify"
Reported by andrew wittenberg on 16 Aug 2012 00:31 UTC
With v6.82, after "go magnify", y-axis labels are automatically adjusted to an invisible size. There's also a cryptic note about magnify.jnl.
NOAA/PMEL TMAP
FERRET v6.82
Linux 2.6.32-279.1.1.el6.x86_64 64-bit - 08/03/12
15-Aug-12 20:22
yes? set view lr; go magnify; plot/vs {-1,1},{-1,1}
*** NOTE: Adjusting Y-axis label size from 0.100 to 0.001, to avoid running off page.
*** NOTE: /home/atw/ferret/exe/x86_64/ferret-6.8.2/go/magnify.jnl
Contrast this with v6.725:
NOAA/PMEL TMAP
FERRET v6.725
Linux 2.6.18-274.12.1.el5 64-bit - 04/10/12
15-Aug-12 20:27
yes? set view lr; go magnify; plot/vs {-1,1},{-1,1}
I'm attaching the resulting plots from each version.
Migrated-From: http://dunkel.pmel.noaa.gov/trac/ferret/ticket/1958
Comment by ansley.b.manke on 16 Aug 2012 19:29 UTC
This comes from our fix for ticket #1896, where the numeric axis labels are so long that they run the ylab label off the edge of the page. With GO MAGNIFY, you don't care about that but want to zoom in on a part of the plot. Orr even in general if the plot is in a viewport so that there's space for the labels then you'd probably want to leave them alone.
I'll think about this one. The issue came up in LAS where we don't know what the data values are going to be like - maybe a setting which would allow the new behavior when we do want it, but otherwise the previous behavior would still be in effect.
Comment by andrew.wittenberg on 16 Aug 2012 21:02 UTC
I'd suggest a "MODE SHRINK_LABELS_TO_FIT" or something like that, which would be unset by default to match the previous behavior.
I often generate viewports whose axis labels, by design, extend well beyond the viewport borders. An example is aligning the plot boundaries to a grid, using a series of viewport specifications with "go magnify". In this case there's actually zero room for the labels inside the viewport, but plenty of room in the margins surrounding the viewport.
Comment by andrew.wittenberg on 17 Aug 2012 17:46 UTC
Changing to severity to "blocker", since this causes labels to vanish in all of our custom-paneled plots.
Comment by ansley.b.manke on 17 Aug 2012 22:38 UTC
This is fixed with a new MODE SHRINK_YLAB
! The lower plot has its vertical axis labels intact and visible (classic behavior)
! The upper one will have them shrink away to nothing (mode shrink_ylab behavior)
set view lr
go magnify
plot/vs {-1,1},{-1,1}
set mode shrink_ylab
set view ur
go magnify
plot/vs {-1,1},{-1,1}
Attachment from andrew.wittenberg on 16 Aug 2012 00:32 UTC
REPLACE THIS TEXT WITH UPLOADED FILE ./attachments/TRAC_1958_GIT_1230/plot_v6.725.gif
Attachment from andrew.wittenberg on 16 Aug 2012 00:32 UTC
REPLACE THIS TEXT WITH UPLOADED FILE ./attachments/TRAC_1958_GIT_1230/plot_v6.82.gif
Attachment from ansley.b.manke on 17 Aug 2012 22:38 UTC
REPLACE THIS TEXT WITH UPLOADED FILE ./attachments/TRAC_1958_GIT_1230/fix_1958.gif
| gharchive/issue | 2017-11-23T10:57:01 | 2025-04-01T04:32:49.412506 | {
"authors": [
"karlmsmith"
],
"repo": "NOAA-PMEL/Ferret",
"url": "https://github.com/NOAA-PMEL/Ferret/issues/1230",
"license": "Unlicense",
"license_type": "permissive",
"license_source": "github-api"
} |
2475952487 | QOS tables are not easy to find on the new documentation
Example: https://oriondocs.rdhpcs.noaa.gov/wiki/index.php?title=Running_Jobs_on_MSU-HPC#Specifying_a_Partition
or
https://rdhpcs-common-docs.rdhpcs.noaa.gov/wiki/index.php?title=Running_and_Monitoring_Jobs#Specifying_a_Partition
In the old docs it's easy to find the tables and I can't find them in the new docs.
There are two issues here that make discovery of this information difficult:
The table is in a strange location — it is on the policy page. I'm not sure why it was placed there other than the table was in the common docs and not on the site-specific pages on the wikis.
2 The sidebar menu has the text "Specifying a Quality of Service (QOS)" and "Changing QOS's." With this text in the menu, searching for "quality of service" and "QOS" will show every page as a valid, good result.
We are still learning to optimize the search (SEO) and will apply items as we understand them. We can experiment with the data-nosnippet HTML attribute to hide the menus from the Google search snippet.
However, both of the above will only do so much. The actual resolution, which should have also been done this way on the wikis, is to replace the host-specific content in the "common" areas with instructions to the user on how to get the information on the systems and place host-specific information on the host pages.
In this example, the Slurm page should indicate how a user can discover the valid QOSes on the systems (sacctmgr show qos), and then on the hera and jet pages list the QOSes with more specific information This helps both the users who search, and others who are reading the pages.
Hi Seth,
Thank you for the quick reply. I agree the placement is not correct and
that needs to be updated. I've copied Forrest, David, and Raghu so we can
bring more people into this discussion as to where to place within the new
document context.
The sidebar is nice, but we will need users to be able to find this via
'search' as that's what most users do out of the gate.
*Renn Valo *NOAA Affiliate
NOAA Systems Administrator – HPC
RIVA/Cherokee
1(508)-495-2303
On Tue, Aug 20, 2024 at 12:39 PM Seth Underwood @.***>
wrote:
There are two issues here that make discovery of this information
difficult:
The table is in a strange location — it is on the policy page. I'm
not sure why it was placed there other than the table was in the common
docs and not on the site-specific pages on the wikis.
2 The sidebar menu has the text "Specifying a Quality of Service
(QOS)" and "Changing QOS's." With this text in the menu, searching for
"quality of service" and "QOS" will show every page as a valid, good result.
We are still learning to optimize the search (SEO) and will apply items as
we understand them. We can experiment with the data-nosnippet HTML
attribute
https://developers.google.com/search/docs/crawling-indexing/robots-meta-tag#data-nosnippet-attr
to hide the menus from the Google search snippet.
However, both of the above will only do so much. The actual resolution,
which should have also been done this way on the wikis, is to replace the
host-specific content in the "common" areas with instructions to the user
on how to get the information on the systems and place host-specific
information on the host pages.
In this example, the Slurm page should indicate how a user can discover
the valid QOSes on the systems (sacctmgr show qos), and then on the hera
and jet pages list the QOSes with more specific information This helps both
the users who search, and others who are reading the pages.
—
Reply to this email directly, view it on GitHub
https://github.com/NOAA-RDHPCS/noaa-rdhpcs.github.io/issues/183#issuecomment-2299504011,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/A7YZFEZQTIFEHG5VKX2EJGDZSOEMPAVCNFSM6AAAAABM2FMBBOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDEOJZGUYDIMBRGE
.
You are receiving this because you authored the thread.Message ID:
@.***>
Forrest and User Support team request this information be transferred to the Slurm pages, and Partition information as we
PR#195 and PR#196 will update the Hera and Jet user guides to include the QOS tables, as Seth instructed.
Closing issue, Slurm information has been collected under the slurm/ location. The Jet QOS table that was (incorrectly) part of common-docs is available in the Jet area at https://docs.rdhpcs.noaa.gov/systems/jet_user_guide.html#jet-partitions
| gharchive/issue | 2024-08-20T15:34:04 | 2025-04-01T04:32:49.425265 | {
"authors": [
"RVNOAA",
"chanwilson",
"elisabethpatterson",
"underwoo"
],
"repo": "NOAA-RDHPCS/noaa-rdhpcs.github.io",
"url": "https://github.com/NOAA-RDHPCS/noaa-rdhpcs.github.io/issues/183",
"license": "CC0-1.0",
"license_type": "permissive",
"license_source": "github-api"
} |
1553677800 | Adjust check's output behavior
Adjust's adam check to print the output of its check command even if it passes. I had to restructure things a little bit to prevent me from duplicating a lot of code in two places
why's the handling for stderr and stdout different?
| gharchive/pull-request | 2023-01-23T19:50:18 | 2025-04-01T04:32:49.429763 | {
"authors": [
"imlazyeye",
"sanbox-irl"
],
"repo": "NPC-Studio/adam",
"url": "https://github.com/NPC-Studio/adam/pull/10",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
1081514597 | Deprecation: csv as input for sampling and inference
STAC items are coming GDL's way and it's time to consider deprecating csv as input in favor of a STAC item, json-like, input. Jsons may not be as friendly if a user needs to create a custom list of images to input, but a simple utility could be used to convert a csv to the json expected as input.
french or English the attribute field?
This is just an example. The "attribute field" column (and all other columns actually) would accept any utf-8 character.
french or English the attribute field?
This is just an example. It could be any string
I think it's a fair assumption to "define an attribute name and value to use to filter all ground truth data in the dataset config". W would then remove the attribute field and the attribute values from the csv.
| gharchive/issue | 2021-12-15T21:07:23 | 2025-04-01T04:32:49.446277 | {
"authors": [
"mpelchat04",
"remtav"
],
"repo": "NRCan/geo-deep-learning",
"url": "https://github.com/NRCan/geo-deep-learning/issues/221",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
1008370386 | Combine cases
This is the completed Combined Cases feature as migrated from the macro. An example usage would be to:
Open new project
Add a default PVWatts - Single Owner case
Add a default Parabolic Trough - Empirical - Single Owner case
Add a Generic System - Single Owner case
Select: Generic System -> Power Plant -> Generate production profiles from open cases -> Select cases...
This has been tested to combine all technology and financial models, and works with both the Generic System and the Generic System-Battery models.
Help documentation has been sent to Paul.
@sjanzou There is one problem when using this with the Power Tower Molten Salt model. The progress bar showing the simulation of the flux maps fails to close on its own and requires the user to press the Close button. I think this has to do with the last call in Simulation::Invoke(), but I wasn't able to rectify it. Do you have any thoughts here? Thanks for reviewing.
Increased height of degradation input
Power Tower - simulating flux maps per comment
Only remaining power tower message (which should probably remain):
These changes are in pull request https://github.com/Matthew-Boyd/SAM/pull/4
Please update this pull request after you have a chance to review my pull request - then go ahead and merge - looks great!
@sjanzou Thank you for the review. I think you may not have seen the latest code, however. For example, I changed the wxSpinCtrlDouble that was too short to a AFSchedNumeric . Maybe try pulling again? Regardless, the new AFSchedNumeric control appears to be the same height as the other same controls in SAM.
Regarding the fix for the Power Tower, adding 'true' to bcsim.Invoke(true) suppresses all of the progress bars, which can make it appear that SAM hangs. I think making the user press 'OK' just for the power tower model is a better tradeoff? So I'm thinking not add the 'true' argument.
@sjanzou Thank you for the review. I think you may not have seen the latest code, however. For example, I changed the wxSpinCtrlDouble that was too short to a AFSchedNumeric . Maybe try pulling again? Regardless, the new AFSchedNumeric control appears to be the same height as the other same controls in SAM.
Regarding the fix for the Power Tower, adding 'true' to bcsim.Invoke(true) suppresses all of the progress bars, which can make it appear that SAM hangs. I think making the user press 'OK' just for the power tower model is a better tradeoff? So I'm thinking not add the 'true' argument.
@Matthew-Boyd , sounds good... I wanted to provide you with the options in SAM and adding the wxBusyCursor to, at least, show that SAM was potentially not hanging. The pull request is good and approved by me ;-)
Thanks @sjanzou !
| gharchive/pull-request | 2021-09-27T16:47:15 | 2025-04-01T04:32:49.485134 | {
"authors": [
"Matthew-Boyd",
"sjanzou"
],
"repo": "NREL/SAM",
"url": "https://github.com/NREL/SAM/pull/706",
"license": "BSD-3-Clause",
"license_type": "permissive",
"license_source": "github-api"
} |
658510500 | Does CSVDiagnosticOutput only handle 1D arrays?
Does the CSVDiagnosticOutput class in diagnostics only handle 1D arrays? (It appears so.)
If so, should it be rewritten to handle 1D or 2D numpy arrays?
The CSVDiagnosticOutput class can be used for 0D data (like a "trace" of a value vs time), or for 1D data vs time. As written, it can't output anything higher, e.g., 2D arrays vs time because it uses the numpy savetxt function which is limited to 2D data.
I like the idea of allowing higher dimensional output, but maybe csv is not a good format for that.
Agree CSV is probably not the right format for higher dimension data.
| gharchive/issue | 2020-07-16T19:49:33 | 2025-04-01T04:32:49.500215 | {
"authors": [
"arichar6",
"padamson"
],
"repo": "NRL-Plasma-Physics-Division/turbopy",
"url": "https://github.com/NRL-Plasma-Physics-Division/turbopy/issues/23",
"license": "CC0-1.0",
"license_type": "permissive",
"license_source": "github-api"
} |
54209746 | DOC: Add sphinx documentation.
This adds sphinx documentation based on the scikit-xray documention. I drafted an introduction to databroker. It can be expanded and improved, but it's a start. There is no other substantive content in this PR.
I notice that ophyd is documented in NSLS-II/docs, and I think this should be too. That means we have two docs to maintian: /docs and /scikit-xray/docs. We can split up /docs later if needed, but that seems like a far-future problem. Closing this for now.
| gharchive/pull-request | 2015-01-13T15:52:47 | 2025-04-01T04:32:49.517012 | {
"authors": [
"danielballan"
],
"repo": "NSLS-II/dataportal",
"url": "https://github.com/NSLS-II/dataportal/pull/4",
"license": "bsd-3-clause",
"license_type": "permissive",
"license_source": "bigquery"
} |
247700921 | Default SSLSessionCache path not changeable
At least, seen in Centos7. Default path for ssl session cache is hardcoded in ssl.erb. While that path is valid for some mod_ssl installation as a default it seems to have been changed in Centos7 at least.
[root@apacheserver1 ~]# rpm -ql mod_ssl.x86_64
/etc/httpd/conf.d/ssl.conf
/etc/httpd/conf.modules.d/00-ssl.conf
/usr/lib64/httpd/modules/mod_ssl.so
/usr/libexec/httpd-ssl-pass-dialog
/var/cache/httpd/ssl
Fixed in PR97
| gharchive/issue | 2017-08-03T13:09:04 | 2025-04-01T04:32:49.531153 | {
"authors": [
"jordiprats",
"kusw3"
],
"repo": "NTTCom-MS/eyp-apache",
"url": "https://github.com/NTTCom-MS/eyp-apache/issues/96",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
2692967590 | [NUsight] Move to ES modules
This basically moves away from JS scripts and defaults all JS files as proper ES modules.
This deprecates the require() import syntax, which unfortunately protobufjs still needs a shim to function (although I hear we are planning to move to Protobuf ES!).
Also upgrades us to minimum node version of 20.
https://nodejs.org/api/packages.html#type
https://blog.logrocket.com/es-modules-in-node-today/
This is gonna be great to have. Last I tried was about a year ago and many of our dependencies were incompatible with an ESM-first approach then.
| gharchive/pull-request | 2024-11-26T03:27:30 | 2025-04-01T04:32:49.537326 | {
"authors": [
"BrendanAnnable",
"JosephusPaye"
],
"repo": "NUbots/NUbots",
"url": "https://github.com/NUbots/NUbots/pull/1502",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
2620852 | save HTML edits?
Can HTML edits also be saved?
No. Not yet. https://bugs.webkit.org/show_bug.cgi?id=74755
| gharchive/issue | 2011-12-21T02:04:07 | 2025-04-01T04:32:49.538454 | {
"authors": [
"NV",
"fabritw"
],
"repo": "NV/chrome-devtools-autosave",
"url": "https://github.com/NV/chrome-devtools-autosave/issues/11",
"license": "mit",
"license_type": "permissive",
"license_source": "bigquery"
} |
1753245713 | Update to reflect new default benchmark settings
It seems that this commit https://github.com/NVIDIA-Merlin/HierarchicalKV/commit/921e9b82d231ad2d62aa67a1f0458af26876ffd5 updates the benchmark numbers but not the settings for the new largest KV.
i.e. dim = 64, capacity = 1024 Million-KV, HBM = 56 GB, HMEM = 200 GB in https://github.com/NVIDIA-Merlin/HierarchicalKV/commit/921e9b82d231ad2d62aa67a1f0458af26876ffd5#diff-b335630551682c19a781afebcf4d07bf978fb1f8ac04c6bf87428ed5106870f5L159
should now be dim = 64, capacity = 512 Million-KV, HBM = 32 GB, HMEM = 96 GB
My numbers with the new benchmark defaults:
## Benchmark
* GPU: 1 x NVIDIA A100 80GB PCIe: 8.0
* Key Type = uint64_t
* Value Type = float32 * {dim}
* Key-Values per OP = 1048576
* Evict strategy: LRU
* `λ`: load factor
* `find*` means the `find` API that directly returns the addresses of values.
* `find_or_insert*` means the `find_or_insert` API that directly returns the addresses of values.
* ***Throughput Unit: Billion-KV/second***
### On pure HBM mode:
* dim = 4, capacity = 64 Million-KV, HBM = 32 GB, HMEM = 0 GB
| λ | insert_or_assign | find | find_or_insert | assign | find* | find_or_insert* | insert_and_evict |
|-----:|-----------------:|-------:|---------------:|-------:|-------:|----------------:|-----------------:|
| 0.50 | 1.169 | 3.019 | 1.837 | 1.974 | 4.469 | 1.843 | 1.033 |
| 0.75 | 1.019 | 2.981 | 0.681 | 0.888 | 1.980 | 1.288 | 0.899 |
| 1.00 | 0.370 | 3.031 | 0.372 | 0.507 | 0.942 | 0.393 | 0.324 |
* dim = 64, capacity = 64 Million-KV, HBM = 16 GB, HMEM = 0 GB
| λ | insert_or_assign | find | find_or_insert | assign | find* | find_or_insert* | insert_and_evict |
|-----:|-----------------:|-------:|---------------:|-------:|-------:|----------------:|-----------------:|
| 0.50 | 0.863 | 2.132 | 0.926 | 1.116 | 4.483 | 1.841 | 0.771 |
| 0.75 | 0.672 | 2.119 | 0.571 | 0.792 | 1.989 | 1.292 | 0.579 |
| 1.00 | 0.324 | 2.179 | 0.324 | 0.470 | 0.931 | 0.393 | 0.232 |
### On HBM+HMEM hybrid mode:
* dim = 64, capacity = 128 Million-KV, HBM = 16 GB, HMEM = 16 GB
| λ | insert_or_assign | find | find_or_insert | assign | find* | find_or_insert* |
|-----:|-----------------:|-------:|---------------:|-------:|-------:|----------------:|
| 0.50 | 0.094 | 0.129 | 0.100 | 0.131 | 4.094 | 1.803 |
| 0.75 | 0.092 | 0.128 | 0.098 | 0.128 | 1.925 | 1.141 |
| 1.00 | 0.077 | 0.115 | 0.080 | 0.105 | 0.930 | 0.392 |
* dim = 64, capacity = 512 Million-KV, HBM = 32 GB, HMEM = 96 GB
| λ | insert_or_assign | find | find_or_insert | assign | find* | find_or_insert* |
|-----:|-----------------:|-------:|---------------:|-------:|-------:|----------------:|
| 0.50 | 0.048 | 0.067 | 0.048 | 0.065 | 3.551 | 1.423 |
| 0.75 | 0.047 | 0.068 | 0.047 | 0.064 | 1.872 | 1.257 |
| 1.00 | 0.043 | 0.067 | 0.043 | 0.060 | 0.916 | 0.391 |
/blossom-ci
Hi @Quentin-Anthony, thank you for taking the time to contribute and improve our project!
| gharchive/pull-request | 2023-06-12T17:46:01 | 2025-04-01T04:32:49.541992 | {
"authors": [
"Quentin-Anthony",
"rhdong"
],
"repo": "NVIDIA-Merlin/HierarchicalKV",
"url": "https://github.com/NVIDIA-Merlin/HierarchicalKV/pull/142",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
1405097788 | How would I pad inputs during batched inference?
Hi,
I am running GPT-J with FasterTransformer with triton (using this backend). Triton supports dynamic batching, so if I want to batch inputs, how would I pad inputs so that they are the same length? Ideally, we would pad inputs from the right and create an attention mask to give info to the model saying the first n tokens are pad tokens. Is there a way of doing so with FasterTransformer?
You only need to pad inputs to same length and provide the real sequence length like
input:
[ [a, b, x],
[a, x, x],
[a, b, c]
]
seq_len:
[ 2, 1, 3]
Hi @byshiue, thanks for your quick reply! I tried padding inputs like following:
original input: [27, 91, 29412, 91, 29, 198, 2, 8818, 284, 3440, 257, 2298, 293, 2393, 198, 4299, 3440]
padded input (to length of 50): [27, 91, 29412, 91, 29, 198, 2, 8818, 284, 3440, 257, 2298, 293, 2393, 198, 4299, 3440, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256]
input_len: 17
Here, 50256 is the pad token. Since I am using this backend, I pass seq_len as input_lengths tensor as shown here and padded input as input_ids. However, doing so gives significantly worse results than one without padding at the end. Am I doing it correctly?
Do you pass sequence length as 17 (length of original input)?
yes, I set the input_lengths tensor to 17
Can you post the results of
original input + output
padding input + output
decoded input (same input ids as above): <|python|>\n#function to load a pickle file\ndef load
Original input + output
[ 27 91 29412 91 29 198 2 8818 284 3440 257 2298 293 2393 198 4299 3440 62 27729 293 7 34345 2599 198 50284 4480 1280 7 34345 11 705 26145 11537 355 277 25 198 50280 7783 2298 293 13 2220 7 69 8 198]
decoded_output: _pickle(filename):\n with open(filename, 'rb') as f:\n return pickle.load(f)\n
2. Padding input + output
[ 27 91 29412 91 29 198 2 8818 284 3440 257 2298 293 2393 198 4299 3440 62 796 2298 293 13 2220 7 9654 7203 27729 293 14 27729 293 13 79 41582 1600 366 26145 48774 198 198 2 8818 284 3613 257 2298 293 50256 50256 50256 50256 50256 50256 50256 50256 50256 50256 50256 50256 50256 50256 50256 50256 50256 50256 50256 50256 50256 50256 50256 50256 50256 50256 50256 50256 50256 50256 50256 50256 50256]
decoded_output: '_ = pickle.load(open("pickle/pickle.pkl", "rb"))\n\n#function to save a pickle'
Let me know if you have any further questions!
I use your inputs to test on GPT-J model and cannot reproduce your issue.
The results of my side:
topk 1, topp 0
[[[ 27 91 29412 91 29 198 2 8818 284 3440 257
2298 293 2393 198 4299 3440 62 27729 293 7 34345
2599 198 220 220 220 1303 9654 262 2393 198 220
220 220 277 796 1280 7 34345 11 705 26145 11537
198 220 220]]]
[[[ 27 91 29412 91 29 198 2 8818 284 3440 257
2298 293 2393 198 4299 3440 62 27729 293 7 34345
2599 198 220 220 220 1303 9654 262 2393 198 220
220 220 277 796 1280 7 34345 11 705 26145 11537
198 220 220 50256 50256 50256 50256 50256 50256 50256 50256
50256 50256 50256 50256 50256 50256 50256 50256 50256 50256 50256
50256 50256 50256 50256 50256 50256 50256 50256 50256 50256 50256
50256 50256 50256]]]
topk 0, topp 0.75
[[[ 27 91 29412 91 29 198 2 8818 284 3440 257
2298 293 2393 198 4299 3440 62 27729 293 7 34345
2599 198 220 220 220 277 796 1280 7 34345 11
705 26145 11537 198 220 220 220 1366 796 2298 293
13 2220 7]]]
[[[ 27 91 29412 91 29 198 2 8818 284 3440 257
2298 293 2393 198 4299 3440 62 27729 293 7 34345
2599 198 220 220 220 277 796 1280 7 34345 11
705 26145 11537 198 220 220 220 1366 796 2298 293
13 2220 7 50256 50256 50256 50256 50256 50256 50256 50256
50256 50256 50256 50256 50256 50256 50256 50256 50256 50256 50256
50256 50256 50256 50256 50256 50256 50256 50256 50256 50256 50256
50256 50256 50256]]]
after setting the FIXED_START_IDS of tools/gpt/identity_test.py to
[27, 91, 29412, 91, 29, 198, 2, 8818, 284, 3440, 257, 2298, 293, 2393, 198, 4299, 3440]
and
[27, 91, 29412, 91, 29, 198, 2, 8818, 284, 3440, 257, 2298, 293, 2393, 198, 4299, 3440, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256, 50256]
(Note that input_len are fixed to 17 for both cases)
Do you use the latest codes of main branch? If yes, please provide the reproduce steps to reproduce your issue.
Close this bug because it is inactivated. Feel free to re-open this bug if you still have any problem.
| gharchive/issue | 2022-10-11T19:00:12 | 2025-04-01T04:32:49.562280 | {
"authors": [
"byshiue",
"sdpmas"
],
"repo": "NVIDIA/FasterTransformer",
"url": "https://github.com/NVIDIA/FasterTransformer/issues/344",
"license": "apache-2.0",
"license_type": "permissive",
"license_source": "bigquery"
} |
2081278901 | Pallas images don't have triton/jax-triton installed
Running the following seems to show that the pallas image doesn't actually include the dependencies. Is something additional required to use this image?
docker run -i --runtime=nvidia --gpus all -t ghcr.io/nvidia/jax:nightly-pallas-2024-01-13
...
$ pip freeze
absl-py==2.0.0
annotated-types==0.6.0
build==1.0.3
chex==0.1.85
click==8.1.7
etils==1.6.0
-e git+https://github.com/google/flax.git@d58e6dde1b659017e7fb8af5989c22e2e86b9f00#egg=flax
fsspec==2023.12.2
importlib-resources==6.1.1
-e git+https://github.com/google/jax.git@b8b119d9b9df4d4d3ca5522a2063e90dcffccb34#egg=jax
jaxlib @ file:///opt/jax/dist/jaxlib-0.4.24.dev20240113-cp310-cp310-manylinux2014_x86_64.whl#sha256=81e26e1f2f4608a9d6f14d823965bd2cdd10d42e2ebbe240b90f20cb23c352e3
markdown-it-py==3.0.0
mdurl==0.1.2
ml-dtypes==0.3.2
msgpack==1.0.7
nest-asyncio==1.5.8
numpy==1.26.3
opt-einsum==3.3.0
optax==0.1.7
orbax-checkpoint==0.4.8
packaging==23.2
pip-tools==7.3.0
protobuf==4.25.2
pydantic==2.5.3
pydantic_core==2.14.6
Pygments==2.17.2
pyproject_hooks==1.0.0
PyYAML==6.0.1
rich==13.7.0
scipy==1.11.4
tensorstore==0.1.52
tomli==2.0.1
toolz==0.12.0
transformer-engine @ file:///opt/transformer-engine/dist/transformer_engine-1.3.0.dev0+daad219-cp310-cp310-linux_x86_64.whl#sha256=efd18bb195aa81ad093153c9410f9cabd68a5c7d9ec182252582c039c2ae6714
typing_extensions==4.9.0
zipp==3.17.0
Thanks for the report. You're right that there's a glitch here, which the CI test has managed not to flag.
Unfortunately this is hiding some other issues, including the one you reported at https://github.com/google/jax/discussions/19353#discussioncomment-8129678, which we'll need to find a solution for.
This specific issue was fixed in https://github.com/NVIDIA/JAX-Toolbox/pull/471, and the other issue I linked to above was fixed in https://github.com/google/jax/pull/19377.
Unfortunately there are still issues running the Pallas tests with the current HEADs (but it's possible that the current nightly Pallas wheels will work for some things).
| gharchive/issue | 2024-01-15T05:29:10 | 2025-04-01T04:32:49.570447 | {
"authors": [
"davisyoshida",
"olupton"
],
"repo": "NVIDIA/JAX-Toolbox",
"url": "https://github.com/NVIDIA/JAX-Toolbox/issues/470",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
678659816 | [BUG] getting one more parquet files than given out_files_per_proc
Describe the bug
I am working on criteo day_0.parquet file. I run the steps below:
...
proc = nvt.Workflow(
cat_names=['C01', 'C02', 'C09'],
cont_names=CONTINUOUS_COLUMNS,
label_name=LABEL_COLUMNS)
proc.add_cont_feature([ZeroFill(), LogOp()])
proc.add_cont_preprocess(Normalize())
# categorification with frequency thresholding
proc.add_cat_preprocess(Categorify(freq_threshold=8))
proc.finalize()
train_dataset = nvt.Dataset(train_paths, engine='parquet', part_mem_fraction=0.12)
valid_dataset = nvt.Dataset(valid_paths, engine='parquet', part_mem_fraction=0.12)
proc.apply(train_dataset, shuffle=nvt.io.Shuffle.PER_PARTITION, output_path=output_train_dir, out_files_per_proc=3)
This generates four parquet files, of which one is 4 bytes size in the output_path. Because of this file, I got Parquet file size is 4 bytes, smaller than the minimum file footer (8 bytes) error when training Keras model. I have to manually remove it to be able to run the model.
Expected behavior
Number of processed parquet files stored in the output_path should be equal to out_files_per_proc param value.
Environment details (please complete the following information):
Environment location: Conda
Method of NVTabular install: Pip install -e
@rjzamora for viz.
@rnyak - Where you able to confirm that this is still a problem for you with the latest main branch?
| gharchive/issue | 2020-08-13T18:48:07 | 2025-04-01T04:32:49.590892 | {
"authors": [
"rjzamora",
"rnyak"
],
"repo": "NVIDIA/NVTabular",
"url": "https://github.com/NVIDIA/NVTabular/issues/211",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
867041765 | [BUG] Total number of concatenated rows exceeds size_type range
Describe the bug
Hello, after seeing your RecSys2020 tutorial, I've attempted to use NVTabular for preprocessing the data in the RecSysChallenge 2021. However when i try to use fit on my workflow i get a Runtime Error: Total number of concatenated rows exceeds size_type range.
Steps/Code to reproduce bug
Notebook containing the code can be found here - https://github.com/RostislavStoyanov/recsys/blob/master/Preprocessing.ipynb
Expected behavior
I expect the fit to complete successfully.
Environment details (please complete the following information):
Environment location: Bare-metal
Method of NVTabular install: conda
Additional context
Stack trace
---------------------------------------------------------------------------
RuntimeError Traceback (most recent call last)
<timed exec> in <module>
~/anaconda3/envs/rapids-0.18/lib/python3.8/site-packages/nvtabular-0.5.0-py3.8.egg/nvtabular/workflow.py in fit(self, dataset)
140
141 if self.client:
--> 142 results = [r.result() for r in self.client.compute(stats)]
143 else:
144 results = dask.compute(stats, scheduler="synchronous")[0]
~/anaconda3/envs/rapids-0.18/lib/python3.8/site-packages/nvtabular-0.5.0-py3.8.egg/nvtabular/workflow.py in <listcomp>(.0)
140
141 if self.client:
--> 142 results = [r.result() for r in self.client.compute(stats)]
143 else:
144 results = dask.compute(stats, scheduler="synchronous")[0]
~/anaconda3/envs/rapids-0.18/lib/python3.8/site-packages/distributed/client.py in result(self, timeout)
218 if self.status == "error":
219 typ, exc, tb = result
--> 220 raise exc.with_traceback(tb)
221 elif self.status == "cancelled":
222 raise result
~/anaconda3/envs/rapids-0.18/lib/python3.8/contextlib.py in inner()
73 def inner(*args, **kwds):
74 with self._recreate_cm():
---> 75 return func(*args, **kwds)
76 return inner
77
~/anaconda3/envs/rapids-0.18/lib/python3.8/site-packages/nvtabular-0.5.0-py3.8.egg/nvtabular/ops/categorify.py in _mid_level_groupby()
~/anaconda3/envs/rapids-0.18/lib/python3.8/site-packages/cudf/core/dataframe.py in from_arrow()
5338 index_col = table.schema.pandas_metadata["index_columns"]
5339
-> 5340 out = super().from_arrow(table)
5341
5342 if index_col:
~/anaconda3/envs/rapids-0.18/lib/python3.8/contextlib.py in inner()
73 def inner(*args, **kwds):
74 with self._recreate_cm():
---> 75 return func(*args, **kwds)
76 return inner
77
~/anaconda3/envs/rapids-0.18/lib/python3.8/site-packages/cudf/core/frame.py in from_arrow()
2210 libcudf.table.Table()
2211 if data.num_columns == 0
-> 2212 else libcudf.interop.from_arrow(data, data.column_names)
2213 )
2214
cudf/_lib/interop.pyx in cudf._lib.interop.from_arrow()
RuntimeError: cuDF failure at: /opt/conda/envs/rapids/conda-bld/libcudf_1618503955512/work/cpp/src/copying/concatenate.cu:368: Total number of concatenated rows exceeds size_type range
Output in console
distributed.worker - WARNING - Compute Failed
Function: _mid_level_groupby
args: ([pyarrow.Table
tweet_id: string
tweet_id_count: int32, pyarrow.Table
tweet_id: string
tweet_id_count: int32, pyarrow.Table
tweet_id: string
tweet_id_count: int32, pyarrow.Table
tweet_id: string
tweet_id_count: int32, pyarrow.Table
tweet_id: string
tweet_id_count: int32, pyarrow.Table
tweet_id: string
tweet_id_count: int32, pyarrow.Table
tweet_id: string
tweet_id_count: int32, pyarrow.Table
tweet_id: string
tweet_id_count: int32, pyarrow.Table
tweet_id: string
tweet_id_count: int32, pyarrow.Table
tweet_id: string
tweet_id_count: int32, pyarrow.Table
tweet_id: string
tweet_id_count: int32, pyarrow.Table
tweet_id: string
tweet_id_count: int32, pyarrow.Table
tweet_id: string
tweet_id_count: int32, pyarrow.Table
tweet_id: string
tweet_id_count: int32, pyarrow.Table
tweet_id: string
tweet_id_count: int32, pyarrow.Table
tweet_id: string
tweet_id_count: int32, pyarrow.Table
tweet_id: string
tweet_id_count: int32, pyarrow.Table
tweet_id: string
tweet_id_count: int32, pyarrow.Table
tweet_id: st
kwargs: {}
Exception: RuntimeError('cuDF failure at: /opt/conda/envs/rapids/conda-bld/libcudf_1618503955512/work/cpp/src/copying/concatenate.cu:368: Total number of concatenated rows exceeds size_type range')
Hi @RostislavStoyanov - thanks for giving NVTabular a go!
I think the bug you're hitting here is because of an underlying restriction of cudf - each dataframe can only have a max of 2GB of data in a single string column.
Can you try reducing the partition sizes you're processing? You can do this by setting the part_size or part_mem_fraction parameters you pass to the nvtabular.Dataset object
Reducing the part_size from 1GB to 256MB had no effect. I will try further reducing that parameter and report here. Not sure if it helps, but i am also using a dask.distributed client connected to a LocalCudaCluster for the Workflow object and i also run client.run(cudf.set_allocator, "managed") on that client (I do this because of memory issues).
Hello @RostislavStoyanov ,
thanks for trying out - sorry for my delayed response, I was on vacation.
The BUG, you discovered, is related to Categorifying the tweet_id and userids. We use Categorify op to map the original IDs to a continuous Integer, as they require less bytes to represent and easier to process. Categorify creates a mapping table from ID->Continuous Integer. In the RecSys2021 dataset, we have ~200M unique IDs and creating the mapping table for so many unique Strings gets difficult.
How to fix that? As we want to represent the tweet_id/user_id for processing reason as Integer, we can hash instead of using Categorify operation.
gdf.hash_columns output is a int32 - we have 2**64 buckets, the probability of a collusion should be pretty small.
After the IDs are int32, they requires less space and are easier to process (Categorify Ops is may possible to apply afterwards on it)
I tested the first step of recsys2020 workflow with ~150 of 253 files and it worked for me. It took ~2-3min for .fit and 4-5min for .transform with 8 GPUs.
From
cat_features = (['tweet_id', 'language', 'engaging_user_id', 'engaged_with_user_id'] + fill_miss >> nvt.ops.Categorify())
To
cat_features = (['language',] + fill_miss >> nvt.ops.Categorify())
tweet_id = (
nvt.ColumnGroup("tweet_id") >>
(lambda col, gdf: gdf.hash_columns(['tweet_id']))
)
engaging_user_id = (
nvt.ColumnGroup("engaging_user_id") >>
(lambda col, gdf: gdf.hash_columns(['engaging_user_id']))
)
engaged_with_user_id = (
nvt.ColumnGroup("engaged_with_user_id") >>
(lambda col, gdf: gdf.hash_columns(['engaged_with_user_id']))
)
Hi @bschifferer,
Thank you for taking the time to help me. Unfortunately, I won't be able to test this until next week, but I will make sure to update this issue once I get back.
Also, considering this is happening, most likely, because of the cudf restriction, mentioned before, should this be considered a bug and have I made a mistake with labeling it as such?
| gharchive/issue | 2021-04-25T15:38:49 | 2025-04-01T04:32:49.601339 | {
"authors": [
"RostislavStoyanov",
"benfred",
"bschifferer"
],
"repo": "NVIDIA/NVTabular",
"url": "https://github.com/NVIDIA/NVTabular/issues/770",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
2117344225 | fix scheduler when not supplied
What does this PR do ?
when user does not supply learning rate scheduler, nemo aligner will crash. but now we use a dummy scheduler object
Just some small suggestion to make everything consistent.
In addition to the comments below, we should also remove the if self.scheduler is not None: in supervised.py
done!
| gharchive/pull-request | 2024-02-04T22:26:09 | 2025-04-01T04:32:49.603745 | {
"authors": [
"gshennvm"
],
"repo": "NVIDIA/NeMo-Aligner",
"url": "https://github.com/NVIDIA/NeMo-Aligner/pull/98",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
716543698 | some questions about how to use network
Hello, I am using HRnet-OCR as my own data set for ARCH training, and I do not find the seg and attn modules mentioned in your paper, are they in the deeper.DeeperS8 module?In addition, if you want the network to be able to achieve multi-scale training, should you modify the parameter scales?
您好，我使用HRnet-OCR作为我自己的ARCH训练数据集，但找不到您的论文中提到的SRG和ATTn模块，它们是否在drexdr8模块中？此外，如果您希望网络能够要实现多尺度训练，是否应该修改参数尺度？
| gharchive/issue | 2020-10-07T13:43:41 | 2025-04-01T04:32:49.770608 | {
"authors": [
"icey-zhang"
],
"repo": "NVIDIA/semantic-segmentation",
"url": "https://github.com/NVIDIA/semantic-segmentation/issues/77",
"license": "BSD-3-Clause",
"license_type": "permissive",
"license_source": "github-api"
} |
1216499275 | Dockerfile should derive from cudf Java Dockerfile
Since this project builds libcudf and libcudfjni, ideally the Dockerfile used for this project should derive from the Dockerfile used for the nightly cudf Java jar builds. Doing so would require publishing the cudf Java Docker image so it can be referenced in this repository's Dockerfile.
Where should we publish the cudf java docker? dockerhub? quay.io?
dockerhub/gpuci seems like the right place to publish the cudf Java jar builds.
cc: @raydouglass for visibility
follow-up to figure out a good way to support auto-trigger in cudf repo w/o cost too much
manually triggered branch-22.06 build at https://github.com/rapidsai/cudf/runs/6379507642?check_suite_focus=true
new image is ready at https://hub.docker.com/r/rapidsai/cudf-jni-build/tags
follow-up to figure out a good way to support auto-trigger in cudf repo w/o cost too much
I'm confused why this repository would track an issue with the cudf repository. The Dockerfile in question is in cudf, and ideally changes to that Dockerfile should trigger a rebuild and push of the new image to the Docker repository, possibly with a workflow action in the cudf repository. It feels odd to "fix" that problem in this repository, especially since it's not in the rapidsai domain.
Also fixing this particular issue, where we base the spark-rapids-jni Dockerfile on the published cudf Dockerfile, is somewhat orthogonal. We can get one working without the other, and thus they should be handled as separate issues.
Since the requirement of cudfjni build image of rapidsai is originally coming from this one, I would put the context here as a reminder. Potentially we could use a workaround like scan cudf commits in this github action of this repo, if found related dockerfile changes from upstream submodule, then notify us to do manual trigger.
This is more about we have less access of rapidsai repo. It would be not that easy for us to make changes or debugging stuff by leveraging their CICD resources in the future. And yes, I totally agree this should be a separate issue in cudf repo.
| gharchive/issue | 2022-04-26T21:31:41 | 2025-04-01T04:32:49.780507 | {
"authors": [
"GaryShen2008",
"jlowe",
"pxLi",
"sameerz"
],
"repo": "NVIDIA/spark-rapids-jni",
"url": "https://github.com/NVIDIA/spark-rapids-jni/issues/203",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
1977458165 | [auto-merge] branch-23.10 to branch-23.12 [skip ci] [bot]
auto-merge triggered by github actions on branch-23.10 to create a PR keeping branch-23.12 up-to-date. If this PR is unable to be merged due to conflicts, it will remain open until manually fix.
SUCCESS - auto-merge
| gharchive/pull-request | 2023-11-04T18:25:57 | 2025-04-01T04:32:49.782170 | {
"authors": [
"nvauto"
],
"repo": "NVIDIA/spark-rapids-ml",
"url": "https://github.com/NVIDIA/spark-rapids-ml/pull/502",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
1336517831 | [FEA] Add UI support to the Profiling AutoTuner
Is your feature request related to a problem? Please describe.
This is a followup to render the output of the auto-tuner as HTML.
Related Context
https://github.com/NVIDIA/spark-rapids/issues/6300
Added a patch for UI for the AutoTuner.
Patch:
Initial_Prototype_of_WebUI_for_AutoTuner.patch.zip
Sample WebUI:
| gharchive/issue | 2022-08-11T21:32:23 | 2025-04-01T04:32:49.785372 | {
"authors": [
"amahussein",
"parthosa"
],
"repo": "NVIDIA/spark-rapids",
"url": "https://github.com/NVIDIA/spark-rapids/issues/6304",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
1517111238 | [BUG] GDS cufile test failed in elder cuda runtime
Describe the bug
As we started build JNI in cuda 11.8 environment,
the GDS UT could fail w/ new JNI artifact, undefined symbol: cuFileDriverClose_v2
/usr/lib/jvm/java-8-openjdk-amd64/jre/bin/java: symbol lookup error: /raid1/jenkins/workspace/rapids_gds-nightly/plugin/tests/target/spark321/tmp/cufilejni2395913045181707900.so: undefined symbol: cuFileDriverClose_v2
when running in ENV w/ elder GDS (like libcufile 11-5). Upgrade local all nvidia-fs pkgs to cuda118 resolved this issue
cuFileDriverClose_v2 comparing to cuFileDriverClose
I found in https://github.com/NVIDIA/DALI/blob/main/dali/operators/reader/gds_mem_test.cu#L35-L38
they added some compatibility support for V1 and V2.
Not sure if we want to simply doc the requirement to run GDS or we add some compatible part cc @sameerz
Steps/Code to reproduce bug
in a machine w/ elder GDS lib (e.g. libcufile-11-5)
mvn -Dbuildver=321 clean package \
-DwildcardSuites=com.nvidia.spark.rapids.RapidsDeviceMemoryStoreSuite,com.nvidia.spark.rapids.RapidsGdsStoreSuite \
-Dtest.exclude.tags=""
closed by #7448. We could file another issue if we would like to provide compatibilty to GDS on elder runtime
| gharchive/issue | 2023-01-03T09:11:46 | 2025-04-01T04:32:49.788484 | {
"authors": [
"pxLi"
],
"repo": "NVIDIA/spark-rapids",
"url": "https://github.com/NVIDIA/spark-rapids/issues/7448",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
747712460 | Add in support for null type
This adds in null type support for almost all operators. The only operators that are missing is python UDF support and window operations. I didn't consider these critical right now. Window support is likely to work, but there have been issues with nulls in the past. I am not sure what the conversion to arrow would look like so I avoided it in the first take on this.
build
build
| gharchive/pull-request | 2020-11-20T18:35:16 | 2025-04-01T04:32:49.789746 | {
"authors": [
"revans2"
],
"repo": "NVIDIA/spark-rapids",
"url": "https://github.com/NVIDIA/spark-rapids/pull/1176",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
1087225173 | Exhaustively iterate ColumnarToRow iterator to avoid leaks
Fixes #4422. ColumnarToRowIterator will close its current batch via a task completion listener, but during unit tests there is no task context to leverage and thus the test must ensure the iterator is exhausted to ensure the batch is closed. In each of these cases the iterator was almost exhausted, but it was lacking a final hasNext to trigger the batch closing. Arguably this was a missed check in each of these tests anyway.
build
| gharchive/pull-request | 2021-12-22T22:15:10 | 2025-04-01T04:32:49.791006 | {
"authors": [
"jlowe"
],
"repo": "NVIDIA/spark-rapids",
"url": "https://github.com/NVIDIA/spark-rapids/pull/4428",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
1472998223 | [auto-merge] branch-22.12 to branch-23.02 [skip ci] [bot]
auto-merge triggered by github actions on branch-22.12 to create a PR keeping branch-23.02 up-to-date. If this PR is unable to be merged due to conflicts, it will remain open until manually fix.
SUCCESS - auto-merge
| gharchive/pull-request | 2022-12-02T14:56:44 | 2025-04-01T04:32:49.792317 | {
"authors": [
"nvauto"
],
"repo": "NVIDIA/spark-rapids",
"url": "https://github.com/NVIDIA/spark-rapids/pull/7224",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
2086455417 | texture format utility
There is a selection of formats but sometimes you want to know how many samples there are or the bit depth of the format. would by nice to have some utilities similar to tiny_imageformat?
https://github.com/DeanoC/tiny_imageformat/blob/master/include/tiny_imageformat/tinyimageformat.h
Hi! Thanks for your suggestion, it does make sense. There is const FormatProps& GetFormatProps(nri::Format format) (link) which is used internally and returns:
stride between pixels in bytes
block width in pixels
"is integer" flag
It can be extended to include:
NRI name as const char*
"is normalized" flag
"is float" flag
"is signed" flag
"is sRGB" flag
"is compressed" flag
channel count
This can be exposed. Which other data bits are you interested in? TIF is nice (at least in terms of functionality). If more is needed why not use TIF? Sample count requires querying... and that's the code, which I would prefer to not have.
Some device specific texture properties are exposed in DeviceDesc:
// Multi-sampling
NRI_NAME(Sample_t) colorSampleMaxNum;
NRI_NAME(Sample_t) depthSampleMaxNum;
NRI_NAME(Sample_t) stencilSampleMaxNum;
NRI_NAME(Sample_t) zeroAttachmentsSampleMaxNum;
NRI_NAME(Sample_t) textureColorSampleMaxNum;
NRI_NAME(Sample_t) textureIntegerSampleMaxNum;
NRI_NAME(Sample_t) textureDepthSampleMaxNum;
NRI_NAME(Sample_t) textureStencilSampleMaxNum;
NRI_NAME(Sample_t) storageTextureSampleMaxNum;
you have a good point I've kind of rolled out my own code. that is good to know, I'll have to think about it.
@pollend Hi! I have implemented (?) your request in v1.133, because @vertver also requested it, and I realized the usefulness. This info is available now:
NRI_STRUCT(FormatProps)
{
const char* name; // format name
NRI_NAME(Format) format; // self
uint8_t redBits; // R (or depth) bits
uint8_t greenBits; // G (or stencil) bits (0 if channels < 2)
uint8_t blueBits; // B bits (0 if channels < 3)
uint8_t alphaBits; // A (or shared exponent) bits (0 if channels < 4)
uint32_t stride : 6; // block size in bytes
uint32_t blockWidth : 4; // 1 for plain formats, >1 for compressed
uint32_t blockHeight : 4; // 1 for plain formats, >1 for compressed
uint32_t isBgr : 1; // reversed channels (RGBA => BGRA)
uint32_t isCompressed : 1; // block-compressed format
uint32_t isDepth : 1; // has depth component
uint32_t isExpShared : 1; // shared exponent in alpha channel
uint32_t isFloat : 1; // floating point
uint32_t isPacked : 1; // 16- or 32- bit packed
uint32_t isInteger : 1; // integer
uint32_t isNorm : 1; // [0; 1] normalized
uint32_t isSigned : 1; // signed
uint32_t isSrgb : 1; // sRGB
uint32_t isStencil : 1; // has stencil component
uint32_t unused : 7;
};
I could have fit into 16 bytes (currently 24), but the current layout is compatible with 64-bits per channel formats and mystical ASTC compression
let me know if something is missing
please, validate my mega-table, I could make a mistake ;)
| gharchive/issue | 2024-01-17T15:39:40 | 2025-04-01T04:32:49.805252 | {
"authors": [
"dzhdanNV",
"pollend"
],
"repo": "NVIDIAGameWorks/NRI",
"url": "https://github.com/NVIDIAGameWorks/NRI/issues/22",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
523805771 | nvcc error in Windows
Setup includes :
Cuda 10.1 and cudnn7.5
Pytorch 1.3.1
Hello,
This is likely an NVCC bug, which was supposed to be fixed in the latest version. Can you try loading the latest cuda compiler from https://developer.nvidia.com/cuda-10.1-download-archive-update2
This one worked fine for me with PyTorch 1.3.1.
Hi, thank you for using Kaolin! The library went through a major rewrite with the version 0.9.0, and this issue is no longer current. Please refer to our latest documentation:
https://kaolin.readthedocs.io/en/latest/
| gharchive/issue | 2019-11-16T07:06:34 | 2025-04-01T04:32:49.809571 | {
"authors": [
"Caenorst",
"arozantsev",
"ronakkaoshik42"
],
"repo": "NVIDIAGameWorks/kaolin",
"url": "https://github.com/NVIDIAGameWorks/kaolin/issues/12",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
1266700620 | Test Bug Fix
Status
[ ] Ready to go
Brief Description
Fixes a potential failure in the tests when these values are not defaulted.
Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.You have signed the CLA already but the status is still pending? Let us recheck it.
| gharchive/pull-request | 2022-06-09T21:43:42 | 2025-04-01T04:32:49.825701 | {
"authors": [
"CLAassistant",
"jwaldrop107"
],
"repo": "NWChemEx-Project/PluginPlay",
"url": "https://github.com/NWChemEx-Project/PluginPlay/pull/258",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
386714037 | Problem with local port
Pastebin is automatically putting a space between 127.0.0.1: and whatever port. Attached is a picture. When clicking check, it keeps saying wrong format. Its the wrong format because pastbin adds a space. Pastebin may take too much time to fix this. So the best solution is make the program read it with a space also.
I used that just now, and this is the error I got. That port is in my settings though.
Because 127 is not your real ip. Please check source code before posting anything.
Because 127 is not your real ip. Please check source code before posting anything.
it is my local ip, the same local Ip that everyone has when running it on themselves, like in your video. The error should be corrected so that it does not say it doesnt work. 127 should always work when testing on oneself.
| gharchive/issue | 2018-12-03T08:49:26 | 2025-04-01T04:32:49.832568 | {
"authors": [
"NYAN-x-CAT",
"anonymous11102"
],
"repo": "NYAN-x-CAT/Lime-RAT",
"url": "https://github.com/NYAN-x-CAT/Lime-RAT/issues/66",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
1191003653 | 运行前端时一直无法进入
所有的环境都是默认的。
win10，已运行host。
但在运行前端时，一直无法进入登录界面。
命令行有以下提示：
然后
可以试下前端官方的项目能不能跑起来
我看到的区别好像是你的vite版本有点高，下图是我运行的情况
| gharchive/issue | 2022-04-03T18:10:15 | 2025-04-01T04:32:49.885437 | {
"authors": [
"NameIsBad",
"seamusic"
],
"repo": "NameIsBad/abp-vue",
"url": "https://github.com/NameIsBad/abp-vue/issues/1",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
155736695 | Assistance with unit testing a module with a service locator
Prerequisites
[x] I have written a descriptive issue title
[x] I have verified that I am running the latest version of Nancy
[x] I have verified if the problem exist in both DEBUG and RELEASE mode
[x] I have searched open and closed issues to ensure it has not already been reported
Description
So I currently have a module that derives from a base module that is using a service locator.
The service locator needs the TinyIocContainer set and that gets set in ConfigureRequestContainer in the Bootstrap.
Here is a snippit of what I am talking about:
public abstract class BaseModule : NancyModule
{
protected ServiceLocator Locator => ServiceLocator.Instance;
protected ISettingsService<PlexRequestSettings> Settings => Locator.Resolve<ISettingsService<PlexRequestSettings>>();
protected string BaseUrl { get; set; }
protected BaseModule()
{
var settings = Settings.GetSettings();
var baseUrl = settings.BaseUrl;
BaseUrl = baseUrl;
var modulePath = string.IsNullOrEmpty(baseUrl) ? string.Empty : baseUrl;
ModulePath = modulePath;
}
}
So all of my modules are deriving from the BaseModule, but due to this I cannot seem to unit test it anymore and I am getting the following:
System.InvalidOperationException : Something went wrong when trying to satisfy one of the dependencies during composition, make sure that you've registered all new dependencies in the container and inspect the innerexception for more details.
----> Nancy.TinyIoc.TinyIoCResolutionException : Unable to resolve type: Nancy.NancyEngine
----> Nancy.TinyIoc.TinyIoCResolutionException : Unable to resolve type: Nancy.Routing.DefaultRequestDispatcher
----> Nancy.TinyIoc.TinyIoCResolutionException : Unable to resolve type: Nancy.Routing.DefaultRouteResolver
----> Nancy.TinyIoc.TinyIoCResolutionException : Unable to resolve type: Nancy.Routing.RouteCache
----> Nancy.TinyIoc.TinyIoCResolutionException : Unable to resolve type: PlexRequests.UI.Modules.ApiModule
----> System.NullReferenceException : Object reference not set to an instance of an object.
Here is an example of the Unit test i am referring to (Using NUnit):
[TestFixture]
public class ApiModuleTests
{
private ConfigurableBootstrapper Bootstrapper { get; set; }
[SetUp]
public void Setup()
{
var requestMock = new Mock<IRequestService>();
var settingsMock = new Mock<ISettingsService<PlexRequestSettings>>();
Bootstrapper = new ConfigurableBootstrapper(with =>
{
with.Module<ApiModule>();
with.Dependency(requestMock.Object);
with.Dependency(settingsMock.Object);
with.ApplicationStartup(
(c, a) =>
{
var loc = ServiceLocator.Instance; // Here I am trying to set the container for the service locator
loc.SetContainer(c);
});
});
}
[Test]
public void GetAllRequests()
{
var browser = new Browser(Bootstrapper);
var result = browser.Post("/api/requests", with =>
{
with.HttpRequest();
with.Header("Accept", "application/json");
with.Query("apikey","a");
});
Assert.That(HttpStatusCode.OK, Is.EqualTo(result.StatusCode));
}
Any ideas how I can do this?
Please do not use the ServiceLocator pattern, it's cancer for code. Nancy was specifically designed to be service locator kryptonite. If you need dependencies in your module, please take them as constructor parameters instead, If you are using the default bootstrapper then your dependencies will automatically be registered for you. if you are using another bootstrapper then you can override the ConfigureApplicationContainer or ConfigureRequestContainer methods and setup your dependencies
| gharchive/issue | 2016-05-19T13:28:42 | 2025-04-01T04:32:49.892804 | {
"authors": [
"thecodejunkie",
"tidusjar"
],
"repo": "NancyFx/Nancy",
"url": "https://github.com/NancyFx/Nancy/issues/2446",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
123491245 | A bug of WebStorm(maybe, all Intellij IDE) cause the watcher no effect.
I spend half a day to resolve a bug, the watcher.js no effect when I modify any file in app directory.
I thought this bug is an issue in the watcher.js, and I set console.log on the the watcher.js.
/**
* This is the watcher callback to verify the file actually changed
* @param dir
* @returns {Function}
*/
function getWatcher(dir) {
return function (event, fileName) {
console.log('event is: ' + event + " and file is " + fileName);
if (event === "rename") {
verifyWatches();
Then, run the watcher, and modify a file.
logout message:
event is: rename and file is main-page.xml___jb_bak___
event is: rename and file is main-page.xml
event is: rename and file is main-page.xml___jb_old___
A change event become to a rename event with wired tail.___jb_bak___
Then I search it in google.
This problem.
For conclusion.
If you use Webstorm or other IDE from jetbrains and want to watch files, Should close the safe write option.
Not a bug of LIveedit.
| gharchive/issue | 2015-12-22T14:34:29 | 2025-04-01T04:32:49.926853 | {
"authors": [
"gordonpro"
],
"repo": "NathanaelA/nativescript-liveedit",
"url": "https://github.com/NathanaelA/nativescript-liveedit/issues/11",
"license": "mit",
"license_type": "permissive",
"license_source": "bigquery"
} |
1833868529 | Text rendering issues
I don't know if this wasn't the case in some old Ghidra versions, but it's been like that for some time and I've confirmed both of the problems in the 10.3.2 version. I've got 3 mainly unrelated complaints.
The text in the Ghidra Help often gets corrupted like that:
Or that:
The effects can be temporarily removed by selecting the whole section where the error occurred or scrolling iirc.
For some reason Ghidra really struggles with the Korean symbols:
The setting "Render non-ASCII Unicode" is set to "all" in this image. It really is Korean...
There's more than that... ™, △, ＋, ～, 【 etc. were rendered as squares too the last time I saw them in Ghidra.
We also can see here that the labels for some reason don't support JP and KR hieroglyphs. I think there's even more they don't support knowing that the JP symbols are rendered fine. When I saw the addr pointers to these strings, I thought these two were of length 0.
Describe the solution you'd like
The symbols issue looks like an edge case that was never dealt with. I really hope it's not related to the Java GUI templates from the prehistoric age or things like that... It's a standard Unicode, we should be able to do something. The Help text corruption may be harder to fix, but I'm not a pro so I'm leaving this to the discretion of more experienced people.
Describe alternatives you've considered
There are no alternatives, we can just technically live without the proper text, but it's not gonna be very pleasant.
There are some options in the support/launch.properties that can modify how java does GUI stuff, and may address your first issue:
# The following options affect rendering on different platforms. It may be necessary to play
# with these settings to get Ghidra to display and perform optimally on HiDPI monitors or in VM's.
VMARGS=-Dsun.java2d.opengl=false
VMARGS_LINUX=-Dsun.java2d.pmoffscreen=false
VMARGS_LINUX=-Dsun.java2d.xrender=true
VMARGS_LINUX=-Dsun.java2d.uiScale=1
VMARGS_LINUX=-Dawt.useSystemAAFontSettings=on
VMARGS_WINDOWS=-Dsun.java2d.d3d=false
As far as the issue with displaying strings with characters from Korean, etc alphabets, that typically is controlled by your installed fonts. The bytes that I can see in your screenshot (f0, c5, b0, etc) display just fine as Korean glyphs, on my machine with the fonts I have.
That's my current font settings. Can you tell me yours? I really doubt this has to do something with my Windows fonts bank. Doesn't Ghidra bundle the necessary stuff with it?
No, Ghidra doesn't bundle any fonts. It relies on your operating system, and the java jdk's configuration.
Now that you specified windows, I can say that you do have a bit more hurdles than non-windows when it comes to how fonts are handled in a java program.
Typically a single font (eg. Arial), won't specify a glyph for every character in the unicode standard. When the font a program is using doesn't handle a specific character in a string, the drawing subsystem can fall-back to other fonts to provide a glyph for that character.
Java provides a jdk-level config file that specifies that information: the <jdk_dir>/lib/fontconfig.properties.
The jdk I'm using on my windows box has a fontconfig.properties.src file, which you have to rename to fontconfig.properties before it is used.
By default, it specifies, by name, several fall-back fonts for different alphabets (scripts), and for hangul it has:
monospaced.plain.korean=GulimChe
I don't have GulimChe installed, but I do have Malgun Gothic, so I changed my file to reflect that and all is good (for "monospaced" stuff in Ghidra).
If you did install the Gulim family of fonts, you probably wouldn't need to mess with this file.
Microsoft says that gulim is something they distribute.
I've enabled Direct3D and so far it looks fine. Continuing to monitor.
Now.... the fonts... Well, it seems that the Gulim font family is not installed in Windows 10 by default. Downloading the additional Korean fonts solved the hieroglyphs issue (no need to touch the files in <jdk_dir>!)
Not the label issue though...
...As for the other symbols, turns out I'm not very great at handling Unicode with Python2. I've fixed the script and the symbols are correct!
Sounds good. Feel free to close this issue if you think everything is resolved.
I would've done this by now if not for the one last part... Namely, the labels! I wrote about them in the 3rd paragraph.
Ghidra purposely restricts the contents of automatically generated labels to only include characters that are ascii and limits the length (typically 32 characters). (its not uncommon to find char[] buffers that aren't used as strings and are initialized with non-string data)
You can put any unicode string you want into a manually created label (except for the space ' ' character, and maybe a couple of others?)
I don't seem to have found any major issues with this since then. As for the labels, I've decided to make a Python script that would recreate the label at the current program address and use the whole string (if there is one and the cursor is at the first symbol, of course).
I'm closing this. Would it make sense to PR the script here in some way or not?
| gharchive/issue | 2023-08-02T20:33:26 | 2025-04-01T04:32:49.941656 | {
"authors": [
"Nemoumbra",
"dev747368",
"ryanmkurtz"
],
"repo": "NationalSecurityAgency/ghidra",
"url": "https://github.com/NationalSecurityAgency/ghidra/issues/5632",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
1289189129 | Assess whether warning icons should use alert role
assess whether or not the alert icons (such as insufficient project points) should use the alert role causing their rendering to interrupt any other text being read by a screenreader.
the warning icons already use the alert role, although the chrome plugin screen reader only seems to read them when the page is initially loaded
| gharchive/issue | 2022-06-29T19:21:54 | 2025-04-01T04:32:49.943200 | {
"authors": [
"mosefer",
"rmmayo"
],
"repo": "NationalSecurityAgency/skills-service",
"url": "https://github.com/NationalSecurityAgency/skills-service/issues/1442",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
241694860 | "getCurrentLocation" function is giving error in nativescript 2.5.0(IOS).
Hi,
I am working on a nativescript-Angular2 app. I'm trying to use nativescript-geolocation plugin to fetch the user's location. But when I am calling the "getCurrentLocation()" method, it is giving error, saying "New location is older than requested maximum age!"
Here is my Environment Info:
nativescript 2.5.0
tns-core-modules 2.4.4
tns-android 2.5.0
tns-ios 2.5.0
nativescript-geolocation 0.0.19
Here is my code sample.
ngOnInit() {
if (!isEnabled()) {
enableLocationRequest();
}
}
buttonTap() {
var location = getCurrentLocation({ desiredAccuracy: 3, updateDistance: 10, maximumAge: 20000, timeout: 20000 }).
then(function (loc) {
if (loc) {
console.log("User location is: " + loc);
}
}, function (e) {
console.log("Error: " + e.message);
});
}
Not sure this is an "unexpected" error. A device may have a cached location sensor reading, and if that location is returned with an age older than 20 seconds, you will get this error. Can also happen if the device is unable to update its location (offline, unable to use GPS, etc).
To remedy, try calling getCurrentLocation with a longer maximumAge...say 60000 milliseconds (1 min). Unless you need a super accurate, up-to-the-second location, extending maximumAge reduces the likelihood of getting this error.
Hi @anuragd7,
I would like to ask if the answer above solves your issue or it still exists?
Best regards,
Veselina
just remove all the options of the getCurrentLocation except desiredAccuracy and it should work fine
Due to no activity from the author and multiple solution provided in the comments, we consider this issue solved.
| gharchive/issue | 2017-07-10T12:25:57 | 2025-04-01T04:32:49.989181 | {
"authors": [
"anuragd7",
"deveix",
"radeva",
"toddanglin"
],
"repo": "NativeScript/nativescript-geolocation",
"url": "https://github.com/NativeScript/nativescript-geolocation/issues/66",
"license": "apache-2.0",
"license_type": "permissive",
"license_source": "bigquery"
} |
194052655 | Conflict with nativescript-facebook-login
I have nativescript-imagepicker installed, version 2.4.1, and nativescript-facebook-login, version 0.2.0. At first I login with Facebook, then I take the picture. Block of code doesn’t work:
.then(function(selection) {
console.log("Selection done:");
selection.forEach(function(selected) {
console.log(" - " + selected.uri);
});
}).catch(function (e) {
console.log(e);
});
I reviewed source code and found out that the handler onResult(args) for event:
application.android.on(application.AndroidApplication.activityResultEvent, onResult);
doesn’t work. That doesn't start.
hi @Denton09 i just read by chance your post. I had got your same problem and i solved it reported it to the author of the facebook login plugin.
Here is the link with the answer: https://github.com/AntonioCuevaUrraco/nativescript-facebook-login/issues/28
Maybe it's too late for you, but i hope the answer can help others.
Hi @Denton09,
Do you still have a problem with nativescript-imagepicker? As @mounirka suggested, most probably this issue has been related to nativescript-facebook-login plugin and there is nothing to do in imagepicker plugin.
Regards,
@tsonevn
Yes, thanks. This bug is in the nativescript-facebook-login. The issue is closed.
Regards,
@Denton09
| gharchive/issue | 2016-12-07T13:43:10 | 2025-04-01T04:32:49.993615 | {
"authors": [
"Denton09",
"mounirka",
"tsonevn"
],
"repo": "NativeScript/nativescript-imagepicker",
"url": "https://github.com/NativeScript/nativescript-imagepicker/issues/51",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
2504471967 | Add a license to the repository
I recently came across your project hamiltonian-evo-simulator-uts41173 and I noticed that the repository github.com/Navrin/hamiltonian-evo-simulator-uts41173 currently does not have a license specified.
Why Adding a License is Important?
As GitHub documents
Public repositories on GitHub are often used to share open source software. For your repository to truly be open source, you'll need to license it so that others are free to use, change, and distribute the software.
Would be possible to add a valid license to this repository? Probably something really OpenSource such as the Apache 2.0 license would fit really well.
Thanks!
Sure, I can add a license to this. I think I originally didn't include an open source license because it was for an assignment for my UTS class, but enough time has passed where it shouldn't matter anymore.
Thanks!
| gharchive/issue | 2024-09-04T06:59:25 | 2025-04-01T04:32:50.031729 | {
"authors": [
"Navrin",
"cbjuan"
],
"repo": "Navrin/hamiltonian-evo-simulator-uts41173",
"url": "https://github.com/Navrin/hamiltonian-evo-simulator-uts41173/issues/1",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
1821054949 | Containers: Allow containers to open ports for communication with com…
…pute nodes
This change allows for container workflows to open ports. These ports are opened on the host nodes (i.e. NNF nodes) where the containers are running. This enables traffic from outside of the network through the IP address of the NNF node and the port. An application on the compute node can contact the container with <NNF_NODE_IP>:.
The port number(s) can be retrieved via the NNF_CONTAINER_PORT environment variable. This environment variable is available inside of the containers. It is also provided to the Workflow so that Flux can inform the application on the compute node of which port(s) to use. If multiple ports are desired, the environment variable will provide a comma separated list of port numbers.
Ports are requested via the NnfContainerProfile's numPorts. A system admin must enable the Ports port range in the SystemConfiguration before ports can be requested. If not, the NnfPortManager will not allocate any ports.
More details:
Enabled default NnfPortManager to manage port allocation
Port allocation occurs in the Setup State
Port de-allocation occurs in the Teardown State
User Container Pods are now destroyed in the Teardown State prior to Port de-allocation
Added example-mpi-webserver NnfContainerProfile to show use of envionrment variable with a simple webserver
Added container teardown + port allocation to workflow deletion
I'm waiting on #215 and then I will rebase and refactor all the new error paths.
I've rebased on top of the new error reporting and accounted for the new errors. Please take another look, especially the latest commit.
| gharchive/pull-request | 2023-07-25T20:03:16 | 2025-04-01T04:32:50.054924 | {
"authors": [
"bdevcich-hpe"
],
"repo": "NearNodeFlash/nnf-sos",
"url": "https://github.com/NearNodeFlash/nnf-sos/pull/214",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
1112505623 | Cannot get response to requests on RPi 3B+ target device
I'm trying to test my branch and could not get response from the hm-diag application on a Raspberry Pi 3B+ test device. I'm aware that some hardware is missing on this device but the problem is worse because the app doesn't even respond to /version route.
The branch under test: https://github.com/NebraLtd/hm-diag/tree/murat/feat-add-short-git-sha
Here's the log I've collected:
testfleet_ursavaspi-24.01.22_14 06 22_(+0300).zip
Edit Note: I can't access this device even locally, therefore running necessary commands from the terminal of Balena dashboard with the following command on Host:
$ curl -s localhost/version
The same command works on my dev machine while debugging the app locally.
The problem is, diagnostics app is not working correctly if the VARIANT and FREQ variables cannot be found in environment variables. This can be easily avoided and reported in the config.
| gharchive/issue | 2022-01-24T11:13:03 | 2025-04-01T04:32:50.070348 | {
"authors": [
"MuratUrsavas"
],
"repo": "NebraLtd/hm-diag",
"url": "https://github.com/NebraLtd/hm-diag/issues/289",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
972528158 | Basic JSONRPC implementation
Do not merge, this is just for testing / evaluation.
This should ultimately be included in pyhelper so that there is a consistent way that we access gateway info across projects.
This is now being done in hm-pyhelper. Closed by: https://github.com/NebraLtd/hm-pyhelper/pull/8
| gharchive/pull-request | 2021-08-17T10:17:11 | 2025-04-01T04:32:50.071891 | {
"authors": [
"marvinmarnold",
"robputt"
],
"repo": "NebraLtd/hm-diag",
"url": "https://github.com/NebraLtd/hm-diag/pull/113",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
2436829713 | 🛑 Becpl Angular UI is down
In 928923a, Becpl Angular UI ($BECPL_ANGULAR_URL) was down:
HTTP code: 0
Response time: 0 ms
Resolved: Becpl Angular UI is back up in f4b209a after 16 minutes.
| gharchive/issue | 2024-07-30T03:58:08 | 2025-04-01T04:32:50.087586 | {
"authors": [
"NehalDamania"
],
"repo": "NehalDamania/becpl-uptime",
"url": "https://github.com/NehalDamania/becpl-uptime/issues/12124",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
1936745715 | 🛑 Becpl Angular UI is down
In 6f78406, Becpl Angular UI ($BECPL_ANGULAR_URL) was down:
HTTP code: 0
Response time: 0 ms
Resolved: Becpl Angular UI is back up in 0318075 after 23 minutes.
| gharchive/issue | 2023-10-11T03:55:37 | 2025-04-01T04:32:50.089681 | {
"authors": [
"NehalDamania"
],
"repo": "NehalDamania/becpl-uptime",
"url": "https://github.com/NehalDamania/becpl-uptime/issues/152",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
1988398017 | 🛑 Becpl Angular UI is down
In 6389f75, Becpl Angular UI ($BECPL_ANGULAR_URL) was down:
HTTP code: 0
Response time: 0 ms
Resolved: Becpl Angular UI is back up in 28692b3 after 7 minutes.
| gharchive/issue | 2023-11-10T21:06:19 | 2025-04-01T04:32:50.092288 | {
"authors": [
"NehalDamania"
],
"repo": "NehalDamania/becpl-uptime",
"url": "https://github.com/NehalDamania/becpl-uptime/issues/1721",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
1989814902 | 🛑 Becpl Angular UI is down
In 009c0c6, Becpl Angular UI ($BECPL_ANGULAR_URL) was down:
HTTP code: 0
Response time: 0 ms
Resolved: Becpl Angular UI is back up in 3280470 after 5 minutes.
| gharchive/issue | 2023-11-13T03:22:13 | 2025-04-01T04:32:50.094389 | {
"authors": [
"NehalDamania"
],
"repo": "NehalDamania/becpl-uptime",
"url": "https://github.com/NehalDamania/becpl-uptime/issues/1838",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
2113606736 | 🛑 Becpl Angular UI is down
In 852ee7e, Becpl Angular UI ($BECPL_ANGULAR_URL) was down:
HTTP code: 0
Response time: 0 ms
Resolved: Becpl Angular UI is back up in 55e3b12 after 17 minutes.
| gharchive/issue | 2024-02-01T22:21:58 | 2025-04-01T04:32:50.096508 | {
"authors": [
"NehalDamania"
],
"repo": "NehalDamania/becpl-uptime",
"url": "https://github.com/NehalDamania/becpl-uptime/issues/5615",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
154370757 | mapproject option
Recently I think the --tr option to mapproject went missing, and only --mpp and --ppd is available. That is confusing to me, as --tr is very useful, and is consistent with how point2dem uses it. It would be nice to get it back.
--tr has been re-added.
| gharchive/issue | 2016-05-12T00:15:16 | 2025-04-01T04:32:50.122728 | {
"authors": [
"ScottMcMichael",
"oleg-alexandrov"
],
"repo": "NeoGeographyToolkit/StereoPipeline",
"url": "https://github.com/NeoGeographyToolkit/StereoPipeline/issues/141",
"license": "apache-2.0",
"license_type": "permissive",
"license_source": "bigquery"
} |
107861080 | How to implement bi-directional RNN/LSTM with Neon?
How to implement bi-directional RNN/LSTM with Neon?
If you look at the RNN/LSTM implementation, you can see internal buffers to store activations for each time step of the sequence. You can follow the same design, but allocate separate buffers for forward passing the sequence and backward passing the sequence.
We are working on and planning to add that feature to neon in the near future. Stay tuned.
| gharchive/issue | 2015-09-23T06:52:47 | 2025-04-01T04:32:50.177411 | {
"authors": [
"buriy",
"yinyinl"
],
"repo": "NervanaSystems/neon",
"url": "https://github.com/NervanaSystems/neon/issues/102",
"license": "apache-2.0",
"license_type": "permissive",
"license_source": "bigquery"
} |
125427772 | x[:] = y proceeds with incompatible shapes
from neon.backends import gen_backend
ng = gen_backend(backend='gpu', device_id=0)
x = ng.zeros((2, 3))
y = ng.zeros((5, 7))
x[:] = y # no exceptions
x[:] = y + 0 # raises an error
Perhaps related to https://github.com/NervanaSystems/neon/issues/182
A fix for this will be part of the imminent v1.5.0 neon release.
| gharchive/issue | 2016-01-07T16:26:01 | 2025-04-01T04:32:50.178939 | {
"authors": [
"jennifermyers",
"oleg-trott"
],
"repo": "NervanaSystems/neon",
"url": "https://github.com/NervanaSystems/neon/issues/183",
"license": "apache-2.0",
"license_type": "permissive",
"license_source": "bigquery"
} |
169732966 | Grid LSTM in Neon
Hi,
Just wondering if Multi-dimensionnal Tensor for LSTM ( size= (100,4,6,3)) is available in Neon
(this is the input tensor) as a template and it can be done (multi-dimensionnal cell memory can cause issues).
Those ones are available in TensorFlow and Torch:
https://github.com/coreylynch/grid-lstm
https://github.com/tensorflow/tensorflow/issues/2560
Original papers:
https://arxiv.org/abs/1507.01526
Thanks
We currently don't support LSTM that takes multi-dimensional tensors like Grid LSTM. The various RNN layers in neon use this data layout.
Our backends supports multi-dimensional tensors that you can do numerical operations on. So if you decides to construct your own layer type, you can write out the forward and backward processes to handle those multi-dimensional tensors.
| gharchive/issue | 2016-08-06T06:37:37 | 2025-04-01T04:32:50.182071 | {
"authors": [
"deeplearning-ai-research",
"yinyinl"
],
"repo": "NervanaSystems/neon",
"url": "https://github.com/NervanaSystems/neon/issues/274",
"license": "apache-2.0",
"license_type": "permissive",
"license_source": "bigquery"
} |
1563941867 | [Bug]: When the source table is not configured with watermark, double write does not take effect
What happened?
If the configuration arctic.emit.mode is auto means flink writer operators will detect if the logstore is enabled automatically. If yes, the writer operators will double-write data into the logstore and filestore.
But if the writer operator didn't receive the watermark from the source operator(the source table didn't configure the watermark spec), the flink writers would not write records into the logstore. It does not meet the expectation.
Affects Versions
master
What engines are you seeing the problem on?
Flink
How to reproduce
create table arctic.flink_test.kafka_pk
(id int, name string, dt timestamp, primary key(id) not enforced)
with (
'log-store.enabled'='true',
'log-store.type'='kafka',
'log-store.topic'='arctic-testing',
'log-store.address'='*:*');
CREATE TABLE cdc_source(
id INT,
name STRING,
op_time STRING
) WITH (
'connector' = 'socket',
'hostname' = 'localhost',
'port' = '9999',
'format' = 'changelog-csv',
'changelog-csv.column-delimiter' = '|'
);
insert into arctic.flink_test.kafka_pk SELECT id,
name,
CAST(TO_TIMESTAMP(op_time) AS TIMESTAMP(6)) ts
FROM cdc_source;
Relevant log output
No response
Anything else
No response
Code of Conduct
[X] I agree to follow this project's Code of Conduct
As I know, the auto-writing to LogStore is dependent on the source watermark spec.
So how do you think the auto emit mode should behave when the source watermark spec is lack of?
@zhoujinsong If the watermark spec is lack of and Flink SQL Job does not configure the arctic.emit.auto-write-to-logstore.watermark-gap, I think the flink writer should execute the double-write operation.
| gharchive/issue | 2023-01-31T09:19:51 | 2025-04-01T04:32:50.210427 | {
"authors": [
"YesOrNo828",
"zhoujinsong"
],
"repo": "NetEase/arctic",
"url": "https://github.com/NetEase/arctic/issues/1036",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
185458107 | readme: Syntax highlight code snippets
This patch syntax highlights the JavaScript code snippets in the
README.
small nit but LGTM
| gharchive/pull-request | 2016-10-26T17:25:17 | 2025-04-01T04:32:50.226942 | {
"authors": [
"ejj",
"tejasmanohar"
],
"repo": "NetSys/quilt",
"url": "https://github.com/NetSys/quilt/pull/566",
"license": "mit",
"license_type": "permissive",
"license_source": "bigquery"
} |
128324433 | bundle install giving problems
I am getting following error
An error occurred while installing nokogiri (1.5.11), and Bundler cannot
continue.
Make sure that gem install nokogiri -v '1.5.11' succeeds before bundling.
while running bundle install
Regards
Hello,
I'm not sure I can help without more information. Are you following the instructions from the wiki? Have you installed all the dependencies listed? Are you on Ubuntu or another operating system? What version of ruby are you using? Can you please post the full output from your bundle install command?
Thanks!
Hello I am on ubuntu 14.04 LTS Desktop and using main repo as source list. I am using it on server now but getting broken dependencies. Is there any distro available for your software code.Best Tahir
On Sunday, 24 January 2016, 20:35, Andy Hoernecke <notifications@github.com> wrote:
Hello,I'm not sure I can help without more information. Are you following the instructions from the wiki? Have you installed all the dependencies listed? Are you on Ubuntu or another operating system? What version of ruby are you using? Can you please post the full output from your bundle install command?Thanks!—
Reply to this email directly or view it on GitHub.
Hi Tahir,
There is a docker distribution available here: https://github.com/ahoernecke/docker_scumblr
However, it's not currently officially maintained or supported. If you have issues with that, please make sure you're following the directions from the wiki when installing from source. https://github.com/Netflix/Scumblr/wiki
Thanks!
| gharchive/issue | 2016-01-23T12:06:41 | 2025-04-01T04:32:50.236965 | {
"authors": [
"ahoernecke",
"xsallowed"
],
"repo": "Netflix/Scumblr",
"url": "https://github.com/Netflix/Scumblr/issues/68",
"license": "apache-2.0",
"license_type": "permissive",
"license_source": "bigquery"
} |
651763536 | Bugfix/correlation id performance fix for issue #1681
This is a performance improvement in searching for workflows by correlation id. It adds a method to the ExecutionDAO interface that takes the workflow name so that implementations of ExecutionDAO can use it to join in other tables in order to get a smaller subset of workflows that only contain the ones with that workflow name.
This is a fix for issue #1681.
Pull Request Test Coverage Report for Build 4311
15 of 33 (45.45%) changed or added relevant lines in 6 files are covered.
7 unchanged lines in 3 files lost coverage.
Overall coverage decreased (-0.05%) to 69.55%
Changes Missing Coverage
Covered Lines
Changed/Added Lines
%
cassandra-persistence/src/main/java/com/netflix/conductor/dao/cassandra/CassandraExecutionDAO.java
0
1
0.0%
redis-persistence/src/main/java/com/netflix/conductor/dao/dynomite/RedisExecutionDAO.java
0
1
0.0%
core/src/main/java/com/netflix/conductor/core/orchestration/ExecutionDAOFacade.java
9
13
69.23%
core/src/main/java/com/netflix/conductor/service/ExecutionService.java
6
10
60.0%
mysql-persistence/src/main/java/com/netflix/conductor/dao/mysql/MySQLExecutionDAO.java
0
4
0.0%
postgres-persistence/src/main/java/com/netflix/conductor/dao/postgres/PostgresExecutionDAO.java
0
4
0.0%
Files with Coverage Reduction
New Missed Lines
%
cassandra-persistence/src/main/java/com/netflix/conductor/dao/cassandra/CassandraEventHandlerDAO.java
2
78.95%
cassandra-persistence/src/main/java/com/netflix/conductor/dao/cassandra/CassandraMetadataDAO.java
2
59.59%
mysql-persistence/src/main/java/com/netflix/conductor/dao/mysql/MySQLBaseDAO.java
3
69.32%
Totals
Change from base Build 4310:
-0.05%
Covered Lines:
12496
Relevant Lines:
17967
💛 - Coveralls
@rickfish Thank you very much for the contribution.
Given that the method getWorkflowsByCorrelationId in ExecutionDAOFacade is only being invoked from the ExecutionService, and in turn the getWorkflowsByCorrelationId method in ExecutionDAO is only being invoked from the ExecutionDAOFacade, I would vote to modify these method implementations with your changes rather than introduce a new method. Please let me know if you think otherwise.
Agreed. I just removed the getWorkflowsByCorrelationId methods that don't take workflowName as a parameter.
Good point. I removed it.
| gharchive/pull-request | 2020-07-06T19:18:27 | 2025-04-01T04:32:50.252752 | {
"authors": [
"apanicker-nflx",
"coveralls",
"rickfish"
],
"repo": "Netflix/conductor",
"url": "https://github.com/Netflix/conductor/pull/1766",
"license": "apache-2.0",
"license_type": "permissive",
"license_source": "bigquery"
} |
1004758934 | fix/inner-layout-overlap
Fix inner layout overlap
I'm no front-end engineer, but the menu overlapping the content layers bugs me a bit. So, any chance this could be the fix? Please end my suffering. Thank you in advance.
The said overlap:
Expected behavior after fix:
Hey @julianoborba, thanks for the PR! Interestingly, this change caused the behavior you're seeing for me. I just pushed a change with a few tweaks -- can you try it out and share some screenshots of what it looks like for you?
Thank you very much.
Looks like my soul found the inner peace with that tweaks. Works for me, if it's also working for you, seems you hit the spot (neat):
| gharchive/pull-request | 2021-09-22T20:45:48 | 2025-04-01T04:32:50.256436 | {
"authors": [
"julianoborba",
"patricksanders"
],
"repo": "Netflix/consoleme",
"url": "https://github.com/Netflix/consoleme/pull/9249",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
2249592118 | bug: Updates is_bot to support bots defined in apps.event.authorizations.list
Issue
Immediately upon creating a channel, our company invites a compliance bot to the channel. This bot advertises itself to the channel using the is_bot property in apps.events.authorizations.list. Because this heuristic is not used in is_bot, it is not recognized as a bot & throws an error when attempting to read the profile / email:
ERROR:slack_bolt.App:'email'
Traceback (most recent call last):
File "/usr/local/lib/python3.11/site-packages/slack_bolt/app/app.py", line 536, in dispatch
middleware_resp, next_was_not_called = listener.run_middleware(req=req, resp=resp)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/slack_bolt/listener/listener.py", line 51, in run_middleware
resp = m.process(req=req, resp=resp, next=next_)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/slack_bolt/middleware/custom_middleware.py", line 34, in process
return self.func(
^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/dispatch/decorators.py", line 134, in wrapper
result = func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.11/site-packages/dispatch/plugins/dispatch_slack/middleware.py", line 281, in user_middleware
email = client.users_info(user=user_id)["user"]["profile"]["email"]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^
KeyError: 'email'
Change
This augments the logic of is_bot to read the authorizations list to determine if the user in question is a bot while being backwards compatible with its existing heuristics.
Cheers 💯
| gharchive/pull-request | 2024-04-18T02:14:38 | 2025-04-01T04:32:50.259326 | {
"authors": [
"Meandmybadself",
"wssheldon"
],
"repo": "Netflix/dispatch",
"url": "https://github.com/Netflix/dispatch/pull/4635",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
331994443 | Merge values
Does dynomite support merge values after network issue occurs?
I have 2 dynomite instaces connected to 2 redis servers. After connection breaks, i add on one dynomite instance "test" -> "test1", on another i add "test" -> "test2". After connection re-established, values swaps between instances. I expected that one should be dropped, and use last updated. Do i mistake?
The last write should win. During the network disconnectivity, the data will be lost. After the connection is re-established there should be no problem. If the data have the same key it is up to Redis to merge them not Dynomite.
So it seems like Redis merged values incorrect?
Dynomite will not merge values on a network disconnect. But you can use
quorum settings to get better consistency in the system.
On Mon, Jun 25, 2018 at 9:52 PM, Anatolii Botov notifications@github.com
wrote:
So it seems like Redis merged values incorrect?
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
https://github.com/Netflix/dynomite/issues/565#issuecomment-400177933,
or mute the thread
https://github.com/notifications/unsubscribe-auth/ALle55hLub7-I80J-hCRvR4Nuwy8b5Bxks5uAb4CgaJpZM4UmMg6
.
Ok, i understand. Thanks.
| gharchive/issue | 2018-06-13T13:27:20 | 2025-04-01T04:32:50.263687 | {
"authors": [
"ipapapa",
"shailesh33",
"tolikr"
],
"repo": "Netflix/dynomite",
"url": "https://github.com/Netflix/dynomite/issues/565",
"license": "apache-2.0",
"license_type": "permissive",
"license_source": "bigquery"
} |
291239468 | Improve florida html parser
These changes help to deal with a case when HTML response might be unexpectedly splitted into parts directly on HTML content delimiter ("\r\n\r\n" sequence). The general idea here is below:
before every parsing iteration read socket until getting all data or filling RX buffer
if we don't get all data at current iteration and HTML content beginning still wasn't detected, last 3 received bytes will be copied to the beginning of the RX buffer and will be used on the next parsing iteration
@shailesh33
Looks like some tests failed. Any idea how to fix things?
| gharchive/pull-request | 2018-01-24T14:59:44 | 2025-04-01T04:32:50.265474 | {
"authors": [
"sobkovych"
],
"repo": "Netflix/dynomite",
"url": "https://github.com/Netflix/dynomite/pull/525",
"license": "apache-2.0",
"license_type": "permissive",
"license_source": "bigquery"
} |
592083546 | Remove equivalent destinations when cleaning certificates
Remove equivalent destinations when cleaning certificates. This will prevent Lemur from attempting to re-upload a certificate after it has been cleaned.
Coverage decreased (-0.02%) to 58.819% when pulling 67d24caef586920ee0d91f790596dead5f119bb5 on castrapel-patch-1 into 0c02ddec3c2f0d84962f7527834450b72078acd0 on master.
| gharchive/pull-request | 2020-04-01T17:31:18 | 2025-04-01T04:32:50.267473 | {
"authors": [
"castrapel",
"coveralls"
],
"repo": "Netflix/lemur",
"url": "https://github.com/Netflix/lemur/pull/2938",
"license": "apache-2.0",
"license_type": "permissive",
"license_source": "bigquery"
} |
2515046143 | Log messages sometimes do not appear
The code uses Python’s print function to log messages. However, print is buffered by default, so messages may not appear immediately or at all.
@fumoboy007 which code?
@savingoyal All code in this repository. Fixed in #438.
I think the existing print statements might be leftovers from way back. The preferred way to fix this issue should probably be to utilize the preconfigured logger from services.utils wherever possible.
from services.utils import logging
logger = logging.getLogger("ComponentName")
...
logger.info("some important message")
| gharchive/issue | 2024-09-09T22:36:52 | 2025-04-01T04:32:50.269386 | {
"authors": [
"fumoboy007",
"saikonen",
"savingoyal"
],
"repo": "Netflix/metaflow-service",
"url": "https://github.com/Netflix/metaflow-service/issues/437",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
183542995 | No details after upgrade
After upgrading to the latest version, the detail pages are blank. The jumplist still has dates, but they are disabled.
Also, reports just hang on loading. I don't see anything useful in the logs under /var/log... Where else can I check for errors that may point to the issue?
I followed these steps for the upgrade: https://github.com/Netflix/security_monkey/issues/205
Here is a screenshot
Hey @mugzy - Did you recompile the web UI (or download the static.tar.gz attached to the release?)
Could you open the Chrome inspector and then right click on the refresh icon and click on "Empty Cache and Hard Reload" ?
Are there any errors in the Javascript console?
Thanks! The "Empty Cache and Hard Reload" did the trick. Thanks!
| gharchive/issue | 2016-10-17T22:16:37 | 2025-04-01T04:32:50.272077 | {
"authors": [
"monkeysecurity",
"mugzy"
],
"repo": "Netflix/security_monkey",
"url": "https://github.com/Netflix/security_monkey/issues/429",
"license": "apache-2.0",
"license_type": "permissive",
"license_source": "bigquery"
} |
222622574 | Fix for issue #680 🚑
Should hopefully resolve #680 🤞
Also brings back env-config/config-local.py
Coverage increased (+0.2%) to 60.457% when pulling 7f9b4151cc21c1250741866d735a8563be5c46b3 on mikegrima:issue680 into 34d72462f4618c5afdea839b54b39f303a67e21a on Netflix:develop.
| gharchive/pull-request | 2017-04-19T05:24:03 | 2025-04-01T04:32:50.273985 | {
"authors": [
"coveralls",
"mikegrima"
],
"repo": "Netflix/security_monkey",
"url": "https://github.com/Netflix/security_monkey/pull/687",
"license": "apache-2.0",
"license_type": "permissive",
"license_source": "bigquery"
} |
939298753 | Added Pod status container statuses for multiple containers
Ever inching towards the multi-container world.
This plubming makes it so we can track the status
of the other non-main containers.
I'm not sending additional pod updates, however.
I'm concerned about overwhelming the API server.
But, when it is updated, I should be able to see the podstatus
via kubectl and actually see how my sidecars are doing.
Next steps after this are:
Add docker event watchers to actually update the status after start.
Optionally killing and restarting containers
Update the Titus API so that we can actually get the containerstatus's
all the way back to the user (so one doesn't have to use kubectl)
Bump: This is ready for review. These updates (while not yet full plumbed to docker healthcheck updates yet) will help power the titus api visibility that @amitaekbote is building.
| gharchive/pull-request | 2021-07-07T22:06:26 | 2025-04-01T04:32:50.276695 | {
"authors": [
"solarkennedy"
],
"repo": "Netflix/titus-executor",
"url": "https://github.com/Netflix/titus-executor/pull/681",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
245754293 | When a node is clicked, a connection is selected
When clicking on a node, there is a good change that the node is not selected, but the connection. It looks like the click event for the connection is fired because that is the one which is selected.
There was an overlapping issue that should be fixed in the latest. (eda7d404e4d9ccd3a380bf2b2b16d934d48a40b1)
| gharchive/issue | 2017-07-26T15:06:48 | 2025-04-01T04:32:50.277839 | {
"authors": [
"fervanrijswijk",
"jrsquared"
],
"repo": "Netflix/vizceral",
"url": "https://github.com/Netflix/vizceral/issues/81",
"license": "apache-2.0",
"license_type": "permissive",
"license_source": "bigquery"
} |
2333732314 | 🛑 NetsBlox Documentation is down
In de70aef, NetsBlox Documentation (https://editor.netsblox.org/docs) was down:
HTTP code: 502
Response time: 2 ms
Resolved: NetsBlox Documentation is back up in a99588c after 9 minutes.
| gharchive/issue | 2024-06-04T14:46:40 | 2025-04-01T04:32:50.305947 | {
"authors": [
"brollb"
],
"repo": "NetsBlox/uptime",
"url": "https://github.com/NetsBlox/uptime/issues/27",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
1229464882 | Low severity threshold doesn't fail test on detecting Low issue
Steps to reproduce:
Run the following scan inside the test (it) block:
const severity = 'LOW';
it(`BC -${severity}`, async () => {
await scan.threshold(Severity[severity]).run({
method: 'GET',
url: 'https://qa.brokencrystals.com/',
});
});
Observe results
Actual: test passed, no exception - despite the fact that there are 5 low issues detected
Expected: test fails, exception that Low issue is found is thrown
Note: same result if no threshold passed
@wawaqa this issue has nothing in common with sec-tester. It looks like it is the same scan status transition issue as in #72.
Look at the runtime notifications:
2022-05-09 16:19:23.000 [INFO] Scan done, shutting down engine
2022-05-09 16:19:26.716 [WARNING] Found new ‘Missing X-Frame-Options Header’ (Low) vulnerability at: https://qa.brokencrystals.com/
2022-05-09 16:19:27.716 [WARNING] Found new ‘Missing Access-Control-Allow-Origin Header’ (Low) vulnerability at: https://qa.brokencrystals.com/
2022-05-09 16:19:28.687 [WARNING] Found new ‘Missing Strict-Transport-Security Header’ (Low) vulnerability at: https://qa.brokencrystals.com/
2022-05-09 16:19:29.613 [WARNING] Found new ‘Missing X-Content-Type-Options Header’ (Low) vulnerability at: https://qa.brokencrystals.com/
2022-05-09 16:19:30.556 [WARNING] Found new ‘Missing Content-Security-Policy Header’ (Low) vulnerability at: https://qa.brokencrystals.com/
2022-05-09 16:21:24.908 [FATAL] Terminated! exit code: `137`, finished at `Time(2022-05-09T12:21:24Z)` reason: `Error`
The engine sends the done status first and then it pushes issues (a few seconds later).
| gharchive/issue | 2022-05-09T09:57:05 | 2025-04-01T04:32:50.318043 | {
"authors": [
"derevnjuk",
"wawaqa"
],
"repo": "NeuraLegion/sec-tester-js",
"url": "https://github.com/NeuraLegion/sec-tester-js/issues/78",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
1939331577 | BlackRock data format 3.0
Describe the bug
Read *.ns6 data from BlackRock causes a "Too many open files" error in __read_nsx_data_variant_b function nearly immediately.
To Reproduce
Something like this:
import neo
data_dir = "data/test_data/HubTests/"
path_gemini = data_dir + "233117-001.ns6" # new data
reader = neo.io.BlackrockIO(filename=path_gemini)
(Data cannot be shared publicly.)
Expected behaviour
Data is read into memory
Environment:
OS: Windows 10
Python version: 3.7.8
Neo version: momentarily not available, most recent downloadable
NumPy version: 1.21.6
Additional context
New Blackrock device uses nsx-format version 3.0.
The neo code seems to have execute branches for version 3.0,
but states in comments that it handles nsx-formats up to version 2.3.
Prior devices (using nsx version 2.3) produced data that were read without problem,
so it seems that either blackrock is implementing the 3.0 format incorrectly, or neo
is handling it incorrectly.
There could be a marker, or signal, in the v3.0-data that is interpreted as
a "recording pause" triggering a segment split by neo code?
The arrival of nsx v3.0 is recent for us, with new equipment, but the version
may have been available for a long while.
Is this a known problem?
Further input, in case it helps: a stacktrace from execution as above:
---------------------------------------------------------------------------
OSError Traceback (most recent call last)
Cell In[5], line 1
----> 1 reader = neo.io.BlackrockIO( filename="data/20230710-031124-001.ns6")
File ~/VirtualEnvs/neotest/lib/python3.11/site-packages/neo/io/blackrockio.py:17, in BlackrockIO.__init__(self, filename, **kargs)
15 def __init__(self, filename, **kargs):
16 BlackrockRawIO.__init__(self, filename=filename, **kargs)
---> 17 BaseFromRaw.__init__(self, filename)
File ~/VirtualEnvs/neotest/lib/python3.11/site-packages/neo/io/basefromrawio.py:74, in BaseFromRaw.__init__(self, *args, **kargs)
72 def __init__(self, *args, **kargs):
73 BaseIO.__init__(self, *args, **kargs)
---> 74 self.parse_header()
File ~/VirtualEnvs/neotest/lib/python3.11/site-packages/neo/rawio/baserawio.py:179, in BaseRawIO.parse_header(self)
166 def parse_header(self):
167 """
168 This must parse the file header to get all stuff for fast use later on.
169
(...)
177
178 """
--> 179 self._parse_header()
180 self._check_stream_signal_channel_characteristics()
File ~/VirtualEnvs/neotest/lib/python3.11/site-packages/neo/rawio/blackrockrawio.py:366, in BlackrockRawIO._parse_header(self)
364 for nsx_nb in self.nsx_to_load:
365 spec = self.__nsx_spec[nsx_nb]
--> 366 self.nsx_datas[nsx_nb] = self.__nsx_data_reader[spec](nsx_nb)
368 sr = float(main_sampling_rate / self.__nsx_basic_header[nsx_nb]['period'])
369 self.sig_sampling_rates[nsx_nb] = sr
File ~/VirtualEnvs/neotest/lib/python3.11/site-packages/neo/rawio/blackrockrawio.py:975, in BlackrockRawIO.__read_nsx_data_variant_b(self, nsx_nb)
971 offset = \
972 self.__nsx_data_header[nsx_nb][data_bl]['offset_to_data_block']
974 # read data
--> 975 data[data_bl] = np.memmap(
976 filename, dtype='int16', shape=shape, offset=offset, mode='r')
978 return data
File ~/VirtualEnvs/neotest/lib/python3.11/site-packages/numpy/core/memmap.py:267, in memmap.__new__(subtype, filename, dtype, mode, offset, shape, order)
265 bytes -= start
266 array_offset = offset - start
--> 267 mm = mmap.mmap(fid.fileno(), bytes, access=acc, offset=start)
269 self = ndarray.__new__(subtype, shape, dtype=descr, buffer=mm,
270 offset=array_offset, order=order)
271 self._mmap = mm
OSError: [Errno 24] Too many open files
I have been in communication with @gedwald and I have been able to reproduce. I am investigating.
The problem is because the file comes from one of Blackrock's newer PTP-enabled systems. Previously, the presence of a header would indicate a new section due to a dropped sample, and all samples within a section would be assumed to be exactly 1/30_000 s apart. Now, on PTP-enabled systems only, each sample has a header containing its timestamp (nanoseconds). If the data reader thinks that the header indicates a new section, then it will create a new section for each sample.
This has been fixed in Blackrock's Python-Utilities (see change here: https://github.com/BlackrockNeurotech/Python-Utilities/pull/22). I am working on a fix for neo. While I do so, @gedwald and other readers may wish to use Blackrock's file importer. https://github.com/BlackrockNeurotech/Python-Utilities/blob/main/examples/extract_continuous_data.ipynb
| gharchive/issue | 2023-10-12T07:00:54 | 2025-04-01T04:32:50.325973 | {
"authors": [
"cboulay",
"gedwald"
],
"repo": "NeuralEnsemble/python-neo",
"url": "https://github.com/NeuralEnsemble/python-neo/issues/1332",
"license": "bsd-3-clause",
"license_type": "permissive",
"license_source": "bigquery"
} |
692165989 | Fixes to make tests pass with Pegasus event files only directory.
This set of changes fixes a bug which prevented the Pegasus .Nev event only files from passing the test suite.
It also allows filename extensions to be in any letter case. For the InputInverted header parameter to be missing,
meaning not inverted. Checks for 8 hashes at the start of the header as a check on header contents. Allows
any whitespace between a header key and value.
Internal changes are addition of a large number of explanatory comments and refactoring of the code to handle
the header into a separate class. Many of these in preparation for handling earlier versions of neuralynx file headers
and more flexible handling of segments withins .Ncs files.
Hello @PeterNSteinmetz! Thanks for opening this PR. We checked the lines you've touched for PEP 8 issues, and found:
In the file neo/rawio/neuralynxrawio.py:
Line 100:30: E261 at least two spaces before inline comment
Line 168:100: E501 line too long (103 > 99 characters)
Line 181:53: E231 missing whitespace after ','
Line 201:100: E501 line too long (107 > 99 characters)
Line 245:39: E261 at least two spaces before inline comment
Line 588:1: E302 expected 2 blank lines, found 1
Line 600:1: E302 expected 2 blank lines, found 1
Line 678:100: E501 line too long (100 > 99 characters)
Line 679:100: E501 line too long (100 > 99 characters)
Line 700:49: E231 missing whitespace after ','
Many thanks for this PR. Please also add your name and affiliation to doc/source/authors.rst
Hi Peter,
thank you very much for this.
Except the comment I have made in the code, I think I am OK with this patch.
Lets see if Andrew or Julia will have time to review this.
Cheers
Samuel
It is definitely easy to revert the change of the local argument name to channel_indices if that is what people want. Since this is a name simply local to the function and makes no difference to the caller, I would suggest just gradually changing this spelling in functions as they are changed for other reasons.
Bumping this. Can I get some advice on use of channel_indices for local argument names in non-public functions, i.e., those beginning with _ ? I can then change this in both of my current pull requests.
Hi @PeterNSteinmetz, thanks for the the PR . Look good to me. Please just revert the name changes for channel_indices and I would be happy to merge.
Appears now to past all tests and should be ready to merge.
| gharchive/pull-request | 2020-09-03T17:57:24 | 2025-04-01T04:32:50.337519 | {
"authors": [
"JuliaSprenger",
"PeterNSteinmetz",
"apdavison",
"pep8speaks",
"samuelgarcia"
],
"repo": "NeuralEnsemble/python-neo",
"url": "https://github.com/NeuralEnsemble/python-neo/pull/851",
"license": "bsd-3-clause",
"license_type": "permissive",
"license_source": "bigquery"
} |
1105436854 | added mriqc tutorial and images to static/mriqc_cvl
Hey folks :)
I have added an mriqc tutorial. Let me know if you have any issues with it.
BTW I am having issues with inserting images. I have added them to static/mriqc_cvl as per the workflow template instructions, and I believe the .md document refers to the images correctly. However, when I build the pages in my repo I do not see the images on the page. Can you let me know if I have missed something?
Cheers :)
Dear @kel-github,
This is fantastic :) Thank you! The images show up nicely for me - not sure what caused the issue on your repo?
Ha, strange. Glad it came through good for you's :)
On Mon, 17 Jan 2022 at 16:46, Steffen Bollmann @.***>
wrote:
Dear @kel-github https://github.com/kel-github,
This is fantastic :) Thank you! The images show up nicely for me - not
sure what caused the issue on your repo?
[image: image]
https://user-images.githubusercontent.com/4021595/149720720-975b667f-0674-44f5-86e4-37671ecebc10.png
—
Reply to this email directly, view it on GitHub
https://github.com/NeuroDesk/neurodesk.github.io/pull/93#issuecomment-1014190435,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/ABXC342O4V3GOUHUJFBF5F3UWO3MVANCNFSM5MD3LK6Q
.
Triage notifications on the go with GitHub Mobile for iOS
https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675
or Android
https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.
You are receiving this because you were mentioned.Message ID:
@.***>
--
Kelly Garner, PhD, she/her
Marie Curie Research Fellow
Google Scholar
https://scholar.google.co.uk/citations?user=nmeSlZkAAAAJ&hl=en
ResearchGate https://www.researchgate.net/profile/Kelly_Garner
U**Q ALL**Y :: Supporting the diversity of sexuality, sex and
gender at UQ.
--I acknowledge the Traditional Owners of the land upon which I live and
work. I pay my respects to Elders past, present and emerging.--
--I seek to maintain a good work-life balance. If you receive this email
outside office hours then don’t feel you have to respond. Please expect the
like from me.--
| gharchive/pull-request | 2022-01-17T06:34:51 | 2025-04-01T04:32:50.347231 | {
"authors": [
"kel-github",
"stebo85"
],
"repo": "NeuroDesk/neurodesk.github.io",
"url": "https://github.com/NeuroDesk/neurodesk.github.io/pull/93",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
1151191646 | Finding value of weights
Hi @shuheng-liu
How can I access the initial values of weights? As well as the final value of weights after max_epochs=1000 fixed epochs when using solver1D class?
Solver1D(
ode_system = lambda u, t: [diff(u, t, order=2) + u],
conditions=[IVP(0, 1, 0)],
t_min=0.0,
t_max=2.0,
train_generator=Generator1D(100, 0.0, 2),
valid_generator=Generator1D(100, 0.0, 2),
)
from neurodiffeq.callbacks import MonitorCallback
monitor = Monitor1D(t_min=0.0, t_max=3.0, check_every=100)
monitor_callback = MonitorCallback(monitor)
solver_system.fit(max_epochs=1000, callbacks=[monitor_callback])
solution_system = solver_system.get_solution()```
Hi dear @shuheng-liu is there any way to find value of weights?
Hello,
We can find the weights by accessing the nets parameter of the solver object and then its NN field (for a FCNN object)(https://github.com/NeuroDiffGym/neurodiffeq/blob/master/neurodiffeq/networks.py#L6). This gives a torch.nn.Sequential object and its weights can easily be found out by indexing the layer and accessing weight field. For your given example,
solver_system.nets[0].NN[0].weight
The output I got when I ran the code:
Parameter containing:
tensor([[ 0.8768],
[-0.6987],
[-0.5978],
[ 0.3867],
[-0.3607],
[-0.6576],
[-0.4753],
[-0.1397],
[-0.6517],
[ 0.4632],
[-0.5898],
[-0.3324],
[ 0.8168],
[-0.8792],
[-0.7331],
[-0.1540],
[ 0.1956],
[ 0.7815],
[ 0.2859],
[-0.6865],
[ 0.8375],
[ 0.8586],
[-0.8746],
[-0.5371],
[-0.7475],
[ 0.4840],
[ 0.5717],
[-0.1432],
[ 0.2203],
[-0.3002],
[ 0.2296],
[-0.3601]], requires_grad=True)
| gharchive/issue | 2022-02-26T04:23:41 | 2025-04-01T04:32:50.350882 | {
"authors": [
"Arup-nit",
"sathvikbhagavan"
],
"repo": "NeuroDiffGym/neurodiffeq",
"url": "https://github.com/NeuroDiffGym/neurodiffeq/issues/166",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
911465952 | MCS-640 moved ai2thor json writing to recorder
This is a small update to make some of the json writing threaded.
It be nice to thread the debug log file writer, but it creates a json string in a custom way. We could create yet another record to just write text?
Any unit tests?
Added a bunch of unit tests
| gharchive/pull-request | 2021-06-04T12:42:47 | 2025-04-01T04:32:50.381951 | {
"authors": [
"deanwetherby",
"kdrumm"
],
"repo": "NextCenturyCorporation/MCS",
"url": "https://github.com/NextCenturyCorporation/MCS/pull/354",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
1643305095 | id: 'npmuser': no such user
emqx_nginx-proxy | ❯ Configuring npmuser ...
emqx_nginx-proxy | id: 'npmuser': no such user
emqx_nginx-proxy | ❯ Checking paths ...
emqx_nginx-proxy | ❯ Setting ownership ...
emqx_nginx-proxy | s6-rc: fatal: timed out
emqx_nginx-proxy | s6-sudoc: fatal: unable to get exit status from server: Operation timed out
emqx_nginx-proxy | /run/s6/basedir/scripts/rc.init: warning: s6-rc failed to properly bring all the services up! Check your logs (in /run/uncaught-logs/current if you have in-container logging) for more information.
run in Docker
nginx-proxy: image: 'jc21/nginx-proxy-manager:latest' container_name: emqx_nginx-proxy hostname: nginx-proxy.emqx.io restart: unless-stopped ports: - '6808:80' - '8818:81' - '7443:443' # volumes: # - ./nginx/nginx-proxy-manager/data:/data # - ./nginx/nginx-proxy-manager/letsencrypt:/etc/letsencrypt
Please try the github-develop branch to resolve the fatal problem - there are other issues reporting this.
The title id: 'npmuser': no such user is not part of the problem and is informational in the logs only.
Please try the github-develop branch to resolve the fatal problem - there are other issues reporting this.
The title id: 'npmuser': no such user is not part of the problem and is informational in the logs only.
I tried github-develop but still had the same issue.
Using github-pr-2661 seemed to fix the issue for me.
| gharchive/issue | 2023-03-28T06:46:09 | 2025-04-01T04:32:52.078259 | {
"authors": [
"AceTheSpaceCase",
"JosonJiang",
"jc21"
],
"repo": "NginxProxyManager/nginx-proxy-manager",
"url": "https://github.com/NginxProxyManager/nginx-proxy-manager/issues/2760",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
1567618444 | Adding to the catalogue
I add some new papers to the catalogue and I would like to do a test to ensure the viability of the data firstly.Moreover,I will check the data again in the following work. Please inform me in time if there is any issue.
Thanks for this @harvey0908. Please allow me to have edit rights (basically just tick the box in the last picture) and I'll start making edits.
Looks like the problem is that you have put 0.0 for a large number of the bandwidth values, which will cause errors (can't divide by zero), so basically, the code wasn't able to fit anything. I fixed one of them, so I'll push it once you've given me push permission then you can start working on the rest.
I've pushed my changes. I renamed some of the YAML files so make sure to use pip uninstall pulsar_spectra to remove the old YAML files
Thanks,l will try to reconfirm the the bandwidth values.
Our current policy is if there are no bandwidth values, do not include that paper (would you agree with this, @cplee1?).
If there are any that you are struggling with, let me know which ones they are, and I'll see if I can at least work out a likely bandwidth.
Ok,thanks
As you said,zero as the value of the bandwidth value affects the result.l have reconfirmed it for a part of papers. However, I feel apologetic as a newcomer.It's a little difficult for me to ensure they are correct Including the rest that can't be determined.Can I contact you in other places or leave your email address so that I can contact you in time.
There is no need to be apologetic! I am excited that you are the first person to make a pull request of an addition to the catalogue. Because you are the first, it will be difficult as we learn the best way to help newcomers.
You can continue to message me on this pull request, there is no message limit. If you prefer, you can email me at nickaswainston@gmail.com.
I have emailed you,including some papers I have confirmed the value of the bandwidth.Sorry to trouble you again.
Hi ,I noticed you deleted some date. Becauese those flux density measurements from previous publications. Thank you for reminding me , I ignored these situations before. I will check the rest papers.
All fixed up and passing all the tests. Thanks again for working with me to get these papers merged!
OK , Thanks for your help！
| gharchive/pull-request | 2023-02-02T09:06:49 | 2025-04-01T04:32:52.096135 | {
"authors": [
"NickSwainston",
"harvey0908"
],
"repo": "NickSwainston/pulsar_spectra",
"url": "https://github.com/NickSwainston/pulsar_spectra/pull/73",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
62957437 | Possible bug in Grid and Canvas element reordering? Or is it just me... :)
I've recently added support for reordering Setters, and I'm currently working on refactoring all the element reordering code into a separate service class.
This ensures that it will be a breeze to add new reorder functionality in the future ;)
The code is all all done, but the output is NOT identical to the original Grid & Canvas reorder code.
The difference is how whitespace and especially comments are handled (or actually NOT handled) by the current Grid/Canvas reorder code.
My new service assumes that all whitespace/comments "belongs" to the following element. In the original code it seems to belong to the previous element.
This has this effect:
Given the following snippet
<Grid>
<Node1 Grid.Row="2">
<Node1 Grid.Row="1">
</Grid>
With current XamStyler would produce:
<Grid>
<Node1 Grid.Row="1">
<Node1 Grid.Row="2">
</Grid>
However the new code would produce:
<Grid>
<Node1 Grid.Row="1">
<Node1 Grid.Row="2">
</Grid>
This at first might seem wrong, but if you add a few comments you see the problem:
Given:
<Grid>
<!-- Comment for row 2 -->
<Node1 Grid.Row="2">
<!-- Comment for row 1 -->
<Node1 Grid.Row="1">
</Grid>
Current XamStyler would produce:
<Grid>
<!-- Comment for row 2 -->
<Node1 Grid.Row="1">
<Node1 Grid.Row="2">
<!-- Comment for row 1 -->
</Grid>
New code produces:
<Grid>
<!-- Comment for row 1 -->
<Node1 Grid.Row="1">
<!-- Comment for row 2 -->
<Node1 Grid.Row="2">
</Grid>
I always write comments in XAML before an element so it seems logical that they should "belong" to the following element, but what about whitespaces?
What do you think the correct output from my test snippet should be?
I've kept it just like this. works good enough for me
if others find it not working well, please open up a new issue
| gharchive/issue | 2015-03-19T11:10:08 | 2025-04-01T04:32:52.113285 | {
"authors": [
"NicoVermeir",
"PolarbearDK"
],
"repo": "NicoVermeir/XamlStyler",
"url": "https://github.com/NicoVermeir/XamlStyler/issues/47",
"license": "apache-2.0",
"license_type": "permissive",
"license_source": "bigquery"
} |
2060808637 | 🛑 Hacker News is down
In e608b0e, Hacker News (https://news.ycombinator.com) was down:
HTTP code: 502
Response time: 147 ms
Resolved: Hacker News is back up in f96fcb3 after 11 minutes.
| gharchive/issue | 2023-12-30T17:34:28 | 2025-04-01T04:32:52.115813 | {
"authors": [
"NicolasAbihaggle"
],
"repo": "NicolasAbihaggle/serviciosestados",
"url": "https://github.com/NicolasAbihaggle/serviciosestados/issues/743",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
2743797420 | 🛑 cedro is down
In 1030a09, cedro (https://www.cedro.org) was down:
HTTP code: 0
Response time: 0 ms
Resolved: cedro is back up in 12f5cb3 after 7 hours, 26 minutes.
| gharchive/issue | 2024-12-17T02:42:21 | 2025-04-01T04:32:52.118165 | {
"authors": [
"NicolasAbihaggle"
],
"repo": "NicolasAbihaggle/serviciosestados",
"url": "https://github.com/NicolasAbihaggle/serviciosestados/issues/832",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
1381043434 | [Feature] - Adicionando fluxo de Compras
Motivo do PR
É necessário que o aplicativo cadastre quais foram os itens que o usuário comprou, consiga buscar todas as compras e também buscar uma compra por ID
O que foi implementado
[x] - Criação do Controller para Compras
[x] - Criação do Repository para Compras
[x] - Criação das rotas para fluxo de compras
Como Testar
Para Criar uma Lista de Compras
1 - Copiar esta branch.
2 - Subir a aplicação docker-compose up para subir o BD, npm install para instalar os pacotes e npm run dev para iniciar a aplicação no modo desenvolvedor.
É necessário ter criado um usuário no sistema
3 - Enviar uma requisição POST para http://localhost:3000/purchases/ com o header Authorization contendo o token do usuário autenticado com o payload abaixo:
{
"title": "lista de compras testes"
}
4 - O status code de retorno deve ser 201 - CREATED payload retornado devera ser:
{
"id": 1,
"title": "lista de compras testes",
}
Para Buscar as listas de compras de um usuário
1 - Enviar uma requisição GET para http://localhost:3000/users/me/purchases/ com o header Authorization contendo o token do usuário autenticado
2 - O status code de retorno deve ser 200 - OK payload retornado devera ser:
[
{
"id": 3,
"title": "lista teste select",
"createdAt": "2022-09-21T04:44:18.407Z"
},
{
"id": 2,
"title": "lista compras",
"createdAt": "2022-09-21T04:38:10.096Z"
},
{
"id": 1,
"title": "lista 1",
"createdAt": "2022-09-21T04:38:03.738Z"
}
]
Para buscar uma compra por ID especifico
1 - Enviar uma requisição GET para http://localhost:3000/purchases/:idPurchase com o header Authorization contendo o token do usuário autenticado
2 - O status code de retorno deve ser 200 - OK payload retornado devera ser:
{
"id": 1,
"title": "lista teste",
"createdAt": "2022-09-21T04:44:18.407Z",
"PurchasedProducts": []
}
Show mano! Só deixei anotado uma nomenclatura inadequado ali no model, o restante tá show.
Um feedback na descrição do PR: vc tem focado bastante no “o que”/“como”, mas não deu nenhum detalhe sobre o “por que”/“como”. Além de explicar os passos para usar o código que está sendo introduzido, é muito importante vc explicar a motivação do pq introduzir esse código. Quais problemas ele resolve? O que ele possibilita o usuário fazer? Como isso (o código introduzido) se encaixa no objetivo final da feature/programa?
Ex:
Introduzindo Compras.
Agora, os usuário poderão comprar os items salvos em suas listas de compras. O usuário deverá informar qual a lista de compras a qual deseja comprar e uma forma de pagamentos.
Caso a lista não exista, será retornado um erro X;
Caso os dados de pagamento sejam inválidos, será retornado um error Y;
Caso o pagamento seja recusado, será retornado um erro Z;
E em caso de sucesso, a compra será efetuada e o usuário receberá um e-mail de confirmação.
Do ponto de vista técnico, não vejo nenhum problema no que vc implementou (além da nomenclatura não-convencional de uma foreign key). No entanto, falta contexto da feature (contexto do que vc está tentando construir — qual objetivo vc pretende alcançar entregando esse código) para que eu possa te ajudar quanto a se o que vc construiu faz sentido ou não.
Me parece que o que vc está entregando aqui deveria ser um carrinho de compras e não uma compra em si.
Muito obrigado pelo feedback, realmente faz sentido a descrição de implementação ser mais detalhada, vou melhorar isso nos próximos PRs
| gharchive/pull-request | 2022-09-21T14:37:28 | 2025-04-01T04:32:52.130440 | {
"authors": [
"NicolasPereira"
],
"repo": "NicolasPereira/nodejs-api-tcc",
"url": "https://github.com/NicolasPereira/nodejs-api-tcc/pull/21",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
144492640 | SQL Error on latests results page
On page: https://html5test.com/results/latest.html, you are currently getting sql syntax error.
You have an error in your SQL syntax; check the manual that corresponds to your MariaDB server version for the right syntax to use near 'ORDER BY timestamp DESC LIMIT 100' at line 7
Errors shouldn't be visible to user in production environment and you should investigate what what is wrong in generated $where, because there is potential SQL Injection vector.
Fixed on alpha.html5test.com
| gharchive/issue | 2016-03-30T07:50:11 | 2025-04-01T04:32:52.134096 | {
"authors": [
"BetaFrey",
"NielsLeenheer",
"Szpadel"
],
"repo": "NielsLeenheer/html5test",
"url": "https://github.com/NielsLeenheer/html5test/issues/431",
"license": "mit",
"license_type": "permissive",
"license_source": "bigquery"
} |
326368672 | Handle overlapping tooltips
Fixing a visibility issue when tooltips overlap. Tooltips should 'merge' to a single tooltip to allow visibility of data.
See #171
wow, this is an amazing feature.
I have a few ideas before i merge it. If you are free, you can help me realize it, or I can do it later.
Provides a property (default: true) to control whether the merge tooltips In the case of overlapping.
Provides a property(Similar to the formatter, default: {value1} - {value2}) to customize merge rules.
The above is just my suggestion. I look forward to your feedback.
I have completed the points mentioned yesterday. Version 2.7.0 has been released
I really appreciate your contribution.
Is there a way to trigger merge on load?
| gharchive/pull-request | 2018-05-25T03:29:22 | 2025-04-01T04:32:52.147811 | {
"authors": [
"NightCatSama",
"aniwng",
"suavelizard"
],
"repo": "NightCatSama/vue-slider-component",
"url": "https://github.com/NightCatSama/vue-slider-component/pull/192",
"license": "mit",
"license_type": "permissive",
"license_source": "bigquery"
} |
2608496613 | [FEATURE] Bring in support for Python 3.13
Is your feature request related to a problem? Please describe.
The koheesio framework does not currently support Python 3.13, limiting developers from using the latest features and improvements.
Describe the solution you'd like
[ ] Add support for Python 3.13 by updating the pyproject.toml file
[ ] ensuring all dependencies and tests are compatible.
[ ] Update the CI/CD pipeline to include Python 3.13 in the testing matrix.
Describe alternatives you've considered
Not Supporting Python 3.13: Continue supporting up to Python 3.12, but this prevents using new features in Python 3.13.
Partial Support: Update the pyproject.toml file without updating the CI/CD pipeline, allowing use of Python 3.13 without guaranteed compatibility.
Additional context
Supporting Python 3.13 ensures the koheesio framework remains up-to-date, providing developers access to new features and improvements.
Checked the current state (based around 0.9.0rc0):
Tableau dependencies are incompatible with Python 3.13
Snowflake dependencies are incompatible with Python 3.13
JDBCReader test is failing with FAILED tests/spark/readers/test_jdbc.py::TestJdbcReader::test_execute_w_dbtable_and_query - AssertionError: assert 'foo' is None
WIth that, marking this as blocked for now. Will have to revisit this in the future.
| gharchive/issue | 2024-10-23T12:51:32 | 2025-04-01T04:32:52.185854 | {
"authors": [
"dannymeijer"
],
"repo": "Nike-Inc/koheesio",
"url": "https://github.com/Nike-Inc/koheesio/issues/80",
"license": "Apache-2.0",
"license_type": "permissive",
"license_source": "github-api"
} |
2587845420 | Want to make look of the website
Add more CSS properties to make more user friendly like addining transition, opacity , ...etc
hi @Dipanita45 , where are you adding css
| gharchive/issue | 2024-10-15T07:05:07 | 2025-04-01T04:32:52.189949 | {
"authors": [
"Dipanita45",
"Niketkumardheeryan"
],
"repo": "Niketkumardheeryan/ML-CaPsule",
"url": "https://github.com/Niketkumardheeryan/ML-CaPsule/issues/1122",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
2186397311 | chafa v1.14.0 messes up sixel image previews
In the new versions of chafa, some extra sixel escape sequences are sent which messes up image previews in lf. The fix is to add an extra --polite=on to avoid sending them.
Is there any way I can append these args to chafa?
For now, I wrote a wrapper for chafa and placed in in my ~/.local/bin folder.
#!/bin/sh
exec /usr/sbin/chafa --polite=on "$@"
This should be resolved by #90 when it's merged.
This has not been merged yet, neither has the PR updating the README(lf supports sixel now). Has the project been abandoned?
After spending an hour researching this and looking through the lf/ctpv/chafa repo issues, in the end I simply downgraded chafa to 1.12.4.
Safe to say the project is unmaintained at this point
| gharchive/issue | 2024-03-14T13:32:11 | 2025-04-01T04:32:52.192722 | {
"authors": [
"BaconIsAVeg",
"UtkarshVerma",
"elvindesouza",
"fdnt7"
],
"repo": "NikitaIvanovV/ctpv",
"url": "https://github.com/NikitaIvanovV/ctpv/issues/94",
"license": "MIT",
"license_type": "permissive",
"license_source": "github-api"
} |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.