added
stringdate 2025-04-01 04:05:38
2025-04-01 07:14:06
| created
timestamp[us]date 2001-10-09 16:19:16
2025-01-01 03:51:31
| id
stringlengths 4
10
| metadata
dict | source
stringclasses 2
values | text
stringlengths 0
1.61M
|
|---|---|---|---|---|---|
2025-04-01T04:10:38.320059
| 2017-01-04T13:32:40
|
198707981
|
{
"authors": [
"camohub",
"f3l1x"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14898",
"repo": "Markette/Gopay",
"url": "https://github.com/Markette/Gopay/issues/55"
}
|
gharchive/issue
|
How to set SK lang
Hi, I would like to have SK lang in GoPay form, and as I see GoPay support SK language https://doc.gopay.com/en/?php#lang
But in the Markete are allowed only CZ/EN.
There is a method:
public function setLang($lang)
{
if (!in_array($lang, $this->allowedLang)) {
throw new InvalidArgumentException('Not supported language "' . $lang . '".');
}
$this->lang = $lang;
return $this;
}
which does not allow any other language. If I change the code ($allowedLang property), to support SK lang it works. In my opinion condition in setLang method is useless. Everybody can ensure the valid lang code and in other way it throws an exception (as it works also for now). Is it really necessary to check allowed langs which are hardly coded in AbstractService->allowedLang?
Thanks.
EDIT: Not valid lang code falls to default EN, what is expectable and desirable, so that condition looks really useless.
Hi. Thank you for your time.
What about stay strict and append only missing languages?
Language code Language
CS Czech
EN English
SK Slovak
DE German
RU Russian
It si your dicision. But what about ES, PL, HU, ................................
I think it is not necessary to be strict here. Fallback to default is common pattern and I thing Gopay will support it also in future. So setLang can be so simple as possible.
Hmm. Good point. But I still thing that strictness should be on library side. It's no difference for programmers, if they'll type properly values, otherwise it'll throw an exception.
I think this way is more clean. Also it's more predictable, because you know what you send to server. Fallbacks are not straightforward solutions, especially not for the beginners.
Do you agree? Would you mind prepare a PR for it?
I don know how to make PR from GH fork. It is not so simple if I remember. But I am still against the strict mode here. Exception as exception. It is the same. If Gopay will throw an exception in future I believe it will be clean also for beginners. And every new language will need new release of this library. Strict mode is mainly about safety. This is not user friendly.
THX.
So I did PR. Now it is on you.
And what about change the exception to EN fallback in this method?
@camohub Sorry, I would like to maintain it as closed (strict) as possible. So no, no fallback. But thank you for your time.
Closed via #57
|
2025-04-01T04:10:38.323027
| 2023-01-24T16:04:58
|
1555252266
|
{
"authors": [
"Claes65"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14899",
"repo": "MarkusAmshove/natls",
"url": "https://github.com/MarkusAmshove/natls/pull/123"
}
|
gharchive/pull-request
|
Added some variants of NOT
NOT <, NOT >, NOT LT, NOT GT, NOTEQUAL
and added new tests and extended other tests
@MarkusAmshove please advise about the code smell, I gave it a read, but I don't understand how I can alter it...
|
2025-04-01T04:10:38.402255
| 2016-05-08T17:25:56
|
153667369
|
{
"authors": [
"MarshallOfSound",
"berryelectronics"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14900",
"repo": "MarshallOfSound/Google-Play-Music-Desktop-Player-UNOFFICIAL-",
"url": "https://github.com/MarshallOfSound/Google-Play-Music-Desktop-Player-UNOFFICIAL-/issues/1051"
}
|
gharchive/issue
|
Bug: Snap feature fails in the mini player when using multiple Virtual Desktops
Win10 x64
When using multiple virtual desktops and trying to snap the mini player onto the right side on the first desktop it wont get the borders right. The anchor point of the mini player doesnt seem to be correct. This results into the player being of the screen
Steps to reproduce:
-Create at least two virtual desktops
-move the mini player onto the first desktop
-press win+right arrow so it "snaps" to the right side
Also what i noticed: You cant snap it in the lower corner. It just snaps to the upper right or upper left. When pressing win+up it moves the player into its normal position (cant maximize it). When pressing win+down doesnt do anything if its already snapped, not moving to the lower corner or into the tray. It only moves to the tray if it isnt snapped
The snapping bug is an issue with WIndows itself not handling windows with a maximum size correctly.
The anchor point issue can't be produced on my Windows 10 machine running the latest insider preview so I'm putting that down to a Windows bug that will be fixed soon in the mainstream channels
|
2025-04-01T04:10:38.405449
| 2016-03-22T08:21:13
|
142579311
|
{
"authors": [
"GPMDP-Magical-Unicorn",
"webshark25"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14901",
"repo": "MarshallOfSound/Google-Play-Music-Desktop-Player-UNOFFICIAL-",
"url": "https://github.com/MarshallOfSound/Google-Play-Music-Desktop-Player-UNOFFICIAL-/issues/723"
}
|
gharchive/issue
|
Uncaught Exception: Attempting to call a function in a renderer window that has been closed or released. Function provid
Nothing wrong actually happens... I just get this popup like every 5 minutes in the lower left hand corner which is really annoying. It would be awesome if you could remove the message! Thanks in advanced
An uncaught exception was reported.
Info:
OS: win32
Arch: ia32
GPMDP Version: 3.0.1
Time: Tue Mar 22 2016 01:20:35 GMT-0700 (US Mountain Standard Time)
Error:
Attempting to call a function in a renderer window that has been closed or released. Function provid
Error: Attempting to call a function in a renderer window that has been closed or released. Function provided here: Object.<anonymous> (C:\Users\Michael\AppData\Local\GPMDP_3\app-3.0.1\resources\app\build\inject\GPMInject\interface\mouseButtonNavigation.js:8:8.
at BrowserWindow.callIntoRenderer (C:\Users\Michael\AppData\Local\GPMDP_3\app-3.0.1\resources\atom.asar\browser\lib\rpc-server.js:201:19)
at emitTwo (events.js:92:20)
at BrowserWindow.emit (events.js:172:7)
This issue was created automatically inside the uncaughtException handler
Closing this automatically reported issue as a duplicate of #470
|
2025-04-01T04:10:38.414384
| 2024-01-08T19:10:07
|
2071062231
|
{
"authors": [
"MartinBernstorff"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14902",
"repo": "MartinBernstorff/herbarium",
"url": "https://github.com/MartinBernstorff/herbarium/pull/63"
}
|
gharchive/pull-request
|
dev: remove redundant ruff target version
dev: remove redundant ruff target version
dev: remove redundant ruff target version
[!WARNING]
This pull request is not mergeable via GitHub because a downstack PR is open. Once all requirements are satisfied, merge this PR as a stack on Graphite.
Learn more
Current dependencies on/for this PR:
main
PR #53
PR #55
PR #56
PR #57
PR #63 👈
PR #59
PR #60
PR #61
This stack of pull requests is managed by Graphite.
Merge activity
Jan 8, 8:13 PM: @MartinBernstorff merged this pull request with Graphite.
|
2025-04-01T04:10:38.415511
| 2024-06-01T23:53:41
|
2329394204
|
{
"authors": [
"dathbe"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14903",
"repo": "MartinGris/MMM-GoogleMaps-Tracking",
"url": "https://github.com/MartinGris/MMM-GoogleMaps-Tracking/issues/18"
}
|
gharchive/issue
|
API change 6/1/24?
As of this morning I have a "This page can't load Google Maps correctly" error. I've rebooted and no change. I have a hunch Google changed something on their end? Is it just me, or is this a thing?
I was finally able to look into this. It turns out my free trial had expired, and I hadn't completed billing information in some respect. So, my problem, not a problem with the module or Google Maps.
|
2025-04-01T04:10:38.419333
| 2021-08-31T13:38:22
|
983874938
|
{
"authors": [
"JkmAS",
"MartinMalinda"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14904",
"repo": "MartinMalinda/vue-concurrency",
"url": "https://github.com/MartinMalinda/vue-concurrency/issues/53"
}
|
gharchive/issue
|
AbortError is called even though task ended with success status
AbortError is called even though the task ended with success status. This happens to me on route change.
Maybe it's related to this issue https://github.com/MartinMalinda/vue-concurrency/issues/13
const task = useTask(function* (signal) {
const revert = (error?) => {
console.log('status': task.last?.status); // success
notifyError();
};
signal.pr.catch(revert); // <-- abort('cancel')
try {
yield apiTask.perform();
} catch (e) {
revert(e);
}
}).drop();
Fixed in 2.1.3 https://github.com/MartinMalinda/vue-concurrency/commit/76af0d55549d1ca39c0dedfd082af3fab007a907
|
2025-04-01T04:10:38.432017
| 2017-03-15T00:10:52
|
214244978
|
{
"authors": [
"MarvinTeichmann",
"baidut"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14905",
"repo": "MarvinTeichmann/KittiSeg",
"url": "https://github.com/MarvinTeichmann/KittiSeg/issues/24"
}
|
gharchive/issue
|
FileNotFoundError: [Errno 2] No such file or directory: 'DATA\data_road/val3.txt'
Run: python evaluate.py to evaluate the pretrained model
2017-03-15 08:05:37,981 INFO Graph loaded succesfully. Starting evaluation.
2017-03-15 08:05:37,981 INFO Output Images will be written to: RUNS\KittiSeg_pretrained\analyse\images/
Traceback (most recent call last):
File "evaluate.py", line 122, in <module>
tf.app.run()
File "D:\Python35\lib\site-packages\tensorflow\python\platform\app.py", line 44, in run
_sys.exit(main(_sys.argv[:1] + flags_passthrough))
File "evaluate.py", line 110, in main
ana.do_analyze(logdir)
File "incl\tensorvision\analyze.py", line 94, in do_analyze
hypes, sess, image_pl, inf_out)
File "RUNS\KittiSeg_pretrained\model_files\eval.py", line 61, in evaluate
with open(data_file) as file:
FileNotFoundError: [Errno 2] No such file or directory: 'DATA\\data_road/val3.txt'
This is a Windows path error. You can the file val3.txt in the folder data. And the script download_data should have copied val3.txt to the right location (line 133).
My guess would be, that you are able to fix this, by adjusting the paths in hypes/kittiseg.json to comply with Windows path notation (line 14).
Other then that, I would strongly recommend using Linux for deep learning ;).
Thank you
For windows users, please copy
$KITTISEG/DATA/val3.txt to $KITTISEG/DATA/data_road/val3.txt
$KITTISEG/DATA/testing.txt to $KITTISEG/DATA/data_road/testing.txt
Now both the demo and evaluate works fine ( Window 10 TensorFlow 1.0 cuda8.0 cudnn v5.1 python3.5)
[x] python demo.py --input_image data/demo/demo.png
[x] python evaluate.py
|
2025-04-01T04:10:38.435397
| 2022-06-09T07:24:05
|
1265709541
|
{
"authors": [
"MaskRay",
"Mic92"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14906",
"repo": "MaskRay/telegramircd",
"url": "https://github.com/MaskRay/telegramircd/pull/28"
}
|
gharchive/pull-request
|
Upgrade to working telethon version
The current version is no longer functional. The port might still have some bugs but at least I can read/write messages, download images and see the contact list.
(Thank you so much for porting. Will read..)
Sorry. It's a bit of a nightmare to review due to size. But I couldn't think of a better way to port it.
Some errors:
Traceback (most recent call last):
File "/home/maskray/projects/telegramircd/./telegramircd.py", line 299, in init
await web.channel_list()
File "/home/maskray/projects/telegramircd/./telegramircd.py", line 277, in channel_list
date = min(msg.date for msg in r.messages)
TypeError: '<' not supported between instances of 'datetime.datetime' and 'NoneType'
ERROR: resolve fwd_from TypeError("unhashable type: 'PeerChannel'")
Traceback (most recent call last):
File "/home/maskray/.local/lib/python3.9/site-packages/telethon/client/updates.py", line 467, in _dispatch_update
await callback(event)
File "/home/maskray/projects/telegramircd/./telegramircd.py", line 2059, in on_telegram_update
await getattr(TelegramUpdate, name)(self, update)
File "/home/maskray/projects/telegramircd/./telegramircd.py", line 1685, in UpdateNewChannelMessage
await server.on_telegram_update_message(update, update.message)
File "/home/maskray/projects/telegramircd/./telegramircd.py", line 2065, in on_telegram_update_message
sender, to = await self.resolve_from_to(msg)
File "/home/maskray/projects/telegramircd/./telegramircd.py", line 2049, in resolve_from_to
from_ = await server.ensure_special_user(msg.from_id, None)
File "/home/maskray/projects/telegramircd/./telegramircd.py", line 1986, in ensure_special_user
if user_id in self.user_id2special_user:
TypeError: unhashable type: 'PeerUser'
INFO: Sleeping for 24s (0:00:24) on GetParticipantsRequest flood wait
ERROR: resolve fwd_from TypeError("unhashable type: 'PeerChannel'")
ERROR: resolve fwd_from TypeError("unhashable type: 'PeerChannel'")
ERROR: resolve fwd_from TypeError("unhashable type: 'PeerChannel'")
ERROR: resolve fwd_from TypeError("unhashable type: 'PeerChannel'")
|
2025-04-01T04:10:38.450514
| 2021-01-18T22:31:44
|
788582534
|
{
"authors": [
"josephmancuso"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14907",
"repo": "MasoniteFramework/debugbar",
"url": "https://github.com/MasoniteFramework/debugbar/issues/5"
}
|
gharchive/issue
|
Javascript debugbar
Need a way to attach a debugbar to all templates of an application. Should just render an HTML snippet with some vanilla JS
done
|
2025-04-01T04:10:38.463351
| 2024-06-25T10:16:40
|
2372287851
|
{
"authors": [
"MastanSayyad",
"rajdesai17"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14908",
"repo": "MastanSayyad/Visual-Sort",
"url": "https://github.com/MastanSayyad/Visual-Sort/issues/23"
}
|
gharchive/issue
|
Excessive Whitespace in Footer - Visual Disruption
This issue reports excessive whitespace present at the bottom of the webpage's footer. This whitespace creates a visually unappealing layout and disrupts the overall design coherence.
@rajdesai17 someone already assigned for it, pls cgdck for other issues!
@rajdesai17 assigned it to you
@rajdesai17
changes you have to do are:
Remove the excessive whitespacing of footer
add transform: scale(1.2) hover to the links
Add links to my social icons , only for Linkdin and Github
Have you understand the task?? @rajdesai17
yes
|
2025-04-01T04:10:38.554402
| 2022-05-12T14:54:57
|
1234112575
|
{
"authors": [
"solomonshorser"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14909",
"repo": "MaterializeInc/pulumi-fivetran",
"url": "https://github.com/MaterializeInc/pulumi-fivetran/issues/2"
}
|
gharchive/issue
|
connection_type is not a part of ConnectorConfigArgs
It appears that connection_type is in DestinationConfigArgs, but based on the documentation (https://fivetran.com/docs/rest-api/connectors/config#configparameters_107) I think it should be in ConnectorConfigArgs.
UPDATE: It looks connection_type moved between v0.2.0 and the current version (v0.4.1) of the Terraform provider for Fivetran. Do you have any plans to rebuild/rerelease the pulumi-fivetran plugin?
I think this has been fixed in https://github.com/fivetran/terraform-provider-fivetran so syncing https://github.com/benesch/terraform-provider-fivetran/tree/main/fivetran would probably fix it.
This seems to have been resolved in more recent releases of this project.
|
2025-04-01T04:10:38.559266
| 2020-09-23T20:02:30
|
707643211
|
{
"authors": [
"MathewSachin",
"tigerdragon71"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14910",
"repo": "MathewSachin/Fate-Grand-Automata",
"url": "https://github.com/MathewSachin/Fate-Grand-Automata/issues/465"
}
|
gharchive/issue
|
Start quest button doesn't function until you switch to a new team
Before opening an issue:
[x] I tested the latest release
[x] I looked at other issues (even the closed ones)
[x] I read the Troubleshooting Guide
[x] I'll fill this template properly
Your issue will get closed/locked by the bot in case proper info isn't provided
If you open unnecessary/dumb issues, you'll be rewarded with appropriate tags.
Describe the bug
Everything else about the app works but whenever it tries to start up a quest after selecting the support, the start quest button will not function until you switch to a new team. Even manually pressing the button does nothing.
Screenshots
Video
**Device Info (please fill at-least 'Device Model's"):
Device model: Samsung galaxy s10+
Screen size: 6.4 in
Android Version: Android 10
RAM: 8gb
https://www.youtube.com/watch?v=q_QZFBIWgDU&feature=youtu.be
First of all, having to switch to a different team is a bug in FGO itself.
And, the app handles it well, but, you have configured the regions wrong.
The regions are totally wrong and the app is working by chance. Turn OFF Ignore notch calculation as it is not needed for your phone. Troubleshooting Guide/Game Area detection details on when you need to turn it ON.
|
2025-04-01T04:10:38.564560
| 2020-09-25T15:49:57
|
709041276
|
{
"authors": [
"HayanShiro",
"MathewSachin"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14911",
"repo": "MathewSachin/Fate-Grand-Automata",
"url": "https://github.com/MathewSachin/Fate-Grand-Automata/issues/473"
}
|
gharchive/issue
|
Support screen error (3rd time)
Before opening an issue:
[O] I tested the latest release
[O] I looked at other issues (even the closed ones)
[O] I read the Troubleshooting Guide
[O] I'll fill this template properly
Your issue will get closed/locked by the bot in case proper info isn't provided
If you open unnecessary/dumb issues, you'll be rewarded with appropriate tags.
Describe the bug
It gets stuck on supports screen even when FGA is full screen. I'm sorry for coming back with same issue for 3rd time but it keeps on stopping on me (×-×)
Screenshots
Video
https://youtu.be/_Gy_YYijvLI
Device Info (please fill at-least 'Device model'):
Device model: Samsung Galaxy S20+
Screen size: 6.7 inch
Android Version: Android 10
RAM: 12GB
Is it not the same video?
Post a video of when you made FGA full-screen.
https://youtu.be/6tIkXFqlKGc
Here it is, but it looks the same.
I'm sorry I uploaded the wrong vid(it'sthe same thing rhough). Here's the one with FGA full screen. https://youtu.be/NMkDJ5sL6rg
Something new happened when I recorded it with same settings. https://youtu.be/j2iod6Zaasc
Seems like that happens whenever I flip the phone in other direction so the black border is on ghe right and not left
I hope this issue gets solved soon so I could make the best out of this Roulette event.
I'll post one again in more organized form.
|
2025-04-01T04:10:38.573582
| 2020-08-06T14:39:15
|
674365903
|
{
"authors": [
"MathewSachin",
"homu-gamepress",
"reconman"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14912",
"repo": "MathewSachin/Fate-Grand-Automata",
"url": "https://github.com/MathewSachin/Fate-Grand-Automata/pull/236"
}
|
gharchive/pull-request
|
Option to take Screenshot of drops (Experimental)
for #188
When this option is ON, the Result screen is handled by a separate branch of code.
@homu-gamepress can you try the APK from this page: https://github.com/MathewSachin/Fate-Grand-Automata/actions/runs/197847815
There's a new option in More options you have to turn ON.
I got it to take a couple drop screenshots, but the current NA event mechanics is making it difficult to make a mass of runs. Will report back once I have a bit more to test with.
I tested a dozen runs on Ice Bridge and it seemed to screencap correctly:
https://imgur.com/a/HoaLTOB
Is everything supposed to be greyscale?
Will try to test whether it scrolls down correctly on large drop counts, once I can get a auto-able setup working for the event.
Is everything supposed to be greyscale?
yup, grayscale is used everywhere in the script because it's faster than having to deal with 3 colour channels.
I think that the guys over at Atlas Academy can't deal with grayscale at the moment.
The second drop screenshot is a bit high, maybe move the scroll down click down a few pixels?
Tried out the new version and it seemed to fixed the issue!
Click version is much better!
Previous scroll version:
@homu-gamepress Do you think this is good enough?
I just feel that it is too slow, not that I can think of any way to make it faster.
It's fast enough.
Bigger issue to testing whether greyscale screenshots causes any misreading for the drop image parser. Any way to take color screenshots?
Bigger issue to testing whether greyscale screenshots causes any misreading for the drop image parser. Any way to take color screenshots?
I can take a color screenshot, if it is full-screen including those blue-borders and notch.
Cropping it to just FGO for color screenshots too would require some parts of the code to be rewritten.
Yeah, full screen works just fine for AA.
Haven't checked screenshot using root, but for the MediaProjection one, I've checked that color screenshots are being taken.
Yep, it's in color! Awesome job!
I've confirmed that this works fine on Nox with Screenshot using root and JP server.
I guess if I match the scrollbar, we can avoid the unnecessary scroll and duplicate screenshot if there aren't that many drops.
|
2025-04-01T04:10:38.579428
| 2023-03-08T10:33:24
|
1615043755
|
{
"authors": [
"astromandev",
"bagelskat"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14914",
"repo": "MatrixTM/OutlookGen",
"url": "https://github.com/MatrixTM/OutlookGen/issues/31"
}
|
gharchive/issue
|
import errors
raceback (most recent call last):
File "/root/OutlookGen/main.py", line 3, in
from ctypes import windll
ImportError: cannot import name 'windll' from 'ctypes' (/usr/lib/python3.9/ctypes/init.py)
In linux I comment out line 3 in main.py seems to fix the problem.
|
2025-04-01T04:10:38.626413
| 2024-07-25T20:29:42
|
2430928697
|
{
"authors": [
"MatteoMagnini",
"mikiungolo"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14915",
"repo": "MatteoMagnini/hyperTensionBot",
"url": "https://github.com/MatteoMagnini/hyperTensionBot/pull/3"
}
|
gharchive/pull-request
|
Changed library for plot.
Added OxyPlot and removed ScottPlot for incompatibility. Advice: add legend to plot.
Now the plots are correctly generated. Well done!
|
2025-04-01T04:10:38.628674
| 2024-04-24T19:34:25
|
2262033445
|
{
"authors": [
"ivanistheone"
],
"license": "CC-BY-4.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14916",
"repo": "MatthewBJane/artifact-corrections-for-effect-sizes",
"url": "https://github.com/MatthewBJane/artifact-corrections-for-effect-sizes/issues/2"
}
|
gharchive/issue
|
Wrong labels on y-axis in Figure 6.3 panel
In Figure 6.3, the true correlation between X an Y is 0.5 (top-left panel of the figure).
But in the bottom left graph, it seems you're saying that when reliability of X goes up, $\rho_XY$ approaches 1.0
I believe the y-axis labels should be up to 0.5 to match the other panels in this figure.
Oh I think I know what's going on. The different curves represent different values of $\rho_{XY}$ [0,0.1,...,1.0].
So the figure is not wrong.
However, I would suggest just showing the case $\rho_{XY}=0.5$ (show only a single curve), to match the other three panels in the Figure.
|
2025-04-01T04:10:38.645462
| 2024-03-30T19:40:01
|
2216640516
|
{
"authors": [
"MatthiasValvekens",
"Salamek",
"blocher",
"rfabbri"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14917",
"repo": "MatthiasValvekens/pyHanko",
"url": "https://github.com/MatthiasValvekens/pyHanko/issues/416"
}
|
gharchive/issue
|
Signature invisible in Adobe Reader but visible in other viewers
Describe the bug
When I create a visible signature, I can see it in Preview and Skim, but not o Adobe Reader.
The visible signature display should be about halfway down, centralized in the page.
To Reproduce
pyhanko sign addsig --field 1/500,530,100,485/Signature1 pkcs12 in.pdf out.pdf cert.pfx
Open in Skim and Preview, the signature is visible, but Acrobat Reader does not display it.
Expected behavior
I expect all PDF viewers to display the visible part of the signature.
The pyhanko signature validates digitally, but it might pose bureaucratic impediments if the officer cannot see any visual mark.
Environment (please complete the following information):
OS: Mac OS 12.6.6 Monterey
Adobe Reader 2024.001.20615
Hi, this kind of problem is unfortunately impossible to diagnose without a copy of the specific input document that's giving you issues. Can you provide one?
I'm also assuming that the actual configuration you're using to populate the visible field is somewhere in your configuration file, can you share that as well?
I have encountered this behavior when AcroForm ->NeedAppearances is present and set to 'true', when document is not signed PDF acrobat reader displays signature field ok, but when i sign it using pyhanko, NeedAppereances is copied over with signature to this block:
<ORIGINAL_PDF>
%%EOF
2 0 obj
<<
/Producer (WeasyPrint 61\0562\073 pyHanko 0\05623\0561)
/ModDate (D\07220240408220437\05302\04700\047)
>>
endobj
3 0 obj
<<
/Type /Catalog
/Pages 1 0 R
/AcroForm <<
/Fields [ 7 0 R ]
/DR 4 0 R
/NeedAppearances true
/SigFlags 3
>>
<REST_OF_SIGNATURE>
And signature field become invisible, it also disappear from list of signatures... (So PDF in PDF acrobat reader looks like unsigned)
All other PDF readers are fine, just PDF acrobat reader is doing this..., when i remove NeedAppearances or set it to false in unsigned PDF and sign it using pyhanko all is well in PDF acrobat reader...
From my testing it looks like /NeedAppearances true carried over to signature block is the issue?
I'm also attaching my testing PDFs:
signed-61.2-need-app-false.pdf: Signed PDF where NeedAppearances was set to 'false' before signing (Work ok in PDF acrobat reader)
signed-61.2-need-app-removed.pdf: Signed PDF where NeedAppearances was removed before signing (Work ok in PDF acrobat reader)
signed-61.2-uncompressed.pdf: Signed PDF where NeedAppearances was set to 'true' (default behavior of my PDF * generator) this is broken in PDF acrobat reader
unsigned-61.2-uncompressed.pdf: Unsigned PDF where NeedAppearances was set to 'true' (default behavior of my PDF generator) this is displayed correctly in PDF acrobat reader
unsigned-54.2.pdf: Unsigned PDF where NeedAppearances is missing (Work ok in PDF acrobat reader)
signed-54.2.pdf: Signed PDF where NeedAppearances is missing (Work ok in PDF acrobat reader)
signed-61.2-uncompressed-app-removed-from-sig.pdf: Signed PDF where NeedAppearances was manually removed from signature block, signature is displayed correctly in PDF acrobat reader but PDF is corrupted (i did simple delete in text editor)
signed-61.2-need-app-false.pdf
signed-61.2-need-app-removed.pdf
signed-61.2-uncompressed.pdf
unsigned-61.2-uncompressed.pdf
unsigned-54.2.pdf
signed-54.2.pdf
signed-61.2-uncompressed-app-removed-from-sig.pdf
54.2 means PDF generated using Weasyprint 54.2
61.2 means PDF generated using Weasyprint 61.2
(I was dissecting missing signatures in PDF acrobat reader after weasyprint update)
Huh, interesting. If NeedAppearances is the culprit, that kind of tracks. It's deprecated in PDF 2.0 and probably ignored by all non-Acrobat processors anyhow. The PDF 2.0 requirement is that all fields need to supply appearance streams, so NeedAppearances is pretty much obsolete.
Nonetheless, I would assume that NeedAppearances only kicks in on fields that don't already have an appearance stream... The fact that that's not what you observe is indeed strange.
If systematically removing NeedAppearances solves the problem, I'm happy to do that (or perhaps hiding it behind a toggle). I don't really have the energy to hunt down all the quirks of Acrobat these days :)
Hmm i'm going to check if /AP is set for signature field...
Ok affected field does have AP specified...
@MatthiasValvekens i think that not passing NeedAppearances to signature block after EOF as a option should do the trick (not sure what should be the default behavior tho, up to you)
Thanks for that analysis. I think removing NeedAppearances by default is probably a sane thing to do. I plan to cut a release later this month, this can probably be squeezed in.
Thanks for this amazing library. I just wanted to check if there is any update on this?
|
2025-04-01T04:10:38.654391
| 2022-09-13T03:26:24
|
1370797457
|
{
"authors": [
"MauriceNino",
"lhoggatt17"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14918",
"repo": "MauriceNino/dashdot",
"url": "https://github.com/MauriceNino/dashdot/issues/370"
}
|
gharchive/issue
|
[Feature] Add Disk % Used to UI (without hovering)
Description of the feature
It would be cool if we could embed the % of disk used into the plots:
Then we wouldn't have to hover over the element. It would also make viewing the dashboard on a mobile device easier. You could make this the default behavior for the DASHDOT_ALWAYS_SHOW_PERCENTAGES = true option. If there was less than x% disk, you'd have to have it on the grey side. Otherwise, it would be nice to have it inside the green bar.
Additional context
No response
Hey, thanks for creating this issue. This sounds like a really great idea in theory, but I am having some issues with it:
The bars are responsive and change their size with the amount of drives as well as the amount of space available (display size). How would I make sure that the font is always viewable, centered and looks nice?
If there are some percentages on the green side and some on the gray side, it would look a bit inconsistent. But if I put them all on the left, they might overlap the border between green and gray + they might be completely on the gray side. This might not look too bad, though, so I think it could be a non-issue.
this is a good question. I suppose you could just turn the %s off when there are too many drives. You could also add a scroll bar here if this setting is enabled. Or just keep scaling the text smaller: even if it's too small to read, it's still there.
Very true, I guess having them overlap wouldn't be the end of the world. Personally, I think having the % information is worth it since the difference between 80% and 90% might be hard to tell visually, but the numbers help make decisions like "it's time to buy a bigger drive" or whatever. I know people have all these fancy systems to monitor this stuff, but a beautiful, yet simple dashboard is more than enough for a simple home server with a few drives.
Maybe I will make the bars limit to 3 per page (same as drives) and change them as well when the drives go to the next page. That would make sure there is enough space for this.
I will try this out, but currently it is a bit lower on my priority list (lots of bug-reports).
:tada: This issue has been resolved in version 4.5.0
Please check the changelog for more details.
|
2025-04-01T04:10:38.705264
| 2020-06-24T20:52:04
|
644955080
|
{
"authors": [
"MaybeShewill-CV",
"lev1khachatryan"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14920",
"repo": "MaybeShewill-CV/bisenetv2-tensorflow",
"url": "https://github.com/MaybeShewill-CV/bisenetv2-tensorflow/issues/14"
}
|
gharchive/issue
|
Explanation of the data augmentation strategy
Hi,
Could you explain the data augmentation strategy, what the following parameters mean:
FIX_RESIZE_SIZE: [720, 720] # (width, height), for unpadding
INF_RESIZE_VALUE: 500 # for rangescaling
MAX_RESIZE_VALUE: 600 # for rangescaling
MIN_RESIZE_VALUE: 400 # for rangescaling
MAX_SCALE_FACTOR: 2.0 # for stepscaling
MIN_SCALE_FACTOR: 0.75 # for stepscaling
SCALE_STEP_SIZE: 0.25 # for stepscaling
If you fix crop size for training and evaluating (by TRAIN_CROP_SIZE and EVAL_CROP_SIZE parameters), then why do you need above parameters ?
@lev1khachatryan I believe the comment has been pretty clear. Those params are used when you apply resize with different strategy (eg. rangescaling and stepscaling). Cropping operation was applied when the image was resized properly:)
|
2025-04-01T04:10:38.759197
| 2024-04-29T08:17:02
|
2268405968
|
{
"authors": [
"Iambestfeed",
"jeonsworld",
"vaibhavad"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14921",
"repo": "McGill-NLP/llm2vec",
"url": "https://github.com/McGill-NLP/llm2vec/issues/33"
}
|
gharchive/issue
|
Error when reproducing training
Because I want to make some tweaks to test with other models like Qwen, I tried running your repo and I encountered some problems as follows:
Train_config:
{
"model_name_or_path": "princeton-nlp/Sheared-LLaMA-1.3B",
"dataset_name": "iambestfeed/wiki-text-subset",
"per_device_train_batch_size": 1,
"per_device_eval_batch_size": 1,
"gradient_accumulation_steps": 1,
"do_train": true,
"do_eval": true,
"max_seq_length": 512,
"mask_token_type": "blank",
"data_collator_type": "default",
"mlm_probability": 0.2,
"overwrite_output_dir": true,
"output_dir": "output/mlm/Sheared-LLaMA-1.3B",
"evaluation_strategy": "steps",
"eval_steps": 100,
"save_steps": 200,
"stop_after_n_steps": 1000,
"lora_r": 16,
"gradient_checkpointing": true,
"torch_dtype": "bfloat16",
"attn_implementation": "flash_attention_2"
}
Code: Colab
You can go to my notebook shared above and take a look for more clarity. Dataset I created 1 subset from wiki-text to perform for testing.
@Iambestfeed Try transformer version==4.39.1
@jeonsworld's suggestion is correct. The attention implementation changes based on transformer version. We have followed the latest implementation present in 4.39.1 and above. This requirement is specified in setup.py
The llm2vec package has now been made compatible with transformers 4.40.0 and 4.40.1. Here is the fix.
You can use the latest version from pip (0.1.5) or install from source using the latest code.
Feel free to re-open the issue if you still encounter this error
|
2025-04-01T04:10:38.773170
| 2024-08-15T22:10:05
|
2469062223
|
{
"authors": [
"Mctalian",
"polki92"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14922",
"repo": "Mctalian/RPGLootFeed",
"url": "https://github.com/Mctalian/RPGLootFeed/issues/9"
}
|
gharchive/issue
|
Configuration for fonts
The font face and font size are locked to a static value.
Add a configuration option for the font face and/or font size.
I nearly had this done, but I kind of didn't like how things felt. I may revisit this later unless there are people who really want this.
I nearly had this done, but I kind of didn't like how things felt. I may revisit this later unless there are people who really want this.
I do personally since I can't read default font :-/
I modify your files after each update to add my own font tho :>
Sounds good! I'll take another look at this.
|
2025-04-01T04:10:38.782301
| 2017-10-24T17:26:34
|
268119070
|
{
"authors": [
"jeremy-lq",
"stuartlangridge"
],
"license": "apache-2.0",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14923",
"repo": "MeasureOSS/Measure",
"url": "https://github.com/MeasureOSS/Measure/issues/16"
}
|
gharchive/issue
|
Exclude forks from all statistics
GHCrawler on default settings fetches data for all forks of our core repos. Ignore all such data everywhere. (Repository dashboards already ignore it because they explicitly query for the repo name we care about, but contributor and root dashboards don't.)
(Ideally this would be fixed by not fetching that data at all, once https://github.com/MeasureOSS/Measure/issues/4 is implemented.)
After a chat with @stuartlangridge I am moving this to milestone .2. While it is an issue, the scope is constrained and well understood, and the fix entails not only a significant refactor but also some upstream GHCrawler work.
|
2025-04-01T04:10:38.797262
| 2023-10-24T12:38:13
|
1959163893
|
{
"authors": [
"TheocharisPetrou",
"emylonas",
"gabrielbodard"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14924",
"repo": "MedCyprus/MedCyprusEFES",
"url": "https://github.com/MedCyprus/MedCyprusEFES/issues/1"
}
|
gharchive/issue
|
Index of Ligatures
pulling from wrong attribute value (clarify)
Future: ligatures across words --Do this first
Future: accented letters (need explanation, example)
043.17 - Μήτηρ shouldn't be the ligature. It's ΜΗΡ. check
I agree. I checked the mark up of all the records and I think I marked up them correct. Perhaps the problem has to do with the fact that ΜΗΡ is ligature and at the same time it is also an abbreviation: Μή(τη)ρ / Μῆ(τε)ρ.
In other words, each entry in the ligature index is triggered by one of the following occurances:
<hi rend="ligature">ABC</hi>
<hi rend="ligature" xml:id="h001">MH</hi>TH<hi rend="ligature" xml:id="h002">P</hi><join targets="#h001 #h002"/>
non-diplomatic elements like <ex> and <corr> won't appear, but <supplied reason="lost"> should show up in the index.
Correct. However…
For the particular of the encoding of ligatures, our practice needs to be that if two parts of a ligature are divided in the edition, by any markup, spaces, restored/expanded/corrected characters, or anything else, it should be tagged as two <hi> elements, with a <join>. (As in my second example above.)
See e.g. 044.001: markup is <hi rend="ligature" xml:id="lig1"><choice><orig>Η</orig><reg>Ἡ</reg></choice></hi> <addName type="epithet" key="Pantanassa"><hi rend="ligature" xml:id="lig2">Π</hi><join target="#lig1 #lig2"/> (CORRECT), but H and Π appear as two separate ligatures in the index (and in the EpiDoc display, NB).
However in 060.001 the markup is incorrect, so not a good diagnostic of broken index. To check with Theo and Dora.
I have fixed the incorrect markup in 043,017 to read: <expan><abbr rend="supraline"><hi rend="ligature">Μ<choice><orig>η</orig><reg>ή</reg></choice></hi></abbr><ex>τη</ex><abbr rend="supraline"><hi rend="ligature">ρ</hi></abbr></expan> which should now be correct. We can test the index against this now.
Gabby, with the above updated markup the whole ligature ΜΗΡ of the abbreviation Μή(τη)ρ was separated in two upper ligature lines instead of one as in other examples we' ve done, e.g. 043.031
@emylonas Can you see if you can fix in (a) the XSLT and CSS and (b) the index, the issue of ligatures broken across words, such as two examples in line 5 of 044.053 (<hi rend="ligature" xml:id="lig1">ν</hi> <hi rend="ligature" xml:id="lig2">μ</hi><join target="#lig1 #lig2"/> and hi rend="ligature" xml:id="lig3">ὰ</hi>), which appear in diplomatic as separate loops above N, M, A and Π. They are also indexed as four separate, single-character ligatures.
This is likely to be hard, so if not fixable, (1) we might talk to Irene; (2) we might have to tolerate this bug (in the latter case we can still fix the index display for examples of <ex> inside <hi>, as above, which is a rarer instance of this problem).
|
2025-04-01T04:10:38.799681
| 2020-07-02T08:14:49
|
649731523
|
{
"authors": [
"flynnamy",
"hxcai"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14925",
"repo": "Media-Smart/flexinfer",
"url": "https://github.com/Media-Smart/flexinfer/issues/2"
}
|
gharchive/issue
|
core dump
I checked that volksdep has been installed and could be used. When I run python3 examples/classifier.py test.jpg,it shows core dump. How to fix this?
@flynnamy Hi, could you please provide your code and environment for analysis?
@flynnamy Hi, could you please provide your code and environment for analysis?
I use original code without modify. My environmet is Ubuntu 16.04, cuda 10.2, cudnn 8.0, pytorch 1.4.
@flynnamy We have updated the code and please download the newest one to use.
|
2025-04-01T04:10:38.839406
| 2024-05-31T17:35:39
|
2328300816
|
{
"authors": [
"khai-meetkai",
"localmind-ai",
"morgendigital"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14926",
"repo": "MeetKai/functionary",
"url": "https://github.com/MeetKai/functionary/issues/200"
}
|
gharchive/issue
|
Streaming with vLLM doesn't seem to work
I'm trying to use functionary (v2.4-medium-AWQ) with the vLLM OpenAI-compatible endpoint, however it seems that streaming doesn't work on the OpenAI-compatible Endpoint.
I have written a short script to evaluate if everything is working correctly, and it works perfectly without streaming, but with streaming, I don't get any output or I get an error when tools are involved.
✅ Without Streaming
Python Code
from openai import OpenAI
client = OpenAI(base_url="http://localhost:8000/v1", api_key="some-key")
completion = client.chat.completions.create(
model="meetkai/functionary-medium-v2.4-AWQ",
messages=[{"role": "user",
"content": "Hey, how are you doing?"}
],
tools=[{
"type": "function",
"function": {
"name": "get_current_weather",
"description": "Get the current weather",
"parameters": {
"type": "object",
"properties": {
"location": {
"type": "string",
"description": "The city and state, e.g. San Francisco, CA"
}
},
"required": ["location"]
}
}
}],
tool_choice="auto"
)
print(completion)
Response
ChatCompletion(id='cmpl-67bed551b2194741ae7325493a553abe', choices=[Choice(finish_reason='stop', index=0, logprobs=None, message=ChatCompletionMessage(content="I'm a computer program, so I don't experience emotions or have a personal life. However, I'm here to assist you with information, answer questions, and provide information to the best of my knowledge and abilities. How can I assist you today?", role='assistant', function_call=None, tool_calls=[], tool_call_id=None, name=None))], created=1717176536, model='meetkai/functionary-medium-v2.4-AWQ', object='chat.completion', system_fingerprint=None, usage=CompletionUsage(completion_tokens=58, prompt_tokens=141, total_tokens=199))
🛑 With Streaming
Python code
from openai import OpenAI
client = OpenAI(base_url="http://localhost:8000/v1", api_key="some-key")
completion = client.chat.completions.create(
model="meetkai/functionary-medium-v2.4-AWQ",
messages=[{"role": "user",
"content": "Hey, how are you doing?"}
],
tools=[{
"type": "function",
"function": {
"name": "get_current_weather",
"description": "Get the current weather",
"parameters": {
"type": "object",
"properties": {
"location": {
"type": "string",
"description": "The city and state, e.g. San Francisco, CA"
}
},
"required": ["location"]
}
}
}],
tool_choice="auto",
stream=True
)
for chunk in completion:
print(chunk.choices[0].delta)
Response
Traceback (most recent call last):
File "/usr/local/lib/python3.10/dist-packages/httpx/_transports/default.py", line 69, in map_httpcore_exceptions
yield
File "/usr/local/lib/python3.10/dist-packages/httpx/_transports/default.py", line 113, in __iter__
for part in self._httpcore_stream:
File "/usr/local/lib/python3.10/dist-packages/httpcore/_sync/connection_pool.py", line 367, in __iter__
raise exc from None
File "/usr/local/lib/python3.10/dist-packages/httpcore/_sync/connection_pool.py", line 363, in __iter__
for part in self._stream:
File "/usr/local/lib/python3.10/dist-packages/httpcore/_sync/http11.py", line 349, in __iter__
raise exc
File "/usr/local/lib/python3.10/dist-packages/httpcore/_sync/http11.py", line 341, in __iter__
for chunk in self._connection._receive_response_body(**kwargs):
File "/usr/local/lib/python3.10/dist-packages/httpcore/_sync/http11.py", line 210, in _receive_response_body
event = self._receive_event(timeout=timeout)
File "/usr/local/lib/python3.10/dist-packages/httpcore/_sync/http11.py", line 220, in _receive_event
with map_exceptions({h11.RemoteProtocolError: RemoteProtocolError}):
File "/usr/lib/python3.10/contextlib.py", line 153, in __exit__
self.gen.throw(typ, value, traceback)
File "/usr/local/lib/python3.10/dist-packages/httpcore/_exceptions.py", line 14, in map_exceptions
raise to_exc(exc) from exc
httpcore.RemoteProtocolError: peer closed connection without sending complete message body (incomplete chunked read)
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "/root/tests/vllm-mixed-functions-streaming.py", line 31, in <module>
for chunk in completion:
File "/usr/local/lib/python3.10/dist-packages/openai/_streaming.py", line 46, in __iter__
for item in self._iterator:
File "/usr/local/lib/python3.10/dist-packages/openai/_streaming.py", line 58, in __stream__
for sse in iterator:
File "/usr/local/lib/python3.10/dist-packages/openai/_streaming.py", line 50, in _iter_events
yield from self._decoder.iter_bytes(self.response.iter_bytes())
File "/usr/local/lib/python3.10/dist-packages/openai/_streaming.py", line 280, in iter_bytes
for chunk in self._iter_chunks(iterator):
File "/usr/local/lib/python3.10/dist-packages/openai/_streaming.py", line 291, in _iter_chunks
for chunk in iterator:
File "/usr/local/lib/python3.10/dist-packages/httpx/_models.py", line 829, in iter_bytes
for raw_bytes in self.iter_raw():
File "/usr/local/lib/python3.10/dist-packages/httpx/_models.py", line 883, in iter_raw
for raw_stream_bytes in self.stream:
File "/usr/local/lib/python3.10/dist-packages/httpx/_client.py", line 126, in __iter__
for chunk in self._stream:
File "/usr/local/lib/python3.10/dist-packages/httpx/_transports/default.py", line 112, in __iter__
with map_httpcore_exceptions():
File "/usr/lib/python3.10/contextlib.py", line 153, in __exit__
self.gen.throw(typ, value, traceback)
File "/usr/local/lib/python3.10/dist-packages/httpx/_transports/default.py", line 86, in map_httpcore_exceptions
raise mapped_exc(message) from exc
httpx.RemoteProtocolError: peer closed connection without sending complete message body (incomplete chunked read)
vLLM Output
ERROR: Exception in ASGI application
Traceback (most recent call last):
File "/usr/local/lib/python3.10/dist-packages/uvicorn/protocols/http/httptools_impl.py", line 426, in run_asgi
result = await app( # type: ignore[func-returns-value]
File "/usr/local/lib/python3.10/dist-packages/uvicorn/middleware/proxy_headers.py", line 84, in __call__
return await self.app(scope, receive, send)
File "/usr/local/lib/python3.10/dist-packages/fastapi/applications.py", line 1106, in __call__
await super().__call__(scope, receive, send)
File "/usr/local/lib/python3.10/dist-packages/starlette/applications.py", line 122, in __call__
await self.middleware_stack(scope, receive, send)
File "/usr/local/lib/python3.10/dist-packages/starlette/middleware/errors.py", line 184, in __call__
raise exc
File "/usr/local/lib/python3.10/dist-packages/starlette/middleware/errors.py", line 162, in __call__
await self.app(scope, receive, _send)
File "/usr/local/lib/python3.10/dist-packages/starlette/middleware/cors.py", line 83, in __call__
await self.app(scope, receive, send)
File "/usr/local/lib/python3.10/dist-packages/starlette/middleware/exceptions.py", line 79, in __call__
raise exc
File "/usr/local/lib/python3.10/dist-packages/starlette/middleware/exceptions.py", line 68, in __call__
await self.app(scope, receive, sender)
File "/usr/local/lib/python3.10/dist-packages/fastapi/middleware/asyncexitstack.py", line 20, in __call__
raise e
File "/usr/local/lib/python3.10/dist-packages/fastapi/middleware/asyncexitstack.py", line 17, in __call__
await self.app(scope, receive, send)
File "/usr/local/lib/python3.10/dist-packages/starlette/routing.py", line 718, in __call__
await route.handle(scope, receive, send)
File "/usr/local/lib/python3.10/dist-packages/starlette/routing.py", line 276, in handle
await self.app(scope, receive, send)
File "/usr/local/lib/python3.10/dist-packages/starlette/routing.py", line 69, in app
await response(scope, receive, send)
File "/usr/local/lib/python3.10/dist-packages/starlette/responses.py", line 270, in __call__
async with anyio.create_task_group() as task_group:
File "/usr/local/lib/python3.10/dist-packages/anyio/_backends/_asyncio.py", line 597, in __aexit__
raise exceptions[0]
File "/usr/local/lib/python3.10/dist-packages/starlette/responses.py", line 273, in wrap
await func()
File "/usr/local/lib/python3.10/dist-packages/starlette/responses.py", line 262, in stream_response
async for chunk in self.body_iterator:
File "/root/functionary/functionary/vllm_inference.py", line 254, in completion_stream_generator
async for response in generate_openai_format_from_stream_async(
File "/root/functionary/functionary/inference_stream.py", line 238, in generate_openai_format_from_stream_async
state, response = prompt_template.update_response_state_from_delta_text(
File "/root/functionary/functionary/prompt_template/prompt_template_v2.py", line 446, in update_response_state_from_delta_text
func_name = tool_choice.function.name
AttributeError: 'str' object has no attribute 'function'
INFO 05-31 17:30:19 async_llm_engine.py:120] Finished request cmpl-6dda581d394e4fa3a10189d83cbb13fa.
🛑 With Streaming, but with Tools removed
Here, I get a slightly different error.
INFO 05-31 17:34:35 metrics.py:229] Avg prompt throughput: 0.4 tokens/s, Avg generation throughput: 0.2 tokens/s, Running: 1 reqs, Swapped: 0 reqs, Pending: 0 reqs, GPU KV cache usage: 0.1%, CPU KV cache usage: 0.0%
INFO 05-31 17:34:37 async_llm_engine.py:120] Finished request cmpl-b548002a12f04b12b7f2a9bbc67634a2.
INFO 05-31 17:34:37 async_llm_engine.py:154] Aborted request cmpl-b548002a12f04b12b7f2a9bbc67634a2.
@morgendigital Thank you for reporting this issue, we are fixing this in a PR, will let you know when it is done
@morgendigital We have merged the PR fixing your issue, can you clone the latest code and try again?
Thank you very much @khai-meetkai , will check it later today!
@khai-meetkai works flawlessly, thank you very much! 👍 Will close it now.
Thank you @khai-meetkai - works flawlessly! Will close this issue now.
|
2025-04-01T04:10:38.848867
| 2017-11-29T16:57:14
|
277830373
|
{
"authors": [
"AlexHGaming",
"DeadLord1337",
"Luka967"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14927",
"repo": "Megabyte918/MultiOgar-Edited",
"url": "https://github.com/Megabyte918/MultiOgar-Edited/issues/1336"
}
|
gharchive/issue
|
1000th closed issue
I had achieved the 500th, now 1000th, congrats to me on the dead repo i guess
#1336
off by one
awww
eeee boy!!
|
2025-04-01T04:10:38.849723
| 2017-06-06T07:06:57
|
233797827
|
{
"authors": [
"AlexHGaming",
"ExobyteZOfficial"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14928",
"repo": "Megabyte918/MultiOgar-Edited",
"url": "https://github.com/Megabyte918/MultiOgar-Edited/issues/908"
}
|
gharchive/issue
|
Error.. At Team Mode
No Leaderboard at TeamMode?
Refer to #482.
@Megabyte918, you need to fix this lmao
|
2025-04-01T04:10:38.893119
| 2021-08-04T00:59:20
|
959724691
|
{
"authors": [
"WANGCHAO1996",
"kdpan",
"ruinmessi"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14929",
"repo": "Megvii-BaseDetection/YOLOX",
"url": "https://github.com/Megvii-BaseDetection/YOLOX/issues/345"
}
|
gharchive/issue
|
AssertionError: Caught AssertionError in DataLoader worker process 0.
Original Traceback (most recent call last):
File "/home/panfulu/anaconda3/envs/atom3d/lib/python3.6/site-packages/torch/utils/data/_utils/worker.py", line 287, in _worker_loop
data = fetcher.fetch(index)
File "/home/panfulu/anaconda3/envs/atom3d/lib/python3.6/site-packages/torch/utils/data/_utils/fetch.py", line 44, in fetch
data = [self.dataset[idx] for idx in possibly_batched_index]
File "/home/panfulu/anaconda3/envs/atom3d/lib/python3.6/site-packages/torch/utils/data/_utils/fetch.py", line 44, in
data = [self.dataset[idx] for idx in possibly_batched_index]
File "/home/panfulu/YOLOX/yolox/data/datasets/datasets_wrapper.py", line 121, in wrapper
ret_val = getitem_fn(self, index)
File "/home/panfulu/YOLOX/yolox/data/datasets/mosaicdetection.py", line 91, in getitem
img, _labels, _, _ = self._dataset.pull_item(index)
File "/home/panfulu/YOLOX/yolox/data/datasets/coco.py", line 105, in pull_item
assert img is not None
AssertionError
Img is not read correctly. Plz check your dataset
Img is not read correctly. Plz check your dataset
how to solve it
|
2025-04-01T04:10:38.896071
| 2021-08-09T08:31:35
|
963765026
|
{
"authors": [
"CallmeAG",
"FateScript",
"gyh420",
"kail85"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14930",
"repo": "Megvii-BaseDetection/YOLOX",
"url": "https://github.com/Megvii-BaseDetection/YOLOX/issues/426"
}
|
gharchive/issue
|
ImportError: cannot import name 'COCODataset' from 'yolox.data'
in yolox/exp/yolox_base.py, at below line
def get_data_loader(self, batch_size, is_distributed, no_aug=False):
from yolox.data import (
COCODataset,
TrainTransform,
YoloBatchSampler,
DataLoader,
InfiniteSampler,
MosaicDetection,
)
Got an error 'ImportError: cannot import name 'COCODataset' from 'yolox.data''
Also tried
from yolox.data.datasets.coco import COCODataset
also got an error
cannot find module 'yolox.data.datasets.coco'.
Anyone please suggest? Thanks.
I Got this error too. Do you solved it yet?
I have tried it on my machine and it works.
In [1]: from yolox.data import COCODataset
In [2]: COCODataset?
Init signature: COCODataset(*args, **kwds)
Docstring: COCO dataset class.
Init docstring:
COCO dataset initialization. Annotation data are read into memory by COCO API.
Args:
data_dir (str): dataset root directory
json_file (str): COCO json file name
name (str): COCO data name (e.g. 'train2017' or 'val2017')
img_size (int): target image size after pre-processing
preproc: data augmentation strategy
File: /data/workspace/YOLOX/yolox/data/datasets/coco.py
Type: type
Subclasses:
Did you install the package through python3 -m pip install -v -e .?
how to fix it?
|
2025-04-01T04:10:38.927214
| 2020-07-04T12:19:13
|
650881887
|
{
"authors": [
"MelbourneDeveloper",
"ep0x",
"matejcik"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14931",
"repo": "MelbourneDeveloper/Trezor.Net",
"url": "https://github.com/MelbourneDeveloper/Trezor.Net/issues/34"
}
|
gharchive/issue
|
After latest Trezor firmware update, not possible to sign TX.
We have problem to sign tx for both models.
Far as i inspected issue, i found :
While sending TxRequest to Trezor One, new requirement is needed such as TxRequest.RequestType.Txmeta ( TXMETA).
Also while doing debugging in our service we noticed when we send TxRequest to device, before device ask us for TXMETA data, now in request.Details.TxHash is not empty like before. There is new things happening now that we need to change.
Now we need to verify prevHas or something. I did a bit inspection in https://github.com/trezor/trezor-firmware/blob/82c0c403677a14bde56c1a6b440a46593a57267d/python/src/trezorlib/btc.py and i found out there is changes in code for txmeta and tx.bin.output added python: Add get_ownership_id() and get_ownership_proof() to trezorlib.
In TrezorLib, inputs value(amount) is not mandatory now if i understand python correctly...
def from_json(json_dict):
def make_input(vin):
i = messages.TxInputType()
if "coinbase" in vin:
i.prev_hash = b"\0" * 32
i.prev_index = 0xFFFFFFFF # signed int -1
i.script_sig = bytes.fromhex(vin["coinbase"])
i.sequence = vin["sequence"]
else:
i.prev_hash = bytes.fromhex(vin["txid"])
i.prev_index = vin["vout"]
i.script_sig = bytes.fromhex(vin["scriptSig"]["hex"])
i.sequence = vin["sequence"]
return i
def make_bin_output(vout):
o = messages.TxOutputBinType()
o.amount = int(Decimal(vout["value"]) * (10 ** 8))
o.script_pubkey = bytes.fromhex(vout["scriptPubKey"]["hex"])
return o
t = messages.TransactionType()
t.version = json_dict["version"]
t.lock_time = json_dict.get("locktime")
t.inputs = [make_input(vin) for vin in json_dict["vin"]]
t.bin_outputs = [make_bin_output(vout) for vout in json_dict["vout"]]
return t
About TX_META
def copy_tx_meta(tx):
tx_copy = messages.TransactionType(**tx)
# clear fields
tx_copy.inputs_cnt = len(tx.inputs)
tx_copy.inputs = []
tx_copy.outputs_cnt = len(tx.bin_outputs or tx.outputs)
tx_copy.outputs = []
tx_copy.bin_outputs = []
tx_copy.extra_data_len = len(tx.extra_data or b"")
tx_copy.extra_data = None
return tx_copy
if res.request_type == R.TXMETA:
msg = copy_tx_meta(current_tx)
res = client.call(messages.TxAck(tx=msg))
And here is trick with request for TXOUTPUT. Now he is checking if there is request.Details.TxHash and if is has is presented then bin_outputs must be set...
elif res.request_type == R.TXOUTPUT:
msg = messages.TransactionType()
if res.details.tx_hash:
msg.bin_outputs = [current_tx.bin_outputs[res.details.request_index]]
else:
msg.outputs = [current_tx.outputs[res.details.request_index]]
res = client.call(messages.TxAck(tx=msg))
In Trezor Model T there is error in firmware message when you try to sign TX.
I am willing to help but idk from where to start...
I noticed there is a lot problems with new firmware for many wallets and services like Wasabi and BtcPayServer, but idk did he fixed issue, he used HWI for his Wasabi project.
I am willing to help if i can in anyway.
It would be awesome if we can join all together and maintain one Lib for Trezor instead 5 different libs at once and having the same issue...
I'd be glad if we could join this project, our respected colleagues like @NicolasDorier and @lontivero @bitcoinbrisbane
Any help are welcome!
Thank you in advance.
@ep0x you probably upgraded the trezor firmware right?
Trezor.net needs to upgrade. You can help by regenerating the protobuf code. You can get that from Satoshi labs.
Otherwise I will do it eventually
@MelbourneDeveloper Correct, all our devices has been updated and we encounter problems with signing TX.
protobuf will fix problems with signing TX?
For regenerating protobuf coode, did you used protobuf compiler to regenerate? I am not exactly sure how to do it.
@ep0x yes. Time to learn!
Here is an old article about it. It's out of date but it's a starting point.
https://christianfindlay.com/2018/01/31/developing-software-for-satoshi-labs-trezor-hardware-wallet-part-1/
Otherwise you can wait for me.
@ep0x BTW I figured this out with no documentation and no support so you can as well.
@ep0x the other option is that you can pay me to fix it.
I didn't ask for payment from your side in this case https://github.com/MelbourneDeveloper/KeepKey.Net/issues/6
I did donate more than 400$...
but np i can donate again :) fix Tx Signing and donation is coming is that ok for you?
@ep0x you donated?
Sorry, can you point me to the transaction please? If you sent $400 that's a big donation.
@ep0x please DM me through my website or Twitter and we can negotiate
@MelbourneDeveloper I sent an email to you from prev conversation :) u have tx in email ;) but sure for your work i will donate again.
@MelbourneDeveloper my mistake, around 60$ was donation. Sorry whole transaction was around 500$
Hi @ep0x, i've been talking with @MelbourneDeveloper and I'm willing to get paid to fix this.
However I am somewhat unclear on the details. There is no specific "sign transaction" method in the Trezor.NET library, and this issue only talks about a particular test case failing. ISTM that Trezor.NET as it stands is fully capable of signing transactions, if the caller can do it.
I can fix the test case, of course. But is that it? I'm assuming there is some other caller code in your application that will need modification as well?
|
2025-04-01T04:10:38.934308
| 2019-03-06T21:03:25
|
418007637
|
{
"authors": [
"malytomas",
"p3tronex"
],
"license": "BSD-2-Clause",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14932",
"repo": "Melown/vts-browser-unity-plugin",
"url": "https://github.com/Melown/vts-browser-unity-plugin/issues/4"
}
|
gharchive/issue
|
Collisions
When i drive my car on a road then at some parts the collisions are like a ramp and the trees are blocking the way in cities.
Hi, the vts plugin is designed to work with data that you provide to it.
Can you share url to the mapconfig you have used and vts position at which you have problems?
A screenshot may help too.
Thanks.
Well, its not one location. The roads and the trees have bad collisions in Krnov > Opava > Ostrava ( Whole cities ) I'ma get the position
Closing for inactivity
|
2025-04-01T04:10:39.073766
| 2024-02-25T14:17:34
|
2152786274
|
{
"authors": [
"Skyxim",
"daiaji",
"hamjin"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14934",
"repo": "MetaCubeX/mihomo",
"url": "https://github.com/MetaCubeX/mihomo/pull/1061"
}
|
gharchive/pull-request
|
Make private ip the same as geoip:private
Currently ip.IsPrivate and netip.Addr.IsPrivate only matches a part of private IPs. This pr uses the CIDR list from geoip:private to fix it.
https://datatracker.ietf.org/doc/html/rfc1918#section-3
https://datatracker.ietf.org/doc/html/rfc1918#section-3
Some network (eg in office) resolves websites to <IP_ADDRESS>/16, which breaks dns.fallback-filter's ip.IsPrivate
其实geoip:private里面的地址不应该叫做Private IP Addresses，而应该是Reserved IP addresses。 个人认为排除所有的Reserved IP addresses相对来说应该更好一些。
确实在理
Reserved IP addresses
这看上去更应该修改geoip-code，让它支持多个code
|
2025-04-01T04:10:39.087992
| 2023-01-04T17:54:19
|
1519366996
|
{
"authors": [
"Gudahtt",
"mcmire"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14935",
"repo": "MetaMask/core",
"url": "https://github.com/MetaMask/core/pull/1038"
}
|
gharchive/pull-request
|
Expose prev state in state change event payloads
ADDED:
When listening for controller state changes, it is useful to be able to tell whether an individual property in state changed. However, this is not possible currently because the state change event only exposes the state after it has been changed, not before. This commit adds a third argument to the stateChange listener so that the previous state can be accessed.
Checklist
[x] Tests are included if applicable
[x] Any added code is fully documented
The patches were intended to show whether a specific property has changed. Could those serve this purpose instead?
@Gudahtt Good question. I suppose you could use patches, but if all you're interested in is "did this property change in some way", you'd have to figure out how to test for that given how patches fundamentally work, and I worry not only that this would be a footgun but also how readable that check would end up being. I believe this would be the correct way to make that check:
const promiseForCountDidChange = new Promise<boolean>((resolve) => {
controllerMessenger.subscribe(
'CountController:stateChange',
(_newState, patches) => {
const countDidChange = patches.some(
(patch) => patch.path.length === 0 || patch.path[0] === 'count',
);
resolve(countDidChange);
},
);
});
// ... do the thing ...
await promiseForCountDidChange;
If we feel that everyone should be familiar with how patches work in Immer, then I suppose we don't need another way to do this. But for the sake of argument, compare the above to the simplicity of:
const promiseForCountDidChange = new Promise<boolean>((resolve) => {
controllerMessenger.subscribe(
'CountController:stateChange',
(newState, _patches, prevState) => {
resolve(newState.count !== prevState.count);
},
);
});
// ... do the thing ...
await promiseForCountDidChange;
I see, yeah that makes sense. That is quite a footgun.
I don't love the idea of exposing the same information in two different ways. And it's unfortunate that we'd be introducing another positional parameter that might be useful without the previous ones. I'm also uncertain how referencing the previous state would affect memory usage.
As an alternative, perhaps we could create a function for checking whether a property has changed, e.g.
// in our "utils" module
function patchPropertyHasChanged(property: string, patches: Patch[]): boolean {
return patches.some(
(patch) => patch.path.length === 0 || patch.path[0] === property,
);
}
// somewhere else
import { patchPropertyHasChanged } from '@metamask/utils';
controllerMessenger.subscribe(
'CountController:stateChange',
(_newState, patches) => {
if (patchPropertyHasChanged('count', patches)) {
onCountChanged();
}
},
);
Or we could build this into the base controller if it's common enough to be worth the added complexity:
controllerMessenger.subscribe(
'CountController:statePropertyChange',
'count',
(_newState, _patches) => {
onCountChanged();
},
);
Thoughts on either of those approaches?
I'm not totally opposed to adding a previousState but if we could avoid duplicate information that would be ideal. And making previousState nicer by updating the stateChanged callback to take an object rather than positional params would be quite a substantial breaking change (because the first parameter, state, is also an object so breakages might go unnoticed).
@Gudahtt Okay, that's fair. I am also loathe to expose the same information two different ways. I think we can start out by introducing a function in @metamask/utils as you recommended and maybe link to that in the JSDocs for subscribe. I will close this in the meantime!
|
2025-04-01T04:10:39.095990
| 2022-07-26T10:21:32
|
1318022924
|
{
"authors": [
"Akatori-Design",
"SaraCheikh"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14936",
"repo": "MetaMask/design-tokens",
"url": "https://github.com/MetaMask/design-tokens/issues/202"
}
|
gharchive/issue
|
[GF UI] Create Infotip Button component
Description
Create a Figma component infotip button. The button toggles the visibility of the overlay. The overlay contains the info content.
Technical Details
-create infotip button: text+Button Icon (with "info" icon)
-create tooltip component
-variants (states): default, hover, pressed
-Variant: Large screen=true/false
-guide (use case): usage
-guide (use case): positioning
Spec Acceptance Criteria
[ ] Guide has three sections : Anatomy, Dimension, Use Cases. ( Can omit boards not applicable)
[ ] Guide’s anatomy contains all necessary component references & links.
[ ] Each visual variants ( types, states) have a visual representation, so dev can inspect.
[ ] Visual references have correct tokens & styles applied (color, shadow, text).
[ ] Component is added to Notion table with matching name, with Figma link.
[ ] Resolve old/outstanding comments.
All Figma components & Boards have consistent & correct Namings
(Component Naming System : https://www.notion.so/Component-Naming-System-Draft-6d59eb23d42b483f9f4648506fdb1c29)
Design Component Acceptance Criteria
[ ] Components have necessary variants with consistent naming, and clear labeling.
[ ] Components have correct Design Tokens assigned: Color, typography, shadows and spacing.
[ ] Components are responsive with text overflow handled.
[ ] Components are tested in a prototype and viewed in its respective context (mobile or extension)
[ ] Resolve old/outstanding comments.
[ ] Component is added to Notion table with status set to “ Figma Done”, and has a Figma link.
[ ] Add update details to the “Upcoming” release note.
References
https://ebay.gitbook.io/mindpatterns/disclosure/infotip
-https://clarity.design/documentation/signposts
Created a Data component for figma design
component called "Key + Value"
https://www.figma.com/file/HKpPKij9V3TpsyMV1TpV7C/DS-Components?node-id=9790%3A37079
|
2025-04-01T04:10:39.099192
| 2019-08-17T18:24:24
|
481913294
|
{
"authors": [
"alexhidalgo",
"calvinh8",
"estebanmino",
"rekmarks",
"tkostantinov-digihey"
],
"license": "ISC",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14937",
"repo": "MetaMask/eth-contract-metadata",
"url": "https://github.com/MetaMask/eth-contract-metadata/pull/353"
}
|
gharchive/pull-request
|
feat(add-token): MM-000 add STORM ERC-20 token + linting
- add storm token logo
- add address and description
- remove trailing spaces
https://www.stormx.io/
https://etherscan.io/token/0xd0a4b8946cb52f0661273bfbc6fd0e0c75fc6433
@tkostantinov-digihey could you fix conflicts please? Thanks
@estebanmino conflict resolved.
Hi @estebanmino, I wonder if we can get this merged in so we can have STORM showing in MetaMask? Thanks!
Hi @estebanmino @pi0neerpat @ososco could we please have you review this PR? We are about to launch our Rewards Program which has MetaMask as our preferred method for users to connect to our app.
Checks out.
@rekmarks thanks for the quick response. This will really help us with our Rewards Program launch this week.
You got it! This should be in production by early next week at the latest.
|
2025-04-01T04:10:39.160750
| 2018-06-15T12:10:34
|
332750254
|
{
"authors": [
"NicolasDorier",
"shwuhk"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14938",
"repo": "MetacoSA/NBitcoin",
"url": "https://github.com/MetacoSA/NBitcoin/issues/445"
}
|
gharchive/issue
|
The value of Difficulty calculated in Target class is wrong
https://github.com/MetacoSA/NBitcoin/blob/9843ab333f93394d70287bec3f3754610abe7897/NBitcoin/Target.cs#L90-L111
All leading zeros in the decimal part will be omited due to decimalPart.Multiply(BigInteger.Ten); as decimalPart is initialized as zero.
You are right. I am surprised no tests caught this. Luckily this number is only used for human consumption.
I wonder where I took this code from. Even the double.Parse is really fishy.
Reviewing https://www.youtube.com/watch?v=6TDLMkOCQkU that's embarassing... ;p
fixed in <IP_ADDRESS>
|
2025-04-01T04:10:39.161971
| 2017-12-23T12:09:56
|
284303331
|
{
"authors": [
"NicolasDorier",
"cryptodeveloperin"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14939",
"repo": "MetacoSA/QBitNinja",
"url": "https://github.com/MetacoSA/QBitNinja/issues/45"
}
|
gharchive/issue
|
Please check there is some error on during gettransaction and all method call
Please check there is some error on during gettransaction and all method call.
Getting one or more error. Please can you check?
no errror on my side, please send me the repro.
Resolved in 15 minute...getting longer time to connect and so error generation.
|
2025-04-01T04:10:39.163222
| 2020-09-02T11:24:27
|
690940257
|
{
"authors": [
"hedgehog1833"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14940",
"repo": "MetalDetectorRocks/metal-detector-main",
"url": "https://github.com/MetalDetectorRocks/metal-detector-main/pull/204"
}
|
gharchive/pull-request
|
dateTo cannot be set without dateFrom in ReleasesRequests
I put another method in the WithTimeRangeValidation interface, please check the naming. I wasn't so sure there...
Bad Request if we pass only dateFrom
|
2025-04-01T04:10:39.166075
| 2020-11-14T13:39:30
|
743005480
|
{
"authors": [
"Eric-Sunny-Fish",
"pudepiedj",
"songhuaixu"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14941",
"repo": "MetalKit/metal",
"url": "https://github.com/MetalKit/metal/issues/9"
}
|
gharchive/issue
|
Running Playground Chapters from MetalKit in MacOS
Marius, I've very much enjoyed your MetalKit sequence and learned much from it, so thank you.
From the Chapter 8 code accompanying the blogs onwards where you use playgrounds rather than projects I find that the path search in Metal.swift using guard let path = never manages to find the Shaders.metal file inside Resources, so the code won't build. Replacing the whole line with an absolute path fixes it, but I am not sure whether this is my setup or something that has changed in Swift 5.3, XCode 12.0 or somewhere else both of which I am using.
All the fixes for this error on SO seem to relate to XCode projects rather than playgrounds, and I haven't found any suggestions about how to fix the problem directly in the playground files and avoid introducing an absolute path.
Thanks.
The reason to add the .metal extension to the file is so that Xcode can recognize the file is a Metal file, and display the file with syntax highlighting. Nothing in the process of running the file requires the extension. An alternative way to mark this as a Metal file is to set the type to Metal Shader Source in the File Inspector.
Bundle.main.path(forResource:, ofType:) appears to be broken. It will not find a file of type "metal" it will find a file of type "meta" or "mmmmm" or no type. It appears to do something wrong with .metal files.
So a work around for this issue is to manually set the type of Shaders.metal to Metal Shader Source, remove the .metal extension, and change the type to nil in Bundle.main.path(forResource:, ofType:).
In Xcode 14, device.makeDefaultLibrary() can function normally
|
2025-04-01T04:10:39.183368
| 2023-01-26T01:08:05
|
1557483921
|
{
"authors": [
"CrytoPal"
],
"license": "CC0-1.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14942",
"repo": "MeteorDevelopment/meteor-addon-template",
"url": "https://github.com/MeteorDevelopment/meteor-addon-template/issues/29"
}
|
gharchive/issue
|
Can't Build
Can't Build the Addon
"Could not find meteor-client-0.5.2-SNAPSHOT.jar (meteordevelopment:meteor-client:0.5.2-SNAPSHOT:20230123.224945-43).
Searched in the following locations:
https://maven.meteordev.org/snapshots/meteordevelopment/meteor-client/0.5.2-SNAPSHOT/meteor-client-0.5.2-20230123.224945-43.jar"
Fixed the issue by deleting the cache folder in Home directory.
|
2025-04-01T04:10:39.282747
| 2016-08-13T16:10:18
|
171015289
|
{
"authors": [
"Lunamick",
"MichMich"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14943",
"repo": "MichMich/MagicMirror",
"url": "https://github.com/MichMich/MagicMirror/issues/398"
}
|
gharchive/issue
|
Magic Mirror wont boot properly
It seems like its running. For support Please use http://forum.magicmirror.builders.
|
2025-04-01T04:10:39.329902
| 2015-04-01T20:33:16
|
65773432
|
{
"authors": [
"MichaelHatherly",
"peter1000"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14944",
"repo": "MichaelHatherly/Lexicon.jl",
"url": "https://github.com/MichaelHatherly/Lexicon.jl/pull/38"
}
|
gharchive/pull-request
|
Fixes the example: file not found: docs/api/Lexicon.md
in mkdocs.yml it is specified with capital L. https://github.com/MichaelHatherly/Lexicon.jl/blob/master/mkdocs.yml#L10
Thanks!
|
2025-04-01T04:10:39.332155
| 2017-01-17T13:26:41
|
201282926
|
{
"authors": [
"MichaelRocks",
"abraaocaldas"
],
"license": "apache-2.0",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14945",
"repo": "MichaelRocks/paranoid",
"url": "https://github.com/MichaelRocks/paranoid/issues/1"
}
|
gharchive/issue
|
Not compatible with Windows
Hello, I´m trying to use it on Android Studio for windows, but I cannot because it generate a long path that windows don´t recognize.
On linux this problem do not occurs.
Hi @abraaocaldas, thank you for reporting the issue. I've just publish a new version that should work fine on Windows. Unfortunately, I don't have Windows installed and cannot ensure if the issue is really fixed. It would be great if you can check the new version and tell me whether or not the issue is fixed.
Yep! It's working, thank you very much!
|
2025-04-01T04:10:39.338156
| 2018-05-18T09:30:25
|
324336636
|
{
"authors": [
"MichaelSolati",
"morrislaptop"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14946",
"repo": "MichaelSolati/geofirestore",
"url": "https://github.com/MichaelSolati/geofirestore/pull/2"
}
|
gharchive/pull-request
|
Add ability to attach data to locations. Fixes #1
Inspired by firebase/geofire-js#45
@MichaelSolati looks like connection errors with Travis. Not sure how to trigger rebuild
@morrislaptop I'll trigger a rebuild whenever you give me a thumbs up (also I'm going to probably merge this into another branch as I'll want to make some tweaks before I green-light it into master)
@morrislaptop I did run a rebuild where it still failed. I'll be personally working on this over the weekend though.
@morrislaptop make two small changes for me por favor, instead of calling the field data please call it document, and I don't want it to be an optional field (it'll be mandatory). These are to make this better inline with version 2.0. (thank you!)
@MichaelSolati ok thanks:
Should I keep setWithData and getWithData or you want me to make document mandatory in the existing get and set methods?
The setWithData method has data optional so that you can specify batch writes with the first parameter - see usage at https://github.com/MichaelSolati/geofirestore/pull/2/files#diff-c93f0b9b40c8b2599e5845d94ec20139R587 - is this OK?
@morrislaptop I see your point here... hrm... ok well keep it optional, but rename it to document for now. In v2.0 it'll be mandatory and there will be a slew of changes across the board. But for now this should work just fine.
Keep the setWithData and getWithData methods, however rename any place you use data/Data to document/Document (setWithData => setWithDocument). The interface with data/document being listed as just d is fine. I just want to use the verbiage of document to be used across the package as it best aligns with the whole Firestore collection/document verbiage.
@MichaelSolati all done :)
|
2025-04-01T04:10:39.365649
| 2018-03-12T10:30:04
|
304315775
|
{
"authors": [
"DominikDitoIvosevic",
"gudrunleos",
"hrolfurh"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14947",
"repo": "Microsoft/AL",
"url": "https://github.com/Microsoft/AL/issues/1770"
}
|
gharchive/issue
|
Error in translation file when label variables have the same name
There seems to be a bug in the translation file when variables have the same name, even though they are in different actions. Below is a sample where I have a Label variable in three different actions.
action("Import Contracts")
{
[...]
trigger OnAction();
var
DialogMsg : Label 'Import Contracts';
begin
[...]
end;
}
action("Export Contracts")
{
[..]
trigger OnAction();
var
DialogMsg : Label 'Export Contracts';
begin
[...]
end;
}
action("Download Contract Template")
{
[...]
trigger OnAction();
var
DialogMsg : Label 'Download Contracts Template';
begin
[...]
end;
}
The XLIFF file that generates has three <trans-unit> nodes with the same ID. This causes error in translation tools and services and might have some other weird interactions. When I changed the variable names to be different in each action the problem was fixed.
@hrolfurh thank you for this nasty edge case!
Right now we identify a variable by it's name, it's parent name (this case trigger) and the application object.
Indeed we need to remember the action name as well.
Is this solved in build 21229?
|
2025-04-01T04:10:39.378658
| 2018-03-20T09:32:45
|
306788008
|
{
"authors": [
"StanislawStempin",
"TKapitan",
"Tharangac",
"qutreson",
"srinivasdes",
"wcld"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14948",
"repo": "Microsoft/AL",
"url": "https://github.com/Microsoft/AL/issues/1816"
}
|
gharchive/issue
|
SchemaUpdateMode = synchronize failed with code 422. Reason: Value cannot be null.
Hello,
Facing the following error while publishing the app to sandbox.
SchemaUpdateMode=synchronize failed with code 422. Reason: Value cannot be null. Parameter name: value
Al Version : 0.14.17461
Can someone suggest me on what i am missing
Launch.json file :
{
"version": "0.2.0",
"configurations": [
{
"name": "Microsoft cloud sandbox",
"type": "al",
"request": "launch",
"startupObjectId": 22,
"schemaUpdateMode": "Synchronize",
}
]
}
Hi Qutreson,
Thanks for your response.
what is meant by repro steps. Can you please explain bit more
Thanks
By repro steps, I mean a sequence of actions for us to do in order to trigger this error. Or, if you cannot find a clear one, could you provide us with more details about the context in which the error happened? (For instance, the code you had that trigger this error, what where you trying to do, if it was the first time you were publishing, if it happened only once or still reproduces regularly etc)
This way, it makes it easier for us to reproduce, understand and fix the issue you are describing
I have done development with AL and fixed the compile related issues. while i am trying to publish my application to the sandbox i am facing this error.
SchemaUpdateMode=synchronize failed with code 422. Reason: Value cannot be null. Parameter name: value
Please find the attachment..
Hi @srinivasdes, what happens if you created a new empty extension with AL:Go using the same launch.json and try to publish it to the same sandbox environment? Do you also get the same error? If not, can you try to cut down your extension until you get a minimal extension that reproduces the error?
Hi Qutreson,
I can able to publish my test application with the same launch.json . I didn't get any error , But as you said i can not cut down my extension until i get a minimal extension, Because there are some dependencies between the al objects.
Can i have any other alternate
Thanks
So, you can either try to update to a newer version of the extension (in case this problem has already been fixed), or, you can also try to publish using the schema update mode recreate in your launch.json (but you will lose the data you have in your tables, see Retaining table data after publishing).
Current extension version is 0.14.17461. I couldn't find the update option to update the extension.
I have modified the launch.json and using the schema update mode recreate , Still having the same issue
any other suggestion..
Hello,
I have the same issue
Testing against NAV 2018/CU03 with AL 15.18771
[2018-04-03 11:23:16.68] Sending request to http://localhost:7059/ArtexAddOn-NAV2018/dev/apps?SchemaUpdateMode=recreate
[2018-04-03 11:23:17.80] The request for path /ArtexAddOn-NAV2018/dev/apps?SchemaUpdateMode=recreate failed with code 422. Reason: Value cannot be null.
Parameter name: value
@TKapitan - if you're using NAV 2018 you should be using the AL Language extension that comes with the product installer (version 0.12.XXX). Newer versions are for Business Central only.
See the FAQ: https://github.com/Microsoft/AL/wiki/Frequently-Asked-Questions#which-version-of-the-al-language-extension-should-i-use
@StanislawStempin
Maybe there is something wrong?
The new AL Extension (version 0.15.18771) has different XLIFF file than the older one. The old version had only source-language - then the XLIFF editor (we use "Transolution Xliff Editor") allows defining the target language at the start of translating. The new version has source-language and target language. Both contain same values (en-US). Than the XLIFF editor presumes the target language will be en-US and is not possible to redefine the language (so the only possibility how to translate is to change target-language manually)
When I change the target-language to something different (like cz-CS), the XLIFF editor is OK.
Another thing is the error "failed with code 422. Reason: Value cannot be null. Parameter name: value". I resolved it with the same - the error is happening because of the same values of source and target languages - it can be solved by changing the target-language in xliff.
Hello Friends;
I want to tell about of this Error : The request for path /DynamicsNAV110/dev/apps?SchemaUpdateMode=synchronize failed with code 422. Reason: TableExtension 50111 :: Unsupported field change. Field:MyComment; Change:Remove
We can solve This error
Please Go to Launch.json
Change "schemaUpdateMode": "Synchronous" to "schemaUpdateMode": "Recreate"
Save and Create Package:
Thank you
Had the same issue. Managed to resolve by deleting translation file (.xlf)
|
2025-04-01T04:10:39.382491
| 2018-04-07T18:10:44
|
312229520
|
{
"authors": [
"StanislawStempin",
"vitaliizz"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14949",
"repo": "Microsoft/AL",
"url": "https://github.com/Microsoft/AL/issues/1961"
}
|
gharchive/issue
|
Action Groups in menu on Role Center
Is there any way to group menu actions similar to "Manual Setup"?
I expect to see "Sales Group" as an Action Group which I can expand/collapse to show/hide "Sales Orders Test" action when I write the code below. Right now "Sales Group" is simply ignored.
Thanks
I reached out for input to the owning team
Here is the related documentation from MSDN.
https://docs.microsoft.com/en-us/dynamics-nav/setting-up-the-home-button-and-home-items
An item is automatically positioned as a child to another item according to the following conditions:
The child item targets a list page that displays filtered data from a table.
The parent item targets the same list page as the child item, except that list page data is not filtered.
The parent item is defined by an action, not a Cue.
|
2025-04-01T04:10:39.383428
| 2017-08-02T14:29:22
|
247408883
|
{
"authors": [
"NicolaiEsbensen",
"esbenk"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14950",
"repo": "Microsoft/AL",
"url": "https://github.com/Microsoft/AL/issues/450"
}
|
gharchive/issue
|
The type or method 'ClearLastError' cannot be used for 'Extension' development.
The header says it all, why is this method not usable in extension development?
I will be.
|
2025-04-01T04:10:39.396635
| 2019-04-05T03:29:48
|
429568497
|
{
"authors": [
"JohanStenberg100",
"adrianordonez"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14951",
"repo": "Microsoft/AL",
"url": "https://github.com/Microsoft/AL/issues/4845"
}
|
gharchive/issue
|
code unit 0145 E-Invoice Mgt. and codeunit 10147 E-Invoice Object Factory
Hello i need to view the code for this two code units in order to create the best way to integrate standar funcionality of Mexico E-Invoice to another PACs
Hi @adrianordonez, thanks for filing this issue. There will be improvements in this area after we stop shipping C/SIDE, but you can today reference the code unit inside your code, make sure you have a running server connected to your VS Code instance, and then right click and "Go To Definition" to see the code. Another option would be to use C/SIDE.
Thanks for the reply, one last doubt on cside i got the error that my user do not have permissions to design on the code unit, this enviroment was created with Sandbox container option from BC hosted locally, where can i assign the correct permissions to my user?
Are you unsuccessful using "Go To Definition"? I think that's the easiest way to do it.
|
2025-04-01T04:10:39.398399
| 2019-04-07T03:26:04
|
430100460
|
{
"authors": [
"TharunMuthu"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14952",
"repo": "Microsoft/ALAppExtensions",
"url": "https://github.com/Microsoft/ALAppExtensions/issues/2007"
}
|
gharchive/issue
|
How to modify existing field caption (Source Expression is variables ) in Pages !
Hi ,
Can you please help me to modify existing field caption (Source Expression is variables ) in Pages !
(Highlighted in yellow colour in the pic , needs to be change )
Note:
Using W1 Version (CU 4)
I did some workaround so consider as case closed !
|
2025-04-01T04:10:39.400366
| 2018-08-12T23:31:59
|
376365703
|
{
"authors": [
"JesperSchulz",
"geeknikolai"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14953",
"repo": "Microsoft/ALAppExtensions",
"url": "https://github.com/Microsoft/ALAppExtensions/issues/220"
}
|
gharchive/issue
|
Allow the Payment Reconciliation Journal to match against existing Bank Ledger Entries
I put an idea onto the ideas page (https://experience.dynamics.com/ideas/idea/?ideaid=89d21a1e-bc9a-e811-b96f-0003ff68935d) but I was suggested to put it here instead/as well.
The Payment Reconciliation Journal has a severe limitation in that it does not recognise when a Bank Ledger has already been posted - the most common example is Vendor Payments. We should be able to match against these unreconciled Bank Ledger entries, and ideally it should try to auto-match as well.
In comparison the Bank Acc. Reconciliation does do this - including the auto-match. Behind the scenes it uses the same Table and Posting Codeunit as the other Bank Rec so this should be possible?
Having the issue logged on the bcideas site is sufficient. Your request is getting tracked through that site. Closing this issue.
|
2025-04-01T04:10:39.406212
| 2017-03-14T09:25:12
|
214011471
|
{
"authors": [
"cijothomas",
"dnduffy",
"follesoe",
"xperiandri"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14954",
"repo": "Microsoft/ApplicationInsights-aspnetcore",
"url": "https://github.com/Microsoft/ApplicationInsights-aspnetcore/issues/359"
}
|
gharchive/issue
|
Update Getting Started page in Wiki for the 2.0 version of the library
The Getting Started page says that you should call services.AddApplicationInsightsTelemetry(Configuration); in ConfigureServices, while the method is marked as obsolete. The change log states that you can simply remove the methods.
So; if I remove the call to AddApplicationInsightsTelemetry, and builder.AddApplicationInsightsSettings(developerMode: true); is wrapped in a if (env.IsDevelopment()) check, is there really no need to do any registration in Startup.cs?
The methods that are obsolete and which should be removed are UseApplicationInsightsRequestTelemetry and UseApplicationInsightsExceptionTelemetry from the Configure method in Startup.cs.
You still need to call one of the services.AddApplicationInsightsTelemetry(***) overloads in the ConfigureServices method in Startup.cs.
Yes and the need of .UseApplicationInsights() in Program.cs is not clear.
Official Azure documentation redirects to wiki and wiki is outdated. Very bad!
I need to update the documentation, it is important, but unfortunately I haven't been able to get to it yet.
@xperiandri @follesoe
I have put together a wiki for 2.0. Please let us if this helps!
https://github.com/Microsoft/ApplicationInsights-aspnetcore/wiki/Getting-Started-for-a-ASP.NET-CORE2.0-WebApp
Sorry for not keeping documentation up-to-date. Will prioritize this for all further changes.
|
2025-04-01T04:10:39.416204
| 2018-11-16T00:03:22
|
381391633
|
{
"authors": [
"oco27"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14955",
"repo": "Microsoft/ApplicationInsights-dotnet",
"url": "https://github.com/Microsoft/ApplicationInsights-dotnet/pull/1006"
}
|
gharchive/pull-request
|
Fix a typo in a method name
Any idea why check #20181116.3 failed ? I don't have access to the details.
|
2025-04-01T04:10:39.427457
| 2018-03-12T07:06:21
|
304260720
|
{
"authors": [
"JasonSowers",
"Kartik1607"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14956",
"repo": "Microsoft/BotBuilder",
"url": "https://github.com/Microsoft/BotBuilder/issues/4299"
}
|
gharchive/issue
|
[Feature Request] Prompt choice with images
I am trying to make a feedback dialog. As a current solution we send a keyword from bot to our custom directline webpage. The page intercepts the keyword as shows the form like this
But this approach limits the bot to a single platform and adds a dependency to Directline.
As a second approach I tried to implement the same using an adaptive card. This is when we ran into some of shortcomings.
How do we show images that change when selected? For instance when a user clicks a rating, we would like to update the same image. Currently the bot can send an action, and we can find out the rating clicked but there was no way to notify the user for same on frontend.
Each image needs to have a Action to implement it. We now have an additional task of handling action clicked on image and on submit button. An updated choiceset API would be optimal for this use case.
[Update] To handle the situation gracefully, I broke the flow into 2 parts. First for getting the rating. Then, handle the comments totally separately.
Thank you for opening an issue against the Bot Framework SDK v3. As part of the Bot Framework v4 release, we’ve moved all v3 work to a new repo located at https://github.com/microsoft/botbuilder-v3. We will continue to support and offer maintenance updates to v3 via this new repo.
From now on, https://github.com/microsoft/botbuilder repo will be used as hub, with pointers to all the different SDK languages, tools and samples repos.
As part of this restructuring, we are closing all tickets in this repo.
For defects or feature requests, please create a new issue in the new Bot Framework v3 repo found here:
https://github.com/microsoft/botbuilder-v3/issues
For Azure Bot Service Channel specific defects or feature requests (e.g. Facebook, Twilio, Teams, Slack, etc.), please create a new issue in the new Bot Framework Channel repo found here:
https://github.com/microsoft/botframework-services/issues
For product behavior, how-to, or general understanding questions, please use Stackoverflow.
https://stackoverflow.com/search?q=bot+framework
Thank you.
The Bot Framework Team
|
2025-04-01T04:10:39.429535
| 2019-01-29T22:18:54
|
404512371
|
{
"authors": [
"cwhitten",
"ikivanc"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14957",
"repo": "Microsoft/BotFramework-Emulator",
"url": "https://github.com/Microsoft/BotFramework-Emulator/issues/1274"
}
|
gharchive/issue
|
Adding a service via Bot Emulator should be more easy to discover
In order to edit encrypted .bot configuration file Bot Emulator is a great place to modify .bot configuration file. This is a good alternative to msbot cli tool.
However this should be more easy to discover on Emulator. Maybe a UI tutorial at first run would be great.
hello @ikivanc - we hope our efforts to remove direct interaction with .bot files in our targeted 4.4 release will alleviate the pain you're experiencing. closing for now
|
2025-04-01T04:10:39.437550
| 2016-03-03T01:14:42
|
138044117
|
{
"authors": [
"Yongqu",
"boingoing"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14958",
"repo": "Microsoft/ChakraCore",
"url": "https://github.com/Microsoft/ChakraCore/pull/432"
}
|
gharchive/pull-request
|
module runtime change #3
Implement the remaining operations in SourceTextModuleRecord, including mainly ResolveExport and GetExportNames operations. Start implementation of ModuleNamespace. Implemented GetModuleNamespace. Detailed namespace implementation is not in yet. Filled in some holes in earlier implementation, like error handling for top level operations, notify the parent module if the current module failed in parse or ModuleDeclarationInitialization. Export method for bytecode generator to retrieve the localexport slot information. Fixed some potential memory leak by implementing the Finalize method for SourceTextModuleRecord. Temporary disable ModuleEvaluation for non-root modules, pending bytecode generator change.
@pleath @boingoing @abchatra can you take a look?
Looks good to me! Big change to module record, almost support everything besides namespace object. Cool!
|
2025-04-01T04:10:39.441835
| 2019-02-27T19:21:55
|
415285916
|
{
"authors": [
"TheFusion21",
"pkristof",
"stanard"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14959",
"repo": "Microsoft/DirectX-Graphics-Samples",
"url": "https://github.com/Microsoft/DirectX-Graphics-Samples/issues/534"
}
|
gharchive/issue
|
HRRESULT 0x887A0005
I have an Issue that seem to be that m_pResource is nullptr when Texture::Create is called
D3D12-capable hardware found: NVIDIA GeForce GTX 1080 Ti (11127 MB)
Changing native resolution to 1280x720
Changing display resolution to 1280x720
HRESULT failed in "d:\downloads\directx-graphics-samples-master\miniengine\core\texturemanager.cpp" @ 55
--> hr = 0x887A0005
That HRESULT is DXGI_ERROR_DEVICE_REMOVED. Have you run in Debug with Validation enabled to see if there were usage errors preceding this call?
I hope you meant SetEnableGPUBasedValidation
if yes here's the output:
Exception thrown at 0x00007FFFE30EA388 in D3D12RaytracingMiniEngineSample.exe: Microsoft C++ exception: _com_error at memory location 0x00000066B13BC7A8.
Exception thrown at 0x00007FFFE30EA388 in D3D12RaytracingMiniEngineSample.exe: Microsoft C++ exception: _com_error at memory location 0x00000066B13BCC00.
Exception thrown at 0x00007FFFE30EA388 in D3D12RaytracingMiniEngineSample.exe: Microsoft C++ exception: _com_error at memory location 0x00000066B13BC7A8.
Exception thrown at 0x00007FFFE30EA388 in D3D12RaytracingMiniEngineSample.exe: Microsoft C++ exception: _com_error at memory location 0x00000066B13BCC00.
D3D12: Removing Device.
Exception thrown at 0x00007FFFE30EA388 in D3D12RaytracingMiniEngineSample.exe: Microsoft C++ exception: _com_error at memory location 0x0000004D56D9C8A8.
Exception thrown at 0x00007FFFE30EA388 in D3D12RaytracingMiniEngineSample.exe: Microsoft C++ exception: _com_error at memory location 0x0000004D56D9CD00.
Exception thrown at 0x00007FFFE30EA388 in D3D12RaytracingMiniEngineSample.exe: Microsoft C++ exception: _com_error at memory location 0x0000004D56D9C8A8.
Exception thrown at 0x00007FFFE30EA388 in D3D12RaytracingMiniEngineSample.exe: Microsoft C++ exception: _com_error at memory location 0x0000004D56D9CD00.
D3D12: Removing Device.
D3D12 ERROR: ID3D12Device::RemoveDevice: Device removal has been triggered for the following reason (DXGI_ERROR_DEVICE_HUNG: The Device took an unreasonable amount of time to execute its commands, or the hardware crashed/hung. As a result, the TDR (Timeout Detection and Recovery) mechanism has been triggered. The current Device Context was executing commands when the hang occurred. The application may want to respawn and fallback to less aggressive use of the display hardware). [ EXECUTION ERROR #232: DEVICE_REMOVAL_PROCESS_AT_FAULT]
Exception thrown at 0x00007FFFE30EA388 in D3D12RaytracingMiniEngineSample.exe: Microsoft C++ exception: _com_error at memory location 0x00000088418FBE58.
Exception thrown at 0x00007FFFE30EA388 in D3D12RaytracingMiniEngineSample.exe: Microsoft C++ exception: _com_error at memory location 0x00000088418FC2B0.
Exception thrown at 0x00007FFFE30EA388 in D3D12RaytracingMiniEngineSample.exe: Microsoft C++ exception: _com_error at memory location 0x00000088418FBE58.
Exception thrown at 0x00007FFFE30EA388 in D3D12RaytracingMiniEngineSample.exe: Microsoft C++ exception: _com_error at memory location 0x00000088418FC2B0.
D3D12: Removing Device.
D3D12 ERROR: ID3D12Device::RemoveDevice: Device removal has been triggered for the following reason (DXGI_ERROR_DEVICE_HUNG: The Device took an unreasonable amount of time to execute its commands, or the hardware crashed/hung. As a result, the TDR (Timeout Detection and Recovery) mechanism has been triggered. The current Device Context was executing commands when the hang occurred. The application may want to respawn and fallback to less aggressive use of the display hardware). [ EXECUTION ERROR #232: DEVICE_REMOVAL_PROCESS_AT_FAULT]
Exception thrown at 0x00007FFFE30EA388 in D3D12RaytracingMiniEngineSample.exe: Microsoft C++ exception: _com_error at memory location 0x000000546E4FC248.
Exception thrown at 0x00007FFFE30EA388 in D3D12RaytracingMiniEngineSample.exe: Microsoft C++ exception: _com_error at memory location 0x000000546E4FC6A0.
Exception thrown at 0x00007FFFE30EA388 in D3D12RaytracingMiniEngineSample.exe: Microsoft C++ exception: _com_error at memory location 0x000000546E4FC248.
Exception thrown at 0x00007FFFE30EA388 in D3D12RaytracingMiniEngineSample.exe: Microsoft C++ exception: _com_error at memory location 0x000000546E4FC6A0.
D3D12 MESSAGE: Device Debug Layer Startup Options: GPU-Based Validation is enabled (disabled by default). This results in new validation not possible during API calls on the CPU, by creating patched shaders that have validation added directly to the shader. However, it can slow things down a lot, especially for applications with numerous PSOs. Time to see the first render frame may take several minutes. [ INITIALIZATION MESSAGE #1016: CREATEDEVICE_DEBUG_LAYER_STARTUP_OPTIONS]
Exception thrown at 0x00007FFFE30EA388 in D3D12RaytracingMiniEngineSample.exe: Microsoft C++ exception: _com_error at memory location 0x000000F4DBCF9A60.
Exception thrown at 0x00007FFFE30EA388 in D3D12RaytracingMiniEngineSample.exe: Microsoft C++ exception: _com_error at memory location 0x000000F4DBCFA898.
Exception thrown at 0x00007FFFE30EA388 in D3D12RaytracingMiniEngineSample.exe: Microsoft C++ exception: _com_error at memory location 0x000000F4DBCFAE00.```
There's a known issue with never NVIDIA drivers causing a TDR in the Raytracing Miniengine sample. Use 398.92 driver or older as a workaround as per the sample's limitations section.
|
2025-04-01T04:10:39.464310
| 2016-05-22T17:40:05
|
156163336
|
{
"authors": [
"walbourn",
"yaocatez"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14960",
"repo": "Microsoft/DirectXTK",
"url": "https://github.com/Microsoft/DirectXTK/issues/34"
}
|
gharchive/issue
|
Wrong Measurements from MeasureString
Hi Developers,
I have notice that the MeasureString returns a wrong size. The string "W" is used to measure the string size. The spritefont file is generated using MakeSpriteFont tool. The font setting is
FontName: Arial
FontStyle: Bold
FontSize: 185
The returned result for the width and the height are 233 and 283.642578 respectively. "W" is placed at (0,0) I have edited Image 1 to include a black dotted line and marked a red circle to indicate the position where my mouse positioned which reflects the values shown in the Window Title. Clearly, there is a big difference between the Vector2 returned and the size which I had measured.
Image 1:
In addition, when I used SpriteBatch to draw "W" on the screen (0,0) using origin (0,0), it does not draw at the correct place. Instead, there is some sort of spacing as shown in Image 2.
Image 2:
To test it further, I have used SpriteBatch to draw "W" on the screen at the bottom left (i.e. making it right justified) of the screen using the origin calculated via MeasureString, but it does not draw at the correct place as shown in Image 3. Instead, it is lifted and cut off.
Image 3:
Which version of DirectX Tool Kit are you using? There was a bug with MeasureString that was fixed in the Febraury 2016 release.
I am using:
Id: directxtk_windowsstore_8_1
Version: 2<IP_ADDRESS>
For your additional information:
I am developing Win32 application using Visual Studio 2013, but instead I am using directxtk_windows_8_1 (Version: 2<IP_ADDRESS>) as using the directxtk_desktop_2013 (Version: 2<IP_ADDRESS>) gives me error LNK1104:cannot open file 'x3daudio.lib'
The different NuGet packages for DirectX Tool Kit are built for specific platforms, and make different assumptions. If you are writing a "classic" Win32 desktop application, you should not use the directxtk_windowsstore_8_1 NuGet package.
For Windows Store 8.1, the application requires Windows 8.1 or later, and so directxtk_windowsstore_8_1 makes use of XAudio 2.8 in the Windows 8.1 SDK. Similarly for UWP, the application requires Windows 10, and so directxtk_uwp makes use of XAudio 2.9 in the Windows 10 SDK. If you are making a Win32 classic desktop application, then I assume you are wanting Windows 7 compatibility. Therefore, if you use the ```directxtk_desktop_201x`` NuGet package, any use of DirectX Tool Kit for Audio requires the legacy DirectX SDK. If you want specifically to write a Win32 classic desktop application that requires Windows 8 or later and thus uses XAudio 2.8, you can do so but you should use project-to-project references--this is a rare scenario, so I don't have a NuGet package specifically for this one. This is covered in detail on the wiki.
In any case I'm able to repro your issue and I'm investigating.
I think this was a bad fix, but I need to go back and figure out why I needed this change to make other things work.
Chuck Walbourn, thanks for the explanations. Actually, I am shifting from DirectX9 to DirectX11 and happened to found DirectXTK. At the same time, I am not going to use the legacy SDK (June 2010) to support my game engine even though it is only used for building very simple 2D games. Using DirectXTK makes the conversion easier as I am not very technical or familiar with the Pipeline related knowledge/experience.
I do have the intention of converting it to Windows Store App style instead of Win32. However, based on my research, I found that I may also have to pick up C++/CX. Currently, I decided to convert it from DirectX9 (Win32) to DirectX11 (Win32) using DirectXTK first before I further my research on Windows Store App and the Pipeline. This may explains why my 2D Game Engine (Win32) uses directxtk_windowsstore_8_1 instead of directxtk_desktop_2013 for DirectXTK's Audio as it works for this code snippet.
Another question aside from Audio and Spritefont is: Do you have the intention of making DirectXTK's documentation similar to the MSDN style or is it advisablefor us to cross reference DirectXTK with XNA?
As for the SpriteFont, I will wait for you to fix it.
Hi Chuck,
Has the said issue been resolve?
I'm wrapping a robust fix this weekend. You can download the library and make the source edit yourself as well.
Ok noted, in that case, i will wait for you to release all the fix at one go. Thanks,
Sync to the master branch to pick up the fixes. Reverted MeasureString to old behavior, and added MeasureDrawString.
RECT MeasureDrawBounds(_In_z_ wchar_t const* text, XMFLOAT2 const& position) const;
RECT XM_CALLCONV MeasureDrawBounds(_In_z_ wchar_t const* text, FXMVECTOR position) const;
Can you roughly estimate when will this fix be pushed to the NuGet Package?
I generally wait until enough work has accumulated to justify a full release, but I can accelerate a publish to NuGet if that's required. I'll look at doing one this week.
alright, noted with thanks.
reminder for NuGet release. Thanks.
Now that June 2016 is done, I'll be doing all the NuGet publishing for it tomorrow.
Published to NuGet:
This version is also available on NuGet as version 2<IP_ADDRESS>
Universal Windows Platform apps using Windows Tools 1.4 (10586)
Windows desktop app using VS 2013 Update 5
Windows desktop app using VS 2015 Update 3
Windows 8.1 Store
Windows phone 8.1
yes thanks!
Hi Chuck,
I am getting this error message: "This project references NuGet package(s) that are missing on this computer. Enable NuGet Package Restore to download them. For more information, see http://go.microsoft.com/fwlink/?LinkID=322105. The missing file is packages\directxtk_windowsstore_8_1.2<IP_ADDRESS>\build\native\directxtk_windowsstore_8_1.targets."
The solution I had discovered is to remove the lines that contains the April NuGet release. This is what is being removed:
<Import Project="packages\directxtk_windowsstore_8_1.2<IP_ADDRESS>\build\native\directxtk_windowsstore_8_1.targets" Condition="Exists('packages\directxtk_windowsstore_8_1.2<IP_ADDRESS>\build\native\directxtk_windowsstore_8_1.targets')" />
<Error Condition="!Exists('packages\directxtk_windowsstore_8_1.2<IP_ADDRESS>\build\native\directxtk_windowsstore_8_1.props')" Text="$([System.String]::Format('$(ErrorText)', 'packages\directxtk_windowsstore_8_1.2<IP_ADDRESS>\build\native\directxtk_windowsstore_8_1.props'))" />
<Error Condition="!Exists('packages\directxtk_windowsstore_8_1.2<IP_ADDRESS>\build\native\directxtk_windowsstore_8_1.targets')" Text="$([System.String]::Format('$(ErrorText)', 'packages\directxtk_windowsstore_8_1.2<IP_ADDRESS>\build\native\directxtk_windowsstore_8_1.targets'))" />
Is there a better way to do it rather than editing it manually in the *****.vcxproj?
As I said earlier in this thread, you shouldn't be using directxtk_windowsstore_8_1 for a Windows desktop project. In my experience, the NuGet Manager is able to do the Update as long you have restored the old package.
|
2025-04-01T04:10:39.485140
| 2017-10-19T10:29:56
|
266794756
|
{
"authors": [
"Almost-Done",
"omaralcheikh"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14961",
"repo": "Microsoft/HoloJS",
"url": "https://github.com/Microsoft/HoloJS/issues/104"
}
|
gharchive/issue
|
Promise does not resolve
Hello,
It seems that promises are not resolving when calling Promise.resolve(), an error is being thrown.
Failure in file {PATH}}\holojs\holojshost\ScriptErrorHandling.h, line 22
Here is a sample of what I am trying to do
TestPromise().then(function (text) {
console.log('resolved');
console.log(text);
});
function TestPromise() {
return new Promise(function (resolve, reject) {
console.log('entered promise');
resolve('some text');
});
}
'entered promise' is being logged and then the error is thrown.
Best Regards,
Omar.
There was no promise continuation handler in native code. Please let me know if the fix is not complete or if you encounter more issues with promises.
Thank you Cristi. It works as fine.
|
2025-04-01T04:10:39.490359
| 2018-07-26T13:31:50
|
344845383
|
{
"authors": [
"codingdna2",
"dishantmshah"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14962",
"repo": "Microsoft/InventorySample",
"url": "https://github.com/Microsoft/InventorySample/issues/34"
}
|
gharchive/issue
|
Main Window switch Page during Secondary Window Consolidation
Repro steps
If I break on source code while a secondary window is open, it navigates the main window to a blank page. I've traced the issue to this piece of code:
private void OnViewConsolidated(ApplicationView sender, ApplicationViewConsolidatedEventArgs args)
{
ViewModel.Unsubscribe();
ViewModel = null;
Bindings.StopTracking();
frame.Navigate(typeof(Page));
var appView = ApplicationView.GetForCurrentView();
appView.Consolidated -= OnViewConsolidated;
ServiceLocator.DisposeCurrent();
}
As a workaround I've commented the frame.Navigate line.
Actual Behavior
Following is a video on how to reproduce the issue:
System
Visual Studio 2017
Windows 10 Build 1803
Related to this issue there's one more (let me know if I should create an additional issue).
The secondary Window is kept on the Live Tree and it's not disposed. I've added to the above method a call to Window.Current.Close(); in order to fix it but I don't know if it's the right solution. Please advise.
@codingdna2 Solution you suggest seems to be working correctly, you can raise a PR with the fix.
|
2025-04-01T04:10:39.494141
| 2017-04-24T14:12:53
|
223833724
|
{
"authors": [
"Laurae2",
"guolinke",
"nnormandin"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14963",
"repo": "Microsoft/LightGBM",
"url": "https://github.com/Microsoft/LightGBM/issues/455"
}
|
gharchive/issue
|
Multi-class R demo does not run (api error)
When trying to use the lightGBM R-package for some work I encountered the error "Number of classes should be specified and greater than 1 for multiclass training." I confirmed that my factor labels were converted to numeric and started at 0, but the error still occurred. I ran the exact code from the demo here, and received the same error.
Environment info
Operating System: Ubuntu 16.04
CPU: i7, 7th gen
R version: 3.3.3
Error Message:
Error in lgb.call("LGBM_DatasetCreateFromMat_R", ret = handle, private$raw_data, : api error: Number of classes should be specified and greater than 1 for multiclass training
I produced this error by running /lightGBM/R-class/demo/multiclass.R
I started up a parallel R session and am unable to reproduce the error, even after loading the same packages. I cleared the environment on the session producing the error and it continues to occur. Closing issue due to inability to reproduce.
@nnormandin use lgb.unloader to fix this. If you try to train a model with wrong/unexpected inputs, it will always error until you unload and restart the package.
@Laurae2 ahhhh thanks! rookie mistake on my part.
@Laurae2 should we add this into document?
|
2025-04-01T04:10:39.497141
| 2018-04-09T19:48:32
|
312662679
|
{
"authors": [
"cre8ivepark"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14964",
"repo": "Microsoft/MixedRealityToolkit-Unity",
"url": "https://github.com/Microsoft/MixedRealityToolkit-Unity/pull/1930"
}
|
gharchive/pull-request
|
[Dev→master] Input test scene updates
Overview
Bringing scene updates from Dev branch. Visual (layout & material) updates. No code changes.
Updated FloorMaterial to have a clear visualization of the borders.
Before & After images
Updated Floor material with MRTK/Standard shader
Merged Patch4_Dev and may18_dev. Now, this PR only shows the changes for the two input test scenes.
|
2025-04-01T04:10:39.605966
| 2017-05-31T23:19:57
|
232724493
|
{
"authors": [
"erictraut",
"msftclas"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14965",
"repo": "Microsoft/NoSQLProvider",
"url": "https://github.com/Microsoft/NoSQLProvider/pull/28"
}
|
gharchive/pull-request
|
Added missing exported types that prevented this module from being us…
…ed in TypeScript projects without manually defining the missing types.
@erictraut,
Thanks for your contribution as a Microsoft full-time employee or intern. You do not need to sign a CLA.
Thanks,
Microsoft Pull Request Bot
|
2025-04-01T04:10:39.607909
| 2017-09-04T11:49:05
|
255018474
|
{
"authors": [
"joergplewe",
"uve"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14966",
"repo": "Microsoft/PowerBI-visuals",
"url": "https://github.com/Microsoft/PowerBI-visuals/issues/246"
}
|
gharchive/issue
|
Visualize data from multiple, independent tables
Hi community!
I need to visualize data (time series in my case) coming from different (well-known, not arbitrary), independent tables.
What is the best way to do it?
I cannot make it with capabilities.json so far, so I desparately need your help.
Hello @joergplewe,
You have to use one common table in Custom Visuals.
You can try to use PowerQuery for your case: https://powerbi.microsoft.com/en-us/documentation/powerbi-desktop-query-overview/
|
2025-04-01T04:10:39.622426
| 2016-04-07T20:15:22
|
146730678
|
{
"authors": [
"MikhailArkhipov",
"huguesv"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14967",
"repo": "Microsoft/RTVS",
"url": "https://github.com/Microsoft/RTVS/issues/1483"
}
|
gharchive/issue
|
Unexpected reformatting
Reformat this code:
# retrieve all available packages from all repos
base.fields <- c(
'Package',
'Version',
'Depends',
'Imports',
'Suggests',
'Enhances',
'License',
'NeedsCompilation',
'Repository'
)
and it becomes:
# retrieve all available packages from all repos
base.fields <- c(
'Package',
'Version',
'Depends',
'Imports',
'Suggests',
'Enhances',
'License',
'NeedsCompilation',
'Repository')
How did you reformat? ^K^D leaves it as it is for me
Sigh. Also works for me on this afternoon's build.
|
2025-04-01T04:10:39.624597
| 2017-11-09T16:09:29
|
272620072
|
{
"authors": [
"arekku",
"jtarquino"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14968",
"repo": "Microsoft/Reporting-Services",
"url": "https://github.com/Microsoft/Reporting-Services/issues/56"
}
|
gharchive/issue
|
Custom authentication extension issue
I seem to be having an issue with my custom extension after following the sample. The issue is that I can log into the report server but get an error 500: unable to validate data, when I try and view the portal (localhost/reports). I have include my config files in case i made a mistake within.
Logs and configs.zip
This was because of the hosting service failing to copy the key from RSReportServer to Web.Config. Default service user must have write permission to web.Config. Please correct me if im wrong.
Yes it needs to copy the key into the web.config
|
2025-04-01T04:10:39.629068
| 2015-12-03T16:13:31
|
120208982
|
{
"authors": [
"aksharpatel47",
"mhegazy",
"msftclas",
"zhengbli"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14969",
"repo": "Microsoft/TSJS-lib-generator",
"url": "https://github.com/Microsoft/TSJS-lib-generator/pull/27"
}
|
gharchive/pull-request
|
Added correct definition for createIndex method of IDBObjectStore
Fixes Microsoft/Typescript#5832
Hi @aksharpatel47, I'm your friendly neighborhood Microsoft Pull Request Bot (You can call me MSBOT). Thanks for your contribution!
In order for us to evaluate and accept your PR, we ask that you sign a contribution license agreement. It's all electronic and will take just minutes. I promise there's no faxing. https://cla.microsoft.com.
TTYL, MSBOT;
@aksharpatel47, Thanks for signing the contribution license agreement so quickly! Actual humans will now validate the agreement and then evaluate the PR.
Thanks, MSBOT;
:+1:
Other than the comment, looks good to me :+1:. Thanks for contributing @aksharpatel47 !
@zhengbli There are a lot of other IndexedDB definitions missing such as multiEntry property of an Index, interfaces for options when creating an object store and an index. I can add them, but should I create a new issue and a pull request or add them here?
@aksharpatel47 An issue should be created, the missing types can all belong to a single issue to avoid too much efforts. Some of them may be corrected next time we take an updated XML spec files from the Edge team, but it would be great if we can get a quick fix before that (which might be a painful process with many breaking changes). Thanks!
|
2025-04-01T04:10:39.636199
| 2017-08-09T11:13:25
|
248996674
|
{
"authors": [
"cto1",
"dmt",
"markmssd",
"michalstocki",
"welkie"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14970",
"repo": "Microsoft/TypeScript-Node-Starter",
"url": "https://github.com/Microsoft/TypeScript-Node-Starter/issues/37"
}
|
gharchive/issue
|
supertest tests always pass
changing any of the expected status codes to other values does not seem to have any impact on whether or not the tests pass.
Digging a bit further, I can't get any done callback based test to work. The only thing that seems to work right is returning the result of the expect call. (After also adjusting the supertest initialisation to point to port 3000).
I'm new to TypeScript, Node and these test frameworks so I might just be missing something but the fact that you can change the status codes for expectations and the tests still pass seems wrong either way.
I agree with you, I had to modify the way supertest is configured to make it work.
e.g:
import * as supertest from 'supertest';
import * as sinon from 'sinon';
import { Neo4J } from '../../src/db';
const app = require('../../src/server');
const agent = supertest.agent(app);
describe('Health GET /health', () => {
it('should return 417 Failed', (done) => {
const stub = sinon.stub(Neo4J, 'checkConnection').rejects('error');
agent.get('/health')
.expect(417)
.then(() => {
stub.restore();
done();
})
.catch(error => {
stub.restore();
done.fail();
});
});
I noticed the same behaviour!
Same. Also not sure if jest import is required. I tried adding callback (done) and returning the promise. Both approaches didn't resolve this issue.
A
Same to me. The issue is because the Jest API is used incorrectly.
First of all this cannot work because when running tests (npm test) there is no place where the server is set up – literally there is nothing like http://localhost:8000 because there is no such server.
Moving further, all the tests incorrectly uses the Jest done() callback. As long as in Mocha or Jasmine, we can simply call done(err) to make the test fail, in Jest we must call done.fail(err) instead.
So, after replacing:
request.get("/contact")
.expect(200, done);
with
request.get("/contact")
.expect(200, (err, res) => {
if (err) {
done.fail(err);
} else {
done();
}
});
or better with
request.get("/contact")
.expect(200)
.end((err, res) => {
if (err) {
done.fail(err);
} else {
done();
}
});
all the tests fail.
At the end I must say, that I spend several hours trying to figure out how the tests are configured that they work, without any configuration of the server under localhost:8000, so I'm a bit frustrated. 😞
BTW @dmt that's why after your change, the tests are failing in #38
It's possible to return the result of setting up the expectation and that seems to behave as it should. I did also update the PR to separate the setup of the app and actually binding to the port. This does make it possible to run the tests, requiring that extracted app setup instead of assuming there is an instance of the server running somewhere.
I'm still not super happy with the changes for various design reasons and the fact that I don't actually know anything about typescript. But at least the tests now actually do something.
This has to do with how Promises are used in JavaScript tests. If you're testing something asynchronous, you must return a promise from the callback in the it function. Alternatively, you add a done parameter and call it at the end of the test. Note that some asynchronous tests will need longer than the default time limit most testing frameworks allow per test, so it's normal to increase that time limit for them.
|
2025-04-01T04:10:39.640326
| 2018-10-10T15:41:38
|
368729586
|
{
"authors": [
"chinanf-boy",
"mdluo",
"msftclas"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14971",
"repo": "Microsoft/TypeScript-Node-Starter",
"url": "https://github.com/Microsoft/TypeScript-Node-Starter/pull/148"
}
|
gharchive/pull-request
|
🇨🇳Chinese for README.md 😊
#130
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.:x: chinanf-boy sign nowYou have signed the CLA already but the status is still pending? Let us recheck it.
Poor quality translation: a lot of unreadable sentences (looks like translated by Google), multiple mistranslations, random spaces and mixed types of punctuation marks, bad formatting and some missing content.
I would suggest the collaborators not to merge this PR before improvements are made upon it. Otherwise there’s no significant difference between the translated document and the translation by the Chrome browser.
Btw, can you please try to avoid China national flag when talking about Chinese (the language)?
|
2025-04-01T04:10:39.645432
| 2018-12-20T17:49:06
|
394521988
|
{
"authors": [
"mjbvz",
"saoudrizwan",
"sheetalkamat"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14972",
"repo": "Microsoft/TypeScript-TmLanguage",
"url": "https://github.com/Microsoft/TypeScript-TmLanguage/issues/700"
}
|
gharchive/issue
|
Add TextMate scopes for variable.other.object.method and support.variable.method (instead of property)
I'm trying to modify the syntax color for method/functions that are members of a variable, no matter if they are being invoked or referenced.
Here's what I have so far:
{
"scope": [
"meta.method.declaration support.function",
"meta.method.declaration entity.name.function",
"meta.definition.method support.function",
"meta.definition.method entity.name.function",
"meta.definition.property support.function",
"meta.definition.property entity.name.function",
"meta.function.declaration support.function",
"meta.function.declaration entity.name.function",
"meta.definition.function support.function",
"meta.definition.function entity.name.function",
"meta.definition.property support.function",
"meta.definition.property entity.name.function"
],
"settings": {
"foreground": "#fff"
}
},
{
"scope": [
"meta.function-call support.function",
"meta.function-call entity.name.function",
"meta.method.declaration meta.function-call support.function",
"meta.method.declaration meta.function-call entity.name.function"
],
"settings": {
"foreground": "#8182F8"
}
},
In this example, I'd like to be able to have fref() and completion() be white instead of purple, but that's not an option because they have the exact same scope as .listen(.
As of now, VSCode exposes variable.other.object.property and support.variable.property, so I can make fref white and app purple, since app is a property of a variable and not a local variable. I'd like to be able to do something similar with functions that are specifically members of a variable or class.
variable.other.object.method and support.variable.method
Update
I believe variable.function is the scope I was looking for, however it doesn't seem to be supported by VSCode. It is supported in other editors!
Moving upstream to get more feedback. However some parts of this sound like they would require semantic colorization which textmate grammars do not support
The only way to syntactically detect function at its reference location is function invocation so that part is out of scope for this repo. The other part of detecting function as variable.function was something we reverted as part of #443 where we concluded that entity.name.function is more widely used and hence we would use that.
|
2025-04-01T04:10:39.654303
| 2017-01-11T11:27:07
|
200068604
|
{
"authors": [
"OleksandrNechai",
"sandersn"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14973",
"repo": "Microsoft/TypeScript",
"url": "https://github.com/Microsoft/TypeScript/issues/13409"
}
|
gharchive/issue
|
Incorrect type inference of object spread, when spreaded object is referenced via interface and created by class instantiation
TypeScript Version: 2.1.X (Playground)
Code
interface I { x: number, y: number, z: () => void }
class MyClass implements I {
constructor(public x: number, public y: number) {
}
z() { alert(this.x + this.y) }
}
var x: I = new MyClass(5, 7);
var o = { ...x }
alert(o.z); // undefined
Expected behavior:
Compile error. Either MyClass incorrectly implements I interface requiring to define z as z=()=>alert(this.x + this.y) or at alert(o.z) reporting there is no such a property on o, or report something similar to FlowType.
Actual behavior:
No compiler warnings and runtime error when trying to invoke o.z().
Note:
When I change a definition of I this way: interface I { x: number, y: number, z(): void }, I get an error Property 'z' does not exist on type '{ x: number; y: number; }' as expected.
This is almost a duplicate of #13148, although there the desired behaviour was different. You can read that bug for the full resolution, but here are a few points specific to this example:
Basically, don't spread class instances. Spread isn't designed to work well with classes.
Unfortunately, because the compiler doesn't track own or enumerable on properties, it guesses when it comes to spread.
After the fix #13365 for #13148, Typescript now assumes that any method that doesn't definitely come from a class is still present after a spread. That's because spread should be used with object literals, not class instances.
Note that if you change the interface to be interface I { x: number, y: number, z(): void }, then flow fails to report the error and hits a runtime error. Flow just chose a different default than Typescript did. It might be easier to add own/enumerable inference to flow than to typescript, but I don't think either compiler does right now.
|
2025-04-01T04:10:39.657050
| 2017-03-12T13:48:38
|
213605710
|
{
"authors": [
"IAmNoP",
"ahejlsberg",
"kitsonk"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14974",
"repo": "Microsoft/TypeScript",
"url": "https://github.com/Microsoft/TypeScript/issues/14609"
}
|
gharchive/issue
|
Typescript Error No best common type exists among return expressions.
Hello, I feel like there is a bug here. please let me know :).
A bit of class inception.
The principals:
the class Number is a CalcNode;
push requires a CalcNode return;
push local variable is a CalcNode;
I have a multiple "if" condition with "else" at the end. It seems like a "no condition" return is attended.
Typescript Error
No best common type exists among return expressions.
src/pages/calculator/engine/number.class.ts
push(new_value: CalcNode){
if(new_value.get_priority() === Priority.THREE){
export class Number implements CalcNode {
value: string = "";
priority: Priority = Priority.THREE;
constructor(new_value: string){
this.value = new_value;
}
get_priority(){
return this.priority;
}
push(new_value: CalcNode){
if(new_value.get_priority() === Priority.THREE){
this.value += ""+new_value.evaluate();
return this;
}
else if (new_value.get_priority() === Priority.ZERO){
var temp: CalcNode = new Multiply();
temp.push(this);
temp.push(new Parenthese());
return temp;
}
else {
return new_value.push(this);
}
}
Try annotating your return of the method push(), because it will help highlight where the compiler is getting confused. I suspect it is the circular logic in trying to identify the return type of push() that is causing it to be confused.
If you post a complete example we can take a look, however I don't think there is a bug here.
|
2025-04-01T04:10:39.660477
| 2017-05-23T15:34:20
|
230752298
|
{
"authors": [
"asvetliakov",
"mhegazy"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14975",
"repo": "Microsoft/TypeScript",
"url": "https://github.com/Microsoft/TypeScript/issues/16028"
}
|
gharchive/issue
|
Generic argument is being narrowed to {}
TypeScript Version: 2.3.3 / nightly (2.4.0-dev.20170523)
Code
// A *self-contained* demonstration of the problem follows...
interface TestProp<T> {
a: number;
b: T;
}
declare function TestC<T>(props: TestProp<T>): void;
declare function TestW<Props>(comp: (props: Props) => void): Props;
const p = TestW(TestC);
Expected behavior:
p is type TestProp<any> or TestProp<{}>
Actual behavior:
p is type {}
Duplicate of https://github.com/Microsoft/TypeScript/issues/9366
Automatically closing this issue for housekeeping purposes. The issue labels indicate that it is unactionable at the moment or has already been addressed.
|
2025-04-01T04:10:39.665660
| 2017-07-20T12:55:13
|
244351980
|
{
"authors": [
"RyanCavanaugh",
"johnnyreilly"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14976",
"repo": "Microsoft/TypeScript",
"url": "https://github.com/Microsoft/TypeScript/issues/17319"
}
|
gharchive/issue
|
Type Narrowing not being all it can be
TypeScript Version: 2.4.1
Hey all, I stumbled upon an oddity around type narrowing. I've been meaning to report for ages but... well better late than never!
Code
interface OneThing {
oneThing: {};
}
interface OtherThing {
otherThing: {};
}
const returnOneThingOrAnother = () =>
'5' === `${6}` // ignore this - what's produced is what matters
? { oneThing: {} } as OneThing
: { otherThing: {} } as OtherThing;
const iHaveOneThingOrAnother = returnOneThingOrAnother();
if (iHaveOneThingOrAnother.oneThing) {
// iHaveOneThingOrAnother should be narrowed to OneThing
} else {
// iHaveOneThingOrAnother should be narrowed to OtherThing
}
Expected behavior:
I would expect type narrowing to occur in the manner suggested in the comments.
Actual behavior:
Actually the expected type narrowing does not occur. Rather, the if (iHaveOneThingOrAnother.oneThing) { expression throws the following exception:
Property 'oneThing' does not exist on type 'OneThing | OtherThing'.
Property 'oneThing' does not exist on type 'OtherThing'.
Why? Obviously I could solve this by introducing a user defined type guard but that seems like it ought not to be necessary.
Is there a reason the TypeScript compiler does not perform the type narrowing in the way I had hoped?
Possibly related to https://github.com/Microsoft/TypeScript/issues/12457 but not certain.
Much :heart: for all your work BTW 🌻
Good question. We consider this to be an unsafe operation because we can't guarantee something like this doesn't happen:
interface GotString {
x: string;
}
interface GotNumber {
y: number;
}
const fooled = { x: 100, y: 200 };
const gn: GotNumber = fooled;
const j: GotString | GotNumber = Math.random() > 0.5 ? gn : { x: 'ok' };
if (j.x) {
// In theory: j: GotString
j.x.substr(0); // Crash - x was a number. Bamboozled!
}
See also #14094 where it's proposed to add syntax to create these sorts of unions automatically.
Thanks for the response @RyanCavanaugh - that makes sense.
I'm quite keen on exclusive unions so at present I'm landed with either introducing a type guard or casting back to any briefly - neither of which is attractive. Suffice to say I like the look of #14094!
Do feel free to close this issue if you think it's covered by #14094. Good to know the reason for this 👍
|
2025-04-01T04:10:39.668719
| 2017-08-09T18:21:05
|
249118558
|
{
"authors": [
"Syderr93",
"j-oliveras",
"kitsonk"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14977",
"repo": "Microsoft/TypeScript",
"url": "https://github.com/Microsoft/TypeScript/issues/17700"
}
|
gharchive/issue
|
Import without asterisk
Is there any way to import npm modules without using the ES6 asterisk import and without using import fs = require("fs") ?
Example:
import * as fs from "fs"
I'd like to write:
import fs from "fs"
like with local project files.
Maybe set allowSyntheticDefaultImports is that you need.
These type of questions should be asked on StackOverflow or Gitter.
@j-oliveras allowSyntheticDefaultImports will use ES6 export.default for every import I use, and not all of npm modules uses ES6 export.default
@kitsonk I really don't think that someone will answer this on SO
@Syderr93 Then, maybe, this is a duplicated of #16093.
|
2025-04-01T04:10:39.674874
| 2017-08-17T19:55:18
|
251053902
|
{
"authors": [
"amcasey",
"mjbvz"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14978",
"repo": "Microsoft/TypeScript",
"url": "https://github.com/Microsoft/TypeScript/issues/17869"
}
|
gharchive/issue
|
Extract method refactoring not returned when selection is on binary expression operator
TypeScript Version: 2.5.0-dev20170816
Code
class Foo {
bar () {
return 1 + 1;
}
}
Place cursor on the + sign and try triggering refactorings:
Expected behavior:
Extract method for the expression 1 + 1 is returned
Actual behavior:
No refactor returned.
here's the ts server trace
[Trace - 12:51:03 PM] Sending request: getApplicableRefactors (14). Response expected: yes. Current queue length: 0
Arguments: {
"file": "/Users/matb/projects/san/castle.js",
"startLine": 3,
"startOffset": 17,
"endLine": 3,
"endOffset": 17
}
[Trace - 12:51:03 PM] Response received: getApplicableRefactors (14). Request took 1 ms. Success: true
Result: []
To be honest, I'm surprised it works with any empty selection. Is this part of a broader request or do you just want support for binary operators?
https://github.com/Microsoft/TypeScript/blob/40e459117aeb0b792a23d6805af884b594dd42c4/src/services/refactors/extractMethod.ts#L169 does the wrong thing for empty spans. @mjbvz, I'm assuming you'd prefer we attempt to handle empty spans, rather than forbidding them explicitly (which would also resolve the inconsistency in the current behavior).
With #18060 now fixed, I don't think this request makes sense anymore. It originally struck me as feeling inconsistent but I now think we should require a selection for extracting in this case
Thanks, @mjbvz! I'll create a PR explicitly disabling that (since there are a few cases where it works by accident).
One other question: what if the selection contains exactly +? Should that work? Presently, it doesn't.
https://github.com/Microsoft/TypeScript/pull/18427
Hmm, I don't think so. I'd say that we should require the binary expression itself to be selected for now. If many users start reporting that extract method does not for cases like this, we can always make the rules looser again
|
2025-04-01T04:10:39.677978
| 2015-02-05T05:01:19
|
56627823
|
{
"authors": [
"RyanCavanaugh",
"basarat"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14979",
"repo": "Microsoft/TypeScript",
"url": "https://github.com/Microsoft/TypeScript/issues/1938"
}
|
gharchive/issue
|
instanceof Error typeguard not functioning
Wonder if this is something that should work. I think it should:
try {
throw new Error('awesome');
} catch (ex) {
if (ex instanceof Error) {
console.log(ex.foo); // compiles. Should be error
}
}
The reason is that ex is still of type any. I suspect its in the spec that instanceof guard will not work on type any (I haven't checked).
Would be great if we could have this though.
See #1425. We do not narrow values of type any.
The first version of type guards did narrow any values, but there was too much code in the wild that did stuff like
var x: any; // x is a string or an HTML element I know is an image
if(x instanceof HTMLElement) {
x.width = 100;
}
@RyanCavanaugh sure its #1425 ?
Typo'd and updated, meant #1426
|
2025-04-01T04:10:39.683599
| 2015-06-06T14:22:41
|
85757013
|
{
"authors": [
"bahrus"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14980",
"repo": "Microsoft/TypeScript",
"url": "https://github.com/Microsoft/TypeScript/issues/3401"
}
|
gharchive/issue
|
Provide a special decorator in which you can define a type
Without instantiating and populating an object, it doesn't seem possible with the available reflection built into JavaScript to determine the type of a property or method. For example, how could you determine that the property Surname of the Employee class / prototype below returns a string?
class Employee{
get Surname() : string{
...
}
}
Assuming there isn't a way to do runtime reflection to determine the type of the "Surname" property in this example, and now that decorators are integrated into the language, why not provide an alternative, slightly clunkier way, to specify type for properties and method. E.g.,
class Employee{
@type(String)
get Surname(){
}
}
Typescript would essentially be reserving a special decorator name for this purpose, and would add the type information into the metadata associated with the Surname property, which could then be inspected at run time.
In this proposal, typescript would treat these two class definitions as the same for compiling / intellisense purposes, and support use of both ways of type annotation, but would reject attempts to use both syntaxes simultaneously on the same property or method to avoid ambiguity. The second way would be used when run time reflection is needed.
I know this starts to resemble Closure, but Closure's annotations were in comments, and now that decorators are integrated directly into the language, I think it's worth revisiting whether the two approaches could be supported together.
Alternatively, if there was a way to set the TypeScript compiler to add type information into the metadata for the property or method in the generated js file, that would be equally effective, I think.
Update: I overlooked the flag --emitDecoratorMetadata , which seems to provide the information I need.
I overlooked the flag --emitDecoratorMetadata , which seems to provide the information I need.
|
2025-04-01T04:10:39.685323
| 2014-08-20T21:24:16
|
40746591
|
{
"authors": [
"mhegazy",
"sedwards2009"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14981",
"repo": "Microsoft/TypeScript",
"url": "https://github.com/Microsoft/TypeScript/issues/494"
}
|
gharchive/issue
|
Request: Compiler option to specify an alternate lib.d.ts
The compiler has an option noLib which disables the default inclusion of lib.d.ts during compilation. Unfortunately there is no compiler option to specify a .d.ts file or files to automatically include before compiling each .ts file. Such an option would make it much easier to use an alternate lib.d.ts.
The work around now is to use the noLib option and then to add a line to each source file to manually pull in your alternate lib.d.ts.
this should be covered by the proposal in https://github.com/Microsoft/TypeScript/issues/6974
|
2025-04-01T04:10:39.688800
| 2016-06-17T16:55:20
|
160931173
|
{
"authors": [
"OlegTsyba",
"sandersn"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14982",
"repo": "Microsoft/TypeScript",
"url": "https://github.com/Microsoft/TypeScript/issues/9231"
}
|
gharchive/issue
|
Compiler doesn't understand type restrictions in function declaration
TypeScript Version:
1.8.10
Code
function map<T>(list: Array<T>,
fn: (T) => any): Array<any>
{
let res = [];
for(let i = 0; i < list.length; i++)
res.push(fn(list[i]));
return res;
}
console.log(map<string>(['fuck', 'the', 'police'], (x: number) => x.toFixed()));
Expected behavior:
I expected the typechecker rejected (x: number) => x.toFixed(), since it doesn't fit to the type declaration.
Actual behavior:
Typechecker didn't reject this code. So, I have runtime error instead of compile error
The syntax for functions is different than what you expected, probably:
fn: (T) => any): Array<any>
should be
fn: (t:T) => any): Array<any>
You have to give the parameter a name. Otherwise, it uses T as the name and any as the type.
|
2025-04-01T04:10:39.690350
| 2018-04-10T00:59:11
|
312737407
|
{
"authors": [
"ajafff",
"weswigham"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14983",
"repo": "Microsoft/TypeScript",
"url": "https://github.com/Microsoft/TypeScript/pull/23297"
}
|
gharchive/pull-request
|
Add support for destructuring well-known and late-bound names
Fixes #23257
I guess there will now be an error if the late bound computed property is not present in the destructured object?
declare function get<T>(): T;
let { [get<'foo'>()]: computed } = get<{ bar: string }>(); // error: property 'foo' does not exist in type '{bar: string}'?
Maybe add a test for that case?
Btw.: why is there no implicit any error if the computed property name is not a literal type and the destructured type has no index signature? Using element access there is an implicit any error with --noImplicitAny.
|
2025-04-01T04:10:39.697772
| 2019-02-27T14:05:41
|
415141384
|
{
"authors": [
"jeschu1",
"pmj",
"wilbaker"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14984",
"repo": "Microsoft/VFSForGit",
"url": "https://github.com/Microsoft/VFSForGit/pull/856"
}
|
gharchive/pull-request
|
Mac kext: Fix for vnode_put() panic bug, tighter invariant assertion & improved root search logic.
The first commit in this series fixes #797, the second one strengthens one of the related assert()s and the third further cleans up the function that contained the bug.
I explained the nature & cause of the bug in this comment so I won't repeat that here. The fix is of course to use the inode of each directory level to try to match a virtualisation root as we walk up the hierarchy.
In addition to the new asserts I added in PR #805, this tightens up a much older assert in a way that would have caught this bug in action. If the vnode/vid pair matches a root vnode/vid pair, their fsid/inode must also match, or something has gone badly wrong.
There has so far been no proper error handling for the vnode_get() call in VirtualizationRoot_FindForVnode(). I don't think that call can fail in that situation, but just in case, I've now changed the logic such that the iocount remains correct. In cases where the incoming vnode is not a directory however, it actually makes more sense to start the search with the parent directory, so we can call vnode_getparent() in that case instead.
I am planning to write some unit tests for VirtualizationRoot_FindForVnode() that would have caught this bug. However, these will depend on the substantial upgrades to the unit testing infrastructure I added in PR #839, and as that PR is still some way away from being merged, I won't delay the fix until then.
I am planning to write some unit tests for VirtualizationRoot_FindForVnode() that would have caught this bug. However, these will depend on the substantial upgrades to the unit testing infrastructure I added in PR #839, and as that PR is still some way away from being merged, I won't delay the fix until then.
I agree we shouldn't delay the fix, I just filed issue #862 for adding unit tests
I agree we shouldn't delay the fix, I just filed issue #862 for adding unit tests
@wilbaker I'm fine with that but let's prioritize that for next week, unless there are objections. It would definitely give us a sense of security if test for cases like this in the future.
@pmj nice job again tracking this down!
@wilbaker I'm fine with that but let's prioritize that for next week, unless there are objections.
I completely agree, adding the unit tests is highest priority (after merging this in).
/ cc: @pmj @jrbriggs
I completely agree, adding the unit tests is highest priority (after merging this in).
Sounds good. Perhaps, @wilbaker you could take a high-to-mid-level look at #839 in that case please, as I'm reluctant to head further down that path with the tests for VirtualizationRoot_FindForVnode() unless I'm reasonably sure I'm not going to have to rework it all in a major way.
#819, the fix for the race condition related to bug #797 is ready to go too (assuming all the tests pass) and should be a fairly quick review if anyone would like to jump on that. Thanks!
Sounds good. Perhaps, @wilbaker you could take a high-to-mid-level look at #839 in that case please, as I'm reluctant to head further down that path with the tests for VirtualizationRoot_FindForVnode() unless I'm reasonably sure I'm not going to have to rework it all in a major way.
#819, the fix for the race condition related to bug #797 is ready to go too (assuming all the tests pass) and should be a fairly quick review if anyone would like to jump on that. Thanks!
@pmj sure thing! I will look at #819 and #839 next.
|
2025-04-01T04:10:39.702054
| 2015-07-31T03:22:42
|
98308884
|
{
"authors": [
"MarcosMeli",
"timsneath"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14985",
"repo": "Microsoft/VSSDK-Extensibility-Samples",
"url": "https://github.com/Microsoft/VSSDK-Extensibility-Samples/pull/1"
}
|
gharchive/pull-request
|
Fixing links
Maybe you need also to remove Visual Studio 2015 RC Download because is the unique that appear on search and sure many people will tend to download it, or provide also final SDK as standalone
http://www.microsoft.com/en-us/download/details.aspx?id=46850
Thanks for the catch, Marcos - I see broken links in a few other places also so we'll have to fix them too. For anyone else watching this request, we're now including the VS SDK as an integrated part of Visual Studio - go to Custom Setup and select the Extensibility Tools checkbox to install the SDK.
|
2025-04-01T04:10:39.707543
| 2018-09-05T23:52:40
|
357444210
|
{
"authors": [
"Brian-Perkins",
"dmex"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14986",
"repo": "Microsoft/WSL",
"url": "https://github.com/Microsoft/WSL/issues/3525"
}
|
gharchive/issue
|
GetExtendedTcpTable returns invalid ProcessIds for WSL processes
The GetExtendedTcpTable function returns invalid ProcessIds for listen sockets created by WSL processes.
A number of people use tools such as netstat, TCPView, Process Hacker and others (including similar tools from 'security' vendors) while checking machines for malware/spyware/rootkits and the GetExtendedTcpTable function returns fake ProcessIds for listen sockets created by WSL processes giving users an incorrect impression the machine has been compromised.
This becomes an additional problem when Windows assigns a new process that same fake ProcessId and these tools will show the completely wrong process that owns the listen socket.
The problem is reproducible on the latest versions of Windows 10 (all branches):
10.0.17134.254 - [RS4]
10.0.17754.1 - [RS5_release]
10.0.18219.1000 - [RS_prerelease]
You can easily reproduce the issue by running the following commands:
WSL:
sudo apt install nginx
sudo nginx
Windows:
Netstat: https://i.imgur.com/Fkd0V5l.png
TCPView: https://i.imgur.com/paS7FYp.png
Process Hacker: https://i.imgur.com/X15GoVu.png
What you are seeing is a pico process. In short, a NT process is used as a kind of container to run a WSL process. That NT process has a PID and that is what the tools are reporting. We also plumbed through reasonable names, but that depends on which APIs are being used. For instance, if the tool tries to inject a thread or directly access process memory to obtain additional information, those techniques will not work very well for pico processes. We have been improving some of the diagnostic API's so that tools don't have to do much work to get better WSL information, but if there is a specific API that has room for improvement, we will certainly investigate.
Ah -- got it, the "fake ProcessId" part led me to incorrect assumptions. It should be true that most APIs returning process information, like GetExtendedTcpTable, are holding a reference to the process they return. This should keep the process ID from being reused. This still leaves the problem of identifying which process is currently using a port when all you are given is the process that originally created/connected the associated socket. That is an interesting problem that I don't have a simple solution for.
|
2025-04-01T04:10:39.714123
| 2017-01-25T19:18:56
|
203196380
|
{
"authors": [
"edvv",
"rajsesh-msft"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14987",
"repo": "Microsoft/WinObjC",
"url": "https://github.com/Microsoft/WinObjC/issues/1819"
}
|
gharchive/issue
|
Need performance benchmark tests for CoreText
Would be useful to time these scenarios. The goal is to have a simple commandline test similar to the unit tests, but rather than pass fail, these would run the perf scenarios and log the various metrics. Here are some basic metrics we want to have:
Drawing a frame with 100 sentences.
Drawing 10000 lines.
Drawing 100,000 runs.
Drawing 10,000,000 glyphs.
Drawing using NSString drawAtPoint and variations.
Drawing using NSAttributedString drawAtPoint.
Once we have this, we could plug this into nightly build for build-over-build tracking of perf numbers to catch future regressions.
@rajsesh-msft Suggestion: If you set a baseline against Release 41 ( https://github.com/Microsoft/WinObjC/issues/1500 ) then that would give pass/fail criterion and the unit tests would be more fudge proof.
|
2025-04-01T04:10:39.716003
| 2016-08-29T21:39:13
|
173879134
|
{
"authors": [
"DHowett-MSFT",
"MSFTFox",
"aballway",
"rajsesh-msft"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14988",
"repo": "Microsoft/WinObjC",
"url": "https://github.com/Microsoft/WinObjC/pull/938"
}
|
gharchive/pull-request
|
Fix regular expression test on long running operations.
Regular expression matching can potentially take longer to run.
On a matching option of type NSMatchingReportProgress, this will
call the block passed in to be called more times than expected.
Modify this test case so that the test calls the block at least
an expected number of times.
Fix #766
:shipit:
:shipit:
:shipit:
|
2025-04-01T04:10:39.718192
| 2016-11-04T16:36:52
|
187380983
|
{
"authors": [
"martinsuchan",
"oldnewthing"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14989",
"repo": "Microsoft/Windows-universal-samples",
"url": "https://github.com/Microsoft/Windows-universal-samples/issues/519"
}
|
gharchive/issue
|
Add sample for the App Extensions API
The collection of samples does not contain any sample how to use the new App Extension API, that was shown here:
https://channel9.msdn.com/Events/Build/2016/B808
This keynot has related sample available here
https://github.com/Microsoft/Build2016-B808-AppExtensibilitySample
but it was not even updated to the Anniversary Update SDK version and needs some maintenance before it actually works.
It would be handy to have out-of-the-box working sample for the App Extension API here.
I've forwarded your feedback to the AppExtension team. thanks.
|
2025-04-01T04:10:39.727506
| 2019-03-18T10:38:30
|
422145872
|
{
"authors": [
"adrianhill102",
"lumaxis"
],
"license": "CC-BY-4.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14990",
"repo": "Microsoft/appcenter",
"url": "https://github.com/Microsoft/appcenter/issues/186"
}
|
gharchive/issue
|
Update plugin to allow a new app to be created as part of the method call
Describe the solution you'd like
We are using this plugin in our CI pipeline. This means that interactive mode is not supported. We want to be able to create a new app and push it to App Center without having to create the app in the console first. This functionality is supported by the HockeyApp plugin.
Describe alternatives you've considered
At present out only option is to manually create an app with a specific name in the App Center console before pushing it from our CI pipeline.
Additional context
APPCENTER_APP_NAME - App name. If there is no app with such name, you will be prompted to create one
We are using the plugin in our CI pipeline so interactive prompts are not suitable
Can you update the plugin to allow a new app to be created as part of the method call?
@adrianhill102 Just to clarify, you're talking about our fastlane plugin, right?
Yes, the fastlane plugin. So in the sample below, 'app_name' would create the app without prompting the user.
appcenter_upload(
api_token: "",
owner_name: "",
app_name: "",
apk: ""
)
|
2025-04-01T04:10:39.735186
| 2018-06-22T14:45:47
|
334905748
|
{
"authors": [
"brcrista",
"damccorm"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14991",
"repo": "Microsoft/azure-pipelines-task-lib",
"url": "https://github.com/Microsoft/azure-pipelines-task-lib/pull/374"
}
|
gharchive/pull-request
|
Add getPlatform function
I've been copy-pasting this in various tasks and feel like it could be a good addition to task-lib.
This improves on the existing osType function by using an enum instead of a string (more type safe) and makes it obvious what the possible return values are.
@brcrista this looks helpful to me, do you want to resolve conflicts and merge?
This looks good, thanks for cleaning it up. I'll go ahead and merge!
|
2025-04-01T04:10:39.736517
| 2019-02-12T21:30:45
|
409509325
|
{
"authors": [
"NathanielRose",
"dtzar"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14992",
"repo": "Microsoft/bedrock",
"url": "https://github.com/Microsoft/bedrock/issues/95"
}
|
gharchive/issue
|
Parity with Azure Dev Ops & Travis for CI
Add Azure Devops yaml file in root.
Allows user to deploy bedrock in ADO for their own CI
This is now working via #199 at https://dev.azure.com/epicstuff/bedrock/_build?definitionId=54
|
2025-04-01T04:10:39.741729
| 2018-04-18T21:52:56
|
315655516
|
{
"authors": [
"JonathanFingold",
"chrimc62",
"cleemullins",
"emadelwany",
"garypretty",
"yochay"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14993",
"repo": "Microsoft/botbuilder-dotnet",
"url": "https://github.com/Microsoft/botbuilder-dotnet/issues/458"
}
|
gharchive/issue
|
Some of the LUIS types are still in the Microsoft.Cognitive.LUIS namespace
When the LUIS types were moved to the Microsoft.Bot.Builder.Ai.LUIS namespace, nearly all of the ones defined in the Client and Client/Models subdirectories were skipped and still live in the Microsoft.Cognitive.LUIS and Microsoft.Cognitive.LUIS.Models namespaces.
Is assumed that as they were already in their own namespace not consistent with the project that this was intentional which is why I left them where they were. Happy to rectify if needed.
Interesting, I'll defer to the architects. @cleemullins ?
This is not intentional. @emadelwany , can you fix?
@garypretty - this is not intentional. The old LUIS client was called "Microsoft.Cognitive.LUIS" and we nuked it some time ago. Apparently some of the pieces escaped. Emad, on the LUIS team, should be able to clean this up asap.
This client code is just temporarily part of the Microsoft.Bot.Builder.Luis project until the identical code ships as part of the Microsoft.Cognitive.LUIS nuget. Once Microsoft.Cognitive.LUIS nuget is released by the Cognitive Services team, the whole /Client directory should be nuked, and the nuget should be a drop-in replacement (requiring 0 code changes). I can move it to the project's namespace and then we can revert all the "using" statements that uses the /client after we grab the nuget if you prefer that.
If changing the namespace for these will introduce a breaking change later, I think it may be better to not make the change. I can add a comment in the doc XML remarks that these will be part of a later version of the Microsoft.Cognitive.LUIS NuGet package.
@emadelwany can we get an ETA?
@Kaiqb heads up we'll need to update LUIS docs in BF
@chrimc62 can you take this one
From Emad's comment sounds like we are waiting for CS to publish a nuget. I've started a thread to see where that process is.
@cleemullins , @yochay , This looks like it was fixed at some point. The SDK doesn't appear to define any types in the Microsoft.Cognitive.LUIS namespace anymore. Please reopen if there is still an issue that needs to be resolved.
|
2025-04-01T04:10:39.757841
| 2015-09-03T18:55:41
|
104759393
|
{
"authors": [
"ddobrev",
"richlander"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14994",
"repo": "Microsoft/dotnet",
"url": "https://github.com/Microsoft/dotnet/issues/230"
}
|
gharchive/issue
|
Marshalling a returned struct with a boolean field crashes
This code:
using System.Runtime.InteropServices;
namespace TestMarshal
{
public struct QVariant
{
[StructLayout(LayoutKind.Explicit, Size = 16)]
public struct Internal
{
[FieldOffset(0), MarshalAs(UnmanagedType.I1)]
public bool d;
}
[UnmanagedFunctionPointerAttribute(CallingConvention.ThisCall)]
public delegate Internal _Interpolated_0Delegate(System.IntPtr instance);
public static _Interpolated_0Delegate _Interpolated_0DelegateInstance;
public static Internal _Interpolated_0DelegateHook(System.IntPtr instance)
{
return new Internal();
}
}
class Program
{
static void Main()
{
QVariant._Interpolated_0DelegateInstance += QVariant._Interpolated_0DelegateHook;
var functionPointerForDelegate = Marshal.GetFunctionPointerForDelegate(QVariant._Interpolated_0DelegateInstance);
functionPointerForDelegate.GetHashCode();
}
}
}
crashes with:
"Method's type signature is not PInvoke compatible."
It crashes the same way if MarshalAs(UnmanagedType.I1) is added to the field.
I have tested this with .NET 4.0 and .NET 4.5.1, no difference in behaviour.
It works properly on Mono with or without a MarshalAs attribute.
/cc @gkhanna79
Please re-activate if this is still an issue.
|
2025-04-01T04:10:39.766221
| 2018-12-08T15:48:07
|
388933079
|
{
"authors": [
"eanders-MS",
"willneedit"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14995",
"repo": "Microsoft/mixed-reality-extension-sdk",
"url": "https://github.com/Microsoft/mixed-reality-extension-sdk/issues/40"
}
|
gharchive/issue
|
Event listeners don't work in other world spaces
When trying to have an actor listening to events (like onHover and onClick), it only works if the item is created in the home space, but not if the item is created in a new world space.
Reproducible with hello-world example: Works okay (growth/shrink on hover, sideflip on click) in Homespace, but no reaction on user interaction when created in world space (just spins)
(UPDATE): Error is more generalized: When an item is used in more than one worldspace, the first instance I encountered worked correctly, not the second one. "Re-enter space" and "Reset Space" doesn't affect that bug; one has to completely exit AltspaceVR and restart the client.
Dup of #14
|
2025-04-01T04:10:39.768941
| 2017-08-11T23:45:23
|
249772108
|
{
"authors": [
"SoftTimur",
"alexandrudima",
"juventus18"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14996",
"repo": "Microsoft/monaco-editor",
"url": "https://github.com/Microsoft/monaco-editor/issues/547"
}
|
gharchive/issue
|
Height of the Monaco Editor
I am still struggling with the height of a Monaco Editor, the behaviour is quite different from CodeMirror or ACE.
I want to make a very simple Monaco Editor: JSBin:
<!DOCTYPE html>
<html>
<head>
<script src="https://code.jquery.com/jquery.min.js"></script>
<script src="https://ajax.googleapis.com/ajax/libs/angularjs/1.6.4/angular.min.js"></script>
<style>
.me {
height: 100vh;
}
</style>
</head>
<body>
<div class="me" id="container"></div>
<script src="https://www.matrixlead.com/monaco-editor/min/vs/loader.js"></script>
<script>
require.config({ paths: { 'vs': 'https://www.matrixlead.com/monaco-editor/min/vs' }})
require(["vs/editor/editor.main"], function () {
var editor = monaco.editor.create(document.getElementById('container'), {
value: 'function x() {\n\tconsole.log("Hello world!");\n}',
language: 'javascript',
minimap: { enabled: false },
scrollBeyondLastLine: false
});
});
</script>
</body>
</html>
When I see it in Chrome and scroll up and down, there is a scroller for the whole window. It seems that it is because the height of the editor is larger than the height of the window. I just don't want to see any scrollers. Does anyone know how to achieve this?
Thank you.
related: #29
Dup of #29.
|
2025-04-01T04:10:39.805593
| 2016-04-09T10:35:43
|
147117496
|
{
"authors": [
"JunielKatarn",
"Sarabeth-Jaffe-Microsoft",
"akoeplinger",
"cdmihai",
"radical",
"rainersigwald"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14997",
"repo": "Microsoft/msbuild",
"url": "https://github.com/Microsoft/msbuild/issues/567"
}
|
gharchive/issue
|
Can't build on UNIX
Hi. I can't build the latest xplat commit (See 5a2dcc67bff636fa8107bbf08e1852a2e34121b8) on Mac OS X 10.11.
After checking out the latest xplat, I run ./cibuild and always get the following output:
Unsupported target detected: . Configuring as if for CoreCLR
Unsupported host detected: . Configuring as if for CoreCLR
Installing dotnet cli...
Restoring BuildTools version 1.0.25-prerelease-00231-07...
ERROR: Could not restore build tools correctly. See '/Volumes/DATA/Build/msb/buildme/init-tools.log' for more details.
Initializing BuildTools...
./init-tools.sh: line 84: /Volumes/DATA/Build/msb/buildme/packages/Microsoft.DotNet.BuildTools/1.0.25-prerelease-00231-07/lib/init-tools.sh: No such file or directory
Done initializing tools.
Build Command: ./Tools/corerun ./Tools/MSBuild.exe ./build.proj /t:Build /p:OS=OSX /p:Configuration=Debug-NetCore /verbosity:minimal /m /fl "/flp:v=diag;logfile=./msbuild.log"
./cibuild.sh: line 42: ./Tools/corerun: No such file or directory
After looking at init-tools.log, found the following:
unning: /Volumes/DATA/Build/msb/buildme/init-tools.sh
Installing 'https://dotnetcli.blob.core.windows.net/dotnet/beta/Binaries/1.0.0-beta-002173/dotnet-dev-osx-x<IP_ADDRESS>-beta-002173.tar.gz' to '/Volumes/DATA/Build/msb/buildme/Tools/dotnetcli/dotnet.tar'
Running: /Volumes/DATA/Build/msb/buildme/Tools/dotnetcli/dotnet restore "/Volumes/DATA/Build/msb/buildme/Tools/1.0.25-prerelease-00231-07/project.json" --packages /Volumes/DATA/Build/msb/buildme/packages --source https://dotnet.myget.org/F/dotnet-buildtools/api/v3/index.json
log : Restoring packages for /Volumes/DATA/Build/msb/buildme/Tools/1.0.25-prerelease-00231-07/project.json...
error: Failed to retrieve information from remote source '/Volumes/DATA/Build/msb/buildme/packages'.
error: Unable to load the service index for source https://dotnet.myget.org/F/dotnet-buildtools/api/v3/index.json.
error: The type initializer for 'Crypto' threw an exception.
error: Unable to load DLL 'System.Security.Cryptography.Native': The specified module could not be found.
error: (Exception from HRESULT: 0x8007007E)
Running: /Volumes/DATA/Build/msb/buildme/packages/Microsoft.DotNet.BuildTools/1.0.25-prerelease-00231-07/lib/init-tools.sh /Volumes/DATA/Build/msb/buildme /Volumes/DATA/Build/msb/buildme/Tools/dotnetcli/dotnet /Volumes/DATA/Build/msb/buildme/Tools
I notice a few things:
The packages subdirectory never got created.
I can browse to https://dotnet.myget.org/F/dotnet-buildtools/api/v3/index.json even when the scripts says its index can't be loaded.
Maybe package System.Security.Cryptography.Native is outdated in the dependency mappings?
The same commit was built successfully on Jenkins at http://dotnet-ci.cloudapp.net/job/Microsoft_msbuild/job/innerloop_xplat_OSX/87/ .
Is there any pre-configuration I need to make in order to build locally?
I just verified the issue also appears on Ubuntu.
I ran ./cibuild.sh --scope Compile (same as latest successful build at http://dotnet-ci.cloudapp.net/job/Microsoft_msbuild/job/innerloop_xplat_Ubuntu/lastBuild/console), and found the same error.
CoreRun is not being downloaded/installed.
That's interesting; like the CI servers I can build on both OS X and Linux. Sadly, ERROR_MOD_NOT_FOUND is not particularly informative!
I don't think this is related to our project.json, since it appears to be a crypto failure when connecting over HTTP. It looks to me like you have a bad on-disk deployment of the dotnet CLI. But I don't know why that would be failing for you.
Do you have a System.Security.Cryptography.Native.so anywhere under Tools/dotnetcli (on Ubuntu--on OS X it's .dylib)?
On a fresh install of Ubuntu 15.10, it seems to failing running dotnet with ..
Failed to initialize CoreCLR, HRESULT: 0x80131500
Looking at it with strace, it seems to be looking for libicuuc.so.52 and libicui18n.so.52, and at least on Ubuntu 15.10 that I tried this on, libicu*so.55 is available.
But I don't have any trouble building on OSX 10.11.4 , but my OSX install might not really be "clean"!
@radical Ubuntu 15.10 isn't supported by .NET Core/CLI yet (https://github.com/dotnet/cli/issues/2072), so this is expected.
@JunielKatarn which version of Ubuntu are you running?
For OSX, I would rm -Rf bin/ Tools/ packages/ and then try the build again.
@radical It's 15.10. I guess it's not supported, then.
Also, for OS X (10.11.4), I tried using a clean checkout (nothing built previously).
Fails every time.
Maybe I misread instructions? Do I need to install something before?
Or, should the cibuild.sh script resolve/fetch all dependencies by itself?
My Mac is also not clean, but I'm pretty sure cibuild.sh should do everything you need. Do you see System.Security.Cryptography.Native.dylib under the Tools/dotnetcli folder when starting from scratch on OS X?
I do see that dylib under ./Tools/dotnetcli/shared/Microsoft.NETCore.App/1.0.0-rc2-23931/System.Security.Cryptography.Native.dylib.
I'm afraid cibuild.sh is not working for me :(
If it's any use, I here is a ZIP file containing the generated Tools subdirectory, along with the build log:
https://www.dropbox.com/sh/var5r5pbn5hwn38/AAAKitK_BwuWKvF9121TlmLma
@JunielKatarn Any progress resolving this issue? I'm running into it myself.
No progress, I'm afraid. I run into this issue both in OS X (El Capitan), and the last 3 releases of Ubuntu (including brand new 16.04).
In short, the .NET tools are not getting downloaded correctly.
@cdmihai, @rainersigwald, do you guys have any pointers?
Try to update to the xplat HEAD, and try again. Your logs suggest that it is trying to fetch an older version:
EXEC : warning : Unable to find version '1.2.0-beta-20151112-01' of package 'Microsoft.Net.Compilers'
This was updated in https://github.com/Microsoft/msbuild/commit/9dd9195913e49459d40e9e5c38977f52a856a09a
Your original post says that you are building - 5a2dcc6 but still getting the older version? I would update to the latest xplat, git-clean -xdf my checkout and try again.
@radical Updating to the head commit and cleaning the local repository goes without saying :). It's always my first course of action, when things like this don't work.
These were my exact steps as of 5 minutes ago:
Fresh clone from xplat.git clone git@github.com:Microsoft/msbuild.git --branch xplat
Verify the latest commit matches the branch head (c93ad0d05139c4b37b6448dfc89fc293d2cc5f65 as of this post).
Get into the clone's working directory, and run the build script../cibuild.sh --Target Compile
The result is the exact same error. No ./Tools/corerun file (See my comment from last week).
And can you share the logs from this build?
Ah, so the problem is that something transitively depends on an old DNX package:
EXEC : warning : Unable to find version '1.0.0-rc2-16177' of package 'dnx-coreclr-darwin-x64'. [/Volumes/DATA/Build/msb/buildme/build.proj]
find: /Volumes/DATA/Build/msb/buildme/packages/dnx-coreclr-darwin-x<IP_ADDRESS>-rc2-16177: No such file or directory
And indeed that doesn't seem to be present on https://www.myget.org/feed/aspnetvnext/package/nuget/dnx-coreclr-darwin-x64
. . . but I tried deleting my ~/.nuget folder, and the build passed without that package. And NuGet doesn't explain what the chain of dependencies was that led to it. So I'm still pretty confused.
After performing the same steps @JunielKatarn followed, this is my output:
Macintosh:msbuild appleprouduser$ ./cibuild.sh Unsupported target detected: . Configuring as if for CoreCLR Unsupported host detected: . Configuring as if for CoreCLR Installing dotnet cli... Restoring BuildTools version 1.0.25-prerelease-00231-07... ERROR: Could not restore build tools correctly. See '/Users/appleprouduser/Dropbox/MSBuild/msbuild/init-tools.log' for more details. Initializing BuildTools... ./init-tools.sh: line 84: /Users/appleprouduser/Dropbox/MSBuild/msbuild/packages/Microsoft.DotNet.BuildTools/1.0.25-prerelease-00231-07/lib/init-tools.sh: No such file or directory Done initializing tools. Build Command: ./Tools/corerun ./Tools/MSBuild.exe ./build.proj /t:Build /p:OS=OSX /p:Configuration=Debug-NetCore /verbosity:minimal /m /fl "/flp:v=diag;logfile=./msbuild.log" ./cibuild.sh: line 42: ./Tools/corerun: No such file or directory
And this is my init-tools.log file:
Running: /Users/appleprouduser/Dropbox/MSBuild/msbuild/init-tools.sh Installing 'https://dotnetcli.blob.core.windows.net/dotnet/beta/Binaries/1.0.0-beta-002173/dotnet-dev-osx-x<IP_ADDRESS>-beta-002173.tar.gz' to '/Users/appleprouduser/Dropbox/MSBuild/msbuild/Tools/dotnetcli/dotnet.tar' Running: /Users/appleprouduser/Dropbox/MSBuild/msbuild/Tools/dotnetcli/dotnet restore "/Users/appleprouduser/Dropbox/MSBuild/msbuild/Tools/1.0.25-prerelease-00231-07/project.json" --packages /Users/appleprouduser/Dropbox/MSBuild/msbuild/packages --source https://dotnet.myget.org/F/dotnet-buildtools/api/v3/index.json log : Restoring packages for /Users/appleprouduser/Dropbox/MSBuild/msbuild/Tools/1.0.25-prerelease-00231-07/project.json... error: Failed to retrieve information from remote source '/Users/appleprouduser/Dropbox/MSBuild/msbuild/packages'. error: Unable to load the service index for source https://dotnet.myget.org/F/dotnet-buildtools/api/v3/index.json. error: The type initializer for 'Crypto' threw an exception. error: Unable to load DLL 'System.Security.Cryptography.Native': The specified module could not be found. error: (Exception from HRESULT: 0x8007007E) Running: /Users/appleprouduser/Dropbox/MSBuild/msbuild/packages/Microsoft.DotNet.BuildTools/1.0.25-prerelease-00231-07/lib/init-tools.sh /Users/appleprouduser/Dropbox/MSBuild/msbuild /Users/appleprouduser/Dropbox/MSBuild/msbuild/Tools/dotnetcli/dotnet /Users/appleprouduser/Dropbox/MSBuild/msbuild/Tools
@Sarabeth-Jaffe-Microsoft that looks like the original error @JunielKatarn had. Do you see ./Tools/dotnetcli/shared/Microsoft.NETCore.App/1.0.0-rc2-23931/System.Security.Cryptography.Native.dylib?
I created a new user on my Mac and was able to clone + ./cibuild with no trouble.
@rainersigwald I'm not with my Mac right now, but at least on Ubuntu, I do see Tools/dotnetcli/shared/Microsoft.NETCore.App/1.0.0-rc2-23931/System.Security.Cryptography.Native.so.
Please note, Ubuntu has exactly the same symptoms as OS X.
@rainersigwald I do have that file. (Using Mac OSX 10.11.4 in case that matters.)
I found this in the CLI issues: https://github.com/dotnet/cli/issues/1161
Do you have openssl? We might have found an implicit system dependency :)
@cdmihai That does look awfully suspicious. brew list --versions tells me I have openssl 1.02e_1 installed--and that would be machine wide, explaining my non-repro with a new user.
I have not installed brew on my Mac.
My Ubuntu installation is also pretty out-of-the-box, so it might also lack it.
Should a third-party software like brew be required at all to get MSBuild to compile?
@JunielKatarn MSBuild doesn't (or at least shouldn't) have any dependencies on homebrew. However, since we use the dotnet CLI tool to download the packages we need (for compilers and build-and runtime references), we're at the mercy of any dependencies they choose to take.
In this case, the dependency is on having a dynamic OpenSSL library available; it shouldn't care about how that came about. brew install openssl and apt-get install openssl are easy, but you could also try building from source.
Downloaded brew & updated openssl:
Macintosh:msbuild appleprouduser$ openssl version -a
OpenSSL 1.0.2g 1 Mar 2016
built on: reproducible build, date unspecified
@rainersigwald suggested:
git clean -xdf
And it looks like it runs with no errors! Not sure if that command is all that's needed or if updating Openssl is necessary.
apt-get install openssl are easy, but you could also try building from source.
@rainersigwald Ubuntu comes with openssl installed by default.
Actually, I have the exact same version as @Sarabeth-Jaffe-Microsoft in OS X (1.0.2g).
I'll try it on OS X later tonight.
I have confirmed that upgrading OpenSSL in OS X via homebrew solved the issue.
I would keep this issue upen until the workaround steps are added to the README file, or the transitive dependency goes away.
In short (@Sarabeth-Jaffe-Microsoft , please confirm):
Install homebrew from http://brew.sh.
brew install openssl
brew link openssl --force (to put homebrew's OpenSSL in front of Apple's OpenSSL).
The issue on Ubuntu is still not resolved.
@Sarabeth-Jaffe-Microsoft git clean -xdf only cleaned up files that were downloaded/created in your repo. It worked around a problem (#590) that the initialization process decided (erroneously) that it was already complete, then failed later.
@JunielKatarn
I would keep this issue upen until the workaround steps are added to the README file, or the transitive dependency goes away.
Agreed. Since we're probably stuck with the dependency, a clear error on our part might be nice too.
brew link openssl --force (to put homebrew's OpenSSL in front of Apple's OpenSSL).
I don't think this is required (on my working Mac, openssl version returns OpenSSL 0.9.8zh 14 Jan 2016 which is the system one).
The issue on Ubuntu is still not resolved.
Ok, let's dig in further. On my working Ubuntu 14.04 system, I did this
diff --git a/init-tools.sh b/init-tools.sh
index 033a652..7fa299d 100755
--- a/init-tools.sh
+++ b/init-tools.sh
@@ -75,7 +75,7 @@ if [ ! -e $__PROJECT_JSON_FILE ]; then
if [ ! -e $__BUILD_TOOLS_PATH ]; then
echo "Restoring BuildTools version $__BUILD_TOOLS_PACKAGE_VERSION..."
echo "Running: $__DOTNET_CMD restore \"$__PROJECT_JSON_FILE\" --packages $__PACKAGES_DIR --source $__BUILDTOOLS_SOURCE" >> $__init_tools_log
- $__DOTNET_CMD restore "$__PROJECT_JSON_FILE" --packages $__PACKAGES_DIR --source $__BUILDTOOLS_SOURCE >> $__init_tools_log
+ strace -f -e trace=file $__DOTNET_CMD restore "$__PROJECT_JSON_FILE" --packages $__PACKAGES_DIR --source $__BUILDTOOLS_SOURCE >> $__init_tools_log 2>&1
if [ ! -e "$__BUILD_TOOLS_PATH/init-tools.sh" ]; then echo "ERROR: Could not restore build tools correctly. See '$__init_tools_log' for more details."; fi
fi
And I see this around loading crypto:
[pid 13479] open("/home/vagrant/src/msbuild/Tools/dotnetcli/shared/Microsoft.NETCore.App/1.0.0-rc2-23931/System.Security.Cryptography.Native.so", O_RDONLY|O_CLOEXEC) = 120
[pid 13479] open("/etc/ld.so.cache", O_RDONLY|O_CLOEXEC) = 120
[pid 13479] access("/etc/ld.so.nohwcap", F_OK) = -1 ENOENT (No such file or directory)
[pid 13479] open("/lib/x86_64-linux-gnu/libcrypto.so.1.0.0", O_RDONLY|O_CLOEXEC) = 120
[pid 13479] access("/etc/ld.so.nohwcap", F_OK) = -1 ENOENT (No such file or directory)
[pid 13479] open("/lib/x86_64-linux-gnu/libssl.so.1.0.0", O_RDONLY|O_CLOEXEC) = 120
[pid 13479] open("/dev/urandom", O_RDONLY|O_NOCTTY|O_NONBLOCK) = 120
[pid 13479] open("/usr/lib/ssl/openssl.cnf", O_RDONLY) = 120
Can you make a similar change and post init-tools.log? Hopefully there's a glaringly obvious difference . . .
On a clean Ubuntu 14.04 VM, I get this failure:
$ /home/vagrant/msbuild/Tools/dotnetcli/dotnet restore "/home/vagrant/msbuild/Tools/1.0.25-prerelease-00231-07/project.json" --packages /home/vagrant/msbuild/packages --source https://dotnet.myget.org/F/dotnet-buildtools/api/v3/index.json
Failed to load /home/vagrant/msbuild/Tools/dotnetcli/shared/Microsoft.NETCore.App/1.0.0-rc2-23931/libcoreclr.so, error: libunwind.so.8: cannot open shared object file: No such file or directory
Failed to bind to coreclr
This appears to have been resolved with sudo apt-get install libunwind8, so I think we have another hidden dependency (and needed update to the docs).
Until the hidden dependencies are resolved, I updated the build instructions in the README. Feel free to update it if I left anything out!
@Sarabeth-Jaffe-Microsoft
Is libunwind8 for both mac and linux, or only linux?
@cdmihai Looks like it's Linux and FreeBSD only, though I'm not sure how stack walking works on the Mac:
https://github.com/dotnet/coreclr/blob/775003a4c72f0acc37eab84628fcef541533ba4e/Documentation/building/osx-instructions.md mentions OpenSSL but not libunwind
https://github.com/dotnet/coreclr/blob/775003a4c72f0acc37eab84628fcef541533ba4e/Documentation/building/linux-instructions.md mentions libunwind8.
@cdmihai @rainersigwald Got it, just clarified it in the documentation.
I still see the issue under Ubuntu. Can you reopen?
@JunielKatarn What issue, exactly? Did you install libunwind8?
@rainersigwald
Yes, I have libunwind8 installed. Still getting the same error.
That said, I AM using Ubuntu 16.04, which may not be supported.
I'll try again with 14.04 and post the results here.
@JunielKatarn, can you please paste in the error you're getting? Hard to follow the thread to figure out which error you're referring to.
@JunielKatarn I'm also interested in the error, but I bet it will be that you're using 16.04. AFAIK, .NET Core is only currently supported on 14.04 (though it looks like folks might be working on modernizing.
I added that explicitly to the xplat build instructions: https://github.com/Microsoft/msbuild/wiki/Building-Testing-and-Debugging-on-.Net-Core-MSBuild/7884f4d239eedb6a47e170530d031c7d7ab28313
@rainersigwald, @cdmihai
I just set up a brand new Ubuntu 14.04 instance, installed libunwind8, and MSBuild got built successfully off the shelf.
I think it's safe to assume the issue lies with 16.04/corerun incompatibility.
|
2025-04-01T04:10:39.808741
| 2015-10-22T18:18:07
|
112860411
|
{
"authors": [
"ValMenn",
"rainersigwald"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14998",
"repo": "Microsoft/msbuild",
"url": "https://github.com/Microsoft/msbuild/pull/310"
}
|
gharchive/pull-request
|
Implement xplat Processor Architecture query
GetSystemInfo from in kernel32.dll was used for one purpose - to
obtain processor architecture. Refactor code to query for processor
architecture instead of the system info. Added appropriate property
to NativeMethodsShared and implemented equivalent for Unix (common
for Linux & OSX).
Made the changes Rainer sugested
LGTM with comment and Lazy.
Added the comment and Lazy
|
2025-04-01T04:10:39.832841
| 2017-07-11T18:33:18
|
242142936
|
{
"authors": [
"beweedon",
"msftclas"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:14999",
"repo": "Microsoft/opengcs",
"url": "https://github.com/Microsoft/opengcs/pull/29"
}
|
gharchive/pull-request
|
Fix issue in runC tests where container list not initialized
This caused the list to leak between tests, resulting in data races and
test failures.
@beweedon,
Thanks for your contribution as a Microsoft full-time employee or intern. You do not need to sign a CLA.
Thanks,
Microsoft Pull Request Bot
|
2025-04-01T04:10:39.851064
| 2017-01-20T23:03:37
|
202264464
|
{
"authors": [
"mousetraps",
"msftclas"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0004.json.gz:15000",
"repo": "Microsoft/tolerant-php-parser",
"url": "https://github.com/Microsoft/tolerant-php-parser/pull/32"
}
|
gharchive/pull-request
|
Update README with travis badge and info on where to find validation …
…test results
Hi @mousetraps, I'm your friendly neighborhood Microsoft Pull Request Bot (You can call me MSBOT). Thanks for your contribution!
It looks like you're a Microsoft contributor (Sara Itani). If you're full-time, we DON'T require a Contribution License Agreement. If you are a vendor, please DO sign the electronic Contribution License Agreement. It will take 2 minutes and there's no faxing! https://cla.microsoft.com.
TTYL, MSBOT;
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.