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
|
|---|---|---|---|---|---|
2025-04-01T06:36:59.359764
| 2019-12-23T10:12:20
|
541682507
|
{
"authors": [
"ArtRog",
"Salakar",
"Smux1demo"
],
"license": "BSD-3-Clause",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:978",
"repo": "FirebaseExtended/flutterfire",
"url": "https://github.com/FirebaseExtended/flutterfire/issues/1701"
}
|
gharchive/issue
|
[firebase_ml_vision] Cloud text recognition does not return coordinates blocks, lines and elements
When using the cloud text recognizer, the coordinates of the recognized text do not come. However, text detection works fine.
The problem does not appear when using text recognition on both platforms on the device.
See attached screenshots.
Steps to reproduce the behavior:
Use FirebaseVision.instance.cloudTextRecognizer()
Progress image File
Look at coordinates
Supporting Code:
final FirebaseVisionImage visionImage =
FirebaseVisionImage.fromFile(imageFile);
final TextRecognizer textRecognizer =
FirebaseVision.instance.cloudTextRecognizer();
VisionText visionText;
textRecognizer.processImage(visionImage).then((vText) {
if (visionText != null && visionText.blocks.isNotEmpty)
visionText.blocks.forEach((block) {
block.lines.forEach((line) {
line.elements.forEach((element) {
print(element);
});
});
});
});
Additional context
Original Image:
The result in the debugger if you use FirebaseVision.instance.textRecognizer():
The result in the debugger if you use FirebaseVision.instance.cloudTextRecognizer():
As you can see in the photo, if you use cloudTextRecognizer (), then the text is recognized, but the empty result of the cornerPoints field for both platforms is returned
Logs
λ flutter doctor -v
[√] Flutter (Channel stable, v1.12.13+hotfix.5, on Microsoft Windows [Version 10.0.18362.535], locale en-US)
• Flutter version 1.12.13+hotfix.5 at D:\Android\flutter_sdk\flutter
• Framework revision 27321ebbad (12 days ago), 2019-12-10 18:15:01 -0800
• Engine revision 2994f7e1e6
• Dart version 2.7.0
[√] Android toolchain - develop for Android devices (Android SDK version 29.0.2)
• Android SDK at D:\Android\Sdk
• Android NDK location not configured (optional; useful for native profiling support)
• Platform android-29, build-tools 29.0.2
• Java binary at: C:\Program Files\Android\Android Studio\jre\bin\java
• Java version OpenJDK Runtime Environment (build 1.8.0_202-release-1483-b03)
• All Android licenses accepted.
[√] Android Studio (version 3.5)
• Android Studio at C:\Program Files\Android\Android Studio
• Flutter plugin version 42.1.1
• Dart plugin version 191.8593
• Java version OpenJDK Runtime Environment (build 1.8.0_202-release-1483-b03)
[√] VS Code (version 1.41.1)
• VS Code at C:\Users\Arro\AppData\Local\Programs\Microsoft VS Code
• Flutter extension version 3.7.1
[√] Connected device (1 available)
• Redmi Note 4 • <IP_ADDRESS>:5555 • android-arm64 • Android 7.0 (API 24)
• No issues found!
pubspec.yaml
firebase_ml_vision: ^0.9.3+5
Given the age of this issue and it's inactivity I'm going to go ahead and close it in favour of trying the latest versions of ML Vision which ships with newer on device models. Thanks
Hi Salakar,
I've stumbled across the exact same issue (I'm using the newest plugin version: firebase_ml_vision 0.12.0+1).
I was wondering, did you manage to find a solution to this issue?
|
2025-04-01T06:36:59.375860
| 2020-10-16T13:24:08
|
723221387
|
{
"authors": [
"Fitz1991",
"TahaTesser",
"Tihan-Nico"
],
"license": "BSD-3-Clause",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:979",
"repo": "FirebaseExtended/flutterfire",
"url": "https://github.com/FirebaseExtended/flutterfire/issues/3882"
}
|
gharchive/issue
|
🐛 [firebase_storage 4.0.1] Can't upload image to firebase in flutter web version, is this possible today? Everything is fine on Android
Bug report
Describe the bug
Overflow on channel: plugins.flutter.io/firebase_storage. Messages on this channel are being discarded in FIFO fashion. The engine may not be running or you need to adjust the buffer size if of the channel.
Error: MissingPluginException(No implementation found for method StorageReference#putData on channel plugins.flutter.io/firebase_storage)
at Object.throw_ [as throw] (http://localhost:55439/dart_sdk.js:4328:11)
at MethodChannel._invokeMethod (http://localhost:55439/packages/flutter/src/services/system_channels.dart.lib.js:942:21)
at _invokeMethod.next (<anonymous>)
at http://localhost:55439/dart_sdk.js:37593:33
at _RootZone.runUnary (http://localhost:55439/dart_sdk.js:37447:58)
at _FutureListener.thenAwait.handleValue (http://localhost:55439/dart_sdk.js:32424:29)
at handleValueCallback (http://localhost:55439/dart_sdk.js:32971:49)
at Function._propagateToListeners (http://localhost:55439/dart_sdk.js:33009:17)
at _Future.new.[_completeWithValue] (http://localhost:55439/dart_sdk.js:32852:23)
at async._AsyncCallbackEntry.new.callback (http://localhost:55439/dart_sdk.js:32874:35)
at Object._microtaskLoop (http://localhost:55439/dart_sdk.js:37708:13)
at _startMicrotaskLoop (http://localhost:55439/dart_sdk.js:37714:13)
at http://localhost:55439/dart_sdk.js:33226:9
Steps to reproduce
mydependencies:
dependencies:
flutter:
sdk: flutter
cloud_firestore: ^0.14.1+2
firebase_core: ^0.5.0+1
cloud_functions: 0.7.0-dev.2
firebase_storage: ^4.0.1
firebase_auth: ^0.18.1+1
index.html:
<script src="https://www.gstatic.com/firebasejs/7.14.4/firebase-app.js"></script>
<script src="https://www.gstatic.com/firebasejs/7.14.4/firebase-firestore.js"></script>
<script src="https://www.gstatic.com/firebasejs/7.14.4/firebase-auth.js"></script>
<script src="https://www.gstatic.com/firebasejs/7.14.4/firebase-functions.js"></script>
<script src="https://www.gstatic.com/firebasejs/7.14.4/firebase-storage.js"></script>
my code:
class WebImagePicker implements ImagePickerUniversal {
Future<Map<String, dynamic>> pickImage() async {
Map<String, dynamic> data = {};
final html.FileUploadInputElement input = html.FileUploadInputElement();
input..accept = 'image/*';
input.click();
await input.onChange.first;
if (input.files.isEmpty) return null;
final reader = html.FileReader();
reader.readAsArrayBuffer(input.files[0]);
await reader.onLoad.first;
final encoded = reader.result as Uint8List;
final imageName = input.files?.first?.name;
data.addAll({'name': imageName, 'file': encoded, 'museumDir': 'Иванов'});
StorageRefere
Firebase Storage currently doesn't support Web functionality. I'd advise you to use Firebase Dart plugin for web. It has all the functionality you need.
Hi @Fitz1991
As stated above, there is no web yet in firebase_storage package
https://github.com/FirebaseExtended/flutterfire/issues/1972
Closing as duplicate
Thank you
|
2025-04-01T06:36:59.381558
| 2021-03-04T07:06:52
|
821830336
|
{
"authors": [
"NebiSarigul",
"Zeynal7",
"bhagirath",
"cbenhagen",
"dsgriffin",
"dukaric1991",
"markusaksli-nc"
],
"license": "BSD-3-Clause",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:980",
"repo": "FirebaseExtended/flutterfire",
"url": "https://github.com/FirebaseExtended/flutterfire/issues/5210"
}
|
gharchive/issue
|
🐛 [firebase_analytics] Firebase analytics web dependency problem
Firebase analytics WEB dependency problem. Need to update WEB dependency from firebase to latest version
Firebase analytics WEB depends on firebase ^7.3.0 (latest is 9.0.0)
firebase ^7.3.0 depends on http >=0.11.3 <0.13.0
And I am using http 0.13.0
If someone can, please update version of firebase in analytics_web to latest (null safe) version. Thanks.
I don't use web version, if there is any way to exclude web version, please, let me know.
Same issue here.
@dukaric1991 please just add a 👍 instead of adding "same issue here" comments. This just sends unnecessary notifications to everyone subscribed to an issue. Thanks :)
Because every version of firebase_analytics_web depends on firebase ^7.3.0 which depends on http >=0.11.3 <0.13.0, every version of firebase_analytics_web requires http >=0.11.3 <0.13.0.
And because firebase_analytics >=7.1.0 depends on firebase_analytics_web ^0.2.0, firebase_analytics >=7.1.0 requires http >=0.11.3 <0.13.0.
Important issue.
any temporary solution for this?
Hi @Zeynal7
I see there's an open issue addressing the case you described https://github.com/FirebaseExtended/flutterfire/issues/5202.
Please follow up on that issue, I'm closing the current one as a duplicate.
If you disagree, please write in the comments and I will reopen it.
Thank you
|
2025-04-01T06:36:59.408971
| 2023-01-08T06:03:56
|
1524346775
|
{
"authors": [
"Fizzadar",
"handcraftedcomputers"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:983",
"repo": "Fizzadar/pyinfra",
"url": "https://github.com/Fizzadar/pyinfra/issues/939"
}
|
gharchive/issue
|
Fact results inaccurate (distro module issue)?
Describe the bug
This could be a me issue (I am a neophyte to pyinfra). I have several "older Linux distributions" under my care. When running pyinfra inventory.py fact server.LinuxDistribution, the facts returned for these servers match the host I'm running pyinfra from, not the target. Although the files appear to be returned successfully, the distro module is unable to interpret correctly (see the following output):
$ pyinfra -vvv inventory.py fact server.LinuxDistribution
--> Loading config...
--> Loading inventory...
--> Connecting to hosts...
[server1] Connected
[server2] Connected
--> Gathering facts...
[server2] >>> sh -c 'cd /etc/ && for file in $(ls -pdL *-release | grep -v /); do echo "/etc/${file}"; cat "/etc/${file}"; echo ---; done'
[server1] >>> sh -c 'cd /etc/ && for file in $(ls -pdL *-release | grep -v /); do echo "/etc/${file}"; cat "/etc/${file}"; echo ---; done'
[server1] /etc/lsb-release
[server1] DISTRIB_ID=Ubuntu
[server1] DISTRIB_RELEASE=9.10
[server1] DISTRIB_CODENAME=karmic
[server1] DISTRIB_DESCRIPTION="Ubuntu 9.10"
[server1] ---
[server1] Loaded fact server.LinuxDistribution
[server2] /etc/centos-release
[server2] CentOS release 6.10 (Final)
[server2] ---
[server2] /etc/redhat-release
[server2] CentOS release 6.10 (Final)
[server2] ---
[server2] /etc/system-release
[server2] CentOS release 6.10 (Final)
[server2] ---
[server2] Loaded fact server.LinuxDistribution
--> Fact data for: server.LinuxDistribution
{
"server1": {
"name": "Debian",
"major": 11,
"minor": null,
"release_meta": {
"PRETTY_NAME": "Debian GNU/Linux 11 (bullseye)",
"NAME": "Debian GNU/Linux",
"VERSION_ID": "11",
"VERSION": "11 (bullseye)",
"VERSION_CODENAME": "bullseye",
"ID": "debian",
"HOME_URL": "https://www.debian.org/",
"SUPPORT_URL": "https://www.debian.org/support",
"BUG_REPORT_URL": "https://bugs.debian.org/",
"CODENAME": "bullseye"
}
},
"server2": {
"name": "Debian",
"major": 11,
"minor": null,
"release_meta": {
"PRETTY_NAME": "Debian GNU/Linux 11 (bullseye)",
"NAME": "Debian GNU/Linux",
"VERSION_ID": "11",
"VERSION": "11 (bullseye)",
"VERSION_CODENAME": "bullseye",
"ID": "debian",
"HOME_URL": "https://www.debian.org/",
"SUPPORT_URL": "https://www.debian.org/support",
"BUG_REPORT_URL": "https://bugs.debian.org/",
"CODENAME": "bullseye"
}
}
}
The result should definitely not be Debian 11 for a Centos 6 and Ubuntu 9.04 (I did say they were old) servers. While I'm not expecting facts to necessarily identify these older servers, defaulting to the underlying host's OS is unexpected behaviour.
To Reproduce
$ pyinfra inventory.py fact server.LinuxDistribution
(where inventory.py points to a couple of archaic distributions).
Host is an up to date Debian 11 laptop.
Expected behavior
Either undefined (if distro module was unable to find an appropriate match), or the appropriate response.
Meta
Include output of pyinfra --support.
--> Support information:
If you are having issues with pyinfra or wish to make feature requests, please
check out the GitHub issues at https://github.com/Fizzadar/pyinfra/issues .
When adding an issue, be sure to include the following:
System: Linux
Platform: Linux-5.10.0-20-amd64-x86_64-with-glibc2.31
Release: 5.10.0-20-amd64
Machine: x86_64
pyinfra: v2.6
Executable: /home/dt/Work/pyinfra/venv/bin/pyinfra
Python: 3.9.2 (CPython, GCC 10.2.1 20210110)
How was pyinfra installed (source/pip)?
Via pip in venv.
Consider including output with -vv and --debug.
As noted above
Additional info
Version of distro is v1.8.0. Output captured while debugging from distro's return result (note: showing os_release_file pointing inappropriately to the localhost's /usr/lib/os-release):
LinuxDistribution(os_release_file='/usr/lib/os-release', distro_release_file='/tmp/tmpis7_en3u/etc/centos-release', include_lsb=False, include_uname=False, include_oslevel=True, root_dir=None, _os_release_info={'pretty_name': 'Debian GNU/Linux 11 (bullseye)', 'name': 'Debian GNU/Linux', 'version_id': '11', 'version': '11 (bullseye)', 'version_codename': 'bullseye', 'id': 'debian', 'home_url': 'https://www.debian.org/', 'support_url': 'https://www.debian.org/support', 'bug_report_url': 'https://bugs.debian.org/', 'codename': 'bullseye', 'release_codename': 'bullseye'}, _lsb_release_info={}, _distro_release_info={'name': 'CentOS', 'version_id': '6.10', 'codename': 'Final', 'id': 'centos'}, _uname_info={}, _oslevel_info='')
Additional additional info
I am enjoying pyinfra much more than Ansible (irrespective of this issue :) )
So, poking at this further (as such is my wont), it appears that patching distro._UNIXCONFDIR as pyinfra currently does breaks the result of distro.LinuxDistribution() since at least distro v1.6.0 when the root_dir named parameter was introduced. With the attached test script, you can see the difference in the return result (using the files returned by Centos 6 fact retrieval) when using the patched distro._UNIXCONFDIR variable versus using the named root_dir parameter. The fix is relatively simple, but I'm not sure what (if anything) this would break. At a bare minimum, distro would need to be pinned >= 1.6.0 for the root_dir named parameter to work, but given that the distro package was moved subordinate to distro in v1.7.0 (i.e. from distro import distro) this would need to be done anyway.
new_rootdir_distro:
{'id': 'centos',
'major': 6,
'minor': 10,
'name': 'CentOS',
'release_meta': {}}
old_rootdir_distro:
{'id': 'debian',
'major': 11,
'minor': None,
'name': 'Debian '
'GNU/Linux',
'release_meta': {'BUG_REPORT_URL': 'https://bugs.debian.org/',
'CODENAME': 'bullseye',
'HOME_URL': 'https://www.debian.org/',
'ID': 'debian',
'NAME': 'Debian '
'GNU/Linux',
'PRETTY_NAME': 'Debian '
'GNU/Linux '
'11 '
'(bullseye)',
'SUPPORT_URL': 'https://www.debian.org/support',
'VERSION': '11 '
'(bullseye)',
'VERSION_CODENAME': 'bullseye',
'VERSION_ID': '11'}}
distro_test.py.txt
Thank you for digging into this @handcraftedcomputers - indeed it looks like there's another variable that needs patching to cover this, but I think it's a good time to bump the minimum version up to 1.6 and remove the patching hack entirely.
Made this change: https://github.com/Fizzadar/pyinfra/commit/06d61c9e9fb6980228afe02cec44a3f8cd815a1a.
Fix was released in 2.6.2!
|
2025-04-01T06:36:59.414340
| 2020-11-18T23:51:13
|
746130524
|
{
"authors": [
"BeatSkip",
"Roemer",
"sharpordie"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:984",
"repo": "FlaUI/FlaUI",
"url": "https://github.com/FlaUI/FlaUI/pull/380"
}
|
gharchive/pull-request
|
.NET 5.0 platform
Added conditional SDK selection
(As .NET 5 uses 'Microsoft.NET.Sdk' instead
of 'Microsoft.NET.Sdk.WindowsDesktop')
Added .NET 5.0 Package includes
currently validating everything
but yet to see any problems
Awesome thank you. I will also do some checks soon so we can merge this change and do a new release.
i updated the versioning and cleaned the packagereferences to the newer format that you already used for the rest. will run functional tests today in a personal project completely built on .net 5.0
Do you think we can soon use FlaUI with .NET5?
Do you think we can soon use FlaUI with .NET5?
I am actually using it on .net 5.0 currently. The project isn't finalized yet, but everything seems to work just fine @sharpordie
I am actually using it on .net 5.0 currently. The project isn't finalized yet, but everything seems to work just fine @sharpordie
I made a new commit on your behalf which should work fine: https://github.com/FlaUI/FlaUI/commit/8d736cbd36bb6b6e1957ee7765f3ab852c520326
So this PR should be obsolete. Thanks for your help!
|
2025-04-01T06:36:59.417277
| 2024-08-12T19:45:08
|
2461767020
|
{
"authors": [
"SpuckwasserDE",
"Zephyrlawg"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:985",
"repo": "FlafyDev/auto-nong-indexes",
"url": "https://github.com/FlafyDev/auto-nong-indexes/issues/2757"
}
|
gharchive/issue
|
Add Polyriddim by phonon
Song Name
Polyriddim
Artist Name
phonon
Source
Youtube
Youtube Link
zsnLglY48YrlZIK2
Direct File Link
No response
Song ID
467339
Start Offset [ms]
54
Extra
Created from Auto Nong.
Created for level: 108370062
reject
invalid YouTube link
|
2025-04-01T06:36:59.420437
| 2024-08-25T02:53:35
|
2485015074
|
{
"authors": [
"SpuckwasserDE",
"kesito533"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:986",
"repo": "FlafyDev/auto-nong-indexes",
"url": "https://github.com/FlafyDev/auto-nong-indexes/issues/3266"
}
|
gharchive/issue
|
Add Going for the kill (Toji) by Rustage
Song Name
Going for the kill (Toji)
Artist Name
Rustage
Source
Youtube
Youtube Link
oR4i3akwWY0
Direct File Link
No response
Song ID
467339
Start Offset [ms]
0
Extra
Created from Auto Nong.
Created for level: 109256508
Downloaded: Yes
reject
popular songs like "At the Speed of light" are too inconvenient to be replaced by smth unrelated and are therefore usually rejected.
Note that you can still use the song yourself, but it won't be public
|
2025-04-01T06:36:59.423110
| 2024-07-05T07:59:09
|
2392068811
|
{
"authors": [
"GDFainter",
"SpuckwasserDE"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:987",
"repo": "FlafyDev/auto-nong-indexes",
"url": "https://github.com/FlafyDev/auto-nong-indexes/issues/989"
}
|
gharchive/issue
|
Add That music by Crucified
Song Name
That music
Artist Name
Crucified
Source
Youtube
Youtube Link
DM4AZJqW5nE
Direct File Link
No response
Song ID
467339
Start Offset [ms]
5000
Extra
Created from Auto Nong.
Created for level: 107412538
reject
popular songs like "At the speed of light" are not good songs to be replaced by smth unrelated and therefore usually rejected
|
2025-04-01T06:36:59.455650
| 2016-04-27T19:31:25
|
151471491
|
{
"authors": [
"liiight"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:988",
"repo": "Flexget/Flexget",
"url": "https://github.com/Flexget/Flexget/issues/1102"
}
|
gharchive/issue
|
Movie list cli doesn't have forget option
Add forget option to fire forget event
Not actually relevant, should only fire forget events for exact titles, and movie list has none of those (just movie titles)
|
2025-04-01T06:36:59.471373
| 2015-03-10T02:37:45
|
60442301
|
{
"authors": [
"btipling"
],
"license": "apache-2.0",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:989",
"repo": "Floobits/floobits-intellij",
"url": "https://github.com/Floobits/floobits-intellij/issues/202"
}
|
gharchive/issue
|
Any kind of intellij activity causes people to follow in web editor
I think something isn't working right.
Can't reproduce now. Maybe I was in follow mode in some way.
|
2025-04-01T06:36:59.505086
| 2023-11-19T11:40:29
|
2000813858
|
{
"authors": [
"FlorianWoelki",
"MsHazelArabia"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:990",
"repo": "FlorianWoelki/obsidian-iconize",
"url": "https://github.com/FlorianWoelki/obsidian-iconize/issues/313"
}
|
gharchive/issue
|
Restart each time Obsidian restarts???
Hello there! I love this plugin - especially for jts functionality in adding custom icon sets. My problem, however, is that every single time I restart Obsidian (multiple times per day) I return to find zero icons in my file explorer. Despite having chosen and allocated icons to each folder, and despite being able to choose new icons upon restart, the ones I've already chosen are missing.
Holding down a folder (bringing up the menu) it offers the options Change and Remove Icon. Any reason for this? Im at my wit's end and will ultimately decide not to use the plugin at all if i can't make it work.
Screenshots
First instance: Icons chosen and displayed:
After reload:
Menu options after reload:
Hey, sorry to hear that. Do you use any sort of syncing provider? If so, what kind?
Does the issue only exist on mobile?
Closing because of inactivity.
|
2025-04-01T06:36:59.517709
| 2022-05-26T10:43:42
|
1249405196
|
{
"authors": [
"FloweyTheFlower420",
"TheButterbrotMan"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:991",
"repo": "FloweyTheFlower420/rebark-fabric",
"url": "https://github.com/FloweyTheFlower420/rebark-fabric/issues/2"
}
|
gharchive/issue
|
Conflicts with Simple Voice Chat
Hello, this mod prevents the client initialization of Simple Voice Chat.
Minecraft: 1.18.2
Fabric: 0.14.6
API: 0.53.4+1.18.2
Rebark: rebark-fabric-1.18.2-1.3.4
Simple Voice Chat: 1.18.2-2.2.44
Logs please.
Serverlog: https://bytebin.lucko.me/JZClT9p2R1
Clientlog: https://bytebin.lucko.me/yTyFd92a7W
I don't see any errors issues related to rebark.
Like I said, there is nothing in the log. It took 2 hours for me to identify the causing mod, your one is the causing.
What do you want me to do about it then? If I don't have any logs to show why a crash occurs, I can't reasonably debug it. Do you want me to pray to the minecraft modding gods in order to be enlightened on the patch for this bug?
Like I said, there is nothing in the log.
You have also never mentioned nothing existing in the logs.
Besides, I don't even see any crashes in the logs at all, just normal server/client shutdown, but I could be missing something.
I don't know what you did to YOUR code. But if you don't want to fix it I just remove your mod from my modpack. No need to bother with you here.
I don't know what you did to YOUR code. If praying to modding god help, does it. But if you don't want to fix it I just remove your mod from my modpack. No need to bother with you here.
I never talked about a crash, have you at least looked at the code of SVC to see if you could cause this?
|
2025-04-01T06:36:59.524425
| 2024-04-02T16:04:07
|
2220895660
|
{
"authors": [
"HenryHengZJ",
"haimco50"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:992",
"repo": "FlowiseAI/Flowise",
"url": "https://github.com/FlowiseAI/Flowise/issues/2086"
}
|
gharchive/issue
|
[BUG] RTL (right to left) content support in chatbot UI
Not sure whether this is a bug / missing feature etc...- but is it possibe to support right to left langauges (Hebrew, Arabic etc..) in the chat UI - basically adding <html dir="rtl"... in the UI container body - is it possible to have some "JS/html hook" for this ? not supporting this makes the UI looks bad mainly with response text that has some bullets numbers etc...
Describe the bug
A clear and concise description of what the bug is.
To Reproduce
Steps to reproduce the behavior:
Go to '...'
Click on '....'
Scroll down to '....'
See error
Expected behavior
A clear and concise description of what you expected to happen.
Screenshots
If applicable, add screenshots to help explain your problem.
Flow
If applicable, add exported flow in order to help replicating the problem.
Setup
Installation [e.g. docker, npx flowise start, pnpm start]
Flowise Version [e.g. 1.2.11]
OS: [e.g. macOS, Windows, Linux]
Browser [e.g. chrome, safari]
Additional context
Add any other context about the problem here.
we're using react-markdown library, you can modify to add RTL support - https://stackoverflow.com/questions/2258028/rtl-in-markdown
ok - thanks for the reference - you mean - doing it in a custom private branch ?
yeah, you can try on your forked branch
ok...thanks - but if you plan to fix/enhance it at some point then it would be nice to have a default behaviour of auto direction based on the content language and then the ability to override the default behaviour with some toggle button (RTL/LTR).
|
2025-04-01T06:36:59.528396
| 2024-12-16T09:06:26
|
2741775431
|
{
"authors": [
"HenryHengZJ",
"isikepalaku"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:993",
"repo": "FlowiseAI/Flowise",
"url": "https://github.com/FlowiseAI/Flowise/issues/3715"
}
|
gharchive/issue
|
[BUG] assistant always error?
While using Flowise via Docker, I created an assistant successfully. However, after chatting for some time, an error occurs when attempting to send a message. The error message displayed is:
Error: predictionsServices.buildChatflow - Unauthorized.
Steps to Reproduce
Set up Flowise using Docker as described in the documentation.
Create an assistant/chatbot with a valid configuration.
Begin interacting with the assistant.
After some time, attempt to send another message.
Observe the error.
Expected Behavior
The chat should proceed without interruptions, and responses should be generated successfully.
Actual Behavior
An error occurs during the interaction, preventing further communication with the assistant. The following error message appears:
Error: predictionsServices.buildChatflow - Unauthorized.
can you check if you have any API key on the chatflow? if yes, try selecting No Authorization:
|
2025-04-01T06:36:59.631016
| 2023-04-18T11:24:47
|
1672908672
|
{
"authors": [
"Clare72",
"admclachlan"
],
"license": "CC-BY-4.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:994",
"repo": "FlyBase/drosophila-anatomy-developmental-ontology",
"url": "https://github.com/FlyBase/drosophila-anatomy-developmental-ontology/issues/1624"
}
|
gharchive/issue
|
New Term Request: LHAV4
For new term requests, please provide the following information:
Preferred term label
adult lateral horn AV4 neuron (paper needs a general term for LHAV4 neurons)
Synonyms
LHAV4
Textual definition
FBrf0255018 - parent class for all LHAV4 neurons (the line in the paper probably labels some subset)
Suggested parent term
adult lateral horn neuron
Attribution
If you would like a nanoattribution, please indicate your ORCID id
I haven't done this in the past - all LH neuron terms are specific ones rather than having grouping terms for each tract. I can see the advantages of having the grouping terms, but always a balance for useful terms v bloat.
If we make this, we probably want to make all of the other grouping terms by tract too.
|
2025-04-01T06:36:59.653985
| 2024-12-01T22:17:19
|
2709944845
|
{
"authors": [
"FlysonBot"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:995",
"repo": "FlysonBot/Mastermind",
"url": "https://github.com/FlysonBot/Mastermind/issues/36"
}
|
gharchive/issue
|
refactor: Resume Game Meun Complexity
The empty message can be turned into an attribute:
https://github.com/FlysonBot/Mastermind/blob/3eb4cd2002709815a6e4ffd6bec99494c2e7b9d5/src/mastermind/ui/menu/concrete_menus.py#L104-L108
Unnecessary else statement after guard condition:
https://github.com/FlysonBot/Mastermind/blob/3eb4cd2002709815a6e4ffd6bec99494c2e7b9d5/src/mastermind/ui/menu/concrete_menus.py#L121-L124
Try-Except for integer conversion can be replaced with is-digit condition:
https://github.com/FlysonBot/Mastermind/blob/3eb4cd2002709815a6e4ffd6bec99494c2e7b9d5/src/mastermind/ui/menu/concrete_menus.py#L125-L131
Refactored: 0d411a94497e110e2d378dabfeea3a8776626fc7
|
2025-04-01T06:36:59.695700
| 2022-01-30T20:33:27
|
1118696756
|
{
"authors": [
"GregariousImpala",
"typhonrt"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:996",
"repo": "Forien/foundryvtt-forien-quest-log",
"url": "https://github.com/Forien/foundryvtt-forien-quest-log/issues/129"
}
|
gharchive/issue
|
Entity Content Links don't work
Using a quest entity content link @Quest[ID] results in a broken link, which, when hovered over, reads, Unknown Tag "Quest"
Using Foundry VTT Version 9, build 245
DnD 5e - Fifth Edition System ver. 1.5.7
HI @GregariousImpala. The maintained FQL repo is here. This is a known issue with v9 as there was a last minute breaking change ~10 days before v9 stable dropped that prevented the solution for "entity content links" w/ @Quest[ID]. A fix is to change @Quest[ID] to @JournalEntry[ID], but unfortunately that is a manual step through various areas where the link is referenced.
You can see this issue for a discussion
There is a continuation module for the quest log in the works for late Q1 release w/ an upgrade path and things will be fixed there.
|
2025-04-01T06:36:59.701960
| 2016-06-13T18:35:26
|
160016119
|
{
"authors": [
"boygirl",
"nfcampos",
"ryan-roemer"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:997",
"repo": "FormidableLabs/builder-victory-component",
"url": "https://github.com/FormidableLabs/builder-victory-component/pull/64"
}
|
gharchive/pull-request
|
opt-in lodash features
cc/ @ryan-roemer or @exogen
I audited victory for usages of lodash, and based on what I found it seems like these opt-ins should be required for LodashModuleReplacementPlugin.These weren't enabled, and nothing seemed to break in development, which gives me pretty low confidence that I understand what is being included with these feature sets. Any insights you have would be greatly appreciated.
@exogen -- Can you shepherd this?
@boygirl -- Which parts would you like input on? For example, with paths you just need to check if any of our usage of get() or similar things is using a deep (foo.bar.baz) path or just shallow (foo) paths...
I started looking into this while hunting down this bug https://github.com/FormidableLabs/victory-chart/issues/241
I am using methods like flatten and partialRight which I would expect to fail without the feature sets included in this PR, and they don't fail which tells me I'm missing something big about how this plugin is working.
@boygirl -- Interesting. Might be nice to post a gist of the diff in the unminified bundle for victory.js itself in this PR so we can look at that here when hashing this out...
I'm not entirely sure how the webpack lodash plugin works but this bit in the code https://github.com/lodash/lodash-webpack-plugin/blob/master/src/index.js#L36 seems to suggest that if you import some module directly the features it needs will be turned on
eg. if i understood it correctly
import {flatten} from 'lodash/flatten'
flatten([])
will work because the flatten module is imported directly. might explain why flatten is working even though you didn't whitelist flattening in the options
After a little more digging, the bug I was seeing was due to iteratee is not a function which would be caused by missing the shorthands feature set.
|
2025-04-01T06:36:59.855491
| 2024-10-11T03:14:52
|
2580329916
|
{
"authors": [
"danikiev",
"mrava87"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:998",
"repo": "FraCSPy/FraCSPy",
"url": "https://github.com/FraCSPy/FraCSPy/pull/71"
}
|
gharchive/pull-request
|
Implementation of Zenodo release integration
Added json file with zenodo metadata
Updated citing, changelog and index pages
Added an automatic zenodo badge to README.md
Updated date in CHANGELOG
Looks good to me, nice to have this integration in place 😀
|
2025-04-01T06:36:59.864594
| 2019-10-09T13:58:41
|
504671381
|
{
"authors": [
"AndreVarandas",
"hghazni",
"xcyberpunkx0"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:999",
"repo": "FranAmorim/teatime",
"url": "https://github.com/FranAmorim/teatime/issues/12"
}
|
gharchive/issue
|
Docs: add an awesome logo!
We need a cool logo, use your imagination! 😎
Added it in a PR to the readme + did some linting :)
https://github.com/FranAmorim/teatime/pull/17
|
2025-04-01T06:36:59.871160
| 2017-03-13T18:54:17
|
213863836
|
{
"authors": [
"Franiac",
"joelang1699",
"obs-x"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1000",
"repo": "Franiac/TwitchLeecher",
"url": "https://github.com/Franiac/TwitchLeecher/issues/113"
}
|
gharchive/issue
|
Save Favorite Broadcaster/Save to Broadcaster folder
Hello
Wonderful piece of software here!
Could you add an option to save a list of favorite broadcasters and also when saving it auto generates a folder specific to the broadcaster
Cool idea. I like 😄
A "Save favorite broadcaster seems to be what I want too, so I'm posting here.
OP mentions a "list", which is more than 1. 10 should be perfectly fine (enough), but the very simple way I imagined it first does not permit for a user-chosen amount - so maybe 5?
The way I thought of it first was basically like this: https://vgy.me/7NgXRn.png - Doing it this way there's no fuss about adding more buttons to the top/resizing them so an addition fits neatly etc, and the Search tab is displayed by default on opening so the buttons are there anyway vs having to go to a different pane between search and favorites etc. Clicking one of the buttons would simply launch a channel search for the fav.
Maybe there's a better, cleaner way to do this (the UI of this is gorgeous). I had just imagined it like this because it seems really simple/easy to add. Cheers!
Nice attempt... don't like it :D
Here's mine:
In the preferences you will be able to click some +/- buttons next to the "channel name" textbox. This way you can add unlimited favourite streamers to a list. The Textbox will be changed to a dropdownbox. This way you are able to type in the control like it was and in addition you can choose one of your favourite streamers.
The same applies to the textbox on the search panel. Just click the dropdown, choose a favourite streamer and it "search".
Pros:
Not many additional controls
Unlimited favourites
Current behaviour isn't changed
Ability to choose a favourite streamer as the "search on application startup" thingy.
Cons:
3
2
None
|
2025-04-01T06:36:59.873618
| 2020-04-05T20:09:35
|
594621532
|
{
"authors": [
"Franiac",
"Keshenatech1"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1001",
"repo": "Franiac/TwitchLeecher",
"url": "https://github.com/Franiac/TwitchLeecher/issues/482"
}
|
gharchive/issue
|
Corrupted output file on long VOD's
I was wondering if there's a maximum time or size length for a VOD to be downloaded? The reason I ask is that the streamer I mod for, Danarachel, just completed her 1st 48 hour 1 year anniversary sub-athon stream. I tried downloading the vod 3 different times with the latest version of Twitch Leacher. It took close to 8 hours to complete and didn't give any error messages, but the output mp4 file won't play and seems to be corrupted. It is 83.1 GB long but won't even show a time length when clicking on it like others do. I finally worked around the issue by using the built in crop feature and split it into 2 different 24 hour videos. Is 48 hours or 83.1 GB too large to process, or should have it worked?
In a 48hr stream it is very likely that there are some dropouts and chokes. FFMPEG cannot handle those kind of hickups very well and gives up pretty fast.
As you relized, cropping is the 1st workaround. Also you could disable mp4 conversion in the preferences and go with the original .ts file.
To my knowledge there is no hard size limit to anything in TL.
|
2025-04-01T06:36:59.977263
| 2023-03-10T06:06:54
|
1618417874
|
{
"authors": [
"Franklalalala"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1002",
"repo": "Franklalalala/AutoSteper",
"url": "https://github.com/Franklalalala/AutoSteper/issues/1"
}
|
gharchive/issue
|
A better arrangement of auxiliary project files
Currently, the auxiliary files in this project is disordered.
Several auxiliary project files need to be changed.
The change log is proposed to use the release-please github action.
The pyproject set a limit on the size of project files, therefore some crucial test files have been deleted.
The installation procedure needs to be optimized with setup.py, etc. The optimal scenario is to install with a single commandline.
The doc file could be decorated with Sphinx.
I believe these code optimization could promote AutoSteper to a broader userbase.
The pyproject.toml controls the building process. After a suitable setup of pyproject.toml, now this project could be install in a single commandline.
However, the FullereneDataParser package could only be installed from source code.
And the usenauty module requires a high version of gcc.
Both of them should be prepared with instruction in readme.
Unfortunately, the building process of pyproject will compile all the test files.
They needs to be squeezed with no better solution.
I was wrong about last comment. The Source Distribution could be controled by MANIFEST.in.
Now all the test files are restored.
As to the latest commit, the release please bot has been sucessfully implemented.
New commits should stay in line with Conventional Commit messages.
In the latest version of autosteper, documentation could be found in https://autosteper.readthedocs.io/
|
2025-04-01T06:37:00.084522
| 2016-09-06T16:06:12
|
175286945
|
{
"authors": [
"DaMexican",
"atjonathan",
"computerluca"
],
"license": "BSD-3-Clause",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1003",
"repo": "FreeCodeCamp/FreeCodeCamp",
"url": "https://github.com/FreeCodeCamp/FreeCodeCamp/issues/10490"
}
|
gharchive/issue
|
What is it that I'm doing wrong?
Challenge Use CSS Selectors to Style Elements has an issue.
User Agent is: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_7_5) AppleWebKit/537.78.2 (KHTML, like Gecko) Version/6.1.6 Safari/537.78.2.
Please describe how to reproduce this issue, and include links to screenshots if possible.
Happy Coding!
|
2025-04-01T06:37:00.087860
| 2015-12-08T18:17:25
|
121066923
|
{
"authors": [
"SaintPeter",
"ltegman",
"smiller171"
],
"license": "BSD-3-Clause",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1004",
"repo": "FreeCodeCamp/FreeCodeCamp",
"url": "https://github.com/FreeCodeCamp/FreeCodeCamp/issues/5083"
}
|
gharchive/issue
|
Waypoint Use Comments to Clarify Code - Test does not catch multiple comments
http://www.freecodecamp.com/challenges/waypoint-use-comments-to-clarify-code
supposed to create a specific comment. does not check proper comment syntax. The following code passes even though the comment syntax is wrong and it is rendering the intended comment as text:
<!-->Only change code above this line.</!-->
<div class="container-fluid">
<h3 class="text-primary text-center">jQuery Playground</h3>
<div class="row">
<div class="col-xs-6">
<h4>#left-well</h4>
<div class="well" id="left-well">
<button class="btn btn-default target" id="target1">#target1</button>
<button class="btn btn-default target" id="target2">#target2</button>
<button class="btn btn-default target" id="target3">#target3</button>
</div>
</div>
<div class="col-xs-6">
<h4>#right-well</h4>
<div class="well" id="right-well">
<button class="btn btn-default target" id="target4">#target4</button>
<button class="btn btn-default target" id="target5">#target5</button>
<button class="btn btn-default target" id="target6">#target6</button>
</div>
</div>
</div>
</div>
Confirmed.
Tests should likely be updated to check for one and only one comment opener and comment closer.
Possibly test for the presence of the text between the comment markers.
Seems like a fun regex problem. If someone else really wants it they can take it, but if not I'd like to work on this tonight.
Haha - you can have dibs. Hope you have your COM1000 up and running.
|
2025-04-01T06:37:00.092283
| 2016-07-15T14:51:17
|
165804879
|
{
"authors": [
"Bouncey",
"dungtran211096",
"erictleung",
"mytreasures"
],
"license": "BSD-3-Clause",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1005",
"repo": "FreeCodeCamp/FreeCodeCamp",
"url": "https://github.com/FreeCodeCamp/FreeCodeCamp/issues/9737"
}
|
gharchive/issue
|
Stand in Line
Challenge Stand in Line has an issue.
User Agent is: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/51.0.2704.106 Safari/537.36.
Please describe how to reproduce this issue, and include links to screenshots if possible.
My code:
function nextInLine(arr, item) {
// Your code here
arr.push(item);
arr.shift([0]);
return item; // Change this line
}
// Test Setup
var testArr = [1,2,3,4,5];
// Display Code
console.log("Before: " + JSON.stringify(testArr));
console.log(nextInLine(testArr, 6)); // Modify this line to test
console.log("After: " + JSON.stringify(testArr));
@dungtran211096 Can you please provide more information regarding this issue
@dungtran211096 your code is incorrect. Please review how to use the .shift() method. If you get stuck with a challenge, please first ask questions to the Help Room. Happy coding!
|
2025-04-01T06:37:00.114305
| 2015-03-26T17:37:11
|
64590268
|
{
"authors": [
"bmiklautz",
"shattars3d"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1006",
"repo": "FreeRDP/FreeRDP",
"url": "https://github.com/FreeRDP/FreeRDP/issues/2500"
}
|
gharchive/issue
|
Update my older freerdp commit
hey, i was just wondering if i was working on a older commit of freerdp, and i can update it to the latest master without having over writing my fixes i had for the older freerdp core files..?
is that possible?..
this url link is the comparison of the master and the one i am using..
https://github.com/FreeRDP/FreeRDP/compare/90936e64845195e5bea396ed90c63635cce2f0a3...master
@shattars3d I would recommend to read some howto or book on git.
Depending on what you want todo you can either just merge the latest master changes ore rebase your changes onto master. In both cases if there a conflicting changes you need to resolve the conflict.
If you need help join our irc channel or send a mail to the mailing list but pleas don't use the issue tracker as help forum.
|
2025-04-01T06:37:00.168139
| 2016-02-17T08:33:37
|
134214644
|
{
"authors": [
"HenryJacques",
"akallabeth",
"awdAvenger",
"bmiklautz",
"ghabxph",
"gtjbtits",
"nfedera",
"olucaslrc",
"sachinh1980"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1007",
"repo": "FreeRDP/FreeRDP",
"url": "https://github.com/FreeRDP/FreeRDP/issues/3139"
}
|
gharchive/issue
|
Connection from linux to Windows 10 not working after latest Win10 update (ERRCONNECT_CONNECT_TRANSPORT_FAILED)
After a reboot due to an update last week, I can no longer connect to my Windows 10 Professional computer from my linux computer using xfreerdp.
I get the following output when trying to connect using the latest version from git (2a3e9996):
[09:19:16:178] [21254:21255] [ERROR][com.freerdp.core.transport] - BIO_read returned a system error 104: Connection reset by peer
[09:19:16:178] [21254:21255] [ERROR][com.freerdp.core] - freerdp_set_last_error ERRCONNECT_CONNECT_TRANSPORT_FAILED [0x2000D]
[09:19:16:178] [21254:21255] [ERROR][com.freerdp.client.x11] - Freerdp connect error exit status 1
This was with a simple command line with /u: and /v:. I have tried with /gt:rpc as well because I've seen somewhere else, but it didn't work.
Note that I can connect to this Windows 10 machine using the Remote Desktop Connection application in Windows XP, and xfreerdp is able to connect to the windows XP machine.
Winver.exe on the Windows 10 machine reports: Version 1511 (OS Build 10586.104).
Let me know if there's anything else that I can do.
I can also note that the connection works fine with the 'rdesktop' utility.
@awdAvenger i this via a gateway? Otherwise /gt wont bring anything. I just tested against a windows 10 test machine and did't see this problem.
Can you check if you see any error in the windows log after connecting? Also it might help us if you run xfreerdp with the environment variable WLOG_LEVEL set to DEBUG:
WLOG_LEVEL=DEBUG
Also no issue here connecting to Windows 10 V.1511 Build 10586.104, neither with 2a3e999 nor with current master (b4b8239)
This is not though a gateway, it's just a simple Win 10 Pro laptop, connected to a domain.
The output from xfreerdp with DEBUG is:
$ WLOG_LEVEL=DEBUG xfreerdp /u:JOTRON\knutid /v:knutid-laptop
[09:44:47:990] [10115:10115] [DEBUG][com.freerdp.client.common.cmdline] - windows: 0/1 posix: -1000/0 compat: 1/0
[09:44:47:993] [10115:10116] [DEBUG][com.freerdp.client.x11] - Searching for XInput pointer device
[09:44:47:993] [10115:10116] [DEBUG][com.freerdp.client.x11] - Pointer device: 10
[09:44:47:994] [10115:10116] [DEBUG][com.freerdp.core.nego] - Enabling security layer negotiation: TRUE
[09:44:47:994] [10115:10116] [DEBUG][com.freerdp.core.nego] - Enabling restricted admin mode: FALSE
[09:44:47:994] [10115:10116] [DEBUG][com.freerdp.core.nego] - Enabling RDP security: TRUE
[09:44:47:994] [10115:10116] [DEBUG][com.freerdp.core.nego] - Enabling TLS security: TRUE
[09:44:47:994] [10115:10116] [DEBUG][com.freerdp.core.nego] - Enabling NLA security: TRUE
[09:44:47:994] [10115:10116] [DEBUG][com.freerdp.core.nego] - Enabling NLA extended security: FALSE
[09:44:47:994] [10115:10116] [DEBUG][com.freerdp.core.nego] - state: NEGO_STATE_NLA
[09:44:47:994] [10115:10116] [DEBUG][com.freerdp.core.nego] - Attempting NLA security
[09:44:47:018] [10115:10116] [DEBUG][com.freerdp.core.nego] - RequestedProtocols: 3
[09:44:47:021] [10115:10116] [DEBUG][com.freerdp.core.nego] - RDP_NEG_RSP
[09:44:47:021] [10115:10116] [DEBUG][com.freerdp.core.nego] - selected_protocol: 2
[09:44:47:021] [10115:10116] [DEBUG][com.freerdp.core.nego] - state: NEGO_STATE_FINAL
[09:44:47:021] [10115:10116] [DEBUG][com.freerdp.core.nego] - Negotiated NLA security
[09:44:47:021] [10115:10116] [DEBUG][com.freerdp.core.nego] - nego_security_connect with PROTOCOL_NLA
[09:44:47:030] [10115:10116] [DEBUG][com.winpr.utils] - Could not open SAM file!
Password:
[09:44:52:034] [10115:10116] [DEBUG][com.winpr.sspi] - InitSecurityInterfaceExA
[09:44:52:034] [10115:10116] [DEBUG][com.freerdp.core.nla] - Sending Authentication Token
[09:44:52:034] [10115:10116] [DEBUG][com.freerdp.core.nla] - 0000 4e 54 4c 4d 53 53 50 00 01 00 00 00 b7 82 08 e2 NTLMSSP.........
[09:44:52:034] [10115:10116] [DEBUG][com.freerdp.core.nla] - 0010 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
[09:44:52:034] [10115:10116] [DEBUG][com.freerdp.core.nla] - 0020 06 01 b1 1d 00 00 00 0f ........
[09:44:52:036] [10115:10116] [DEBUG][com.freerdp.core.nla] - Sending Authentication Token
[09:44:52:036] [10115:10116] [DEBUG][com.freerdp.core.nla] - 0000 4e 54 4c 4d 53 53 50 00 03 00 00 00 18 00 18 00 NTLMSSP.........
[09:44:52:036] [10115:10116] [DEBUG][com.freerdp.core.nla] - 0010 78 00 00 00 30 01 30 01 90 00 00 00 0c 00 0c 00 x...0.0.........
[09:44:52:036] [10115:10116] [DEBUG][com.freerdp.core.nla] - 0020 58 00 00 00 0c 00 0c 00 64 00 00 00 08 00 08 00 X.......d.......
[09:44:52:036] [10115:10116] [DEBUG][com.freerdp.core.nla] - 0030 70 00 00 00 10 00 10 00 c0 01 00 00 35 b2 88 e2 p...........5...
[09:44:52:036] [10115:10116] [DEBUG][com.freerdp.core.nla] - 0040 06 01 b1 1d 00 00 00 0f af 62 f8 f2 a5 de 14 70 .........b.....p
[09:44:52:036] [10115:10116] [DEBUG][com.freerdp.core.nla] - 0050 17 46 60 e6 8b 06 92 78 4a 00 4f 00 54 00 52 00 .F`....xJ.O.T.R.
[09:44:52:036] [10115:10116] [DEBUG][com.freerdp.core.nla] - 0060 4f 00 4e 00 6b 00 6e 00 75 00 74 00 69 00 64 00 O.N.k.n.u.t.i.d.
[09:44:52:036] [10115:10116] [DEBUG][com.freerdp.core.nla] - 0070 6e 00 6f 00 6e 00 65 00 47 a9 da 2b d1 05 86 59 n.o.n.e.G..+...Y
[09:44:52:036] [10115:10116] [DEBUG][com.freerdp.core.nla] - 0080 ac f5 16 c9 b4 f6 ad 09 4e 96 36 3a 3c ef b0 d1 ........N.6:<...
[09:44:52:036] [10115:10116] [DEBUG][com.freerdp.core.nla] - 0090 2a bd 2d 3b ac ba 22 65 9d 18 3d 30 31 1f 01 67 .-;.."e..=01..g
[09:44:52:036] [10115:10116] [DEBUG][com.freerdp.core.nla] - 00a0 01 01 00 00 00 00 00 00 4c ac 94 9f 96 73 d1 01 ........L....s..
[09:44:52:036] [10115:10116] [DEBUG][com.freerdp.core.nla] - 00b0 4e 96 36 3a 3c ef b0 d1 00 00 00 00 02 00 0c 00 N.6:<...........
[09:44:52:036] [10115:10116] [DEBUG][com.freerdp.core.nla] - 00c0 4a 00 4f 00 54 00 52 00 4f 00 4e 00 01 00 1a 00 J.O.T.R.O.N.....
[09:44:52:036] [10115:10116] [DEBUG][com.freerdp.core.nla] - 00d0 4b 00 4e 00 55 00 54 00 49 00 44 00 2d 00 4c 00 K.N.U.T.I.D.-.L.
[09:44:52:036] [10115:10116] [DEBUG][com.freerdp.core.nla] - 00e0 41 00 50 00 54 00 4f 00 50 00 04 00 18 00 4a 00 A.P.T.O.P.....J.
[09:44:52:036] [10115:10116] [DEBUG][com.freerdp.core.nla] - 00f0 6f 00 74 00 72 00 6f 00 6e 00 2e 00 6c 00 6f 00 o.t.r.o.n...l.o.
[09:44:52:036] [10115:10116] [DEBUG][com.freerdp.core.nla] - 0100 63 00 61 00 6c 00 03 00 34 00 6b 00 6e 00 75 00 c.a.l...4.k.n.u.
[09:44:52:036] [10115:10116] [DEBUG][com.freerdp.core.nla] - 0110 74 00 69 00 64 00 2d 00 6c 00 61 00 70 00 74 00 t.i.d.-.l.a.p.t.
[09:44:52:036] [10115:10116] [DEBUG][com.freerdp.core.nla] - 0120 6f 00 70 00 2e 00 4a 00 6f 00 74 00 72 00 6f 00 o.p...J.o.t.r.o.
[09:44:52:036] [10115:10116] [DEBUG][com.freerdp.core.nla] - 0130 6e 00 2e 00 6c 00 6f 00 63 00 61 00 6c 00 05 00 n...l.o.c.a.l...
[09:44:52:036] [10115:10116] [DEBUG][com.freerdp.core.nla] - 0140 18 00 4a 00 6f 00 74 00 72 00 6f 00 6e 00 2e 00 ..J.o.t.r.o.n...
[09:44:52:036] [10115:10116] [DEBUG][com.freerdp.core.nla] - 0150 6c 00 6f 00 63 00 61 00 6c 00 07 00 08 00 4c ac l.o.c.a.l.....L.
[09:44:52:036] [10115:10116] [DEBUG][com.freerdp.core.nla] - 0160 94 9f 96 73 d1 01 06 00 04 00 02 00 00 00 0a 00 ...s............
[09:44:52:036] [10115:10116] [DEBUG][com.freerdp.core.nla] - 0170 10 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
[09:44:52:036] [10115:10116] [DEBUG][com.freerdp.core.nla] - 0180 00 00 09 00 2a 00 54 00 45 00 52 00 4d 00 53 00 .....T.E.R.M.S.
[09:44:52:036] [10115:10116] [DEBUG][com.freerdp.core.nla] - 0190 52 00 56 00 2f 00 6b 00 6e 00 75 00 74 00 69 00 R.V./.k.n.u.t.i.
[09:44:52:036] [10115:10116] [DEBUG][com.freerdp.core.nla] - 01a0 64 00 2d 00 6c 00 61 00 70 00 74 00 6f 00 70 00 d.-.l.a.p.t.o.p.
[09:44:52:036] [10115:10116] [DEBUG][com.freerdp.core.nla] - 01b0 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
[09:44:52:036] [10115:10116] [DEBUG][com.freerdp.core.nla] - 01c0 29 a6 82 81 50 78 6d 3c c0 13 b9 24 2c 1d f1 93 )...Pxm<...$,...
[09:44:55:976] [10115:10116] [ERROR][com.freerdp.core.transport] - BIO_read returned a system error 104: Connection reset by peer
[09:44:55:976] [10115:10116] [DEBUG][com.freerdp.core.transport] - transport_check_fds: transport_read_pdu() - -1
[09:44:55:976] [10115:10116] [DEBUG][com.freerdp.core.rdp] - transport_check_fds() - -1
[09:44:55:976] [10115:10116] [ERROR][com.freerdp.core] - freerdp_set_last_error ERRCONNECT_CONNECT_TRANSPORT_FAILED [0x2000D]
[09:44:55:976] [10115:10116] [ERROR][com.freerdp.client.x11] - Freerdp connect error exit status 1
On the windows side, I get an Audit Success message in the Security logs with a Login and then a Logoff right after that. Nothing else can be seen in the event viewer.
Shoudn't you escape the domain separator? Maybe simply try xfreerdp /u:knutid /d:JOTRON
I actually did escape it, not sure why that didn't show on the pasting. I tried with /d: as well, but it still doesn't work and I don't think it's directly related to authentication credentials, as Windows says it has authenticated successfully.
Since you have a working and non-working revision you could use git bisect to identify the bad commit.
Well, they are different applications, I know freerdp is a fork of rdesktop, but I think going back that far probably won't be helpful?
I compared Wireshark output of rdesktop and freerdp and found that both receive the same RST from Windows, but the rdesktop application then initiates another connection which then proceeds to complete the connection process. I cannot make much sense of the actual data sent though, as most appears to be binary structures and/or encrypted.
@awdAvenger just to clarify FreeRDP < 1.0 was initially a fork but with 1.0 FreeRDP it was completely rewritten from scratch.
Regarding your problem can you try with /sec:rdp and/or with /sec:tls?
With /src:rdp and /sec:tls it works!
@awdAvenger so it seems that there is a problem with NLA in your case.
So it would appear. I can work with the other security protocols, so thanks a lot for your help. If you decide to track down this issue more, let me know if you need some additional information.
@awdAvenger may I ask what client you are on (distribution, system architecture)?
@awdAvenger Do you also have the issue if you log on as a local user instead of a domain user? (Use something like xfreerdp /u:knutid /d:knutid-laptop /v:knutid-laptop. Also what have you configured in System Properties -> Remote Tab ?
@awdAvenger You might also want to post a registry dump of the following key: HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Terminal Server\WinStations\RDP-Tcp
I am on Arch Linux 64 bit, I used the package in the repositories, but I also tried to build the package from git. I tried to downgrade the package but it did not help, so I am pretty confident that that the change happened on the Windows side, unless a library freerdp depends on changed.
In the remote settings I have allowed remote connections and have not checked the 'Allow connections only from computers running with NLA'. No change from the defaults have been done to the users list. It says my user has access and the list is empty.
Connecting as my non-domain user works as well, even with NLA security.
I have attached the registry dump as requested.
regdump.zip
@awdAvenger thx, this reg key is fine. However, if it works with a local user you seem to have some domain policies in place. Ask your Domain Administrator and/or run rsop.msc (elevated) to check the applied policies. Concentrate on entries under Computer Configuration -> Administrative Templates -> Windows Compnents -> Remote Desktop Services
@awdAvenger @nfedera
I've already encountered such behaviour and indeed it had something to do with session allowed only on specific workstations.
In dsa.msc -> User properties window -> Account Tab -> Log on to
See
https://msdn.microsoft.com/en-us/library/bb742516.kerb02_big(l=en-us).gif
and
http://windowsitpro.com/site-files/windowsitpro.com/files/archive/windowsitpro.com/content/content/104709/fig4_this_user_can_log_on_to.jpg
I've just tested it with a Windows 2012 R2 DC and a Windows 7 Entreprise.
If I put the netbios name of the workstation and I cannot login anymore using NLA (ERRCONNECT_TRANSPORT_FAILED).
If I use rdp, it works.
@HenryJacques Thanks! Confirmed. @awdAvenger can you also confirm that this setting is enabled for your user? @HenryJacques However, even with Windows mstsc I'm not able to login via NLA if "Logon to .." is configured.
I will try to check the user settings for my user. In the mean time, I can tell that I don't know if rdesktop used NLA at all. And the remote client on windows that worked was on Windows XP, and I don't think NLA is supported on that either.
@nfedera with mstsc I cannot use NLA myself.
What kind of error do you have?
I've created a rdp file to configure specific option with mstsc:
with negotiate security layer:i:0 I get an error message and cannot connect at all
with negotiate security layer:i:1 I get no error, the remote session starts but do not automatically login. Instead I have the Windows session prompt. I have to re-enter the password and then it works
@awdAvenger rdesktop doesn't support NLA.
Windows XP SP3 supports NLA but not from scratch, you have to do a special configuration.
See https://support.microsoft.com/en-en/kb/951608
@nfedera to complete my previous post, I ran a wireshark session and I saw that mstc talks directly to the DC. I get a Kerberos Error (STATUS_INVALID_WORKSTATION) in both cases.
If negotiate security layer:i:1 is set, it only fall back to standard RDP security.
@HenryJacques mstsc under Windows 10 shows better error messages. If I rdp from Windows 10 domain workstation A to win10 domain workstation B and the user has only A in the "Logon to" list mstsc reports: `The system administrator has limited the computers you can log on with. Try logging on at a different computer``. This does not seem to make any sense because the computer B is in the list. If I add both computers A and B to the "Logon to" list it works however.
It even works for non-domain, non-windows clients. If I add both, the target computer (in domain) and the dns name of a (non domain) linux machine, to a user's "Logon To ..." list in dsa.msc, xfreerdp /sec:nla from this linux pc to that windows 10 domain computer works perfectly fine.
So I guess we can close this issue - @awdAvenger ok ?
@nfedera great finding! I'd never had thought to put the A computer in the allowed workstations list. Maybe in Microsoft's mind if you want to login on B and you're logged on A (with the same account though), you have to be able to open a session on A first...
I also have the same conclusion concerning your very last post, I can make it work with Linux and xfreerdp. If I put the hostname of my linux client in the authorized workstations list.
However, I think there is a bug: I use the /hostname:XXX option, and setting XXX in the workstation list doesn't work. I have to put the real hostname.
@HenryJacques elaborate on the last sentence. There is no /hostname: option
@nfedera sorry ;) I meant the /client-hostname: option
@HenryJacques Ok, you were trying to fake it. It would have been embarrassing (for MS) if that actually worked. I don't think this is a bug ;)
@nfedera Ok, this option is just meant as a fake? I used it in conjunction with the /drive: so that there is "friendly" name.
So, this is a normal behaviour for you?
@HenryJacques In the client core data block (https://msdn.microsoft.com/en-us/library/cc240510.aspx) you can see the clientName field which is describe as "name of the client computer.". FreeRDP sets this value to the hostname here and overwrites it here if /client-hostname is specified. So this isn't a secure setting and should not be trusted on the server and only be used as a client provided friendly name. And yes, for me it is normal behaviour that the server gives the peer name/address of the connection priority over the name the client says that he has.
Seems that this is not a FreeRDP bug, closing this issue.
@awdAvenger if it turns out it wasn't the problem you where hitting please re-open.
@nfedera thanks for the clarification!
For people who may have this kind of issue:
In your Windows 10 Machine, Go to: System Properties > Remote > Uncheck "Allow connections only from computers running Remote Desktop with Network Level Authentication (recommended)"
Then in your linux machine's terminal, type:
xfreerdp /sec:tls /u: /v:
I have this kind of issue for a very very long time, and disabling NLA in my windows local machine, and forcing connection to TLS solves my problem.
I like to use xfreerdp because it has multi-monitor support, unlike remmina.
Hope this helps everyone.
For people who may have this kind of issue:
In your Windows 10 Machine, Go to: System Properties > Remote > Uncheck "Allow connections only from computers running Remote Desktop with Network Level Authentication (recommended)"
Then in your linux machine's terminal, type:
xfreerdp /sec:tls /u:{username} /v:{IP}
I have this kind of issue for a very very long time, and disabling NLA in my windows local machine, and forcing connection to TLS solves my problem.
I like to use xfreerdp because it has multi-monitor support, unlike remmina.
Hope this helps everyone.
Thanks for the hint. But does it not mean that we are disabling the recommended settings for Remote-Connection on Windows Server?
In my case, i just configure the RDP settings on Windows with Services, starting the Remote Access Auto Connection Manager service. After that i can using FreeRDP or other Linux RDP apps to connect on Windows.
since I´ve been reminded due to comments to this issue:
DO NOT DISABLE NLA if you do not have a very good reason for that. FreeRDP works fine with it, if you have issues check your network and credentials first.
For the newbies, like me:
The reason that you'll get an "BIO_read returned a system error 104: Connection reset by peer" error right after "Sending Authentication Token / Client: Sending PubKeyAuth Token" might be in wrong credentials. Check them carefully one more time.
P.S.: Download freerdp through homebrew might be a bad idea, cause homebrew tries to download a half of the internet to satisfy all of dependencies. Using vm with Desktop Ubuntu (for example) safe a lot time for you.
|
2025-04-01T06:37:00.175358
| 2022-03-14T10:09:07
|
1168151723
|
{
"authors": [
"akallabeth",
"rvigliotti-sf"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1008",
"repo": "FreeRDP/FreeRDP",
"url": "https://github.com/FreeRDP/FreeRDP/issues/7723"
}
|
gharchive/issue
|
RDP through an http proxy that does not support socks5
I am attempting to connect to a RDP host through a HTTP Public Proxy and from the xfreerdp manual it says I need to specify socks5:// instead of http:// when attempting to connect to a HTTP proxy, but I've been told the proxy I'm connecting through does not support socks5 - is there a way to force xfreerdp to use http proxies that uses HTTP CONNECT instead of SOCKS ? Thanks!
This is FreeRDP version 2.1.1 (n/a)
Which manual are you referring to? Here is help output for xfreerdp:
/proxy:[<proto>://][<user>:<password>@]<host>:<port>
Proxy settings: override env. var (see
also environment variable below). Protocol
"socks5" should be given explicitly where
"http" is default.
@akallabeth I'm referring to that manual, correct
@rvigliotti-sf please add some log, your usage of crash indicates you mean something different.
Also note that 2.1.1 is a very old release of the 2.0 series.
This is using SOCKS5
This is using HTTP (thats what I mean by crash)
@rvigliotti-sf ok. You tried a newer release already? (there were 10 bugfix releases in between)
@akallabeth not yet, I will update this one once I've updated - thanks!
@rvigliotti-sf also please note for asking questions please contact us on matrix/irc first, it might already be answered elsewhere. (see https://freerdp.com for addresses)
|
2025-04-01T06:37:00.180022
| 2023-06-22T23:03:23
|
1770502951
|
{
"authors": [
"akallabeth",
"hardening",
"hlefebvre"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1009",
"repo": "FreeRDP/FreeRDP",
"url": "https://github.com/FreeRDP/FreeRDP/pull/9088"
}
|
gharchive/pull-request
|
Adding CredSSP with Early User Auth support (sec:ext)
Recently I wasn't able to connect to my work servers using FreeRDP.
After digging I found that without using /sec:ext the server was closing the connection.
But with the command line switch it wasn't working, I found out that HYBRID_EX wasn't handled in nego.c.
Afterward I found out that we were receiving 4 bytes between the end of NLA and the MCS connect, I found out that it was the Early User Auth Result PDU (https://learn.microsoft.com/en-us/openspecs/windows_protocols/ms-rdpbcgr/d0e560a3-25cb-4563-8bdc-6c4cc625bbfc).
I don't know RDP protocol neither the FreeRDP source code but I attempted to add support the best I can, trying to respect the way the code is architectured and using clang_format on the modified sources.
With this patch I can now connect to my work servers again, if it can help other people then I'm happy to share the patch.
@freerdp-bot test please
LGTM, nice patch, thank you for your contribution!
if you could squash the commits into one and fix the (minor) remarks it would be really great!
ok I fixed everything and squashed it
Fix my remark and we're ok (we'll ask for test bots after that)
@freerdp-bot can you test buddy ?
I've fixed the wrong indentation for the enum
@freerdp-bot test please
@freerdp-bot test please
|
2025-04-01T06:37:00.237983
| 2022-02-09T21:37:29
|
1129046183
|
{
"authors": [
"AlOneill",
"AnneV-Learn"
],
"license": "apache-2.0",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1010",
"repo": "FreeUKGen/FreeCENMigration",
"url": "https://github.com/FreeUKGen/FreeCENMigration/issues/1372"
}
|
gharchive/issue
|
Update Help page for new Records section pages
Have received updated text from @AnneV-Learn:
"Select RECORDS from the main menu to see statistics relating to the number of census Pieces currently in the database for each census year and details of recent additions (i.e. Pieces added to the database in the previous 30 days). Censuses are broken into Pieces by the National Archive (a Piece consists of a related set of census pages). From here you can choose a county and view the statistics for that county or select a specific place in the county and view that detail. At the County and Place level you can view details of the Pieces (including Civil Parish names). Where a Piece is online (i.e. has been transcribed by our volunteers and uploaded into the database) you can also get a list of names (surnames/forenames) in the database. We hope this detail will help you to refine your SEARCH criteria to get the best from the database."
Work in progress on DRAFT Help.
@AnneV-Learn With a few tweaks to the text as given, the 'Database contents' section has been updated on t3:
https://test3.freecen.org.uk/cms/draft-help#dbContents
Thank you @AlOneill, it looks good to me
Page promoted. It is sync-ready.
Looks ok to me so probably ready for deployment now.
|
2025-04-01T06:37:00.439560
| 2017-07-18T13:03:04
|
243708082
|
{
"authors": [
"Nyholm",
"jkobus"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1013",
"repo": "FriendsOfApi/localise.biz",
"url": "https://github.com/FriendsOfApi/localise.biz/issues/19"
}
|
gharchive/issue
|
Client does not support keys containing "spaces"
Keys like "submit button" are not supported.
Translation asset is created (its sent in POST body), but when we try to add translation it fails:
Translation.php:62
$response = $this->httpPostRaw(sprintf('/api/translations/%s/%s?key=%s', urlencode($id), $locale, $projectKey), $translation);
this becomes:
/api/translations/some translation key/pl?key=12
Request that is actually beeing sent is directed to:
/api/translations/some
... and that results in an error response from api: "Missing API key"
This might be related to how the http client implementation handles such cases, I used the php-http/curl-client.
PR is on the way.
Fixed by #20
|
2025-04-01T06:37:00.458877
| 2018-03-14T18:22:40
|
305276189
|
{
"authors": [
"SpacePossum",
"keradus",
"ntzm"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1014",
"repo": "FriendsOfPHP/PHP-CS-Fixer",
"url": "https://github.com/FriendsOfPHP/PHP-CS-Fixer/pull/3600"
}
|
gharchive/pull-request
|
Fix strict_param when importing
Ref #3599
closes #3599
thanks @ntzm looking good :)
I would place the check before calling fixFunction as here we are not fixing a function but dealing with an import,
in other words; for single responsibility principle sense leaving fixFunction to fix a function and the caller being responsible for detecting if it is a function that needs fixing (hope it makes sense ;) )
Ah thanks @SpacePossum, that's a good point. Will do
Thank you @ntzm.
|
2025-04-01T06:37:00.474014
| 2020-07-03T16:51:17
|
650695076
|
{
"authors": [
"GrahamCampbell",
"SpacePossum",
"kubawerlos"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1015",
"repo": "FriendsOfPHP/PHP-CS-Fixer",
"url": "https://github.com/FriendsOfPHP/PHP-CS-Fixer/pull/5045"
}
|
gharchive/pull-request
|
BacktickToShellExecFixer - add priority relation to NativeFunctionInvocationFixer and SingleQuoteFixer
This is actually a bug fix, because at 2.16 NativeFunctionInvocationFixer has priority 10, which is greater that BacktickToShellExecFixer priority (2).
So, for the person who will be merging this up to 2.16: on the conflict in NativeFunctionInvocationFixer for priorities 10 and 1 - choose 1 or tests would fail.
This PR doesn't take into account the new level of this fixer on the master branch. Can this level be adjusted so that its priority is correct wrt the GlobalNamespaceImportFixer fixer?
So you need BacktickToShellExecFixer: 12, SingleQuoteFixer: 11, NativeFunctionInvocationFixer: 10.
Why? What's wrong with what I wrote in the first comment?
Because if you apply your changes to the 2.17.x version (master), the integration tests fail.
The priorities I provided work on all branches upwards and including 2.15.x.
Because if you apply your changes to the 2.17.x version (master), the integration tests fail.
Which test exactly?
The priorities I provided work on all branches upwards and including 2.15.x.
I don't think so. You want SingleQuoteFixer to have priority 11 and it must run after EscapeImplicitBackslashesFixer, EscapeImplicitBackslashesFixer has priority 1 and last time I've checked 11 is greater than 1 ;)
Well, with those values I have set, all the integration tests pass, including the new ones in this PR. If there is still an issue, it is missing a test?
Well, with those values I have set, all the integration tests pass
No, they don't, I've explained the reason in the previous comment.
No, they don't, I've explained the reason in the previous comment.
Well, they do, because I ran them?
Well, they do
Well, they don't, see: https://github.com/FriendsOfPHP/PHP-CS-Fixer/pull/5071
Hmm, well more specifically, the tests pass after I applied the following PRs, plus the changes I described:
Oh, sorry! Leave the single quote fixer at priority 0.
This is the src diff I applied (plus the tests changes from this PR):
diff --git a/src/Fixer/Alias/BacktickToShellExecFixer.php b/src/Fixer/Alias/BacktickToShellExecFixer.php
index 09d63df50e834bad0e8bfa418bdb6e92e6079eca..a01b7c99d11ec2b82393ef3daa8421f14769056c 100644
--- a/src/Fixer/Alias/BacktickToShellExecFixer.php
+++ b/src/Fixer/Alias/BacktickToShellExecFixer.php
@@ -56,11 +56,11 @@ public function getDefinition()
/**
* {@inheritdoc}
*
- * Must run before EscapeImplicitBackslashesFixer, ExplicitStringVariableFixer.
+ * Must run before EscapeImplicitBackslashesFixer, ExplicitStringVariableFixer, NativeFunctionInvocationFixer, SingleQuoteFixer.
*/
public function getPriority()
{
- return 2;
+ return 12;
}
/**
diff --git a/src/Fixer/FunctionNotation/NativeFunctionInvocationFixer.php b/src/Fixer/FunctionNotation/NativeFunctionInvocationFixer.php
index e118e066205d98a266aa2ce937dac678507f87e9..7ad76a832de3b3198c9ad60a6557aeec00781b4f 100644
--- a/src/Fixer/FunctionNotation/NativeFunctionInvocationFixer.php
+++ b/src/Fixer/FunctionNotation/NativeFunctionInvocationFixer.php
@@ -164,7 +164,7 @@ function baz($options)
* {@inheritdoc}
*
* Must run before GlobalNamespaceImportFixer.
- * Must run after StrictParamFixer.
+ * Must run after BacktickToShellExecFixer, StrictParamFixer.
*/
public function getPriority()
{
diff --git a/src/Fixer/StringNotation/SingleQuoteFixer.php b/src/Fixer/StringNotation/SingleQuoteFixer.php
index 46055b478fe3c6d7069e8e6cbe572abacd596b0f..b9c987e04e1e791efbde6ef2785f8da185cc6345 100644
--- a/src/Fixer/StringNotation/SingleQuoteFixer.php
+++ b/src/Fixer/StringNotation/SingleQuoteFixer.php
@@ -55,7 +55,7 @@ public function getDefinition()
/**
* {@inheritdoc}
*
- * Must run after EscapeImplicitBackslashesFixer.
+ * Must run after BacktickToShellExecFixer, EscapeImplicitBackslashesFixer.
*/
public function getPriority()
{
Ok, then the tests are passing.
But still, I don't know why your idea suppose to be better than changes in this PR - both modify priorities of 2 fixers.
Also, I don't like that NativeConstantInvocationFixer got priority 10 here: https://github.com/FriendsOfPHP/PHP-CS-Fixer/pull/4355 - it should get 1 there as the only fixer it must run before has 0.
here: #4355 - it should get 1 there as the only fixer it must run before has 0.
did you test this, if so this PR is RTM I think?
did you test this, if so this PR is RTM I think?
Yes, it works with the updated priorities I'd recommended.
did you test this, if so this PR is RTM I think?
Yep, works if you follow the conflict resolution (choose 2 over 11 for StrictParamFixer, which should be obvious knowing the first conflict solution).
It works with the updated priorities I'd recommended, but not the ones in this PR.
@GrahamCampbell how did you test it this time, so it's not working? It's "funny" that your changes works for you when in fact they don't and mine doesn't when in fact they do...
https://travis-ci.org/github/GrahamCampbell/PHP-CS-Fixer/jobs/712728528 shows what can happen if someone merges upwards but keeps the priorities of the master branch version. If you use the priorities that are already there, then there can be no opportunity for error, because merging your PR upwards would then be as simple as: https://github.com/GrahamCampbell/PHP-CS-Fixer/compare/ae91c78d3465ffd14f8bac52da6d88f5db37d5be...9054cece9dcec5e5935a1ae04e0fc0d6e3f2a15b.diff.
what can happen if someone merges upwards but keeps the priorities of the master branch version
That's why I put in the description how to solve conflicts...
Thank you @kubawerlos.
|
2025-04-01T06:37:00.483184
| 2015-04-22T12:48:11
|
70115810
|
{
"authors": [
"JEDIBC",
"jubianchi"
],
"license": "bsd-2-clause",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1016",
"repo": "FriendsOfPHP/pickleweb",
"url": "https://github.com/FriendsOfPHP/pickleweb/issues/22"
}
|
gharchive/issue
|
CS & tests not working
Since @jubianchi PR, composer cs and composer test don't work at all :cry:
@JEDIBC I just tested it and it works:
composer cs . => OK
composer test => OK
composer run-script cs . => OK
composer run-script test => OK
I've deleted vendor, git pull & composer install. then I got :
➜ pickleweb git:(master) composer cs
[LogicException]
You must call one of in() or append() methods before iterating over a Finder.
fix [--config="..."] [--config-file[="..."]] [--dry-run] [--level="..."] [--fixers="..."] [--diff] [--format="..."] [path]
Script vendor/bin/php-cs-fixer --ansi fix --config=sf23 handling the cs event returned with an error
[RuntimeException]
Error Output:
[LogicException]
You must call one of in() or append() methods before iterating over a Finder.
fix [--config="..."] [--config-file[="..."]] [--dry-run] [--level="..."] [--fixers="..."] [--diff] [--format="..."] [path]
cs [--dev] [--no-dev] [args1] ... [argsN]
➜ pickleweb git:(master) composer test
Script vendor/bin/atoum -ft handling the test event returned with an error
[RuntimeException]
Error Output:
test [--dev] [--no-dev] [args1] ... [argsN]
Ok, I haven't seen the . in the command composer cs .
But I still got the problem with atoum
In syslog :
Apr 23 10:47:32 chalopin-ubuntu php: PHP Fatal error: Class 'PickleWeb\Tests\Atoum\Report\Fields\Runner\Pickle' not found in /mnt/jedibc/www/pickleweb/.atoum.php on line 9
Apr 23 10:47:32 chalopin-ubuntu php: PHP Stack trace:
Apr 23 10:47:32 chalopin-ubuntu php: PHP 1. mageekguy\atoum\scripts\runner::mageekguy\atoum\scripts\{closure}() /mnt/jedibc/www/pickleweb/vendor/atoum/atoum/classes/scripts/runner.php:0
Apr 23 10:47:32 chalopin-ubuntu php: PHP 2. mageekguy\atoum\scripts\runner->run() /mnt/jedibc/www/pickleweb/vendor/atoum/atoum/classes/scripts/runner.php:589
Apr 23 10:47:32 chalopin-ubuntu php: PHP 3. mageekguy\atoum\script\configurable->run() /mnt/jedibc/www/pickleweb/vendor/atoum/atoum/classes/scripts/runner.php:203
Apr 23 10:47:32 chalopin-ubuntu php: PHP 4. mageekguy\atoum\script\configurable->useDefaultConfigFiles() /mnt/jedibc/www/pickleweb/vendor/atoum/atoum/classes/script/configurable.php:87
Apr 23 10:47:32 chalopin-ubuntu php: PHP 5. mageekguy\atoum\scripts\runner->useConfigFile() /mnt/jedibc/www/pickleweb/vendor/atoum/atoum/classes/script/configurable.php:60
Apr 23 10:47:32 chalopin-ubuntu php: PHP 6. mageekguy\atoum\script\configurable->includeConfigFile() /mnt/jedibc/www/pickleweb/vendor/atoum/atoum/classes/scripts/runner.php:228
Apr 23 10:47:32 chalopin-ubuntu php: PHP 7. mageekguy\atoum\includer->includePath() /mnt/jedibc/www/pickleweb/vendor/atoum/atoum/classes/script/configurable.php:178
Apr 23 10:47:32 chalopin-ubuntu php: PHP 8. mageekguy\atoum\scripts\runner->mageekguy\atoum\scripts\{closure}() /mnt/jedibc/www/pickleweb/vendor/atoum/atoum/classes/includer.php:56
Apr 23 10:47:32 chalopin-ubuntu php: PHP 9. include_once() /mnt/jedibc/www/pickleweb/vendor/atoum/atoum/classes/scripts/runner.php:228
It was a case sensitive problem
|
2025-04-01T06:37:00.495589
| 2023-08-12T09:46:55
|
1847893900
|
{
"authors": [
"marcmanusch",
"tinect"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1017",
"repo": "FriendsOfShopware/FroshPlatformThumbnailProcessor",
"url": "https://github.com/FriendsOfShopware/FroshPlatformThumbnailProcessor/issues/104"
}
|
gharchive/issue
|
Cannot install
PHP Version
8.2
Shopware Version
<IP_ADDRESS>
Actual behaviour
Expected behaviour
Installation is successful
Steps to Reproduce?
Install
Hi,
I tried installing version 4.0.0 at shopware <IP_ADDRESS> without problem.
Could you please verify the following things:
you don't have the plugin on multiple locations (custom/plugins, custom/static-plugins, vendor without link)
you don't have a manipulated \Shopware\Core\Content\Media\Thumbnail\ThumbnailService
Hi, yes I am sure, that I only have the directory custom/plugins and have no manipulation on this service. I tried to install from store and zip file.
how can I get more information?
I see you might have FroshTools installed?
please check the tab "Shopware files" for possible manipulated files.
Sure...
Is there any way you can give me SSH access to the server?
You are welcome to contact me at http://slack.shopware.com/
Sorry thats not possible. I noticed a strange thing. When I try to install it over SSH, it tells me its already installed, but its not.
Let me check the Database...
Hmmm...
There is a similar issue for an other project, in this case the composer.json makes the problem.
https://github.com/phpstan/phpstan/issues/5263
https://github.com/phpstan/phpstan-src/commit/0f8ead706ba4b514e4a511c743645c065f52f195
FROM: "nikic/php-parser": "4.11.0",
TO: "nikic/php-parser": "dev-master#c758510a37218d631fd10f67bca5bccbfef864fa as 4.11.0",
Solved.
|
2025-04-01T06:37:00.526063
| 2024-05-27T08:59:39
|
2318631389
|
{
"authors": [
"Carnewal",
"ZekiJohn",
"bakhirev"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1018",
"repo": "Frikallo/stargazerz",
"url": "https://github.com/Frikallo/stargazerz/issues/1"
}
|
gharchive/issue
|
ValueError: range() arg 3 must not be zero
I just run the example provided. didn't change anything but got this error
This happens because of the floor division to calculate the chunk_size returning 0 which is not valid
chunk_size = len(self.stargazers) // self.threads
To work around this you need to set threads to a value that would produce something > 1, for example
crawler = stargazerz.Crawler(threads=7, target="Frikallo/stargazerz")
It work for me too. May be add "threads=7" like default value?
|
2025-04-01T06:37:00.536886
| 2024-02-06T22:36:40
|
2121782482
|
{
"authors": [
"Joy4634"
],
"license": "BSD-3-Clause-Clear",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1019",
"repo": "Frogbots4634/FtcRobotController",
"url": "https://github.com/Frogbots4634/FtcRobotController/pull/13"
}
|
gharchive/pull-request
|
This has a great park auto
Before issuing a pull request, please see the contributing page.
my code is awesome
|
2025-04-01T06:37:00.565144
| 2014-12-20T23:11:50
|
52579694
|
{
"authors": [
"Derrick91",
"anothergituser"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1020",
"repo": "FrozenNode/Laravel-Administrator",
"url": "https://github.com/FrozenNode/Laravel-Administrator/issues/783"
}
|
gharchive/issue
|
password match
Hi,
I use these rules in my User model when i create an user
public static $rules = [
'name' => 'required',
'email' => 'required|email|unique:users',
'password' => 'required',
'password_repeat' => 'required|same:password',
'language' => 'required'
];
In my administrator user model i have
[ 'password' => [
'title' => trans('backend.password'),
'type' => 'password',
],
'password_repeat' => [
'title' => trans('backend.password'),
'type' => 'password',
],
but when creating it says they do not match.
Using just password without password_repeat works as expected.
Question is, can we use administrator to create fields that need to match other fields?
May i know how did you solved this?
|
2025-04-01T06:37:00.573987
| 2016-05-11T13:30:56
|
154244163
|
{
"authors": [
"fubar-coder",
"zevsst"
],
"license": "BSD-2-Clause",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1021",
"repo": "FubarDevelopment/restsharp.portable",
"url": "https://github.com/FubarDevelopment/restsharp.portable/issues/66"
}
|
gharchive/issue
|
Restsharp portable parallel requests are sent successively by one instance of RestClient
Hello, Mark.
I have some issue with parallelism of my rest calls. Please, have a look at:
stackoverflow
Thanks.
HttpClient didn't support parallel requests, which is the only reason AsyncLock exists. Maybe they've changed its implementation, but I haven't tested it yet.
I personally don't like the AsyncLock stuff, and I'll change the code soon to allow parallel requests when using HttpWebRequest instead of HttpClient.
A possible solution around this limitation would be to use pooling for IRestClient instances.
|
2025-04-01T06:37:00.579293
| 2024-05-30T19:56:15
|
2326438724
|
{
"authors": [
"DefiCake"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1022",
"repo": "FuelLabs/fuel-bridge",
"url": "https://github.com/FuelLabs/fuel-bridge/pull/197"
}
|
gharchive/pull-request
|
docs: update ARCHITECTURE.md
Closes #167
Added additional context around the raised points
+1 from me, the diagrams makes sense and are clear.
Some amendments to consider for future iterations:
Some terms could use further explanation to help users with less context on the use-case of the bridge. For example, "sequencers" and "epochs" could be briefly defined.
The think this sentence could be elaborated on for more context: The system features sending messages in both directions (L1 to L2, and L2 to L1), though the mechanisms involved for each direction are different and almost independent.
It would be nice to add some bullet points in the Message Passing from L1 to L2 portion, to be able to understand the flow a little better, eg:
-- An entity on L1 calls sendMessage on the Message Portal. -- The sendMessage function emits a MessageSent event. -- Fuel's sequencers detect the event and include the message in the next L2 block as a UTXO. -- The L2 block header's inbox reflects the message.
LGTM, just some docs nits
I 'll refreain from using apostrophes on written text from here on. I think I have done enough harm already.
|
2025-04-01T06:37:00.590379
| 2022-10-07T15:05:48
|
1401330640
|
{
"authors": [
"Voxelot",
"adlerjohn",
"mohammadfawaz"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1023",
"repo": "FuelLabs/fuel-specs",
"url": "https://github.com/FuelLabs/fuel-specs/issues/413"
}
|
gharchive/issue
|
Should predicates allow backward jumps if a gas limit is imposed?
Predicates do not currently allow backward jumps. This is a protocol restriction:
During predicate mode, hitting any of the following instructions causes predicate verification to halt, returning Boolean false:
Any contract instruction.
JI, JNEI, JMP or JNE with jump-to value less than or equal to $pc (these would allow loops). In other words, $pc must be strictly increasing.
The downside of the above is that function calls cannot be properly supported because they require jumps to and from the functions being called, which evidently leads to having backward jumps at some point. For now, the compiler will inline all function calls in predicates but that can lead to bytecode size bloat.
@adlerjohn mentioned that it may be possible to remove the restriction above if we impose a gas limit that guarantees that the predicate eventually halts execution.
If predicates have a guaranteed upper bound on cost (e.g. by using gas metering), then sure. One thing we have to be careful of is that changes to the gas schedule will require re-verifying predicates in the mempool, along with the possibility that some predicates may become unsatisfiable.
@mitch-fuel assigning this to you as it's related to #440
Closing as this issue is resolved and backward jumps are allowed.
|
2025-04-01T06:37:00.593446
| 2024-06-17T21:20:13
|
2358280029
|
{
"authors": [
"Br1ght0ne",
"Rcontre360",
"calldelegation"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1024",
"repo": "FuelLabs/fuels-rs",
"url": "https://github.com/FuelLabs/fuels-rs/issues/1437"
}
|
gharchive/issue
|
Revert OutOfGas error when using the produce_blocks function in the provider to go forward in time
The bug happens when trying to go forward in time using the provider.produce_blocks function
After doing that, the calls performed to the same contract fail with a Revert OutOfGas error
This issue was found by the protofire team
Repository to reproduce the bug
https://github.com/0xLucca/fuels-bug
Hi! is there any progress with this?
Reproduced on both fuels 0.62 and fuels 0.64, looking into it.
@calldelegation There's a way to fix the problem in Sway, without the TxPolicieschange, by moving the state lookups from inside the require call to separate calls, like this:
let start_time = storage.start_time.read();
let end_time = storage.end_time.read();
require(asset_amount > 0, "Asset amount must be greater than 0");
require(current_time >= start_time && current_time < end_time, "Action can only be executed within the time window")
|
2025-04-01T06:37:00.597964
| 2022-01-19T20:41:30
|
1108535768
|
{
"authors": [
"digorithm",
"vnepveu"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1025",
"repo": "FuelLabs/fuels-rs",
"url": "https://github.com/FuelLabs/fuels-rs/pull/48"
}
|
gharchive/pull-request
|
feat: return receipts in data returned by contract calls
Introduce the CallResult struct, that holds the call return in its
value field, and the receipts in the receipts field.
This closes issue #35
Note: failing tests are passing when I run them individually, for some weird reason (probably config issue?)
Note: Sometimes some tests are failing but passing when I run them individually, for some weird reason (probably config issue?).
The tests were designed to be self-contained enough so that state doesn't leak between tests. E.g. we spawn a short-lived fuel-core per test case in harness.rs.
Please provide repro steps if you found this not to be the case. Also, check it against master to see if this was introduced on this branch.
So far I haven't experienced any flakiness in fuels-rs tests, and if that's the case now we have to do the best we can to avoid flakiness in tests.
Tell me if you think I should add more tests!
It's true that it doesn't happen on master… But I have little to no idea how to investigate, since the behavior seems erratic at best, and seems really linked to a performance issue (some tests don't complete in less than 60 secs and are killed).
CI passing as well! :tada:
Changing docs strings broke CI? wth
Haha, that's actually what I'm currently investigating. It's failing to compile Sway code when we call forc::build during a test. So weird.
|
2025-04-01T06:37:00.599387
| 2023-01-17T17:56:49
|
1536802957
|
{
"authors": [
"digorithm",
"iqdecay"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1026",
"repo": "FuelLabs/fuels-rs",
"url": "https://github.com/FuelLabs/fuels-rs/pull/784"
}
|
gharchive/pull-request
|
style: use C-style getters naming for SDK methods
This PR closes #783 by changing the naming of contract instance getter methods.
It also renames two functions whose name wasn't really adapted.
It also adds a comment about a .gitignore file.
Let's be extra careful and considerate when making big changes like this one to public-facing APIs. Although this is obviously a good change, it warrants a discussion first!
|
2025-04-01T06:37:00.602415
| 2022-12-21T16:05:36
|
1506550584
|
{
"authors": [
"arboleya"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1027",
"repo": "FuelLabs/fuels-ts",
"url": "https://github.com/FuelLabs/fuels-ts/pull/703"
}
|
gharchive/pull-request
|
chore: inverting symlink origin and destination
The <rootDir>/CONTRIBUTING.md is the actual Markdown file.
The <rootDir>/docs/CONTRIBUTING.md is the symlink, pointing to the former.
This should guarantee that:
Users are presented with the original rendered Markdown when clicking on it (on <rootDir>/)
The documentation can still follow the symlink at build time (from <rootDir>/docs/)
Close #693
This is a re-fix on top of #694
@camsjams Thank you. I've reopened the issue.
|
2025-04-01T06:37:00.604619
| 2022-08-26T04:17:18
|
1351723233
|
{
"authors": [
"pedronauck"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1028",
"repo": "FuelLabs/fuels-wallet",
"url": "https://github.com/FuelLabs/fuels-wallet/pull/66"
}
|
gharchive/pull-request
|
feat(app): add AssetItem and AssetList components
✨ Fix #8
✨ Fix #9
@luizstacio I ended up choosing react-content-loader because is possible to just copy SVG from Figma and paste it into their website to generate the right loader, that the result:
https://fuels-wallet-bdzlegv2g-fuel-labs.vercel.app/storybook/?path=/story/asset-assetlist--loading
And that's the code
|
2025-04-01T06:37:00.613444
| 2023-04-27T21:51:21
|
1687598731
|
{
"authors": [
"Braqzen",
"nfurfaro"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1029",
"repo": "FuelLabs/sway-libs",
"url": "https://github.com/FuelLabs/sway-libs/pull/147"
}
|
gharchive/pull-request
|
Add test for cross-contract reentrancy
Type of change
Improvement (refactoring, restructuring repository, cleaning tech debt, ...)
Changes
The following changes have been made:
added a new test case
added supporting new contract & abi
added new functions on existing contracts & abis
Notes
The test case added performs the following steps:
The Attack contract's launch_thwarted_attack_3() function is triggered by a contract call from the SDK.
The Attack contract calls into the Target contract's cross_contract_reentrancy_denied() function.
The Target contract calls back to the Attack contract's evil_callback_4() function.
The Attack contract calls into the Attack Helper contract's attempt_cross_contract_reentrancy() function.
(attack vector) The Attack Helper contract attempts to re-enter by calling into the Target contract's cross_contract_reentrancy_denied() function(which hasn't finished execution of the initial call from the Attack contract).
The attack is thwarted by the reentrancy_guard() function in place.
Seems fine but is there something preventing a 2 contract interaction over 3 contracts?
@Braqzen I'm not quite sure what you're getting at here.
If the reentrancy_guard() function is not used, given 2 Contracts A & B:
A can call into B can call into A can call into B can call into A can call into B ... until the gas runs out.
The compiler just warns us if the CEI (Checks, Effects, Interactions) pattern is violated. It doesn't strictly prevent reentrant function calls.
Seems fine but is there something preventing a 2 contract interaction over 3 contracts?
@Braqzen I'm not quite sure what you're getting at here. If the reentrancy_guard() function is not used, given 2 Contracts A & B:
A can call into B can call into A can call into B can call into A can call into B ... until the gas runs out.
The compiler just warns us if the CEI (Checks, Effects, Interactions) pattern is violated. It doesn't strictly prevent reentrant function calls.
By "non-reentrant func" I meant to use the guard in that function, I should have been clearer, my mistake.
|
2025-04-01T06:37:00.615685
| 2023-04-10T00:58:46
|
1660069294
|
{
"authors": [
"chlenc",
"mohammadfawaz"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1030",
"repo": "FuelLabs/sway",
"url": "https://github.com/FuelLabs/sway/issues/4415"
}
|
gharchive/issue
|
Internal compiler error: Unable to determine ID for log instance. during building of predicate
Hi guys build this predicate
predicate;
use std::{
inputs::input_owner,
logging::log,
};
const RECEIVER_CONFIG: b256 = 0x5d99ee966b42cd8fc7bdd1364b389153a9e78b42b7d4a691470674e817888d4e;
fn main() -> bool {
const RECEIVER = Address::from(RECEIVER_CONFIG);
log::<Address>(input_owner(0).unwrap());
log::<Address>(input_owner(1).unwrap());
true
}
Gives me this error
^^/p/f/l/s/p/o/vGx8NIeAtmcrErOEa3Z4rIORovU1i0 >>> forc build (1) (*master+78) 01:54:15
WARNING! unused manifest key: project.target
WARNING! unused manifest key: project.target
WARNING! unused manifest key: project.target
WARNING! unused manifest key: project.target
WARNING! unused manifest key: project.target
WARNING! unused manifest key: project.target
WARNING! unused manifest key: constants
WARNING! unused manifest key: constants
Compiling library core (/Users/alexey/.forc/git/checkouts/std-9be0d6062747ea7/607ac50176db8bef936f91bacf435d0ea37d041e/sway-lib-core)
Compiling library std (git+https://github.com/fuellabs/sway?tag=v0.37.0#607ac50176db8bef936f91bacf435d0ea37d041e)
Compiling predicate swap-predicate (/Users/alexey/projects/fuel/limit_orders/services/predicate-orders-builder/orders/vGx8NIeAtmcrErOEa3Z4rIORovU1i0)
error
--> /Users/alexey/.forc/git/checkouts/std-9be0d6062747ea7/607ac50176db8bef936f91bacf435d0ea37d041e/sway-lib-std/src/logging.sw:8:5
|
6 |
7 | pub fn log<T>(value: T) {
8 | __log::<T>(value);
| ^^^^^^^^^^^^^^^^^ Internal compiler error: Unable to determine ID for log instance.
Please file an issue on the repository and include the code that triggered this error.
9 | }
|
____
Aborting due to 1 error.
Error: Failed to compile swap-predicate
Versions of all tools latest
@chlenc thanks for flagging this. Logs are not allowed in a predicates. That being said, a better error message should be emitted in this case.
|
2025-04-01T06:37:00.621057
| 2022-06-23T14:33:59
|
1282507413
|
{
"authors": [
"tritao",
"vaivaswatha"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1031",
"repo": "FuelLabs/sway",
"url": "https://github.com/FuelLabs/sway/pull/2096"
}
|
gharchive/pull-request
|
Fix checking for too many function call arguments.
This PR adds explicit checking for the case of too many function call arguments in function application instantiation.
This needs to be done before type checking because after type checking we drop the extra function call arguments due to having no type to check against.
During this PR development, it was found that Sway was allowing associated functions to be called with method call syntax,
which should not be allowed (Rust also throws an error in this case), and previously was just working by chance and a side
effect of the zip calls dropping the extra arguments.
So this adds a new error and check for this and updates some tests that were relying on this non-Rust-like behaviour.
There was another issue found, where the IR (at least the IR parser) currently cannot deal with empty structs, which was triggered by an existing IR test after being updated to work with the above changes.
Also re-enables the should_fail/excess_fn_arguments test.
Closes https://github.com/FuelLabs/sway/issues/2081.
This needs to be done before type checking because after type checking we drop the extra function call arguments due to having no type to check against.
Does this mean that, with your changes, instantiate_function_application_inner can never see TooManyArgumentsForFunction? That means we could also check for TooFewArgumentsForFunction in the outer functions and remove the checks entirely from instantiate_function_application_inner ?
Updated the PR with fixes to the IR parsing issue instead of changing the test.
Meanwhile I've also found a couple more issues with associated methods while reworking the checks I had earlier on this PR.
The new checking code is more robust and found a couple more issues related to traits, as such I've decided to just split it from this one, and will be sending a different PR later to deal with that with more tests once I handle those extra edge cases.
|
2025-04-01T06:37:00.630042
| 2024-07-18T03:50:19
|
2415157319
|
{
"authors": [
"aluminumbox",
"dataangel"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1032",
"repo": "FunAudioLLM/CosyVoice",
"url": "https://github.com/FunAudioLLM/CosyVoice/issues/168"
}
|
gharchive/issue
|
如何训练方言
需要多大的显存还有数据结构是啥样子的
prepare some dialect data, and use a dialect label at sentence start. for example <|chuan|> for 四川话. check whisper special token to see if it has already covered some dialect label
|
2025-04-01T06:37:00.632200
| 2021-03-01T11:18:39
|
818764346
|
{
"authors": [
"arlapin",
"kathyxchen"
],
"license": "BSD-3-Clause-Clear",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1033",
"repo": "FunctionLab/selene",
"url": "https://github.com/FunctionLab/selene/pull/162"
}
|
gharchive/pull-request
|
Fix a typo and documentation
What does this implement/fix? Explain your changes.
The dictionary name is wrong.
The returned vector is always of size n_features, so the docstring is misleading.
What testing did you do to verify the changes in this PR?
Added unit tests
PTAL @kathyxchen
LGTM! Thanks for this bugfix and for adding more tests! woohoo
|
2025-04-01T06:37:00.639796
| 2021-05-24T09:15:42
|
899474594
|
{
"authors": [
"hijamoya"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1034",
"repo": "FunnyDevs/hilt-conductor",
"url": "https://github.com/FunnyDevs/hilt-conductor/issues/1"
}
|
gharchive/issue
|
Injected field is null
Hi, I used your library and everything is ok, but when I access the injected field, it throws a npe. I dont know what step I missed, can you give some suggestions?
Thanks!
I found the problem, the injection happens in onCreateView, if the parent class implement the method, the child class can not inject the field correctly.
|
2025-04-01T06:37:00.708876
| 2016-08-04T14:16:29
|
169385179
|
{
"authors": [
"FutoRicky",
"oskarrough"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1038",
"repo": "FutoRicky/ember-cli-dropzonejs",
"url": "https://github.com/FutoRicky/ember-cli-dropzonejs/pull/38"
}
|
gharchive/pull-request
|
Add support for Ember FastBoot
As per http://www.ember-fastboot.com/docs/addon-author-guide
Thanks for this!
Please make the PR to the dev branch.
|
2025-04-01T06:37:00.725404
| 2024-04-07T03:14:16
|
2229537306
|
{
"authors": [
"cbebe"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1039",
"repo": "Fysh-Fyve/vscode-fysh",
"url": "https://github.com/Fysh-Fyve/vscode-fysh/issues/4"
}
|
gharchive/issue
|
Bundle JS code
https://code.visualstudio.com/api/working-with-extensions/bundling-extension
Solved in 1.0.7
|
2025-04-01T06:37:00.776887
| 2020-08-26T17:25:36
|
686491494
|
{
"authors": [
"dbauszus-glx"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1040",
"repo": "GEOLYTIX/xyz",
"url": "https://github.com/GEOLYTIX/xyz/issues/339"
}
|
gharchive/issue
|
Layer Style panel & themes
With a style.theme defined in the configuration that theme will be applied when the layer is loaded. Otherwise the first theme from the style.themes object will be assigned as style.theme.
themes: {} needed to be removed from the defaults to make the themes check falsy if no theme is defined in the layer's style configuration.
Switch all control to be added with corresponding themes not always and then toggled when a theme is applied.
Flatten the layers.view.style directory in lib.
In order to see the 'Basic' theme legend which allows to modify the default style configuration for the layer a theme type: 'basic' needs to be defined.
polyCategorized.mjs and clusterCategorized.mjs have now been merged to categorized.mjs.
The legend should not depend on whether the layer is cluster or mvt format but only whether styles are defined as symbol, line, or polygon.
polyGraduated.mjs and clusterGraduated.mjs have now been merged to graduated.mjs.
The legend should not depend on whether the layer is cluster or mvt format but only whether styles are defined as symbol, line, or polygon.
hyperHTML construction has now been superseeded with microHTML.
Grid element items are now longer in intermediate container but aligned and justified in a grid proper.
The grid theme is now type: grid
The 'basic' type theme can now be configured.
With only the theme type in the configuration this will result in a panel for Stroke (colour, width) and Fill (colour, width).
Individual fields can be defined in a fields array.
Field types can be text or range (for now).
The field name must be provided.
The style can be set to highlight, default is default.
"Basic": {
"type": "basic",
"fields": [
{
"title": "Color",
"style": "highlight",
"type": "text",
"field": "strokeColor"
}
]
}
|
2025-04-01T06:37:00.795238
| 2022-03-24T15:05:21
|
1179647436
|
{
"authors": [
"mathomp4"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1041",
"repo": "GEOS-ESM/GOCART",
"url": "https://github.com/GEOS-ESM/GOCART/pull/131"
}
|
gharchive/pull-request
|
Fixes #130. Initialize allocatable variables in Process Library
Closes #130
TL;DR: This PR fixes the "changing klid halts GOCART" issue seen by @wmputman.
My testing shows that it is zero-diff if klid is set to 0.01 (aka k=1), and lets the model run at other klid values.
Runs with:
[x] Intel Release
[x] Intel Debug
[x] GNU Release
[x] GNU Debug
The solution was found by running with Intel in Debug mode where it pointed to line 1412:
https://github.com/GEOS-ESM/GOCART/blob/4d4f27c51f49ccebb2df5e167582d033eea4d0d4/Process_Library/GOCART2G_Process.F90#L1400-L1414
The issue is that vsettle is filled by Chem_CalcVsettle but only for vsettle(:,:,klid:km). But then this line:
vsettle = max(1.0e-9, vsettle - v_upwardMaring)
operates on all of vsettle. Since vsettle was only allocated and never initialized to anything, boom.
To be safe, I'm initializing all local allocatable arrays that I can see to zero. (I'm not a scientist, so someone like @pcolarco or @amdasilva might want to make sure that's a "safe" or "valid" initialization value to use.)
Also, in one subroutine, I changed the name of the array dp to d_p. The issue is that there was an array:
real(kind=DP), dimension(:,:), allocatable :: dp
So dp was kind=DP. The compiler sort of went a bit nuts when I tried source=0.0_DP in that routine. So rather than change the kind, we just change dp to d_p as was done in another routine.
Releasing from draft. It seems to work with both Intel and Debug
@adarmenov Okay. Let me change that code. Give me a second...
|
2025-04-01T06:37:00.797874
| 2021-05-26T19:09:03
|
902817140
|
{
"authors": [
"mathomp4",
"tclune"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1042",
"repo": "GEOS-ESM/mepo",
"url": "https://github.com/GEOS-ESM/mepo/pull/172"
}
|
gharchive/pull-request
|
Add allrepos option to clone
Based on a workflow suggested by @rtodling, this PR adds a new --allrepos option to mepo clone. With this flag, not only is the tag/branch passed to the --branch/-b option to mepo clone checked out, but it also checks out that tag in all sub repositories!
To wit, if you do:
mepo clone -b <tag> --allrepos fixture-repo
what this will do is:
git clone -b fixture-repo
cd fixture-repo
mepo clone
mepo checkout <tag> [on all subrepos]
Note this means it will fail if the tag requested is not on all repos!
Does this mean it will ignore the actual tags that the top repo has stored for the sub repos? What is the example use case?
Never mind- I had misunderstood.
|
2025-04-01T06:37:00.848478
| 2018-09-13T19:31:54
|
360035088
|
{
"authors": [
"phlptp"
],
"license": "BSD-3-Clause",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1043",
"repo": "GMLC-TDC/HELICS-src",
"url": "https://github.com/GMLC-TDC/HELICS-src/issues/462"
}
|
gharchive/issue
|
Function naming for C and language API's
there are a set of functions in the C shared library named like
helicsPublicationPublishXXX(pubid, value) or similar
I am wondering if it would be inconsistent to drop the Publication so they are
helicsPublishXXX(... ) instead.
likewise
helicsInputGetXXX or helicsInputSetDefualt(...)
to helicsGetXXX or helicsSetDefault(...)
Basically dropping the Publication or Input for some functions that will be commonly used and have some redundancy in the name.
My reluctance is that it isn't totally consistent with the function names being helics but on the other hand it is simpler and I don't think changes the information in the name, and might be a little more obvious for the transition to 2.0 as many functions will go from helicsSubscriptionGetXXX to helicsInputGetXXX which might be non obvious if you generated with the subscription convenience function.
Thoughts?
It sounded like the consensus was to leave the functions names as is to maintain consistency in the naming conventions.
|
2025-04-01T06:37:00.948853
| 2019-06-18T10:12:35
|
457382323
|
{
"authors": [
"awav",
"st--"
],
"license": "apache-2.0",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1044",
"repo": "GPflow/GPflow",
"url": "https://github.com/GPflow/GPflow/pull/990"
}
|
gharchive/pull-request
|
Make ScipyOptimizer reusable (without growing tf graph)
Currently, the ScipyOptimizer only exposes a minimize() method that adds ops to the default graph, thereby slowing down performance when run repeatedly (e.g. in online learning or backtesting). This is an updated version of awav/reuse-scipyoptimizer.
I know this is a bit hacky, but it's something that I actually need and it'd be much easier if I wouldn't have to copy these files manually/work off a feature branch ... and I know other people need this feature as well. Could we please get this merged? @awav if you've got any ideas for how to clean it up before merging please say!
Changes made in original branch
|
2025-04-01T06:37:00.954624
| 2016-01-11T16:42:02
|
125985924
|
{
"authors": [
"GRbit",
"aglyzov"
],
"license": "BSD-3-Clause",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1045",
"repo": "GRbit/go-pcre",
"url": "https://github.com/GRbit/go-pcre/pull/1"
}
|
gharchive/pull-request
|
Fixed tests, better UTF-8 support
Hi, I fixed the tests which did not compile.
And also I added a UCP = C.PCRE_UCP constant (stands for Unicode Character Properties) and modified the ParseFlags() function to treat "(?u)" as UTF8 | UCP. This is needed to make \b \w \s \B \W \S to work properly in Unicode patterns.
Just one example:
TEXT: "мощность 25 Вт"
REG: (?ux) \b \d+ \s* Вт \b
Without UCP in flags the pattern won't match "25 Вт" which is counterintuitive.
Thank you! Tests really needed to be updated, I paid very little attention to them. UCP flag is very important too!
BTW it's my first pull request on github, so that makes me twice as happy =)
|
2025-04-01T06:37:01.012028
| 2020-11-24T00:26:39
|
749225370
|
{
"authors": [
"GSculerlor",
"MLisDreaming"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1046",
"repo": "GSculerlor/Cytosu",
"url": "https://github.com/GSculerlor/Cytosu/issues/13"
}
|
gharchive/issue
|
Crash on Beatmap Load
The game crashes as soon as beatmap finishes loading if the latest compile of Cytosu is used on the latest version of osu!lazer (2020.1121.0).
Thank you for reporting. I'll check it later today
should not be an issue anymore. feel free to reopen if you somehow can reproduce the issue
|
2025-04-01T06:37:01.042036
| 2016-12-05T19:07:23
|
193583012
|
{
"authors": [
"ollwenjones",
"pstockley"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1048",
"repo": "GWTReact/gwt-react",
"url": "https://github.com/GWTReact/gwt-react/issues/2"
}
|
gharchive/issue
|
Native JS React Components
Do you have any examples of rendering 3rd party / non GWT-React class inside a GWT-React construct?
You mention how to bundle them with webpack, so I assume you've done it? From what I can see, since your React.createElement overrides only take class extending ReactClass<P> which is a non-native @JsType - so a native @JsType can't extend it.
Am I missing something in the utils somewhere or should I contribute another overload of that method?
Anything helps. :slightly_smiling_face:
Currently working around this using a utility in my JavaScript bundle that calls React.createElement on the JavaScript side, and just using the return of that on the GWT/Java side. Might be cleaner to handle it that way anyway.
An example you could look at it is MobXDevTools. This is a React component that renders some development tools. To access this I created a JsType interface as follows
@JsType(isNative = true, namespace = JsPackage.GLOBAL) public class MobXDevTools { @JsProperty(name = "default") public static ReactClass<BaseProps> component; }
You then use it as follows:
ReactDOM.render( div(null, //Show Dev tools toolbar React.createElement(MobXDevTools.component, $(new BaseProps(), "hightlightTimeout", 4000)), React.createElement(App.component, $(new AppStateProps(), "appState", appState)) ), Document.get().getElementById("mainCont")); }
You can additionally create a factory method to avoid using React.createElement. The docs go into more details on this.
Closing this. Let me know if you need more examples.
|
2025-04-01T06:37:01.085787
| 2024-10-18T13:25:40
|
2597517531
|
{
"authors": [
"adrn",
"nstarman"
],
"license": "BSD-3-Clause",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1050",
"repo": "GalacticDynamics/unxt",
"url": "https://github.com/GalacticDynamics/unxt/issues/234"
}
|
gharchive/issue
|
Astropy equivalencies not fully supported
We have this doctest example:
x = Quantity([1, 2, 3], "Kelvin")
with u.add_enabled_equivalencies(u.temperature()):
y = x.to(u.deg_C)
which works, but this one doesn't work:
x = Quantity([1, 2, 3], "radian")
with u.add_enabled_equivalencies(u.dimensionless_angles()):
y = x.to(u.one)
I think this is because the physical type (dimension) changes in the latter case and not the first. See https://github.com/GalacticDynamics/unxt/blob/main/src/unxt/_src/quantity/base_parametric.py#L43
Actually I think the main issue is with this line - I don't think we want to do a replace here (that line just replaces the unit and value of a quantity), because conversions with equivalencies can change the physical type (dimensions) of a quantity (such as with dimensionless_angles()).
Right. Though do we expect other arguments? If we don't mind type coercion, one option would be to swap the replace for:
return Quantity(_apy7_unit_to(x.unit, unit, x.value), unit)
We can support all fields by using the dataclassish.field_items()
Preventing unnecessary type coercion is a bit more fiddly. I think we can do this by checking the dimensions of the output unit against the input units and only doing type coercion if they are different.
|
2025-04-01T06:37:01.098927
| 2018-12-27T10:06:09
|
394352133
|
{
"authors": [
"AlexeyRaga",
"TikhonJelvis",
"TomMD"
],
"license": "BSD-3-Clause",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1051",
"repo": "GaloisInc/avro",
"url": "https://github.com/GaloisInc/avro/issues/75"
}
|
gharchive/issue
|
Enable CCI to use orbs
@TomMD Hi Tom,
I'd like to use CircleCI Orb (https://circleci.com/orbs/registry/orb/haskell-works/haskell-build) for building this project.
You can read more about orbs here: https://circleci.com/docs/2.0/using-orbs/
Would it be possible to enable this settings for the Galois organisation in CircleCI?
Hurm, that is a good question @AlexeyRaga . I'll need to ask the admin, which I can do on Tuesday.
Is this urgent? If this is presently a significant issue then it might be the straw that moves the project over to your namespace after all.
Thinking for more than a second on it - we should probably transfer to your namespace. Care to click the button? I don't think I can since I can't create repos in your org. That said, I'm also happy to talk to the admin on Tuesday if you'd rather leave it here.
Why? Well obviously I haven't had work projects pull me back to this code like I had expected. Certainly you are using it heavily and should be able to take necessary actions without an intermediary. If anyone objects then I'm all ears, I just don't see any strong reason to keep it under the GaloisInc organization.
Fair enough! I am OK with transferring it under haskell-works.
Unless there are no major objections (@TikhonJelvis?) I will make a transfer in a couple of days.
Transferring sounds perfectly reasonable to me.
We should update the links on the Hackage page when we do.
Thanks everyone, the transfer has happened, a new release with the updated links has been published to Hackage.
|
2025-04-01T06:37:01.103274
| 2018-03-24T20:09:15
|
308294929
|
{
"authors": [
"foxnoodles",
"mc-gulu"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1052",
"repo": "GameBuildingBlocks/PerfAssist",
"url": "https://github.com/GameBuildingBlocks/PerfAssist/issues/1"
}
|
gharchive/issue
|
Unity 2018.1 is coming out next week
Would be nice to update the project. It's still on 5.x
haven't got 2018 installed yet :)
Lol how about 2017.1 .2 .3 ?))))
There are some absolete api warnings popping out in 2017.x so yeah
would do it later (should be soon)
|
2025-04-01T06:37:01.114876
| 2022-06-25T14:07:17
|
1284613370
|
{
"authors": [
"Zettovec",
"kazumasa200"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1054",
"repo": "GamerJoep/MinetopiaVehicles",
"url": "https://github.com/GamerJoep/MinetopiaVehicles/pull/88"
}
|
gharchive/pull-request
|
Update Japanese translation
Updated some new keys to Japanese translation
Thanks! Do you think it would be possible to wait a tad and translate even more messages, when PR #86 is merged?
understood. I will wait for a while.
Great! I'll let you know
@kazumasa200 the new messages are on lines 66, 67, 86 and 171! ;)
Thanks again!
|
2025-04-01T06:37:01.132785
| 2015-11-24T12:17:58
|
118596631
|
{
"authors": [
"k6Labz",
"sergey-filipenko"
],
"license": "BSD-3-Clause",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1055",
"repo": "Gapminder/vizabi",
"url": "https://github.com/Gapminder/vizabi/issues/1010"
}
|
gharchive/issue
|
v0.8.9_15: Cross symbol ‘x’ is missing
SEVERITY: Major
MODULE: Mountain Chart
Platform: MAC / Safari 9.0
RELEASE: v0.8.9
STEPS TO REPRODUCE:
Navigate URL: http://static.gapminderdev.org/vizabi/release/v0.8.9/preview/mountainchart.html
Mountain chart page will be loaded
Select a country. For example we select China.
Selected country is shown on top left with the population of that country
Take the mouse to the population. A small circle is shown to remove this population
Note that cross symbol ‘x’ is missing there
ACTUAL RESULTS:
Cross symbol ‘x’ is missing on population of the country
EXPECTED RESULTS:
Cross symbol ‘x’ should be there for better understanding to deselect the country
SCREENSHOT:
https://github.com/Gapminder/vizabi/pull/1042
|
2025-04-01T06:37:01.135425
| 2016-09-19T03:49:42
|
177699907
|
{
"authors": [
"greenfire27"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1056",
"repo": "GarageGames/Torque2D",
"url": "https://github.com/GarageGames/Torque2D/issues/369"
}
|
gharchive/issue
|
CompositeSprites Animates Continue When Scene Is Paused
If you pause a scene with normal sprites that are animating, the animation freezes as you'd expect. If you pause the scene with animated composite sprites, then the animations continues. This is easy to test in the toy box. Open the AudioToy and pause the scene.
SandboxScene.setScenePause(true);
The animations will freeze. Now open the CompoundObjectsToy and click around to throw in a few objects. Now try pausing the scene again. If any of the objects are moving, they'll freeze in place, but their animations will continue.
This issue hits my game hard, where I have multiple scenes with lot's of composite sprites. When I hide one scene I pause it so it's not hogging resources, but with this bug, it still is using some juice.
Fixed with pull request #375.
|
2025-04-01T06:37:01.138137
| 2017-04-27T12:54:08
|
224774017
|
{
"authors": [
"Duion",
"dottools"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1057",
"repo": "GarageGames/Torque3D",
"url": "https://github.com/GarageGames/Torque3D/issues/1995"
}
|
gharchive/issue
|
Cannot edit multiple objects
I remember being able to select and edit the values of multiple objects before in Torque, but it seems to be broken now.
To reproduce: Place multiple objects in the default template, like some rocks, select them all and edit some values like scale and you will notice it will only update the item you selected last.
You really should be able to select and edit multiple objects at once, imaging having placed hundreds of objects and in the end you realize you have to switch collision type to visible mesh or something like that on all of them and it would be just insane to select each one separately and change the value and then repeat it a hundred times, when you just could select all and change it and done.
Hm I found that for some things it seems to work, for example if you select multiple audio emitters and change the volume, it will change the volume on all at the same time.
Makes sense you can only do multiple edit for the same object type and possibly only for some fields too.
I was never trying to edit different object types, it was always the same like static shapes etc.
The question now is where the reason is, why it only works for certain types, since it should work with any object type that has the same fields like coordinates or the same flags, since the only task that needs to be done is to write the same values in the same field in the mission file.
The engine does account for that already, for example if you edit multiple objects with different values in a certain field, the field will show as empty, signalizing that not all objects have the same value and when you enter some value in there, it will apply for all objects.
|
2025-04-01T06:37:01.149112
| 2024-07-03T16:31:36
|
2389035640
|
{
"authors": [
"OwenPriceSkelly",
"codecov-commenter"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1058",
"repo": "Garden-AI/garden",
"url": "https://github.com/Garden-AI/garden/pull/495"
}
|
gharchive/pull-request
|
local data toggle
closes #494
Overview
Very small PR to sanity check approach before actually replacing all of the local_data calls. This turns import local_data into an error in the presence of a special environment variable, and adds the basic CRUD methods to the backend client.
Discussion
There should be exactly zero behavioral difference for anyone without the GARDEN_DISABLE_LOCAL_DATA variable set except for one new (currently optional) function_text field on RegisteredEntrypoint for compatibility with the backend response schema.
This is always the same as the function text otherwise stored in .steps[0].function_text, it's only necessary because the backend doesn't have steps.
Testing
Without the variable I can import garden_ai just like before, but with the environment variable set it fails with flying colors.
Documentation
n/a
📚 Documentation preview 📚: https://garden-ai--495.org.readthedocs.build/en/495/
:warning: Please install the to ensure uploads and comments are reliably processed by Codecov.
Codecov Report
Attention: Patch coverage is 54.34783% with 21 lines in your changes missing coverage. Please review.
Project coverage is 75.18%. Comparing base (1d34837) to head (20cb989).
Files
Patch %
Lines
garden_ai/backend_client.py
33.33%
20 Missing :warning:
garden_ai/local_data.py
93.33%
1 Missing :warning:
:exclamation: Your organization needs to install the Codecov GitHub app to enable full functionality.
Additional details and impacted files
@@ Coverage Diff @@
## main #495 +/- ##
==========================================
- Coverage 75.50% 75.18% -0.32%
==========================================
Files 55 55
Lines 3695 3732 +37
==========================================
+ Hits 2790 2806 +16
- Misses 905 926 +21
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.
|
2025-04-01T06:37:01.164702
| 2022-04-21T09:22:29
|
1210755058
|
{
"authors": [
"Garulf",
"mrmetrix4"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1059",
"repo": "Garulf/Flox",
"url": "https://github.com/Garulf/Flox/pull/11"
}
|
gharchive/pull-request
|
Fixing empty jsonRPC.
As answer to my https://github.com/Garulf/Flox/issues/10:
Reverting the commit https://github.com/Garulf/Flox/commit/01bb1daf3de78c3699862ef62b751fed78e3efe1.
Changing __call__ back to __init__, solves the problem.
I was attempting to step away from the "hacky" use of __del__ and have the user explicitly call the Launcher class. But in hind sight I should have held off on this change.
You'd be better off re-adding the __del__ method instead of __init__.
I was attempting to step away from the "hacky" use of __del__ and have the user explicitly call the Launcher class. But in hind sight I should have held off on this change.
You'd be better off re-adding the __del__ method instead of __init__.
It does work, but I'm not sure I know why.
Why don't you use __init__ anyway?
The original reason was to allow plugins to use their own __init__ and still inherit Flox's __init__.
class MyPlugin(Flox):
def __init__(self, *args, **kwargs):
super().__init__()
# this line is never ran if the launcher is run on init
The original reason was to allow plugins to use their own __init__ and still inherit Flox's __init__.
class MyPlugin(Flox):
def __init__(self, *args, **kwargs):
super().__init__()
# this line is never ran if the launcher is run on init
Ofcourse, but that's how python works.
I claim that if someone implements it's own init, it should be done as following:
class MyPlugin(Flox):
def __init__(self, *args, **kwargs):
self.something = "something"
# any other initializers.
super().__init__()
# the super should be last.
def query(self, query)
[...]
And ofcourse it works (with Launcher.init).
BTW. Thanks for responding so fast, very appriciated!
Flox is getting a complete re-write to avoid these issues. I'm working on this now just finalizing the design.
Flox is getting a complete re-write to avoid these issues. I'm working on this now just finalizing the design.
Cool, I'm looking forward for it!
I reverted the change that caused this issue. Sorry for the trouble!
|
2025-04-01T06:37:01.167548
| 2022-09-26T07:52:15
|
1385602882
|
{
"authors": [
"GateteVerde",
"GoldyFredy9",
"mariocraft31"
],
"license": "BSD-3-Clause",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1060",
"repo": "GateteVerde/Mario-Worlds-Delta",
"url": "https://github.com/GateteVerde/Mario-Worlds-Delta/issues/5"
}
|
gharchive/issue
|
Mario stays in free fall state for all levels
Regardless if i obtain a power-up, Mario stays in falling state when i enter a level and am unsure how to fix this myself. along with the ground being more pixelated than usual.
having the same issue, i can't pinpoint what does this but i think it has something to do with gravity?
Sorry for the late reply but this is caused by GameMaker Studio 2
You need to tick on "Collision Compatibility Mode" in the Game Options
Thank you! As for the pixelated ground, I found that you can fix it by increasing the Texture Page Size in the Windows Settings
|
2025-04-01T06:37:01.194349
| 2016-03-26T19:52:50
|
143743729
|
{
"authors": [
"Geal",
"coveralls",
"lu-zero"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1062",
"repo": "Geal/nom",
"url": "https://github.com/Geal/nom/pull/240"
}
|
gharchive/pull-request
|
doc: Add a contribution guideline
It is basically a copy and cut from rust-av.
Coverage decreased (-0.08%) to 84.262% when pulling 4df96c4ee4fdefadfeda29a7b355140780713be5 on lu-zero:master into 41f87f896ad03d1f2344be1146c19363b0e9172a on Geal:master.
Thanks!
|
2025-04-01T06:37:01.222997
| 2018-05-25T01:10:56
|
326347987
|
{
"authors": [
"Enalmada",
"SupriyaKalghatgi",
"erick2014",
"jannikbuschke",
"joshuahenson",
"kentrh",
"rosemarystanley"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1065",
"repo": "GeekyAnts/NativeBase",
"url": "https://github.com/GeekyAnts/NativeBase/issues/1932"
}
|
gharchive/issue
|
There is not plenty of space to use in Header - title
Hey guys, I have reviewed some issues related with this problem, but in all cases people are using a short text.
I'm no able to change the width of the Title within the Body component. There is plenty of space to use, but the Body component is using a very small portion of the available width.
Also in IOS: if I dont use a right button, the text gets lost:
react-native, react and native-base version
"react-native": "0.55.1",
"native-base": "^2.4.5",
Expected behaviour
To be able to use more width within the Body component, and align the text when there is not right button in IOS
Actual behaviour
Not using plenty of space available for the title and the alignment when there is not right button gets lost
Steps to reproduce (code snippet or screenshot)
const buildHomeBtn = navigation => {
return (
<Left>
<Button transparent onPress={() => navigation.navigate('PatientHome')}>
<Icon name='ios-home-outline' style={{ color: PRIMARY_DARK }} />
</Button>
</Left>
)
}
const buildLeftArrowIcon = leftIconHandler => {
console.log('leftIconHandler', leftIconHandler)
return (
<Left>
<Button transparent onPress={leftIconHandler}>
<Icon name='ios-arrow-back' style={{ color: PRIMARY_DARK }} />
</Button>
</Left>
)
}
const buildRightSideIcon = navigation => {
return (
<Right>
<Button transparent onPress={() => navigation.navigate('DrawerOpen')}>
<Icon name='menu' style={{ color: PRIMARY_DARK }} />
</Button>
</Right>
)
}
<Header style={typo.bgWhite}
androidStatusBarColor={getStatusBarColor(statusBarColor)}
iosBarStyle='light-content'>
{leftIconType === 'home' ? buildHomeBtn(navigation) : buildLeftArrowIcon(leftIconClickHandler)}
<Body>
<Title style={{ color:'blue', fontSize: 16 }}> RESUMEN DEL SERVICIO </Title>
</Body>
{showRightIcon ? buildRightSideIcon(navigation) : null}
</Header>
Is the bug present in both ios and android or in any one of them?
both
@erick2014
Also in IOS: if I dont use a right button, the text gets lost:
If you include Left / Body / Right, to align your elements, then you have to wrap the rest of the elements similarly for alignment
@SupriyaKalghatgi do you mean use Left or right within the Body?
I believe she means that if you are using the Left and Body component, you also have to use the Right component but you don't have to include anything inside it. Something like this should work for you.
{showRightIcon ? buildRightSideIcon(navigation) : <Right />}
Alright I'm going to try, thx for the explanation @joshuahenson
@erick2014 Did you try?
This issue will be closed tomorrow, in case of no response
Did anyone find a solution for this? I'm having the same issue. I'm using all three components Left / Body / Right and all of them seem to be cutting off like there's not enough room when really there is plenty of space.
<Header>
<Left>
<Button transparent onPress={() => this.props.navigation.navigate('Accounts')}>
<Icon active name='ios-arrow-back' type='Ionicons' />
<Text>Back</Text>
</Button>
</Left>
<Body>
<Title>EDIT ACCOUNTS</Title>
</Body>
<Right>
<Button transparent onPress={() => this.setState({ reorder: !this.state.reorder })}>
{(this.state.reorder ? <Icon active name='close' type='AntDesign' /> : <Icon name='ios-menu' type='Ionicons' />)}
</Button>
</Right>
</Header>
I have the same problem
<Header >
<Left>
<Button transparent onPress={() => navigation.goBack()}>
<Icon name='arrow-back' />
</Button>
</Left>
<Body style={{ backgroundColor: "black" }}>
<Title>{item.level3}</Title>
<Subtitle>{item.level4}</Subtitle>
</Body>
<Right />
</Header>
I tried-and-error setting some styles on the Body and Right component , and also removed the Right component without any good results.
I solved the truncated titles by putting style={{ flex: 2 }} on the <Body> tag
@SupriyaKalghatgi style={{ flex: 2 }} on the <Body> tag seems to workaround the problem. Can GeekyAnts get a more permanent fix?
|
2025-04-01T06:37:01.225626
| 2018-09-20T01:08:31
|
361980930
|
{
"authors": [
"SupriyaKalghatgi",
"jblevins1991"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1066",
"repo": "GeekyAnts/NativeBase",
"url": "https://github.com/GeekyAnts/NativeBase/issues/2287"
}
|
gharchive/issue
|
Floating Label not visible when building app with expo.
I have a very simple component that uses floatingLabels and when I use expo start they appear properly. When I use expo build:android the labels do not appear at all.
Here is more details from a forum thread. Unfortunately the solution the person proposed did not work for me I also wanted to post this issue here because using an older version is not a solution, its a workaround.
http://discuss.nativebase.io/t/nativebase-form-is-not-visible-properly-in-release-build-apk-generated-with-gradlew-assemblerelease/1142/7
Did you check #2264 before creating this issue?
I did a while ago, but didn't think to check today. A new version will be available this week. Thank you for the link.
|
2025-04-01T06:37:01.228449
| 2017-02-07T12:25:18
|
205872304
|
{
"authors": [
"appjitsu",
"gustavorozolin",
"sandropf",
"sankhadeeproy007"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1067",
"repo": "GeekyAnts/NativeBase",
"url": "https://github.com/GeekyAnts/NativeBase/issues/474"
}
|
gharchive/issue
|
InputMask or add a prop mask to Input component #feature
A suggestion
Add to the NativeBase a component InputMask or a prop in the Input component.
Example:
<Input mask = "99999-99" />
Or
<InputMask mask = "99999-99"/>
Thank you
Hi thanks for reporting this! NativeBase being a UI library, we'd like to keep business logic separate from NativeBase. You can use this feature of redux form for that.
What feature are you referring to exactly? I'd also like a input mask feature.
@sankhadeeproy007 Could you please reconsider adding this as input-mask is really common stuff in any of the UI frameworks.²
|
2025-04-01T06:37:01.240248
| 2018-09-05T18:15:13
|
357345556
|
{
"authors": [
"dalehenrich",
"ericwinger"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1068",
"repo": "GemTalk/Jadeite",
"url": "https://github.com/GemTalk/Jadeite/issues/148"
}
|
gharchive/issue
|
Poison pill for Jadeite alpha
The Jadeite alpha which was released in May will not work with the rowan services & JadeServer classes in Oscar.
We need to (potentially) add a poison pill so that the user cannot login to an Oscar-rowan server with the Jadeite alpha (2.0.1).
@dalehenrich Since no one is using Jadeite Alpha anymore, should we close this issue?
... not for Alpha1, but do we need one for Jade? There is some public information about how to install Jade persistently in an image that would cause problems for Jadeite and if anyone were to try to use Jade in a Rowan image (changing Rowanized methods or classes) that would definitely cause trouble ... I would think that creating a object that cannot be written to in one of the spots used by Jade might do the trick?
don't reassign to me unless you need it for your peace of mind ... I don't look at issues assigned to me on the Jadeite list in the normal course of business, but I do monitor the github mail traffic and that is usually how I recognize that I need to respond:)
@dalehenrich I assigned to you for a response to my question. Nothing more.
Putting in the poison pill for Jade is a good idea as much of the Jade service code would get overwritten. I'll reassign back to myself and mark as an April fix.
@dalehenrich I looked at the JadeServer installation code and Jade creates a new symbol list on the server then installs the JadeServer client code into the that symbol list. If persisted, this could make Jadeite non-functional as indicated earlier in this issue.
In thinking about a poison pill ... JadeServer is currently installed in UserGlobals. Installing it in Globals would eliminate any potential damage by anyone but SystemUser.
Another thought ... If Rowan controls all class definition compiles then couldn't Rowan just throw an error if JadeServer was recompiled? Any attempts to login by Jade would fail and code integrity would be preserved.
The "poison pill" idea is to force an error during installation so that the user does not get the idea that Jade can be safely used, or is inadvertently logging into a Rowanized stone ... eventually it will be legal to use the "standard" class and method creation calls at which point it won't hurt for Jade to be installed, so this is only an "issue" as long as a Rowanized stone is "vulnerable" ... if there is no practical way to prevent installation, then we'll just have to take our chances ...
@dalehenrich Is this still needed? We have a form of poison pill in that versions are checked on login - giving a warning message if the user is out of sync.
the alpha jadeite is no longer an issue
Retiring
|
2025-04-01T06:37:01.261218
| 2019-09-24T22:49:34
|
497960550
|
{
"authors": [
"LisaAlmarode",
"dalehenrich",
"ericwinger"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1069",
"repo": "GemTalk/Jadeite",
"url": "https://github.com/GemTalk/Jadeite/issues/530"
}
|
gharchive/issue
|
On one hand, it's nice that there was no walkback. On the other, it's frustrating because it's not clear how the class's dictionary could be empty.
It's possible the class needed an update from the server, but without a reproduction case it's going to be difficult to track down. What was done before the browser got into this state?
Turning on logging could provide a nice 'paper trail' to follow. A reproduction case would be better.
@LisaAlmarode If you can get logs of this happening or a reproduction case, I'd appreciate it. I've had no luck so far.
@ericwinger where is the code that defines the "containing dictionary"? is it in the Services code? If so, add a pointer to the place where you determine the containg dictionary and I shoul be able to provide you with a test case ... the problem reported above fails consistently, and it is in a private stone of mine:
If you want to login, you have to use rogue-vpn to find my machine when I am at home working ... let me know if you can't reproduce the problem and I will do the debugging on my end
@dalehenrich Take a look at RowanClassService>>setDictionary:. See if that sheds any light.
Note that Rowan knows which dictionary the class is is (Globals), when it creates the class definition template:
Looking at senders of RowanClassService>>setDictionary:, I see that both RowanClassService>>basicRefreshFrom: and RowanClassService>>minimalRefreshFrom: can return the meta class and with a little bit of testing in my image, I see that (Rowan image symbolList dictionariesAndSymbolsOf: GsQueryPredicate) first first name produces#'Globals' ... which is the correct answer and (Rowan image symbolList dictionariesAndSymbolsOf: GsQueryPredicate class) produces anArray( ) ... which is the wrong answer.
Anything other than a class returns an empty list, so I would suspect that Jadeite is either passing in a non-Class theClass or perhaps metaClass is defaulting to true. I'm pretty sure that this happens when I have just opened up fresh project browser on the class.
Here's an example where explicitly selecting the class button creates and empty list:
Ahh, you may have found it. Try this - If you click on the Class tab, I'll bet you don't see the dictionary. That would be a nice reproduction case.
Take a look at the final image in my previous comment ... I did select the 'Class' button ... howver, in the original report, the 'Instance' button is clearly selected, which is why I mentioned the the metaClass iv might default to true or you might be passing in a non-class object ...
This image? The class side is selected.
Exactly ... and as predicted ... the containing dictionary is empty ... it shouldn't be ... at a minimum this is wrong ... have you tried connecting to my stone? then you can do whatever experiments you need to do here ... AFAICT, both class tabs are already selected ... is there another class tab I am missing?
The problem is the meta class that you pointed out. I've got a fix ready for 3.0.78 for you to try.
... your fix won't necessarily fix the original problem (or at best it will mask the problem) as in the original bug report the instance tab is clearly selected ... so I think it might be an initial condition problem as well ...
Now that I've been workin gin this project ... I've added two more packages and the problem does not reproduce (instance tab ending up with empty string for dictionary name) ... so there is clearly something else going wrong here ...
I am suspicious of this line of code:
classOrMeta := meta == true ifTrue:[theClass class] ifFalse:[theClass].
if meta is nil, you will return the meta class for theClass, which is what I am seeing ... and the fact that you aren't using meta directly implies that in some circumstances meta will be set to a non-boolean value (nil perhaps) ... you've got an initialize method that sets meta to false, but the behavior that I'm seeing in my initial case, implies that someone is using the setter (meta:) to set meta to a non-boolean value --- or to true ... I've lost my "reproducable test case" ... so I cannot test your code for you ... so I think that unless additional work is done this bug will still be lurking around waiting to hit us again ...
Looking at all references in the ServiceClass and verifying that the setters all look good and then looking at all senders of meta: is probably a good idea, since your fix will just cause another part of the code to go haywire, if meta is not being handled correctly --- which appears to be the case ...
Yes, meta can legitimately be nil in some cases especially when the class service originates in Dolphin. e.g. Doing a find class doesn't yet know it's meta so it assumes false.
fyi - None of the images you've posted on this issue show the instance tab selected with the problem issue. If you can send me a picture of the instance tab selected & the problem dictionary I'll look into it further.
Ahh, I did not notice that ... I don't believe that I did anything other than open the project browser and select the class in question (without touching any of the class tabs) I assumed that I was looking at the instance side in my original bug report. So presumably there is another condition (closely related to my original bug report, which does not reproduce now:) where the class side of a class gets selected when I open a fresh project browser and simply select the class ...
I see that the class tab is sticky:
... this is unexpected to me ... is this "stickiness" expected behavior?
Earlier today I wasted a fair amount of time writing code on the class side of a class when I thought I was on the instance side ... I couldn't understand why the project browser "put code on the class-side of a class" ...
I expect to be on the instance side when I switch classes, so I guess I'm going to have to start training myself to worry about the instance/class tab whenever I am working with Jadeite, unless this is considered to be a bug by you and/or @LisaAlmarode :) ... a couple more times of wasting time will probably be enough ...
It's a bit of a surprise, that I just found this out after working in Jadeite for a year and a half, but clearly I haven't been bit by this assumption until today:)
So... it is very possible that I was working on another class when I switched to the bad boy and I just assumed that the instance side was selected... sorry about that ...
... this is unexpected to me ... is this "stickiness" expected behavior?
The class tab "stickiness" is actually intended in some scenarios.
For example, if you have an open browser on a class with class tab selected then use Jadeite menu option New Projects Browser (Ctrl+N). It will open up the new browser on that class with the class tab selected. The assumption is that you want to open a new browser in the same place you are currently rather than starting from a fresh browser.
Alternatively, if you do a Find Class ... from the console's Jadeite>Browse menu, it will open on the instance side. The assumption is that you are starting your work from scratch and most work is done on the instance side of classes.
Yet, it's possible that there might be a bug or enhancements in this area. Since I've got a fix for the specific problem coming in 3.0.78, can we close this issue for review by @LisaAlmarode and open a new issue to examine class tab "stickiness" behavior?
Well the case where I have observed the "stickiness" is when I switch to a different class in the same browserr ... I also did a Find Class ... earlier today and ended up with the class tab selected ... of course it didn't reproduce --- neither of these use cases fall under your "intentional Use case" ... so there does appear to be some more bugs in this area ...
This issue got a bit muddled. The bulk of this is discussion about the selection of class or instance doesn't seem to have any problems (other than new issue #555 reported).
The case where RowanClassService on the class side showed an empty dictionary is fixed in 3.0.79.
There were cases long ago in which some menu operations acted wonky (e.g. drag and drop moved methods to the class side), due to internal defaults to meta. It seems possible that the initial report on this case may be due to a leftover case of this kind. However, we'd need some details as to the steps preceding to hope to reproduce this... I tried a few cases but there are too many possibilities.
Given the one issue fixed, and the large amount of digressions, if we have another case of a class's go-to-dictionary turning up empty, a new issue should be opened.
|
2025-04-01T06:37:01.264691
| 2023-12-07T21:53:23
|
2031591863
|
{
"authors": [
"LisaAlmarode",
"ericwinger"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1070",
"repo": "GemTalk/JadeiteForPharo",
"url": "https://github.com/GemTalk/JadeiteForPharo/issues/55"
}
|
gharchive/issue
|
Request changing the background color on edit/error
In JforD, when you start editing a method, the background turns green. If you save with an error, it turns pink.
I really like these color changes (the green, at least). The pharo corner color is completely inadequate. The GBS big red dot on the Source tab is at least right there in your face, where the code is that you are staring at, not way off to the far side of the windows.
This mentioned in #44 but deserves its own issue.
Should have the green background working in https://github.com/GemTalk/JadeiteForPharo/commit/c9bd0ba6c9fcc584b2db711500ccc454a1b4a9ad
Added red dot that shows up if method source, class comment, or class definition panes changed.
https://github.com/GemTalk/JadeiteForPharo/commit/f006867fe29dadab863b0774c83096bc4b32040e
https://github.com/GemTalk/JadeiteForPharo/commit/919183710b732385e041cd13899918807d98b4a7
|
2025-04-01T06:37:01.311921
| 2019-11-21T03:17:19
|
526339429
|
{
"authors": [
"mikebronner",
"saernz"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1071",
"repo": "GeneaLabs/laravel-model-caching",
"url": "https://github.com/GeneaLabs/laravel-model-caching/issues/305"
}
|
gharchive/issue
|
Updates fail with concurrent query / updates
Describe the bug
Hi I believe I have found a bug which appears to rear its head when queries and updates happen at the same time, I have also identified an area on the code which may be causing this problem and a simple fix which seems to prevent this behaviour and may also help save up to 50% of memory storage in redis. This is potentially loosely related to #201 and may suggest behaviour described in this issue is not intended.
I found when polling a certain resource on my server that if I do an update to that resource the cache will fail to update 100% of the time even though the database field clearly has changed. When polling is disabled the cache works as expected.
I found the problem exists in GeneaLabs\LaravelModelCaching\Traits\Buildable in the retrieveCachedValue(...) function. It appears a race condition can happen in this function if a query and update are performed at the same time. The problem happens when a first() query is performed which calls cachedValue() > retrieveCachedValue(). This actually results in two results being stored in the cache. When the cache is found to be empty at rememberForever() the parent::{$method}(...$arguments) function is called which resolves to the first() method in Illuminate\Database\Concerns\BuildsQueries as intended. The breakdown happens here as the first() method in this class actually runs $this->take(1)->get($columns)->first(), which unintentionally calls back GeneaLabs\LaravelModelCaching\Traits\Buildable::get() as the CachedBuilder is overriding the get() method. This actually results in an additional call to retrieveCachedValue() which adds in another key/val pair to redis, but most dangerously this tries to recall the cache when we actually intended it to skip the cache and call the database. Because of this recursive behaviour a dirty read from the cache can happen at exactly the same time as a flush and the dirty read ends up becoming the new value instead of the updated value in the database.
This problem can be fixed by simply adding $this->disableModelCaching(); to the anonymous function in the retrieveCachedValue() method as below:
return $this->cache($cacheTags)
->rememberForever(
$hashedCacheKey,
function () use ($arguments, $cacheKey, $method) {
$this->disableModelCaching();
return [
"key" => $cacheKey,
"value" => parent::{$method}(...$arguments),
];
}
);
This prevents any recursive calls back to the CachedBuilder, or at least causes the calls to fall back to the EloquentBuilder instead. As a result of adding this it also prevents the extra get() query from being unintentionally cached saving some space in redis.
Stack Trace
A stack trace of the recursive behaviour as explained above:
[0] => Array
(
[file] => /srv/application/vendor/laravel/framework/src/Illuminate/Cache/Repository.php
[line] => 422
[function] => GeneaLabs\LaravelModelCaching\Traits\{closure}
[class] => GeneaLabs\LaravelModelCaching\CachedBuilder
[type] => ->
)
[1] => Array
(
[file] => /srv/application/vendor/genealabs/laravel-model-caching/src/Traits/Buildable.php
[line] => 300
[function] => rememberForever
[class] => Illuminate\Cache\Repository
[type] => ->
)
[2] => Array
(
[file] => /srv/application/vendor/genealabs/laravel-model-caching/src/Traits/Buildable.php
[line] => 231
[function] => retrieveCachedValue
[class] => GeneaLabs\LaravelModelCaching\CachedBuilder
[type] => ->
)
[3] => Array
(
[file] => /srv/application/vendor/genealabs/laravel-model-caching/src/Traits/Buildable.php
[line] => 100
[function] => cachedValue
[class] => GeneaLabs\LaravelModelCaching\CachedBuilder
[type] => ->
)
[4] => Array
(
[file] => /srv/application/vendor/laravel/framework/src/Illuminate/Database/Concerns/BuildsQueries.php
[line] => 77
[function] => get
[class] => GeneaLabs\LaravelModelCaching\CachedBuilder
[type] => ->
)
[5] => Array
(
[file] => /srv/application/vendor/genealabs/laravel-model-caching/src/Traits/Buildable.php
[line] => 293
[function] => first
[class] => Illuminate\Database\Eloquent\Builder
[type] => ->
)
[6] => Array
(
[file] => /srv/application/vendor/laravel/framework/src/Illuminate/Cache/Repository.php
[line] => 422
[function] => GeneaLabs\LaravelModelCaching\Traits\{closure}
[class] => GeneaLabs\LaravelModelCaching\CachedBuilder
[type] => ->
)
[7] => Array
(
[file] => /srv/application/vendor/genealabs/laravel-model-caching/src/Traits/Buildable.php
[line] => 300
[function] => rememberForever
[class] => Illuminate\Cache\Repository
[type] => ->
)
[8] => Array
(
[file] => /srv/application/vendor/genealabs/laravel-model-caching/src/Traits/Buildable.php
[line] => 231
[function] => retrieveCachedValue
[class] => GeneaLabs\LaravelModelCaching\CachedBuilder
[type] => ->
)
[9] => Array
(
[file] => /srv/application/vendor/genealabs/laravel-model-caching/src/Traits/Buildable.php
[line] => 79
[function] => cachedValue
[class] => GeneaLabs\LaravelModelCaching\CachedBuilder
[type] => ->
)
[10] => Array
(
[file] => /srv/application/vendor/cloudcreativity/laravel-json-api/src/Eloquent/AbstractAdapter.php
[line] => 218
[function] => first
[class] => GeneaLabs\LaravelModelCaching\CachedBuilder
[type] => ->
)
[11] => Array
(
[file] => /srv/application/vendor/cloudcreativity/laravel-json-api/src/Store/Store.php
[line] => 245
[function] => find
[class] => CloudCreativity\LaravelJsonApi\Eloquent\AbstractAdapter
[type] => ->
)
[12] => Array
(
[file] => /srv/application/vendor/cloudcreativity/laravel-json-api/src/Store/Store.php
[line] => 257
[function] => find
[class] => CloudCreativity\LaravelJsonApi\Store\Store
[type] => ->
)
[13] => Array
(
[file] => /srv/application/vendor/cloudcreativity/laravel-json-api/src/Routing/Route.php
[line] => 93
[function] => findOrFail
[class] => CloudCreativity\LaravelJsonApi\Store\Store
[type] => ->
)
Environment
PHP: 7.3.9
OS: Alpine
Laravel: 5.8.19
Model Caching: 0.7.0
Hope this isn't too lengthy. Thanks for your help.
@saernz Thanks for this detailed report! Very interesting find, indeed! I will implement the fix you suggest and run it against the unit tests and report back. Please give me a few days to get back to you on this, I will try to get to it on Friday at the latest.
@mikebronner Thanks man. Hopefully I'm not wrong, though I'll leave it up to the experts to decide :)
Unfortunately I don't think my fix has worked as I ran into the same problem again. I believe my DB lock is not working some how and when the cache updates after a flush it does a dirty read of the DB some how as it ends up caching the old value. I think the bug I described above still exists, though I'm not sure if it's causing the race condition or not. From looking at the code and stack traces I believe the recursive call may still be a bug but I'm not sure how it's all relating to the race condition I'm having, will need to investigate this next week to know for sure.
I've looked into this a bit more now and I believe I understand the problem a lot better. There isn't a race condition as I initially described in my original post, though I believe stopping that recursive behaviour will still help stop an intermediate value being stored in the cache caused by the eloquent builders call to get().
I have found I'm getting the wrong value returned from cache after an update if I use transactions. What seems to happen is when my update is performed in a transaction the value is correctly updated in the DB, and the cache gets flushed, but if I don't commit the transaction quick enough before the next read the cache will still get the old DB value as the transaction has not committed.
I was a bit confused before as I saw the cache was flushed only after the DB had been changed and what I was seeing should technically not be possible, though because I'm also using transactions technically the flush can happen before the transaction has been committed to the database. This mainly happens because I have to dispatch an event after my record has been updated, unfortunately the framework I'm using only provides a hook to update events within the database transaction, as a result dispatching the event delayed the commit of the transaction so a read could sneak in between the database flush and the transaction being committed.
The workaround for this was to perform another flush by using $myModel->flushCache() either after the transaction or as close as you can get to the transaction being closed. I'm not sure if a more permanent fix for this library would be to somehow run a flush after a transaction has been committed, though I'm not sure how easy this will be. Possibly just documenting how to use the library with transactions might be enough.
@saernz Thanks for the follow-up. I will update the documentation to explain the work-around with transactions.
Also, I was unable to implement your suggested fix in your first post, as it breaks unit tests. It might have worked well for you specific use-case, but it seems to break other areas.
|
2025-04-01T06:37:01.339047
| 2022-02-12T13:55:11
|
1134093358
|
{
"authors": [
"DoomPtrl",
"GenjiApp",
"audioses"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1072",
"repo": "GenjiApp/RingerVolume",
"url": "https://github.com/GenjiApp/RingerVolume/issues/17"
}
|
gharchive/issue
|
how to use?
Hello there. How do we use this program, can you provide some basic steps, I'M not a programmer.
You must have a Mac and build this app by yourself. You must have an experience of develop iOS app.
Download this project
Open this project on Xcode
Build an app and install to iOS devices
Is there any other way I can use this app?
|
2025-04-01T06:37:01.357423
| 2020-05-13T10:27:52
|
617332346
|
{
"authors": [
"lcx19950201",
"rom1v"
],
"license": "apache-2.0",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1073",
"repo": "Genymobile/scrcpy",
"url": "https://github.com/Genymobile/scrcpy/issues/1387"
}
|
gharchive/issue
|
How to delete the attached adb script
I wrote some adb scripts to control the phone, but the accompanying adb scripts conflicted, such as using 'w''s''d''a' to control, and the right mouse button. At the same time I found that my adb script has a short time (~ 0.05s) delay, and sometimes does not work, I use the python + adb method.
Please tell me how to delete the internal adb, if you can, can you try to solve my problem? Thank you
What do you mean by "adb script"? What do you mean by "conflicted"?
adb code：db shell input tap 1280 560
Because scrpy comes with some adb operations, such as the right mouse button is the return button, how do I delete them
scrcpy --no-control
I want to assign a new adb command to the right mouse button （
Scrcpy does not use adb shell input … to inject events: https://github.com/Genymobile/scrcpy/issues/231#issuecomment-414111753
For now, it is not possible to reassign mouse buttons. But there is already a feature request for that: https://github.com/Genymobile/scrcpy/issues/1302
I created a new python script to output adb commands through os, but the right mouse button is occupied.
Could you share your script, because I don't understand what you mean by "the right mouse button is occupied". If you execute adb shell input ... commands, it's totally independant of scrcpy.
https://github.com/lcx19950201/Python-/blob/master/adb.py
|
2025-04-01T06:37:01.362681
| 2022-03-08T12:12:46
|
1162582908
|
{
"authors": [
"rom1v",
"smoqmilus"
],
"license": "apache-2.0",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1074",
"repo": "Genymobile/scrcpy",
"url": "https://github.com/Genymobile/scrcpy/issues/3096"
}
|
gharchive/issue
|
Scrcpy starts and then stops, no error is being thrown
[x] I have read the FAQ.
[x] I have searched in existing issues.
Environment
OS: Windows
scrcpy version: 1.23
installation method: extracted from the package
device model: Google Pixel 2
Android version: 11
Describe the bug
The scrcpy executable runs and stops without any indication of the problem.
I attached the logcat.log file as well to help you with finding a root cause.
.\scrcpy.exe
scrcpy 1.23 <https://github.com/Genymobile/scrcpy>
C:\scrcpy\scrcpy-server: 1 file pushed, 0 skipped. 27.6 MB/s (41123 bytes in 0.001s)
[server] INFO: Device: Google Pixel 2 (Android 11)
logcat.log
Hi,
I ran the command a second time and redirected the logcat output to the file.
Let me know if it is useful for you.
Thanks,
Martin
logcat.log
Try with another encoder: https://github.com/Genymobile/scrcpy#encoder
|
2025-04-01T06:37:01.389581
| 2022-10-21T14:10:18
|
1418403178
|
{
"authors": [
"GeorgeCiesinski",
"Wissididom"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1075",
"repo": "GeorgeCiesinski/poke-guesser-bot",
"url": "https://github.com/GeorgeCiesinski/poke-guesser-bot/issues/79"
}
|
gharchive/issue
|
Migration command
Describe the new feature
A simple discord command for migrating the replit database to the new postgres database by providing the value of the REPLIT_DB_URL env var from replit.
Describe alternatives you've considered
Another option would be to simply get all settings and such and resetup them in the new version, but that has a higher expense depending on how much users might need to do that.
Planning
I thought of some command like /migrate <replit_db_url> which can be enabled in the .env for migrating the old replit database to postgres. We can probably even write the command to get it (echo $REPLIT_DB_URL) in the description of the command.
API
Simply like with every other command to give them their own file and just call their main function (in this case probably migrate in the file migrate.js)
Useful links
Replit Database FAQ
Using Databases in Replit
Replit's Database Module on NPM
I was just going through the issues and wanted to see if we need to keep this one open. As far as I know, there is one user who is officially using the replit version. If that is the case, it might not make sense to write a feature that is only going to be used once.
Assuming this user is going to switch to the new version once it is released, is data going to need to be transferred, or can it be done following a "championship" so that a new game can start from scratch in the new version?
Also, is there any changes that need to happen on the discord bot settings? I know that the slash commands need to be installed, but besides that, is there anything else that needs done?
If that guy ends it with a championship it would be resetting up the allowed mods and allowed channels although the guy could also simply set it within Discords Command Permission system. Other than that I think running setupCommands once and running the bot would be enough.
If he/she does not want to reset the scores he'd need to use the /mod score subcommands for each user.
As I don't think people using the replit bot have a lot of members in their server I'm fine with not creating the migration command for the reason you mentioned.
Also the guy at some point has to move over to the new version because discord.js v12 won't be supported by Discord (for the API version they talk to) forever.
Got it. In that case I will go ahead and close this issue.
|
2025-04-01T06:37:01.403871
| 2022-06-21T19:42:15
|
1278967901
|
{
"authors": [
"DSLituiev"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1076",
"repo": "Georgetown-IR-Lab/QuickUMLS",
"url": "https://github.com/Georgetown-IR-Lab/QuickUMLS/issues/86"
}
|
gharchive/issue
|
[BUG] SSLError workarounds?
This seems to be a firewall related error, which has known workarounds for standard PIP install.
python -m quickumls.install . ~/repos/umls-files
Error:
python -m quickumls.install . ~/repos/umls-files
Determining if SpaCy for language "ENG" is installed...
SpaCy is not available! Attempting to download and install...
⚠ As of spaCy v3.0, shortcuts like 'en' are deprecated. Please use the
full pipeline package name 'en_core_web_sm' instead.
Traceback (most recent call last):
File "/home/dlituiev/anaconda3/envs/spacy/lib/python3.10/site-packages/quickumls/install.py", line 130, in install_spacy
spacy.load(SPACY_LANGUAGE_MAP[lang])
File "/home/dlituiev/anaconda3/envs/spacy/lib/python3.10/site-packages/spacy/__init__.py", line 51, in load
return util.load_model(
File "/home/dlituiev/anaconda3/envs/spacy/lib/python3.10/site-packages/spacy/util.py", line 426, in load_model
raise IOError(Errors.E941.format(name=name, full=OLD_MODEL_SHORTCUTS[name])) # type: ignore[index]
OSError: [E941] Can't find model 'en'. It looks like you're trying to load a model from a shortcut, which is obsolete as of spaCy v3.0. To load the model, use its full name instead:
nlp = spacy.load("en_core_web_sm")
For more details on the available models, see the models directory: https://spacy.io/models. If you want to create a blank model, use spacy.blank: nlp = spacy.blank("en")
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/home/dlituiev/anaconda3/envs/spacy/lib/python3.10/site-packages/urllib3/connectionpool.py", line 703, in urlopen
httplib_response = self._make_request(
File "/home/dlituiev/anaconda3/envs/spacy/lib/python3.10/site-packages/urllib3/connectionpool.py", line 386, in _make_request
self._validate_conn(conn)
File "/home/dlituiev/anaconda3/envs/spacy/lib/python3.10/site-packages/urllib3/connectionpool.py", line 1040, in _validate_conn
conn.connect()
File "/home/dlituiev/anaconda3/envs/spacy/lib/python3.10/site-packages/urllib3/connection.py", line 414, in connect
self.sock = ssl_wrap_socket(
File "/home/dlituiev/anaconda3/envs/spacy/lib/python3.10/site-packages/urllib3/util/ssl_.py", line 449, in ssl_wrap_socket
ssl_sock = _ssl_wrap_socket_impl(
File "/home/dlituiev/anaconda3/envs/spacy/lib/python3.10/site-packages/urllib3/util/ssl_.py", line 493, in _ssl_wrap_socket_impl
return ssl_context.wrap_socket(sock, server_hostname=server_hostname)
File "/home/dlituiev/anaconda3/envs/spacy/lib/python3.10/ssl.py", line 512, in wrap_socket
return self.sslsocket_class._create(
File "/home/dlituiev/anaconda3/envs/spacy/lib/python3.10/ssl.py", line 1070, in _create
self.do_handshake()
File "/home/dlituiev/anaconda3/envs/spacy/lib/python3.10/ssl.py", line 1341, in do_handshake
self._sslobj.do_handshake()
ssl.SSLCertVerificationError: [SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: unable to get local issuer certificate (_ssl.c:997)
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/home/dlituiev/anaconda3/envs/spacy/lib/python3.10/site-packages/requests/adapters.py", line 489, in send
resp = conn.urlopen(
File "/home/dlituiev/anaconda3/envs/spacy/lib/python3.10/site-packages/urllib3/connectionpool.py", line 785, in urlopen
retries = retries.increment(
File "/home/dlituiev/anaconda3/envs/spacy/lib/python3.10/site-packages/urllib3/util/retry.py", line 592, in increment
raise MaxRetryError(_pool, url, error or ResponseError(cause))
urllib3.exceptions.MaxRetryError: HTTPSConnectionPool(host='raw.githubusercontent.com', port=443): Max retries exceeded with url: /explosion/spacy-models/master/compatibility.json (Caused by SSLError(SSLCertVerificationError(1, '[SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: unable to get local issuer certificate (_ssl.c:997)')))
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/home/dlituiev/anaconda3/envs/spacy/lib/python3.10/runpy.py", line 196, in _run_module_as_main
return _run_code(code, main_globals, None,
File "/home/dlituiev/anaconda3/envs/spacy/lib/python3.10/runpy.py", line 86, in _run_code
exec(code, run_globals)
File "/home/dlituiev/anaconda3/envs/spacy/lib/python3.10/site-packages/quickumls/install.py", line 233, in <module>
main()
File "/home/dlituiev/anaconda3/envs/spacy/lib/python3.10/site-packages/quickumls/install.py", line 171, in main
install_spacy(opts.language)
File "/home/dlituiev/anaconda3/envs/spacy/lib/python3.10/site-packages/quickumls/install.py", line 134, in install_spacy
spacy.cli.download(SPACY_LANGUAGE_MAP[lang])
File "/home/dlituiev/anaconda3/envs/spacy/lib/python3.10/site-packages/spacy/cli/download.py", line 67, in download
compatibility = get_compatibility()
File "/home/dlituiev/anaconda3/envs/spacy/lib/python3.10/site-packages/spacy/cli/download.py", line 78, in get_compatibility
r = requests.get(about.__compatibility__)
File "/home/dlituiev/anaconda3/envs/spacy/lib/python3.10/site-packages/requests/api.py", line 73, in get
return request("get", url, params=params, **kwargs)
File "/home/dlituiev/anaconda3/envs/spacy/lib/python3.10/site-packages/requests/api.py", line 59, in request
return session.request(method=method, url=url, **kwargs)
File "/home/dlituiev/anaconda3/envs/spacy/lib/python3.10/site-packages/requests/sessions.py", line 587, in request
resp = self.send(prep, **send_kwargs)
File "/home/dlituiev/anaconda3/envs/spacy/lib/python3.10/site-packages/requests/sessions.py", line 701, in send
r = adapter.send(request, **kwargs)
File "/home/dlituiev/anaconda3/envs/spacy/lib/python3.10/site-packages/requests/adapters.py", line 563, in send
raise SSLError(e, request=request)
requests.exceptions.SSLError: HTTPSConnectionPool(host='raw.githubusercontent.com', port=443): Max retries exceeded with url: /explosion/spacy-models/master/compatibility.json (Caused by SSLError(SSLCertVerificationError(1, '[SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: unable to get local issuer certificate (_ssl.c:997)')))
**Environment **
OS: WSL Ubuntu
QuickUMLS version [e.g. 1.4]
UMLS version [e.g. 2022AA]
Python 3.10
spacy: spacy-3.3.1
en_core_web_sm is installed
it would be great to have --trusted-host argument
also it seems like an antiquated requirement for en language instead of a specific spacy model like en_core_web_sm (see spacy warning in the log)
I've patched it by replacing in constants.py: en -> en_core_web_sm
|
2025-04-01T06:37:01.408655
| 2018-02-20T12:35:26
|
298578123
|
{
"authors": [
"ianneilmacleod"
],
"license": "bsd-2-clause",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1077",
"repo": "GeosoftInc/gxapi",
"url": "https://github.com/GeosoftInc/gxapi/issues/16"
}
|
gharchive/issue
|
memory grid has an invalid kx, asserts baseimg.c {1919}
See issues/issue 16 be4be3a57bd4fca42ef128
Options:
(prefered) ensure memory grids behave like persistent grids, in which case they can have a KX +1 or -1.
(easier, but the bug may appear somewhere else) baseimg can tread KX == 0 as KX == 1, does not need to assert.
See https://github.com/GeosoftInc/gxapi/tree/master/issues/issue 16 for a failing program.
See also https://github.com/GeosoftInc/gxapi/blob/master/tests/python/test_grid.py, which has a skipped test that will fail until this is fixed.
Jacques found that is call img.opt_kx(kx) the kx is set and assertions go away. Still should not assert though...
Resolved fda2032d3ef25067be57fc344d06
|
2025-04-01T06:37:01.410268
| 2023-01-25T22:41:19
|
1557368141
|
{
"authors": [
"ccmothes",
"dhunt22"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1078",
"repo": "GeospatialCentroid/NASA-prison-EJ",
"url": "https://github.com/GeospatialCentroid/NASA-prison-EJ/pull/11"
}
|
gharchive/pull-request
|
superfund, power plants, init weight calculation functions
Review this pull request carefully for compatibility. I had to stash and sync my fork a few times.
In this commit, I have split point_data_factors.R into npl_super and power_plants. I saved the new npl_superfund geocoded file to the data/processed folder. I also created the framework for some weighting functions.
I think it would be a good idea to restructure the R folder to have subdirectories of "functions", "factors", and 'standalone scripts'.
Looks great! I'm definitely on board with a file restructure, let's revisit once most of the processing code is developed.
|
2025-04-01T06:37:01.477415
| 2016-07-29T08:01:34
|
168271311
|
{
"authors": [
"Ghosh",
"i-break-codes"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1079",
"repo": "Ghosh/uiGradients",
"url": "https://github.com/Ghosh/uiGradients/pull/163"
}
|
gharchive/pull-request
|
Add gradient, remove unused method in Gradients.js + minor linting
Added 3 new gradients to gradients.json
getGradients() in Gradients.js was not used anymore, so removed
Minor refactor
Minor linting, adding missing semi colons, spaces between braces, removed unnecessary white space
@i-break-codes Thanks for the PR, this is super.
One minor edit though, can you just please revert the getGradients() method. It is basically like a utility method I use sometimes. It's left there intentionally.
@Ghosh aah, my bad, sorry, I didn't realized that! Well, reverted back.
Thanks 👍
|
2025-04-01T06:37:01.501031
| 2021-02-12T06:33:58
|
806987660
|
{
"authors": [
"ghost",
"kenkit",
"ufoDziner"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1081",
"repo": "GiorgioAresu/FanController",
"url": "https://github.com/GiorgioAresu/FanController/issues/15"
}
|
gharchive/issue
|
How to stop the fans?
I can't seem to find a way to stop the fans. The obvious setDutyCycle(0) doesn't work.
In forums I've read that you can turn it off by switching pin mode to OUTPUT and digitalWrite(pin, LOW).
But I don't know what effect that would have while this libraries code is running (problems along the line, how to start the fans again?)
I would highly appreciate if someone could implement a method stop() and maybe restart()
Just started to use this library. It's great, but I can't stop the fans either. Closest I can get is ~5% without side effects. Any chance there is a fix in the works? Thanks!
You can use another input and a mofset as mentioned in this other diagram.
You only need the mofset part
https://github.com/sker65/esphome-fan-controller
|
2025-04-01T06:37:01.536941
| 2017-01-05T10:34:36
|
198923919
|
{
"authors": [
"Soreine"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1082",
"repo": "GitbookIO/feedback",
"url": "https://github.com/GitbookIO/feedback/issues/295"
}
|
gharchive/issue
|
Custom 404 Not Found pages for books
Most books would benefit from having relevant 404 pages. I can see several solutions:
Book setting to setup a redirection URL on 404 (simple and flexible)
404 pages defined in the gitbook (not great because it introduces a new convention in the toolchain for 404.md files)
GitBook.com could add a relevant link to the book homepage on 404 pages
Related ticket: https://gitbook.zendesk.com/agent/tickets/4891
I think we will opt for solution 3. Improving the 404 page for books, by providing a call to action:
It looks like the page you accessed does not exist, or was moved.
Go back to the content: <root_url_of_the_book>
|
2025-04-01T06:37:01.668113
| 2023-08-15T11:27:19
|
1851282660
|
{
"authors": [
"erikmannerfelt",
"rhugonnet"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1126",
"repo": "GlacioHack/xdem",
"url": "https://github.com/GlacioHack/xdem/issues/404"
}
|
gharchive/issue
|
Coreg.fit_pts(..., mask_high_curv=True) does not seem to work.
I'm getting an error when enabling the mask_high_curv=True flag (after adhering to the warning: "Warning: There is no curvature in dataframe. Set mask_high_curv=True for more robust results").
File /nix/store/kyrwb500hfcyyabb66msvq2whpvhmh0d-python3-3.10.10-env/lib/python3.10/site-packages/xdem/coreg/base.py:907, in Coreg.fit_pts(self, reference_dem, dem_to_be_aligned, inlier_mask, transform, samples, subsample, verbose, mask_high_curv, order, z_name, weights)
903 ref_dem = reference_dem[ref_valid]
905 if mask_high_curv:
906 maxc = np.maximum(
--> 907 np.abs(get_terrain_attribute(tba_dem, attribute=["planform_curvature", "profile_curvature"])), axis=0
908 )
909 # Mask very high curvatures to avoid resolution biases
910 mask_hc = maxc.data > 5.0
TypeError: bad operand type for abs(): 'Raster'
The code in question is here:
https://github.com/GlacioHack/xdem/blob/02902c095ffb2bfbf422e90c13d4fce227da6583/xdem/coreg/base.py#L905-L910
I don't know if this piece of code was ever tested, or if something new has broken it! Either way, it doesn't seem to work right now, as np.abs(Raster) doesn't work it seems. @adehecq or @rhugonnet, do you know if this has ever worked? I see three potential solutions:
Make np.abs(Raster) work in GeoUtils.
Change the problematic line to: np.max(np.abs(get_terrain_attribute(tba_dem, attribute=[...]).data), axis=0) (note the .data addition to make it a masked array instead of a raster).
Remove this functionality here altogether and make it a Filter
Also, I don't think it's great that a maxc limit of 5.0 is hardcoded. Should that be made a keyword argument?
What do you think @rhugonnet?
Yes the tests are still poor for these new point functions (I intended to work a bit on that in the next PR on coregistration, but I'm not there yet). We didn't insist too much on tests in #346 as we knew we were about to rework some of the module.
I have no idea if it has ever worked, never used it.
I agree, actually I think both directions should be done:
We should definitely move this to coreg/filters in time, with the limit a keyword argument,
It'd be great if we added np.abs to GeoUtils in single-input handled functions: https://github.com/GlacioHack/geoutils/blob/main/geoutils/raster/raster.py#L68.
To keep new features cleanly tested in the future (and ensure they are added to API, we update the package history, etc...), we could add an automated "PR" checklist in .github? Something like this: https://github.com/pyproj4/pyproj/blob/main/.github/PULL_REQUEST_TEMPLATE.md.
What do you think @erikmannerfelt @adehecq?
I added np.abs and np.absolute (aliases) in https://github.com/GlacioHack/geoutils/pull/393, it was just a couple lines!
I remember now: the reason that line of code might have worked before is because the Raster class had the __array_interface__. Unfortunately we had to deactivate it for now because it created infinite loops when called by np.ma functions, and messed up some priorities for arithmetic functions. They are working on fixing that in NumPy by adding an interface for masked arrays (...eventually!) :sweat_smile:
Yes the tests are still poor for these new point functions (I intended to work a bit on that in the next PR on coregistration, but I'm not there yet). We didn't insist too much on tests in #346 as we knew we were about to rework some of the module. I have no idea if it has ever worked, never used it.
I agree, actually I think both directions should be done:
* We should definitely move this to `coreg/filters` in time, with the limit a keyword argument,
* It'd be great if we added `np.abs` to GeoUtils in single-input handled functions: https://github.com/GlacioHack/geoutils/blob/main/geoutils/raster/raster.py#L68.
To keep new features cleanly tested in the future (and ensure they are added to API, we update the package history, etc...), we could add an automated "PR" checklist in .github? Something like this: https://github.com/pyproj4/pyproj/blob/main/.github/PULL_REQUEST_TEMPLATE.md. What do you think @erikmannerfelt @adehecq?
Awesome @rhugonnet, thanks for the input! Yeah a checklist would be great. Even though there's no check that the listed parts are actually implemented, I think it could work in our favour in the long term.
Also, thanks a lot for fixing the problem on the GU side!!
Oh, and perhaps internally we could have a routine for PRs like #346. I was the one who pushed it despite it not being absolutely finished, so I'm mostly to blame. But I suspect there may be times when this could happen again, such as when @adehecq needed to quickly implement some functionality for his workshop a year ago (?).
It would be good to have an "express train" routine to make sure that the essentials are merged, and cleanup happens soon thereafter!
|
2025-04-01T06:37:01.679881
| 2023-08-02T14:45:06
|
1833332838
|
{
"authors": [
"greyscaled",
"scsmithr",
"sunfishcode"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1127",
"repo": "GlareDB/glaredb",
"url": "https://github.com/GlareDB/glaredb/issues/1460"
}
|
gharchive/issue
|
Investigate: Building 0.3.0 image failed the first time
Context
It failed quickly (~48seconds)
https://github.com/GlareDB/glaredb/actions/runs/5739955754/job/15556727166#step:6:164
with:
error[E0433]: failed to resolve: could not find FsConfigCmd in types
--> /usr/local/cargo/registry/src/index.crates.io-6f17d22bba15001f/rustix-0.38.5/src/backend/linux_raw/mount/syscalls.rs:214:27
|
214 | super::types::FsConfigCmd::Create,
| ^^^^^^^^^^^ could not find FsConfigCmd in types
error[E0433]: failed to resolve: could not find FsConfigCmd in types
--> /usr/local/cargo/registry/src/index.crates.io-6f17d22bba15001f/rustix-0.38.5/src/backend/linux_raw/mount/syscalls.rs:228:27
|
228 | super::types::FsConfigCmd::Reconfigure,
| ^^^^^^^^^^^ could not find FsConfigCmd in types
Compiling serde_json v1.0.104
Compiling camino v1.1.6
Compiling crypto-common v0.1.6
error[E0277]: the trait bound reg::ArgReg<'_, A3>: From<MountFlagsArg> is not satisfied
--> /usr/local/cargo/registry/src/index.crates.io-6f17d22bba15001f/rustix-0.38.5/src/backend/linux_raw/arch/mod.rs:258:17
|
258 | $a3.into(),
| ^^^^ the trait From<MountFlagsArg> is not implemented for reg::ArgReg<'_, A3>
|
::: /usr/local/cargo/registry/src/index.crates.io-6f17d22bba15001f/rustix-0.38.5/src/backend/linux_raw/mount/syscalls.rs:23:13
|
23 | ret(syscall_readonly!(
| ______-
24 | | NR_mount,
25 | | source,
26 | | target,
... |
29 | | data
30 | | ))
| |- in this macro invocation
|
= help: the following other types implement trait From<T>:
<reg::ArgReg<'a, Num> as From<&'a CStr>>
<reg::ArgReg<'a, Num> as From<&'a mut MaybeUninit>>
<reg::ArgReg<'a, Num> as From<&'a mut [MaybeUninit]>>
<reg::ArgReg<'a, Num> as From<(backend::fs::types::Mode, backend::fs::types::FileType)>>
<reg::ArgReg<'a, Num> as From<*const T>>
<reg::ArgReg<'a, Num> as From<*mut T>>
<reg::ArgReg<'a, Num> as From>
<reg::ArgReg<'a, Num> as From<BorrowedFd<'a>>>
and 18 others
= note: required for MountFlagsArg to implement Into<reg::ArgReg<'_, A3>>
= note: this error originates in the macro syscall_readonly (in Nightly builds, run with -Z macro-backtrace for more info)
Compiling block-buffer v0.10.4
Compiling term_size v0.3.2
Specifically on cargo install just. Not sure why that would intermittently fail.
This was a bug in rustix 0.38.5, which is now fixed in rustix 0.35.6.
🙏 Really appreciate the info and getting a fix out quickly.
|
2025-04-01T06:37:01.681803
| 2015-10-21T13:01:02
|
112582689
|
{
"authors": [
"Glavin001",
"kasik96"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1128",
"repo": "Glavin001/atom-beautify",
"url": "https://github.com/Glavin001/atom-beautify/issues/615"
}
|
gharchive/issue
|
Code Folding error
When i have folded some code. And if press Beautify, folding open its self and doesnt come back to closed. It will be handfull after Beautify ends, load the folding. :)
Duplicate of #116
|
2025-04-01T06:37:01.693111
| 2015-07-30T07:06:54
|
98100278
|
{
"authors": [
"Glavin001",
"jhansche",
"meustice"
],
"license": "mit",
"license_source": "bigquery",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1129",
"repo": "Glavin001/atom-beautify",
"url": "https://github.com/Glavin001/atom-beautify/pull/488"
}
|
gharchive/pull-request
|
Add Gherkin grammar support
This resolves Glavin001/atom-beautify#377
Uses the https://github.com/cucumber/gherkin/tree/master/js node.js
package to use the official Lexer class.
The only known issue I've seen so far is that it does not format
tables properly -- the columns are not resized to match the widest
cell in each column.
@Glavin001 When can we expect v0.29.0 with Gherkin support? very excited about this update! :+1:
This is great! Thank you for this Pull Request. Only thing that is missing is tests: place original and expected files in a gherkin directory inside of examples/nested-jsbeautifyrc/ directory: https://github.com/Glavin001/atom-beautify/tree/master/examples/nested-jsbeautifyrc
Thanks again!
@meustice I hope to merge this soon. I have been set back with v0.29.0 because of work and I do not have the time that I would like to review all of the Pull Requests properly and add new features under the v0.29.0 milestone. Hopefully this week goes well and I have a little bit of time this weekend to review and merge everything.
If the tests are all there, I would like to merge all of the currently open Pull Requests and then make a new release.
Thank you for your patience!
@Glavin001 thanks, I'll see if I can get around to writing tests for this, and I'll make the change for debug_lexer also. The main reason I didn't want it to run initially is that it could slow things down if it's executed when it's not needed. Is there a Logger method to check if a certain type is enabled, before doing the rest of the work? I.e., only do the work if log level is verbose?
Is there a Logger method to check if a certain type is enabled, before doing the rest of the work? I.e., only do the work if log level is verbose?
Good idea. You could probably do something like:
loggerLevel = atom?.config.get('atom-beautify._loggerLevel')
if loggerLevel is 'verbose':
# Log stuff here
I'll see if I can get around to writing tests for this, and I'll make the change for debug_lexer also.
The tests should be very easy: an example file with some styling problems in the original directory, and then the correctly styled output in the expected directory. You can disable a test by adding _ to the front of the filename inside of original. The _ prefix in the original example file will cause it to be ignored / skipped. So if you could add a completely working test and then maybe a disabled test for the tables, then that would be great! Should only take a few minutes. Just something rough. Thanks!
@Glavin001 Are the existing test failures expected? When I run build-package.sh, I get a bunch of failures:
Finished in 4.637 seconds
55 tests, 395 assertions, 201 failures, 0 skipped
@Glavin001 I ran into an issue with Atom automatically inserting a newline at the end of the file upon saving (due to the whitespace package). This was causing problems because atom-beautify appears to specifically strip the trailing newlines from resolved string, as even trying to explicitly add a call to @write_blank(), or concatenating \n to the resolved string, and the test continued to fail due to the trailing newline in the expected output.
The only way to get the tests to pass was by turning off the "ensure newline" option from the whitespace package, so that I could delete the trailing newline from the original and expected files. Then it would work.
Is that a known issue? Anything I should have done differently to account for that?
The only way to get the tests to pass was by turning off the "ensure newline" option from the whitespace package, so that I could delete the trailing newline from the original and expected files. Then it would work.
Is that a known issue? Anything I should have done differently to account for that?
Nope, you are correct. Disabling the whitespace package, at least temporarily, if also what I do. Alternatively you could use the Right-Click context menu item for the file in the file tree called Beautify File which will not trigger the whitespace package.
Are the existing test failures expected? For instance, to run the tests for beautifying PHP code Atom Beautify requires PHP-CS-Fixer, which I assume you may not have installed. As long as your tests for Gherkin pass, then I can review this and still merge it. Thanks.
Published to v0.28.9
|
2025-04-01T06:37:01.710221
| 2023-10-06T20:52:44
|
1930919628
|
{
"authors": [
"rsglobal"
],
"license": "Apache-2.0",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1130",
"repo": "GloDroidCommunity/raspberry-pi",
"url": "https://github.com/GloDroidCommunity/raspberry-pi/pull/30"
}
|
gharchive/pull-request
|
Migrate to AOSPv14
Boots to UI.
The same issue here: https://github.com/waydroid/waydroid/issues/1070
We also have this issue:
https://github.com/raspberry-vanilla/android_local_manifest/issues/29
I ended up reverting wpa_supplicant to A13.
Ready for merging.
|
2025-04-01T06:37:01.775596
| 2024-02-29T14:18:32
|
2161375856
|
{
"authors": [
"BrokAnkle",
"Malcolmnixon"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1131",
"repo": "GodotVR/godot-xr-tools",
"url": "https://github.com/GodotVR/godot-xr-tools/issues/618"
}
|
gharchive/issue
|
button_pressed and button_released signals of InteractableAreaButton always generate errors
The button_pressed and button_released signals send this error when emited:
interactable_area_button.gd:87 @ _on_button_entered(): Error calling from signal 'button_pressed' to callable: 'Node3D(Button.gd)::on_button_pressed': Cannot convert argument 1 from Object to NodePath.
<C++ Source> core/object/object.cpp:1140 @ emit_signalp()
<Stack Trace> interactable_area_button.gd:87 @ _on_button_entered()
By changing emit(self) by emit(get_path()) it resolve the problem but I don't know if this is correct.
Could you provide a bit more information. The complaint is indicating a mismatch with some "on_button_pressed" handler in a Button.gd file - could you provide the code of that callback function?
Yes sorry it was in fact my callback and not the signal the problem, for some reason I typed the "button" argument as a NodePath instead of a Variant. I was confused looking at the code that has a "button" variable wich is a NodePath.
|
2025-04-01T06:37:01.793960
| 2024-07-05T13:22:45
|
2392604763
|
{
"authors": [
"EvergreenTheTree",
"bwanshoom",
"luX0r-reload"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1132",
"repo": "Gogh-Co/Gogh",
"url": "https://github.com/Gogh-Co/Gogh/issues/450"
}
|
gharchive/issue
|
apply-colors.sh has a bug
Line 126 is currently this:
if [[ -z "${GS}"]] &&[[ -z "${DCONF}" ]] && [[ -z "${GCONF}" ]]; then
it should be this:
if [[ -z "${GS}" ]] && [[ -z "${DCONF}" ]] && [[ -z "${GCONF}" ]]; then
It's missing a couple spaces
Yes, Yesterday update break apply script.
On my Ubuntu workstation with gnome terminal:
/tmp/gogh.apply.EEXYdO: line 126: conditional binary operator expected
/tmp/gogh.apply.EEXYdO: line 126: syntax error near `-z'
/tmp/gogh.apply.EEXYdO: line 126: ` if [[ -z "${GS}"]] &&[[ -z "${DCONF}" ]] && [[ -z "${GCONF}" ]]; then'
Fixed in #449, should be closed.
|
2025-04-01T06:37:01.796398
| 2022-09-07T20:22:47
|
1365142059
|
{
"authors": [
"Gogo1951"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1133",
"repo": "Gogo1951/Groupie",
"url": "https://github.com/Gogo1951/Groupie/issues/86"
}
|
gharchive/issue
|
Bulletin Board Tab Selection Broken
Tabs aren't showing which is selected any more.
Also, PVP should come to the left of "Other" and "All"
Also we lost the "Send Spec" button in the lower right.
Also needs "Created" time and "Updated" time still as columns, default sort to Created oldest on top. Sorry I can't remember if that's in another ticket yet.
Seems to work now... maybe I had some bad data in the groups? Dunno.
Probably a Dupe of #87
|
2025-04-01T06:37:01.808307
| 2018-11-07T12:25:24
|
378266314
|
{
"authors": [
"ZerakPalani",
"ayberkanilatsiz",
"jskidd3"
],
"license": "MIT",
"license_source": "github-api",
"license_type": "permissive",
"provenance": "gharchive-dolma-0000.json.gz:1134",
"repo": "GoldenOwlAsia/react-native-twitter-signin",
"url": "https://github.com/GoldenOwlAsia/react-native-twitter-signin/issues/119"
}
|
gharchive/issue
|
RNTwitterSignIn.logIn() sometimes requires authorising twice
When invoking RNTwitterSignIn.init followed by RNTwitterSignIn.logIn(), most of the time only the screen on the left will show. I press connect, and this then dismisses the view and I handle the promise call using the authToken and authTokenSecret.
However, I would say roughly 1 in 3 times, the screen on the right will show immediately after pressing connect and the screen dismissing.
Why is this happening? The screen on the left is from the Twitter app. I am logged in and as far as I know once I hit connect that should finish the process. The package should just pass the relevant info through the promise and that be the end. Why does it sometimes load this second window afterwards prompting login?
Thanks in advance.
@jskidd3 @nabylb Did you find any solution? I've same issue.
If I remember rightly the Twitter login SDK is deprecated so we pulled
support from our app. 👍
On Thu, 24 Oct 2019 at 13:39, Ayberk<EMAIL_ADDRESS>wrote:
@jskidd3 https://github.com/jskidd3 @nabylb https://github.com/nabylb
Did you find any solution? I've same issue.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
https://github.com/GoldenOwlAsia/react-native-twitter-signin/issues/119?email_source=notifications&email_token=AAP6LL3RE73DE7TDV3IQYL3QQGJPJA5CNFSM4GCJKTPKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOECE36CI#issuecomment-545898249,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/AAP6LL4QSXVZG364M2KB6STQQGJPJANCNFSM4GCJKTPA
.
--
Regards,
Joel.
@GoldenowlConsultingCompany
Having same issue. Why does this even happen? Like the first auth through Twitter app doesn't even matter. You can just close it and then the browser version is opened and that's the one that matters (if auth or decline). Also why do I have to authorize everytime? Shouldn't I just be logged in already everytime the Twitter auth screen opens? I am writing my Twitter username & password everytime I Want to login with Twitter for some reasons.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.