Open
Conversation
MonoGame.Packaging.Flatpak is a package I wrote a while ago to quickly package up a MonoGame game into a flatpak package for Linux during `dotnet pack` process. However its long since been broken and I don't see any benefit in keeping it around in its current state.
…8636) Changed the name of the **save** option from 'Save...' to 'Save' and changed 'Save As' to 'Save As...' under the **file** menu of the menu strip of the MGCB Editor for convention. As **save as** opens a dialog while **save** does not. --------- Co-authored-by: Simon (Darkside) Jackson <darkside@zenithmoon.com>
Fixes a documentation typo. ### Description of Change This is a very simple change that fixes a single typo in the `MonoGame.Framework.Content.Pipeline/Processors/ModelProcessor.cs` file. Before: `MipMpas` After: `MipMaps` Co-authored-by: Simon (Darkside) Jackson <darkside@zenithmoon.com>
Requires MonoGame/monogame.github.io#198 to fix the Linux builds
Reference specific commit for the Vulkan setup, so things build. We'll need to update this once the setup-vulkan-sdk 1.2.1 release it official. As per temporary suggested fix there - humbletim/setup-vulkan-sdk#28 (comment)
Wine 10.0 removed the wine64 executable and now only has wine. This change detects which one is available and uses it. It also fixes an issue where the working directory could not be set. ``` An error occurred trying to start process 'wine64' with working directory '/Users/runner/work/MonoGame/MonoGame/Artifacts/Tests/Tools/Release'. ``` We do this by providing an actual working directory which we know exists. This is the location of the WineHelper executable.
I started addressing #8104 and those changes should make MonoGame mostly safe in AOT contexts (it fixes all the errors of loading assets and removes the need to force MonoGame to not be trimmed). Working on this actually made visible more AOT issues. I fixed the most trivial IL warnings but some of them persists. Also, since BRUTE hasn't fully retired yet, I'm trying to write this in a ```net452``` backward compatible way (some of the newer and better reflection stuff isn't available on older .NET Frameworks). Any help is welcome! Here are my notes on the remaining issues (none really are blockers, just enhancing the AOT experience and avoid misleading warnings for users): - [x] ```ReflectiveReader```, ```DictionaryReader```, ```ArrayReader```, and ```MultiArrayReader``` need to be flagged as not compatible with AOT with the ```RequiresDynamicCode``` attribute. Those types are generic and can't be made safe for AOT. The warning inside the attribute must encourage users to rewrite their content serialization in a way that doesn't rely on reflection is they wish to use AOT. - [ ] WindowsDX will [fail loading an embedded icon](https://github.com/MonoGame/MonoGame/blob/develop/MonoGame.Framework/Platform/Windows/WinFormsGameWindow.cs#L261) because ```assembly.Location``` is always null in AOT contexts. Those ones aren't really issues and will likely not trigger runtime errors. We could just disable warnings on those, but it would be nice to try to fix them in a clean way to avoid unforeseen consequences: - [ ] [VectorConversion](https://github.com/MonoGame/MonoGame/blob/develop/MonoGame.Framework/Design/VectorConversion.cs) uses reflection to convert to/from ```IPackedVector```. This needs to be rewritten reflection-free. We can also assume that ```IPackedVector``` is never trimmed and just disable the warnings, but that's an easy fix and likely less error-prone to not rely on disabling warnings. - [ ] There are a couple of instances of type instance creation that we should try to get rid of if possible: https://github.com/MonoGame/MonoGame/blob/2e72fcfa4144ea5bc413963af8f97a2d5e902338/MonoGame.Framework/Content/ContentReaders/MultiArrayReader.cs#L36 https://github.com/MonoGame/MonoGame/blob/2e72fcfa4144ea5bc413963af8f97a2d5e902338/MonoGame.Framework/Graphics/Vertices/VertexDeclaration.cs#L214 - [ ] ```ContentTypeReaderManager``` has a warning here (might be fixable): https://github.com/MonoGame/MonoGame/blob/2e72fcfa4144ea5bc413963af8f97a2d5e902338/MonoGame.Framework/Content/ContentTypeReaderManager.cs#L154 - [ ] ```ContentManager``` has 2 warnings here (can be disabled, this will likely never fail, and this is only useful for Android): https://github.com/MonoGame/MonoGame/blob/2e72fcfa4144ea5bc413963af8f97a2d5e902338/MonoGame.Framework/Content/ContentManager.cs#L592 - [ ] DesktopGL has ```VertexBuffer.GetBufferData()``` which might fail in AOT depending on the vertex declaration (we can't fix this, but it would be best to flag this method as incompatible, but like other warnings, this should actually be fine): https://github.com/MonoGame/MonoGame/blob/2e72fcfa4144ea5bc413963af8f97a2d5e902338/MonoGame.Framework/Platform/Graphics/Vertices/VertexBuffer.OpenGL.cs#L92
Partially addresses #8165 ### XML Documentation This PR aims to add new and update existing XML documentation to various public members with the intent to reduce build warnings
Fixes #8408 We'll need to update SDL to latest 2.x release after that for maximum profit.
…atform (#8660) Fixes #8657 FileDrop in a WindowsDX project cuts off the last character in the path ### Description of Change Adding a byte to the final call of DragQueryFile as it expects the final byte to be the destination of the null string terminator. Ie, '.png' is 4 chars, with buffSize = 4, result would be '.pn`null`' > [out] lpszFile > > Type: LPTSTR > > The address of a buffer that receives the file name of a dropped file when the function returns. This file name is a null-terminated string. If this parameter is NULL, DragQueryFile returns the required size, in characters, of this buffer. Co-authored-by: Simon (Darkside) Jackson <darkside@zenithmoon.com>
…trix (#8671) Fixes #8661 ### Description of Change Fixes `OpenAssetImporter.GetRelativeTransform` to now correctly return a row-major matrix. Appears to have been broken when #8607 removed the `ToXna(Matrix4x4 matrix)` method, resulting in the return of an incorrect column-major matrix. While this fixes the rendering of ShipGame models, I just realised other changes from that PR might have broken other things. I'll try some other 3d models and import options next I guess. --------- Co-authored-by: Simon (Darkside) Jackson <darkside@zenithmoon.com>
…8676) This commit removes an early return in the WineHelper.cs file to allow the process to retry with "wine" as the fallback executable instead of "wine64". Removed the return statement in the catch block so that after changing the executable to "wine", the program will continue to start the process. This change enables the intended retry logic for process execution.
### Description of Change This PR: a) fixes the typo docline in `ContentManager.cs` introduced in 7b84682#diff-37705868f9e9f6e1ee0f3b9f8d532d01f861d3e26a3050b30435ac545abb81ceR301 b) adds **empty** doclines for readers that were made public in a23eccd (this fixes all [CS1591](https://learn.microsoft.com/en-us/dotnet/csharp/language-reference/compiler-messages/cs1591) warnings in DesktopGL and partly in WindowsDX): `Content/ContentReaders/ArrayReader.cs` `Content/ContentReaders/DictionaryReader.cs` `Content/ContentReaders/EnumReader.cs` `Content/ContentReaders/ListReader.cs` `Content/ContentReaders/MultiArrayReader.cs` `Content/ContentReaders/NullableReader.cs` `Content/ContentReaders/ReflectiveReader.cs` Feel free to revert this change, if actual documentation is to be added to these files.
…itives() (#8580) Fixes #3125 ## Description of Change Added vertex size check to GraphicsDevice.DrawUserPrimitives() Now method throws an exception when the size of a vertexDeclaration.VertexStride is less than ReflectionHelpers.FastSizeOf<T>() as suggested here: #3125 (comment)
<!-- Are you targeting develop? All PRs should target the develop branch unless otherwise noted. --> Fixes #6298 <!-- Please try to make sure that there is a bug logged for the issue being fixed if one is not present. Especially for issues which are complex. The bug should describe the problem and how to reproduce it. --> ### Description of Change Implemented the ContainsPrecise method for BoundingBox, as suggested by Member (#8564) - ContainsPrecise performs an exact containment check between a BoundingBox and BoundingFrustum, avoiding false positives - Implemented 6 unit tests covering various scenarios for the ContainsPrecise method - Tests include cases for full containment (1), intersection (3), and disjoint (2) relationships <!-- Enter description of the fix in this section. Please be as descriptive as possible, future contributors will need to know *why* these changes are being made. For inspiration review the commit/PR history in the MonoGame repository. -->
Comit 5595e50 added changes to the working directory that caused a problem when compiling effects. So lets back out that change. ``` The input file '/Users/xxx/.nuget/packages/dotnet-mgcb/3.8.2.1922-develop/tools/net8.0/any/shaders/NormalMapping.fx' was not found! ```
For some reason, ```ConsoleCheck``` has been unplugged from the CI and as a result some PRs introduced incompatibilities with .NET 4.5 / C# 5. This PR fixes the incompatibilities and restores ```ConsoleCheck```. For now, the ```ConsoleCheck``` only tests against the current code architecture. It doesn't test the newer ```Native``` structure but some consoles have already been ported to that structure, so we need something like ```ConsoleCheck.Native```. Additionnaly, it would be necessary to have the ```DesktopGL``` tests to run using ```PublishAot``` to make sure that no PR uses runtime reflection (which we can't test with building ```ConsoleCheck```) to avoid regressions on the ```PublishAot``` compatibility. ----- If anyone wonders "why is this limitation still in place? didn't you change the console runtime?", the answer is: yes and no. Yes, the consoles have been moved toward using ```PublishAot``` instead of BRUTE. All but one, because ```PublishAot``` can't be used there. Which means that there is one single platform currently limiting the entirety of MonoGame. We're trying to figuring it out, but it's a process we don't have a power over and involves bureaucracy (or possibly to rewrite entirely BRUTE). So no ETA on fixing this, but it's expected to take a very long time.
This PR updates DesktopGL to SDL 2.32.2
…-platform apps. (#8663) Fixes #8664 ### Description of Change Moved the SensorData classes up to the top level (still don't like that the namespaces are `Microsoft.Devices.*`), then refactored the Platform classes to remove the iOS IFDEF, following more consistent patterns in the codebase. Then added the classes to the non supported classes (the throw `PlatformNotSupported` exceptions if called), but this now allows game devs to use Accelerometer and Compass classes in a more cross-platform way. Since Microsoft has pulled the WinPhone, in MG 4, we should consider renaming the namespace to `MonoGame.Devices.*`, IMHO.
Fixes #8683 ### Description of Change Fixes importing contentproj files that reference files outside the project directory. Visual Studio puts a `<Link>` in a `<Content>`, `<Compile>` or `<None>` element to show the desired position in the structure, which MGCB was not respecting.
As Per subject line. Something I missed with the previous PR 🤦♂️ .
Hi! First time contributing, please be gentle! I ran into a problem where I couldn't set the Pitch of a sound effect instance below `-1.0f` or above `1.0f` without throwing an exception. This is on purpose, according to the docs, but 1 octave up/down is really limiting! I saw discussion of it in [this ancient PR](#4852) and it seems there's no technical reason for this limitation. The best justification I can imagine is "That's how XNA did it." (I have no evidence of this, total guess). My proposal is pretty straightforward: we just get rid of the range check and let AL handle it. I'm not changing any of the math for backwards compatibility, plus 1.0 == 1 octave is actually pretty elegant, I just wish I could have 2.0, or 10.0! I'd be happy to add/modify a test for this, but I wasn't able to run the unit tests because I can't get Monogame to build on my machine (attempting to build via Visual Studio 2019 using `MonoGame.Framework.DesktopGL.sln` and/or `MonoGame.Tools.Windows.sln`).
This pull request is to fix DynamicSoundEffectInstance requesting buffers too late. More details on the issue can be found here: #7565 For demonstration purposes please see the OpenALPendingBufferCount project within here: [https://github.com/squarebananas/MonoGameSamplesForIssues](https://github.com/squarebananas/MonoGameSamplesForIssues) --------- Co-authored-by: Simon (Darkside) Jackson <darkside@zenithmoon.com>
Prevents a "Skipping XYZ..." message for each asset which is _not_ converted during the content build process. On projects with a lot of content, this creates a lot of noise. I left the normal output message enabled since the conversion can be slow, and it gives the user indication of what is happening. The "Skipping XYZ..." messages just tend to drown out everything else.
…8690) Just a quick one - saw a few trivial warnings when building in debug that were due to invalid XMLDOC comments. Thought I'd fix them up.
Updating the csproj files for the solution with README references fo r the NuGet packages to make them clearer, including links to required destinations such as issue fixing - resolves: #8691
Stuff done: - Split out the base class logger from the new content builder logger - Hooked up the `LogLevel` parameter to the content logger - Hooked up the `LoggerRootDirectory` so we got some proper pathing showing up - Reworked the default appearance of the log thats: - Friendly for msbuild to parse the warnings / errors from - Friendly appearance for threaded logging if we ever get to that - Honestly even better for looking at for me :D, technically if anyone liked the previous one they can just assign the `ContentBuilder.Logger` property to a custom made logger Bellow are some examples on appearance based on log level. Debug: <img width="648" height="398" alt="debug" src="https://github.com/user-attachments/assets/534cb671-5616-438b-b005-4db79b2bce52" /> Info: <img width="532" height="187" alt="info" src="https://github.com/user-attachments/assets/0cd5da1d-7fb9-47ae-92f0-2c38b00c874d" /> Warning: <img width="397" height="101" alt="warning" src="https://github.com/user-attachments/assets/3393db82-0186-44d7-ab22-bc6250b27d43" />
We renamed `WindowsGDK` `TargetPlatform` and `MonoGamePlatform` to `WindowsDX12`. Also fixed shader, texture, and audio profiles for `WindowsDX12`.
# Summary Updating license year for 2026 and a revision of published docs in the repository
# Summary Update templates to use the 3.8.5-preview releases for CB templates
More Fixes going all the way back to #[8101](#8101) ### Description of Change Updates remaining Headers with the previous **The MonoGame Team** copyright to match the new Foundation Copyright standard. I happened to be playing around with the newer native implementations and noticed the CType generators had the older copyright text and then I went down this rabbit hole of finding more and more. I apologize for the big PR! It really is just comment updates with the exception of these files that have impactful changes to the comments created on newly generated CTypes files: Tools/MonoGame.Generator.CTypes/EnumWritter.cs Tools/MonoGame.Generator.CTypes/PInvokeWritter.cs Tools/MonoGame.Generator.CTypes/StructWritter.cs
I **suspect** this is supposed to fix the following issues (the circumstances when this reproduces are a bit different for me, but the stack is similar to what #9164 has reported; I suspect they all are relatively about the same, and it would make sense to at least ask the folks to re-check their issues after merging this fix): - fix #9164 - fix #9161 - fix #9156 ### Description of Change Previously, the pointers `lastFeature` and `deviceFeatures2` could outlive the variables they are pointing to. These pointers are then exposed as part of the `deviceCreateInfo` — which could cause crash if the stack memory is reused for any purpose. The problem was reproduced for me when debugging a MonoGame program by Rider on Windows, but not under any other circumstances (e.g. debugging via VS or just running the program — no luck, doesn't reproduce). Also note that it was only reproducing if using the **Release** build of `mgruntime.dll` — not the **Debug** one. Rebuilt MonoGame locally, can confirm I have stable repro before the fix (on the official binaries and on the ones I build myself), and after the fix the issue no longer reproduces. --------- Co-authored-by: Simon (Darkside) Jackson <darkside@zenithmoon.com>
This is the initial implementation of Song for the native backend including a unit test as well.
Added a contributor declaration to certify no LLMs were used for the contributed code.
## Description Fixes two bugs in `Color.FromHSV()` 1. Fixes fractional hue calculation within color wheel segments. The fractional position `f` within each 60deg segment was calcualted as `int f = (int)(h / 60.0f) - i`, which always resulted in a `0` due to integer arithmetic. Changed to `float f = (h % 60.0f) / 60.0f` to correctly calcualte fractional position within [0,1] range. 2. Removed incorrect normalization of saturation and value parameters. The documentation for the method states that these values should be in the [0,1] range. By normalizing them by dividing by 100, this causes colors to be nearly black in all cases (e.g. passing `s=1.0, v=1.0` would result in RGB values of `2` instead of `255`) Additionally, `Color.FromHSV()`, `Color.FromHSL()` were changed to `static` methods. These are factory methods for creating a `Color` value from a different structure, it doesn't make sense to have them as instance members that require an instance before using. ## Breaking Changes Since `Color.FromHSV()` and `Color.FromHSL()` are changed to static factory methods, this would be considered a breaking change as it changes the public API that already existed. Not sure if you want to keep this change since it's technically a breaking change or we can obsolete the existing ones with a warning and forward them to the new static ones, then remove the instance methods in a future release where breaking changes are good. ## Related Issues - Closes #9131
# Summary Cleanup for the builder `.targets`, does not need exception for windows for launching the builder project. Cleaned up
Disable VSCode based CoPilot features so that contributors who need to use CoPilot for work or other projects do not need to disable it manually. ### Contributor Declaration - [x] I certify that no LLM's were used to generate any code or documentation in this contribution.
Add a new `FontImporter` in which you can directly specify the properties that would otherwise go into a .spritefont file. ### Contributor Declaration - [x] I certify that no LLM's were used to generate any code or documentation in this contribution.
Fixes #9011 ### Description of Change The `StringBuilder` in `HandleDropMessage` was allocated with a buffer size of `buffSize`, but the `cch` parameter passed to `DragQueryFile` is `buffSize + 1` (to account for the null terminator). This mismatch could cause the last character of file paths to be truncated on some configurations, as reported in the issue. The fix ensures the `StringBuilder` capacity matches the expected buffer size passed to the Win32 API. **Before:** ```csharp StringBuilder builder = new StringBuilder((int)buffSize); DragQueryFile(hdrop, i, builder, buffSize + 1); ``` **After:** ```csharp StringBuilder builder = new StringBuilder((int)buffSize + 1); DragQueryFile(hdrop, i, builder, buffSize + 1); ``` Co-authored-by: Tom Spilman <tom@sickheadgames.com>
Fixes #9189 ### Description of Change Takes the same logic used in Load and handles converting \\\ to the / format when trying to UnloadAssets. Verified that it works using a simple Repro project based on the provided code in the issue. ### Contributor Declaration - [x] I certify that no LLM's were used to generate any code or documentation in this contribution. Co-authored-by: Jake Kirkbride <jakelegendxiii@rjtechnologies.com>
- Fix `TokenType` to be internal to pipeline. - Updated CppNet submodule for doc fixes. ### Contributor Declaration - [x] I certify that no LLM's were used to generate any code or documentation in this contribution.
Also added some improved docs for FrameworkDispatcher. ### Contributor Declaration - [x] I certify that no LLM's were used to generate any code or documentation in this contribution.
- Validate controllers have valid gamepad profiles before reporting as connected - Fix GetCapabilities() to set IsConnected = true when profiles exist - Applies to both iOS and tvOS implementations Addresses #9099 ### Description of Change Adds defensive validation to iOS/tvOS gamepad detection in response to reports of phantom controllers appearing in iOS 18. ## Changes Made 1. **Profile validation in `PlatformGetState()`**: Before reporting a controller as connected, verify it has a valid gamepad profile (ExtendedGamepad, Gamepad, or MicroGamepad for tvOS). 2. **`GetCapabilities()`**: Now sets `IsConnected = true` when a controller has a valid profile. ## Rationale Users report phantom gamepads appearing on iOS 18, even when no gamepad is connected. Disabling Bluetooth or unpairing devices resolves it, suggesting Apple's `GCController.Controllers` array may include paired-but-disconnected controllers or misclassified Bluetooth devices. This validation ensures only controllers with actual gamepad input profiles are reported as connected by adding a defensive layer without breaking legitimate controller support. ## Testing Notes - Non-breaking change—legitimate controllers with valid profiles continue to work - Filters devices without gamepad profiles (keyboards, mice, disconnected controllers)
### Description of Change Refactored Matrix.CreateBillboard for better readability. Renamed local variables: vector -> cameraDir vector2 -> up vector3 -> right
…ementations to avoid confusion. (#9203) # Summary It was repoted there was confusion regardind the description of `Keys.None` as it was different to other implementations. On review, this has now been updated to "No Key Set" which aligns with all other "None" enum definitions, including `Buttons.None` Resolves - #7826
Fixes #9213 Lets pin xcode for 26.0.1 to fix this error ``` error : This version of .NET for iOS (26.0.9783) requires Xcode 26.0. The current version of Xcode is 26.2. Either install Xcode 26.0, or use a different version of .NET for iOS. See https://aka.ms/xcode-requirement for more information. ``` 26.0.1 is already installed on the macos images, we just need to select it until its no longer available. ### Contributor Declaration - [x] I certify that no LLM's were used to generate any code or documentation in this contribution.
<!-- Are you targeting develop? All PRs should target the develop branch unless otherwise noted. --> Fixes #9205 <!-- Please try to make sure that there is a bug logged for the issue being fixed if one is not present. Especially for issues which are complex. The bug should describe the problem and how to reproduce it. --> ### Description of Change This adds `SaveAsPngShouldWork` and `SaveAsJpegShouldWork` to Texture2DTests.cs, plus some static helpers for building test textures. Both tests rely on `Texture2D.FromStream`. They make a sample texture, save to a `MemoryStream` with the method being tested, and then compare the sample against the result of `Texture2D.FromStream` on the same memory stream. The PNG tests require all pixels to be identical between the two images, and the JPEG tests require that all result pixels be within 35 of the source on R, G, and B channels. Alpha is not tested for JPEG. 35 was arbitrary to get my test cases to pass so I won't lose any sleep over it if someone who understands JPEG better than me wants to improve the result comparison. <!-- Enter description of the fix in this section. Please be as descriptive as possible, future contributors will need to know *why* these changes are being made. For inspiration review the commit/PR history in the MonoGame repository. --> ### Contributor Declaration - [X] I certify that no LLM's were used to generate any code or documentation in this contribution. Co-authored-by: Tom Spilman <tom@sickheadgames.com>
#9220) Fixes #9212 ### Description of Change This PR addresses an issue on the Vulkan where textures that were 2D arrays were getting mapped to `VK_IMAGE_VIEW_TYPE_2D` instead of `VK_IMAGE_VIEW_TYPE_2D_ARRAY`. On MacOS, this mismatch causes an application crash. The approach chosen was to add an additional optional `layerCount` parameter to `ToVkImageViewType` that can be used to distinguish between 2D and 2D Array types. Note that there is a similarly named `ToVkImageType` that performs a conversion between `MGTextureType` and Vulkan constants. It does not include this new switching logic because it doesn't seem to affect the MacOS runtime problem which is what motivated the creation of #9212. ### Contributor Declaration - [x] I certify that no LLM's were used to generate any code or documentation in this contribution. Co-authored-by: Adam Carter <adamcarter@umpquabank.com>
Fixes #9206. ### Description of Change This changes [ RenderTarget3D.cs](https://github.com/MonoGame/MonoGame/blob/6058c786ee5b30d9ab91135edd4df29acaefc52c/MonoGame.Framework/Graphics/RenderTarget3D.cs) to add XML comments for the class. ### Contributor Declaration - [x] I certify that no LLM's were used to generate any code or documentation in this contribution.
…ng units from KHz to Hz where appropriate. (#9238) ### Description of Change This pull request changes the error string for `WavImporter`'s `InvalidContentException` when a `.wav` file outside of the supported sample rate range is supplied. Previously, this error string incorrectly displayed the offending file's sample rate in `Hz` yet labeled it as `KHz`. This pull request changes the unit label to `Hz` and, for the sake of consistency, further changes the part of the error string that informs the user of the 8KHz-48KHz bounds to be displayed in `Hz`. ### Contributor Declaration - [x] I certify that no LLM's were used to generate any code or documentation in this contribution.
Pin Xcode to use 26.2 to fix the CI builds ### Contributor Declaration - [x] I certify that no LLM's were used to generate any code or documentation in this contribution.
Turns out we were missing the float * Matrix operator on the Matrix class all these years. We have Matrix * float, but not the other. This adds it. https://learn.microsoft.com/en-us/previous-versions/windows/xna/bb198123(v=xnagamestudio.40) ### Contributor Declaration - [x] I certify that no LLM's were used to generate any code or documentation in this contribution.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
20 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
No description provided.