status
stringclasses 1
value | repo_name
stringclasses 13
values | repo_url
stringclasses 13
values | issue_id
int64 1
104k
| updated_files
stringlengths 11
1.76k
| title
stringlengths 4
369
| body
stringlengths 0
254k
⌀ | issue_url
stringlengths 38
55
| pull_url
stringlengths 38
53
| before_fix_sha
stringlengths 40
40
| after_fix_sha
stringlengths 40
40
| report_datetime
unknown | language
stringclasses 5
values | commit_datetime
unknown |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
closed | godotengine/godot | https://github.com/godotengine/godot | 80,751 | ["servers/rendering/renderer_viewport.cpp"] | Debanding stops applying when the project restarts or TAA is toggled | ### Godot version
4.2 dev 3
### System information
Windows 11, i3-10105f, RX6600 23.7.2, Forward+
### Issue description
Debanding doesn't seem to apply when the Project is loaded. Both in Editor and Running Project.
Haven't found any other settings that seem to behave like that (Soft Shadows, Shadow Resolution, MSAA, FXAA, TAA...)
### Steps to reproduce
Turn Debanding On
Debanding is Applied
Save
Close Editor
Start Editor
Load Project
Debanding is not Applied but Project settings state On
Turn it off and on again
Debanding is Applied
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/80751 | https://github.com/godotengine/godot/pull/82766 | 7a9947f7f52acdb639682f1169a3a9b5d2ba1c56 | ccbb91137f94e62d35b715fd3aea3ec30a29d680 | "2023-08-18T09:53:43Z" | c++ | "2023-10-04T13:45:39Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,750 | ["editor/gui/editor_file_dialog.cpp"] | `FileDialog` considers `/path/to/dir/`'s parent directory to be `path/to/dir` | ### Godot version
4.1.1
### System information
linux
### Issue description
- When favoriting a path in a `FileDialog`, clicking on it travels to `path/to/dir/`.
- When going up a level from there, it doesn't go to `path/to` but to `path/to/dir` (without the `/`).
[simplescreenrecorder-2023-08-18_09.38.37.webm](https://github.com/godotengine/godot/assets/71937306/00c20ad3-292f-4961-84ed-dbc64906df52)
The problem is either (or both) that:
- loaded path has a trailing `/`
- parent dir of `stuff/` is considered to be `stuff`
### Steps to reproduce
favorite a directory and use the feature to navigate to it
### Minimal reproduction project
any (i.e. the project manager's "new project" dialog has one) | https://github.com/godotengine/godot/issues/80750 | https://github.com/godotengine/godot/pull/80821 | 97f3f97cc01205f329b10c1f5103ba2aeb1c6944 | 88269cfb865077a0af3ea2f93a2584910e4a41f7 | "2023-08-18T07:41:29Z" | c++ | "2023-09-07T11:50:55Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,748 | ["editor/editor_node.cpp"] | Dock windows are empty if they were hidden when loading a layout that make them floating | ### Godot version
4.1.1
### System information
linux
### Issue description
If a dock isn't visible in the editor (in an unshown tab), loading an editor layout that undocks it will result in this behavior:
[simplescreenrecorder-2023-08-18_08.58.08.webm](https://github.com/godotengine/godot/assets/71937306/08aea47c-e948-4345-9943-71fc0bbadd3e)
This also happen *sometimes* when the docks are visible before undocking, but I can't reproduce it :upside_down_face:
### Steps to reproduce
See video above:
1. save a layout with a floating dock,
2. restore the default layout or manually close the dock window,
3. hide this particular dock,
4. load the previously created layout.
### Minimal reproduction project
Here's the editor layout I use in the clip, though it needs to be recreated locally because of monitors and coordinates etc.:
<details>
append to `~/.config/godot/editor_layouts.cfg`
```
[test]
dock_1_selected_tab_idx=0
dock_2_selected_tab_idx=0
dock_3_selected_tab_idx=0
dock_4_selected_tab_idx=0
dock_5_selected_tab_idx=0
dock_6_selected_tab_idx=0
dock_7_selected_tab_idx=0
dock_8_selected_tab_idx=0
dock_floating={
"FileSystem": {
"window_rect": Rect2i(2669, 319, 228, 302),
"window_screen": 0,
"window_screen_rect": Rect2i(1990, 27, 2490, 1413)
},
"Inspector": {
"window_rect": Rect2i(2927, 323, 270, 300),
"window_screen": 0,
"window_screen_rect": Rect2i(1990, 27, 2490, 1413)
},
"Scene": {
"window_rect": Rect2i(2388, 315, 252, 300),
"window_screen": 0,
"window_screen_rect": Rect2i(1990, 27, 2490, 1413)
}
}
dock_split_2=0
dock_hsplit_1=0
dock_hsplit_2=337
dock_hsplit_3=-337
dock_hsplit_4=0
dock_filesystem_split=0
dock_filesystem_display_mode=0
dock_filesystem_file_sort=0
dock_filesystem_file_list_display_mode=1
dock_filesystem_selected_paths=PackedStringArray("res://")
dock_filesystem_uncollapsed_paths=PackedStringArray("res://")
dock_3="Import,Scene"
dock_5="Node,History,Inspector"
dock_4="FileSystem"
dock_split_3=0
```
</details> | https://github.com/godotengine/godot/issues/80748 | https://github.com/godotengine/godot/pull/82742 | 85e508b5700382809c5bdbf406914663177aa7d3 | 230bdda872bd01a776f30f2c15ab76b354b6e900 | "2023-08-18T07:09:37Z" | c++ | "2023-10-13T09:43:35Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,744 | ["scene/gui/file_dialog.cpp"] | FileDialog automatically selects first folder in Open Folder mode | ### Godot version
4.1.1
### System information
Windows 10
### Issue description
When using a FileDialog in Open Folder mode to browse and select a folder, the file/directory browser seems to select the first directory in the list by default. This is confusing because if you double click a folder to enter it and then click "select this folder" it will actually select the first folder in the list (which is now selected by default) instead of the one you just entered. In Godot 3.x, there was no folder selected in the list until explicitly clicked on.
This appears to have the same underlying cause as #62890
### Steps to reproduce
* Run a scene containing a FileDialog node which has set its File Mode to Open Folder and the Access to File System to ensure you will have access to folders with subdirectories in them.
* Double click to select and enter a directory.
* Observe that the first directory in the list is selected.
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/80744 | https://github.com/godotengine/godot/pull/81034 | 55282ddc106cb0b2fd8750e710fbd49efab5f56a | 49cdd330be770cbc25e126d0020d5496a76022d9 | "2023-08-18T02:34:07Z" | c++ | "2023-10-10T20:49:16Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,734 | ["modules/gdscript/editor/gdscript_docgen.cpp"] | In Custom Class Documentation a Function cannot return 'Variant' | ### Godot version
v4.1.1.stable.official [bd6af8e0e]
### System information
Windows 10
### Issue description
As said in the Title Custom Class Documentation a Function cannot return 'Variant' and will instead be shown to return void, even though Built-in Classes are shown to return 'Variant', an example would be 'Array'
### Steps to reproduce
```gdscript
extends Node
class_name CustomClass
func VariantFunction() -> Variant:
return 10
```
Now when checking the Generated Documentation it will appear like: `void VariantFunction()`
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/80734 | https://github.com/godotengine/godot/pull/80745 | 56dd0edf588fb72f38c02d81dfda08bc3df8e313 | 9becff0c54ffb5aee65d0abef6443cdf2b01dfbb | "2023-08-17T19:20:23Z" | c++ | "2023-08-21T06:22:31Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,728 | ["modules/gdscript/gdscript_analyzer.cpp", "modules/gdscript/tests/scripts/analyzer/errors/engine_singleton_instantiate.gd", "modules/gdscript/tests/scripts/analyzer/errors/engine_singleton_instantiate.out", "modules/gdscript/tests/scripts/analyzer/errors/extend_engine_singleton.gd", "modules/gdscript/tests/scripts/analyzer/errors/extend_engine_singleton.out"] | Misleading error message when attempting to create an instance of a singleton | ### Godot version
0511f9d9a7d56c742d87fafdcea8785d40ad14b3 4.2.dev (current master as of writing)
### System information
macOS 13.4.1 (this issue affects all systems)
### Issue description
When trying to construct a built-in singleton (ex: `DisplayServer.new()`), this error message is misleading:
<img width="800" alt="Screenshot 2023-08-17 at 11 30 46 AM" src="https://github.com/godotengine/godot/assets/1646875/74965bb3-ba26-43dc-8b1b-12046551f421">
The error message says it's abstract, which seems like it should indicate that you need to use a derived class. However, this is a singleton, and you are meant to use its methods directly. Instead of the current error text, it should say something about how it's a singleton and you should use its methods directly.
### Steps to reproduce
Write `DisplayServer.new()` in the script editor.
### Minimal reproduction project
Since it's one line I am not going to bother making a MRP. | https://github.com/godotengine/godot/issues/80728 | https://github.com/godotengine/godot/pull/82098 | f795e458680ce5d7879e5556f8606aa96a6832a8 | 1c10ff37b34e557afe0da0bdd6f7ae180201a533 | "2023-08-17T16:37:28Z" | c++ | "2023-09-22T20:05:34Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,718 | ["scene/2d/tile_map.cpp"] | filter not works in Tilemap.get_used_cells_by_id() | ### Godot version
4.2.dev3.official[013e8e3af]
### System information
Godot v4.2.dev3 - Arch Linux #1 SMP PREEMPT_DYNAMIC Fri, 11 Aug 2023 11:03:36 +0000 - X11 - GLES3 (Compatibility) - NVIDIA GeForce RTX 2060 (nvidia; 535.98) - Intel(R) Core(TM) i7-10875H CPU @ 2.30GHz (16 Threads)
### Issue description
`TileMap.get_used_cells_by_id()` seems not work as expected.
It seems that this method will return the same result as "TileMap.get_used_cells()" since Godot 4.2dev2.
The arguments (`source_id: int`, `atlas_coords: Vector2i`, `alternative_tile: int`) of `get_used_cells_by_id()` seems not make the filter working as expected. It would only return the same result as `get_used_cells()`, contains all the tiles on the layer.
### Steps to reproduce
1. create a tilemap scene
2. attach a tileset, that contains at least 2 tiles , paint some tiles.
3. calls Tilemap.get_used_cells_by_id(layer, source_id, atlas_coords,alternative_tile)
4. run it
### Minimal reproduction project
run tile_map.tscn by Godot4dev3 (or custom build)
[tilemap_debug.zip](https://github.com/godotengine/godot/files/12369631/tilemap_debug.zip)
| https://github.com/godotengine/godot/issues/80718 | https://github.com/godotengine/godot/pull/80729 | 8c0e08481d69b4114f46506c72acda044875c28b | 5f247652deb730b98468514236e8411d7228dfed | "2023-08-17T13:11:32Z" | c++ | "2023-08-17T22:42:25Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,707 | ["editor/scene_tree_dock.cpp"] | Dragging node in place creates useless undo action | ### Godot version
4.2 f7bc653
### System information
Windows 10.0.19045 - Vulkan (Forward+) - dedicated NVIDIA GeForce GTX 1060 (NVIDIA; 30.0.15.1403) - Intel(R) Core(TM) i7-7700HQ CPU @ 2.80GHz (8 Threads)
### Issue description
https://github.com/godotengine/godot/assets/2223172/1f014ed1-771e-4fc1-b889-b366b8a0397c
Nothing changed, but the scene was "modified".
### Steps to reproduce
1. Have at least 2 nodes on a scene
2. Drag one node on the position between the nodes (doesn't work if drop position is too close to the center)
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/80707 | https://github.com/godotengine/godot/pull/80817 | d80ce0c52a097c35bb4d441e7a9c0a76e6342ba4 | fc64dfedfbb2513e8c07de9d59d058b1683c850b | "2023-08-17T09:12:40Z" | c++ | "2023-10-10T13:06:06Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,685 | ["editor/plugins/gizmos/collision_shape_3d_gizmo_plugin.cpp"] | Position moves when resizing BoxShape3D using grab handles | ### Godot version
4.2.dev3
### System information
Windows 10
### Issue description
Testing 4.2 dev 3, when I resize a BoxShape3D (as seen in https://github.com/godotengine/godot/pull/71092) , the position of the BoxShape moves as well. If the parent of the BoxShape3D is located at (0,0,0), then the BoxShape resizes (but doesn't move) as expected.
If the parent of the BoxShape3D is not (0,0,0), resizing the shape sees the position change unexpectedly.
Resize works as expected if the parent node is at (0,0,0)
Resize works correctly if you edit the size values in the inspector, it's only grabbing the handles with the mouse that causes the position to change.
### Steps to reproduce
Create a new 3D scene.
Move the Node3D root to any location (eg. (1,0,1))
Create a CollisionShape3D child of the Node3D
Configure the CollisionShape3D as a BoxShape3D
Pull any of the grab handles to resize the Box
The Box will move as well as resize when it should only resize
### Minimal reproduction project
| https://github.com/godotengine/godot/issues/80685 | https://github.com/godotengine/godot/pull/80689 | c7356c795b3c8cec2453061299e85cb315a9a6a7 | a42f95f9e26fcd8be952b7db5cd85961114fe335 | "2023-08-16T13:39:31Z" | c++ | "2023-08-17T09:27:28Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,664 | ["editor/gui/editor_spin_slider.cpp", "scene/gui/spin_box.cpp"] | Some editor fields don't accept comma as a decimal separator | ### Godot version
v4.1.1.stable.official [bd6af8e0e]
### System information
Windows 10
### Issue description
Most numerical fields in editor like inspector properties and settings accept comma as input (eg. `5,5` gets parsed to `5.5`). However, there's some inconsistency in this support. Color picker doesn't support this and also Input Map's deadzone fields.
### Steps to reproduce
Open any color picker in inspector or Input Map settings, try to input value like `0,5` in it.
### Minimal reproduction project
Any project | https://github.com/godotengine/godot/issues/80664 | https://github.com/godotengine/godot/pull/80699 | 813cd1dfc8fb7e6b8cfbc6945ecdf550a7f673a2 | 4f0e2ea86e84e369b249bdcdffec254f28e948e2 | "2023-08-15T16:48:08Z" | c++ | "2023-09-28T18:04:02Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,646 | ["servers/rendering/renderer_rd/renderer_canvas_render_rd.cpp", "servers/rendering/renderer_rd/renderer_canvas_render_rd.h", "servers/rendering/renderer_rd/storage_rd/texture_storage.cpp", "servers/rendering/renderer_rd/storage_rd/texture_storage.h"] | 2d hdr + font msdf causes broken text rendering | ### Godot version
4.2.dev3
### System information
Godot v4.2.dev3 - Arch Linux #0~20230811.g489fabdc SMP PREEMPT_DYNAMIC Sat Aug 12 02:04:20 UT - X11 - Vulkan (Forward+) - dedicated NVIDIA GeForce GTX 1650 (nvidia; 535.98) - Intel(R) Core(TM) i5-9300H CPU @ 2.40GHz (8 Threads)
### Issue description
After turning on 2d hdr and font msdf at the same time, there is a problem with ui text rendering.
2d hdr off:
2d hdr on:
### Steps to reproduce
1. Open msdf for default font or custom font.
2. Turn on 2d hdr.
3. Add a Label node and enter text, the text rendering is destroyed.
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/80646 | https://github.com/godotengine/godot/pull/80651 | f8aa7caf1d438f1620f0ebce5585ddff5186676c | 623156a05c07cc7625679ce730a16129a06edc8b | "2023-08-15T04:06:19Z" | c++ | "2023-08-16T07:17:01Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,640 | ["scene/animation/animation_player.cpp"] | AnimationPlayer seek() does not skip Call Method Track Keys | ### Godot version
4.1.1 stable 4.2. dev3
### System information
Win 10
### Issue description
[seek()](https://docs.godotengine.org/en/stable/classes/class_animationplayer.html?rtd_search=add_color_override#class-animationplayer-method-seek) is supposed to skip the keys inbetween current Timestamp and it's argument value Timestamp.
When playing an Animation in the AnimationPlayer which has a Call Method track key somewhere in the middle section for example, and then the user tries to "skip" that middle section using `seek()`, the Call Method track keys in the middle section are still handled, even though animation has skipped the middle section and the `update` argument is `false`.
So it basically behaves like [advance()](https://docs.godotengine.org/en/stable/classes/class_animationplayer.html?rtd_search=add_color_override#class-animationplayer-method-advance) should behave.
However, advance also does not behave as it should, since it advances further than it should and when setting Update in the seek() method to `true`, the Call Method track key is actually called twice!
### Steps to reproduce
Create an AnimationPlayer Animation of 5 seconds,
Add a script with two functions: One to seek(4.0) and another one to print("hi")
Add a Call Method Track and at Timestamp second 1.0 add a key to the seek() method and on Timestamp 2.5 add a key to print().
Play the scene and observe "hi" being printed
Set seek(4.0, true) to enable Update (like in advance method), and observe "hi" being printed twice!
Maybe related: https://github.com/godotengine/godot/issues/76773
### Minimal reproduction project
4.2. dev3 (issue is the same in 4.1.1 stable)
[animation_seek_bug.zip](https://github.com/godotengine/godot/files/12340066/animation_seek_bug.zip) | https://github.com/godotengine/godot/issues/80640 | https://github.com/godotengine/godot/pull/80708 | e73a4a382ee57c9eb006d7e8f11038643081b12b | 0511f9d9a7d56c742d87fafdcea8785d40ad14b3 | "2023-08-14T22:48:56Z" | c++ | "2023-08-17T13:44:03Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,636 | ["servers/rendering/renderer_rd/storage_rd/material_storage.cpp"] | Sprite 3D doesnt work with Viewport texture on android. | ### Godot version
4.1.1
### System information
Windows 10, Godot 4.1.1 Latest, Vulkan API 1.3.206 - Forward Mobile.
### Issue description
I was trying to make ProgressBar for 3d Object. I used Sprite3D with SubViewPort as a texture. In edit page it works correctly, in run page too. After exporting to android device its not. It works not only with ProgressBar, with all textures to sprite3D with SubViewPort as a texture.
### Steps to reproduce
1. Create godot(4.1.1) project Mobile.
2. Add 3D Node and any 2d stuff, for ex, ColorRect, u can choos button, progressbar, sprite2d etc.
3.Join it as a child to SubViewport
4.Create Sprite3D and pick texture to it as SubViewport, and choose SubViewport with 2d stuff as a child.
5.place camera 3D so when u run scene u can see our 2d stuff in 3d space.
6. As we can see, all works properly. Then export it to android, and check. We see nothing.
### Minimal reproduction project
- | https://github.com/godotengine/godot/issues/80636 | https://github.com/godotengine/godot/pull/81607 | b42aa5ad9d185a5263885755a17dc9d4e93314d5 | c255acebe867b65bd7cd81a900c0e3144e2d472a | "2023-08-14T20:06:49Z" | c++ | "2023-10-05T08:07:16Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,625 | ["scene/resources/particle_process_material.cpp"] | In the particle shader the deterministic sequence of random values is not guaranteed | ### Godot version
v4.1.1.stable.official [bd6af8e0e]
### System information
Windows 10.0.22621 - Vulkan (Forward+) - dedicated NVIDIA GeForce GTX 1650 SUPER (NVIDIA; 31.0.15.3130) - Intel(R) Core(TM) i5-10400F CPU @ 2.90GHz (12 Threads)
### Issue description
I think I found a bug in the particle shader. It revolves around the following line:
```
// apply linear acceleration
force += length(VELOCITY) > 0.0 ? normalize(VELOCITY) * tex_linear_accel * mix(linear_accel_min, linear_accel_max, rand_from_seed(alt_seed)) : vec3(0.0);
```
As long as VELOCITY is greater than 0, "rand_from_seed(alt_seed)" is called among other things. Thereby "alt_seed" is changed.
But if now damping is active, VELOCITY reaches 0 at some point. Then "rand_from_seed(alt_seed)" is NOT called anymore.
That means: all following random values, which depend on "rand_from_seed()", change once abruptly.
### Steps to reproduce
The error probably manifests itself in very different ways. As soon as damping brings the speed to 0, there will be one-time jumpy changes in all aspects that use edge_fom_seed:
- radial acceleration
- tangential acceleration
- orbit velocity
- damping
- angular velocity
- animation offset and speed
- hue
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/80625 | https://github.com/godotengine/godot/pull/80638 | 1d852cb75ea004054621fe7976c44adc63d572b3 | ed1180a97e5e1bd333895e7e3937c83622674c34 | "2023-08-14T15:41:21Z" | c++ | "2023-08-16T07:16:14Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,593 | ["editor/export/editor_export_platform.cpp"] | Export plugin not working with GdExtensions | ### Godot version
4.1.1-stable
### System information
macOS 13.4.1 arm64
### Issue description
Implementing `_export_file`, `_export_begin` and `_export_end` in a gdextension export plugin does nothing.
In engine we have this code:
```cpp
EditorExportPlatform::ExportNotifier::ExportNotifier(EditorExportPlatform &p_platform, const Ref<EditorExportPreset> &p_preset, bool p_debug, const String &p_path, int p_flags) {
HashSet<String> features = p_platform.get_features(p_preset, p_debug);
Vector<Ref<EditorExportPlugin>> export_plugins = EditorExport::get_singleton()->get_export_plugins();
//initial export plugin callback
for (int i = 0; i < export_plugins.size(); i++) {
export_plugins.write[i]->set_export_preset(p_preset);
if (export_plugins[i]->get_script_instance()) { //script based
PackedStringArray features_psa;
for (const String &feature : features) {
features_psa.push_back(feature);
}
export_plugins.write[i]->_export_begin_script(features_psa, p_debug, p_path, p_flags);
} else {
export_plugins.write[i]->_export_begin(features, p_debug, p_path, p_flags);
}
}
}
```
We can see it will call `_export_begin_script` if a script instance is attached to the export plugin, otherwise it calls `_export_begin`. `_export_begin_script` does a `GDVIRTUAL_CALL` call, so it works within a script language. On another side `_export_begin` rely on classic cpp virtual call.
In the GdExtension case, we have no script instance attached, so we end up in classic cpp vtable case, but gdextensions does not use cpp vtable for inheritance. This ends up with no call to gdextension implementation.
Currently, the only work around to make a gdextension export plugin work is to create a script proxy for it, here is my example.
gdextension:
```cpp
namespace godot {
class FmodEditorExportPlugin : public EditorExportPlugin {
GDCLASS(FmodEditorExportPlugin, EditorExportPlugin)
public:
void _export_begin(const PackedStringArray &features, bool is_debug, const String &path, uint32_t flags) override;
void _export_file(const String& path, const String& type, const PackedStringArray& features) override;
String _get_name() const override;
static void _bind_methods();
FmodEditorExportPlugin() = default;
~FmodEditorExportPlugin() = default;
};
}
```
Proxy script:
```gdscript
@tool
class_name FmodEditorExportPluginProxy
extends FmodEditorExportPlugin
```
I think we should always rely on `GDVIRTUAL_CALL` and avoid check on script instance, so that both script and gdextensions are handled.
### Steps to reproduce
Create a gdextension export plugin and add it using editor plugin.
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/80593 | https://github.com/godotengine/godot/pull/80999 | ea3c5f96e5254de94263e05d031d7db17c59e035 | 9a140f92be687cb9c79e0a354e915372bd5b0645 | "2023-08-13T16:44:47Z" | c++ | "2023-08-28T10:06:39Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,588 | ["editor/editor_node.cpp", "editor/plugins/script_editor_plugin.cpp", "editor/plugins/script_editor_plugin.h"] | CollisionShape3D Shape Size handles missing when Script editor floating | ### Godot version
4.1.1.stable
### System information
Godot v4.1.1.stable - Windows 10.0.19045 - Vulkan (Mobile) - dedicated NVIDIA GeForce RTX 2070 SUPER (NVIDIA; 30.0.15.1295) - Intel(R) Core(TM) i5-8400 CPU @ 2.80GHz (6 Threads)
### Issue description
When undocking the script editor so that it's floating, CollisionShape3D Shape Size handles stop working.
If they were visible on a Shape, they remain visible on that shape and still work for that shape; but they will not appear on any other collision object when selecting it.
If you launch the editor with the Script panel already in a floating state, no shape handles ever become visible.
If you dock the Script editor again, everything goes back to normal and works fine.
### Steps to reproduce
Create a 3D scene with some collision shapes.
Undock the Script editor
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/80588 | https://github.com/godotengine/godot/pull/80647 | 04306777a7c75d49bfb47f484cda29ef939a92b7 | c7356c795b3c8cec2453061299e85cb315a9a6a7 | "2023-08-13T09:50:38Z" | c++ | "2023-08-17T09:27:04Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,574 | ["platform/android/java/lib/src/org/godotengine/godot/Godot.kt"] | Immersive mode no longer works on android, both the notification bar and the navigation bar are still visible. | ___
***Bugsquad note:** This issue has been confirmed several times already. No need to confirm it further.*
___
### Godot version
v4.2.dev2.official [da81ca62a]
### System information
Windows 11 - Vulkan (Forward+)
### Issue description
Immersive mode seems to no longer works on android, both the notification bar and the navigation bar are still visible. The pinhole cutout is has also gained a black bar.
Tried creating a completely blank project and the problem still arises.
Godot 4.2 dev 2:
Godot 4.1:
I am using a Pixel 7 with Android 13
### Steps to reproduce
Export to android
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/80574 | https://github.com/godotengine/godot/pull/84102 | b096174259e1db12ae7cd30ea5fe3a3176c51029 | fd49c428a3eb0b05b15d656e50d00df198abb161 | "2023-08-13T01:16:17Z" | c++ | "2023-10-28T16:35:45Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,570 | ["drivers/vulkan/vulkan_context.cpp"] | Vulkan Validation Error when resizing | ### Godot version
master [4714e95896c8db02616ea6ec7f9aff92dec1cae4]
### System information
Godot v4.2.dev (4714e9589) - Ubuntu 20.04.6 LTS (Focal Fossa) - X11 - Vulkan (Forward+) - dedicated AMD Radeon RX 6800 XT - AMD Ryzen 9 5900X 12-Core Processor (24 Threads)
### Issue description
Sometimes when resizing the window on Linux X11 we may get the following validation error:
```
ERROR: VALIDATION - Message Id Number: -370888023 | Message Id Name: VUID-vkAcquireNextImageKHR-semaphore-01286
Validation Error: [ VUID-vkAcquireNextImageKHR-semaphore-01286 ] Object 0: handle = 0xdcc8fd0000000012, type = VK_OBJECT_TYPE_SEMAPHORE; | MessageID = 0xe9e4b2a9 | vkAcquireNextImageKHR: Semaphore must not be currently signaled or in a wait state. The Vulkan spec states: If semaphore is not VK_NULL_HANDLE it must be unsignaled (https://vulkan.lunarg.com/doc/view/1.2.198.1/linux/1.2-extensions/vkspec.html#VUID-vkAcquireNextImageKHR-semaphore-01286)
```
I nailed down the problem. In `VulkanContext::prepare_buffers` we have the following loop (I simplified to highlight the problem):
```cpp
do {
err = fpAcquireNextImageKHR(device, w->swapchain, UINT64_MAX,
w->image_acquired_semaphores[frame_index], VK_NULL_HANDLE, &w->current_buffer);
if (err == VK_SUBOPTIMAL_KHR) {
// Swapchain is not as optimal as it could be, but the platform's
// presentation engine will still present the image correctly.
print_verbose("Vulkan: Early suboptimal swapchain, recreating.");
Error swap_chain_err = _update_swap_chain(w);
if (swap_chain_err == ERR_SKIP) {
break;
}
} while (err != VK_SUCCESS);
```
The problem is that vkAcquireNextImageKHR returned VK_SUBOPTIMAL_KHR but it already signaled the semaphore (because it is possible to continue normally with a VK_SUBOPTIMAL_KHR result).
Then we recreate the swapchain and reuse the w->image_acquired_semaphores[frame_index] which is in an inconsistent state.
I will handle this bug.
### Steps to reproduce
1. Open the Editor on any project
2. Resize the project until it happens
Reproducibility is much more frequent after applying PR #80566
### Minimal reproduction project
Any. | https://github.com/godotengine/godot/issues/80570 | https://github.com/godotengine/godot/pull/80571 | 6c3735cb7c9600abfb18b683b06b4c6d00a09d07 | d1b8e9abd618a21daaec56cf4c71e4eedae51910 | "2023-08-12T23:10:39Z" | c++ | "2023-08-16T07:13:53Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,564 | ["modules/gdscript/gdscript_parser.cpp"] | Mistake in godot debug error message | ### Godot version
4.1.1.stable.official
### System information
Godot v4.1.1.stable - Windows 10.0.19045 - Vulkan (Compatibility) - Intel(R) HD Graphics 4600 (Intel Corporation; 20.19.15.5126) - Intel(R) Core(TM) i7-4790 CPU @ 3.60GHz (8 Threads)
### Issue description
So theres this error message:
Expected error after ">" operator."
there appears to be a useless " at the end. Is this intentional?
### Steps to reproduce
attach a script and write a ternary operator line, But do after the operator do not write the other piece of code.
var test = 1 if a > # leave it out so it gives an debug error message at edit time
you'll see the debug error message
### Minimal reproduction project
go to scenes\enemys\frog\frog.gd
youll see line 24 which includes the bug
[test.zip](https://github.com/godotengine/godot/files/12328367/test.zip)
| https://github.com/godotengine/godot/issues/80564 | https://github.com/godotengine/godot/pull/80568 | 4714e95896c8db02616ea6ec7f9aff92dec1cae4 | 7ba79d68bd0f97797d7cb37452da6a036ba7c7c9 | "2023-08-12T22:01:01Z" | c++ | "2023-08-13T09:23:22Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,556 | ["scene/gui/tree.cpp"] | Control Tree range slider not working (also affects input action deadzone editor) | ### Godot version
4.1.1-stable_win64
### System information
Windows 10
### Issue description
Range cell within Tree control not update when using slider.
Also it would be nice, if I could change few settings for range mode:
1. Visibility of slider. Not always is necessery.
2. Visibility of spin box. As above
3. Slider theme variant. Now it takes default variant, but I would want to set dedicated variant for it.
4. Turn off context menu under RMB. Editable cells in general. Not only range
### Steps to reproduce
1. Create Tree control
2. Add TreeItem. Set it's cell in range mode and editable
3. Run scene. Try to change value of cell using slider.
### Minimal reproduction project
[TreeRange.zip](https://github.com/godotengine/godot/files/12327982/TreeRange.zip)
| https://github.com/godotengine/godot/issues/80556 | https://github.com/godotengine/godot/pull/81174 | 0d4fe74043125f4c7e0291687e2820e8a5c67cb0 | 0c7ac25b470cf4fa44b15dae0e3097a50d998cf6 | "2023-08-12T18:05:53Z" | c++ | "2023-10-02T11:16:03Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,553 | ["scene/gui/rich_text_label.cpp", "scene/gui/scroll_bar.cpp"] | Updating RichTextLabel's Text while using scrollbar prevents release | ### Godot version
v4.1.stable.official [970459615]
### System information
Windows 10
### Issue description
Changing a RichTextLabel's text property while dragging the scrollbar causes releasing the mouse to not actually prevent further dragging. Technically it's more pervasive than this since the scrollbar stops reporting other mouse input events (stops being highlighted when hovered, can't release) in general when updating every frame. This works fine for 3.5. This is still broken for 4.2dev3.
### Steps to reproduce
Change a RichTextLabel's text property while dragging the scrollbar. MRP attached.
https://github.com/godotengine/godot/assets/3820082/99c961af-e31c-4dde-8fd5-061e2e9e5fdc
### Minimal reproduction project
[rtl-scrollbar-mrp.zip](https://github.com/godotengine/godot/files/12327813/rtl-scrollbar-mrp.zip)
| https://github.com/godotengine/godot/issues/80553 | https://github.com/godotengine/godot/pull/80606 | 4ed3f67229d2c8b8dbdc9306e41e22dda7b33889 | da80a033f118b44334a2c0b6d75e1df39236191e | "2023-08-12T16:32:15Z" | c++ | "2023-08-16T07:14:40Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,524 | ["editor/editor_inspector.cpp", "editor/editor_inspector.h"] | Can't reorder OptionButton Items if there are more than five items | ### Godot version
v4.1.stable.official [970459615]
### System information
Windows 10, Forward+, AMD Radeon R9 290X
### Issue description
There's no way to reorder items "across" pages in paged lists in the inspector - items can only be rearranged within the page they're already in. The "Max Array Dictionary Items per Page" setting doesn't seem to affect OptionButton Item lists.
There are two ways around the issue: add a script that orders the items as desired, or delete items until the item you need to reorder is on the right page, then re-add the items you deleted.
### Steps to reproduce
1. Create an OptionButton
2. Add six items
3. See that the last item is on a new page
4. There's no way to change the position of the last item
### Minimal reproduction project
[OptionButtonItemListIssueMRP.zip](https://github.com/godotengine/godot/files/12323899/OptionButtonItemListIssueMRP.zip)
| https://github.com/godotengine/godot/issues/80524 | https://github.com/godotengine/godot/pull/80617 | 3417f34a374ea08720758f067c77436d3bfcb53c | c1e85c65930c515c0d6627a9f6b5e6b131767dea | "2023-08-11T17:38:32Z" | c++ | "2023-08-29T10:43:34Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,509 | ["modules/navigation/nav_mesh_generator_3d.cpp"] | NavigationRegion3D generates inexplicable hole at (0,0,0) against HeightMapShape3D with values at -0.5 to -1.75... | ### Godot version
4.1.1 Stable
### System information
Windows 11, RTX 3070TI, Vulcan
### Issue description
NavigationRegion3D generates inexplicable hole at (0,0,0) against HeightMapShape3D with values at -0.5 to -1.75
Not sure how this hasn't been noticed before, but the NavigationMeshRegion3D bakes a single square unit size hole at (0,0,0) against HeightMapShape3D if it is using values between -0.5 and -1.75... I have attached a MRP that is very easy to use and demonstrate the error happening.
### Steps to reproduce
Create a HeightMapShape3D, set values between -0.5 and -1.75, bake NavigationRegion3D mesh on it. Observe hole in mesh at (0,0,0).
In MRP with the "main" node selected, you can adjust the "height_map_value" export variable to move a dynamically generated heightmapshape3d up and down, the script will automatically rebake the navmesh when the value is changed. When values are between -0.5 and -1.75 a hole will appear at (0,0,0)
[MRP_NavMeshHole.zip](https://github.com/godotengine/godot/files/12319007/MRP_NavMeshHole.zip)
### Minimal reproduction project
In MRP with the "main" node selected, you can adjust the "height_map_value" export variable to move a dynamically generated heightmapshape3d up and down, the script will automatically rebake the navmesh when the value is changed. When values are between -0.5 and -1.75 a hole will appear at (0,0,0) | https://github.com/godotengine/godot/issues/80509 | https://github.com/godotengine/godot/pull/83783 | 17aa5c51705bd67f0c1ab4d532aa884bbc66433d | 912366093d5b665bed1674c746cbec08bf8ddb0f | "2023-08-11T06:23:16Z" | c++ | "2023-10-23T10:42:28Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,508 | ["modules/gdscript/gdscript_analyzer.cpp", "modules/gdscript/tests/scripts/analyzer/features/inner_class_access_from_inside.gd", "modules/gdscript/tests/scripts/analyzer/features/inner_class_access_from_inside.out"] | Inner class in GDScript can no longer refer to its own name | ### Godot version
4.2.dev commit 013e8e3afb982d4b230f0039b6dc248b48794ab9
### System information
Any
### Issue description
The following code from [dialogic](https://github.com/coppolaemilio/dialogic) no longer works in 4.2-dev, but does in 4.1 and before:
```gdscript
class VariableFolder:
var data = {}
var path = ""
var outside
func _init(_data, _path, _outside):
data = _data
path = _path
outside = _outside
func _get(property):
property = str(property)
if property in data:
if typeof(data[property]) == TYPE_DICTIONARY:
return VariableFolder.new(data[property], path+"."+property, outside) ### <====== ERROR ###
else:
return DialogicUtil.logical_convert(data[property])
func _set(property, value):
property = str(property)
if not value is VariableFolder:
outside._set_value_in_dictionary(path+"."+property, outside.dialogic.current_state_info['variables'], value)
return true
elif VariableFolder:
return true
```
On compilation this will give the error
```
SCRIPT ERROR: Compile Error: Identifier not found: VariableFolder
at: GDScript::reload (res://addons/dialogic/Modules/Variable/subsystem_variables.gd:182)
```
(line 182 is the `VariableFolder.new` line)
i bisected it to the following commit, from #79880:
```
d53fc92b4c6b5e4484e8f0bfff6ac55163dde3fb is the first bad commit
commit d53fc92b4c6b5e4484e8f0bfff6ac55163dde3fb
Author: Danil Alexeev <[email protected]>
Date: Tue Jul 25 14:21:49 2023 +0300
GDScript: Fix bug with identifier shadowed below in current scope
```
### Steps to reproduce
i think the issue description is complete enough to allow reproduction, let me know if there's any problems.
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/80508 | https://github.com/godotengine/godot/pull/80510 | 0308422f461dce11339896249e23ff02d978bfa0 | 6b5e44f3ca1664448c1ecc185a1a03efe557b2e9 | "2023-08-11T06:22:21Z" | c++ | "2023-08-14T13:28:57Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,507 | ["editor/plugins/asset_library_editor_plugin.cpp", "editor/plugins/asset_library_editor_plugin.h", "scene/gui/link_button.cpp", "scene/gui/link_button.h"] | AssetLib: Long plugin name break the UI | ### Godot version
v4.1.1.stable.arch_linux
### System information
OpenGL API 4.2 (Core Profile) Mesa 23.1.4 - Compatibility - Using Device: Intel - Mesa Intel(R) HD Graphics 2500 (IVB GT1)
### Issue description
The AssetLib container is set to the size of their children and will automatic resize itself to display all the text, ignore user and windows manager setting.
### Steps to reproduce
Find some plugin that has very long name like _VRM Importer and MToon Shader - Humanoid 3d Avatar Model Format_
### Minimal reproduction project
n/a | https://github.com/godotengine/godot/issues/80507 | https://github.com/godotengine/godot/pull/80555 | 365ee4196cbf8c2d64e118b51473de75ae3a6b6a | 2d6cee48f6a407f171e4802d242fb6731ad020ad | "2023-08-11T05:36:43Z" | c++ | "2023-10-02T13:18:38Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,493 | ["servers/rendering/renderer_rd/effects/tone_mapper.cpp"] | Vulkan Mobile error with OpenXR at Scale 1: "Draw command lacks indices, but pipeline format does not use vertices" | ### Godot version
v4.2.dev3.official [013e8e3af]
### System information
Godot v4.2.dev3 - Windows 10.0.19045 - Vulkan (Mobile) - dedicated NVIDIA GeForce RTX 3080 (NVIDIA; 31.0.15.3667) - AMD Ryzen 7 5800X3D 8-Core Processor (16 Threads)
### Issue description
When I open my OpenXR mobile project in 4.2.dev3 I get the error
> "Draw command lacks indices, but pipeline format does not use vertices".
The 3D Viewport is completely black when this error is going. The error disappears and 3D Viewport works again as soon as scaling (both bilinear and FSR) are changed to anything other than 1. It only happens when Project Settings > Rendering > Scaling 3D > Scale == 1
This error also disappears when switching to the Foward+ renderer. I tried to see if disabling OpenXR Shaders would fix things but it crashes the editor and makes a project unopenable. This occurs with the base Godot XR Template (which is what my XR project is based on).
Additionally, at the beginning of opening the project there will be several errors when loading various scripts from the godot-xr-tools addon:
> Failed to load script "res://addons/godot-xr-tools/functions/movement_sprint.gd" with error "Compilation failed".
This may have been introduced by the fix in: https://github.com/godotengine/godot/pull/80368
### Steps to reproduce
Open the [OpenXR template](https://github.com/GodotVR/godot-xr-template) and set renderer to Mobile
### Minimal reproduction project
[Open XR Template](https://github.com/GodotVR/godot-xr-template) | https://github.com/godotengine/godot/issues/80493 | https://github.com/godotengine/godot/pull/80502 | a8edbcfe134590cc98e4f19855d1f95dbc932564 | 4714e95896c8db02616ea6ec7f9aff92dec1cae4 | "2023-08-10T19:13:34Z" | c++ | "2023-08-11T08:33:47Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,442 | ["editor/debugger/editor_debugger_inspector.cpp"] | Crash on breakpoint if freed node exists | ### Godot version
v4.2.dev2.official [da81ca62a]
### System information
Godot v4.2.dev2 - Windows 10.0.19045 - Vulkan (Forward+) - dedicated NVIDIA GeForce GTX 1060 6GB (NVIDIA; 31.0.15.3640) - Intel(R) Core(TM) i5-3570K CPU @ 3.40GHz (4 Threads)
### Issue description
If a freed node exists when editor breaks on breakpoint, Godot crashes. My guess is that debug view tries to show the freed node and somehow fails. If the freed node is set to null before the breakpoint, crash doesn't happen.
`v4.1.1.stable.official` works fine, but `v4.2.dev2.official` crashes.
### Steps to reproduce
```gdscript
extends Node2D
func _ready() -> void:
var node = Node.new()
node.free()
#node = null # uncomment this line to prevent crash
pass # add breakpoint here to crash
```
### Minimal reproduction project
[g42_crash.zip](https://github.com/godotengine/godot/files/12300368/g42_crash.zip)
| https://github.com/godotengine/godot/issues/80442 | https://github.com/godotengine/godot/pull/80256 | 835957b1fc418c5506c40aa23cc2d9b1d327e06c | 75c979e48df935b15eeb8c7f33b1f7dc50527526 | "2023-08-09T09:06:54Z" | c++ | "2023-08-09T13:20:20Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,434 | ["doc/classes/Callable.xml", "modules/gdscript/gdscript_lambda_callable.cpp", "modules/gdscript/gdscript_lambda_callable.h", "modules/gdscript/gdscript_rpc_callable.cpp", "modules/gdscript/gdscript_rpc_callable.h"] | Can't get method name from named lambda | ### Godot version
4.2 f7bc653
### System information
Windows 10.0.19045 - Vulkan (Forward+) - dedicated NVIDIA GeForce GTX 1060 (NVIDIA; 30.0.15.1403) - Intel(R) Core(TM) i7-7700HQ CPU @ 2.80GHz (8 Threads)
### Issue description
`get_method()`:
```
Returns the name of the method represented by this Callable. If the callable is a lambda function, returns the function's name
```
However calling it on a named lambda results in
```
Can't get method on CallableCustom "test(lambda)".
```
(ironically the error even contains the function's name...)
`get_method()` in GDScriptLambdaCallable is likely unimplemented or is wrong. It should work for named lambdas.
### Steps to reproduce
```GDScript
var fun = func test(): pass
print(fun.get_method())
```
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/80434 | https://github.com/godotengine/godot/pull/80506 | 8b3de35c6431dddd34def48cca491daa994f9411 | e1bb9b523d144fcc8c30aff02270604043ba8ab1 | "2023-08-08T21:55:52Z" | c++ | "2023-08-14T07:36:35Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,425 | ["scene/gui/dialogs.cpp", "scene/gui/popup.cpp"] | OptionButton menu not closable with joypad | ### Godot version
v4.1.stable.official [970459615]
### System information
Godot v4.1.stable - Windows 10.0.22000 - Vulkan (Forward+) - dedicated NVIDIA GeForce RTX 2070 SUPER (NVIDIA; 30.0.15.1295) - Intel(R) Core(TM) i7-9700K CPU @ 3.60GHz (8 Threads)
### Issue description
`OptionButton` menu's do not seem to be closable using a joypad. They can be closed using `ESC` on a keyboard. Setting a `ui_cancel` mapping for the joypad does not fix the issue. Further, the `MenuButton` seems to exhibit the same behaviour based on this [Reddit post](https://www.reddit.com/r/godot/comments/142d8pa/how_to_exit_a_menubutton_with_gamepad/). I would expect to be able to close the menu from the joypad using a button mapped to `ui_cancel`.
### Steps to reproduce
1. Add an OptionButton to project
2. Setup `ui_cancel` input mapping to have a joypad mapping
3. Run the project and open the OptionButton menu.
4. Try to close the window without selecting an option using the joypad. This is the step that does not work. Pressing `ESC` will close the window.
### Minimal reproduction project
[OptionButtonReproProject.zip](https://github.com/godotengine/godot/files/12294459/OptionButtonReproProject.zip) | https://github.com/godotengine/godot/issues/80425 | https://github.com/godotengine/godot/pull/80594 | 122f3730ca28654c173ba4c26c9a0e96909433f6 | 0308422f461dce11339896249e23ff02d978bfa0 | "2023-08-08T18:01:31Z" | c++ | "2023-08-14T07:38:08Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,420 | ["modules/openxr/openxr_api.cpp", "modules/openxr/openxr_api.h", "modules/openxr/openxr_interface.cpp"] | Multiple OpenXR Action Sets in Godot does not work | ### Godot version
4.2 dev 2
### System information
Windows 10 Godot Engine v4.2.dev2.mono.official.da81ca62a - https://godotengine.org OpenGL API 3.3.0 NVIDIA 536.67 - Compatibility - Using Device: NVIDIA - NVIDIA GeForce GTX 980 Ti
### Issue description
If you have multiple openxr action sets it will fail to load all of the action sets after the first one loaded
### Steps to reproduce
Have multiple openxr action sets
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/80420 | https://github.com/godotengine/godot/pull/80419 | a017e9c69877be77744dfb2bff0bbc2a811427cf | 8a02cd088320cb428c5ea157a26ce739aab8f0cd | "2023-08-08T15:17:33Z" | c++ | "2023-08-18T07:30:52Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,417 | ["core/io/resource.cpp", "core/io/resource.h", "scene/main/viewport.cpp", "scene/main/viewport.h"] | When scene is reloaded after changing instance, ViewportTextures break | ### Godot version
4.2 eca6f0e
### System information
Windows 10 x64
### Issue description
https://github.com/godotengine/godot/assets/2223172/a2796263-4e2e-4b14-a080-b3ad034cc141
### Steps to reproduce
1. Add SubViewport
2. Add Sprite2D
3. Set spirte's texture to ViewportTexture
4. Create new scene
5. Instance the new scene inside the scene with viewport
6. Make sure both scenes are opened
7. Go to instance scene and save it
8. Go to viewport scene
You will see error spam. Reloading the scene fixes it.
### Minimal reproduction project
[vip.zip](https://github.com/godotengine/godot/files/12292386/vip.zip)
| https://github.com/godotengine/godot/issues/80417 | https://github.com/godotengine/godot/pull/79201 | 1537452aa94e3fdb19386eac9f8394b391998e8a | e73a4a382ee57c9eb006d7e8f11038643081b12b | "2023-08-08T14:28:35Z" | c++ | "2023-08-17T13:43:38Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,403 | ["editor/editor_themes.cpp", "editor/plugins/texture_region_editor_plugin.cpp", "editor/plugins/texture_region_editor_plugin.h", "scene/resources/atlas_texture.cpp"] | Region Editor not showing the correct image when using Atlas texture in StyleBoxTexture | ### Godot version
v4.1.1.stable.official [bd6af8e0e]
### System information
Godot v4.1.1.stable - Ubuntu 22.04.2 LTS 22.04 - Vulkan (Forward+) - integrated Intel(R) UHD Graphics 615 (AML-KBL) () - Intel(R) Core(TM) i3-10100Y CPU @ 1.30GHz (4 Threads)
### Issue description
Summary: When editing the subregion of a StyleBoxTexture, it shows me the entire spritesheet of the Atlas texture, instead of the selected portion of the atlas, making it difficult (and non-intuitive) to correctly select the subregion I need.
What I expect: I understand that the sub-region may be used to emulate the behavior of a nine-patch. So I would expect to see there the subregion that i selected previously from the Atlas as the texture to be shown.
I have only tested this in a Panel Container
### Steps to reproduce
I found this problem when doing these steps:
- Add a PanelContainer node
- In the inspector select Theme Overrrides -> Style -> new StyleBoxTexture
- In the texture of the StyleBoxTexture select new AtlasTexture
- Dropped the spritesheet in the Atlas of the AtlasTexture
- In Region of the AtlasTexture, selected Edit Region and drew the rectangle on the texture i want:
- Back in the StyleBoxTexture options, in Sub-Region, clicked on Edit Region to select the nine-patch region i need
- Here is where the region editor shows me the entire spritesheet and not the sub-texture i selected from the atlas. The image below shows the axis correctly positioned for the nine-patch to work, but it is not intuitive because it has to be positioned without seeing the correct texture, making it difficult to finetune for some textures.
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/80403 | https://github.com/godotengine/godot/pull/80435 | 2c0a74a14938e053fc44e32b7fc134513c6f3bb5 | 33a3e12fe1a0060197c77ba9531ebe5528543356 | "2023-08-08T07:28:04Z" | c++ | "2023-08-28T10:04:18Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,399 | ["editor/plugins/tiles/tiles_editor_plugin.cpp", "editor/plugins/tiles/tiles_editor_plugin.h"] | TileSet options get hidden after every change | ### Godot version
v4.2.dev2.mono.official
### System information
Godot v4.2.dev2.mono - Windows 10.0.22621 - GLES3 (Compatibility) - NVIDIA GeForce RTX 3060 (NVIDIA; 31.0.15.3179) - Intel(R) Core(TM) i7-10700F CPU @ 2.90GHz (16 Threads)
### Issue description
When making a change on a TileSet, the TileSet tab will close, so you have to reopen it to continue doing what you want to do.
https://github.com/godotengine/godot/assets/88297220/b9f727d8-3ced-404a-9eb0-06f22c66bb99
### Steps to reproduce
1. Add a TileMap
2. Assign a TileSet to the TileMap
3. Add a new layer or make a change to the TileSet.
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/80399 | https://github.com/godotengine/godot/pull/80462 | d76b97e9040fb620ef71160a7ddd118199485461 | 013e8e3afb982d4b230f0039b6dc248b48794ab9 | "2023-08-08T04:50:51Z" | c++ | "2023-08-09T22:44:13Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,397 | ["editor/connections_dialog.cpp"] | Custom signal descriptions are not displayed in Node > Signals tab | ### Godot version
v4.1.1.stable.official [bd6af8e0e]
### System information
v4.1.1.stable.official [bd6af8e0e]
### Issue description
In-script documentation (`##`) for `signal` members is not appearing in the _Node_ tab as expected. Other documentation appears to work correctly in the _Inspector_ tab.
Working example (exported variable):
Non-working example (`signal`)
I have tried deselecting and reselecting the associated Node, changing between the tabs on the left pane, changing active scripts and even restarting Godot - nothing makes it work. According to the [GScript Documentation on Documentation Comments](https://docs.godotengine.org/en/stable/tutorials/scripting/gdscript/gdscript_documentation_comments.html), this **should** work, right? Or is there something completely obvious that I am missing?
### Steps to reproduce
1. In Godot, create a new Node of type `Node`
2. Attach a new script to this Node, accepting all current defaults
3. In the script editor create a `signal` called `Foo` and right above it, create a `##` comment for it
4. Also create an `@export`ed integer variable called `Bar` and also create a `##` comment for it
5. Unselect and reselect the Node in the _Scene_ tab to refresh the _Inspector_ and _Node_ tabs
6. Hover over the variable in the _Inspector_ tab and see that the added description is shown correctly
7. Hover over the signal in the _Node_ tab and see that the added description is **not** shown and "_No description._" is shown instead
This is an example of the script that could be added:
```
## This is signal foo!
signal Foo
## This is exported signal bar!
@export var Bar : int = 0
```
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/80397 | https://github.com/godotengine/godot/pull/80411 | 06eade1bc274c42fa3fb9d27108056ce91780840 | 417239ee6dd795fd064e635b022895a50698fe0e | "2023-08-08T03:29:14Z" | c++ | "2023-08-17T09:25:51Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,396 | ["drivers/gles3/rasterizer_canvas_gles3.cpp", "drivers/gles3/rasterizer_canvas_gles3.h"] | GLES3: Color.WHITE or Color(1, 1, 1, 1) is black when used inside _draw() function | ### Godot version
4.1.1-stable
### System information
Windows 10
### Issue description
this will draw black rect
```gdscript
func _draw():
draw_rect(Rect2i(Vector2i.ZERO, img.get_size()), Color(1, 1, 1, 1), false)
func _draw():
draw_rect(Rect2i(Vector2i.ZERO, img.get_size()), Color.WHITE, false)
```
while this will draw white rect
```gdscript
func _draw():
draw_rect(Rect2i(Vector2i.ZERO, img.get_size()), Color(0.99, 1, 1, 1), false)
```
while this is weird but making circle draws Color(0.99, 1, 1, 1) makes also the _draw_rect to get fixed!
### Steps to reproduce
use Color.WHITE inside any _draw() function or Color(1, 1, 1, 1) and it will draw black color instead of white color
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/80396 | https://github.com/godotengine/godot/pull/80151 | c72b851dfb46606d3d04d00c79abd555309cae33 | 5541567c1a8a8159b7041049a89310616afa0065 | "2023-08-08T03:25:27Z" | c++ | "2023-08-17T09:23:24Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,388 | ["scene/animation/tween.cpp"] | Using SetDelay() in Godot 4 Tweeners makes them start from the value at the start of the delay rather than at the start of the interpolation | ### Godot version
v4.1.stable.mono.official [970459615]
### System information
Godot v4.1.stable.mono - Windows 10.0.19045 - Vulkan (Forward+) - dedicated NVIDIA GeForce GTX 1660 Ti (NVIDIA; 31.0.15.1640) - AMD Ryzen 7 4800H with Radeon Graphics (16 Threads)
### Issue description
When you use SetDelay() with a Tweener the initial value that gets interpolated is the value at the start of the delay, rather than the value at the start of the tweening. If the Tween has been set to run in parallel using SetParallel(true) then this essentially has the same effect as if all the Tweeners were created with FromCurrent(). If the Tween is not set to run in parallel then the Tweeners use the initial value from the end of the last tween that run (ie before the delay) rather than the value when they actually start (ie. after the delay).
I would expect the same behavior from running two tweens in series and running two tweens in parallel where one has a delay set that's equal to the duration of the other one. The documentation isn't 100% clear on whether the intended behavior is to tween from the start of the tweening or from the start of the delay, but this definitely feels like the less natural way for this to work, and it makes a number of things rather more difficult.
An example of where this becomes very important is in a turn based game where a move triggers a series of events where various objects are tweened to a series of new positions (eg. slide, then fall). You could use a different tween to move each object (on non-parallel mode), but then if the turn is interrupted (eg. by an undo command during the turn) then you need to keep track of all the running tweens and stop them one-by-one. Another wat to make this work currently would be to keep track of the previous position of every object being moved and manually feed in that using From(), but this is also a lot of extra overhead.
### Steps to reproduce
Run minimal reproduction project. Left click anywhere and the white square will be tweened to a new position, then a 1 second delay, then tweened back to its original position. Now try doing this again but right-click during the delay and the square will be moved right by 100 pixels. When the tween starts again after the delay it starts from the position when the delay started rather than its new position at the end of the delay.
### Minimal reproduction project
[tweener test.zip](https://github.com/godotengine/godot/files/12284031/tweener.test.zip)
| https://github.com/godotengine/godot/issues/80388 | https://github.com/godotengine/godot/pull/80702 | c28cc5d5ebf1be0b7d22eeaa008942b2e97e9ec0 | c5a7462a00a665a72c7c1a746c105e3c1c475c96 | "2023-08-07T18:28:51Z" | c++ | "2023-08-17T09:28:15Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,385 | ["scene/gui/button.cpp"] | Text in Checkboxes and CheckButtons can go out of Node bounds if "Text Overrun Behavior" is set to anything other than "Trim Nothing" | ### Godot version
v4.1.stable.official [970459615]
### System information
Godot v4.1.stable - Windows 10.0.22621 - Vulkan (Forward+) - dedicated AMD Radeon RX 7900 XTX (Advanced Micro Devices, Inc.; 31.0.21001.45002) - AMD Ryzen 7 7700X 8-Core Processor (16 Threads)
### Issue description
When using CheckBoxes and CheckButtons, the text is able to go beyond node bounds if Text Overrun Behavior is set to any option other than "Trim Nothing", seemingly because it does not account correctly for the space taken by the icon integrated in these two types of Nodes.
The expected behavior is that the text never leaves the Node Size bounds nor overlaps with the icon that is part of the CheckBox or CheckButton. It's alignment and trimming all occur within the expected bounds.
### Steps to reproduce
1. Create a _CheckBox_ or _CheckButton_ node.
2. Within the Node Properties: Button/Text Behavior/Text Overrun Behavior. Set to anything other than "Trim Nothing".
3. Write text in the button until it fills its width entirely, triggering the "trimming". Resizing the button for better visibility is recommended.
It should be visible how the text either:
A) Goes beyond the node's bounds (size).
or
B) Overlaps with the icon of the _CheckBox_ or _CheckButton_, depends on the button used for reproduction.
Which of these two cases occur depends on the text alignment and the type of button chosen, but both are caused by the same issue: trimming not accounting for the integrated icon of _CheckBox_ and _CheckButton_.
### Minimal reproduction project
[Checkbox_Text_Bug.zip](https://github.com/godotengine/godot/files/12283225/Checkbox_Text_Bug.zip)
| https://github.com/godotengine/godot/issues/80385 | https://github.com/godotengine/godot/pull/80402 | 310553f2a30a8937c02aa52d2f57838db0f4a650 | 0ce1ca4677ad4d9b5952747f7940e712d561655e | "2023-08-07T16:58:43Z" | c++ | "2023-08-08T15:00:46Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,379 | ["editor/filesystem_dock.cpp"] | Scene becomes corrupted or Godot crashes when renaming a scene that involves dependencies | ### Godot version
v4.1.stable.official [970459615]
### System information
Godot v4.1.stable - Windows 10.0.22621 - Vulkan (Forward+) - dedicated AMD Radeon RX 7900 XTX (Advanced Micro Devices, Inc.; 31.0.21001.45002) - AMD Ryzen 7 7700X 8-Core Processor (16 Threads)
### Issue description
When renaming a scene in specific conditions regarding inheritance, the scene is wrongly marked as corrupt or Godot just crashes.
The setup for the bug to happen involves having 3 scenes, one being within the other (forming a 1->2->3 order), having all 3 scenes opened in the editor, and then renaming the 3rd scene while scene 1 is open, causing the 2nd become corrupt or Godot to crash. This also occurs if the name of the folder where the 3rd scene is contained is changed, instead of the scene itself. I presume the path of the 3rd scene changing at all while the 1st scene is opened results in the bug.
Given that having multiple scenes with dependencies with each other seems to cause this bug, it is probable that his bug occurs _beyond_ the aforementioned 3-scene setup, but it is a scenario where the bug can be consistently reproduced. It could be that the bug described in issue #80324 is related if not the same as the one described here.
### Steps to reproduce
**Empty Project:**
1. Create an empty Godot project.
2. Create 3 scenes, each just being a Node. The contents of the scenes seem irrelevant beyond the dependencies explained in the following steps.
3. Make the 2nd scene a child of the 1st, and the 3rd a child of the 2nd.
4. Have all three of them opened, but have the 1st scene as the active one in the viewport.
5. Rename the 3rd scene in the File System.
**Minimal Reproduction Project:**
1. Open all 3 scenes in the project, but have the 1st scene as the active one in the viewport.
2. Rename the 3rd scene in the File System OR the folder containing it. Either works.
The 2nd scene should now be marked as corrupt, and the user will be unable to open it until Godot has been restarted. Alternatively, Godot will have crashed.
### Minimal reproduction project
[Rename_Scene_Bug.zip](https://github.com/godotengine/godot/files/12282031/Rename_Scene_Bug.zip)
| https://github.com/godotengine/godot/issues/80379 | https://github.com/godotengine/godot/pull/80503 | 475860839cd0cf2ce3c938290a7c7bbc69c3ecbd | e83fc9c03aaa598c8f8d0526b4d467786569a480 | "2023-08-07T15:26:47Z" | c++ | "2023-09-12T10:00:49Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,371 | ["drivers/vulkan/rendering_device_vulkan.cpp", "drivers/vulkan/rendering_device_vulkan.h"] | OBJ meshes get corrupted on reimport on Android editor | ### Godot version
4.1.stable
### System information
Android13
### Issue description
I'm trying to import the obj model file on Android but there seems to be some compatibility issues? You can do two tests on both the computer and the mobile side to make sure what I'm saying is true.
### Steps to reproduce
Create an obj file and import it into the Godot engine on Android.
### Minimal reproduction project
[test.zip](https://github.com/godotengine/godot/files/12279298/test.zip)
| https://github.com/godotengine/godot/issues/80371 | https://github.com/godotengine/godot/pull/84852 | 8a9e3ad8d5623bad204964795504e5cd10289dad | a9ba8695d4ac453c5edc8339371116ea28f403aa | "2023-08-07T11:48:46Z" | c++ | "2023-12-04T22:06:04Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,347 | ["editor/editor_node.cpp", "editor/plugins/canvas_item_editor_plugin.cpp", "editor/scene_tree_dock.cpp"] | Some Menu Items that Trigger a Secondary Interface are Missing an Ellipsis | ### Godot version
v4.1.1.stable.official [bd6af8e0e]
### System information
Godot v4.1.1.stable - Windows 10.0.22621 - Vulkan (Compatibility) - NVIDIA GeForce GTX 1060 (NVIDIA; 30.0.14.7311) - Intel(R) Core(TM) i7-8750H CPU @ 2.20GHz (12 Threads)
### Issue description
OK, so this is minor, but having a bit of UI experience on Windows it is a bit of a _pet peeve_. Most of Godot's menu items get this right, but some are missing it when they should have it, and most of the context menu items are definitely missing it.
To recap standard Windows UI behavior: if you have a menu item that when activated, raises a modal dialog that **_must be interacted with in order to complete the functionality indicated by the menu item_**, it should have an ellipsis at the end of it. OK, so that is a mouthful so I will provide two examples of when you do and do not include the ellipsis.
When you right-click on a Node and select "**Add Child Node**" from the context menu, the menu item should really have an ellipsis and be shown as "**Add Child Node...**" Why? Because the user **must interact** with the following modal dialog in order to complete the functionality of _adding a child node_. Also, they have the option to _cancel_ the operation entirely.
On the other hand, the ubiquitous **Help --> About** menu item generally does NOT get an ellipsis. Why not? Because once you have activated the menu item and shown the about box, the user does not have to do anything to complete the "about" functionality. It is **_done_** - the showing of the _About_ dialog _completed_ the requested functionality. The only thing they can do is dismiss the dialog. Because the functionality is actually completed by activation of the menu item alone, there is no need for an ellipsis.
That said, most of the main menu items are correct, but there are a few like **Project --> Version Control -->***, whose subitems should likely have the ellipsis. But quite a few of the context or pop-up menu items are missing them.
Also, right-clicking on an empty scene raises a menu with **Add Node Here** and **Instantiate Scene Here**, both of which should have an ellipsis on them.
On the context menu for the Scene pane, the context menu has a few items that are missing an ellipsis including the ones mentioned above as well as **Attach Script** and **Change Type**.
### Steps to reproduce
Open Godot and open the **Project --> Version Control** submenu. Note the subitems are missing an ellipsis but should likely have them.
Create an empty 2D Scene and then right-click on the Node to raise the context menu. The **Add Child Node**, **Instantiate Child Scene**, **Attach Script**, and **Change Type** should have an ellipsis because they all raise an interface that must be used to complete the functionality.
Right-click on the empty 2D view and note that **Add Node Here** and **Instantiate Scene Here** are missing an ellipsis because both commands raise an interface that must be used to complete the functionality.
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/80347 | https://github.com/godotengine/godot/pull/80355 | 11ea4dc46647693eb41179b641b3c4f261b4fef4 | 8375f73680c742d59c7119c430e64a307f52af26 | "2023-08-06T18:37:28Z" | c++ | "2023-08-09T15:46:26Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,333 | ["editor/editor_plugin_settings.cpp"] | make_mesh_previews() crashes Godot | ### Godot version
4.1.1
### System information
Windows 11 - Godot 4.1.1
### Issue description
Using EditorPlugin.make_mesh_previews() crashes Godot.
The MRP contains a minimal EditorPlugin.
### Steps to reproduce
```
@tool
extends EditorPlugin
func _ready():
var _mesh = preload("res://box.tres")
var _mesh_arr : Array[Mesh]
_mesh_arr.append(_mesh)
var _2d_array : Array[Texture2D] = get_editor_interface().make_mesh_previews(_mesh_arr, 128)
```
### Minimal reproduction project
[MakeMeshPreview.zip](https://github.com/godotengine/godot/files/12269663/MakeMeshPreview.zip)
| https://github.com/godotengine/godot/issues/80333 | https://github.com/godotengine/godot/pull/81121 | 356624ce6d59682fa874487bbad43fe163d4ed42 | 102f42a48dd28ee5de9ad850827f00ef0679549e | "2023-08-06T13:08:01Z" | c++ | "2023-09-06T12:49:09Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,302 | ["modules/noise/noise_texture_3d.cpp", "modules/noise/noise_texture_3d.h"] | Editor crashes when trying to create a NoiseTexture3D in Compatibility renderer | ### Godot version
4.1.1
### System information
Godot v4.1.1.stable unknown - Arch Linux #1 SMP PREEMPT_DYNAMIC Thu, 27 Jul 2023 22:02:18 +0000 - Vulkan (Compatibility) - AMD Radeon Graphics (raphael_mendocino, LLVM 15.0.7, DRM 3.52, 6.4.7-arch1-1) () - AMD Ryzen 5 7520U with Radeon Graphics (8 Threads)
### Issue description
When trying to create a NoiseTexture3D while in the Compatability Renderer the editor hard crashes with the following StackTrace:
```
================================================================
handle_crash: Program crashed with signal 4
Engine version: Godot Engine v4.1.1.stable.arch_linux
Dumping the backtrace. Please include this when reporting the bug to the project developer.
[1] /usr/lib/libc.so.6(+0x39ab0) [0x7f065ec4fab0] (??:0)
[2] /usr/bin/godot(+0x6537e0) [0x5563005cd7e0] (??:?)
[3] /usr/bin/godot(+0x2882300) [0x5563027fc300] (??:?)
[4] /usr/bin/godot(+0x28a7a87) [0x556302821a87] (??:?)
[5] /usr/bin/godot(+0x1f932ec) [0x556301f0d2ec] (??:?)
[6] /usr/bin/godot(+0x1f9b289) [0x556301f15289] (??:?)
[7] /usr/bin/godot(+0x1fe2b93) [0x556301f5cb93] (??:?)
[8] /usr/bin/godot(+0x21e0935) [0x55630215a935] (??:?)
[9] /usr/bin/godot(+0x1faf642) [0x556301f29642] (??:?)
[10] /usr/bin/godot(+0x21caf58) [0x556302144f58] (??:?)
[11] /usr/bin/godot(+0x21cb545) [0x556302145545] (??:?)
[12] /usr/bin/godot(+0x4cd9461) [0x556304c53461] (??:?)
[13] /usr/bin/godot(+0x18f7f48) [0x556301871f48] (??:?)
[14] /usr/bin/godot(+0x341598f) [0x55630338f98f] (??:?)
[15] /usr/bin/godot(+0x306f1ae) [0x556302fe91ae] (??:?)
[16] /usr/bin/godot(+0x30b05bc) [0x55630302a5bc] (??:?)
[17] /usr/bin/godot(+0x30b423b) [0x55630302e23b] (??:?)
[18] /usr/bin/godot(+0x30e1b27) [0x55630305bb27] (??:?)
[19] /usr/bin/godot(+0x1201969) [0x55630117b969] (??:?)
[20] /usr/bin/godot(+0x493ecc1) [0x5563048b8cc1] (??:?)
[21] /usr/bin/godot(+0x49401d1) [0x5563048ba1d1] (??:?)
[22] /usr/bin/godot(+0x1204931) [0x55630117e931] (??:?)
[23] /usr/bin/godot(+0x11758c4) [0x5563010ef8c4] (??:?)
[24] /usr/lib/libc.so.6(+0x23850) [0x7f065ec39850] (??:0)
[25] /usr/lib/libc.so.6(__libc_start_main+0x8a) [0x7f065ec3990a] (??:0)
[26] /usr/bin/godot(+0x11843e5) [0x5563010fe3e5] (??:?)
-- END OF BACKTRACE --
================================================================
```
This only happens in the Compatability Renderer and not in either of the vulkan renderers.
### Steps to reproduce
Open/Create a project using the Compatability Renderer.
Create a NoiseTexture3D through the Filesystem window.
Editor crashes
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/80302 | https://github.com/godotengine/godot/pull/80407 | 0ce1ca4677ad4d9b5952747f7940e712d561655e | ff1f50f952c4eb9f7877ef6d4b6d79dc8c947810 | "2023-08-05T17:22:01Z" | c++ | "2023-08-08T15:01:10Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,291 | ["servers/rendering/renderer_rd/renderer_canvas_render_rd.cpp"] | `POINT_SIZE` does not work in canvas item shader (Godot 3 to 4 regression) | ### Godot version
v4.1.1.stable.official [bd6af8e0e]
### System information
Godot v4.1.1.stable - Ubuntu 22.04.2 LTS 22.04 - Vulkan (Forward+) - dedicated NVIDIA GeForce GTX 980 (nvidia; 535.54.03) - Intel(R) Core(TM) i5-4670 CPU @ 3.40GHz (4 Threads)
### Issue description
I'm trying to migrate a project from Godot 3 to 4, but I'm running into an apparent showstopper: The project relies on some fancy 2d shaders based on `PRIMITIVE_POINTS` meshes in combination with adjusting the `POINT_SIZE` in the vertex shader (amongst other things, see the nice [Antialiased 2D Grid, Marker, and Arrow Shaders](https://jcgt.org/published/0003/04/01/paper.pdf) paper for motivation).
This used to work well under Godot 3, but unfortunately it looks like it is no longer possible in Godot 4. The approach relies on being able to set the `POINT_SIZE` in the vertex shader. Reduced to its minimal form:
```glsl
shader_type canvas_item;
void vertex() {
// The vertex shader should be able to control the point size.
POINT_SIZE = 10.0;
}
void fragment() {
COLOR = vec4(1.0, 0.0, 0.0, 1.0);
}
```
Using this minimal bootstrap code:
```gdscript
extends Node2D
func _ready():
var points = PackedVector2Array()
points.push_back(Vector2(10, 30))
points.push_back(Vector2(100, 100))
points.push_back(Vector2(30, 10))
var arrays = []
arrays.resize(ArrayMesh.ARRAY_MAX)
arrays[ArrayMesh.ARRAY_VERTEX] = points
var mesh = ArrayMesh.new()
mesh.add_surface_from_arrays(Mesh.PRIMITIVE_POINTS, arrays)
var mesh_instance = MeshInstance2D.new()
mesh_instance.mesh = mesh
mesh_instance.material = load("res://shader_material.tres")
add_child(mesh_instance)
```
the output in Godot 4 is:
Notice that the three points are just single pixels.
Doing the same in Godot 3 (of course replacing `PackedVector2Array` by `PoolVector2Array`) has the expected outcome, i.e., setting `POINT_SIZE` is not ignored:
### Steps to reproduce
Basically as already described above.
### Minimal reproduction project
[ShaderPointSize.zip](https://github.com/godotengine/godot/files/12266592/ShaderPointSize.zip)
Or on GitHub:
- [Godot 4 version reproduction example](https://github.com/bluenote10/GodotDebugProjects/tree/master/Godot4/ShaderPointSize) (having the issue)
- [Godot 3 version reproduction example](https://github.com/bluenote10/GodotDebugProjects/tree/master/Godot3/ShaderPointSize) (working fine) | https://github.com/godotengine/godot/issues/80291 | https://github.com/godotengine/godot/pull/80323 | 23db0240a2e8ef7edb76616be5cb169c0b06e766 | 10b1f823bc500ca5d915a1b31402b16b64b33e1a | "2023-08-05T08:37:34Z" | c++ | "2023-08-21T06:21:14Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,286 | ["servers/rendering/renderer_rd/environment/gi.cpp"] | Having multiple VoxelGIs in a scene causes Vulkan error: Did not create swapchain successfully. VK_NOT_READY and Vulkan error: Cannot submit graphics queue. VK_ERROR_DEVICE_LOST | ### Godot version
4.1.1
### System information
Windows 10, Nvidia GTX 1660Ti, Vulkan
### Issue description
Having multiple VoxelGI nodes and trying to bake them consistently causes never ending Vulkan error spam as seen in the image below. This completely stops the editor as well as the project itself and it's impossible to work on the project or play it.
For me it's just 5 VoxelGI nodes but it could be higher/lower for you. Considering the bake says it's about 8MB and my GPU has 6GB it shouldn't be not enough vram issue? Sometimes I get to bake 4 of them and it breaks while baking the fifth one. Sometimes jsut having 5 GIs in and baking only one is enough to break it. One way or another it happens everytime, no exceptions.
I've tried this on a number of older drivers as well as brand new ones (yesterday updated to the latest stable) but the issue is the same.
### Steps to reproduce
In the test project it will either be already broken or not.
1. If it's not broken then click on each VoxelGI node and bake them until you get the error, if nothing happens duplicate them and carry on until you get the error.
2. If it's already broken and you want to reproduce then create a new project, 3d scene, add couple voxelGI nodes, move them aside so they don't overla and just bake them one after another until it breaks and your project is no longer usable.
### Minimal reproduction project
[VGITestGD.zip](https://github.com/godotengine/godot/files/12266313/VGITestGD.zip)
| https://github.com/godotengine/godot/issues/80286 | https://github.com/godotengine/godot/pull/80356 | ff70ce158c0d021352ace5579912d524ef471af1 | 8a3a3aaebebf31a604c670d01476c6d9951d2da0 | "2023-08-05T06:26:09Z" | c++ | "2023-08-07T12:55:33Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,280 | ["modules/gdscript/gdscript_editor.cpp"] | Some completion options are placed too much down because there locations default to location others | ### Godot version
16a93563bf
### System information
macOS 13.3.0 - Vulkan (Forward+) - integrated Apple M1 Pro - Apple M1 Pro (8 Threads)
### Issue description
As tittle says some properties are badly located as location others whoch means they are put way below others. One of such example is the property `texture` of `Sprite2D`.
### Steps to reproduce
in any script write
```
var sprite : Sprite2D
sprite.text
```
and autocomplete at the end texture will appears below a lot of other signal and constants starting with texture which should come afterward
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/80280 | https://github.com/godotengine/godot/pull/80283 | 01758ea4cebc6bfe959be0f3f8d049887f81720f | 40363a88bde046c5330781ad6cc7cee05f1b8c8b | "2023-08-05T00:11:10Z" | c++ | "2023-08-07T12:53:10Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,266 | ["modules/gdscript/gdscript_analyzer.cpp"] | Parser Bug While Entering Code | ### Godot version
v4.1.stable.official [970459615]
### System information
Godot v4.1.stable - Windows 10.0.19045 - Vulkan (Forward+) - dedicated NVIDIA GeForce RTX 3070 Ti (NVIDIA; 31.0.15.3179) - AMD Ryzen 9 5900X 12-Core Processor (24 Threads)
### Issue description
While typing code (a match statement in case it's relevant) into the IDE, I started seeing parser errors.
### Steps to reproduce
Opened my project
Started editing code
ran project once
Continued editing code, this is when I started seeing the parser errors
In the test project (attached), it just started happening while making the const values
### Minimal reproduction project
[GodotTest.zip](https://github.com/godotengine/godot/files/12262760/GodotTest.zip)
| https://github.com/godotengine/godot/issues/80266 | https://github.com/godotengine/godot/pull/83257 | 18f12d1a1e06f88b644793ea604a2a13a321de68 | ddb7969cb65f5e068265e200a378c1f75518b2e3 | "2023-08-04T16:44:40Z" | c++ | "2023-10-16T16:44:27Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,255 | ["editor/editor_node.cpp", "editor/import/resource_importer_layered_texture.cpp", "editor/import/resource_importer_layered_texture.h", "editor/import/resource_importer_scene.cpp", "editor/import/resource_importer_scene.h", "editor/import/resource_importer_texture.cpp", "editor/import/resource_importer_texture.h"] | Random crash when trying to open a scene | ### Godot version
4.2 eb4301b
### System information
Windows 10.0.19045 - Vulkan (Forward+) - dedicated NVIDIA GeForce GTX 1060 (NVIDIA; 30.0.15.1403) - Intel(R) Core(TM) i7-7700HQ CPU @ 2.80GHz (8 Threads)
### Issue description
Sometimes when trying to open a scene from filesystem dock, the editor will crash. It happened 3 times to me already, but I don't know what exactly causes it.
```
[0] FileSystemDock::_select_file (C:\godot_source\editor\filesystem_dock.cpp:1115)
[1] FileSystemDock::_select_file (C:\godot_source\editor\filesystem_dock.cpp:1115)
[2] FileSystemDock::_tree_activate_file (C:\godot_source\editor\filesystem_dock.cpp:1187)
[3] call_with_variant_args_helper<FileSystemDock> (C:\godot_source\core\variant\binder_common.h:308)
[4] call_with_variant_args<FileSystemDock> (C:\godot_source\core\variant\binder_common.h:418)
[5] CallableCustomMethodPointer<FileSystemDock>::call (C:\godot_source\core\object\callable_method_pointer.h:105)
[6] Callable::callp (C:\godot_source\core\variant\callable.cpp:51)
[7] Object::emit_signalp (C:\godot_source\core\object\object.cpp:1073)
[8] Node::emit_signalp (C:\godot_source\scene\main\node.cpp:3577)
[9] Object::emit_signal<> (C:\godot_source\core\object\object.h:891)
[10] Tree::gui_input (C:\godot_source\scene\gui\tree.cpp:3867)
[11] Control::_call_gui_input (C:\godot_source\scene\gui\control.cpp:1806)
[12] Viewport::_gui_call_input (C:\godot_source\scene\main\viewport.cpp:1565)
[13] Viewport::_gui_input_event (C:\godot_source\scene\main\viewport.cpp:1794)
[14] Viewport::push_input (C:\godot_source\scene\main\viewport.cpp:3129)
[15] Window::_window_input (C:\godot_source\scene\main\window.cpp:1482)
[16] call_with_variant_args_helper<Window,Ref<InputEvent> const &,0> (C:\godot_source\core\variant\binder_common.h:303)
[17] call_with_variant_args<Window,Ref<InputEvent> const &> (C:\godot_source\core\variant\binder_common.h:418)
[18] CallableCustomMethodPointer<Window,Ref<InputEvent> const &>::call (C:\godot_source\core\object\callable_method_pointer.h:105)
[19] Callable::callp (C:\godot_source\core\variant\callable.cpp:51)
[20] DisplayServerWindows::_dispatch_input_event (C:\godot_source\platform\windows\display_server_windows.cpp:2612)
[21] DisplayServerWindows::_dispatch_input_events (C:\godot_source\platform\windows\display_server_windows.cpp:2577)
[22] Input::_parse_input_event_impl (C:\godot_source\core\input\input.cpp:722)
[23] Input::flush_buffered_events (C:\godot_source\core\input\input.cpp:979)
[24] DisplayServerWindows::process_events (C:\godot_source\platform\windows\display_server_windows.cpp:2296)
[25] OS_Windows::run (C:\godot_source\platform\windows\os_windows.cpp:1479)
[26] widechar_main (C:\godot_source\platform\windows\godot_windows.cpp:182)
[27] _main (C:\godot_source\platform\windows\godot_windows.cpp:204)
[28] main (C:\godot_source\platform\windows\godot_windows.cpp:218)
[29] WinMain (C:\godot_source\platform\windows\godot_windows.cpp:232)
[30] __scrt_common_main_seh (D:\a\_work\1\s\src\vctools\crt\vcstartup\src\startup\exe_common.inl:288)
[31] <couldn't map PC to fn name>
```
### Steps to reproduce
1. Open a project
2. Open a scene
I think the scene needs to be opened early enough.
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/80255 | https://github.com/godotengine/godot/pull/80377 | 8375f73680c742d59c7119c430e64a307f52af26 | 1b132b7c231453fa3c321c41013c3a8da3954b30 | "2023-08-04T14:47:23Z" | c++ | "2023-08-09T15:46:33Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,244 | ["doc/classes/MeshDataTool.xml"] | The description of get_vertex in MeshDataTool documentation can be misleading | ### Godot version
4.1.1
### System information
WIn 11
### Issue description
The description of get_vertex in MeshDataTool documentation can be misleading.
It actually returns the vertex position by index.
### Steps to reproduce
no
### Minimal reproduction project
no | https://github.com/godotengine/godot/issues/80244 | https://github.com/godotengine/godot/pull/81212 | c326914c4919955337822f7c96e361cbc5afc465 | f383249f6fecd5320fa5287e5f8bc531aada0411 | "2023-08-04T08:56:43Z" | c++ | "2023-09-02T12:31:00Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,238 | ["platform/windows/export/export_plugin.cpp"] | "Icon" and "Console Wrapper Icon" are swapped when a wrapper icon is specified (Windows) | ### Godot version
4.1.1stable
### System information
Godot v4.1.1.stable - Windows 10.0.19045 - Vulkan (Compatibility) - NVIDIA GeForce GTX 750 Ti (NVIDIA; 31.0.15.3179) - 12th Gen Intel(R) Core(TM) i5-12400F (12 Threads)
### Issue description
An incredibly minor issue - when exporting with a console wrapper icon specified it gets applied to the wrong one.
### Steps to reproduce
As you can see, the purple godot is the main icon. The gray one is the wrapper icon.
Yet when we export, they are flipped.
They both are the main one when only the main icon is specified.
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/80238 | https://github.com/godotengine/godot/pull/80357 | f6b9d445b58123a169006c34d015fdf626a5d5b9 | 08e5dea630d5671792ff9ad6e9a37258ae738310 | "2023-08-04T05:34:19Z" | c++ | "2023-08-09T13:20:28Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,236 | ["editor/debugger/editor_debugger_inspector.cpp"] | Crash when the debugger encounters FileAccess.open return is null. | ### Godot version
4.2.dev2
### System information
Godot v4.2.dev2 - Windows 10.0.19044 - GLES3 (Compatibility) - AMD Radeon(TM) Vega 8 Graphics (Advanced Micro Devices, Inc.; 31.0.14051.5006) - AMD Ryzen 5 2500U with Radeon Vega Mobile Gfx (8 Threads)
### Issue description
4.2.dev1 will not crash.
When running the following code, because `FileAccess.open`cannot create a folder, it will return a null, and the debugger will trigger a breakpoint, causing the editor to crash.
```gdscript
extends Node
func _ready() -> void:
var f := FileAccess.open("user://log/log.gd",FileAccess.READ_WRITE)
print(f.get_open_error())
#var fok:FileAccess = null # This is ok.
#print(fok.get_open_error())
```
backtrace
```
godot_windows_editor_x86_64!EncodedObjectAsID::get_object_id
godot_windows_editor_x86_64!EditorDebuggerInspector::add_stack_variable
godot_windows_editor_x86_64!ScriptEditorDebugger::_parse_message
godot_windows_editor_x86_64!ScriptEditorDebugger::_notification
godot_windows_editor_x86_64!ScriptEditorDebugger::_notificationv
godot_windows_editor_x86_64!Object::notification
godot_windows_editor_x86_64!SceneTree::_process_group
godot_windows_editor_x86_64!SceneTree::_process
godot_windows_editor_x86_64!SceneTree::process
godot_windows_editor_x86_64!Main::iteration
godot_windows_editor_x86_64!OS_Windows::run
godot_windows_editor_x86_64!widechar_main
godot_windows_editor_x86_64!_main
godot_windows_editor_x86_64!main
godot_windows_editor_x86_64!invoke_main
godot_windows_editor_x86_64!__scrt_common_main_seh
KERNEL32!BaseThreadInitThunk
ntdll!RtlUserThreadStart
```
### Steps to reproduce
1. Add the above code to the project and run it. The debugger will encounter an error and crash.
2. Test the commented code, but the editor will not crash.
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/80236 | https://github.com/godotengine/godot/pull/80256 | 835957b1fc418c5506c40aa23cc2d9b1d327e06c | 75c979e48df935b15eeb8c7f33b1f7dc50527526 | "2023-08-04T03:52:54Z" | c++ | "2023-08-09T13:20:20Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,234 | ["editor/plugins/tiles/tile_set_atlas_source_editor.cpp"] | Alternative Tile 'Plus' Icon incorrectly sized for small tile sizes | ### Godot version
v4.1.stable.mono.official [970459615]
### System information
Godot v4.1.stable.mono - Windows 10.0.22621 - Vulkan (Forward+) - dedicated GeForce GTX 1660 () - AMD Ryzen 7 2700 Eight-Core Processor (16 Threads)
### Issue description
In the TileSet module, the 'plus' icon to add an alternative tiles resizes when *interacting anywhere in the Tileset module* and obscures other alternative tiles, making them hard/impossible to select for tile sizes smaller than the default 16x16 (in my case, 8x8).
I have tried adjusting TileMap values in the inspector (and elsewhere) to see if this changes the behaviour, but it seems like the icon reverts to a 16x16 size if the TileSet Atlas' texture region is smaller than that, as sizes of 15x15, 14x14... demonstrate the overlapping issue. Larger tile sizes (e.g. 20x20, 32x32, etc.) are handled fine and the icon resizes accordingly.
It is possible to get around this by reopening the TileSet module (e.g. switching to TileMap and back) where the icon resizes correctly (even for smaller sizes), but breaks again when interacting with the module.
https://github.com/godotengine/godot/assets/1772788/01d6a942-b5fb-4a6b-8dbb-4a70280dc1ca
### Steps to reproduce
1. Navigate away from the TileSet module (so that it loses focus).
2. Navigate to the TileSet module.
3. Interact with any element that _forces the module to redraw_ (select a tile/tab, or change a property on a tile)
4. (Optional) Adjust the Tile Atlas `Texture Region Size` to a size smaller than 16x16 and repeat Steps 1-3.
### Minimal reproduction project
[AlternativeTileBugProject.zip](https://github.com/godotengine/godot/files/12256488/AlternativeTileBugProject.zip) | https://github.com/godotengine/godot/issues/80234 | https://github.com/godotengine/godot/pull/80968 | b6c4efe1b1a223c113d4c2d7a4b8ca8056bbcd85 | 6da4ad16624484398331f393b503f8b5e2888c51 | "2023-08-04T02:45:53Z" | c++ | "2023-08-25T19:26:50Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,224 | ["scene/gui/popup_menu.cpp"] | PopupMenu scrolling is unusable with keyboard/controller input | ### Godot version
4.2.dev (latest)
### System information
Arch Linux
### Issue description
Selecting items on `PopupMenu`s by using a keyboard/controller is unusable due to a visual bug. The scrolling position jumps constantly back and forth:
https://github.com/godotengine/godot/assets/6501975/a9e995e6-7a73-4027-8961-b63cb25cac25
The selection itself is working properly; only the logic for scrolling to the selected item is wrong. I've bisected it to https://github.com/godotengine/godot/pull/57692 (cc @YeldhamDev), and restoring the `scroll_to_item` code to before this PR fixes the issue:
https://github.com/godotengine/godot/pull/57692/files#diff-7ee2fe488be40211cc85edff72e3668728514546ee530987099f7dc1f057f85cR1294-R1306
### Steps to reproduce
- Have an `OptionButton` with enough options to make it scrollable.
- Attempt to scroll by holding the up/down keys to traverse items.
- Items out of the initial selection will make the scroll position jump back and forth.
### Minimal reproduction project
[Repro.zip](https://github.com/godotengine/godot/files/12252179/Repro.zip) | https://github.com/godotengine/godot/issues/80224 | https://github.com/godotengine/godot/pull/80271 | c6635b4640813cbf0e30da94c334089989c5c9d0 | c1fed539435b61fe5d6b05bee96d3280e746eebe | "2023-08-03T15:08:01Z" | c++ | "2023-10-09T13:30:57Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,221 | ["editor/plugins/tiles/tile_map_editor.cpp", "editor/plugins/tiles/tile_map_editor.h"] | TileMap editor shortcuts are not handled properly | ### Godot version
4.2 725beaa
### System information
Windows 10.0.19045 - Vulkan (Forward+) - dedicated NVIDIA GeForce GTX 1060 (NVIDIA; 30.0.15.1403) - Intel(R) Core(TM) i7-7700HQ CPU @ 2.80GHz (8 Threads)
### Issue description
When you are using TileMap editor, some shortcuts are not working. Most notable is E, which is supposed to toggle the Erase too, but instead if changes to Rotate tool in 2D editor.
The editor might needs something similar to #79529, but the buttons are not inside menu and their actions are called directly from pressed callback.
### Steps to reproduce
1. Open TileMap editor
2. Add TileSet
3. Select a tile
4. Press E to enable erase tool
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/80221 | https://github.com/godotengine/godot/pull/80317 | 0fff0b10543abb4e0efbccaf5865620060d9be77 | 8d0c9a4c872f13cd7df6100f1fa40d1d4474d386 | "2023-08-03T14:00:55Z" | c++ | "2023-08-17T09:24:39Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,201 | ["doc/classes/OptionButton.xml", "scene/gui/option_button.cpp", "scene/gui/option_button.h"] | OptionButton Item Shortcut is broken | ### Godot version
4.1.1-stable
### System information
Windows 10
### Issue description
OptionButton items shortcut is not working
| https://github.com/godotengine/godot/issues/80201 | https://github.com/godotengine/godot/pull/80203 | 12a9ed0fbf9e5be281091d913f8c1fd79b7805c5 | 0fff0b10543abb4e0efbccaf5865620060d9be77 | "2023-08-03T04:20:11Z" | c++ | "2023-08-17T09:24:14Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,186 | ["drivers/gles3/rasterizer_canvas_gles3.cpp", "drivers/gles3/rasterizer_canvas_gles3.h"] | GLES3: Godot game is not loading shadow_atlas_size defined in project settings | ### Godot version
4.1.1-stable
### System information
Windows 10
### Issue description
When the project loads in editor, all the settings defined in project settings will be loaded over the default values, but when running the game nearly 95% of them will be ignored resulting in just loading the default settings.
here it's just loading only 4 variables from project setting and i have added the fifth in this pull request https://github.com/godotengine/godot/pull/80151
| https://github.com/godotengine/godot/issues/80186 | https://github.com/godotengine/godot/pull/80151 | c72b851dfb46606d3d04d00c79abd555309cae33 | 5541567c1a8a8159b7041049a89310616afa0065 | "2023-08-02T19:55:16Z" | c++ | "2023-08-17T09:23:24Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,158 | ["doc/classes/Input.xml"] | when the mouse is moving quickly and clicked, the click event is triggered twice | ### Godot version
4.1.1
### System information
WIn 11
### Issue description
I encountered a strange problem - when the mouse is moving quickly and clicked, the click event is triggered twice.
I tried left click, right click, and middle click, they all have the same problem.
I also tried replacing the mouse but still got the same error no matter using is_action_just_pressed or is_action_just_released.
https://github.com/godotengine/godot/assets/52212525/d1ae00d0-ea1b-486e-9637-1d38c6703445
### Steps to reproduce
mouse is moving quickly and clicked
### Minimal reproduction project
[mouse.zip](https://github.com/godotengine/godot/files/12238065/mouse.zip)
| https://github.com/godotengine/godot/issues/80158 | https://github.com/godotengine/godot/pull/80185 | 8ebb3c7f8580dc03f81dd7fd35ceb0d20e8fe8a0 | 5a0c6a8d694559bf13f81476082e7914ce056a39 | "2023-08-02T08:47:31Z" | c++ | "2023-11-10T20:52:18Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,157 | ["modules/gdscript/gdscript_compiler.cpp", "modules/gdscript/tests/scripts/runtime/features/first_class_callable_and_signal.gd", "modules/gdscript/tests/scripts/runtime/features/first_class_callable_and_signal.out"] | Accessing to the signal of the Control class reports an error. | ### Godot version
4.2 (c5da2e54fe50cad34b5f93fa10538978b33af88d)
### System information
Windows 11
### Issue description
Attempting to run a scene with the following script report Identifier not found error:
This does not happen on 4.1.1 so this is a regression.
### Steps to reproduce
Create a script derived from Control and try to get access to the mouse_entered signal.
### Minimal reproduction project
[TestBug.zip](https://github.com/godotengine/godot/files/12237764/TestBug.zip)
| https://github.com/godotengine/godot/issues/80157 | https://github.com/godotengine/godot/pull/80165 | 8b6c867c81f171f24d3daca5d6eb9d07fbd09f27 | 95d71dbad1287af72d8d77bd249631d9599c9d9c | "2023-08-02T08:21:16Z" | c++ | "2023-08-03T12:46:48Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,152 | ["editor/editor_node.cpp"] | Editor deadlock on exit when using Thread/Semaphore in autoloaded tool script | ### Godot version
v4.1.1.stable.mono.official [bd6af8e0e]
### System information
Godot v4.1.1.stable.mono - macOS 13.4.1 - Vulkan (Forward+) - integrated Apple M1 - Apple M1 (8 Threads)
### Issue description
After creating a `@tool` script that is autoloaded and uses Semaphores and a Thread, when exiting the editor it freezes and stops responding on MacOS.
Editor not responding after closing the editor:
The editor must then be force quit and produces the following error information:
https://gist.github.com/Trikzon/dc3e1ed7a258fa9cc2c363b18b43c353
A full minimal reproducible product is provided, but the important snippets are:
```
@tool
extends Node
var mutex: Mutex
var semaphore: Semaphore
var thread: Thread
var exit_thread := false
func _ready():
mutex = Mutex.new()
semaphore = Semaphore.new()
exit_thread = false
thread = Thread.new()
thread.start(_thread_function)
func _thread_function():
while true:
semaphore.wait()
mutex.lock()
var should_exit = exit_thread
mutex.unlock()
if should_exit:
break
# Actual thread work goes here
func _exit_tree():
mutex.lock()
exit_thread = true
mutex.unlock()
semaphore.post()
thread.wait_to_finish()
```
At first I thought that `_exit_tree()` was not being called when the editor closed, but I found that it is because placing a `ResourceSaver.save()` in `_exit_tree()` creates a file when the editor closes.
### Steps to reproduce
Create a `@tool` script using Semaphores and a Thread and set it as an autoloading script in-editor.
The [Using Multiple Threads#Semaphores](https://docs.godotengine.org/en/stable/tutorials/performance/using_multiple_threads.html#semaphores) guide can be used as a base.
### Minimal reproduction project
[Thread Tools Test.zip](https://github.com/godotengine/godot/files/12237193/Thread.Tools.Test.zip)
Note: I use a mono build of Godot, but this does not contain any C#.
Contains an autoloaded tool script that is a slightly modified version of the [Using Multiple Threads#Semaphores](https://docs.godotengine.org/en/stable/tutorials/performance/using_multiple_threads.html#semaphores) guide.
Also contains a test scene with a tool script where changing a number in-editor will post the semaphore to see that the thread is running correctly in the background. | https://github.com/godotengine/godot/issues/80152 | https://github.com/godotengine/godot/pull/80161 | f415200323d0684942184818ab0fecdb994b54d8 | 79f6ac5cd5809e559477c97e89a2f7efb8d82ab2 | "2023-08-02T07:36:59Z" | c++ | "2023-08-02T19:37:13Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,148 | ["drivers/gles3/rasterizer_canvas_gles3.cpp", "drivers/gles3/rasterizer_canvas_gles3.h"] | GLES3 changing 2d shadow atlas size is broken | ### Godot version
4.1.1 stable
### System information
Windows 10
### Issue description
this is the default size loaded on startup 1024 which seems fine but it actually have loaded the default [2048] and ignored the size in project settings
changing it now to any size will break it
[2048]
[512]
### Steps to reproduce
change 2d shadow atlas size in project settings while using compatibility renderer.
### Minimal reproduction project
_ | https://github.com/godotengine/godot/issues/80148 | https://github.com/godotengine/godot/pull/80151 | c72b851dfb46606d3d04d00c79abd555309cae33 | 5541567c1a8a8159b7041049a89310616afa0065 | "2023-08-02T05:07:18Z" | c++ | "2023-08-17T09:23:24Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,134 | ["platform/linuxbsd/x11/display_server_x11.cpp", "platform/windows/display_server_windows.cpp", "scene/main/window.cpp"] | Random crashes from `Window::_event_callback` | ### Godot version
2992ffd25523cb8b3e0ef1b873c40115b8940178
### System information
Windows 10
### Issue description
Godot crashes randomly within a few seconds after project start when moving with the mouse.
Started after recent merge of https://github.com/godotengine/godot/pull/67791.
I can confirm that reverting the pr fixes the crashes.
### Steps to reproduce
see desc
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/80134 | https://github.com/godotengine/godot/pull/80142 | ba3fb66d71cef59c1152a13eab9491ea25953ff3 | 933281f7ec16af679a63b18746984d9e2c3e7a62 | "2023-08-01T19:18:31Z" | c++ | "2023-08-02T05:42:37Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,132 | ["editor/editor_file_system.cpp"] | MeshInstance3D loses reference to mesh resource | ### Godot version
4.1
### System information
Godot v4.1.1.stable.mono - Windows 10.0.19045 - Vulkan (Forward+) - dedicated NVIDIA GeForce RTX 2070 SUPER (NVIDIA; 31.0.15.2802) - AMD Ryzen 9 3900X 12-Core Processor (24 Threads)
### Issue description
Saving a MeshInstance3D with a mesh resource imported from a blender file will lose the reference to the mesh resource when reopening de project.
When testing I have found that the uid referencing the mesh saved in the .tscn is the same as the one int the .blend.import and not an unique one for the specific mesh
uid in .tscn: 
uid in .blend.import: 
I don't know if that is the cause of the error but I have found a way to force it to generate a unique uid:
1. select the mesh in the MeshInstance3D and "Make Unique"
2. reassign the mesh resource to the MeshInstance3D
now the .tscn has a unique uid for the mesh and remains referenced when reopening the project (this happens randomly sometimes too when working on the scene)
Related: [ #57053](https://github.com/godotengine/godot/issues/57053)
### Steps to reproduce
1. import a .blend file with a mesh (if the .blend has multiple mesh the error doesn't seems reproduce)
2. open the blender importer, set mesh paths and reimport
3. create a new MeshInstance3D node and assign the created .res as a mesh resource
4. save the scene and restart Godot.
### Minimal reproduction project
[mesh_test.zip](https://github.com/godotengine/godot/files/12232647/mesh_test.zip)
| https://github.com/godotengine/godot/issues/80132 | https://github.com/godotengine/godot/pull/85922 | e1d4b3cc07605a28f182e23d1924b6432fda2186 | e551672c2d7e0b4a484a647d5bf07acf08f59766 | "2023-08-01T18:11:00Z" | c++ | "2023-12-11T23:33:25Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,130 | ["core/string/translation.cpp"] | Changing the locale breaks remapped fonts | ### Godot version
4.1.stable.official
### System information
Arch Linux
### Issue description
I'm porting part of an old project that uses remapped font resources over to Godot 4. When the locale is changed, these fonts break, and an error is logged.
In the minimal project I've attached the error looks like:
```
E 0:00:00:0730 main.gd:4 @ _ready(): Condition "!font_owner.get_or_null(p_fonts[i])" is true. Returning: false
<C++ Source> modules/text_server_adv/text_server_adv.cpp:4031 @ _shaped_text_add_string()
<Stack Trace> main.gd:4 @ _ready()
```
In my full project I get many many errors that look like this (but I think the cause is the same):
```
E 0:00:05:0248 _font_draw_glyph: Condition "!fd" is true.
<C++ Source> modules/text_server_adv/text_server_adv.cpp:3370 @ _font_draw_glyph()
```
I remap the font resources to change the font that gets used when playing in Chinese, Japanese, or Korean. (Setting CJK fonts as fallbacks would be suboptimal, since the same unicode code point can often look different in Japanese vs Simplified Chinese.)
### Steps to reproduce
1. Download the attached minimal reproduction project.
2. Run the `main.tscn` scene.
3. Note that the text of the `RichTextLabel` node does not display.
4. Note the error that gets logged.
### Minimal reproduction project
[godot4-font-remap.zip](https://github.com/godotengine/godot/files/12232624/godot4-font-remap.zip)
| https://github.com/godotengine/godot/issues/80130 | https://github.com/godotengine/godot/pull/84873 | 70ec97aadf2be992edf7a95ecccc00f0891d82d2 | 40cdd22d570024d906c9a5795fa079342dce2197 | "2023-08-01T17:49:28Z" | c++ | "2023-11-14T11:17:39Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,120 | ["platform/android/java/lib/src/org/godotengine/godot/Godot.kt"] | Quest 2 export in 4.2.dev2 not working | ### Godot version
4.2.dev2
### System information
MacOS 13.5 - Godot 4.2.dev2 - Mobile
### Issue description
UPDATE: It is a Quest2 specific issue. I tried it using a regular renderer on a Android Samsung phone and it works fine in 4.2.dev2
Dear developers,
I was happy to see the XR action map working in 4.2.dev2. Unfortunately when I try to export to the Quest 2, the executable crashes on the Quest 2. Enabling debugging on the Quest 2 yields more information in version 4.1.1 of Godot than in 4.2.dev2 (see enclosed screenshots).
<img width="998" alt="Godot411" src="https://github.com/godotengine/godot/assets/3133763/219b3cee-65cd-4da8-8cb2-0af27376149e">
Debug messages from the quest 2 in 4.1.1
<img width="998" alt="Godot42dev2" src="https://github.com/godotengine/godot/assets/3133763/58d1af15-f66f-44f1-80fd-fd3b2cec8e80">
Debug messages from the quest 2 in 4.2.dev2
I tried several things to find more information:
- Renamed the export templates from 4.1.1 to 4.2dev2. Prevents crashing, but shows a black windowed screen on the quest 2.
- Used the changed files from 4.2dev2 (SCsub, config.py and OpenXR_api.cpp) to get the XR action map editor working in 4.1.1. This was stupid and did not work at all.
- Recompiled 4.2dev2. Didn't do anything.
- Tried different Android versions (30, 33, 34), which didn't work either.
I am not sure whether it is a MacOS specific issue. I haven't been able to test this on a Windows machine.
### Steps to reproduce
Just make a simple XR project, using a tutorial:
- A Node3D as root.
- XR Origin
- XRCamera3D pointing at the plane and capsule
- MeshInstance3D - Plane (10x10)
- MeshInstance3D - Capsule
- A simple OmniLight3D to light the scene.
Use standard white materials for the plane and capsule.
Loaded: XRTools, XR Loaders, Android build templates (need to change the version in gradle-wrapper.properties to 7.6 in the URL to get a working compile)
Settings: OpenXR Enabled in project settings, ASTC texture compression enabled
Export Settings: Android Export - Gradle build - Android target platform 33 - Godot OpenXR for Meta - XR Mode: OpenXR
Connect Quest 2 headset via USB
Enable "deploy with remote debug" in the editor
Use the run remote to run the apk on the Oculus 2.
In 4.1.1 this yields a basic scene you can navigate. In 4.2.dev2 this crashes on the headset.
### Minimal reproduction project
Just extract the following files in a OpenXR42 directory and follow the steps written earlier.
[minimal project.zip](https://github.com/godotengine/godot/files/12230070/minimal.project.zip)
| https://github.com/godotengine/godot/issues/80120 | https://github.com/godotengine/godot/pull/79681 | 79d3468246939795e82677e9f185b20875a3eaff | 9fbb349a337b76b6559c1c3aeed7379b8049202e | "2023-08-01T14:05:54Z" | c++ | "2023-07-31T19:01:25Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,118 | ["scene/animation/animation_blend_tree.cpp"] | AnimationTree: animation_started signal is always emitted twice | ### Godot version
v4.1.1.stable.official [bd6af8e0e]
### System information
Godot v4.1.1.stable - Windows 10.0.19045 - Vulkan (Forward+) - dedicated NVIDIA GeForce GTX 1060 6GB (NVIDIA; 31.0.15.3179) - Intel(R) Core(TM) i5-8500 CPU @ 3.00GHz (6 Threads)
### Issue description
When connecting the `animation_started` signal from an animation tree, the connected method is always called twice.
This only occurs with the `animation_started` signal, the `animation_finished` signal is only emitted once.
This unfortunately makes relying on the signal impossible.
### Steps to reproduce
(See minimal production project)
- Use `animation_started.connect()`
- Print something out in the connected method (the example project prints the `anim_name` parameter)
- method is called twice -> two prints in the console
### Minimal reproduction project
[anim_tree_signal_bug.zip](https://github.com/godotengine/godot/files/12229864/anim_tree_signal_bug.zip)
The animation tree travels to the 'move' animation when `ui_accept` is pressed (see `anim_tree.gd`) and has print statements inside the methods connected to both the `animation_started` and `animation_finished` signal.
| https://github.com/godotengine/godot/issues/80118 | https://github.com/godotengine/godot/pull/80367 | e71ec0b5ebe05a162bec4e8383d4105d60f56a0d | 06eade1bc274c42fa3fb9d27108056ce91780840 | "2023-08-01T13:46:40Z" | c++ | "2023-08-17T09:25:28Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,099 | ["servers/rendering/renderer_rd/forward_clustered/render_forward_clustered.cpp"] | SSIL has validation errors | ### Godot version
master [75f9c97deaa8378477441e07ea1782ecab6d2eca]
### System information
Godot v4.2.dev (262d1eaa6) - Ubuntu 20.04.6 LTS (Focal Fossa) - X11 - Vulkan (Forward+) - dedicated AMD Radeon RX 6800 XT - AMD Ryzen 9 5900X 12-Core Processor (24 Threads)
### Issue description
I get SSIL spams in the log with:
```
ERROR: VALIDATION - Message Id Number: -564812795 | Message Id Name: VUID-VkDescriptorImageInfo-imageLayout-00344
Validation Error: [ VUID-VkDescriptorImageInfo-imageLayout-00344 ] Object 0: handle = 0xdc9e5a0, type = VK_OBJECT_TYPE_COMMAND_BUFFER; | MessageID = 0xde55a405 | vkCmdDispatch: Cannot use VkImage 0xc3e06d00000048ea[RenderBuffer render_buffers/depth] (layer=0 mip=0) with specific layout VK_IMAGE_LAYOUT_SHADER_READ_ONLY_OPTIMAL that doesn't match the previous known layout VK_IMAGE_LAYOUT_DEPTH_STENCIL_ATTACHMENT_OPTIMAL. The Vulkan spec states: imageLayout must match the actual VkImageLayout of each subresource accessible from imageView at the time this descriptor is accessed as defined by the image layout matching rules (https://vulkan.lunarg.com/doc/view/1.2.198.1/linux/1.2-extensions/vkspec.html#VUID-VkDescriptorImageInfo-imageLayout-00344)
Objects - 1
Object[0] - VK_OBJECT_TYPE_COMMAND_BUFFER, Handle 231335328
Command Buffer Labels - 1
Label[0] - Downsample Depth{ 1, 1, 1, 1 }
at: _debug_messenger_callback (drivers/vulkan/vulkan_context.cpp:267)
ERROR: VALIDATION - Message Id Number: -439258052 | Message Id Name: VUID-vkCmdDispatch-None-02699
Validation Error: [ VUID-vkCmdDispatch-None-02699 ] Object 0: handle = 0x22f9750000004c6b, name = RID:572218492846533, type = VK_OBJECT_TYPE_DESCRIPTOR_SET; | MessageID = 0xe5d1743c | Descriptor set VkDescriptorSet 0x22f9750000004c6b[RID:572218492846533] encountered the following validation error at vkCmdDispatch time: Image layout specified at vkUpdateDescriptorSet* or vkCmdPushDescriptorSet* time doesn't match actual image layout at time descriptor is used. See previous error callback for specific details. The Vulkan spec states: Descriptors in each bound descriptor set, specified via vkCmdBindDescriptorSets, must be valid if they are statically used by the VkPipeline bound to the pipeline bind point used by this command (https://vulkan.lunarg.com/doc/view/1.2.198.1/linux/1.2-extensions/vkspec.html#VUID-vkCmdDispatch-None-02699)
Objects - 1
Object[0] - VK_OBJECT_TYPE_DESCRIPTOR_SET, Handle 2520174109350972523, Name "RID:572218492846533"
at: _debug_messenger_callback (drivers/vulkan/vulkan_context.cpp:267)
ERROR: VALIDATION - Message Id Number: -564812795 | Message Id Name: VUID-VkDescriptorImageInfo-imageLayout-00344
Validation Error: [ VUID-VkDescriptorImageInfo-imageLayout-00344 ] Object 0: handle = 0xd9da440, type = VK_OBJECT_TYPE_COMMAND_BUFFER; | MessageID = 0xde55a405 | vkCmdDispatch: Cannot use VkImage 0xc3e06d00000048ea[RenderBuffer render_buffers/depth] (layer=0 mip=0) with specific layout VK_IMAGE_LAYOUT_SHADER_READ_ONLY_OPTIMAL that doesn't match the previous known layout VK_IMAGE_LAYOUT_DEPTH_STENCIL_ATTACHMENT_OPTIMAL. The Vulkan spec states: imageLayout must match the actual VkImageLayout of each subresource accessible from imageView at the time this descriptor is accessed as defined by the image layout matching rules (https://vulkan.lunarg.com/doc/view/1.2.198.1/linux/1.2-extensions/vkspec.html#VUID-VkDescriptorImageInfo-imageLayout-00344)
Objects - 1
Object[0] - VK_OBJECT_TYPE_COMMAND_BUFFER, Handle 228435008
Command Buffer Labels - 1
Label[0] - Downsample Depth{ 1, 1, 1, 1 }
at: _debug_messenger_callback (drivers/vulkan/vulkan_context.cpp:267)
ERROR: VALIDATION - Message Id Number: -439258052 | Message Id Name: VUID-vkCmdDispatch-None-02699
Validation Error: [ VUID-vkCmdDispatch-None-02699 ] Object 0: handle = 0x22f9750000004c6b, name = RID:572218492846533, type = VK_OBJECT_TYPE_DESCRIPTOR_SET; | MessageID = 0xe5d1743c | Descriptor set VkDescriptorSet 0x22f9750000004c6b[RID:572218492846533] encountered the following validation error at vkCmdDispatch time: Image layout specified at vkUpdateDescriptorSet* or vkCmdPushDescriptorSet* time doesn't match actual image layout at time descriptor is used. See previous error callback for specific details. The Vulkan spec states: Descriptors in each bound descriptor set, specified via vkCmdBindDescriptorSets, must be valid if they are statically used by the VkPipeline bound to the pipeline bind point used by this command (https://vulkan.lunarg.com/doc/view/1.2.198.1/linux/1.2-extensions/vkspec.html#VUID-vkCmdDispatch-None-02699)
Objects - 1
Object[0] - VK_OBJECT_TYPE_DESCRIPTOR_SET, Handle 2520174109350972523, Name "RID:572218492846533"
at: _debug_messenger_callback (drivers/vulkan/vulkan_context.cpp:267)
ERROR: VALIDATION - Message Id Number: -564812795 | Message Id Name: VUID-VkDescriptorImageInfo-imageLayout-00344
Validation Error: [ VUID-VkDescriptorImageInfo-imageLayout-00344 ] Object 0: handle = 0xdd1e910, type = VK_OBJECT_TYPE_COMMAND_BUFFER; | MessageID = 0xde55a405 | vkCmdDispatch: Cannot use VkImage 0xc3e06d00000048ea[RenderBuffer render_buffers/depth] (layer=0 mip=0) with specific layout VK_IMAGE_LAYOUT_SHADER_READ_ONLY_OPTIMAL that doesn't match the previous known layout VK_IMAGE_LAYOUT_DEPTH_STENCIL_ATTACHMENT_OPTIMAL. The Vulkan spec states: imageLayout must match the actual VkImageLayout of each subresource accessible from imageView at the time this descriptor is accessed as defined by the image layout matching rules (https://vulkan.lunarg.com/doc/view/1.2.198.1/linux/1.2-extensions/vkspec.html#VUID-VkDescriptorImageInfo-imageLayout-00344)
Objects - 1
Object[0] - VK_OBJECT_TYPE_COMMAND_BUFFER, Handle 231860496
Command Buffer Labels - 1
Label[0] - Downsample Depth{ 1, 1, 1, 1 }
at: _debug_messenger_callback (drivers/vulkan/vulkan_context.cpp:267)
ERROR: VALIDATION - Message Id Number: -439258052 | Message Id Name: VUID-vkCmdDispatch-None-02699
Validation Error: [ VUID-vkCmdDispatch-None-02699 ] Object 0: handle = 0x22f9750000004c6b, name = RID:572218492846533, type = VK_OBJECT_TYPE_DESCRIPTOR_SET; | MessageID = 0xe5d1743c | Descriptor set VkDescriptorSet 0x22f9750000004c6b[RID:572218492846533] encountered the following validation error at vkCmdDispatch time: Image layout specified at vkUpdateDescriptorSet* or vkCmdPushDescriptorSet* time doesn't match actual image layout at time descriptor is used. See previous error callback for specific details. The Vulkan spec states: Descriptors in each bound descriptor set, specified via vkCmdBindDescriptorSets, must be valid if they are statically used by the VkPipeline bound to the pipeline bind point used by this command (https://vulkan.lunarg.com/doc/view/1.2.198.1/linux/1.2-extensions/vkspec.html#VUID-vkCmdDispatch-None-02699)
Objects - 1
Object[0] - VK_OBJECT_TYPE_DESCRIPTOR_SET, Handle 2520174109350972523, Name "RID:572218492846533"
at: _debug_messenger_callback (drivers/vulkan/vulkan_context.cpp:267)
ERROR: VALIDATION - Message Id Number: -564812795 | Message Id Name: VUID-VkDescriptorImageInfo-imageLayout-00344
Validation Error: [ VUID-VkDescriptorImageInfo-imageLayout-00344 ] Object 0: handle = 0xdd13710, type = VK_OBJECT_TYPE_COMMAND_BUFFER; | MessageID = 0xde55a405 | vkCmdDispatch: Cannot use VkImage 0xc3e06d00000048ea[RenderBuffer render_buffers/depth] (layer=0 mip=0) with specific layout VK_IMAGE_LAYOUT_SHADER_READ_ONLY_OPTIMAL that doesn't match the previous known layout VK_IMAGE_LAYOUT_DEPTH_STENCIL_ATTACHMENT_OPTIMAL. The Vulkan spec states: imageLayout must match the actual VkImageLayout of each subresource accessible from imageView at the time this descriptor is accessed as defined by the image layout matching rules (https://vulkan.lunarg.com/doc/view/1.2.198.1/linux/1.2-extensions/vkspec.html#VUID-VkDescriptorImageInfo-imageLayout-00344)
Objects - 1
Object[0] - VK_OBJECT_TYPE_COMMAND_BUFFER, Handle 231814928
Command Buffer Labels - 1
Label[0] - Downsample Depth{ 1, 1, 1, 1 }
at: _debug_messenger_callback (drivers/vulkan/vulkan_context.cpp:267)
ERROR: VALIDATION - Message Id Number: -439258052 | Message Id Name: VUID-vkCmdDispatch-None-02699
Validation Error: [ VUID-vkCmdDispatch-None-02699 ] Object 0: handle = 0x22f9750000004c6b, name = RID:572218492846533, type = VK_OBJECT_TYPE_DESCRIPTOR_SET; | MessageID = 0xe5d1743c | Descriptor set VkDescriptorSet 0x22f9750000004c6b[RID:572218492846533] encountered the following validation error at vkCmdDispatch time: Image layout specified at vkUpdateDescriptorSet* or vkCmdPushDescriptorSet* time doesn't match actual image layout at time descriptor is used. See previous error callback for specific details. The Vulkan spec states: Descriptors in each bound descriptor set, specified via vkCmdBindDescriptorSets, must be valid if they are statically used by the VkPipeline bound to the pipeline bind point used by this command (https://vulkan.lunarg.com/doc/view/1.2.198.1/linux/1.2-extensions/vkspec.html#VUID-vkCmdDispatch-None-02699)
Objects - 1
Object[0] - VK_OBJECT_TYPE_DESCRIPTOR_SET, Handle 2520174109350972523, Name "RID:572218492846533"
at: _debug_messenger_callback (drivers/vulkan/vulkan_context.cpp:267)
```
### Steps to reproduce
1. Create an empty project
2. Create a WorldEnvironment node
3. Add a custom color background (unknown if this has an effect)
4. Enable SSIL
### Minimal reproduction project
I will handle this bug myself during the weekend. I'm reporting it for tracking purposes.
[EmptyGodotProject.zip](https://github.com/godotengine/godot/files/12224010/EmptyGodotProject.zip)
| https://github.com/godotengine/godot/issues/80099 | https://github.com/godotengine/godot/pull/80315 | 0422e9e2a279744f2805265e31f2b4de50071c47 | ff70ce158c0d021352ace5579912d524ef471af1 | "2023-08-01T02:49:17Z" | c++ | "2023-08-07T12:55:10Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,098 | ["platform/windows/display_server_windows.cpp"] | "window_set_mouse_passthrough" With Transparent Window causes white flickering during runtime | ### Godot version
4.1.1.stable
### System information
Godot v4.1.1.stable - Windows 10.0.19045 - Vulkan (Forward+) - dedicated NVIDIA GeForce GTX 1080 Ti (NVIDIA; 31.0.15.3179) - Intel(R) Core(TM) i5-8600K CPU @ 3.60GHz (6 Threads)
### Issue description
https://github.com/godotengine/godot/assets/75141902/a8fa0a11-3a8b-474b-8a70-80dab030a947
When using window_set_mouse_passthrough with a transparent window in my project (a desktop Tamagotchi-like game) it causes white flickering/flashing across the newly set passthrough area (but not consistently)? I used the same setup in Godot 3.5 without flickering.
The issue is occurring in my project when the player performs an action that would update the mouse passthrough such as dragging the device around on the monitor or adding new charms/etc which change the clickable area (and is very disorienting)!
### Steps to reproduce
Set up a clickthrough area with a transparent window + change the clickable area periodically.
Display > Window settings:
### Minimal reproduction project
This project is set up the way mine is which causes the MOST flickering (setting the window area to cover all monitors), however you can **_comment out set_game_window() to only cover a normal project area_** (flickering happens way less often but still happens).
Pressing TAB while the window is selected will cause the selected area to swap between full coverage and the little red square (present for your sanity). You will likely have to click the red square to actually focus on the window!
[Test_passthrough_Flicker.zip](https://github.com/godotengine/godot/files/12222888/Test_passthrough_Flicker.zip) | https://github.com/godotengine/godot/issues/80098 | https://github.com/godotengine/godot/pull/80153 | f823bccdf889094bd4d13acf653227fba8505ce5 | dca12c2e54bc9a91c490479ec0db38e6f69d122b | "2023-07-31T23:20:39Z" | c++ | "2023-08-02T10:23:12Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,074 | ["core/variant/method_ptrcall.h", "core/variant/typed_array.h"] | GDExtension: `PtrToArg::convert()` creates unnecessary copies | ### Godot version
4.2-dev (branched off 75f9c97deaa8378477441e07ea1782ecab6d2eca)
### System information
Windows 10
### Issue description
Stumbled upon this while debugging equality between `StringName` instances (for unrelated issue #78580).
The GDExtension function pointer for equality calls `OperatorEvaluatorEqual<StringName, StringName>::ptr_evaluate()`.
This one checks `PtrToArg<A>::convert(left) == PtrToArg<B>::convert(right)`.
The function `PtrToArg::convert()` is defined as follows (substituted `m_type` with `StringName`):
```cpp
template <>
struct PtrToArg<m_type> {
_FORCE_INLINE_ static StringName convert(const void *p_ptr) {
return *reinterpret_cast<const StringName *>(p_ptr);
}
}
```
`reinterpret_cast` returns an expression of type `const StringName*`.
Dereferencing that yields `const StringName&` (and **not `StringName`**).
The return type of `convert()` however is `StringName` (by value).
Unless I'm missing something, the C++ compiler cannot apply RVO here because we don't have a value, only a const-reference.
So this calls the copy-constructor `StringName::StringName(const StringName &p_name)` -- I verified this via debugger.
This copy is unnecessary for many operations which only have read access, like equality here. Even if `StringName` is cheap to copy, this still comes with refcount inc/dec + locking, when it would mostly need to compare two hashes.
### Steps to reproduce
Equality-compare two `StringName` objects in a GDExtension binding and step in via debugger. Copy constructor of `StringName` is invoked.
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/80074 | https://github.com/godotengine/godot/pull/80075 | c63e0aefcf9d29dc4684efe893aceea6efac2bb6 | 61df1deb61b97028087dcbca8e68c82d0756d5f1 | "2023-07-30T20:57:51Z" | c++ | "2023-09-17T12:46:13Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,041 | ["scene/gui/code_edit.cpp"] | The engine crashes after changing the path to the node | ### Godot version
v4.1.1-stable_mono_win64
### System information
Windows 11
### Issue description
I removed the node to replace it with another, but when I change the path in the code, the engine crashes for me.
### Steps to reproduce
I did a screen recording as it happens:
https://github.com/godotengine/godot/assets/116289408/b2fd2689-8d82-44e8-8e02-be01674842e1
### Minimal reproduction project
[MixApo.zip](https://github.com/godotengine/godot/files/12208403/MixApo.zip)
| https://github.com/godotengine/godot/issues/80041 | https://github.com/godotengine/godot/pull/80472 | 6eed73788b6b8540c944ec0e0bcdf35f54498570 | 850b9b1f40563dab392babacf1fa3980a5a1d87d | "2023-07-30T02:15:12Z" | c++ | "2023-08-10T12:55:09Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,015 | ["editor/editor_node.cpp", "editor/plugins/script_editor_plugin.cpp", "editor/plugins/script_editor_plugin.h"] | If the script editor is floating, auto switching to 2D view when clicking Control nodes doesn't work | ### Godot version
v4.1.1.stable.official [bd6af8e0e]
### System information
Godot v4.1.1.stable - Ubuntu 23.04 23.04 - Vulkan (Forward+) - dedicated NVIDIA GeForce RTX 3070 Ti (nvidia; 535.54.03) - 12th Gen Intel(R) Core(TM) i9-12900F (24 Threads)
### Issue description
As per the issue title - in a 3D scene which contains a Control node, clicking that node will switch to the 2D view and allow you to edit the controls visually. Likewise, clicking a node that derives from Node3D will switch back to the 3D view so you can continue to visually work with the selected nodes.
When going into the script window and clicking the button to open the script into a floating window, however, the aforementioned actions no longer work. Closing the script window so it is no longer floating will restore the functionality.
### Steps to reproduce
1. Create a 3D scene
2. Add a Node3D
3. Add a Control
4. Click the Node3D node and see that the editor shows the 3D view
5. Click the Control and see that the editor shows the 2D view
6. Go to the Script tab and click the button to open it as a floating window
7. Repeat steps 4 and 5 to see that either the 3D view or 2D view will remain the active view regardless of which node is selected
8. Close the floating script window and repeat steps 4 and 5 to see they are now working again
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/80015 | https://github.com/godotengine/godot/pull/80647 | 04306777a7c75d49bfb47f484cda29ef939a92b7 | c7356c795b3c8cec2453061299e85cb315a9a6a7 | "2023-07-29T14:35:12Z" | c++ | "2023-08-17T09:27:04Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,011 | ["doc/classes/Node.xml", "scene/main/viewport.cpp", "scene/main/viewport.h", "scene/main/window.cpp", "scene/main/window.h", "tests/scene/test_window.h", "tests/test_main.cpp"] | Area2D receives mouse input events from outside root viewport | ### Godot version
v4.1.1.stable.official [bd6af8e0e]
### System information
Godot v4.1.1.stable - Debian GNU/Linux trixie/sid trixie - Vulkan (Forward+) - dedicated NVIDIA GeForce GTX 1060 (nvidia) - Intel(R) Core(TM) i5-8400 CPU @ 2.80GHz (6 Threads)
### Issue description
`Area2D`s which extend beyond the bounds of the root viewport, eg. when using a `canvas_items` stretch mode, receive mouse input events even in the black bands letterboxing the game:
https://github.com/godotengine/godot/assets/441188/3beb962f-c1fa-4098-a375-f3d862013577
### Steps to reproduce
1. Set project stretch mode to either `canvas_items` or `viewport`
2. Have an `Area2D` whose `CollisionShape` extends beyond the bounds of the rendered region
3. Connect the `Area2D`'s `mouse_[entered|exited]` signal to something
4. Wave the mouse around in the black areas outside the rendered game
### Minimal reproduction project
Project used for video demo above:
[area2d-oob.zip](https://github.com/godotengine/godot/files/12207031/area2d-oob.zip)
This only demos `mouse_[entered|exited]`, but I first encountered this with the `Area2D.input_event` signal, where it received mouse clicks as well. | https://github.com/godotengine/godot/issues/80011 | https://github.com/godotengine/godot/pull/80334 | 132b97c325f8b708cab73e544c4b9da3d01271d8 | a7dc4c22a927eb27c81619720548fe2a9516c193 | "2023-07-29T12:00:23Z" | c++ | "2023-08-16T07:12:42Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,010 | [".github/workflows/web_builds.yml", "platform/web/detect.py"] | Compiling Godot web export template with LTO fails with Emscripten 3.1.42+ | ### Godot version
3.5.2
### System information
AMD EPYC 7B13; 32 GB RAM
Ubuntu 22.04.2 LTS; Python 3.11.1; Emscripten 3.1.44; clang 17; Scons 4.0.1
### Issue description
Compiling with LTO gives this error:
```
[Initial build] Compiling ==> core/bind/core_bind.cpp
[Initial build] Creating 'bin/godot.javascript.opt.engine.js'
[Initial build] Install file: "misc/dist/html/offline-export.html" as "bin/.javascript_zip/godot.offline.html"
[Initial build] Install file: "misc/dist/html/service-worker.js" as "bin/.javascript_zip/godot.service.worker.js"
[Initial build] progress_finish(["progress_finish"], [])
[Initial build] Linking Static Library ==> core/libcore.javascript.opt.a
Ranlib Library ==> core/libcore.javascript.opt.a
[Initial build] Linking Program ==> bin/godot.javascript.opt.js
cache:INFO: generating system asset: symbol_lists/df45337bb9df6876919ba6b0bb4825c9a26cbd2f.json... (this will be cached in "/workspace/web/emsdk/upstream/emscripten/cache/symbol_lists/df45337bb9df6876919ba6b0bb4825c9a26cbd2f.json" for subsequent builds)
cache:INFO: - ok
cache:INFO: generating system library: sysroot/lib/wasm32-emscripten/lto/libGL-webgl2-ofb.a... (this will be cached in "/workspace/web/emsdk/upstream/emscripten/cache/sysroot/lib/wasm32-emscripten/lto/libGL-webgl2-ofb.a" for subsequent builds)
system_libs:INFO: compiled 4 inputs
cache:INFO: - ok
cache:INFO: generating system library: sysroot/lib/wasm32-emscripten/lto/libal.a... (this will be cached in "/workspace/web/emsdk/upstream/emscripten/cache/sysroot/lib/wasm32-emscripten/lto/libal.a" for subsequent builds)
system_libs:INFO: compiled 1 inputs
cache:INFO: - ok
cache:INFO: generating system library: sysroot/lib/wasm32-emscripten/lto/libhtml5.a... (this will be cached in "/workspace/web/emsdk/upstream/emscripten/cache/sysroot/lib/wasm32-emscripten/lto/libhtml5.a" for subsequent builds)
system_libs:INFO: compiled 3 inputs
cache:INFO: - ok
cache:INFO: generating system library: sysroot/lib/wasm32-emscripten/lto/libstubs.a... (this will be cached in "/workspace/web/emsdk/upstream/emscripten/cache/sysroot/lib/wasm32-emscripten/lto/libstubs.a" for subsequent builds)
system_libs:INFO: compiled 2 inputs
cache:INFO: - ok
cache:INFO: generating system library: sysroot/lib/wasm32-emscripten/lto/libc.a... (this will be cached in "/workspace/web/emsdk/upstream/emscripten/cache/sysroot/lib/wasm32-emscripten/lto/libc.a" for subsequent builds)
system_libs:INFO: compiled 1023 inputs
cache:INFO: - ok
cache:INFO: generating system library: sysroot/lib/wasm32-emscripten/lto/libdlmalloc.a... (this will be cached in "/workspace/web/emsdk/upstream/emscripten/cache/sysroot/lib/wasm32-emscripten/lto/libdlmalloc.a" for subsequent builds)
system_libs:INFO: compiled 1 inputs
cache:INFO: - ok
cache:INFO: generating system library: sysroot/lib/wasm32-emscripten/lto/libcompiler_rt.a... (this will be cached in "/workspace/web/emsdk/upstream/emscripten/cache/sysroot/lib/wasm32-emscripten/lto/libcompiler_rt.a" for subsequent builds)
system_libs:INFO: compiled 175 inputs
cache:INFO: - ok
cache:INFO: generating system library: sysroot/lib/wasm32-emscripten/lto/libc++-noexcept.a... (this will be cached in "/workspace/web/emsdk/upstream/emscripten/cache/sysroot/lib/wasm32-emscripten/lto/libc++-noexcept.a" for subsequent builds)
system_libs:INFO: compiled 49 inputs
cache:INFO: - ok
cache:INFO: generating system library: sysroot/lib/wasm32-emscripten/lto/libc++abi-noexcept.a... (this will be cached in "/workspace/web/emsdk/upstream/emscripten/cache/sysroot/lib/wasm32-emscripten/lto/libc++abi-noexcept.a" for subsequent builds)
system_libs:INFO: compiled 16 inputs
cache:INFO: - ok
cache:INFO: generating system library: sysroot/lib/wasm32-emscripten/lto/libsockets.a... (this will be cached in "/workspace/web/emsdk/upstream/emscripten/cache/sysroot/lib/wasm32-emscripten/lto/libsockets.a" for subsequent builds)
system_libs:INFO: compiled 21 inputs
cache:INFO: - ok
wasm-ld: error: /workspace/web/emsdk/upstream/emscripten/cache/sysroot/lib/wasm32-emscripten/lto/libc.a(scalbnf.o): attempt to add bitcode file after LTO.
em++: error: '/workspace/web/emsdk/upstream/bin/wasm-ld -o bin/godot.javascript.opt.wasm platform/javascript/audio_driver_javascript.javascript.opt.o platform/javascript/godot_webgl2.javascript.opt.o platform/javascript/http_client_javascript.javascript.opt.o platform/javascript/javascript_singleton.javascript.opt.o platform/javascript/javascript_main.javascript.opt.o platform/javascript/os_javascript.javascript.opt.o platform/javascript/api/javascript_tools_editor_plugin.javascript.opt.o platform/javascript/javascript_runtime.javascript.opt.o main/libmain.javascript.opt.a modules/libmodules.javascript.opt.a platform/libplatform.javascript.opt.a drivers/libdrivers.javascript.opt.a scene/libscene.javascript.opt.a servers/libservers.javascript.opt.a core/libcore.javascript.opt.a modules/freetype/libfreetype_builtin.javascript.opt.a -L/workspace/web/emsdk/upstream/emscripten/cache/sysroot/lib/wasm32-emscripten/lto -lGL-webgl2-ofb -lal -lhtml5 -lstubs -lc -ldlmalloc -lcompiler_rt -lc++-noexcept -lc++abi-noexcept -lsockets -mllvm -combiner-global-alias-analysis=false -mllvm -enable-emscripten-sjlj -mllvm -disable-lsr /tmp/tmpk0a0yca9libemscripten_js_symbols.so --strip-debug --export-if-defined=main --export-if-defined=__start_em_asm --export-if-defined=__stop_em_asm --export-if-defined=__start_em_lib_deps --export-if-defined=__stop_em_lib_deps --export-if-defined=__start_em_js --export-if-defined=__stop_em_js --export-if-defined=__main_argc_argv --export-if-defined=fflush --export=stackSave --export=stackRestore --export=stackAlloc --export=__errno_location --export=__get_temp_ret --export=__set_temp_ret --export=__funcs_on_exit --export=__wasm_call_ctors --export-table -z stack-size=65536 --initial-memory=33554432 --no-entry --max-memory=2147483648 --global-base=1024' failed (returned 1)
scons: *** [bin/godot.javascript.opt.js] Error 1
scons: building terminated because of errors
```
### Steps to reproduce
1. Install the requirements listed on https://docs.godotengine.org/en/3.6/development/compiling/compiling_for_web.html.
2. Run `scons platform=javascript tools=no target=release use_lto=yes`
### Does anyone else have this issue?
### Is it something related to a newer version of Emscripten? Python? Scons?
### I did not have this issue before. I am using a fresh build setup. | https://github.com/godotengine/godot/issues/80010 | https://github.com/godotengine/godot/pull/81340 | 0a7f75ec7b465604b6496c8f5f1d638aed250d6d | bceac8c34f2cd6f9f156b3b1b9cbd012bc45d928 | "2023-07-29T11:30:18Z" | c++ | "2023-09-06T11:50:39Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,004 | ["editor/pot_generator.cpp"] | Impossible to generate well formated POT files when a `String` with `\"` is used | ### Godot version
v4.1.stable.official [970459615]
### System information
Godot v4.1.stable - Windows 10.0.22621 - Vulkan (Forward+) - dedicated NVIDIA GeForce GTX 1050 (NVIDIA; 31.0.15.3667) - Intel(R) Core(TM) i5-8300H CPU @ 2.30GHz (8 Threads)
### Issue description
When generating a POT file using Godot Editor, if the `String` contains double quotes, its entry in the POT file will be wrong (see below).
### Steps to reproduce
Create this script:
```
extends Control
func _ready():
var a: String = tr("\"Problematic\" string")
print(a)
```
And then generate a POT file from it. You will get something like:
```
# LANGUAGE translation for Bugs for the following files:
# res://main.gd
#
# FIRST AUTHOR < EMAIL @ADDRESS>, YEAR.
#
#, fuzzy
msgid ""
msgstr ""
"Project-Id-Version: Bugs\n"
"MIME-Version: 1.0\n"
"Content-Type: text/plain; charset=UTF-8\n"
"Content-Transfer-Encoding: 8-bit\n"
#: main.gd
msgid ""Problematic" string"
msgstr ""
```
The problem cames from the extra double quotes inside the string, they will raise an error when generating a PO file.
A possible solution would be to make:
```
extends Control
func _ready():
TranslationServer.set_locale("fr")
var a: String = tr("\\\"Problematic\\\" string")
print(a)
```
Therefore, the POT file:
```
# LANGUAGE translation for Bugs for the following files:
# res://main.gd
#
# FIRST AUTHOR < EMAIL @ADDRESS>, YEAR.
#
#, fuzzy
msgid ""
msgstr ""
"Project-Id-Version: Bugs\n"
"MIME-Version: 1.0\n"
"Content-Type: text/plain; charset=UTF-8\n"
"Content-Transfer-Encoding: 8-bit\n"
#: main.gd
msgid "\"Problematic\" string"
msgstr ""
```
But then, the string is not recognized. This is the PO file with an extra entry added manually:
```
# LANGUAGE translation for Bugs for the following files:
# res://main.gd
#
# FIRST AUTHOR < EMAIL @ADDRESS>, 2023.
#
msgid ""
msgstr ""
"Project-Id-Version: Bugs\n"
"MIME-Version: 1.0\n"
"Content-Type: text/plain; charset=UTF-8\n"
"Content-Transfer-Encoding: 8bit\n"
"Last-Translator: Automatically generated\n"
"Language-Team: none\n"
"Language: fr\n"
"Plural-Forms: nplurals=2; plural=(n > 1);\n"
#: main.gd
msgid "\"Problematic\" string"
msgstr "TEST0"
msgid "\\\"Problematic\\\" string"
msgstr "TEST1"
```
Then `print(a)` prints `TEST1` instead of `TEST0`.
### Minimal reproduction project
[project.zip](https://github.com/godotengine/godot/files/12207357/project.zip)
| https://github.com/godotengine/godot/issues/80004 | https://github.com/godotengine/godot/pull/80058 | c5903cf381972b1bc909cc9fa3aabd9a5f0f60ac | dfebfd10a863b6df8491933384d00f6757da0a66 | "2023-07-29T09:41:36Z" | c++ | "2023-08-01T15:36:07Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,003 | ["modules/gdscript/gdscript_analyzer.cpp", "modules/gdscript/tests/scripts/analyzer/warnings/shadowning.gd", "modules/gdscript/tests/scripts/analyzer/warnings/shadowning.out"] | Variable name shadowing class_name class doesn't trigger any error/warning | ### Godot version
v4.1.stable.official [970459615]
### System information
Godot v4.1.stable - Windows 10.0.22621 - Vulkan (Forward+) - dedicated NVIDIA GeForce GTX 1050 (NVIDIA; 31.0.15.3667) - Intel(R) Core(TM) i5-8300H CPU @ 2.30GHz (8 Threads)
### Issue description
If you define a `class_name` and then use the same name in a variable, no shadowing error/warning is triggered. This also create some bad references (see below).
### Steps to reproduce
Just do something like:
```
class_name Foo
extends Control
var Foo: Foo
```
In another script you can also do:
```
extends Node
var Foo: Foo #= Foo.new() # this raises an error if uncommented: Parse Error: Could not resolve member "Foo": Cyclic reference
func _ready():
Foo ## This refers to main.gd and not to the local variable
```
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/80003 | https://github.com/godotengine/godot/pull/80587 | 017541bcec6a6887a0e68d6eaa770163ed7b1846 | 10c3941a9a85958ec9fb2bd296a466659fb5363c | "2023-07-29T08:59:00Z" | c++ | "2023-09-26T11:44:56Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,002 | ["core/object/message_queue.cpp"] | set_deferred method is invalid | ### Godot version
4.2 Dev 2
### System information
Win10
### Issue description
An area node, and setting area.set_deferred ('monitoring', true) in gdscript,and then 4.2 Dev 1 and below are no problem.
Updating to Dev2 failed, and tests found that setting other properties was also ineffective, such as Sprite2D.set_deferred('position', Vector2(100,100))
### Steps to reproduce
add area node in scene
write the code in _read callback
```gdscript
func _ready() -> void:
area_2d.set_deferred('monitoring',true)
```
run,observe the node properties in the property inspector
### Minimal reproduction project
[test_bug.zip](https://github.com/godotengine/godot/files/12206623/test_bug.zip)
| https://github.com/godotengine/godot/issues/80002 | https://github.com/godotengine/godot/pull/80081 | 262d1eaa631e9cefc3f6f09845579cef2af37576 | 3fa8fad26b97a8af20e7996b7e17d8f23fc04b89 | "2023-07-29T08:10:26Z" | c++ | "2023-07-31T10:26:41Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 80,001 | ["editor/editor_log.cpp", "editor/editor_log.h"] | BBCode tags are closed prematurely in "print_rich()" method if the provided text contains "\n" in it | ### Godot version
4.2.dev1
### System information
Manjaro Linux (DE: Cinnamon v5.8.4) 64 bits
### Issue description
The title already explains it. Take a look at the steps to reproduce below.
### Steps to reproduce
In a GDScript `_ready()` function write this:
```print_rich("[color=red]Red line\n2nd red line[/color]")```
the second line is not red and the output is as follow:
I tested all the the possible BBCodes for the print_rich() method (as stated in the documentation), and of course they all failed.
The RichTextLabel in my projects works as expected.
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/80001 | https://github.com/godotengine/godot/pull/80476 | 5ba34564b91dfdc3b9b72e2dd6db6615aa04d4d1 | 7df393312fa8eab98231f2c335366a1ef6205367 | "2023-07-29T04:06:03Z" | c++ | "2023-08-10T12:55:56Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 79,994 | ["editor/editor_themes.cpp"] | Editor scroll bar's gutter don't reach its ends | ### Godot version
v4.2.dev2.official [da81ca62a], v4.1.1.stable.official [bd6af8e0e], v4.0.3.stable.official [5222a99f5]
### System information
N/A
### Issue description
When dragging an in-editor scrollbar there's still a visible gap even when dragged to one of its ends.
  
Comparison with v3.5.2.stable.official [170ba337a] (from Project Manager):
 
### Steps to reproduce
Try dragging any (?) scrollbar in the editor.
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/79994 | https://github.com/godotengine/godot/pull/83868 | ba79e53dd06dc9e6ce4a59e711cb72da1865eabb | a725e847ea533e09d9068f155ae59e24767d256e | "2023-07-28T17:31:24Z" | c++ | "2023-10-27T15:45:20Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 79,987 | ["scene/3d/multimesh_instance_3d.cpp", "scene/3d/multimesh_instance_3d.h", "scene/3d/voxel_gi.cpp", "scene/3d/voxelizer.cpp"] | VoxelGI dynamic meshes linger in original position in GI data after bake, have lower energy when moved. | ### Godot version
4.1.1.stable
### System information
Godot v4.1.1.stable - Windows 10.0.19045 - Vulkan (Forward+) - dedicated NVIDIA GeForce RTX 3070 Laptop GPU (NVIDIA; 31.0.15.3667) - 11th Gen Intel(R) Core(TM) i7-11800H @ 2.30GHz (16 Threads)
### Issue description
When baking VoxelGI with a dynamic GI mesh, that mesh appears to stay in place in the GI data when moved. Furthermore, the mesh appears to emit less energy in the GI solution when it does move.
### Steps to reproduce
Open the repro project
Select VoxelGI to rebake
Move the box named Dynamic and notice 1) it's original GI data stays in place and 2) it is somewhat darker.
### Minimal reproduction project
[BugProject.zip](https://github.com/godotengine/godot/files/12195684/BugProject.zip)
| https://github.com/godotengine/godot/issues/79987 | https://github.com/godotengine/godot/pull/81616 | bf46ee144d898017c0777774d0d027de1eb6cb1e | 23b1f213a8bd1e0355f0affcd7f0286a68372c8d | "2023-07-28T14:54:21Z" | c++ | "2023-10-26T13:58:29Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 79,979 | ["editor/editor_node.cpp"] | Extended class shows wrong icon | ### Godot version
4.1.1
### System information
Linux Mint, Godot 4.1.1
### Issue description
Hi,
I've extended my own class, which extends "Node2D".
Now, in the create node menu, the icon from the "Node" class is displayed instead of the icon from the extended "Node2d"-class.
Here is a screenshot to show how it looks:
But after you create the node, the correct Icon is picked:
Regards
abgenullt
### Steps to reproduce
Extend from a custom class and check the icon in the menu.
### Minimal reproduction project
[InheritTest.zip](https://github.com/godotengine/godot/files/12192148/InheritTest.zip)
| https://github.com/godotengine/godot/issues/79979 | https://github.com/godotengine/godot/pull/81336 | 2282fc5de992a278e6a4e9c279f4bf8086203b48 | 8449592d92eaeef990f5502b419d491ee3eeb7a6 | "2023-07-28T08:26:41Z" | c++ | "2023-09-06T14:10:33Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 79,944 | ["editor/editor_node.cpp"] | Error "Index p_idx = -1 is out of bounds (edited_scene.size() = 1)" emitted during opening of a Godot editor | ### Godot version
4.2.dev [202e4b2c1e7f8b25738b93d0e4d5066453d3edf3]
### System information
Godot v4.2.dev (202e4b2c1e) - SteamOS holo - X11 - Vulkan (Forward+) - integrated AMD Custom GPU 0405 (RADV VANGOGH) () - AMD Custom APU 0405 (8 Threads)
### Issue description
Opening any project with Godot COULD emit error message `Index p_idx = -1 is out of bounds (edited_scene.size() = 1).`. This error message will be visible in Godot editor as well as in the terminal (if Godot is opened from terminal).
After some debugging, I have come to following conclusion and found the root cause. Godot editor saves the editor state like the list of all opened scenes and the current edited scene in `.godot/editor/editor_layout.cfg` (project relative path). The problem is that current edited scene (represented as `current_scene` in `EditorNode` section of the configuration file) was not updated when all scenes are closed while the list of all opened scenes is properly updated. This leaves the value for `current_scene` to the value from the previous Godot editor launch:
```
[EditorNode]
open_scenes=PackedStringArray()
current_scene="res://scenes/node2.tscn"
center_split_offset=0
selected_default_debugger_tab_idx=0
selected_main_editor_idx=1
```
Godot editor, when started, will read this configuration file and open all scenes listed in `open_scenes` and select the scene tab represented by `current_scene`. Since `res://scenes/node2.tscn` is not found in the list of opened scenes, the error message is emitted:
```
ERROR: Index p_idx = -1 is out of bounds (edited_scene.size() = 1).
```
Issue was introduced in:
```
commit 454befc880f391448bbd1dce56a38897f5dd3037
Author: Alex Drozd <[email protected]>
Date: Mon Jun 12 20:00:42 2023 +0200
Prevent non-existent scene from being saved to persistent editor config
```
It is only reproducible on master branch.
### Steps to reproduce
This issue can be reproduced with ANY project by following steps:
1. Start Godot editor.
2. Open a scene (any one).
3. Quit Godot editor.
4. Start Godot editor.
5. Close the opened scene. The issue can only be triggered when you closed ALL opened scenes.
6. Quit Godot editor.
7. Start Godot editor.
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/79944 | https://github.com/godotengine/godot/pull/79945 | f15898161a6e7f186a160ae2c3ced3d075ce0867 | fbe7602bd329c9cf4dcbf836e2ad9f81544c2e0a | "2023-07-27T05:31:01Z" | c++ | "2023-07-31T19:01:43Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 79,943 | ["editor/editor_node.cpp"] | Godot editor crash with signal SIGSEGV when "Close All Tabs" is selected | ### Godot version
4.1.1.stable
### System information
Godot v4.1.1.stable (bd6af8e0e) - Freedesktop SDK 22.08 (Flatpak runtime) - Vulkan (Forward+) - integrated AMD Custom GPU 0405 (RADV VANGOGH) () - AMD Custom APU 0405 (8 Threads)
### Issue description
Closing all scene tabs in Godot results with a crash with following signature:
```
Godot Engine v4.1.1.stable.flathub.bd6af8e0e - https://godotengine.org
Vulkan API 1.3.246 - Forward+ - Using Vulkan Device #0: AMD - AMD Custom GPU 0405 (RADV VANGOGH)
handle_crash: Program crashed with signal 11
Engine version: Godot Engine v4.1.1.stable.flathub (bd6af8e0ea69167dd0627f3bd54f9105bda0f8b5)
Dumping the backtrace. Please include this when reporting the bug to the project developer.
ERROR: FATAL: Index p_index = 1 is out of bounds (size() = 0).
at: get (./core/templates/cowdata.h:155)
```
Since this is a release build which is missing debug symbols, the whole backtrace is not visible. Due to that, I have downloaded source code and build myself a debug version (latest commit on master available at the time) of Godot and got the following more insightful backtrace:
```
handle_crash: Program crashed with signal 11
Engine version: Godot Engine v4.2.dev.custom_build (6588a4a29af1621086feac0117d5d4d37af957fd)
Dumping the backtrace. Please include this when reporting the bug to the project developer.
[1] /usr/lib/libc.so.6(+0x38a00) [0x7f11002dea00] (??:0)
[2] /run/media/mmcblk0p1/godot/bin/godot.linuxbsd.editor.dev.x86_64(__dynamic_cast+0x2d) [0x55bffe1e44bd] (??:?)
[3] Node* Object::cast_to<Node>(Object*) (/opt/workspace/./core/object/object.h:764 (discriminator 1))
[4] VariantObjectClassChecker<Node*>::check(Variant const&) (/opt/workspace/./core/variant/binder_common.h:230)
[5] VariantCasterAndValidate<Node*>::cast(Variant const**, unsigned int, Callable::CallError&) (/opt/workspace/./core/variant/binder_common.h:256 (discriminator 2))
[6] void call_with_variant_args_helper<__UnexistingClass, Dictionary, Node*, 0ul, 1ul>(__UnexistingClass*, void (__UnexistingClass::*)(Dictionary, Node*), Variant const**, Callable::CallError&, IndexSequence<0ul, 1ul>) (/opt/workspace/./core/variant/binder_common.h:303 (discriminator 4))
[7] void call_with_variant_args_dv<__UnexistingClass, Dictionary, Node*>(__UnexistingClass*, void (__UnexistingClass::*)(Dictionary, Node*), Variant const**, int, Callable::CallError&, Vector<Variant> const&) (/opt/workspace/./core/variant/binder_common.h:451)
[8] MethodBindT<Dictionary, Node*>::call(Object*, Variant const**, int, Callable::CallError&) const (/opt/workspace/./core/object/method_bind.h:331)
[9] Object::callp(StringName const&, Variant const**, int, Callable::CallError&) (/opt/workspace/core/object/object.cpp:739 (discriminator 1))
[10] Callable::callp(Variant const**, int, Variant&, Callable::CallError&) const (/opt/workspace/core/variant/callable.cpp:62)
[11] CallQueue::_call_function(Callable const&, Variant const*, int, bool) (/opt/workspace/core/object/message_queue.cpp:220)
[12] CallQueue::flush() (/opt/workspace/core/object/message_queue.cpp:326)
[13] SceneTree::physics_process(double) (/opt/workspace/scene/main/scene_tree.cpp:473)
[14] Main::iteration() (/opt/workspace/main/main.cpp:3396)
[15] OS_LinuxBSD::run() (/opt/workspace/platform/linuxbsd/os_linuxbsd.cpp:912)
[16] /run/media/mmcblk0p1/godot/bin/godot.linuxbsd.editor.dev.x86_64(main+0x19f) [0x55bff8d0f4e8] (/opt/workspace/platform/linuxbsd/godot_linuxbsd.cpp:76)
[17] /usr/lib/libc.so.6(+0x23290) [0x7f11002c9290] (??:0)
[18] /usr/lib/libc.so.6(__libc_start_main+0x8a) [0x7f11002c934a] (??:0)
[19] /run/media/mmcblk0p1/godot/bin/godot.linuxbsd.editor.dev.x86_64(_start+0x25) [0x55bff8d0f285] (??:?)
-- END OF BACKTRACE --
```
Based on the fact that I am a total beginner with Godot code base, and that I know that the Godot 4.0.3 doesn't have this issue, and the fact that this is trivial to reproduce, I have decided to use `git bisect` to find the first bad commit which is:
```
commit ddc856702264cf892cea6e19f57d750ab824952a
Author: kobewi <[email protected]>
Date: Fri Oct 14 20:41:05 2022 +0200
Scene tab closing refactor
```
This commit from Kobe helped me tremendously to focus on just a small code path. Together with this information and a couple of sessions with GDB, I have found the following:
```
Thread 1 "godot.linuxbsd." received signal SIGSEGV, Segmentation fault.
0x000055555d3a4e3d in __dynamic_cast ()
(gdb) bt
#0 0x000055555d3a4e3d in __dynamic_cast ()
#1 0x0000555557d4d700 in Object::cast_to<Node> (p_object=0x555560f5eee0) at ./core/object/object.h:754
#2 0x00005555587413e6 in VariantObjectClassChecker<Node*>::check (p_variant=...) at ./core/variant/binder_common.h:220
#3 0x0000555558741290 in VariantCasterAndValidate<Node*>::cast (p_args=0x7fffffffd860, p_arg_idx=1, r_error=...) at ./core/variant/binder_common.h:246
#4 0x00005555598ab795 in call_with_variant_args_helper<__UnexistingClass, Dictionary, Node*, 0ul, 1ul> (p_instance=0x55556382d4c0,
p_method=(void (__UnexistingClass::*)(__UnexistingClass * const, Dictionary, Node *)) 0x555559839456 <EditorNode::_set_main_scene_state(Dictionary, Node*)>, p_args=0x7fffffffd860, r_error=...) at ./core/variant/binder_common.h:293
#5 0x00005555598a9ef1 in call_with_variant_args_dv<__UnexistingClass, Dictionary, Node*> (p_instance=0x55556382d4c0, p_method=(void (__UnexistingClass::*)(__UnexistingClass * const, Dictionary, Node *)) 0x555559839456 <EditorNode::_set_main_scene_state(Dictionary, Node*)>,
p_args=0x7fffffffda80, p_argcount=2, r_error=..., default_values=...) at ./core/variant/binder_common.h:440
#6 0x00005555598a491a in MethodBindT<Dictionary, Node*>::call (this=0x55557108d130, p_object=0x55556382d4c0, p_args=0x7fffffffda80, p_arg_count=2, r_error=...) at ./core/object/method_bind.h:324
#7 0x000055555ce7c423 in Object::callp (this=0x55556382d4c0, p_method=..., p_args=0x7fffffffda80, p_argcount=2, r_error=...) at core/object/object.cpp:733
#8 0x000055555cac8c13 in Callable::callp (this=0x7ffff2753af0, p_arguments=0x7fffffffda80, p_argcount=2, r_return_value=..., r_call_error=...) at core/variant/callable.cpp:62
#9 0x000055555ce76036 in MessageQueue::_call_function (this=0x555561238e60, p_callable=..., p_args=0x7ffff2753b08, p_argcount=2, p_show_error=false) at core/object/message_queue.cpp:229
#10 0x000055555ce7639d in MessageQueue::flush (this=0x555561238e60) at core/object/message_queue.cpp:275
#11 0x000055555ab31046 in SceneTree::physics_process (this=0x5555638337f0, p_time=0.016666666666666666) at scene/main/scene_tree.cpp:430
#12 0x0000555557d44ea0 in Main::iteration () at main/main.cpp:3131
#13 0x0000555557cbab78 in OS_LinuxBSD::run (this=0x7fffffffdd90) at platform/linuxbsd/os_linuxbsd.cpp:889
#14 0x0000555557cb12d8 in main (argc=4, argv=0x7fffffffe2d8) at platform/linuxbsd/godot_linuxbsd.cpp:73
(gdb) frame 9
#9 0x000055555ce76036 in MessageQueue::_call_function (this=0x555561238e60, p_callable=..., p_args=0x7ffff2753b08, p_argcount=2, p_show_error=false) at core/object/message_queue.cpp:229
(gdb) p *p_callable.method._data
$4 = {refcount = {count = {value = {<std::__atomic_base<unsigned int>> = {static _S_alignment = 4, _M_i = 8}, static is_always_lock_free = true}}}, static_count = {value = {<std::__atomic_base<unsigned int>> = {static _S_alignment = 4, _M_i = 1},
static is_always_lock_free = true}}, cname = 0x0, name = {_cowdata = {_ptr = 0x55557108a8d0 U"_set_main_scene_state"}, static _null = 0 U'\000', static invalid_node_name_characters = {_cowdata = {_ptr = 0x555560f14470 U". : @ / \" %"}, static _null = 0 U'\000',
static invalid_node_name_characters = <same as static member of an already seen type>}}, debug_references = 0, idx = 10433, hash = 704063681, prev = 0x0, next = 0x55556319f4a0}
```
From this backtrace, it can be seen that the method `_set_main_scene_state` is called deferred (on the next call to physics_process) for each of the scenes that Godot actually removed. Since a scene is removed (memory holding this scene has been de-allocated), `dynamic_cast` casts an invalid address (address that Godot now doesn't have an access to anymore) which results in SIGSEGV.
This issue is reproducible on 4.1 as well as on master branch.
### Steps to reproduce
Issue is trivial to reproduce with reproduction rate of 100%. No special project configuration is needed. The only precondition is to have multiple scene files available in a project. For your convenience, attached is a test project from which this issue can be reproduced.
Steps to reproduce this issue are following:
1. Open the attached project in Godot.
2. Open two or more scenes. More then one scene has to be open for editing, otherwise the issue will not be triggered.
3. Right click on one of the opened scene tabs and from the popup menu click on "Close All Tabs".
The last step will cause Godot to crash.
### Minimal reproduction project
[test.tar.gz](https://github.com/godotengine/godot/files/12179462/test.tar.gz) | https://github.com/godotengine/godot/issues/79943 | https://github.com/godotengine/godot/pull/79917 | 3de7dd902c3b491b92cad822eb1ce7018001c24b | f15898161a6e7f186a160ae2c3ced3d075ce0867 | "2023-07-27T05:14:50Z" | c++ | "2023-07-31T19:01:40Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 79,942 | ["scene/gui/tree.cpp"] | Tree.propagate_check() does not uncheck parents | ### Godot version
4.2 dev1
### System information
Windows 10.0.19045 - Vulkan (Forward+) - dedicated NVIDIA GeForce GTX 1060 (NVIDIA; 30.0.15.1403) - Intel(R) Core(TM) i7-7700HQ CPU @ 2.80GHz (8 Threads)
### Issue description
Calling `Tree.propagate_check()` after an item is unchecked does not properly uncheck the parents. They are left in partially checked state.
It's easy to observe in theme editor:
https://github.com/godotengine/godot/assets/2223172/71d0479c-0a5e-48d3-957f-d90379a95c22
### Steps to reproduce
1. Edit a Theme resource
2. Go to Manage Items... menu
3. Go to Import Items tab
4. Unfold something
5. Check and uncheck a child checkbox
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/79942 | https://github.com/godotengine/godot/pull/79946 | de0a06ee80794beab79a66fa2fb4489640f58e09 | 88068a0374f01f2cfed140f206429e4d36696dd5 | "2023-07-27T00:21:54Z" | c++ | "2023-08-01T18:03:43Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 79,936 | ["editor/import/resource_importer_texture.cpp"] | Godot Editor crashes whenever I try and open or create a project using the compiled master branch. | ### Godot version
4.0 master commit: 8367152723ebb96b1c799a9c6c9ac1d85178f80a
### System information
```
`-/osyhddddhyso/-` angle@Sager
.+yddddddddddddddddddy+. -----------
:yddddddddddddddddddddddddy: OS: Xubuntu 23.04 x86_64
-yddddddddddddddddddddhdddddddy- Kernel: 6.2.0-26-generic
odddddddddddyshdddddddh`dddd+ydddo Uptime: 2 hours, 14 mins
`yddddddhshdd- ydddddd+`ddh.:dddddy` Packages: 5713 (dpkg), 11 (brew), 21
sddddddy /d. :dddddd-:dy`-ddddddds Shell: bash 5.2.15
:ddddddds /+ .dddddd`yy`:ddddddddd: Resolution: 1920x1080
sdddddddd` . .-:/+ssdyodddddddddds DE: Xfce 4.18
ddddddddy `:ohddddddddd WM: Xfwm4
dddddddd. +dddddddd WM Theme: Default
sddddddy ydddddds Theme: Greybird [GTK2/3]
:dddddd+ .oddddddd: Icons: elementary-xfce-darker [GTK2/
sdddddo ./ydddddddds Terminal: xfce4-terminal
`yddddd. `:ohddddddddddy` Terminal Font: Monospace 12
oddddh/` `.:+shdddddddddddddo CPU: Intel i5-6500 (4) @ 3.600GHz
-ydddddhyssyhdddddddddddddddddy- GPU: NVIDIA GeForce GTX 1050 Ti
:yddddddddddddddddddddddddy: Memory: 6966MiB / 23975MiB
.+yddddddddddddddddddy+.
`-/osyhddddhyso/-`
```
### Issue description
When I compile the latest version of Godot Engine, and then try to run it and open a 4.1 project, or even create a new project, it crashes silently, usually without any kind of error message. I've been having this problem for almost a week now, and it has persisted across me pulling master and recompiling several times. Only once did I manage to get any kind of error message:
```
Godot Engine v4.2.dev.custom_build.29c21e020 - https://godotengine.org
OpenGL API 3.3.0 NVIDIA 525.125.06 - Compatibility - Using Device: NVIDIA - NVIDIA GeForce GTX 1050 Ti
Project is missing: /home/angle/Documents/Programming/GodotEngine/Projects/4.0/Advanced Movement System Godot (AMSG) Template/project.godot
Editing project: /home/angle/Documents/Programming/GodotEngine/Projects/4.0/Omnicraft
Godot Engine v4.2.dev.custom_build.29c21e020 - https://godotengine.org
Vulkan API 1.3.224 - Forward+ - Using Vulkan Device #0: NVIDIA - NVIDIA GeForce GTX 1050 Ti
WARNING: Invalid pipelines cache header.
at: _load_pipeline_cache (drivers/vulkan/rendering_device_vulkan.cpp:9071)
```
### Steps to reproduce
Compile the latest version of Godot from the master branch on a linux machine and then run it and try to create a new project.
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/79936 | https://github.com/godotengine/godot/pull/79954 | 4e22ce827a7cf769df4bc773802c4279b4d2297e | a0f572494e6602e0927bc3de61e9093c37da60f0 | "2023-07-26T18:16:20Z" | c++ | "2023-07-27T13:23:05Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 79,931 | ["servers/rendering/renderer_rd/forward_mobile/render_forward_mobile.cpp"] | Clear color is lighter on mobile renderer | ### Godot version
4.1.1
### System information
windows 11, mobile renderer
### Issue description
On Mobile Vulkan renderer, the clear color is 50% lighter than expected
See steps to reproduce and Minimal reproduction project below.
The box and background have the same color on Forward+ and compatibility, but not on Mobile
happens also in runtime on exported project, not only in editor preview
Images of the same scene from
compatibility:
Forward+:
Mobile:
### Steps to reproduce
- add a CSGBox3D with "unshaded" shading, and albedo of #545454
- add environment with Background = Custom color and color of #545454
Box and background are not the same color
### Minimal reproduction project
[bg_color.zip](https://github.com/godotengine/godot/files/12174822/bg_color.zip)
| https://github.com/godotengine/godot/issues/79931 | https://github.com/godotengine/godot/pull/80933 | 79e87eee1d377504a51fe8256149f62031a13b11 | aa9f3583e83338a7fa8f2c092cbb65996186c805 | "2023-07-26T16:13:43Z" | c++ | "2023-08-31T06:53:41Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 79,927 | ["scene/gui/tree.cpp", "scene/gui/tree.h"] | Too many warnings in the debugger causing editor lag especially in Chinese language and GLES3 (Compatibility) . | ### Godot version
v4.2.dev1.official [0c2144da9]
### System information
Godot v4.2.dev1 - Windows 10.0.19044 - GLES3 (Compatibility) - AMD Radeon(TM) Vega 8 Graphics (Advanced Micro Devices, Inc.; 31.0.14051.5006) - AMD Ryzen 5 2500U with Radeon Vega Mobile Gfx (8 Threads)
### Issue description
**Edit: Bug introduced between 4.1.dev2 and 4.1.dev3. In dev2, the scrollbars can be dragged very smoothly, and even if the scrollbars are at the bottom, there is no performance loss.**
When there are too many warnings in the debugger and render mode is GLES3 (Compatibility).
The lag situation of the editor is positively correlated with the position of the scroll bar. The closer the scroll bar is to the bottom, the more lagging the editor becomes.
Although the English version may be a bit lag, the situation is much better, but the Chinese version will be very poor.
In forward+ mode, both Chinese and English perform similarly, with slight performance impact.
Chinese:
https://github.com/godotengine/godot/assets/45038185/2449fedb-113e-45b6-ba6a-0fbd3652175d
English:
https://github.com/godotengine/godot/assets/45038185/ff3448e5-d08e-4077-93ac-32e180a8cf3b
### Steps to reproduce
1. Run Project.
2. Waiting for warnings of around 1000.
3. When the scroll bar in the error panel of the debugger is at the top, test the editor.
4. Test the editor when the scroll bar is at the bottom.
5. Try Chinese language mode.
6. Try foward+ and GLES3 (Compatibility).
When switching languages, you may need to use arrow keys because of other bugs.
### Minimal reproduction project
[WarningLag.zip](https://github.com/godotengine/godot/files/12174463/WarningLag.zip)
| https://github.com/godotengine/godot/issues/79927 | https://github.com/godotengine/godot/pull/79325 | bf8069e80100b27b390c8e393bba04fcec961000 | ad2295e35b755295eaddf3102567c1e27985c2f4 | "2023-07-26T15:48:31Z" | c++ | "2023-08-04T14:59:07Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 79,910 | ["scene/gui/label.cpp"] | `Label`'s `vertical_alignment` still affects the position even at minimum size | ### Godot version
4.2.dev [202e4b2]
### System information
N/A
### Issue description
Changing the property `vertical_alignment` in a `Label` still changes the position of the letters, even if said `Label` is at its minimum size, implying that it could be shrunk further:
### Steps to reproduce
Create a `Label` node, type anything in it, and the `vertical_alignment` property.
### Minimal reproduction project
N/A, trivially reproducible. | https://github.com/godotengine/godot/issues/79910 | https://github.com/godotengine/godot/pull/79913 | 7fe88c771017cbd1cf17701ce5d149dce6f111f2 | 7305318a884ebc9e8602e1fff7321bb28a900e15 | "2023-07-26T01:20:10Z" | c++ | "2023-07-26T16:40:42Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 79,896 | ["modules/mono/csharp_script.cpp", "modules/mono/csharp_script.h", "modules/mono/editor/Godot.NET.Sdk/Godot.SourceGenerators/ScriptMethodsGenerator.cs", "modules/mono/glue/GodotSharp/GodotSharp/Core/Bridge/ManagedCallbacks.cs", "modules/mono/glue/GodotSharp/GodotSharp/Core/Bridge/ScriptManagerBridge.cs", "modules/mono/mono_gd/gd_mono_cache.cpp", "modules/mono/mono_gd/gd_mono_cache.h"] | CSharp static methods not visible/accessible | ### Godot version
v4.1.stable.mono.official [970459615]
### System information
Windows 10
### Issue description
I want to load an external class that was loaded into the project via nuget package and access the static methods provided.
Accessing the static methods does not work, the methods are not found.
I am using the same approach I used in Godot3, but it seems to be broken in Godot4.
### Steps to reproduce
I build a simple test project with a given c# class containig two methods.
```
public partial class TestClass : Node
{
public static string Foo() {
return "foo";
}
public string Bar() {
return "bar";
}
}
```
And load them via `const TestClass = preload("res://TestClass.cs")`
i can access the method `Bar()` but the static method `Foo` is not accessible.
The only way around this problem is to write a wrapper method to call the static method it contains, but this is extremely inconvenient.
```
// GdUnit4 c# API wrapper
public partial class GdUnit4MonoApi : GdUnit4.GdUnit4MonoAPI
{
public bool _IsTestSuite(string classPath) {
return GdUnit4.GdUnit4MonoAPI.IsTestSuite(classPath);
}
}
```
Load the attached example to reproduce.
### Minimal reproduction project
[csharp-statics.zip](https://github.com/godotengine/godot/files/12164244/csharp-statics.zip)
| https://github.com/godotengine/godot/issues/79896 | https://github.com/godotengine/godot/pull/81783 | 6df12fe1873c7c477454bc87effe1c87a9735ddd | ce0fa4c691638949387d25fc19fcfe10898d218c | "2023-07-25T18:09:18Z" | c++ | "2023-09-20T16:55:10Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 79,869 | ["doc/classes/Tween.xml", "scene/animation/tween.cpp"] | Using `queue_free` after `Tween.Stop` causes a memory leak | ### Godot version
v4.1.1.stable.official [bd6af8e0e]
### System information
Godot v4.1.1.stable - Windows 10.0.22621 - Vulkan (Forward+) - dedicated NVIDIA GeForce RTX 3070 Ti (NVIDIA; 31.0.15.3640) - Intel(R) Core(TM) i7-10700K CPU @ 3.80GHz (16 Threads)
### Issue description
Running the following code causes a memory leak:
```
var tween = get_tree().create_tween()
tween.stop()
queue_free()
```
### Steps to reproduce
1. Create a Tween
2. Use `tween.stop()`
3. `queue_free()` the Node that created the Tween
### Minimal reproduction project
[TweenMemoryLeak.zip](https://github.com/godotengine/godot/files/12156855/TweenMemoryLeak.zip)
| https://github.com/godotengine/godot/issues/79869 | https://github.com/godotengine/godot/pull/79879 | fc1b392e085ff61a729b30722aeaa9b8bfdde13b | a581bf0a24cc2db0b56477101c0bd180333edccd | "2023-07-25T04:11:01Z" | c++ | "2023-07-25T19:51:35Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 79,862 | ["editor/plugins/node_3d_editor_plugin.cpp", "editor/plugins/node_3d_editor_plugin.h"] | Editor 3d transform snap parameters randomly become very small | ### Godot version
since 3.x, still there in 4.1.1
### System information
ubuntu
### Issue description
Resulting from I don't know what, this sometimes happen automatically:
related: #39507
maaaybe related: #77414
### Steps to reproduce
I couldn't isolate it yet.
Update: I can't reproduce it yet, but it just happened just after I pressed the manual snap button. That is:
- I checked before, the value was okay
- I toggled the manual snap
- when I tried to move a mesh, it didn't snap
- I opened the snapping settings and indeed, it was all messed up again.
The way I find out this happened is I try to snap using Ctrl, and the thing doesn't visibly snap.
### Minimal reproduction project
none | https://github.com/godotengine/godot/issues/79862 | https://github.com/godotengine/godot/pull/84049 | 868392faedad2b44c2b2aee6a392855a3ba1f4ee | f1f92bdb30f119cac2ab922b03b54b302c72d75b | "2023-07-24T16:59:38Z" | c++ | "2023-10-31T19:15:30Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 79,861 | ["modules/gdscript/gdscript_compiler.cpp", "modules/gdscript/gdscript_parser.h", "modules/gdscript/tests/scripts/runtime/features/export_group_no_name_conflict_with_properties.gd", "modules/gdscript/tests/scripts/runtime/features/export_group_no_name_conflict_with_properties.out"] | @export_category creates a variable that causes issues if another one has the same name | ### Godot version
4.1.1stable
### System information
Godot v4.1.1.stable - Windows 10.0.22621 - Vulkan (Forward+) - dedicated NVIDIA GeForce RTX 3050 Laptop GPU (NVIDIA; 31.0.15.3179) - 11th Gen Intel(R) Core(TM) i5-11300H @ 3.10GHz (8 Threads)
### Issue description
So, to make my code better I was using `@export_category("HEALTH")` to create categories in the editor.
Also, right under that, I had ```@export var HEALTH = 30; var current_health: float```.
Turns out, the `@export_category` creates a boolean variable with the same name as my health variable with caused issues when I wanted to set `current_health` to `HEALTH`.
Never in the editor this was said or implied by a console error.
I simply fixed it by removing `@export_category` even though I could have changed the name of the category instead.
Just wanted to point that out in case this wasn't intended.
### Steps to reproduce
Try the following code:
```gdscript
@export_category("NAME")
@export var NAME = 1
var my_name: int
func _ready():
my_name = NAME
func _physics_process(delta):
print(NAME)
print(my_name)
```
### Minimal reproduction project
N/A
_Bugsquad edit:_ Fix formatting. | https://github.com/godotengine/godot/issues/79861 | https://github.com/godotengine/godot/pull/78254 | 62b4643d854755b36d7c7ace0047d7a40e812f1a | efbff1369a9eb366667c1d641ed6c853bdd8c830 | "2023-07-24T16:35:48Z" | c++ | "2023-07-31T19:01:03Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 79,850 | ["editor/filesystem_dock.cpp", "editor/filesystem_dock.h"] | Crashing and breaking a scene when moved with parent scene | ### Godot version
4.1.1
### System information
Godot v4.1.1.stable - Ubuntu 22.04.2 LTS 22.04 - Vulkan (Forward+) - dedicated NVIDIA GeForce GTX 1060 3GB (nvidia; 525.125.06) - AMD Ryzen 9 3900X 12-Core Processor (24 Threads)
### Issue description
If I move inherited scene, with the scene that I had inherited from + if those two scenes are open in the editor, it crashes, and the inherited scene is breaking.
### Steps to reproduce
The project is already set up, the only thing that you need to do, is to open two scenes and then select them both and try to drug them to "directory" folder, and it should crash.
### Minimal reproduction project
[Bug report.zip](https://github.com/godotengine/godot/files/12144957/Bug.report.zip) | https://github.com/godotengine/godot/issues/79850 | https://github.com/godotengine/godot/pull/81657 | 25688079413707851c1cc921308b03a531bd6f25 | e5ad3d4d58fff417a9306bd802d622402ba303c7 | "2023-07-24T10:52:23Z" | c++ | "2023-09-15T17:48:33Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 79,838 | ["servers/text_server.cpp"] | Ellipsis in `Button` nodes are drawn without outlines | ### Godot version
4.2.dev [6588a4a]
### System information
N/A
### Issue description
While using `Button`s and when setting `text_overrun_behavior` to anything that adds ellipsis, they are drawn completely ignoring the outline values of the text. `Label` is able to do it correctly:
### Steps to reproduce
Create a `Button` and set `text_overrun_behavior` to anything that has ellipsis in it.
### Minimal reproduction project
N/A, trivially reproducible. | https://github.com/godotengine/godot/issues/79838 | https://github.com/godotengine/godot/pull/79844 | 0cbdad47f9fd0ad5c5161c9d91957d24abd0a25c | ee06d3e83ff16489cd4690dd972bc402ef9400af | "2023-07-23T21:47:45Z" | c++ | "2023-07-24T17:33:45Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 79,835 | ["scene/2d/physics_body_2d.cpp", "scene/2d/physics_body_2d.h", "scene/3d/physics_body_3d.cpp", "scene/3d/physics_body_3d.h"] | [GodotPhysics] Unexpected behavior while interacting with linear_velocity and state.linear_velocity. | ### Godot version
Godot 4 stable and Godot 3.x LTS
### System information
Fedora linux
### Issue description
If you use some custom logic inside _integrate_forces() to alter the state linear velocity and there is any reference to linear_velocity in physics process, the result in godot 3.5 is that it moves just a little bit, while in godot 4 is totally still.
### Steps to reproduce
For the example i created a simple scene in which i apply a force on the linear velocity in _integrate_forces() and added a line that add zero to it in physics process. You will not see any move, but it will work fine if you comment the line adding zero.
### Minimal reproduction project
[PhysicsBugReproduction.zip](https://github.com/godotengine/godot/files/12139772/PhysicsBugReproduction.zip)
| https://github.com/godotengine/godot/issues/79835 | https://github.com/godotengine/godot/pull/79977 | 88068a0374f01f2cfed140f206429e4d36696dd5 | e8106711160686219fc878e09289929472898b44 | "2023-07-23T19:10:05Z" | c++ | "2023-08-01T18:03:47Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 79,804 | ["modules/gdscript/gdscript_compiler.cpp", "modules/gdscript/gdscript_parser.h", "modules/gdscript/tests/scripts/runtime/features/export_group_no_name_conflict_with_properties.gd", "modules/gdscript/tests/scripts/runtime/features/export_group_no_name_conflict_with_properties.out"] | Using the same export_group name in an inherited class and re-saving that class while the project is running crashes the project and sometimes the editor | ### Godot version
4.1.0
### System information
Windows 10, i711800H, Forward+ Vulkan
### Issue description
When using @export_group("Name") in a BaseClass, then inheriting from that BaseClass in another class and there also using @export_group("Name"), with the same group name, then running the project and re-saving the inherited class the project crashes with
```
ERROR: FATAL: Index p_index = 3 is out of bounds (size() = 3).
at: get (./core/templates/cowdata.h:155)
```
sometimes the editor crashes too after a while.
### Steps to reproduce
1. open reproduction project
2. Run the project.
3. Open the "super"script
4. Press CTRL + S to save it.
### Minimal reproduction project
[MRP_.zip](https://github.com/godotengine/godot/files/12137412/MRP_.zip)
| https://github.com/godotengine/godot/issues/79804 | https://github.com/godotengine/godot/pull/78254 | 62b4643d854755b36d7c7ace0047d7a40e812f1a | efbff1369a9eb366667c1d641ed6c853bdd8c830 | "2023-07-22T23:35:46Z" | c++ | "2023-07-31T19:01:03Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 79,778 | ["drivers/gles3/rasterizer_scene_gles3.cpp", "drivers/gles3/storage/render_scene_buffers_gles3.cpp", "drivers/gles3/storage/render_scene_buffers_gles3.h", "servers/rendering/renderer_rd/forward_mobile/render_forward_mobile.cpp"] | Viewport background transparency not working with custom clear color in Mobile and Compatibility renderers | ### Godot version
4.1.1.stable, 4.2.dev [6588a4a]
### System information
Godot v4.1.1.stable - Windows 10.0.19045 - Vulkan (Mobile) - dedicated NVIDIA GeForce GTX 1080 Ti (NVIDIA; 31.0.15.3640) - Intel(R) Core(TM) i7-8700K CPU @ 3.70GHz (12 Threads)
### Issue description
This is related to #72602, but since this also affects the Mobile renderer, this is likely a separate issue.
Setting the "Transparent BG" setting on a viewport is supposed to always keep the background transparent. On the Forward+ renderer, this setting is respected for all Environment Background Modes . On both the Mobile and Compatibility renderers, using the "Clear Color" background mode works with the transparent background, but using a "Custom Color" background does not render it as transparent.
Using Custom Color:
Using Clear Color:
### Steps to reproduce
1. Add a SubViewport to render some 3D scene and a way to view it (eg. a Sprite2D using the viewport as its texture)
2. Enable the "Transparent BG" option on the SubViewport
3. Setup an Environment for the 3D scene being rendered to the SubViewport (either on the Camera3D node or on a WorldEnvironment node)
4. Change the Background Mode of the Environment to Custom Color.
### Minimal reproduction project
[SkyTest.zip](https://github.com/godotengine/godot/files/12135972/SkyTest.zip)
| https://github.com/godotengine/godot/issues/79778 | https://github.com/godotengine/godot/pull/79876 | 0be587521e5eeb6b3d3d057b2f3f57e9a23a2aea | 8b12849fef2059421583e4e5bf2a27f654d8ab42 | "2023-07-22T08:22:51Z" | c++ | "2023-07-31T19:01:32Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 79,769 | ["drivers/gles3/storage/mesh_storage.cpp"] | Skeleton meshes not updating in XR modes under gl_compatibility | ### Godot version
4.1.1
### System information
Meta Quest 2 - GLES3
### Issue description
When writing VR applications on Godot 4 using gl_compatibility, skeleton-driven meshes like hands are often rendered corrupted, or sometimes update slowly.
### Steps to reproduce
The godot-xr-tools demo projects have these issues in the climbing/gliding and pointer-demo scenes if the project is switched to gl_compatibility mode and run on a mobile VR headset.
### Minimal reproduction project
The godot-xr-tools demo project has been the testing project for this issue. | https://github.com/godotengine/godot/issues/79769 | https://github.com/godotengine/godot/pull/79772 | 7dbbf2eef972bd6b81013860dff7c924a5abcc2c | b1f4e58bbf338b74de25ab9181ca414c04001132 | "2023-07-21T23:11:06Z" | c++ | "2023-07-24T17:33:27Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 79,759 | ["scene/main/viewport.cpp"] | Undescriptive "affine_invert: Condition det == 0 is true" error after minimizing a borderless window resized to full screen with display/window/stretch/mode set to viewport. | ### Godot version
v4.1.1.stable.official.bd6af8e0e
### System information
Windows 10.0.19045 - Vulkan (Compatibility) - GeForce GT 540M- Intel Core i7-2670QM CPU @ 2.20GHz (8 Threads)
### Issue description
Making the root window borderless, resizing it to full screen size, and hen minimizing it with display/window/stretch/mode set to viewport in project settings, causes a weird "affine_invert: Condition det == 0 is true" error to appear from one to three times in the console.
https://github.com/godotengine/godot/assets/84321760/e753b201-d411-49d4-94fc-c2d5f331c68c
### Steps to reproduce
0. Use the minimal reproduction project.
1. Run the project.
2. Once the game resizes to full screen, defocus it to make the script minimize it.
3. Observe the error in the console.
### Minimal reproduction project
[Affine-Invert-Window-Problem-Test-4.1.1.zip](https://github.com/godotengine/godot/files/12131401/Affine-Invert-Window-Problem-Test-4.1.1.zip)
| https://github.com/godotengine/godot/issues/79759 | https://github.com/godotengine/godot/pull/79766 | 951691a9a9064fa41dbe8727960692a400753e19 | 712f49ec5770fe46207b5f72b9cbb8406190a111 | "2023-07-21T18:24:51Z" | c++ | "2023-07-26T16:40:26Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 79,751 | ["editor/scene_create_dialog.cpp", "editor/scene_create_dialog.h"] | When creating a new scene the default root node name doesn't respect the `editor/naming/node_name_casing` project setting | ### Godot version
4.1.stable
### System information
Windows 11
### Issue description
The Create New Scene dialog always defaults the scene root node name to the filename. If you're following the recommended (and default) naming conventions for filenames (snake_case) and node names (PascalCase), this default name always needs to be changed. It would be lovely if the default root node name respected the `editor/naming/node_name_casing` project setting.
Also, I have already implemented this in my fork and would be happy to contribute it. This was my "dip my toes" into Godot engine development idea since it was simple and has bothered me for a while. I consider this more of an oversight bug than a feature proposal which is why I'm opening the issue here instead of on the proposals repo.
### Steps to reproduce
1. Create and open new project
2. (only if you've changed this) In project settings set `editor/naming/node_name_casing` to PascalCase
3. In Filesystem view, right click | Create new | Scene
4. Enter a filename like "test_scene" and click Ok
Notice the root node of the new scene will also be named "test_scene" instead of "TestScene"
### Minimal reproduction project
N/A | https://github.com/godotengine/godot/issues/79751 | https://github.com/godotengine/godot/pull/79756 | 3e958cfa8ddae44fc9440ae4841308db464af4d2 | 93c69a2db620224380ad3dac4b7d8da4b17e5002 | "2023-07-21T15:49:12Z" | c++ | "2023-08-02T10:19:11Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 79,697 | ["modules/gdscript/editor/gdscript_translation_parser_plugin.cpp"] | Crash on 'Generate POT' when Dictionary-modifying Method is Present | ### Godot version
4.0 -> latest
**I've tested all this versions:**
4.0, 4.0.1, 4.0.2, 4.0.3, 4.0.4
4.1, 4.1.1
4.2 dev 4
in all this versions problem exists.
### System information
Fedora 38, macOS
### Issue description
I have found a very strange bug while generating translations. The presence of a method modifying the dictionary's content causes the editor to crash during the generation of the translation template.
The mere presence of an uninvoked method causes a crash.
**Edit: The issue occurs not only with singletons. Any script containing a method that modifies any dictionary leads to a crash.**
### Steps to reproduce
**Project -> Project Settings -> Localizations -> Generate POT -> Add (Add your singleton)**
```
extends Node
var session_data = {
"coins" : 0,
"power": 0
}
var shop_items = [
{
"id": "max_hp",
"name": tr("First Shop Item"),
"desc": tr("Great tool."),
}
]
func add_power(val: int):
session_data["power"] += val
```
_⚠️ Problem is in add_power method, not in shop_items._
**Project -> Project Settings -> Localizations -> Generate POT -> Generate**
And... Crash.
[Nagranie ekranu z 2023-09-14 13-10-42.webm](https://github.com/godotengine/godot/assets/5132385/6a00805a-34c8-408e-9b00-eb5875301312)
### Minimal reproduction project
[translation_gm_crash.zip](https://github.com/godotengine/godot/files/12104763/translation_gm_crash.zip)
_Just open project and go to: Project -> Project Settings -> Localizations -> Generate POT -> Generate_
| https://github.com/godotengine/godot/issues/79697 | https://github.com/godotengine/godot/pull/81653 | 1328367abd5d5201537496181e5c70c88a91dd68 | a79955c440b5ec65507994441f5f55a90af4fc96 | "2023-07-20T10:00:26Z" | c++ | "2023-09-14T21:41:21Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 79,689 | ["doc/classes/CPUParticles2D.xml", "doc/classes/CPUParticles3D.xml", "doc/classes/GPUParticles2D.xml", "doc/classes/GPUParticles3D.xml"] | GPUParticles3D: Emitting with oneshot + explosiveness within editor doesn't work if triggered soon after 1st time. | ## Godot version
v4.1.1.stable.mono.official [bd6af8e0e]
## System information
Godot v4.1.1.stable.mono - Windows 10.0.19044 - Vulkan (Forward+) - dedicated NVIDIA GeForce RTX 3070 Laptop GPU (NVIDIA; 31.0.15.3598) - AMD Ryzen 9 5900HX with Radeon Graphics (16 Threads)
# Issue description
### Situation
If One Shot checkbox is enabled, Emitting will uncheck itself after its lifetime. Normal procedure so far.
### Problem
As soon as Emitting boolean is automatically unchecked, hitting it again will not work (it unchecks itself again, but the particles do not show). It seems you have to wait an additional amount of time on top of its Lifetime to preview its emission properly.
It wouldn't be a problem if the particle's Lifetime was less than a second. However, it's time-consuming with long-lasting particle emissions (eg. Lifetime >= 1.5s ). I can tell this extra delay is roughly proportional to the Lifetime value.
### Example:
1. Set Lifetime to 1.0s
2. Set Emitting to true (particles spawned successfully)
3. Emitting automatically sets to false after 1.0s
4. (Wait twice its Lifetime since turning Emitting ON to expectedly emit again.)
5. Set Emitting to true (particles spawned successfully again)
6. Emitting automatically sets to false after 1.0s
7. Set Emitting to true (it fails: close to zero or no particles emitted)
The 4th step has been mandatory ever since I started with Godot 3, regardless of hardware. Note that I'm running an arguable high-end computer and this still happens. At this point, I assume this is an engine issue.
# Steps to reproduce
## Case A: Without explosiveness
1. New project
2. Create GPUParticles3D with basic setup
* Amount = 4
* Time -> One Shot = ON
* Time -> Lifetime = 1 (default)
* Process Material -> New ParticleProcessMaterial
* Draw Passes -> Pass 1 -> New QuadMesh
3. Emitting = true
4. Do it again, as soon as Emitting changes to false.
When activating Emitting **a second time**, immediately after it turns false, one of two happens:
A. Spawns one or two particles, Emitting turns false immediately.
B. No particle spawning at all, Emitting turns false abruptly.
## Case B: Explosiveness ON
1. Same steps above, with an additional in GPUParticles3D
* Time -> Explosiveness = 1 (or higher values than 0.5)
When activating Emitting a second time, while done as soon as it was turned false automatically, no particle spawns at all and Emitting turns false with approx. same delay as Lifetime value. If you don't wait at least twice as long as its Lifetime before turning Emitting ON again, it will still not emit.
# Minimal reproduction project
[Godot411-GPUParticles3D-oneshot.zip](https://github.com/godotengine/godot/files/12103554/Godot411-GPUParticles3D-oneshot.zip) | https://github.com/godotengine/godot/issues/79689 | https://github.com/godotengine/godot/pull/79720 | 0e9e373b870676b69f6378c286ad8b4960652a61 | d60f42a64ed979447c0c8b98029d6eb254ec143b | "2023-07-20T08:06:15Z" | c++ | "2023-07-25T19:26:06Z" |
closed | godotengine/godot | https://github.com/godotengine/godot | 79,684 | ["editor/export/project_export.cpp", "editor/export/project_export.h"] | Exporting specific resources is broken, includes resources it should not include | ### Godot version
4.1.1 stable
### System information
Godot v4.1.1.stable - Windows 10.0.19045 - Vulkan (Compatibility) - NVIDIA GeForce GTX 750 Ti (NVIDIA; 31.0.15.3179) - 12th Gen Intel(R) Core(TM) i5-12400F (12 Threads)
### Issue description
To start, let me explain the setup:
Here I have a nearly empty project. I've added two folders. "Large" is for resources with a large file size, and "Small" is for resources with a small file size.
Contained in the "Large" folder is two needlessly-large images, used to represent actual game assets with a justified file size.
Contained in the "Small" folder is a singular 4 kilobyte image.
Now that I have explained the project, lets export it!
We will be exporting using these four presets: (All have embedded pcks for convenience but issue does not rely on that)
**Small Images Only Dedicated Server** exports only the "Small" folder using the **"Export as dedicated server"** option.
**Everything Dedicated Server** exports the entire project, including the large folder, using the **"Export as dedicated server"** option.
**Small Images Only Desktop** exports only the small folder using **"Export selected resources (and dependencies)"**
**Everything Desktop** exports the entire project with everything selected using "**Export selected resources (and dependencies)"**
The expected file sizes should be:
Everything Desktop > Everything Server > Small Images Desktop > Small Images Server
Small Images Server the smallest because it only has the singular png file, being stripped. Small Images Desktop is next because it has only the small image, but unstripped. Then, Everything Server, because it has all of the images but they are stripped. And finally, Everything Desktop containing every resource unstripped.
But this isn't the case.
(export without embedded pck version)
Everything Desktop > Small Images Only Dedicated Server > Everything Server > Small Images Only Desktop
Having only selected a **4 kilobyte image** in the small images dedicated server, it should not be 106 mb. The everything server has a proper file size, implying the images were stripped properly.
As you can see, only a singular folder with a 4kb image is selected. Yet the file size is the same as the builds with the images unstripped and in their raw form.
The expected behavior is that the dedicated server with only one file selected should only include the data for that one file, not the rest of the project.
### Steps to reproduce
Open the project, and then go to the export project menu. Observe the selected resources for each one to ensure they match the name, and then export them and observe their file sizes.
### Minimal reproduction project
[ResourceExportExample.zip](https://github.com/godotengine/godot/files/12101349/ResourceExportExample.zip)
*Note - I have reduced the file size of the large images from 20~ mb to 6~ mb in order to upload it.
The issue still stands regardless. | https://github.com/godotengine/godot/issues/79684 | https://github.com/godotengine/godot/pull/79725 | fa54bc3eca13fb34bedfd408642ac1c2eda91e3e | ca6b404739566b918ba00a6c16b53ed27f43562d | "2023-07-20T01:52:46Z" | c++ | "2023-10-30T22:16:54Z" |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.