- Oct 07, 2021
-
-
These are Win10 builds with a maximum API support of Windows 1809 = Redstone 5. We use the LLVM "UWP" image as it's the one with UCRT runtime by default. It's not tied to UWP API's only. Only the C runtime is different compared to the win64 LLVM builds we already have.
-
The NTDDI version contains the windows version and the subversion like the Windows 10 flavor or Service Pack. This is supported by mingw64 and MSDK. https://docs.microsoft.com/en-us/windows/win32/winprog/using-the-windows-headers
-
-
Current libplacebo git master has dropped fields deprecated in libplacebo v2, but we remain backwards compatible with libplacebo v2 due to reasons of debian stable / ubuntu LTS. Rather than bumping up to libplacebo v3 (which would allow us to drop these deprecated functions entirely), simply add forward compatibility defines for now. Can be removed whenever there's a new ubuntu/debian release.
-
With the bump to the minimum version, all these #ifdefs can (mostly) be cleaned up.
-
This is packaged everywhere relevant, including Debian 11 (stable). Fedora 33 and Ubuntu 20.10. (Rolling release distros all have v4 anyways) Motivation behind this is twofold: besides the obvious code simplification, this is mainly for simpler forward compatibility with libplacebo v4+.
-
-
-
There are no needs to reset p_sys on error. (Only the callbacks need to be reset - for affected capabilities.)
-
- Oct 06, 2021
-
-
The previous ones had outdated certificates. They can no longer build contribs.
-
-
Captions used to fallback on 608 when 708 is undetected, but as probing can be asynchronous, we have no way to deal with this properly. (and re-updgrade to 708 anyway) refs #26159
-
The texture scaling factor represents how a plane is scaled compared to the first plane, so by definition, the first plane is not scaled (its ratio is 1:1). This simplification was missed by 499b7ef7.
-
- Oct 05, 2021
-
-
-
QOpenGLFramebufferObject::takeTexture is detaching the current texture and transfering ownership to the caller, and the texture was never destroyed afterwards. It was leading to a huge memory consumption and leak when using the sample app. From the documentation[^1] in Qt5: Returns the texture id for the texture attached to this framebuffer object. The ownership of the texture is transferred to the caller. If the framebuffer object is currently bound, an implicit release() will be done. During the next call to bind() a new texture will be created. [1]: https://doc.qt.io/qt-5/qopenglframebufferobject.html#takeTexture Co-authored-by:
Alexandre Janniaux <ajanni@videolabs.io> For the commit message and cleaning the fix mostly. Fixes #25867
-
- Oct 04, 2021
-
-
mainInterface.minimumHeight is no longer available, therefore use topWindow to retrieve minimum height.
-
Now that TexCoordsMap is the same for all planes, initializing the pixel vec4 in several steps is not necessary anymore. Concretely, replace: vec4 texel; vec4 pixel = vec4(0.0, 0.0, 0.0, 1.0); texel = texture2D(Textures[0], tex_coords); pixel[0] = texel.r; texel = texture2D(Textures[1], tex_coords); pixel[1] = texel.r; texel = texture2D(Textures[2], tex_coords); pixel[2] = texel.r; by: vec4 pixel = vec4( texture2D(Textures[0], tex_coords).r, texture2D(Textures[1], tex_coords).r, texture2D(Textures[2], tex_coords).r, 1.0);
-
-
-
-
This allows passing a specific value instead of inheriting these values from the build environment. Fixes #26035
-
- Oct 03, 2021
-
-
this ensure that non-empty media list is added to the playlist fix: #26139
-
- Oct 02, 2021
-
-
-
-
-
The TransformMatrix is provided by Android to apply its internal coordinates transformations to sample the texture at the correct location, so it must be applied last, after the orientation/crop from the picture_t. The fact that their internal transform is exposed as an explicit matrix is an arbitrary design choice from Android. Instead they could have exposed a GLSL function "android_texture2D(vec2 coords)": vec2 tex_coords = TexCoordsMap * OrientationMatrix * pic_coords; vec4 texel = android_texture2D(tex_coords); In that case, it is obvious that their transform would be applied last. Refs <https://developer.android.com/reference/android/graphics/SurfaceTexture#getTransformMatrix(float%5B%5D)>
-
The transform matrix provided by Android is 4x4: <https://developer.android.com/reference/android/graphics/SurfaceTexture#getTransformMatrix(float%5B%5D)> However, the third column is never used, since the input vector is in the form (s, t, 0, 1). Similarly, the third row is never used either, since only the two first coordinates of the output vector are kept. Using a 3x3 matrix simplifies the product with other 3x3 matrices. Note: in theory, a 3x2 matrix would be sufficient, but use a square matrix for wider compatibility with all OpenGL versions.
-
The orientation matrix applies a 2D affine transform, so a 3x3 matrix is sufficient.
-
-
-
-
-
-
Extending spacer widget is not compatible with the central alignment, so it makes sense to hide while this widget is being dragged
-
-
-
toolbar widgets can not be expanded if placed in player toolbar center alignment. However, this behavior was not being reflected in the toolbar editor.
-
-
-
-