80c52493ef
Includes a fix for inaccurate slice range calculation for bases with diacritics.
Also removes unnecessary ChangeLog, and updated unrelated vhacd commit to match
currently used one.
We've had many issues with WebM support and specifically the libvpx library
over the years, mostly due to its poor integration in Godot's buildsystem,
but without anyone really interested in improving this state.
With the new GDExtensions in Godot 4.0, we intend to move video decoding to
first-party extensions, and this would likely be done using something like
libvlc to expose more codecs.
Removing the `webm` module means we can remove libsimplewebm, libvpx and
opus, which we were only used for that purpose. Both libvpx and opus were
fairly complex pieces of the buildsystem, so this is a nice cleanup.
This also removes the compile-time dependency on `yasm`.
Fixes lots of compilation or non-working WebM issues which will be linked
in the PR.
Stop include Bullet headers using `-isystem` for GCC/Clang as it misleads
SCons into not properly rebuilding all files when headers change.
This means we also need to make sure Bullet builds without warning, and
current version fares fairly well, there were just a couple to fix (patch
included).
Increase minimum version for distro packages to 2.90 (this was never released
as the "next" version after 2.89 was 3.05... but that covers it too).
Fixes#43868.
(cherry picked from commit b7901c773c)
* Fixed LODs for shadow meshes.
* Added a merging step before simplification. This helps with tesselated
meshes that were previously left untouched. The angle difference at
wich edges ar considered "hard" can be tweaked as an import setting.
* LODs will now start with the highest decimation possible and keep
doubling (approximately) the number of triangles from there. This
makes sure that very low triangle counts are included when possible.
* Given more weight to normal preservation.
* Modified MeshOptimizer to report distance-based error instead of
including attributes in the reported metrics.
* Added attribute transference between the original mesh and the
various LODs. Right now only normals are taken into account,
but it could be expanded to other attributes in the future.
* Added support to our local copy of SpirV Reflect (which does not support it).
* Pass them on render or compute pipeline creation.
* Not implemented in our shaders yet.
* Shader compilation is now cached. Subsequent loads take less than a millisecond.
* Improved game, editor and project manager startup time.
* Editor uses .godot/shader_cache to store shaders.
* Game uses user://shader_cache
* Project manager uses $config_dir/shader_cache
* Options to tweak shader caching in project settings.
* Editor path configuration moved from EditorSettings to new class, EditorPaths, so it can be available early on (before shaders are compiled).
* Reworked ShaderCompilerRD to ensure deterministic shader code creation (else shader may change and cache will be invalidated).
* Added shader compression with SMOLV: https://github.com/aras-p/smol-v
Since Embree v3.13.0 supports AARCH64, switch back to the
official repo instead of using Embree-aarch64.
`thirdparty/embree/patches/godot-changes.patch` should now contain
an accurate diff of the changes done to the library.
This makes font oversampling work out of the box, while also increasing
the supported character set's size. The default font is now larger
as well to better fit today's screen resolutions.
The OpenSans SemiBold font was chosen for two reasons:
- Small file size, yet its character set supports Latin-1 and Cyrillic
text.
- A heavier font weight looks better in most "game" scenarios and is
more readable against mixed-color backgrounds.
This is considered a breaking change as it changes the default font's
metrics, which will likely affect how Control nodes are laid out in
scenes (unless a custom font is in use).
Avoid unnecessary allocation of temporary buffers for each mip, and creates
only one Image with the compressed data.
Also renames variable and reorders code for clarity.
Clarify that squish is now only used for decompression.
Documented which formats can be decompressed in Image.
We do our own image loading, threading, and memory management in Godot already,
so the only components we need from etcpak (at least as of now) are the
`Compress*` methods defined in `ProcessDxtc.cpp` and `ProcessRGB.cpp`.
So we don't need to compile or vendor the rest.
- `etc` module was renamed to `etcpak` and modified to use the new library.
- PKM importer is removed in the process, it's obsolete.
- Old library `etc2comp` is removed.
- S3TC compression no longer done via `squish` (but decompression still is).
- Slight modifications to etcpak sources for MinGW compatibility,
to fix LLVM `-Wc++11-narrowing` errors, and to allow using vendored or
system libpng.
Co-authored-by: Rémi Verschelde <rverschelde@gmail.com>
This is a tricky one, it used to work, but it was wrong, because in such
a scenario instead of passing NULL as required by the API, it would pass
a buffer containing the `\0` terminator.
This stopped working on a specific miniupnpc version, when they fixed
some network endianess issue on Windows, to which we made a workaround,
which in turn would probably result in failures when the interface is
specified.
This commit address the issue properly, by checking the specified
interface string size, and correctly passing NULL instead of the empty
string when necessary.
Also reverts the commit that introduced the bogus workaround:
e85330231c
One of those PR when the explaination is much longer then code changes
:).
Also include public domain assets in `COPYRIGHT.txt` with Unlicence text or
dual-licensing scheme.
And document commit hashes for most thirdparty code in `thirdparty/README.md`
for clarity, and in case there's no tag matching the included version numbers.
Updated glslang and Vulkan headers/loader following the instructions found in thirdparty/README.
glslang was updated to the 'known good' matching Vulkan SDK version 1.2.162.0. Vulkan headers and loader were updated to the commit tagged with sdk-1.2.162.0.
'vk_mem_alloc.h' and 'vk_mem_alloc.c' are unchanged since there hasn't been a new tagged release since 2.3.0.
Here's the Vulkan release notes for this update:
https://vulkan.lunarg.com/doc/sdk/1.2.162.0/windows/release_notes.html
Reverted and removed the unnecessary fix-mingw-snprintf patch for glslang as well as the mention of it in thirdparty/README.md.
-Happens on import by default for all models
-Just works (tm)
-Biasing can be later adjusted per node or per viewport (as well as globally)
-Disabled AABB.get_support test because its broken
-Reworked how meshes are treated by importer by using EditorSceneImporterMesh and EditorSceneImporterMeshNode. Instead of Mesh and MeshInstance, this allows more efficient processing of meshes before they are actually registered in the RenderingServer.
-Integrated MeshOptimizer
-Reworked internals of SurfaceTool to use arrays, making it more performant and easy to run optimizatons on.
This reverts commit f697e7879b.
Part of the update introduced a regression:
https://github.com/memononen/nanosvg/issues/188.
We could include a local revert of the problematic commit but let's just do a
full revert to our previous version, and I'll re-update once the regression
is fixed upstream.
Fixes#43641.
Actually sdk-1.2.154.1 for Vulkan-Loader.
glslang is updated to bacaef3237c515e40d1a24722be48c0a0b30f75f which is the
known-good version for Vulkan-ValidationLayers 1.2.154.0.
COPYRIGHT.txt was synced with the current version of the glslang LICENSE.txt,
and `glslang/register_types.cpp` now uses the upstream definition for its
default builtin resource instead of hardcoding it.
This code currently isn't compiled (and cannot compile).
We plan to re-add OpenGL ES-based renderer(s) in Godot 4.0 alongside Vulkan
(probably ES 3.0, possibly also a low-end ES 2.0), but the code will be quite
different so it's not relevant to keep this old Godot 3.2 code.
The `drivers/gles2` code from the `3.2` branch can be used as a reference for
a potential new implementation.
Implements exit codes into the engine so tests can return their statuses.
Ideally we don't do this, and we use FIXUP logic to 'begin' and 'end' the engine execution for tests specifically.
Since realistically we're initialising the engine here we don't want to do that, since String should not require an engine startup to test a single header.
This lowers the complexity of running the unit tests and even for
physics should be possible to implement such a fix.
Upgrade OIDN to 1.1.0, the latest stable version that doesn't need
the ISPC compiler.
Documented the changes made during the removal of TBB and added a patch
file for them.
Upstream development restarted after 13 years. Changes:
2020-02-02: Version 0.5.0
Minor speed improvement on the decompressor.
Prevent memory violation when decompressing corrupted input.
2020-01-10: Version 0.4.0
Only code & infrastructure clean-up, no new functionality.
The comment mentioned a conflict with libwebsockets, but we actually
still get this conflict even now that we don't use libwebsockets.
Not sure what component is clashing but we should basically just keep
this patch.
Follow-up to #36823.
Upstream removed the option in KhronosGroup/Vulkan-Loader#260, which
breaks our current use case.
This commit reverts KhronosGroup/Vulkan-Loader#260 is our vendored
loader.
We may need to re-evaluate how we link the loader, but until then,
reverting this PR fixes Windows support after the upgrade to a recent
SDK version in #36932.
