This removes one limitation of the guarded scopes that needed to
not have return statement. Now a dummy return statement is created.
See #139233 for original implementation.
This adds the following features:
- `class` keyword support: checked by C++, mutated to struct for shader.
- `private` and `public` keywords: checked by C++, removed for shader.
- `static` methods.
- `const` and non-const methods.
What is not supported:
- Constructors
- Destructors
- operators
- Method definition outside of class definition
- member reference without `this` keyword.
This is implemented using a very simple lexer/parser allowing semantic traversal.
Pull Request: https://projects.blender.org/blender/blender/pulls/144025
There should be minimal user-measurable changes since the old algorithm
is not the same as OpenSubdiv. For example now creases (both edge and
vertex ones) are properly supported.
This allows to have less dependencies on the old subdivision surface code
which ideally will be removed soon.
Pull Request: https://projects.blender.org/blender/blender/pulls/143712
Previously copying grease pencil strokes implicitly requires materials
to be available in `grease_pencil_copy_strokes_exec` which isn't
necessarily the case sometimes. Now just set material session uid to 0
when material is not found. When pasting, non-existent materials will
be created as a fallback already.
Additionally, this PR corrects material
access by using `BKE_object_material_get` instead of manually getting
the material from grease pencil material array, making it consistent
with the rest of the logic in blender.
Pull Request: https://projects.blender.org/blender/blender/pulls/142702
overlay_wireframe_frag doesn't always set the gl_FragDepth. Depending on
how the driver deals with this it can use uninitialize values or previous
values. This leads to render artifacts as depth testing/storage could be
wrong.
Pull Request: https://projects.blender.org/blender/blender/pulls/144198
The Cryptomatte file picker always samples the first frame in an image
sequence, even if the image editor/backdrop display a different frame.
To fix this, we take the frame into consideration when retrieving the
image to sample from.
Pull Request: https://projects.blender.org/blender/blender/pulls/144195
The preview of the render layers node is set to off in the default
node tree.
The reasoning is:
- It is common to start compositing without rendering, so this will
give a cleaner look for a first impression.
- Previews in general are not working well (currently the preview won't
compute if the node is disconnected, sometimes the preview is shown
with an empty image, etc..).
Pull Request: https://projects.blender.org/blender/blender/pulls/144173
Extend the existing benchmark framework to support the compositor.
Files are added separately.
Example output:
```
$ ./benchmark.py run comp
4.5 main
ghosts 1.2055s 1.0243s
ghosts 1.4936s 1.0454s
ghosts 0.4414s 0.4330s
file:///Users/habib/blender-git/benchmark/comp/results.html
```
Pull Request: https://projects.blender.org/blender/blender/pulls/136600
Improve the layout and listing of image output file format menu.
* Sort the menu alphabetically.
* Remove image icons.
* Remove column layout, use a single list.
* Group first the most commonly used formats.
* Add extension, similar to I/O menus.
See PR for details and screenshots.
Pull Request: https://projects.blender.org/blender/blender/pulls/144162
The title says it pretty much all: the shrinkwrap modifier is no longer
dependent on the legacy DerivedMesh.
The benefits for doing this are:
- The DerivedMesh is closer to be removed from the codebase.
- Only base vertices are subdivided which could result in better
performance and memory usage.
- Proper crease support (for edges and vertices).
The downside of using OpenSubdiv is that the topology refiner might be
expensive to calculate, so for meshes with a lot of extraordinary faces
(such as n-gons, triangles) the performance will be worse.
Pull Request: https://projects.blender.org/blender/blender/pulls/143706
This PR contains the following incremental changes:
- Remove dead code of AO multires baker
- Use C++ primitives in multires baker (and cleanup some comments and naming)
- Remove unused bake user scale
Pull Request: https://projects.blender.org/blender/blender/pulls/143707
* NodeTree space does not refcount the displayed nodegroup
* Collections used by light linking _do_ ID refcount (unlike default
expected behavior for Collection usages).
Swapchain handling of minimized windows wasn't correct. On some
platforms it still tried to create images with no surface.
This PR will discard swapchains of minimized windows, but still being
able to flush the render graph and free resources.
Pull Request: https://projects.blender.org/blender/blender/pulls/144189
The `Add` > `Scene` > `New Scene` operator was behaving as follows:
* If there is no active scene strip, create an empty blank scene and
assign it to a new scene strip.
* Otherwise, use the scene referenced by the active scene strip and
use one of the copy scene methods (blank, copy settings, copy linked,
or full copy) to create a new scene and assign it to a
new scene strip.
This was not ideal for 2 reasons:
1. With no strip selected, creating blank scenes is generally not very
useful as it means that the user has to e.g. update all the render
settings, add a camera, etc.
2. The behavior of copying an existing scene by using the active strip
is very hidden. Also because adding the strip immedialty calls the
move operator, so you cannot even adjust the copy method of the
scene.
This PR changes 3 things:
1. Don't use at the active scene strip. Instead use the active scene.
2. Don't create an empty blank scene, copy the settings of the active
scene by default. This means that e.g. render settings can be reused.
3. The operator entry in the `Add` > `Scene` menu is now called `Empty Scene`.
Part of #144063.
Pull Request: https://projects.blender.org/blender/blender/pulls/144069
This test can't easily be updated because it relies on the old socket shape
design before #144119. Parts of it could be done from scratch again, but
most of this stuff is also tested by the new structure type inferencing test
in `bl_node_structure_type_inference.py`.
Pull Request: https://projects.blender.org/blender/blender/pulls/144185
Mixing dynamic data with other data always results in dynamic now. While this
wasn't true without lists, it is true now and also makes sense generally. Also
mixing lists and single values or fields results in a list now.
Pull Request: https://projects.blender.org/blender/blender/pulls/144183
Node buttons declared with `NodeDeclarationBuilder::add_default_layout`
are not visible in the `uiTemplateNodeView` used for shader nodes among
others. This happens in the Glass BSDF shader node.
The `ui_node_draw_node` was not handling node layout items in case the
node is using a custom socket order.
This was originally fixed in #143781 which was closed accidentally.
Pull Request: https://projects.blender.org/blender/blender/pulls/144159
This removes the `STRIP_DUPE_UNIQUE_NAME` and `STRIP_DUPE_ALL` macros
and replaces them with an `enum StripDuplicate`.
A flag with a value of 0 was implicitly used as "duplicate selected strips" so the
enum uses `Selected = 0` to document this behavior better.
Pull Request: https://projects.blender.org/blender/blender/pulls/144125
Vulkan 1.0 render passes have been replaced by dynamic rendering in 1.2.
Blender Vulkan backend was implemented with dynamic rendering in mind.
All our supported platforms support dynamic rendering. Render pass support
was added to try to work around an issue with legacy drivers. However these
drivers also fail with render passes.
Using render passes had several limitations (blending and some workbench
features were not supported). As no GPU uses it and it is quite some code
to support it is better to remove it.
Pull Request: https://projects.blender.org/blender/blender/pulls/144149
This was disabled because island selection used to be exposed as a
UV selection mode (along with vertex/edge/face) where selecting
parts of other islands would seem like a bug.
Now island selection has been moved to a separate toggle it makes
sense to allow connected geometry to be included in the selection when
other selection options enforce this.
Support limiting the merge by connected geometry.
This is useful for operations such as bevel where it's not desirable
for the merge to connect isolated surfaces.
Accessible from the Python API, no user visible changes.
Ref !144036
Failure to do this can cause the the property default to be incorrect
and some sanity checks fail to run.
Fail when an error if this isn't set.
Ref !144133
Follow up to f646948ace
The repeat zone color is changed back to the color it used in 4.5, to
avoid the need for people to re-learn the color. The closure socket and
zone share the same color, which is made more yellow-ish to distinguish
it from the bundle socket color.
The spec for .usdz permits only a small handful of file formats to be
contained in the archive, and HDR is not among those supported[1]. This
also causes validation errors with the `usdchecker` tool (will be
properly tested in a follow up change).
Ignoring the potential for a user to export a .usdz file with materials
referencing unsupported file formats, Blender itself should use only
the supported formats for its business.
[1] https://openusd.org/release/spec_usdz.html#usdz-specification
Pull Request: https://projects.blender.org/blender/blender/pulls/144101
