The `DeltaProjectionFunc` introduced in cbe2bb6755 was always using a
"worldspace" plane [which ignored the layer rotations for projections --
be it through actual layer transforms or object rotations].
So to resolve, use the appropriate axis of the (already available)
`layer_to_world` matrix.
Pull Request: https://projects.blender.org/blender/blender/pulls/132690
When the brush refactor project was in progress, the "Fake Neighbors"
feature used exclusively by the Pose brush was made more explicit.
However, the migration missed the usage of this data in the `flood_fill`
APIs.
This resulted in the option not affected unconnected topology islands
except when the Pose Origin Offset was non zero and using the Topology
mode of the brush.
To fix this, this commit updates the flood_fill APIs to take in the
pre-calculated fake neighbor data and use it when traversing connected
vertices and applying weights.
Pull Request: https://projects.blender.org/blender/blender/pulls/132610
With the brush asset project, the `Paint` `brush` and `eraser_brush`
properties were effectively turned into a convenient cache of the active
brush. A related operator, `paint.brush_set` was also removed in favor
of `brush.asset_activate`
While this is technically a breaking change to the API, it currently
seems better to align this property with expected usage & other recent
changes rather than allow users to set a property that may not behave as
expected.
There are two currently known side effects that setting this property
via the Python API has that the equivalent call to brush.asset_activate
does not:
* Changing this property via the console or script, peforming a stroke
and then undoing the stroke causes the active brush to change - this
directly contrasts with the normal experience of using the asset
shelf where brush changes are not affected by undo
* The asset shelf itself does not update the currently active brush
until a subsequent mouseover
Pull Request: https://projects.blender.org/blender/blender/pulls/131991
Before, it was only possible to clear the entire cache at once or to rely on
automatic clearing when it gets full. This patch adds the ability to remove
cached data based on a predicate function.
This is useful for #124369 for partially invalidating the cache for some files.
Pull Request: https://projects.blender.org/blender/blender/pulls/132605
The issue was that the propagation of referenced anonymous attributes treated
geometry outputs of the foreach zone as "normal". That means that every
anonymous attributes referenced by the input socket would also be referenced by
the output socket.
However, just like in the repeat zone, this so called "propagate relation" needs
some special behavior, because anonymous attributes references created inside a
zone have to remain inside that zone. Instead, the output node creates a new
anonymous attribute reference that is used outside of the zone.
Note, this is the same as #132560 but also implements the right-to-left pass,
whereas before only the left-to-right pass was implemented.
Pull Request: https://projects.blender.org/blender/blender/pulls/132607
There were two separate issues which have a very similar solution:
* When loading the .blend file, the `scene->rigid_body_world->group` collection
pointer has to be mapped to the actual address. However, that was not done
because `BKE_rigidbody_world_id_loop` was a stub when `WITH_BULLET` is off.
That resulted in the collection pointer having an invalid address.
* When closing the file, there was some issue because of incomplete code for
copying rigid body related stuff for the depsgraph and hence there was some
unexpectedly shared ownership which leads to a use-after-free. Here the fix is
to move the copy code out of a `#ifdef WITH_BULLET` block too.
Since none of the moved code actually needs bullet, it seems fine to move it.
The code should be exactly the same in the common case with `WITH_BULLET` is on.
Pull Request: https://projects.blender.org/blender/blender/pulls/132697
Papercut reported in #132293.
Allow explicitly disabling this behavior for a button, and do so for the
big preview asset shelf popup button. It gets more in the way than it's
useful in this case.
Mistak in 39f7c506b5.
The VBOs need to be allocated! And we can just use a single dummy
type as well, rather than using the type from the attribute request
which is meaningless in this case.
PR #129315 refactored polylines. The shaders now attaches the vertex
attributes as SSBOs. Adding a workaround for polyline shaders to
extract the correct vertex formats when called via Python.
Pull Request: https://projects.blender.org/blender/blender/pulls/132689
Confusing error messages are printed when requesting a clipped builtin
shader via Python that does not exist.
This PR will remove the confusion of the messaging:
- Replaced BLI_assert_unreachable with an assert as it is reachable
code.
- Adding clipped configuration for POLYLINE_UNIFORM_COLOR
Pull Request: https://projects.blender.org/blender/blender/pulls/132686
This renames the struct `Sequence` to `Strip`.
While the motivation for this partially comes from
the "Sequence Design" #131329, it seems like this
is a good refactor whether the design gets implemented
or not.
The `Sequence` represents what users see as strips in the
VSE. Many places in the code already refere to a `Sequence`
as "strip". It's the C-style "base class" of all strip types.
This also renames the python RNA type `bpy.types.Sequence`
to `bpy.types.Strip` which means that this technically breaks
the python API.
Pull Request: https://projects.blender.org/blender/blender/pulls/132179
The max length of the RNA property `ActionSlot.identifier` was set
incorrectly. The setter code did manage the length properly, but the
getter was checking agains that incorrect max length, and rightfully
complained.
Pull Request: https://projects.blender.org/blender/blender/pulls/132691
This PR will add timeline semaphores to be required. It doesn't use
the timeline semaphores yet, but as multiple developments will
rely on it it is better to add the requirement.
Pull Request: https://projects.blender.org/blender/blender/pulls/132683
Framebuffers are getting freed in the GPUContext base class destructor. But
the framebuffer destructors use the MTL/VK/GLContext derived class, whose
destructor has already completed at this point. So these contexts are no
longer valid to use.
Now free the framebuffers earlier.
This caused ASAN warnings, it's not known to cause actual bugs.
Pull Request: https://projects.blender.org/blender/blender/pulls/132504
This code is now only used by annotations but was not updated.
Uses the right properties now and removes the layer mask
one which is not used by annotations.
Pull Request: https://projects.blender.org/blender/blender/pulls/132030
Move the `ID_TAG_ID_LINK_PLACEHOLDER` bit of `id->tag` to
`id->runtime.readfile_data->tags.is_id_link_placeholder`. It also
introduces the necessary stucts and allocation/freeing code.
Old code:
```cpp
if (id_tag & ID_TAG_ID_LINK_PLACEHOLDER) {
```
New code:
```cpp
if (readfile_id_runtime_tags(id).is_id_link_placeholder) {
```
where `readfile_id_runtime_tags(id)` is a getter for
`id->runtime.readfile_data->tags` that is null-safe for
`id->runtime.readfile_data`. The `readfile_data` is not allocated in
these cases:
1. When reading undo steps, because that doesn't have to deal with
versioning or linking (which are the sole purposes for this
struct).
2. When linking from another file (for example from the 'Link...'
operator). The just-linked IDs will have the `readfile_data`
struct, but already-loaded IDs will already have had those freed.
No functional changes intended.
Pull Request: https://projects.blender.org/blender/blender/pulls/132169
Design Task: #131695
Fix bone snapping failure by removing the Armature bounding box check.
That check would go over all the bones, compute the total bounding box,
just to avoid the remaining code (which goes over all the bones anyway).
Not sure why the code is causing these issues, but I'm guessing it's due
to getting the bounding box in the wrong space (could be a bug
introduced in 6212c3c374). Since this call doesn't look like it's an
actual optimization to me, I think it's better to remove it.
Pull Request: https://projects.blender.org/blender/blender/pulls/132602
The compositor crashes if a node inside a node group is connected to a
group input that have a different type and the node group is used
without a connection to that input. That's because the compositor code
assumes the type of the group input without implicit type conversion to
the expected type of the node. To fix this, handle implicit conversion
for unconnected sockets as well.
This patch adds support for using integer sockets in compositor nodes.
This involves updating the Result class, node tree compiler, implicit
conversion operation, multi-function procedure operation, shader
operation, and some operations that supports multiple types.
Shader operation internally treats integers as floats, doing conversion
to and from int when reading and writing. That's because the GPUMaterial
compiler doesn't support integers. This is also the same workaround used
by the shader system. Though the GPU module are eyeing adding support
for integers, so we will update the code once they do that.
Domain realization is not yet supported for integer types, but this is
an internal limitation so far, as we do not plan to add nodes that
outputs integers soon. We are not yet sure how realization should happen
with regards to interpolation and we do not have base functions to
sample integer images, that's why I decided to delay its implementation
when it is actually needed.
Pull Request: https://projects.blender.org/blender/blender/pulls/132599
Decimate used `is_quad_convex_v3` to check if faces should be converted
to quads, adjust the epsilon since small faces were being skipped.
Ref: !132342
The "Blender" menu contains seven items in "System / "Redraw Timer"
that are for troubleshooting and testing. Six of the items take far
less than a second, but one takes a long time. "Animation Play" times
how long it takes to play through your current animation ten times. But
there is no way to guess this and some users have run it accidentally
by finding it in menu search. It gives no feedback at all, and there
are no hints on how long it will take. Once it plays through your
animation once you might guess that it runs forever. This PR shows
the test name and where it is in its ten runs. On platforms that
support it this also shows an app progress bar (on the taskbar icon for
Windows).
Pull Request: https://projects.blender.org/blender/blender/pulls/132648
- Using a main function allows the scripts to be imported without
executing logic.
- Declaring `__all__` lets tools such as "vulture" detect unused code.
Rendering a scene with py-drivers would crash after dropping an
extension onto Blender's window then canceling before the update
completed.
This was caused by context.temp_override pushing arguments into the
context but failing to pop them in the case of an error - which would
happen when the popup was closed while the update ran.
ContextTempOverride objects were not freeing the reference to their
arguments when the context managers __exit__ didn't run.
This would happen if __enter__ failed with an unsupported configuration
or if object was created and not used.
Resolve by making ContextTempOverride a container type with GC support.
This also allows the ContextTempOverride to be reused since it's
arguments are now cleared when it's deallocated instead of __exit__.
