Mistake in fe52284be9.
The attribute filter created by the geoemtry nodes evaluation system
doesn't work in this case because the the conversion from curves to
points is internal to the node, just an implementation detail.
Pull Request: https://projects.blender.org/blender/blender/pulls/137614
For performance reasons render graphs can keep memory allocated so it
could be reused. This PR optimizes the memory usage inside the
rendergraph to keep it within normal usage.
I didn't detect any performance regression with this change but reduces
the memory when performing final image rendering of heavy scenes.
Partial fix for #137382. the amount of memory still increases with 4mb
per render. It fixes the main difference when using large scenes.
Pull Request: https://projects.blender.org/blender/blender/pulls/137660
This adds a new Instance Bounds nodes which provides an easy and
efficient way to get the bounding box of each instance.
Nested instances or not taken into account to compute the bounding
box. That would result in a major performance degredation because
one would have to basically realize all instances to do that.
Also see #135933 for some examples.
Pull Request: https://projects.blender.org/blender/blender/pulls/135933
The Cryptomatte node only works with the first view layer in a setup
where a view later is called ViewLayer and another called ViewLayer.001.
This is because the code used a prefix check to match the view layer to
the Cryptomatte type name, and since ViewLayer is a prefix of the other
ViewLayer.001, the check always matches the first ViewLayer.
To fix this, we just remove that guard condition and rely on later
checks.
Pull Request: https://projects.blender.org/blender/blender/pulls/137659
The goal here is to avoid having to cast to and from `ID` when getting the
evaluated or original ID using the depsgraph API, which is often verbose and not
type safe. To solve this, there are now `DEG_get_original` and
`DEG_get_evaluated` methods which are templated on the type and use a new
`is_ID_v` static type check to make sure it's only used with valid types.
This allows removing quite some verbosity on all the call sites. I also removed
`DEG_get_original_object`, because that does not have to be a special case
anymore.
Pull Request: https://projects.blender.org/blender/blender/pulls/137629
The issue was that there were no handle positions but
`retrieve_visible_bezier_handle_points` still returned a mask
of bezier points.
The fix ensures that ther drawing has handle positions and returns
an empty mask otherwise.
This means that the code that draws the handles now correctly ignores
them when there are no handle positions available.
Pull Request: https://projects.blender.org/blender/blender/pulls/137655
It is more convenient for math reasons, for Blender to treat a pen tilt
toward the user as a positive angle. This PR just reverses all the
comments about the expected values. And also reverses the values from
WinTab, Windows Ink, and Wayland to match.
Pull Request: https://projects.blender.org/blender/blender/pulls/137636
Prevent dissolve leaving duplicate faces when dissolving edges.
Also use the builtin reuse of duplicate faces provided by BM_face_join()
instead of doing it manually.
Ref: !137634
Caused by checking if strip frame index falls into a range instead of
timeline frame. This was introduced in 6a39d79967 and was incorrect
solution to the issue.
Cache key had `timeline_frame` field, which was removed, because it is
not updated when strip moves. The value is now calculated from frame
index.
The issue was not very noticable especially when working with larger
strips, but the behavior was incorrect and confused me when working on
cache related features.
Pull Request: https://projects.blender.org/blender/blender/pulls/137583
In bf18e8f814, `buttons_context_compute()` was moved "upstream" in code
path, as a de-duplication effort. But this wasn't correct, because
function `property_search_for_context()` was executed in a loop where
the tabs were iterated over.
This commit moves `property_search_for_context()` back where it was
before.
Pull Request: https://projects.blender.org/blender/blender/pulls/137576
Since 115d809358 the scroll bars are darker on light theme. But this
makes the circular "handles", in Timeline for example, to not be
visible. This is because we start with the item color and then make
it darker when over a threshold. But we never make them lighter when
they are dark. This PR makes these darker or lighter depending on the
base color.
Pull Request: https://projects.blender.org/blender/blender/pulls/137618
Currently this code uses the `Mesh` struct as a value embedded in the
undo step struct. This causes the need for various hacky workarounds and
abstraction leaks. It's simpler to just allocate it in the regular way
The Shape Key ID can also be freed with the regular ID free function,
avoiding the need to expose the internal free function.
Pull Request: https://projects.blender.org/blender/blender/pulls/137613
This adds support for having viewer nodes in closures. The code attempt to
detect where the closure is evaluated and shows the data from there.
If the closure is evaluated in multiple Evaluate Closure nodes, currently it
just picks the first one it finds. Support for more user control in this case
may be added a bit later. If the Evaluate Closure node is in e.g. the repeat or
foreach zone, it will automatically use the inspection index of that zone to
determine what evaluation to look at specifically.
Overall, not too much had to change conceptually to support viewers in closures.
Just some code like converting between viewer paths and compute contexts had to
be generalized a little bit.
Pull Request: https://projects.blender.org/blender/blender/pulls/137625
The test_draw_submit_only was broken because of the
empty pass optimization #135875. Making the passes
non-empty fixes the test.
`eevee_surfel_list` has non-deterministic result. This
should be investigated in the future.
Tests were calling the submission mutex without
init. Adding functions to expose only setting
up the submission mutex instead of the full
DRW context (which is uneeded here).
This makes no sense to have in the draw namespace.
Also take the opportunity for making the coordinates
a float2 and rename them to something more descriptive.
Viewer nodes in compositor now show the eye icon when they are active,
so after https://projects.blender.org/blender/blender/pulls/134949
both geometry nodes viewers and compositor viewers look the same.
The only remaining difference is that in geometry nodes, clicking on
the eye icon deactivates the viewer and the node tree output
is shown instead. This is not possible in the compositor currently,
because at least one viewer must be active in the node tree.
Pull Request: https://projects.blender.org/blender/blender/pulls/134417
Operators like smooth weights, or transfer weights are exposed in
weight mode. Their definition has `OBJECT_OT_*` prefix, this appends
shortcuts key on those operators to "object mode" keymap. Use
`WM_keymap_guess_from_context` instead. This will allow new shortcut
keys to relevant keymap when object operator is exposed in 3d view.
Pull Request: https://projects.blender.org/blender/blender/pulls/137519
After creating a new shading material, the new node tree might become invisible and the user has to scroll down to see the nodes.
The idea of this PR is to create nodes with x-positions centered around zero, such that they are always visible no matter the screen size or workspace setup. Ideally, the nodes' position should depend on the region's zoom and pan. However, the node creation code currently does not depend on `SpaceNode`, so such solution would complicate the code a bit. Also, this heuristic seem to work well enough for most cases.
Note: this only affects newly created materials. The material of the default cube and the default world material still have invisible nodes sometimes. This is because they are saved in the startup file, which will be addressed in a different patch.
Pull Request: https://projects.blender.org/blender/blender/pulls/136926
For consistency with other parts of Blender, the geometry visibility in the viewport is controlled by the viewport icon. The compositor will still have the eye icon for previews, but viewer nodes will get the viewport icon, also for consistency reasons.
Pull Request: https://projects.blender.org/blender/blender/pulls/134949
