| Age | Commit message (Collapse) | Author |
|---|
|
# Conflicts:
# source/blender/draw/modes/paint_texture_mode.c
|
|
|
|
|
|
This is a revert of a small fraction of commit rB5e332fd700
that introduced the issue according to bisect.
Doing a break here is wrong, because BKE_crazyspace_build_sculpt
assumes that processing stopped at the first deform modifier
without deformMatrices, and thus skips all modifiers until it
finds one like that. Thus this early loop exit makes the behavior
worse, as instead of skipping just Mask and Mirror, it skips all.
|
|
When the frame was zero, the frame number was clamped by error to 1.
|
|
The thickness was using the object scale always, but when the layer is parented, must use the parented object scale.
|
|
|
|
Matches 3D Connexion's reference software.
|
|
|
|
Instead of having the option to show marker lines,
make the marker region optional.
- Added a Show Markers entry in the View menu of the animation editors.
- If the markers region is not active then the Marker menu gets hidden.
- Removed marker menu from the driver editor
and don't allow to use marker operators.
|
|
|
|
|
|
This is a partial rollback of f18ad385dffe70f5e57df00ff9bbb7b42fa05be0. It
turned out to be more tricky to determine animatedness of an object.
This fixes T71986.
|
|
When the final stroke was smaller than original stroke, the weights array must be resized, but by error the function used the original stroke pointer instead of the new stroke pointer and this corrupted the pointers.
|
|
Occurs in edit mode when object has no uniform scale.
|
|
Loading a blendfile allocates one or more windows that need to be freed.
Freeing those windows also calls `BKE_workspace_instance_hook_free()` to
free workspaces. However, in the `BlendfileLoadingBaseTest` test there are
no workspaces allocated. This caused an assertion failure, which was worked
around by not asserting when Blender is running in background mode.
Reviewed by @Severin via pair programming
|
|
This gets rid of the TAA which needs to be added back
|
|
|
|
Currently unused, but needed for coming fix.
|
|
|
|
It started to be a long code of all various cases in a single file,
which started to be really confusing.
|
|
C++ is used more and more, and it is becoming more and more annoying to
keep track of whether header have C++ guard or not.
Is easier and more clear to be consistent in all headers and have such
guards in all headers.
|
|
The visibility of the object is what counts, not just the visibility of
the collection.
|
|
There are no more 'layers' in Blender. I chose 'Visible Objects' rather
than 'Visible Collections' to be consistent with the other
'{Renderable,Selected} Objects Only' options.
No functional changes.
|
|
`AbcTransformWriter::hasAnimation` recently became smarter than just
returning `true`, but wasn't quite smart enough yet. Constraints are now
considered a source of 'animation'.
|
|
list is empty
The problem was the draw function tried to use the material and gpsettings and both were NULL.
Now, the default material is used.
|
|
The `subprocess.call()` function doesn't check the exit status code of the
subprocess. Use `subprocess.check_call()` or `subprocess.run()` instead.
|
|
On CentOS 7, `tar --use-compress-program='xz -9'` tries to run `xz -9` as
executable, rather than running `xz` with `-9` as argument. Passing the
`-9` option via the `XZ_OPT` environment variable, as suggested by
@campbellbarton in D6138, works fine.
|
|
The VERTEX mode was not checked in shaders and must be managed equals to MATERIAL mode.
|
|
Remove redundant defines.
|
|
Number sliders were being used in a menu which doesn't work very well.
|
|
|
|
This was too slow taking around 8 seconds for a full revolution.
|
|
|
|
|
|
Missed one in the previous commit.
|
|
|
|
Material mode.
|
|
Previously some important features like OpenSubdiv were disabled by default,
which caused confusion.
The purpose of disabling some of these features was to avoid potentiall build
errors on Linux. But with precompiled libraries, install_deps.sh and better
library availability checking this is hopefully not much of a problem anymore.
This makes "make full" obsolete, but it's kept to not break docs or shell
scripts that people may have, and the .cmake config file remains useful to
modify an existing build folder.
This also changes some option to only be available on platforms where they
are actually supported (WITH_JACK, WITH_TBB_MALLOC_PROXY and X11 options).
Fixes T69742
Differential Revision: https://developer.blender.org/D6306
|
|
When negative scaled matrices were used, the roll calculation
didn't match, when calling Armature.transform().
|
|
This is caused by the fallback path used by OSX, which is reconfiguring
the same default VAO. But it seems to be an issue on certain drivers.
|
|
This new test class minimally sets up Blender so that it can load blend
files and construct a depsgraph without crashing.
Note that it hasn't been tested on very complex blend files, so it may
still crash when the loaded blend file references/requires uninitialised
data structures.
The test will certainly crash with Blend files created with Blender
older than 2.80, as the versioning code requires space types to be
registered. This is normally done by initialising the window manager,
which is not done in this test. The WM requires Python to run, which in
turn requires that Blender finds the release directory in the same
directory that contains the running executable, which is not the case
for GTest tests (they are written to `bin/tests/executablename`.
Reviewed By: sergey, mont29
Differential Revision: https://developer.blender.org/D6246
|
|
|
|
When encountering an error during context creation, the "OptiXDevice" constructor aborts early.
This means the "cuda_stream" vector is never resized and the destructor iterated over non-existent data.
|
|
Logic for registering and checking properties of registrable classes was
broken, allowing to ignore some errors.
Recent fix rBeb798de101a `broke` the result of the
pyapi_idprop_datablock test, because previously that test would fail
(i.e. suceed, as it is an 'expected to break test') for a reason it was
not designed to check.
This is the problem with that kind of tests - you cannot really check
that they are failing on the expected reason(s)...
|
|
Menu items can be used to find shortcuts instead of keeping
keymap items here.
|
|
Small change do not draw an unsupported renderpass in the viewport
|
|
The acceleration structure built by OptiX may be different between GPUs, so cannot assume the memory size is the same for all.
This fixes that by moving the memory management for all OptiX acceleration structures into the responsibility of each device (was already the case for BLAS previously, now for TLAS too).
|
|
Render using the default view instead of the jittered view.
|
|
xz compresses 25% better than bz2, reducing download times and server load.
The numbers:
blender-2.80-linux-glibc217-x86_64.tar.bz2 (release): 134 886 174 bytes
with xz: 96 181 604 bytes (-28.7%)
with xz -9: 93 871 548 bytes (-30.4%)
blender-2.81-7c1fbe24ca33-linux-glibc217-x86_64.tar.bz2 (beta): 173 600 363 bytes
with xz: 133 100 664 bytes (-23.3%)
with xz -9: 129 534 124 bytes (-25.4%)
xz also decompresses more than twice as fast as bz2, however compression needs
four times as long (on my 7-year-old laptop 3-4 minutes instead of <1).
Also xz has become more common than bz2, e.g. Debian/Ubuntu deb packages have
been xz-compressed for years, so the dpkg package manager as well as systemd
and grub all depend on liblzma being present, whereas bz2 is becoming more and
more optional.
Current Linux archives also include the UID/GID of whatever user account
happens to be used for building by the blender.org infrastructure. If someone
then installs these archives as root e.g. to /usr/local/... and doesn't pay
full attention the files remain owned by a regular user, which is a serious
security issue. This patch fixes that by setting the UID/GID to 0.
Differential Revision: https://developer.blender.org/D6138
|
