This should be purely an implementation change,
for end users there should be no functional difference.
The entire key configuration is in one file with ~5000 lines of code.
Mostly avoiding code duplication and preserve comments and utility
functions from the C code.
It's a bit long but for searching and editing it's also convenient to
have it all in one file.
Notes:
- Actual keymap is shared by blender / blender_legacy
and stored in `keymap_data/blender_default.py`
This only generates JSON-like data to be passed into
`keyconfig_import_from_data`, allowing other presets to load and
manipulate the default keymap.
- Each preset defines 'keyconfig_data'
which can be shared between presets.
- Some of the utility functions for generating keymap items still
need to be ported over to Python.
- Some keymap items can be made into loops (marked as TODO).
See: D3907
That kind of implicit includes should really only be done when totally,
absolutely necessary, and ideally only with rather simple 'second-level'
headers.
Otherwise not being explicit with includes always end up biting in
unexpected ways...
This is a variation of older hach which was setting is_rendering
to truth to tell window manager to not do dependency graph update.
In the nowadays reality window manager is supposed to do dependency
graph update during rendering, that was the whole purpose of CoW
project. This works fine for rendering, since render engines has
their own dependency graphs.
Physics, on the other hand, is using same dependency graph as used
for the viewport, and what's worse: it modifies objects from it.
For example, in a single threaded evaluation ASAN instantly catches
case when cached BVH constructed by smoke is referencing looptri
layer which is freed by viewport's update.
Now we are locking interface, allowing only a subset of navigation
operators to run. This seems to be safest way of dealing with the
problem. There are following variations which we can consider
doing:
- Allow viewport navigation, which will require making it so draw
manager does not write to the objects.
A bit dangerous, since smoke simulation might in theory modify
data which is also used by a draw manager.
- Make physics simulation to have own dedicated dependency graph,
solving all threading conflicts all together.
This fixes crash when baking smoke. Steps to reproduce:
- Call "Quick Smoke"
- In smoke panel, click "Bake".
* Icons in specials menu (Copy and the new Duplicate icon)
* Remove redundant "Settings" label next to ID template widget.
* Remove "Type" label and expand dropdown, since it only has 2 options (emitter/hair), saves 1 click.
* Move "Seed" to and "Hair Segments" to Emission panel
* Rename "Duplicate Particle Systems" to "Duplicate Particle System", since it
only duplicates one (the active one) at a time.
Add tool options to control how select operates (add/sub/set/and/xor).
Note: edit mode armature select still needs to support all options,
this is complicated by how it handles partial end-point selection.
There is now a manual refresh button on the panel to update the list
of objects in case it changes, and it also gets refreshed when changing
the collection or toggling the use count option.
This is a bit more manual but the previous code of refreshing the
list while evaluating the depsgraph was unreliable.
This also fixes it to take properly take into account visibility, and
to work with linked collections for which index writing was missing.
To prevent the pointcache from being copied-on-write too (and requiring
copying back), the cache is now shared between the original and
evaluated scenes. Reading from the cache is always allowed; running the
sim and writing to the cache is only allowed when the depsgraph is
active.
Some pointers have moved from RigidBodyWorld (RBO) to
RigidBodyWorldShared (RBOS). writefile.c copies some pointers back from
RBOS to RBO so that the file can still be opened on older Blenders
without crashing on a segfault.
The RigidBodyWorldShared struct is written to the blend file, because it
refers to the PointCache ID block.
The RigidObjectShared struct is runtime-only, and thus not saved to the
blend file.
An RNA getter-function is used to hide the new 'shared' pointer. As a
result the Python API hasn't changed.
Reviewed by: campbellbarton
Differential Revision: https://developer.blender.org/D3508
This translates the gl calls to the new GPU_ wrappers from D3501.
Given it's tedious and repetitive work, this patch does as much as it can with search + replace, the remainder of the gl calls will need to be manually dealt with on a case by case basis.
This fixes 13 of the 28 failing editors when building without opengl.
For the list of substitutions see D3502
Reviewers: brecht
Differential Revision: https://developer.blender.org/D3502
