Blender Git Loki

Bugfix [#35382] NLA "Multiply" Blend Mode calculated incorrectly

The "Multiply" blending mode for NLA strips worked incorrectly. Instead of
modulating the influence of the current strip, it was in fact scaling the result
of the entire stack (with the strip applied). This caused problems when
influence = 0, as it was in fact muting everything instead of just controlling
the strip we are interested in.

Fix bug in vertex group code after recent changes, did not do proper NULL
pointer check.

RNA: simplify space type items code, no need to have this duplicated.

Replaced a hard-coded array length with MAX_ID_NAME.

correction to typo in r57327, ui-load wasnt working on splash screen.

also minor changes:
- update the load-UI flag when changing in the preferences.
- remove unused initialization values for subdivide.

Correction to previous maya keymap commit

Added UNDOflag to the wrong operator..

Fix for r57316: SPACE_EMPTY *is* needed in RNA enum for now, exported keymaps abuse it...

So for now, static enum is back to its previous form, and the menu one is dynamically generated by a callback. Not nice, but works.

Fix #35661: Maya key config - select operations are skipped by "undo"

Made blender compilable with strict flags again.

Maybe Joshua would want to check on arguments for nlaevalchan_verify.

Fix #35683: Addon working fine in console. But crashed when click GUI button.

Crash happened in ED_view3d_calc_zfac and happend in cases operator was invoked
from a region different from RGN_TYPE_WINDOW.

For a transformation zfac is only used in convertViewVec in cases region is
RGN_TYPE_WINDOW, so solved by just adding extra check in calculateCenter
for this particular case.

Cycles / Wavelength Node:
* Fix for OpenCL compilation, tested with Intel and nVidia. It didn't like an implicit double promotion.
* Some small variable renaming.

optimize customdata lookups for subdivision.

Bugfixes [#35263], [#35382] NLA Animated Influence is ignored if strips below
have zero total influence

Previously, when evaluating the NLA stack at a particular point in time, if a
channel hadn't been encountered before, influence values were simply ignored
when accumulating the values contributed by each strip to the overall stack.

This behaviour simplified the handling of the problem of what "baseline" to
blend relative to (i.e. influence basically scales the magnitude of a scalar
around 0, but we may not exactly want a property to get it's value set to 0 as
baseline). However, the problem was that this meant that you'd get popping
artifacts when the a lower strip finally reaches influence=0 but your upper
strips haven't fully reached maximum yet ([#35382]). Another problem was that
you'd end up with less ability to scale the influence of all strips (as in
[#35263]).

So, as a stop-gap fix now, we will allow influence scaling to work on these
strips too. This still doesn't fix some of the other problems regarding
baselines/rest-poses and deterministic behaviour when some channels are only
keyed in one strip which isn't set to extend it's influence... Fixing those
issues is a bit more involved, and would require a bit of refactoring of how we
keep track of accumulation channels.

Bugfix [#35643] Animated textures are invisible in Graph Editor if it is not
linked via material

Textures linked to modifiers are now shown in the AnimEditor channel hierarchy
under object level now (i.e. on same level as ob-data, shapekeys, and object's
action). This makes it possible to edit such animation data without having to
ensure that these textures are also linked to the object's material so that they
will appear.

As a side-effect of how this is implemented, if playback is slower on scenes
following this commit, disable the "modifier" filter under the filtering
settings in the relevant animation editor header. In particular, it may be
beneficial to disable this when you've got scenes with meshes that have many
modifiers (but none of these have any linked data with settings which can be
animated), as Blender will still try to go through all those modifiers checking
for anything to show.

fix [#35648] Loop cut, smoothness showing wrong symmetrical

this is more of a feature request then a bug, but nicer to use even smoothing offset for loopcut smoothing.

fix [#35669] Snap curvepoints dont work correct when another object is selected

Ctrl+Click extrude curves now supports snapping to faces as with meshes.

fix [#35670] Selectionmode LMB dont change the controlling for putting curve points with strg+lmb

minor edit to previous commit, only show option to reload the file if its been saved.

Python script auto-execution changes:

- script execution is off by default

- if a blend file attempts to execute a script
this shows a message in the header with the action
that was suppressed (script/driver/game-autostart) and 2 buttons to either reload the file trusted, or to ignore the message.

- the file selector will always default to use the trust setting in the user preferences,
but reloading an open file will keep using the current setting (whatever was set before or set on the command-line).

- added SCons setting WITH_BF_PYTHON_SECURITY, this sets the default state for the user prefereces not to trust blend files on load.
... this option was in CMake before, but always off, now its enabled by default for SCons and CMake, and forced on in CMake for now.

Cycles / Code cleanup:
* Move hsv and xyz color functions into the dedicated util files (util_color.h and node_color.h).
* svm_lerp moved into util_math.h and renamed to lerp_interp, as it's used for the wavelength node now as well.