Implemented using GLSL API from OpenColorIO library and
some general functions were added to it's c-api:
- OCIO_setupGLSLDraw prepares OpenGL context for GPU-based
transformation for a giver processor.
This function compiles and links shader, sets up it's
argument. After this transformation would be applied
on an image displaying as a 2D texture.
So, glaDrawPixelsTex called after OCIO_setupGLSLDraw will
do a proper color space transform.
- OCIO_finishGLSLDraw restores OpenGL context after all
color-managed display is over.
- OCIO_freeOGLState frees allocated state structure used
for cacheing some GLSL-related stuff.
There're some utility functions in IMB_colormanagent which
are basically proxies to lower level OCIO functions but
which could be used from any place in blender.
Chacheing of movie clip frame on GPU is also removed now,
and either glaDrawPixelsTex or glaDrawPixelsAuto are used
for display now. This is so no code duplication happens
now and no large textures are lurking around in GPU memory.
Known issues:
- Texture buffer and GLSL are no longer checking for
video card capabilities, possibly could lead to some
artifacts on crappy drivers/cards.
- Only float buffers are displaying using GLSL, byte
buffers will still use fallback display method.
This is to be addressed later.
- If RGB curves are used as a part of display transform,
GLSL display will also be disabled. This is also thing
to be solved later.
Additional changes:
- glaDrawPixelsTexScaled will now use RGBA16F as an
internal format of storing textures when it's used
to draw float buffer. This is needed so LUT are
applied without precision loss.
bookkeeping variables from when the code was shared with texture
painting) Also texonly now is always zero, probably another leftover
from 2d/3d painting entanglement era.
- pass string size to BLI_timestr() to avoid possible buffer overrun.
- quiet warning for mingw.
- include guards for windows utf conversion funcs.
- fix for mistage in edge-angle-selection check.
- some style cleanup.
* Reflect changes stated in prev commit about contexts in py code.
* Add a "Plural" context, to handle cases where english does not mark plural at all (e.g. shorten labels of only one adjective). Not so happy with that, but can't see any other way to do it, for now.
* Abuse "ID_CURVE" context for all falloff curves (this should solve some confusion issues, e.g. "sharp"...).
Couple of precision issues here:
- Interpolation was rounding trunkcating colors, and because
of some precision issues value 254.999 became 254 leading
to troubles later.
Now color interpolaiton will do rounding to nearest int.
- Subtract was setting channels to something negative which
confused color management (especially negative alpha),
Now subtract effect will clamp channels to 0 from bottom
and also do some tricks to prevent small alpha which
could also confuse color management.
To complete previous commit:
New "Activate same type next/prev" operator - replaces the two not working previous ones.
This selects/activates and views the next or previous node of same type.
Shortcuts: Shift ] and Shift [
This is the first half of the commit, it's going to be reviewed further.
It also depends on other work in node drawing (to not reshuffle order on
every select/activate).
At least it works quite nice for the shortcuts shift [ and shift ]
"View selected strip(s)" now includes the active strip (which can be deselected
in cases). Less confusing this way, also because active strip is being drawn
very visible.
Issue was caused by delayed or missing image user frame
number update, which lead to image loading failure in
cases node is updating from image signal callback.
Solved in a way that file from image datablock is used
for sockets detection instead of loading image for
current frame.
The time it was printing was excluding the time spent on color and file saving.
Now it includes this too, plus it prints the time spent on color/files.
Makes quite a difference when you do sequencer conversions or grading in
terminals - time that was printed was very misleading.
For readablity of logs, also added an extra empty line after this print.
In this case edgenet_fill operator failed and it would fallback to filling as unordered vertices which was mostly fine but failed on some concave loops.
Add a new bmesh operator 'edgeloop_fill' fills in closed loops even if they don't make a valid edge-net.
Added new build option WITH_JACK_DYNLOAD for CMake and
WITH_BF_JACK_DYNLOAD for SCons, which means there'll be
no build-time linking against libjack and getting symbols
from libjack will happen runtime using dlopen and dlsym
tricks.
Alternative would be to use weak linking, but it'll require
having wrapper for preloading libjack.
This new options are disabled by default and they only
intended to be used on linux. Other platforms shall not
be using this and there shall be no functional changes
on non-linux platforms at all.
