The problem was that wmPushMatrix/wmOrtho/.. and similar functions did not
work well for offscreen rendering. It would have been possible to make a
fake subwindow for this, but I decided to just remove this extra layer as
it does not seem to have much purpose and has been quite confusing when
trying to fix other bugs. The relevant matrices are already stored in
RegionView3D so there will be no increase in calls to glGetFloat, which may
have been a performance reason to use this system in the past.
The problem was: the Blender default camera has DOF distance as 0.0. Since we are using this as Focal Length for the stereo calculation we had terrible stereo by default.
Fix: whenever DOF == 0.0 we use focal length as eye separation * 30.0 (known to be a reasonable value)
Fixed the operators for DopeSheet/Graph Editors responsible for setting the "auto-clamped". This option is actually per F-Curve instead of per handle, and the code here should function like it did in 2.4x
However, despite this, it still appears to work oddly IMO. Any comments Bassam or animators familiar with the intentions of this?
were cached once at the start, but these can change when modifiers are
executed, now it simply doesn't cache them anymore, that was only really
a performance bottleneck when it was caching individual vertices.
and made that the default for windows software opengl because that
seems to be working better at least on XP. Previously this could only
be specified from the command line.
Operator to add action-clips now takes names without needing the "AC" prefix. The previous way was non-obvious for scripting usage, but did not cause too much trouble to replace.
When the user double-clicks on a document file in the Finder, OSX doesn't simply give the filename as a command-line argument when calling Blender, as it is done in other OSes.
Instead, it launches the app if needed, and then sends an "openFile" event.
The user can also open a document file by dropping its icon on the app dock icon. But as this is not real Drag'n'drop, I've renamed the Ghost event to a less confusing "GHOST_kEventOpenMainFile" name.
DND Ghost wiki page updated : http://wiki.blender.org/index.php/BlenderDev/Blender2.5/DragnDrop
set the draw method to triple buffer or overlap depending on the
configuration. Ideally I could get all cases working well with triple
buffer but it's hard in practice. At the moment there are two cases
that use overlap instead:
* opensource ATI drives on linux
* windows software renderer
Also added a utility function to check GPU device/os/driver.
This was previously only available from the datablocks viewer, but this is the direct replacement to the badly named 'speed' ipo-curve in the past ('speed' implies a rate that must be integrated/added to the results of past frame, rather than a factor).
Also, tweaked the RNA definition so that this shows as a slider (i.e. a factor), since the valid values for this are clamped to the [0, 1].
view, and be stuck there permantenly when leaving the region. Now the
button interaction is cancelled when starting a modal operator, not too
happy about this, but couldn't think of another way to detect this well.
The FEdgeXDetector class for silhouette edge detection and the
ViewMapBuilder class for view map creation are enhanced. All changes
are related to view point handling (i.e., the camera location). In
the perspective projection, the visibility of feature edges and edge
types (e.g., borders and crease lines) are computed based on a fixed
view point, whereas the view point in the orthographic projection is
an arbitrary point on the XY plane at the origin and is determined so
that a line segment from the view point to a point in the 3D space is
perpendicular to the XY plane. All view point related portions of the
class definitions were modified so as to deal with the two different
camera modes.
- disallow setting swizzle's that have duplicate values: vec.xxx = [1,2,3]
- raise an error if the list setting a swizzle's longer.
- mathutils write callback was running on the wrong vector, which could crash blender.
- bpy.app moved into PyStructSequence (used by sys.float_info)
- added buildinfo into bpy.app.build_*
- bpy.ui removed (wasnt used)
- include external example files in Mathutils docs (only Mathutils and Vector are currently written)
- added support to auto document PyStructSequence's
- CMake had "'s inside all its strings.
This also means that only -s and -S are accepted to set start frame and scene (before, it would accept anything that started with s or S, you could have done blender -b file.blend -super 1 -Science "scene 2").
We really need better argument parsing...
account in quad split 3d view. Now WM_OP_EXEC/INVOKE_REGION_WIN
context stays in the current region if it is already a region
of type window, so the operator gets executed in the right context.
the texture would crash because the VBO was still bound. As I
understand it this is not necessarily against the opengl spec,
but might as well unbind it, the driver bug seems to have been
fixed but has not trickled down everywhere yet.
http://bugs.freedesktop.org/show_bug.cgi?id=23859
* Clearly labelled the way that the scrollbar hiding works. Also see the report comments for an overview
* Added another pair of flags for another one of the cases in which scrollbars should also get ignored; when the entire contents of the view are visible, a pair of flags is now set in the view2d data (instead of for the scrollers tempdata only) for detecting this case too
* Fixed the potential for scrollbars without zoom handles shown to have those handles still considered. This still happened in the User Preferences window, but has now been disabled.
--
These changes still don't solve the bug though. Currently after the scrollbar operator passes through, the Outliner's activate-selection operators still fail to start.
- added an option to BLI_convertstringframe and BLI_convertstringframe_range to add digits if not found.
- removed BLI_convertstringframe where its obviously not needed - such as loading movies and sounds.
* Old files imported to 2.5 with curves that got cycles FModifiers added during version patching, would not have their keyframes shown for editing.
* #20893: Can't see/show the Properties area in the NLA
I don't know why this worked on a few of my test files several months ago, and yet now fails for many files today.
