Blender Git Loki

-> Fix for Mesh Properties Python API

When switching Mesh properties over to METH_O from METH_VARARGS certain method definitions
didn't get switched over correctly, causing calls to mesh property functions to fail.

-> Selection History Fix

Old bug in selection history meant that face selections were getting erased from
history on editmode toggle.

Added active drawing in editmode (may need some more work), this hilightes the last selected vert/edge/face, which is needed for moving some of the UV tools into editmode.

Added copy menu for editmode for copying from the active faces material, image, UV's vcols etc,
improved on the current vertex color copy by averaging the 4th color when copying color from a tri to a quad.

And added edge crease and length copy

cerrently verts dont have a copy menu but eventually copy weights would be good.

udates from stable

Bugfix:

Copy/pasting multiple ipo curves in editmode, segfaulted due to some misplaced code.

This does nothing to fix the irix stuff but gets rid of a couple of simple
warnings that showup on irix.

Kent

* AO bugfix - Constant QMC sampling wasn't getting initialised for bake rendering, causing a crash.

-> Multires Editmode fix

Typo in Multires Editmode integration meant that way too much memory
was getting allocated for a pointer array (more than 10 times as much).

-> Added mesh properties to modifiers

Mesh properties layers were not getting copied in the modifier stack.

-> Fixed initialization of Octree Resolution and Tri2Quad threshold

add_scene() in scene.c didn't properly initialize the octree resolution
for new scenes or the join threshold for Triangle to Quad conversions.

some button arranging

(UV Calculation tab is nolonger relevent to FaceSelect draw modes.)
Move FaceSelect draw options into WeightPaint and VertexPaint panels.
Moved UV Calculation panel in with shape keys tab.

draw shadowmesh wasnt working and the button in the header for selecting faces in paint mode was drawn in editmode.

Changes to "Face Select" mode
* Does not indicate that UV's exist, nor does it add UV's when used.
* Only accessible for texturepaint, vertexpaint and weightpaint from a button in the header (Paint Selection Mask)
* Not accessible from the mode menu, this is only an option that applies to paint modes.

This dosnt effect DNA, face select (G_FACESELECT) can be enabled at any time but is only used when paint modes are enabled.

Other changes
* UKey is uv unwrap in editmode, Ukey for undo was editmode only anyway.
* UVCalc in editmode adds a UV Layer if there is not one alredy.
* texture draw in editmode does not draw the face dots (they are get in the way of texturing)
* some missing updates were added.
* removed manipulator from when paint modes are enabled since the manipulator is not drawn in the 3d view.

Version bump to 2.45 and splash screen for stable release

UV Editing is now done in editmode rather then UV/Face Select mode.
Notes
* you cant edit UV's in the image window in "UV Face Select" mode. (removed UV from the name)
* going into Face Select mode no longer adds UV's and does not need UV's to work.
* The UV Calculation menu is now in editmode (Alt+W)

Todo..
* Image replace - partly broken in stable also.
* Rotate/Mirror UV/VCol are still only in Face Select mode.
* Hide/Reveal is not quite right, (issue with editmode flushing)

IPO Copy/Paste Bugfix:

This commit fixes two minor mistakes made in the previous commit for this:
* forgot to advance to next ipo curve from buffer
* handles on the curves that got data pasted were not recalculated (those from the buffer were instead!)

Action Editor - minor cleanups of transform code...

Bugfix #7294:

When resizing an Outliner window, the contents would eventually get 'pushed out of view' when shrinking the view.

I've added a bit of a 'hack', which will ensure that this doesn't happen, by forcing the view to look at the left-side of the outliner tree, when the width of the outliner window decreses due to resizing.

* Raised some hard coded limits such as mist/envmap clipping ranges

Fix for edgesplit crashes reported in bugs 6695 and 7142. The problem, as
found by Riku Palomäki, was that "bridge" vertices (which connect two or more
faces that don't share edges) were not being included in the maximum final
vertex count calculation (used for memory allocation, hence the crashes).

Unfortunately Riku's patch actually stopped bridge vertices from being split
correctly, so I have fixed the problem by adding the maximum number of vertices
that can be generated from bridge vertices to the maximum final vertex count.

Thanks to Riku Palomäki for finding the cause!