Blender Git Loki

* Preferences for displaying the grid floor, X axis, Y axis,
Z axis. (ported from tuhopuu2)

Controls are found in the 3D View Properties panel (with
the rest of the grid settings). This is more flexible for
people like me who hate the grid getting in the way when
modelling - turning off the grid floor and turning on the Z
axis gives a setup like in Wings 3D or Clay.

* Cleaned up the 3D View Properties and Background Image panel's layout while I was adding the new buttons (including change over to UiDefButBit).

- when not F10->"Ray" option is set, materials with "Ray Transp" will
render solid now (no alpha).
- This gives nicer previews, but also makes envmaps look better, since
environment maps are rendered without raytracing
- I decided not to raytrace envmaps mainly because of speed... if you use
environment maps you want something quick... otherwise just use ray_mir
material here!

- forgot to commit this file for fix several days ago which fixed
only_shadow rendering...

- another fix for incorrect Osa vectors during tracing... I've now solved
it by precalculating all needed Osa vectors when rendering glass, this
because it is unpredictable what is exactly going to be needed after.
- a hint that this should be done better is in the code... will mostly
improve sharpness and a bit rendertime though
(thanks bugmaster intrr!)

- when pulldown menus get flipped, the arrow keys were not behaving OK.
also added that left/right arrow is handled, to prevent weird redraw.

Blender's debug mode only worked on startup:
- G.f's G_DEBUG flag was being erased in blenkernel/intern/blender.c's
setup_app_data:

G.f= bfd->globalf

// added a line above it to fix this:

if (G.f & G_DEBUG) bfd->globalf |=G_DEBUG;
G.f= bfd->globalf;

BPython:
- debug info now only shown if Blender is started with '-d' option
- added ~/.blender/scripts to modules sys.path
- added two new functions to Blender.sys: basename and splitext
- added doc for Blender.sys, updated other docs

- fix for raytrace render; when you render ray_transp material, it didn't
recalculate the O.dxno and O.dyno for each pixel. This causes scanline
errors (stripes) when in same scanline something else is called that
calculates O.dxno/dyno...

#ifdef INTERNATIONAL fix for usiblender.c
thankyou Todd Koeckeritz.

- FTF_TTFont didn't initialize fontm,fontl,fonts to NULL, leads to
possible bad free
- FTF_Api.cpp used static C++ object, hard to understand and icky,
was probably source of phase4's crashes, replaced with dynamically
allocated object and added FTF_End to free it and sensible time.

BPython:
- local tentative fix for BLI_gethome(), which returns '.blender' appended only
on some Windows systems. Created bpymenu_gethome() to check and
add '.blender' if needed.
- changed name: .Bpymenus to Bpymenus as suggested by GSR
- trivial additions:
Object module: added methods .set/getSize
Armature/Bone module: bone.set???() methods now accept both n
floats or a list of n floats: fff or (fff). All these additions were requested
by user Carlos Lopez (Klopes).
- New doc: for module Registry.

Bug Fix: bones get updated after using NKEY in pose mode (not really
useful, but anyways ...).

Bug fix: meshes deformed by armatures were not getting updated
when rendering an animation, i.e., the deformation was delayed.
Please test by rendering any animations you have that involve armatures.

I modified it so that the particles on a face's position are affected
by the seed variable. Requested by S68.

Basically they move slightly when you change the seed. Might be
a good idea to split this up so there are two seed options in a
particle effect, however there are already quite a bit of them so I opted
for this.

Kent

- increased max targa size to 8192x8192
was done before by Phase in tuho1, and no error found...

- added clipping value for spothalo rendering. Unified render doesn't like
alpha>1.0, the 'threshold' calculation in vanillaRenderPipe.c then works
wrong... not sure if this should be fixed there.
- for now, the spothalo render function itself clips.
- again; thanks to horrible intrr test scene! :P

- fix for specularity calculus. Because of new area lamps and new ray-shadow
some tests where moved around, causing specularity being calculared when
light actually shines behind a face.
Thanks inttr for the (horrible!) test scene that showed it. :)

New loopcut version. much praise goes to Johnny Matthews who implemented precision cutting.

new features:

- after choosing a loop to cut you go into a second mode that lets you choose where exactly on the edge you want to cut.
the placement is in percentages, so 0% is one side of the edge, and 100% is the other side.

- holding CTRL snaps the placement to whole percentages. 1.00 instead of the standard 0.01 of a percentage.

- Pressing S while you place the cut turns on Smooth-subdivide for the cut (or as I like to call it: Loopcutsubdividesmooth :)

- the percentages and the ON/OFF for smooth cut can be seen in the view3D header.

fixes:

- loop takes into account hidden vertices, this solves bug #895.
- Cutpreviews are drawn correctly for triangular faces.
- renamed the function from loop -> loopoperations. I'm sure someone has a problem with the new name too, but BLAH! :D
- the parameters for the function are defines. loopoperations(LOOP_CUT) cuts, and loopoperation(LOOP_SELECT) selects (duh). this is changes in all the places the function gets called.

if people find new bugs, feel free to yell! :)

Roel

findnearestedge() ignored hidden vertices. both vertices of the edge should be non-hidden and at least 1 of them should be on-screen (visible in the 3D viewport)

some MSVC cleanup.

- removed DNA_script_types from makesdna project.
- fixed a few errors in the projectsettings of the yafray project...it's a wonder people actually got it to compile.

- fix for bug #903
this was an error as reported more, with horizontal lines in raytraced
renderings. It appeared to be an Osa struct being not reset to zero
for normals... only happens when using bumpmapping.