Blender Git Loki

Updated channel listing for help with compiling Blender to reflect (once
again) the existence of #blendercompilers. Also added link to updated
compiling guide found here:

http://www.blender3d.org/cms/Building_with_Cygwin.524.0.html

Thanks to Tom M. and joeedh for pointing out that it should be updated.

Small fix; when pulldown menus are too large to display (they get aligned
with top/bottom to active button) the menu aligns from bottom of blender
window.

This already happened for File->Export menu

- Removed temporal hack that switched manipulator type on G, R, S hotkeys.
That was blocking immediate hotkey access for normal blendering.
- New, experimental, hotkey to switch: CTRLKEY! Added code in such a way
it detects a CTRL key press and release without using it as modifier.
Pressing CTRL will cycle through the 3 manipulator types.

- Fix for yesterday's commit; click on centerpoint for switching
orientation happened too on a click for other widgets. Not nice...

- Nothing to see hear, move along, move along.

Reordered some of the includes to fix STL issues.

For details, see:
http://www.blender.org/modules.php?op=modload&name=phpBB2&file=viewtopic&t=5867

Kent

Added call in creating transform data Mesh editmode to calculate vertex-
normals for the ALT+S "shrink/fatten" option.

add files for Daniel Dunbar's new CCGSUBSURFLIB work to SCons build.
This commit *only* adds the new files.

Since this is an experimental feature, it is NOT enabled by default.

At the moment, you can turn it on by editing the toplevel SConstruct
file. NOTE: this is a hack. it will be done properly later!

To enable this experimental feature,
add the following line after the line that starts with 'extra_flags = ['

extra_flags.append( '-DUSE_CCGSUBSURFLIB' )

- Hacked in blender style creasing support to new subsurfs.

I would like to have proper creasing (following what appears to be
the informal standard in the subdivision surface literature) but
I do not know when I will get to this or how complicated it will
be. With this patch CCGSubSurfs should now have the same capabilities
as regular subsurfs (barring possible bugs) and so replacing the
old implementation is a reasonable possibility.

BPython:
-- Stephane Soppera (thanks) reported libc stat function fails for paths ending with "" under win with free VC++ compiler toolkit 2003:
removed final '/' (BLI_make_file_string changes the '/' to '\\' for win) slashes from relevant paths, that should take care of it . Note: here (linux, glibc, gcc 3.3.3) stat doesn't have this problem. Also checking if U.pythondir ends with a slash and, if so (as long as its length > 2 to), removing the slash, for the same reason.
-- small cosmetic changes in BPY_menus.c for debug msgs and in header_scripts (added a separator in the Scripts win -> Scripts menu).

Transform widget update

- now support for global/local/normal orientation.
- LMB click on center switches orientation mode
- in object mode, local (now) only displays on single object selected
- in editmode, Normal orientation is derived from faces selected
- if no normal can be found, it shows local orientation

Currently implemented, for test, in Mesh editmode and PoseMode.

Note for PoseMode; the 'translation widget' shows on "IK" bones but
doesn't translate, it rotates instead. Pretty interesting to use the
translate widget for it...least cluttered display I think.

Note for Matt; I tried MMB click for switching orientation... it's just
weird that way... such clicks, repeatedly, with mousewheels isnt nice
either.

Added option to use new subsurf module added by Daniel Dunbar. Simply add:

export NAN_NEW_SUBSURF=true

to your user-def.mk file.

Size flipping, for kaito. Move pointer to the other side (horizontal) of the pointer to see. (Does affects size member, so just object position and edit mode)

Fix for possible divide by zero error in Rotate.

Fix for MMB behavior when two axis were exactly on one another or very close. It now defaults like this: X, Y, Z (meaning if as near as X as Y, it chooses X). This could be fixed further.

Created initConstraint. It basicly just checks if the CON_APPLY has been set up (by the BIF_setConstraint calls for example) and calls startConstraint. This must be done because startConstraint uses data initialised when starting transform so it needs to be called after that.
Also changed some strcpy into strncpy.

Added string to BIF_setSingleAxisConstraint() function for headerprint.
Needed for martin to further work on print stuff

Shrink/Fatten behavior change. This time for good I hope.

Technically, it now works by getting the mouse motion in 3D (just like Translation/Grab), projecting it on the vertical view axis, using the vector length as the shrink/fatten factor. If the motion is downward (on the screen), shrink, if it is upward, fatten.
In layman terms: move up to fatten, down to shrink and it adapts to the viewport zoom, like Translation.

I changed to snapgrid factors to match those of Translation to, so it really acts like grabbing.

Transform:

- cleaner code for selection of handles in transform widget. in ortho views
selecting the center handle defaults, to prevent Inf. overflows for
perpendicular handles. While holding SHIFT it works the opposite!

- removed redraws on modifier (CTRL, SHIFT) key release. that ensures you
can apply to grid nicely (functioned that way in all blender releases)

More transform widget goodies;

- added 'ghosting' for while using translate/scaling widget
- added 'pie chart' ghost to denote angle while using rotate widget
- added settings to tweak widget in User menu (InfoWindow) "View & Control"
- Size: total widget size as percentage of window size
- Handle: as percentage of widget radius
- Hotspot: for clicking handles, in pixels
- Fized Size: option to make Widget size independent of window size
Not sure if all of these are useful to keep, but makes for good testing
in this stage.

Also: made #define to use new transform to be set TRUE by default now. :)

Scripts (making some changes to the scripts dir):
- moved bpydata/ to scripts/bpydata/ and added a config/ subdir to it;
- created scripts/bpymodules for py modules (also got rid of those "mod_"'s appended to the files);
- updated scripts accordingly.

This will require you to "reinstall" (just copy the scripts/ dir over your older one) if you have a .blender/scripts/ dir somewhere. Otherwise some scripts won't work. You can use the updated "Help->System->System Information" script here to check all is fine. An installer script yet to be written will help users with these issues, specially to make the user defined dir have the same structure expected from the default scripts dir, so the basic facilities (module search; saved config data; scripts: installer, help browser, config editor) are also available for a user's own collection of written and downloaded scripts.

BPython:
- slikdigit's crash was because he had no <home or blender exe location>/.blender/:
proper check added and also now if all else fails the <cvsblender>/release/scripts/ dir is also searched for scripts. All this registration dirs stuff is a little messy (installation!), so please report any troubles (I only tested on linux).
- slight change in error report in BPY_interface.c's BPY_menu_do_python; remembering to set globaldict pointer to NULL there, too.
- moved bpy_gethome() to EXPP_interface.[ch]
- "//" as user defined python dir is ignored while looking for scripts, considering it's only a default some users use, not really meant for a scripts dir.

- update new conversion to DLM routine to use match subsurf
(means optim, seams, selection should work same now, but I am
not super familiar with all this stuff so can't test very well).

These hacks to the DLM structure are disgusting btw Ton. What
a waste of memory! All the information that is so meticulous to
kept and managed in the old structure is essentially explicit (or
easily made so) in the new one.