Spen't lots of time debugging the misbehaving COLLADASW::URI
constructor. The problem is although I pass correct parameters to it,
the resulting file ends up having empty url attributes. It is either a
bug in OpenCollada or my code - couldn't figure this out.
OpenCollada and pcre are added as 3rd party libraries collada module
depends on.
I now build Blender with:
scons BF_DEBUG=1 BF_OPENCOLLADA_LIBPATH=/home/chin/src/opencollada/ BF_OPENCOLLADA=/home/chin/src/opencollada
- setup SConscript, not sure about these values: libtype='core', priority=150
- only scons build on linux is set up
- not configured to link to OpenCollada yet
Fonts like Type 1, have one file with the glyph image and another
file with metrics and kerning information, this try to search if the font
have this information and load (try open the same file but with the .afm
and .pfm extension).
Also add a function to load the same information from memory, just in case
that in some point we add a font like this.
Now that we only work with Freetype2, I don't see any point to keep
wrapping the functions.
Also remove the reference code, it's something that we don't go to used.
* Replaced a few duplicated enums that were also defined for some RNA properties already with the RNA ones.
* Start of cleanup of markers code in preparation for getting the rest of the editing code involving markers working again
1) Remove WITH_FREETYPE2 from code, so now blender always need freetype2
2) Remove the old bmfont
3) Remove ftfont and bFTGL library
4) Implement a new BLF_draw_default function for place that still need/use
the old BMF api.
I try to update both, scons and cmake, but I only can test with make, so
hope all work fine.
MSVC is broken, but I don't have Windows, things to search and fix are
any reference to WITH_FREETYPE2, FTGL and BMFONT (take in care that
blenkernel also have a BKE_bmfont.h, this don't have anything to do with bmfont).
Always have to link/include the freetype2 library
Remove any reference to libbmfont
Remove any reference to libftfont
Remove any reference to libbftgl (or libbFTGL)
* Added 'Replace Keyframes' entry that appears in addition to delete, since it is often useful to be able to directly replace values instead of having to delete and reinsert keyframes. Internally, this just calls the Insert Keyframe operator.
* Separated the code to for Driver editing from the tests for keyframes, since they're independent.
* Added separators between keyframe and driver operations. This could be removed if the UI-mafia consider it overkill.
* Loading files saved with 2.5 works again now without crashing. Was crashing when trying to load Grease Pencil data (even though there was none)
* Fixed some msvc compiler warnings in own code
Time-Modifying F-Curve Modifiers now get special callbacks to allow them to specify what frame they need to be evaluated on, instead of forcing a re-evaluation of the preceeding curve + modifier-stack. This should be more robust than the old way in general.
It still remains to be seen if some tweaks to this are still needed, as the full consequences of the propogation of modified time-spaces have yet to be fully explored.
For now though, evaluation works by finding the last modifier on the stack which modifies time, and asks it what time it modifies the given time to. This modified time is used to evaluate the F-Curve data only. The modifier stack gets evaluated using the original time instead.
This has been wrapped under Scene.RenderData, and can currently be accessed under the 'Image' panel in the Scene settings.
As in the old code, image types are only included if they are enabled when compiling. Also note that image-type specific settings have not been wrapped yet.
Cleanup of conventions for icon placement in menus.
Old code only allowed 1 icon per menu, forcing to pass on the sublevel
menu icon with exception handling.
Now sublevel icons are drawn automatically, allowing to add an icon
before name too. (Check shift+a menu)
A little experiment in putting icons on this menu as per a mockup from William.
However, it appears that there is a minor bug here though with the menu drawing code. I've deliberately left the Mesh entry without an icon for now to illustrate part of the problem (note the placement of the icons for curves and surfaces)
* Made separator item work horizontal & vertical.
* Add colon (:) automatic for int/float/enum/string.
* Added space variables to uiStyle and use them in the
layout engine.
* Added initial World buttons by Thomas Dinges, thanks!
* Added some code for modifiers in the Object Data context.
This will become a template though.
* Use a common poll() callback in the scripts to reduce code.
Summary of ain features:
- Themes and Styles are now editable.
- CTRL+U "Save user defaults" now goes to new .B25.blend, so you
can use 2.4x and 2.5x next to each other. If B25 doesn't exist, it
reads the regular .B.blend
- Press Tkey in 3d window for (unfinished) toolbar WIP. It now only
shows the last operator, if appropriate.
Nkey properties moved to the other side.
A lot of work was done on removing old themes for good and properly
getting it work with the 2.5 region system. Here's some notes;
- Buttons now all have a complete set of colors, based on button classifications
(See outliner -> user prefs -> Interface
- Theme colors have been extended with basic colors for region types.
Currently colors are defined for Window, Header, List/Channels and
for Button/Tool views.
The screen manager handles this btw, so a TH_BACK will always pick the
right backdrop color.
- Menu backdrops are in in Button theme colors. Floating Panels will be in
the per-space type Themes.
- Styles were added in RNA too, but only for the font settings now.
Only Panel font, widget font and widget-label work now. The 'group label'
will be for templates mostly.
Style settings will be expanded with spacing defaults, label conventions,
etc.
- Label text colors are stored in per-space Theme too, to make sure they fit.
Same goes for Panel title color.
Note that 'shadow' for fonts can conflict with text colors; shadow color is
currently stored in Style... shadow code needs a bit of work still.
