* less calls to BKE_image_get_ibuf when switching between images
* remove thread debug prints
* use own pixel blending funcs (IMB_blend_color isnt needed in some cases), slightly faster without this.
- Added markers (partially) back, ed_markers.c (temporary in spacetime,
where to put general stuff like this?), and added ED_markers.h for api.
- Mkey in timewindow shows marker (no icon, doesnt work yet).
(Adding operators gets easy :)
- Noted some conflict for operator storage... with markers being used all
over, added it to the screen level.
- More todos: view2d is stored in space, for time, etc. With regions we
could move this over, but better todo as post 2.5 project. Multiple 2d
views in area would be cool though.
* Disable editable pointers for now, difficult to support well.
* Swap parameters in RNA_access.h functions to make it more
consistent.
* Rename rna members for operators to wmOperatorType.srna, and
wmOperator.ptr, to make the distincton a bit clearer.
• Removed the RNA_int_default and similar functions, they're too
confusing. RNA_property_is_set can still be used to achieve
the same goal.
* Add functions to create RNA pointers.
Some example code for RNA data access and operator properties:
http://wiki.blender.org/index.php/BlenderDev/Blender2.5/RNAExampleCode
to compile blender with gcc on IRIX, IRIX_USE_GCC needs to be set to true in
user-def.mk.
Other changes related to irix:
* compile solid from extern/
* don't build plugins (yet) with "make release" when using gcc (the shell
script used assumes MIPSpro is installed)
* use statvfs instead of statfs on irix, like done on solaris
* use external libs from $(LCGDIR) instead of /usr/freeware
* use glew header files from $(LCGDIR)/glew instead of the ones installed on
the system (this applies to other platforms as well)
* ffmpeg support currently is disabled on irix
* Wrap most user editable data in DNA_mesh_types.h and DNA_meshdata_types.h.
Still needs to be improved in some areas though, especially how to deal
with data layers (expose per element or as array, or both?), and data in
face corners (bmesh type data structures are more logical here).
Tweaks to RNA defining to make some cases easier:
* Added range callback function for int/float.
* Added 'skip' callback for listbase and array iterators to skip items in
the collection.
* Extra error print when calling wrong define_property_*_sdna functions.
* Also made button code respect non_editable flag, is quick change though,
need to add support for properly graying out etc.
- Added standard "tweak" gesture operator, which can be set per region, to
generate EVT_TWEAK events. You can configure tweaks for any mouse button
and have handlers for such events check for modifiers etc.
It even stores tweak direction (8 directions). Might be fun to experiment
with tweak gestures N, S, etc. :)
In general it can be used to replace the current tweak code in 2.48
(std_rmouse_transform).
Test added: on screen level it now adds LMB tweaks, if tweak-South it splits
the area. Will be removed of course.
- Added to Border operator a property to store event used to end border with.
- Moved the "AZone" triangle drawing to the right context (area). It was on
screen level, not respecting area-redraws. Also cleaned up drawing for it,
and moved the "swap buffers indicator" square to look nicer. Those squares
are only for test!
- event-match function had bad code for checking for event-value. Made a
"KM_ANY" define so keymaps can be defined ignoring event values.
- Gesture todo: lasso, "real gesture" (like blender now has)
* float buffer support for all painting operation and undo.
* only run brush_sample_tex() for textures brushes.
* redraw the brush outline even when nothing is painted.
Reset transform restriction flags when switching to other transformations while running.
Also don't draw constraint if no constraint flag is on (old annoyance).
Display the number of joints in the selected template in the sketch panel
When using selected bones as template, turn off Connected if bone has a parent that is not copied.
The reason for this was with transforming screenspace points outside the face into UV space and when the bucket bounds was enough outside the face.
For faces viewed side on, this transformation would be applied incorrectly (a bit like trying to apply a projection matrix to a point behind the view), the buckets UV space coords would be incorrect and the wrong pixels would be initialized for that face.
solution is to clip the screenspace face with the bucket before getting the UVs. This should also be a bit faster since the clipped polyline will have a smaller bounding box.
* Added RNA for operators. This still uses ID properties internally,
but through the RNA API now. The OP_get/set_* API that was used is
replaced by the RNA API. Currently RNA properties for operators are
defined at runtime since it means operator registration can be done
in a single function.
* Changed the existing operators to use this system, I haven't defined
user interface names yet though. I also think there need to be some
conventions on which properties to expose to make these operators
usable in macros, for example if mouse coordinates should be stored
or not.
* When using ID properties through defined RNA properties, it now
checks that the ID property actually matches the RNA property and
removes/overwrites it otherwise. This ensures that you can safely
get/set arrays for example without having to worry that some
external thing may have changed the length.
* Documentation now has some information on RNA + ID properties.
http://wiki.blender.org/index.php/BlenderDev/Blender2.5/RNA
* faces that were ignored were also not taken into acount when checking UV seams - causng bleed not to work properly in some cases.
* commented early pixel filling loop exit that fails in some cases.
* More ID property support. What was already possible was showing
ID properties as RNA properties. Now it is possible to define
RNA properties and have an ID property automatically created the
first time it is set (if not set it retuns the default).
* Added support for defining RNA structs and properties at runtime.
This is useful for python and plugins, and could also be used
for operators, not sure yet what is best there, they could be done
in preprocess for speed, but not sure how to do that while keeping
operator registration a single function.
* Added quick functions to get/set properties based on names, to be
used for operators.
* Added some simple support for inheritance, was already doing this
but having it as a feature simplifies things. Two things were added
for this: when defining a struct you can give a 'from' struct whose
properties will be copied, and structs like ID, operator, modifier,
can define a refine callback that will return the more specific type
of the struct like ID -> Object, Mesh, .. .
* Added simple windowmanager wrap with only the registered operators
list, used for testing RNA for operators.
- operator definitions, callbacks, registry to WM and handlers for it are
now always in a file xxxx_ops.c or xxxx_operators.c, in the bottom you
will find the registry and handler code.
- fixed some confusing naming conventions "rip_area vs area_join" etc. Now
stick to convention to first name subject, then operation (like UI :).
So it's area_rip, screen_add, and so on.
- Nicely put exported calls (outside module) together in bottom: this using
names such as ED_screen_duplicate().
- Moved Operator-Property API to new C file.
- Simplified and cleaned previous border code
It was a bit too complex, too many data manipulations
Original idea was to have WM API calls to manage border, circle, lines,
lasso, etc. This now means that WM provides callbacks for custom operators,
so it's very easy to make them. Check bottom of screen_edit.c for an
example.
Currently two borders were coded; with and without cross hair.
Press Bkey in any area-region to test it (note: time window has wrong matrix!)
Some specs to note:
- gestures are in region space, and draw 'over'. That latter still needs some
work when we do real composites.
- only the active region is redrawn.
- on todo is the generic gesture engine for 'tweak' or like how currently grab
gestures in Blender work. These will be configurable per area-region, and WM
then will send the proper "Gesture Event" with properties (N, S, E, W, etc)
to which you then can assign operators. Such events will be generated with low
priority, so other handlers who swallowed mouse events have preference.
void bicubic_interpolation(struct ImBuf *in, struct ImBuf *out, float u, float v, int xout, int yout);
void neareast_interpolation(struct ImBuf *in, struct ImBuf *out, float u, float v, int xout, int yout);
void bilinear_interpolation(struct ImBuf *in, struct ImBuf *out, float u, float v, int xout, int yout);
Added...
void bicubic_interpolation_color(struct ImBuf *in, unsigned char *col, float *col_float, float u, float v);
void neareast_interpolation_color(struct ImBuf *in, unsigned char *col, float *col_float, float u, float v);
void bilinear_interpolation_color(struct ImBuf *in, unsigned char *col, float *col_float, float u, float v);
This is needed so for projection painting but generally useful if you want to get the interpolated color of a pixel in an image without having a destination imbuf.
While editing these I noticed the functons are a bit dodgy, they assume the input ImBuf has matching float/chr buffer to the output.
Tested overall speedup is about 5x when scaling 4096x4096 -> 4000x4000 in the sequencer.
There were some artifacts in the resulting image but double checked and the old code gives the same problems.
Added back old code with #if 0's since its a bit more readable.
General optimizations
* Precompute 1/x when dividing by x multiple times.
* Use float constants like 0.0f instead of 0.0, avoids conversions from float to doubles and back.
ProjectPixel
* make pixel (and similar pointers elsewhere) a union with a float and unsigned int pointer to reduce the number of casts a little. generally there are a lot of casts going on in the code, makes it hard to read.
project_paint_begin()
* the perspective case checks with (*projScreenCo)[3] > 0.001) for faces behind the view. - Changed to use the clip start from get_view3d_viewplane
* removed arbitrary check for brush size to disable threads.
imapaint_paint_sub_stroke_project()
* Make clone tool use IMB_blend_color to reduce the code and support blend modes.
imapaint_paint_sub_stroke_project_mt()
* Make threaded and non threaded mode use same function (just dont start threads when its set to 1)
* removed PIL_sleep_ms, was not needed and slowed down threading (my bad!, was copied from bake code).
- moved from WM to Screen code (it uses active area)
- less code :) result of cleaning some calls
- added WM_window_open() to WM API for this
- now opens new window on top of area, and leaves old screen unaffected
(simple, atomic, the 'do not think for user' convention :)
* Added start of lamp wrapping (code by Michael Fox).
* Add back Object.data, was crashing with unknown data type.
* Added support for using consecutive variables like float r, g, b;
as an array without writing a manual get/set function.
* Also note the RNA documentation is updated now to be more
about how to define RNA and use it, including some diagrams.
http://wiki.blender.org/index.php/BlenderDev/Blender2.5/RNA
* use a faster method of getting a pixels screenspace location.
* check if its possible to break out of pixel to bucket/face intersection early - ~7% overall speedup (ignoring redraw time).
* removed scanline intersection function (added back incase they were needed but it looks like there not)
* speedup for painting with only 1 image (use a loop without context switching checks)
* more commenting + cleanup
