Age | Commit message (Collapse) | Author |
|
blocks that were previously missed; and b) greatly increase my
ohloh stats!
|
|
Having splitted a screen almost into 128 subwindows gives crashes when you
try to make a 'full window'. Code was missing NULL pointer checks for when
new window failed to open...
|
|
the amount of subwindows reaches 128 (64 areas+headers).
Added a warning instead of crash!
|
|
- Texture Node: now displays 'intensity values' in node too, and has input,
and shows in buttons when activated in Node editor. (no browsing buttons
yet...)
- New: "Normal Node". This uses a new UI button, which allows to quickly
input a normal vector, based on spherical coordinates.
The Normal Node has optional vector input, and delivers a dot product
then. This can be used as a blending factor between nodes, or for fake
extra light in a certain direction.
- New: "Geometry Node". This actually replaces the Input node. It offers
all coordinates (vectors) as being the starting point for shading and
for textures. Note: for preview render this doesn't give much different
results yet... this is the start for real render support!
- http://www.blender.org/bf/rt5.jpg
The two new nodes in action
- Bugfix: the "Block" button (which delivers popups) did not return a
correct event when nothing happened (mouse moved out), which could
cause mouse clicks to be passed on to the queue.
|
|
cpack(index_to_framebuffer(index)) but that is all that is ever
used.
|
|
NOTE: BLI_winstuff.h was meant to be a wrapper around windows.h to handle
undefining various crap that windows.h defines. Platform specific headers
should only have to be included in a few places. This reduces the number
of inclusions of BLI_winstuff.h to 16 which is a much more reasonable
number (than the 144 or whatever it used to be)
|
|
lot of casts, added prototypes, missing includes and some true errors
|
|
Backbuffer selection codes issue on OSX.
On some h/w configs, with "1000s of colors set", the drivers seem to round
colors up or down... whilst until now rounding always was just truncating
bits. This gives errors in writing color codes and reading it back.
The c code only changed for OSX. For other OS's and HW this test could be
done as well.
|
|
Found out the conversion functions for indices <-> framebuffer only went
up to 18 bits, making selection in zbuffer-select (or vpaint/faceselect
even) not work on large datasets. Provided you have 24 bits display, of
course.
This limit was just dangling in the code since long ago... modern times eh!
|
|
in mywindow_build_and_set_renderwin(...
curswin= 2; was assigned too late
|
|
----- Killed UI frontbuffer draw
The interface toolkit was drawing all live updates (while using menus/buttons)
in the frontbuffer. This isn't well supported cross-platform, so time to be
killed once. Now it uses *only* glReadPixels and glCopyPixels for frontbuffer
access.
Live updates or menus now are drawn in backbuffer always, and copied to
front when needed.
NOTE: it was tested, but needs thorough review! On PC systems I suspects
backbuffer selection to screw up (check!). On SGI/SUN workstations it
should work smooth; but I need evidence
----- Smaller fixes;
- AA fonts were garbled on ATI systems. Now the AA fonts are drawn exact
on pixel positions. Needs the new FTGL libb too, patch is on maillist
- Rounded theme uses antialiased outlines
- Pulldown and popup menus have nice softshadow now
- New button type 'PULLDOWN', thats the one that callsup a pulldown menu.
Should be added to themes, as is the full menu/pulldown drawing
- Screendump for 1 window does the full window now, including header
- Empty pulldowns (for example running blender without scripts) give no
drawing error anymore
For review & fun;
- added curved lines as connectors, for Oops window
|
|
osx with AUX buffers, than on other systems with GL_BACK buffer...
Now i've added a check in force_draw to also redraw the backbuffer, if
needed. ANother weak point in the sublooping system in blender.
|
|
- borderselect draws cleaner info in bottom/left corner window
- OGL render (in view3d header) option didn't work on successive renders.
was due to removing hack that reopened the window each time. But now it
needed to be assigned correctly to blender's 'mywindow' system.
|
|
- Zbuffer clipped selection
Based on same algos as for drawing (blender polygon offset) vertices, edges
and faces now are clipped for draw and selection when Zbuffer is used.
Note that it works for endpoint vertices of edges, and for facecenters.
Also works for border and circle-select
- Optimal draw subsurfs
This now draws optimal with faces and edges selected/unselected, hiding the
original 'cage' (mesh) completely.
TODO: edge select, which still uses original (invisible) edge.
http://www.blender3d.org/cms/Mesh_editing_rewrite.425.0.html
|
|
This is supposed to fix gradual slowdown of Blender interface on ATI cards.
|
|
"Color depth R=8 G=8 B=8" lines.....
Kent
|
|
- added: when start blender in debug mode (blender -b) it prints the depth
of framebuffer.
|
|
colors. This because of the pretty weird (ab)use of load & make editmesh...
For each added undo step, the load_editmesh was fed with an empty mesh
to assign data to, without knowledge of what was in the original mesh.
That way UV and color data got lost.
Solved it in 2 steps:
1. removing the ->tface pointer from EditVlak, and make TFace a builtin
struct inside EditVlak. This didnt cost much extra mem, since it already
stored UV and color. This enabled some pretty cleanup in editmesh.c as
well, storing tface pointers was cumbersome.
2. for each undo step, it then generates always a tface and mcol block to
link to the undo Mesh.
Even when it wasn't in the actual Mesh, at exit editmode the original
Mesh is used as reference anyway, and undo-meshes are freed correctly.
The enormous commit is because I had to change the BLI_editVert.h file, and
found it was included in about every file unnecessary. I removed it there.
ALso found out that subsurf has code ready (unfinished) to make UV coords for
the displaylist in EditMode as well, nice to know for later...
|
|
the 'OB' and 'ME' buttons were not visible in all situations anymore...
|
|
all calls to ortho2 with correctness offset of 0.375 instead of 0.5.
this efficiently solves bug in drawing UV lines as reported.
cvS: ----------------------------------------------------------------------
|
|
using an identical offset; which is 0.375 and official recommended
by OpenGL.
This to further investigate the AA font errors on some systems
with ATI cards.
|
|
chapter in 'opengl programming guide' you have to offset the matrix
with 0.375 for exact & predictable cross platform pixel drawing.
Blender used 0.5....
Committed for test among platforms... if you see weird drawing results
(like polygons not aligning good with lines) just notify me.
|
|
window (type)
- each SpaceData struct (not the window!) can get 'block handlers'
assigned, basically event codes that invoke drawing button panels.
- this is saved in files, and Panels behave in any window like it does now
in buttonswindow
- it also means that a 'space window' should leave with a matrix set for
buttons level
- try it in view3d header menu, 'view'->'backdrop'. this opens the old
viewbuttons
- it all works non blocking! instant updates of viewbuttons visible in
3d window now.
Not done yet:
- checking and fixing frontbuffer drawing (select a wireframe draws over)
- temporally vertices cannot be selected, is my next project
- closing or hiding Panels...
- styling stuff... i committed for others to review as well.
Have fun. this is certainly a huge improvement over the old viewbuttons!
|
|
- removed src/buttons.c and include/BIF_buttons.h
- added src/buttons.txt, which is the old buttons.c for review and adding
code to new panels structure
- changed internal events to match new buttonspace structure
- added tabs for new shading group of buttons
- removed loads of little warnings, -Wall now compiles src/ almost without
error (hint: setenv NAN_QUIET to see it all better)
Now I'm ready to do actual buttons -> panels conversion. I will do the raw
versions first, others then can cleanup
|
|
All modifications are behind compileflag INTERNATIONAL,
and affect these two directories; source/blender/src and
source/blender/ftfont.
Thanks to philipp, a new utf8towchar() function is added
to Shizu's work, which fixes the international support for
OSX. It also makes the iconv library obsolete. This means
all translation files (blender.mo) need to be 'recoded' to
UTF-8, and language support in the fileselect window and
the text editor is not possible. Iconv might be added in
the future to enable these features.
|
|
Might be possible I mised an .h or so, just notify me in that case.
|
|
instead of passing seperate arguments. For when we want to pass 'richer'
events around.
|
|
#ifdef HAVE_CONFIG_H
#include <config.h>
#endif
Just need to finish cpp files now :)
Kent
--
mein@cs.umn.edu
|
|
|