Campbell Barton [Thu, 29 Mar 2012 13:44:30 +0000 (13:44 +0000)]
added Mesh.calc_tessface(), needed to update mesh tessface after bmesh edits.
also add py api BMDeformVert.clear()
Campbell Barton [Thu, 29 Mar 2012 13:09:07 +0000 (13:09 +0000)]
fix [#30715] bmesh: select linked not ignoring hidden verts/edges/faces
add optional flag to ignore hidden elements.
also remove loop mask flag - since it wasnt used and vert/edge/face is enough.
Nicholas Bishop [Thu, 29 Mar 2012 11:31:44 +0000 (11:31 +0000)]
More array modifier fixes.
* Skip calculation of merge indices if merging isn't enabled
* Clean up usage of BMesh operators to fix small memory leak
* Fix harmless BLI_assert in CustomData_bmesh_merge
* Another null-initialization fix in CustomData_bmesh_merge
Nicholas Bishop [Thu, 29 Mar 2012 10:49:17 +0000 (10:49 +0000)]
Fix uninitialized variable in CustomData_bmesh_merge.
Caused a crash in array modifier with caps.
Campbell Barton [Thu, 29 Mar 2012 03:06:42 +0000 (03:06 +0000)]
converting a mesh to an undo mesh now skips using bmesh operator.
Mitchell Stokes [Thu, 29 Mar 2012 01:56:08 +0000 (01:56 +0000)]
BGE: This fixes a problem where lamps in inactive layers were rendering shadows buffers that never got used, which was a huge performance sink.
Campbell Barton [Thu, 29 Mar 2012 01:41:56 +0000 (01:41 +0000)]
skip using bmesh operators for converting to/from undo meshes (gives some speedup)
Mitchell Stokes [Thu, 29 Mar 2012 01:40:19 +0000 (01:40 +0000)]
Windows Blenderplayer: The console now closes when the Blenderplayer launches (like it does with Blender). To get it back, use the -c flag, which is finally working as described: "-c: keep console window open"
Campbell Barton [Wed, 28 Mar 2012 23:53:27 +0000 (23:53 +0000)]
fix [#30402] modules dir missing from sys.path
Campbell Barton [Wed, 28 Mar 2012 22:43:26 +0000 (22:43 +0000)]
fix [#30705] Playing animation in reverse prevents further forward playback
Campbell Barton [Wed, 28 Mar 2012 22:11:19 +0000 (22:11 +0000)]
fix invalid *= on an un-initialized value when adding primitives.
Nicholas Bishop [Wed, 28 Mar 2012 22:03:46 +0000 (22:03 +0000)]
Partial fix for array modifier vertex merge (bug 30195).
Two fixes:
* The find-doubles operator was overriding the "dirty" element
indices, so vertices were not being mapped correctly.
* In some cases a vertex can be set to merge with a vertex that is
also set to be merged, so added a loop to find the first unmerged
vert in this case.
Not yet working is the "First Last" merge option.
Also made some cleanups and added comments in the array/remove doubles
code.
Mitchell Stokes [Wed, 28 Mar 2012 19:56:11 +0000 (19:56 +0000)]
This does two things for cleanup builds made with SCons:
1) Removes unwanted folders from Python/lib instead of just emptying them out.
2) Removes the test folder from Python/lib. This folder contains unit tests for Python (not the unittest module itself), and is about 10~12MB.
Brecht Van Lommel [Wed, 28 Mar 2012 12:18:25 +0000 (12:18 +0000)]
Fix #30702: game engine softbody weld threshold was exposed in UI, but actually
disabled in code because it doesn't work well, so hide the property from the UI.
Brecht Van Lommel [Wed, 28 Mar 2012 12:18:12 +0000 (12:18 +0000)]
Fix #30551: cycles passes combining did not always give identical result combined
with antialiasing/defocus, now divide out color at the very end instead of for each
sample.
Campbell Barton [Wed, 28 Mar 2012 11:53:18 +0000 (11:53 +0000)]
style cleanup
Nicholas Bishop [Wed, 28 Mar 2012 11:48:16 +0000 (11:48 +0000)]
Small cleanup for r45238, remove empty/unused function.
Brecht Van Lommel [Wed, 28 Mar 2012 10:39:21 +0000 (10:39 +0000)]
Cycles: shadow pass support. Note that this only takes into account lamps,
emitting objects or world lighting do not contribute to the shadow pass.
Consider this more as a pass useful for some compositing tricks, unlike
other lighting passes this pass can't be used to exactly reconstruct the
combined pass.
Campbell Barton [Wed, 28 Mar 2012 09:10:19 +0000 (09:10 +0000)]
expose image render_slot so keybindings can be configured to change
Brecht Van Lommel [Wed, 28 Mar 2012 09:07:43 +0000 (09:07 +0000)]
Cycles: add alpha output to image node.
Brecht Van Lommel [Wed, 28 Mar 2012 09:07:35 +0000 (09:07 +0000)]
Cycles: add rendered draw mode option in 3d view header to show the active
render layer rather than the viewport layers.
Brecht Van Lommel [Wed, 28 Mar 2012 09:07:10 +0000 (09:07 +0000)]
Cycles: viewport rendered draw mode now shows background images, also changed the
image editor checkerboard pattern to be the same as cycles viewport.
Campbell Barton [Wed, 28 Mar 2012 08:42:19 +0000 (08:42 +0000)]
bmesh: be more strict with operator string formatting, no tabs and only accept slot=%x style.
Campbell Barton [Wed, 28 Mar 2012 08:00:58 +0000 (08:00 +0000)]
print error if rip does nothing (rather then grabbing the unripped verts as it did before)
Lukas Toenne [Wed, 28 Mar 2012 07:48:08 +0000 (07:48 +0000)]
Fix for the node tree nodes.new API function, this wasn't setting the scene and main context pointers in the node template, used by file output node. Also the file output node itself now works in case of scene==NULL (might happen in some contexts).
Campbell Barton [Wed, 28 Mar 2012 07:44:20 +0000 (07:44 +0000)]
remove dissolve keybinding, conflicts with draw, add back to delete menu.
dissolve is now 'X -> D'
Campbell Barton [Wed, 28 Mar 2012 06:44:19 +0000 (06:44 +0000)]
fix [#30699] blender crashes when dissolving faces
thanks to Francisco De La Cruz for investigating and providing fix.
Campbell Barton [Wed, 28 Mar 2012 06:34:18 +0000 (06:34 +0000)]
code cleanup: Ctrl+Click extrude - replace inline axis/angle to matrix conversion with a call to axis_angle_to_mat3()
Campbell Barton [Wed, 28 Mar 2012 05:44:38 +0000 (05:44 +0000)]
fix for vertex rip
wire edges connected to the vertex would confuse vertex rip.
- was only checking verts first edge had a face, this is incorrect since other vert-edges may have a face.
- wasn't checking of edges connected to the vertex were hidden.
Campbell Barton [Wed, 28 Mar 2012 05:20:48 +0000 (05:20 +0000)]
fix [#30701] bmesh: segfault when ripping partially connected vertices
Campbell Barton [Wed, 28 Mar 2012 05:09:50 +0000 (05:09 +0000)]
style cleanup: for creator.c & more useful assert message (file:line)
Campbell Barton [Wed, 28 Mar 2012 05:03:24 +0000 (05:03 +0000)]
style cleanup
Campbell Barton [Wed, 28 Mar 2012 03:47:33 +0000 (03:47 +0000)]
style cleanup: mainly whitespace around operators and indentation - paint/sculpt tools
Campbell Barton [Wed, 28 Mar 2012 03:17:22 +0000 (03:17 +0000)]
own error r45225
- move defines to the top of the func.
- rename T -> EPS
- undefine at the end.
Campbell Barton [Wed, 28 Mar 2012 01:06:43 +0000 (01:06 +0000)]
code cleanup: (dont include ';' in defines), last commit also missed changes to paint_image.c
Campbell Barton [Wed, 28 Mar 2012 00:42:38 +0000 (00:42 +0000)]
fix [#30666] Texturepaint "Soften" brush gets stuck
use derived mesh tessfaces since the mesh doesnt ensure this, also means non-project-painting in the viewport works with modifiers now.
Campbell Barton [Tue, 27 Mar 2012 23:23:47 +0000 (23:23 +0000)]
fix [#30685] Solidify tool behaves like Shrink/Fatten (editmesh functionality stuff)
solidify had a limit imposed on how far vertices could move which didnt work well, removed (the modifier doesnt do this).
also missed freeing 2 arrays.
Bastien Montagne [Tue, 27 Mar 2012 20:45:06 +0000 (20:45 +0000)]
Fix [#30700] Blend from Shape gives only selection for basis shapekey
RNA_enum_item_add already sets totitem, so it was increased twice…
Brecht Van Lommel [Tue, 27 Mar 2012 18:35:49 +0000 (18:35 +0000)]
Fix #30658: add new uv layer in object mode did not copy coordinates from
existing layer as it does in edit mode.
Brecht Van Lommel [Tue, 27 Mar 2012 18:26:41 +0000 (18:26 +0000)]
Fix tweak for #30638: use same BMesh Face area computation functions as Mesh,
it's a bit fuzzy how to compute area for ngons, but at least should be consistent.
Brecht Van Lommel [Tue, 27 Mar 2012 17:58:50 +0000 (17:58 +0000)]
Fix related to #30674: remove unnecessary global undo push during image paint,
not needed since it already has local image undo stack.
Thomas Dinges [Tue, 27 Mar 2012 17:53:31 +0000 (17:53 +0000)]
Patch [#30683] Editmesh_tools without descriptions by Harley Acheson (harley)
* Some Edit Mesh operators were missing the description after b-mesh merge.
Brecht Van Lommel [Tue, 27 Mar 2012 17:19:13 +0000 (17:19 +0000)]
Fix related to #30674: 2D image painting would do unnecessary full GL texture
delete at the end of the stroke, even when doing partial updates.
Bastien Montagne [Tue, 27 Mar 2012 13:52:30 +0000 (13:52 +0000)]
BMesh TODOs: use loops/polys instead of tessfaces.
Nicholas Bishop [Tue, 27 Mar 2012 13:08:40 +0000 (13:08 +0000)]
Fix bug 30697, strange shading on array modifier with offset object.
* An offset object can rotate output, so need to update normals to
reflect that. Fix by adding a normals recalc, but only if there's an
offset object.
* Added BMESH_TODO comment to check whether there are other cases were
normals need to be updated.
Nicholas Bishop [Tue, 27 Mar 2012 12:34:00 +0000 (12:34 +0000)]
Partial fix for bug 30695, "Array broke crease, weird visibility and slowdown"
* Array modifier creates BMesh from DM; add missing CD_CREASE layer
for edge creases.
* With a modifier stack like mirror+subsurf+array, face normals were
wrong. Fix by removing CD_NORMAL layer from CCGDM output. Previously
the elements in this layer were simply copied, so they did not
reflect subdivision correctly.
* Minor style fixes in bmo_dupe.c.
Issues not yet addressed:
* Subsurf's optimal draw setting for hiding subdivision edges is not
respected by the array output.
* Slowdown issue; array modifier is much slower than in 2.62.
Nicholas Bishop [Tue, 27 Mar 2012 11:03:10 +0000 (11:03 +0000)]
Fix more truncated comments.
Nicholas Bishop [Tue, 27 Mar 2012 10:55:20 +0000 (10:55 +0000)]
Spelling fix, sentine -> sentinel.
'sentine' apparently means "A place for dregs and dirt; a sink; a sewer" :)
Campbell Barton [Tue, 27 Mar 2012 10:30:10 +0000 (10:30 +0000)]
bmesh py api:
added access to deform weights, access to weights acts like a python dict so you can do...
print(group in dvert)
dvert[group] = 0.5
print(dvert[group])
del dvert[group]
print(dvert.items())
Nicholas Bishop [Tue, 27 Mar 2012 10:28:06 +0000 (10:28 +0000)]
Fix bug 30696, circular mesh objects had wrong orientation.
Behavior now matches 2.62, circle, cone, and cylinder get first vertex
at 0 degrees.
Also fixed range for vertex property of add-cone and add-cylinder
operators, changed minumum from two to three.
Nicholas Bishop [Tue, 27 Mar 2012 09:24:00 +0000 (09:24 +0000)]
Fix bug 29123, subdivide smooth distorts shape with numcuts > 1.
Three fixes:
* When smoothing a coord, the original position should be based off
the unsmoothed subdivision (i.e. the vertex coord rather than the
smoothed position in the shapekey)
* The normal of new vertices created by edge split must be updated
(used the same interpolation as pre-bmesh here)
* The vertices created inside the subdivided face should use the
subdivided coords from edges, so copy the shapekey coords back to
vertex coords after subdividing edges
Sergey Sharybin [Tue, 27 Mar 2012 09:10:05 +0000 (09:10 +0000)]
Port Setup Tracking scene to bmesh api
Sergey Sharybin [Tue, 27 Mar 2012 08:59:36 +0000 (08:59 +0000)]
MakeDupliFaces was using mesh.faces which was renamed to mesh.tessfaces.
TODO about porting it to polys is still relevant.
Sergey Sharybin [Tue, 27 Mar 2012 08:54:37 +0000 (08:54 +0000)]
reload_sequence_new_file now works with movie clip strips as well
Campbell Barton [Tue, 27 Mar 2012 05:03:23 +0000 (05:03 +0000)]
minor code/style cleanup
Campbell Barton [Tue, 27 Mar 2012 04:54:52 +0000 (04:54 +0000)]
fix for undoMesh_to_editbtMesh not setting BMEditMesh.ob field.
error found by XercesBlue on IRC
Campbell Barton [Tue, 27 Mar 2012 04:46:52 +0000 (04:46 +0000)]
style/name cleanup: have EDBM_* functions match our style guide and also match BM_ function naming conventions
Campbell Barton [Tue, 27 Mar 2012 01:52:49 +0000 (01:52 +0000)]
fix error in last commit
Campbell Barton [Tue, 27 Mar 2012 01:28:25 +0000 (01:28 +0000)]
bmesh todo, unlikly but possible - entering editmode with faces and no polys.
rather then printing a warning and failing - convert them to polys.
Campbell Barton [Tue, 27 Mar 2012 01:24:16 +0000 (01:24 +0000)]
style cleanup: wm, mosyly adding space around ops
Campbell Barton [Tue, 27 Mar 2012 00:17:57 +0000 (00:17 +0000)]
use cmake WITH_MOD_ * to exclude include paths which are not used.
Campbell Barton [Tue, 27 Mar 2012 00:01:35 +0000 (00:01 +0000)]
fix [#30651] bpy.ops.object.mode_set(...) editmode removes faces.
problem was that BMesh had tessellation call when undo pushes were called.
if python called an operator with no undo push, tessfaces would not be created.
fix this by making it the responsibility of each editmesh operator to re-tessellate, as it is with notifiers and depsgraph.
added EDBM_update_generic() function to add notifier, tag for depsgraph update and optionally re-tessellate.
Sergey Sharybin [Mon, 26 Mar 2012 22:36:32 +0000 (22:36 +0000)]
Return back sfra property to Sequence structure to prevent data loose
until things discussing in #30663 will be resolved.
Sergey Sharybin [Mon, 26 Mar 2012 22:26:30 +0000 (22:26 +0000)]
Clean-up logic of behavior of refresh/reload operators in sequencer
After discussion with Campbell we found much nicer solution which
keeps operation with data much more clear:
- Refresh Sequencer is totally harmless, do not touch actual data
and just removes everything from cache
- Reload Strip will reload data and adjust it's length for all
selected strips without affecting on length of strip itself
- Reload Strip and Adjust length will do the same but will also
adjust length of strip itself.
Campbell Barton [Mon, 26 Mar 2012 20:49:33 +0000 (20:49 +0000)]
quiet some warnings for gcc 4.7
Campbell Barton [Mon, 26 Mar 2012 20:41:54 +0000 (20:41 +0000)]
style cleanup: py api
Nicholas Bishop [Mon, 26 Mar 2012 19:50:45 +0000 (19:50 +0000)]
Fix for position of array-modifier end cap when not using a fixed count.
Moved calculation of final offset so that correct count value has been set.
Nicholas Bishop [Mon, 26 Mar 2012 19:44:25 +0000 (19:44 +0000)]
Fix run-time warning of freeing null pointer in array modifier.
Bastien Montagne [Mon, 26 Mar 2012 17:24:54 +0000 (17:24 +0000)]
Renaming pt_BR to pt, as a recent reform [1] made most country speaking portugese use a same, common version of it, which we want to use here.
[1] http://www.reformaortografica.com/
Nicholas Bishop [Mon, 26 Mar 2012 16:25:21 +0000 (16:25 +0000)]
Fix for truncation of 64-bit CustomDataMasks.
Can't use GET_INT_FROM_POINTER anymore with CD masks, as this
truncates to 32-bit.
Bug: http://projects.blender.org/tracker/index.php?func=detail&aid=30680&group_id=9&atid=498
CR: http://codereview.appspot.com/
5905059/
Sergey Sharybin [Mon, 26 Mar 2012 16:04:10 +0000 (16:04 +0000)]
Style cleanup of movie cache module
Sergey Sharybin [Mon, 26 Mar 2012 15:57:00 +0000 (15:57 +0000)]
Fix crash of Refresh Sequencer when some Movie / MovieClip strips doesn't have animation opened
Sergey Sharybin [Mon, 26 Mar 2012 15:29:27 +0000 (15:29 +0000)]
Fix #30675: Default .blend should have clay strips brush
Sergey Sharybin [Mon, 26 Mar 2012 14:30:30 +0000 (14:30 +0000)]
Default settings for new movie clip tracking now is the same as Default preset
Bastien Montagne [Mon, 26 Mar 2012 14:10:15 +0000 (14:10 +0000)]
Fix for segfault when building proxies in Sequencer.
Sergey Sharybin [Mon, 26 Mar 2012 13:48:45 +0000 (13:48 +0000)]
Request from Mango team which will most probably expected by everyone:
Refresh Sequencer button will now update lengths of strips as well.
Bastien Montagne [Mon, 26 Mar 2012 13:45:06 +0000 (13:45 +0000)]
Fix for own error in sequencer's Select menu (from r40658, makes me wonder if anyone uses that feature except me...), and update to Select Grouped op (taking into account new MovieClip strip type).
Sergey Sharybin [Mon, 26 Mar 2012 13:36:22 +0000 (13:36 +0000)]
- MovieClip strips now have got valid color for older saved files
- Update Strip Length will now properly update lenght of movie clip
strips created from movies before lenght was added there.
Sergey Sharybin [Mon, 26 Mar 2012 13:30:53 +0000 (13:30 +0000)]
Fix for crash when adding image sequence movie clip into a sequencer
Brecht Van Lommel [Mon, 26 Mar 2012 13:21:43 +0000 (13:21 +0000)]
Cycles: fix for ColorRamp node alpha output.
Brecht Van Lommel [Mon, 26 Mar 2012 13:21:13 +0000 (13:21 +0000)]
Fix #30676: UV unwrap did not create UV map with proper default coordinates.
Brecht Van Lommel [Mon, 26 Mar 2012 12:45:28 +0000 (12:45 +0000)]
Fix blenderplayer build, adding stub.
Brecht Van Lommel [Mon, 26 Mar 2012 12:45:14 +0000 (12:45 +0000)]
Cycles: add ColorRamp node.
Bastien Montagne [Mon, 26 Mar 2012 12:43:48 +0000 (12:43 +0000)]
Fix [#30684] Bmesh AddTorus.execute failure.
Done BMESH_TODO (use polys/loops instead of tessfaces).
Sergey Sharybin [Mon, 26 Mar 2012 12:20:46 +0000 (12:20 +0000)]
Fix #30647: Error in wire drawing of subsurf
Patch by Nicholas Bishop, thanks!
Campbell Barton [Mon, 26 Mar 2012 12:02:41 +0000 (12:02 +0000)]
fix [#30257] bmesh: Rip "V" don't work on end vertex
added option to edgesplit bmesh operator to take tagged vertices as well so an edge at a boundary can split without splitting off the boundary vertex.
the behavior/speed of the edge split modifier and tool remainss the same, this is only used for rip.
Sergey Sharybin [Mon, 26 Mar 2012 11:44:03 +0000 (11:44 +0000)]
Fix #30430: Can only alt+scroll image buffer slot once
Issue was caused by mapping old buttons to new buttons. Render slot button in header
holds RenerResult for particular slot, but in N-panel it holds RenderResult from RenderEngine.
So what was happening is: switching render slot to empty slot makes slot button in header
contain NULL as button function's argument, but old button holds RenderResult for rendered
image, so this two buttons aren't equal and so button isn't getting activated and no scrolling
happens.
Making slot button hold RenderResult directly from RenderEngine (as it's happening with
buttons in N-panel) makes old->new buttons mapping work correct and it's possible
to alt-scroll smoothly
Sergey Sharybin [Mon, 26 Mar 2012 10:33:37 +0000 (10:33 +0000)]
Fix #30671: Reset to default values bug
Campbell Barton [Mon, 26 Mar 2012 09:38:38 +0000 (09:38 +0000)]
fix [#30598] Bad 3D view display & render with smooth mesh
Sergey's suggestion - just call calc normals fixes.
Sergey Sharybin [Mon, 26 Mar 2012 09:36:19 +0000 (09:36 +0000)]
Fix for object solver possible pointing to a freed memory when removing
movie clip data block from python/
Campbell Barton [Mon, 26 Mar 2012 08:58:17 +0000 (08:58 +0000)]
style cleanup: use NULL rather than 0 for raytrace code.
Campbell Barton [Mon, 26 Mar 2012 08:10:12 +0000 (08:10 +0000)]
fix [#30653] Wrong image at UV/Image Editor window
render results would be displayed on loading new files if the scene names matches, now free render-results so as not to display stale data - also saves some memory.
Campbell Barton [Mon, 26 Mar 2012 08:06:13 +0000 (08:06 +0000)]
fix for possible crash - freed scene was being used for render result lookup when 'Load UI' was disabled.
Sergey Sharybin [Mon, 26 Mar 2012 07:51:47 +0000 (07:51 +0000)]
Setup tracking scene works again now
Campbell Barton [Mon, 26 Mar 2012 07:28:31 +0000 (07:28 +0000)]
fix for possible crash using freed ObCenter pointer when loaing a blend file without loading the UI.
Campbell Barton [Mon, 26 Mar 2012 06:55:09 +0000 (06:55 +0000)]
style cleanup: python api
Bastien Montagne [Mon, 26 Mar 2012 06:20:59 +0000 (06:20 +0000)]
Minor fixes to recent new UI messages.
Campbell Barton [Mon, 26 Mar 2012 05:28:00 +0000 (05:28 +0000)]
fix [#30658] existing uv mesh coordinates not copied to new created uv map
Campbell Barton [Mon, 26 Mar 2012 04:32:04 +0000 (04:32 +0000)]
disallow adding tessfaces to a mesh with polygons (only allowed case is creating a new mesh with tessfaces and later converting to polygons, which wont work if polygons exist)
Campbell Barton [Mon, 26 Mar 2012 04:23:18 +0000 (04:23 +0000)]
setting the active texture layer from python would get the UV layer out of sync. (entering editmode would show the wrong UV layer)