blender-staging.git
14 years agoProject Files (MSVC 6.0) update:
Martin Poirier [Tue, 19 Jul 2005 23:52:21 +0000 (23:52 +0000)]
Project Files (MSVC 6.0) update:
- drawarmature.c
- modifier.c, BKE_modifier.h
- DNA_modifier_types.h

14 years ago - added ModifierTypeInfo.updateDepgraph function, responsible for building
Daniel Dunbar [Tue, 19 Jul 2005 23:04:34 +0000 (23:04 +0000)]
 - added ModifierTypeInfo.updateDepgraph function, responsible for building
   appropriate relations in dep graph (modifiers respond to scene changes
   correctly now)
 - update modifier buttons to trigger depgraph rebuild if needed

14 years ago - fixed subsurf_make_derived_from_dlm
Daniel Dunbar [Tue, 19 Jul 2005 20:42:34 +0000 (20:42 +0000)]
 - fixed subsurf_make_derived_from_dlm

14 years ago - vertex indexing in softbody was wrong after I switch loop direction
Daniel Dunbar [Tue, 19 Jul 2005 20:29:22 +0000 (20:29 +0000)]
 - vertex indexing in softbody was wrong after I switch loop direction

14 years ago - split {curve,lattice,armature}_deform_verts out of mesh_deform
Daniel Dunbar [Tue, 19 Jul 2005 20:14:17 +0000 (20:14 +0000)]
 - split {curve,lattice,armature}_deform_verts out of mesh_deform
 - removed mesh_deform (merge into mesh_modifier)
 - switch python lattice_apply function to use object_apply_deform,
   this isn't exactly equivalent but the python system shouldn't
   have been calling that deep into the kernel anyway.

New feature: Modifier stack
 - added Object.modifiers (list of ModifierData elements)
 - added DNA_modifier_types.h
     o contains type definition for the file data for the various
       modifier types
 - added BKE_modifier.h
     o contains modifierType_get_info (access to modifier type registry)
     o structs and defines for runtime modifier usage
 - updated mesh_calc_modifiers to evaluate modifier stack (note that
   for the time being it also evaluates the old style modifiers so files
   should load and work as normal).
 - add file handling modifier code (todo: don't replicate on object copy)
 - add modifier stack UI code (lives in object panel)

Only real new feature at the moment is that you can apply lattices and
curves *after* a subdivision surface which was never possible before.

Todo:
 - DEP graph updating does not work correctly yet, so you generally have
   to tab cycle to see results.
 - editmode calculation does not use modifier stack.
 - bug fixes (there must be a few in there somewhere)

14 years agoFix: CTRL+Click adding bones didn't attach them to the parent!
Ton Roosendaal [Tue, 19 Jul 2005 19:25:17 +0000 (19:25 +0000)]
Fix: CTRL+Click adding bones didn't attach them to the parent!

14 years ago__ bug fix__
Joseph Gilbert [Tue, 19 Jul 2005 18:47:25 +0000 (18:47 +0000)]
__ bug fix__
- internal methods for mathutils class need to incref before returning a pointer to self
  * memory allocation errors (freeing bad pointers) were appearing after repeatedly calling internal methods on the same object due to ref counts.

14 years agoCouple of fixes combined;
Ton Roosendaal [Tue, 19 Jul 2005 16:12:56 +0000 (16:12 +0000)]
Couple of fixes combined;

- With Actions on a Pose, ESC in transform restored wrong.
  This is solved similar to ipos now, storing a 'last time evaluated'.
  Could be extended to ghosting... soon.
- Moving the little yellow 'key blocks' in Action window didn't update
  3d window. ALso the 'lock' option didn't work, and flashed header.
- Pose Transform: noticed there were still errors in cases, especially
  with actions. Painfully tried to build the desired matrix now.
- Removed obsolete Bone pointer from TransData

14 years agoFirst step towards restoring / improving subdivide 'beauty'.
Johnny Matthews [Tue, 19 Jul 2005 15:37:18 +0000 (15:37 +0000)]
First step towards restoring / improving subdivide 'beauty'.

Beauty button now is Beauty and Short.

Works as follows

Beauty on: If a face is selected, only subdivide the longest 2 sides
Beauty & Short on: If a face is selected, only subdivide the shortest 2 sides

1 problem atm is when more than 2 sides are equal. Must add code to check for this and disable beauty on that face. Use with caution! :)

Also restoring selection needs to be tweaked here.

Side Note: for most accurate subdividing, use edge mode and select only the edges you wish to cut rather than relying on beauty.

14 years ago-Patch submitted by Campbell Barton
Joseph Gilbert [Tue, 19 Jul 2005 14:17:07 +0000 (14:17 +0000)]
-Patch submitted by Campbell Barton
-Fixed matrix.invert() to return the correct inverse matrix

14 years agoGrms... removed too happily a part of depgraph yesterday. That fix was OK,
Ton Roosendaal [Tue, 19 Jul 2005 11:47:22 +0000 (11:47 +0000)]
Grms... removed too happily a part of depgraph yesterday. That fix was OK,
but doesn't make the removed code obsolete :)
Error was that transform on deformed objects didn't set recalc flags

14 years agoRemoved last remainder of old transform. The function is still called
Ton Roosendaal [Tue, 19 Jul 2005 11:21:01 +0000 (11:21 +0000)]
Removed last remainder of old transform. The function is still called
special_aftertrans_update() though...

14 years agoBugfix #2818
Ton Roosendaal [Tue, 19 Jul 2005 10:50:13 +0000 (10:50 +0000)]
Bugfix #2818

On importing a file (dxf for example), the active filename got set to the
imported filename, causing save-over to use that too. Tracked it down to
be a very old feature... better to remove this, doesn't work according
specs!

14 years ago1) Removed old annoyance in adding armature or bones.
Ton Roosendaal [Tue, 19 Jul 2005 10:06:57 +0000 (10:06 +0000)]
1) Removed old annoyance in adding armature or bones.

- Add Armature or Bone (SHIFT+A) now adds by default a single Bone, view
  aligned and of unit-size 1.
- Then use E-key (extrude) to draw chains, this option now doesn't popup
  the "OK" requester anymore, so works fast
- Another new method is using CTRL+click to add Bones

This makes it working identical to the other edit modes in Blender. The old
modal loop for drawing bones just wasn't working well either.

Related to this; have a real set of useful Armature primitives?

2) Removed event REMAKEALLIPO, which was added in NaN days for testing, and
proved to be extremely slow. Code is moved now to editnla.c.
Thanks Tom Hendrick to pointing to this nasty buggy feature!

14 years agoBug fix; pose mode transform didn't count # of selected bones correctly,
Ton Roosendaal [Tue, 19 Jul 2005 08:02:27 +0000 (08:02 +0000)]
Bug fix; pose mode transform didn't count # of selected bones correctly,
skipping selected bones with parent but without "IK to Parent"

14 years agoAdded DAG flush to NMesh when updating geometry (was needed after zr's latest commits).
Martin Poirier [Tue, 19 Jul 2005 05:35:35 +0000 (05:35 +0000)]
Added DAG flush to NMesh when updating geometry (was needed after zr's latest commits).
I added that in mesh_update, if anyone has a better idea, feel free to fix it.

14 years ago - some missing changes for switch of mesh_modifier to
Daniel Dunbar [Tue, 19 Jul 2005 04:27:43 +0000 (04:27 +0000)]
 - some missing changes for switch of mesh_modifier to
   deform new verts array.

I don't really know how to use softbody, so it would be nice
if someone would test this for me.

14 years agoinitialize simd quad in default constructor
Erwin Coumans [Tue, 19 Jul 2005 04:11:47 +0000 (04:11 +0000)]
initialize simd quad in default constructor
this should not be necessary, its for testing

14 years ago - split mesh_deform off from object_deform
Daniel Dunbar [Tue, 19 Jul 2005 02:36:21 +0000 (02:36 +0000)]
 - split mesh_deform off from object_deform
 - changed mesh_modifier, sbObjectStep, object_deform to take vertexCo
   argument instead of operating on mesh
 - fixed bug where a derived mesh would not be returned in editmode
 - removed object_wave, replaced by init_wave_deform and calc_wave_deform
 - moved cached DerivedMesh to Object, not Mesh... fixes heisenbugs
   with linked objects

14 years agoFix for bug #2848 This flushes selection upward in subdivide based on select mode...
Johnny Matthews [Tue, 19 Jul 2005 02:23:52 +0000 (02:23 +0000)]
Fix for bug #2848 This flushes selection upward in subdivide based on select mode. So faces with all edges selected should now be selected as well.

14 years ago - silly bug, didn't test for hash before freeing
Daniel Dunbar [Tue, 19 Jul 2005 01:12:14 +0000 (01:12 +0000)]
 - silly bug, didn't test for hash before freeing

14 years ago - change mesh_get_derived_render to mesh_create_derived_render (always
Daniel Dunbar [Tue, 19 Jul 2005 00:21:01 +0000 (00:21 +0000)]
 - change mesh_get_derived_render to mesh_create_derived_render (always
   builds new DerivedMesh... caching can come later)
 - split DerivedMesh returning functions into editmesh and mesh groups
 - got rid of DL_NORS displist type (get built on fly for mesh when
   needed)
 - got rid of Mesh.disp (yay!)
 - started to punch DerivedMesh returning functions into shape to introduce
   modifier stack

14 years agoBugfix: clear rotation on Pose bone didn't flush changes to DAG
Ton Roosendaal [Mon, 18 Jul 2005 22:18:03 +0000 (22:18 +0000)]
Bugfix: clear rotation on Pose bone didn't flush changes to DAG

Fix for previous commit today; found other test file with action
constraint that behaved different... found out the old action constraint
used the deform matrix (from restpos to pose) to define the key.

ALso removed unused function from outliner.c

14 years ago - ditch mesh_uses_displist
Daniel Dunbar [Mon, 18 Jul 2005 20:49:19 +0000 (20:49 +0000)]
 - ditch mesh_uses_displist

14 years ago - change convert subsurf to mesh to use mesh_create_derived_no_deform
Daniel Dunbar [Mon, 18 Jul 2005 20:48:16 +0000 (20:48 +0000)]
 - change convert subsurf to mesh to use mesh_create_derived_no_deform
   (converts any modifier now)

14 years ago - remove some extverts related cruft
Daniel Dunbar [Mon, 18 Jul 2005 20:34:28 +0000 (20:34 +0000)]
 - remove some extverts related cruft

14 years ago - miscapitalized ghash...
Daniel Dunbar [Mon, 18 Jul 2005 20:05:30 +0000 (20:05 +0000)]
 - miscapitalized ghash...

 Ok, I'll stop trying to bump up my commits per minute rate for now.

14 years ago - yet another warning fix
Daniel Dunbar [Mon, 18 Jul 2005 20:03:52 +0000 (20:03 +0000)]
 - yet another warning fix

14 years ago - another warning fix
Daniel Dunbar [Mon, 18 Jul 2005 20:00:41 +0000 (20:00 +0000)]
 - another warning fix

14 years ago - minor warning fix
Daniel Dunbar [Mon, 18 Jul 2005 19:59:51 +0000 (19:59 +0000)]
 - minor warning fix

14 years ago - added DerivedMesh.getVertCos function and implementations
Daniel Dunbar [Mon, 18 Jul 2005 19:58:23 +0000 (19:58 +0000)]
 - added DerivedMesh.getVertCos function and implementations
 - added mesh_create_derived_no_deform[_render]
 - mesh_create_orco now always goes through a DerivedMesh, some
   redundant copying atm but can be fixed (and orco generation is
   not a big bottleneck)

New feature: TexMesh (texcomesh) works with subsurf now (are
you listening rob?)

14 years agoTwo small fixes (armatures);
Ton Roosendaal [Mon, 18 Jul 2005 19:36:17 +0000 (19:36 +0000)]
Two small fixes (armatures);

- adding relations to pose-constraints should call DAG on the pose
- prevented constraints to break "IK" chains

14 years ago - need to zero hash pointer on free
Daniel Dunbar [Mon, 18 Jul 2005 19:29:42 +0000 (19:29 +0000)]
 - need to zero hash pointer on free

14 years ago - removed orco pointer from struct Mesh
Daniel Dunbar [Mon, 18 Jul 2005 18:28:16 +0000 (18:28 +0000)]
 - removed orco pointer from struct Mesh
 - switch renderer to store orco's in a hash table instead of
   caching in mesh (I don't like the renderer caching data
   in the actual mesh structure)
 - added mesh_create_orco[_render] function

14 years ago[ #2423 ] Object.getParentSubName
Joseph Gilbert [Mon, 18 Jul 2005 18:11:23 +0000 (18:11 +0000)]
[ #2423 ] Object.getParentSubName
- this is a submitted patch by lordbosh
- adds to ability to get the bone name when and object is parented to a bone

14 years ago - remove unnecessary editmode cycle on convert subsurf to mesh
Daniel Dunbar [Mon, 18 Jul 2005 17:45:30 +0000 (17:45 +0000)]
 - remove unnecessary editmode cycle on convert subsurf to mesh

14 years ago - added boundbox_set_from_min_max function
Daniel Dunbar [Mon, 18 Jul 2005 17:33:51 +0000 (17:33 +0000)]
 - added boundbox_set_from_min_max function
 - fix DerivedMesh.getMinMax implementations to set min & max when
   there are no vertices
 - mesh boundbox calc was wrong in some cases, messed up HOMEKEY
   and localview zooming

14 years agoMade action constraints to become backwards compatible. In the old version
Ton Roosendaal [Mon, 18 Jul 2005 17:27:17 +0000 (17:27 +0000)]
Made action constraints to become backwards compatible. In the old version
it used the 'global' (Pose Space) rotation to define the action timing.
That was of course hardly useful, reason for many weird patched rigs.
Nevertheless, sometimes it can be useful, and it's nice to see old files
nicely work still. So: "Local" is an option now.

Note: if you changed last week files because of the new action constraint,
set the new "Local" option in the object-buttons Constraint panel.

14 years agomore preparations for bullet physics, and some eol issues with SCA_Joystick*
Erwin Coumans [Mon, 18 Jul 2005 05:41:00 +0000 (05:41 +0000)]
more preparations for bullet physics, and some eol issues with SCA_Joystick*

14 years agoIncrease the warning level for the python vcproj
Joseph Gilbert [Mon, 18 Jul 2005 03:52:29 +0000 (03:52 +0000)]
Increase the warning level for the python vcproj

14 years agoHeader file clean up and warning fixes
Joseph Gilbert [Mon, 18 Jul 2005 03:50:37 +0000 (03:50 +0000)]
Header file clean up and warning fixes
- Mostly this cleans up the #includes and header files in the python project.
- Warning fixes are mostly casting issues and misc fixes. General warning clean up.
- #include Python.h MUST come as the first include to avoid the POSIX redefine warning in the unix makefiles
- fno-strict-aliasing flag added to makefile to fix a unavoidable type punning warning in types.c

14 years agoThird attempt to fix crash on leave editmode + delete loadsa bones! :)
Ton Roosendaal [Sun, 17 Jul 2005 22:15:12 +0000 (22:15 +0000)]
Third attempt to fix crash on leave editmode + delete loadsa bones! :)

Now certain pointers get cleared better, prevent reading freed data...

14 years ago - "Text" tfaces should only be drawn from deformed (not subdivided)
Daniel Dunbar [Sun, 17 Jul 2005 21:54:12 +0000 (21:54 +0000)]
 - "Text" tfaces should only be drawn from deformed (not subdivided)
   mesh

14 years ago - convert write_object_stl to use mesh_get_derived_final
Daniel Dunbar [Sun, 17 Jul 2005 21:48:23 +0000 (21:48 +0000)]
 - convert write_object_stl to use mesh_get_derived_final

14 years ago - convert draw_mesh_tfaces to use mesh_get_derived_final instead of
Daniel Dunbar [Sun, 17 Jul 2005 21:43:35 +0000 (21:43 +0000)]
 - convert draw_mesh_tfaces to use mesh_get_derived_final instead of
   calling mesh_uses_displist (becoming obsolete)

14 years ago - missed commit of file with mesh_calc_normals
Daniel Dunbar [Sun, 17 Jul 2005 21:24:43 +0000 (21:24 +0000)]
 - missed commit of file with mesh_calc_normals

14 years ago - commit patch from Matthew Plough to request buffer with SWAP_COPY
Daniel Dunbar [Sun, 17 Jul 2005 20:40:07 +0000 (20:40 +0000)]
 - commit patch from Matthew Plough to request buffer with SWAP_COPY
   behavior

14 years ago - replace displistmesh_calc_normals by mesh_calc_normals
Daniel Dunbar [Sun, 17 Jul 2005 20:12:16 +0000 (20:12 +0000)]
 - replace displistmesh_calc_normals by mesh_calc_normals
 - convert MeshDerivedMesh to calculate new vertex normals and such
   on initialize, means copy free conversion to DispListMesh
 - replace vertex access through function by direct access fo
   MeshDerivedMesh
 - shadeDispList was not getting correct orco's

14 years ago - need to release object DerivedMesh prior to release object data
Daniel Dunbar [Sun, 17 Jul 2005 19:15:06 +0000 (19:15 +0000)]
 - need to release object DerivedMesh prior to release object data

14 years agoMore fixes thanks to irc testing;
Ton Roosendaal [Sun, 17 Jul 2005 19:08:04 +0000 (19:08 +0000)]
More fixes thanks to irc testing;

- On exit editmode and after deleting bones, a pointer in PoseChannels
  was not set NULL correctly, causing potential crashes.
- Action Window: select on 'ipo key' (little icon) didn't select/deselect
  the associated bone (old bug!)
- After unlinking an Action, the blue color in bones was not cleared

Also:

- Improved selection in editmode, doing it in 2 stages. Makes it a lot more
  accurate.

14 years ago - switch shadeDispList to use DLM face normals if available
Daniel Dunbar [Sun, 17 Jul 2005 18:51:50 +0000 (18:51 +0000)]
 - switch shadeDispList to use DLM face normals if available

14 years ago - switch shadeDispList to always using DispListMesh from
Daniel Dunbar [Sun, 17 Jul 2005 18:37:27 +0000 (18:37 +0000)]
 - switch shadeDispList to always using DispListMesh from
   mesh_get_derived_final... a bit slower atm because of extraneous
   copying (will be fixed shortly) but normals are correct for deformed
   meshes now (don't think they ever have been)

14 years ago - assorted fixes for drawing "Text" tfaces
Daniel Dunbar [Sun, 17 Jul 2005 18:31:39 +0000 (18:31 +0000)]
 - assorted fixes for drawing "Text" tfaces

14 years ago - remove extverts argument from vertexnormals_mesh
Daniel Dunbar [Sun, 17 Jul 2005 18:22:23 +0000 (18:22 +0000)]
 - remove extverts argument from vertexnormals_mesh
 - changed new_NMesh_internal to not take extverts argument

14 years ago - removed unused DL_TRIA type
Daniel Dunbar [Sun, 17 Jul 2005 17:43:15 +0000 (17:43 +0000)]
 - removed unused DL_TRIA type

14 years ago - convert mesh_modifier to return deformed verts instead of
Daniel Dunbar [Sun, 17 Jul 2005 17:41:03 +0000 (17:41 +0000)]
 - convert mesh_modifier to return deformed verts instead of
   leaving in a DL_VERTS type displist (and modifying mesh)
 - removed DL_VERTS displist type (woot woot)
 - makeDispListMesh now puts deformed verts in object->derivedDeform
 - switch over other system parts to new deformed vert storage,
   still kinda hacky and maybe some inconsistencies... will be
   sorted out soon enough.
 - moved build_particle_system to makeDispListMesh... this may have
   adverse side effects, needs to be sorted out with depgraph system

14 years agoThree fixes, thanks to irc reporting of Basse & Sten;
Ton Roosendaal [Sun, 17 Jul 2005 17:36:54 +0000 (17:36 +0000)]
Three fixes, thanks to irc reporting of Basse & Sten;

- Bone points for B-Bone drawing were not adjusted for size of Bone, now
  they draw same size as for other bone type
- Deleting bones in editmode didn't check for constraints in Pose, causing
  weird collapsed poses or even crashes
- ALT+S scaling on B-Bones in editmode didn't correct for the Object scale.

14 years ago - made particles use displistmesh not DerivedMesh interface
Daniel Dunbar [Sun, 17 Jul 2005 17:33:09 +0000 (17:33 +0000)]
 - made particles use displistmesh not DerivedMesh interface

14 years ago - back out ton's vertex_duplilist "fix"
Daniel Dunbar [Sun, 17 Jul 2005 16:16:22 +0000 (16:16 +0000)]
 - back out ton's vertex_duplilist "fix"

14 years agoWire selecting in Armature editmode accidentally broke. :)
Ton Roosendaal [Sun, 17 Jul 2005 16:04:53 +0000 (16:04 +0000)]
Wire selecting in Armature editmode accidentally broke. :)

14 years ago - add extern to selboxes variables declaration
Daniel Dunbar [Sun, 17 Jul 2005 16:02:10 +0000 (16:02 +0000)]
 - add extern to selboxes variables declaration

14 years agoFix: outliner crash on deleting bones in editmode
Ton Roosendaal [Sun, 17 Jul 2005 15:45:35 +0000 (15:45 +0000)]
Fix: outliner crash on deleting bones in editmode

Further cleaned up editmesh code to be warning free and to have a little
nicer style :)

14 years agoWhile checking old bugreports on the todo tracker, found weak issue in
Ton Roosendaal [Sun, 17 Jul 2005 12:20:08 +0000 (12:20 +0000)]
While checking old bugreports on the todo tracker, found weak issue in
current depgraph... that's when an object has multiple "parents", which
have different relation types. (one parent moves Object, the other
deforms). In that case the result was undefined.

Solved it nicely with the DAG building code, resulting in quicker flushing
of changes even.
Also noted a bug in the drawview.c 'object_handle_update' call, which
should not check for layers. The DAG flushing clears recalc flags for
invisible layers, if possible. Now need to write that @#!# doc!

14 years agoprovided proper license for lcp solver
Erwin Coumans [Sun, 17 Jul 2005 10:07:46 +0000 (10:07 +0000)]
provided proper license for lcp solver

14 years agoBugfix #2836
Ton Roosendaal [Sun, 17 Jul 2005 09:49:02 +0000 (09:49 +0000)]
Bugfix #2836

Crash in Vertex-duplicator when there's an empty Mesh. Simply added check
for dm pointer. :)

14 years agoNew: B-Bones now support adjusting the 'handle length', with the values
Ton Roosendaal [Sun, 17 Jul 2005 08:58:42 +0000 (08:58 +0000)]
New: B-Bones now support adjusting the 'handle length', with the values
"In" and "Out" in the new posemode Panel, editing buttons.

http://www.blender.org/cms/Armature_draw_modes.629.0.html

Fix in image.c: just an unitialized variable.

14 years ago- The "in" and "out" interpolation values for B-Bones were not copied to
Ton Roosendaal [Sun, 17 Jul 2005 08:56:51 +0000 (08:56 +0000)]
- The "in" and "out" interpolation values for B-Bones were not copied to
and from editmode Armature

- Altered 'axismat' for PoseMode bone transform. Fix for bug #2769, bone
  rotations using hotkey-constraints (R, Y, Y) didn't work well.

14 years ago - add DerivedMesh *derivedDeform to struct Object, this is temporary
Daniel Dunbar [Sun, 17 Jul 2005 05:34:35 +0000 (05:34 +0000)]
 - add DerivedMesh *derivedDeform to struct Object, this is temporary
   until fullblown modifier system comes in, need it now to move forward.

14 years ago - convert vertex_duplilist to use mesh_get_derived_deform
Daniel Dunbar [Sun, 17 Jul 2005 05:20:57 +0000 (05:20 +0000)]
 - convert vertex_duplilist to use mesh_get_derived_deform
 - convert object_apply_deform to use mesh_get_derived_deform
 - convert build_particle_system to use mesh_get_derived_deform
 - remove apply option from mesh_modifier

What do you know, this leaves only one callsite for mesh_modifier
and one "user" site for DL_VERTS...

14 years ago - added DerivedMesh.getMinMax function (calls DO_MINMAX on all vertices)
Daniel Dunbar [Sun, 17 Jul 2005 04:17:33 +0000 (04:17 +0000)]
 - added DerivedMesh.getMinMax function (calls DO_MINMAX on all vertices)
 - static'd boundbox_displist
 - added getMinMax implementations for all existing DerivedMesh
   implementations (mesh, editmesh, displistmesh, and ccgsubsurf)

14 years ago - converted draw_tfaces3d and bbs_mesh_solid to use
Daniel Dunbar [Sun, 17 Jul 2005 03:48:29 +0000 (03:48 +0000)]
 - converted draw_tfaces3d and bbs_mesh_solid to use
   mesh_get_derived_deform... (whats that? more general AND it uses less
   lines? how can that be!!!)

14 years ago - convert give_parvert to use mesh_get_derived_deform
Daniel Dunbar [Sun, 17 Jul 2005 01:29:12 +0000 (01:29 +0000)]
 - convert give_parvert to use mesh_get_derived_deform

14 years ago - converted write_videoscape_mesh to use mesh_get_derived_deform
Daniel Dunbar [Sun, 17 Jul 2005 01:23:43 +0000 (01:23 +0000)]
 - converted write_videoscape_mesh to use mesh_get_derived_deform

14 years ago - added DirectMesh.getVert{Co,No} functions
Daniel Dunbar [Sun, 17 Jul 2005 01:18:59 +0000 (01:18 +0000)]
 - added DirectMesh.getVert{Co,No} functions
 - added mesh_get_derived_deform function (always returns a DerivedMesh
   corresponding to deformed (but not subdivided) mesh). used in places
   where original mesh is to be displayed but with deformed coordinates
   (vpaint for example).
 - added DirectMesh.getVert{Co,No} implementations for MeshDerivedMesh
 - updated vpaint to use mesh_get_derived_deform

14 years agoanother file behind the #USE_BULLET
Erwin Coumans [Sat, 16 Jul 2005 22:13:20 +0000 (22:13 +0000)]
another file behind the #USE_BULLET

14 years agoReverted to good line endings
Alexander Ewering [Sat, 16 Jul 2005 22:02:59 +0000 (22:02 +0000)]
Reverted to good line endings

14 years agopreparation for bullet physics
Erwin Coumans [Sat, 16 Jul 2005 21:47:54 +0000 (21:47 +0000)]
preparation for bullet physics

14 years ago - switch M_NMesh_GetRawFromObject to always get mesh data from
Daniel Dunbar [Sat, 16 Jul 2005 21:20:44 +0000 (21:20 +0000)]
 - switch M_NMesh_GetRawFromObject to always get mesh data from
   DerivedMesh (needs testing)
 - added needsFree argument to mesh_get_derived_final

14 years ago - added dontFreeNors flag to DispListMesh as well
Daniel Dunbar [Sat, 16 Jul 2005 21:16:05 +0000 (21:16 +0000)]
 - added dontFreeNors flag to DispListMesh as well
 - changed mesh_get_derived_render to always return a DerivedMesh (even if
   no subsurf)
 - changed init_render_mesh to always get the mesh data through a
   DerivedMesh

14 years ago - added dontFreeVerts and dontFreeOther flags to displistmesh for
Daniel Dunbar [Sat, 16 Jul 2005 21:03:28 +0000 (21:03 +0000)]
 - added dontFreeVerts and dontFreeOther flags to displistmesh for
   situations where data can be shared easily.
 - added convertDisplistToMesh function for regular mesh DerivedMesh
   interface (how many times can *you* use mesh in one sentence?)
 - do_puno was uninitialized in init_render_mesh
 - added mesh_get_derived_final (temporary), difference from
   mesh_get_derived is it always returns a derived mesh, even if
   no subsurf.

14 years ago - remove call to mesh_modifier from init_render_mesh, shouldn't
Daniel Dunbar [Sat, 16 Jul 2005 20:42:20 +0000 (20:42 +0000)]
 - remove call to mesh_modifier from init_render_mesh, shouldn't
   be needed anymore

14 years agoMore armature goodies;
Ton Roosendaal [Sat, 16 Jul 2005 19:07:02 +0000 (19:07 +0000)]
More armature goodies;

The B-Bones!
(where the B can be read as 'block' or 'bezier' or 'b-spline')

- set option on/off in edit buttons, armature panel
- scaling of B-bones only works in editmode, use ALT+S to make bones fatter
  or thinner. Also works for constrainted transform
- In pose mode, you now have a buttons panel with per-bone settings too
  Here you can find the "segments" button, which allows bones to
  interpolate between previous/next bones, including roll.
- Buttons to control interpolation ("In" and "Out" are disabled, doesn't
  work satisfying yet

NOTE: this doesn't give deform yet! Main purpose for now is to test if this
drawing method serves to animate/pose armatures well.
Still need to review proper interpolation methods... maybe bezier is too
limited.

14 years agoadded Bullet physics binding files
Erwin Coumans [Sat, 16 Jul 2005 10:15:31 +0000 (10:15 +0000)]
added Bullet physics binding files

14 years agoAdded Bullet library.
Erwin Coumans [Sat, 16 Jul 2005 09:58:01 +0000 (09:58 +0000)]
Added Bullet library.
Only windows projectfiles for now.
Will ask Hans to get unix makefiles done.

14 years agoAdded type for Bullet collision detection and physics engine.
Erwin Coumans [Sat, 16 Jul 2005 09:55:22 +0000 (09:55 +0000)]
Added type for Bullet collision detection and physics engine.
This will improve collision detection and physics for the game engine (Ketsji).
Bullet uses optionally uses the ODE quickstep solver.

Hope this commit doesn't break everything.

14 years agoFix radiosity "Collect Meshes" to ignore TFaces set to "Invisible".
Alexander Ewering [Sat, 16 Jul 2005 01:06:55 +0000 (01:06 +0000)]
Fix radiosity "Collect Meshes" to ignore TFaces set to "Invisible".

I hope this doesn't break anything - works fine here.

14 years ago - removed makeDispList, set_displist_onlyzero
Daniel Dunbar [Fri, 15 Jul 2005 20:56:55 +0000 (20:56 +0000)]
 - removed makeDispList, set_displist_onlyzero
 - appropriate callers of makeDispList replaced with depgraph calls
 - unappropriate places just killed... small chance this gives some
   errors in corner cases if dep graph isn't notified (example, font
   family displists) but these can be tracked down as they show up.
 - still a large number of callers of makeDispListCurveTypes, but
   makeDispListMesh has just a few.

14 years ago - remove calls to showkeypos from exit editmode functions, should be
Daniel Dunbar [Fri, 15 Jul 2005 18:15:56 +0000 (18:15 +0000)]
 - remove calls to showkeypos from exit editmode functions, should be
   handled by DAG update now. (ton: please confirm)

14 years agoPart #2 to vert color fix in new subdivide, Please check UV and Vertcolor with new...
Johnny Matthews [Fri, 15 Jul 2005 18:13:21 +0000 (18:13 +0000)]
Part #2 to vert color fix in new subdivide, Please check UV and Vertcolor with new subdivide.

Was checking for a condition < 0 should have been > 1

14 years ago - missed a space in header title text
Daniel Dunbar [Fri, 15 Jul 2005 17:57:40 +0000 (17:57 +0000)]
 - missed a space in header title text

14 years ago - decided it made more sense to make a key_get_active function, switched
Daniel Dunbar [Fri, 15 Jul 2005 17:55:19 +0000 (17:55 +0000)]
 - decided it made more sense to make a key_get_active function, switched
   to that in editmesh as well as for edit{curve,lattice}
 - added a G.editModeTitleExtra string that gets displayed in header info
   string in editmode. currently used to display "(Key)" when editing a
   key (before there was not UI level display of this info).

14 years ago - removed unused actkey field from struct Key
Daniel Dunbar [Fri, 15 Jul 2005 17:35:10 +0000 (17:35 +0000)]
 - removed unused actkey field from struct Key

14 years ago - more signedness warning fixes in editsima
Daniel Dunbar [Fri, 15 Jul 2005 17:31:58 +0000 (17:31 +0000)]
 - more signedness warning fixes in editsima
 - added mesh_get_active_key and replaced code in editmesh to use this
 - removed obsolete code in object_deform

14 years agoPart #1 of the fix for UV colors in new subdivide code
Johnny Matthews [Fri, 15 Jul 2005 16:12:34 +0000 (16:12 +0000)]
Part #1 of the fix for UV colors in new subdivide code

UV colors on Tri's should go right now when subdivivded,
for Quads, 3 of the for points should be right. Still working on the 4th.

14 years ago - remove some redundant drawing code
Daniel Dunbar [Fri, 15 Jul 2005 15:00:08 +0000 (15:00 +0000)]
 - remove some redundant drawing code
 - mousewheel zooming in drawipo used uninitialized var (MSVC run time
   error).

14 years ago - static'd several functions in key.c, just for readability
Daniel Dunbar [Fri, 15 Jul 2005 14:50:48 +0000 (14:50 +0000)]
 - static'd several functions in key.c, just for readability

14 years agoBPY Support for Deform Parenting and Vertex Parenting
Martin Poirier [Fri, 15 Jul 2005 05:30:58 +0000 (05:30 +0000)]
BPY Support for Deform Parenting and Vertex Parenting

This adds two new method to the BPY Object type:
 - makeParentDeform
 - makeParentVertex

Both are based on makeParent. In fact, I splitted off the parenting
part of makeParent into an internal function so that all can reuse it. I
also added DEP_graph refresh flag that where missing.

makeParentDeform sets the parent type accordingly for correct
deformation (armature and curve) and checks if the parent and childs are of the
correct type for deformation.

makeParentVertex takes as additional mandatory argument a tuple of
indices to tell the parent which vertex to use. It does some object type
checking too.

Also included documentation. :)

Test file is in the patch there: http://projects.blender.org/tracker/?func=detail&atid=127&aid=2833&group_id=9

Note to Ton: When doing vertex parenting, the dashed parent line is only updated correctly if the objects are on screen when calling the function. If they aren't, the line is wrong and is only corrected when another recalc is called on the parent or child.

14 years ago - more warning killing: unused vars, double -> float constants, signedness
Daniel Dunbar [Thu, 14 Jul 2005 22:21:12 +0000 (22:21 +0000)]
 - more warning killing: unused vars, double -> float constants, signedness

14 years ago - switch em_{vert,solid,wire}offs to be unsigned
Daniel Dunbar [Thu, 14 Jul 2005 22:16:38 +0000 (22:16 +0000)]
 - switch em_{vert,solid,wire}offs to be unsigned
 - some other twiddles to fix mixed signedness compiler warnings

And from the "Dear lazy programmers" file: After you have extern declared
a variable 5 or more times at the top a function, you would have actually
saved a lot of effort by just putting it in the header where it belonged.
Or perhaps you thought by hiding it no one else would notice you were
using global variables to pass information around. Tsk tsk.

14 years agoAppending armature crashed... nasty stuff, but works again :)
Ton Roosendaal [Thu, 14 Jul 2005 22:04:57 +0000 (22:04 +0000)]
Appending armature crashed... nasty stuff, but works again :)