Another crash with array caps, was caused by not making a deep enough
copy of CD field.
Also fixed the type of the 'mask' parameter, was int where it should
be 64-bit.
- remove unused vars
- no need to hard code version number for collada.
- cleanup some typos in comments.
- movieclip_calc_length was passing arg which should be unsigned to BLI_stringdec()
- memset(..., -1) is used incorrectly even though it worked: MOD_solidify.c - thanks Halley from IRC for noticing this. use fill_vn_i() instead.
- quiet warnings in editmesh_slide.c
- cleanup comments in bmesh and some other minor comment additions.
This report points out thet absolute shape keys are unusable.
The problem is there was no way to adjust the play time of a shape key (all absolte shape keys would start at frame zero with no way to change the speed).
Added an 'eval_time' property to the key block that works like the curve path evaluation time, so the time in the keyblock can be controlled.
When merging bmesh customdata, memset the new blocks to
zero. Otherwise can contain junk values (some layer types with
pointers like MDeformVert can then crash.)
When adding extra context data in a layout using uiLayoutSetContextPointer, this info was not inherited by popup menus generated in this layout. While operators from regular buttons work fine, the data is missing in operators from menus and such. This patch copies the bContextStore from buttons to the new uiLayout used for popups.
Do not recalculate curve's path if displist is building for orco -- in this
case modifiers are not applying on curve which makes path be calculated and
later used by dependent objects with it's non-modified state.
When ORIGSPACE is added, it is now a loop data, so we need to mark tessdata as dirty too. Also fixed DM_ensure_tessface to check DM_DIRTY_TESS_CDLAYERS flag!
This fixes problems with subsurf + dynapaint + particles, however, some modifiers, like remesh, seems to create a new dm from scratch, hence loosing completely those CD layers… Note this bug already existed in 2.62, so this is not a regression.
--debug
--debug-ffmpeg
--debug-python
--debug-events
--debug-wm
This makes debug output easier to read - event debug prints would flood output too much before.
For convenience:
--debug-all turns all debug flags on (works as --debug did before).
also removed some redundant whitespace in debug prints and prefix some prints with __func__ to give some context.
This fix adds a "ref_id" ID pointer to BKE_all_animdata_fix_paths_rename() & co, which is the ID against which prefix+oldName/NewName is "applied", currently only used for drivers' bones targets. Just pass NULL to get same behavior as previously. A bit annoying to make such a change for such a specific case, but there seems to be no other way to go... :/
- Deduplicate patetrn sampling used in esm and lmicklt trackers and
move SamplePattern to image/sample.h - Move computation of Pearson
product-moment correlation into own function in new file image/correlation.h
so all trackers can use it to check final correlation.
- Remove SAD tracker. It's almost the same as brute tracker, with only two differences:
1. It does brute search of affine transformation which in some cases helps to track
rotating features 2. It didn't use common tracker api which probably gave some
speed advantage, but lead to a real headache to use it together with other
trackers leading to duplicated code in blender side.
- Switch blenedr to use brute tracker instead of sad tracker which tracking made
source code much more simple to follow.
Issue was caused by object moved really far away (not just actually issue,
it's just about long mouse gesture and X-axis orientation which projects
position to quite large X-axis value) and for this location start offset
from curve length was calculating iteratively which takes plenty of time for
short curves.
Replace iterative search of offset with formula which seems to be working
in the same way and should be a bit more accurate.
* 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
* 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.
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.
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.
