scanfill remove-doubles pass assumes ordered edges (as with curves), otherwise it can hang.
workaround this problem by skipping removing-doubles for mesh ngons, since this isnt such a common case as it is with curves and we can just not support it.
* MEM_CacheLimitier - Size type to int conversion, should be safe for now (doing my best Bill Gates 640k impression)
* OpenNL CMakeLists.txt - MSVC and GCC have slightly different ways to remove definitions (DEBUG) without the compiler complaining
* BLI_math inlines - The include guard name and inline option macro name should be different. Suppressed warning about not exporting any symbols from inline math library
* BLI string / utf8 - Fixed some inconsistencies between declarations and definitions
* nodes - node_composite_util is apparently not used unless you enable the legacy compositor, so it should not be compiled in that case.
Leaving out changes to BLI_fileops for now, need to do more testing.
This commit makes BKE_image_acquire_ibuf referencing result, which means once
some area requested for image buffer, it'll be guaranteed this buffer wouldn't
be freed by image signal.
To de-reference buffer BKE_image_release_ibuf should now always be used.
To make referencing working correct we can not rely on result of
image_get_ibuf_threadsafe called outside from thread lock. This is so because
we need to guarantee getting image buffer from list of loaded buffers and it's
referencing happens atomic. Without lock here it is possible that between call
of image_get_ibuf_threadsafe and referencing the buffer IMA_SIGNAL_FREE would
be called. Image signal handling too is blocking now to prevent such a
situation.
Threads are locking by spinlock, which are faster than mutexes. There were some
slowdown reports in the past about render slowdown when using OSX on Xeon CPU.
It shouldn't happen with spin locks, but more tests on different hardware would
be really welcome. So far can not see speed regressions on own computers.
This commit also removes BKE_image_get_ibuf, because it was not so intuitive
when get_ibuf and acquire_ibuf should be used.
Thanks to Ton and Brecht for discussion/review :)
Originally issue was discovered when using stabilization and movie distortion
nodes, but in fact issue was caused by render layer node always doing nearest
interpolation. Now made it so this node will respect sampler passed to it's
executePixel function and do an interpolation.
Added two new functions to do bilinear/bicubic interpolation in float buffer
with variable number of components per element, so it could interpolate 1, 3
and 4 component vectors. This functions currently mostly duplicates the same
functions from imageprocess.c and it should actually be de-duplicated. Think
it's ok to leave a bit of time with such duplication, since functions should
be generalized one more time to support byte buffers, which could backfire on
readability.
Also removed mark as complex from stabilization node, which isn't needed sine
int fact this node is not complex.
pBVH nodes buffers could share some pointers from pBVH itself,
and after updating pointers in pBVH we need to recreate this buffers.
This starts looking a bit awkward and seems it's mainly because of
assumptions made when was working on sculpting on deformed mesh.
Would revision such a sculpting and would try to make it more clear
later.
Issue was caused by call ED_sculpt_force_update from WM_file_write which frees
derived mesh and pBVH and forces them to be re-created. After this sculpt session
wasn't repairing properly.
Freeing data from WM_file_write is not nice by itself, but it exposed possible
real issue when sculpting on modifiers. Made it so BLI_pbvh_grids_update will
update pointers to material flags and hidden grids as well.
This isn't idea solution for particular circumstances from the report, since
pBVH would be rebuilt after save which ends up with small interface lags for
while pBVH is rebuilding.
Would be nice to have multires displacement propagation to higher levels
without freeing data.
FLT_EPSILON was too small to use when checking if the edge collapse result is an error.
add invert_m3_m3_ex(), invert_m3_ex() functiosn which take an epsilon to check the determinant, saves calculating it twice per edge collapse.
When this new subtypes is used, then string of property is hidden using
asterisks, e.g.: mysecretpassword -> ****************
This code was reviewed and modified by Brecht. Thanks very much:
- https://codereview.appspot.com/6713044/
This new subtype of string property is intended mostly for Add-on developers
writing Add-on which communicates with some server (http, sql, ftp, verse,
etc.). When this server requires user authentication and user has to type
username and password, then current API didn't allow to type 'hidden' password,
e.g. when you want to demonstrate this script, then everybody can see this
security password. Some examples of Add-on which could use this new subtype:
- On-line database of textures
- Integration of render farm
- Integration of Verse
Security Notes:
- You can copy paste hiddent string of property from text input using (Ctrl-C, Ctrl-V),
but you can do this in other GUI toolkits too (this behavior it is widely used).
- Text of string property is stored in plain text, but it is widely used in other
GUI toolkits (Qt, Gtk, etc.).
Simple examples:
- https://dl.dropbox.com/u/369894/draw_op_passwd.py
- https://dl.dropbox.com/u/369894/blender-password.png
