December 8, 2013, 08:01 (GMT) |
Added new operator for point cache 'Export'. This is basically the same as baking, naming may have to be adjusted later, for now they just exist side-by-side. |
December 8, 2013, 08:01 (GMT) |
Exporter class to handle baking of point cache data over a frame range. This uses a mutex and cancel flag to make it work in the job system. |
December 8, 2013, 08:01 (GMT) |
Added boost includes and a thread utility file for mutex and locking. |
December 8, 2013, 08:01 (GMT) |
Incomplete dummy function for baking. |
December 8, 2013, 08:01 (GMT) |
Removed the writer_create/reader_create callbacks in PTCacheID, will focus on baking first and see about overhauling the API later. |
December 8, 2013, 08:01 (GMT) |
Removed unused Scene argument from reader/writer api functions. |
December 8, 2013, 08:01 (GMT) |
Use the path utility functions for generating the filename inside the new point cache, instead of using the blenkernel code. |
December 8, 2013, 08:01 (GMT) |
Utility functions for setting the cache path. Same procedure as in classic pointcache, can be improved later. |
December 8, 2013, 08:01 (GMT) |
Added reader constructor function for PTCacheID, but disabled reader/writer in DNA for now, this needs consideration. |
December 8, 2013, 08:01 (GMT) |
Basic reader class. |
December 8, 2013, 08:01 (GMT) |
More info in the persistent alembic storage. Now use Writer/Reader structs in PointCache instead of plain Archive. The writer is constructed on write calls currently, because the basic point cache creation function doesn't provide necessary parameters. This will get replaced eventually, for now just testing the alembic code. |
December 8, 2013, 08:01 (GMT) |
Some directory reorganization and a bunch of other stuff ... |
December 8, 2013, 08:01 (GMT) |
Use a dummy struct type PTCArchive instead of a plain void pointer. |
December 8, 2013, 08:01 (GMT) |
Create an OArchive internally for writing out point cache. |
December 8, 2013, 08:01 (GMT) |
Pass cache and object arguments to ptcache_path directly instead of the PTCacheID. This allows using the function for the alembic archive name as well, where we don't construct the cache id struct. Also added a few sanity checks to avoid potential NULL pointer access. |
December 8, 2013, 08:01 (GMT) |
Store an untyped archive pointer in the DNA PointCache struct. This is used to store an alembic archive which constitutes the point cache output. |
December 8, 2013, 08:01 (GMT) |
Optional HDF5 support. This is not directly linked by Alembic, so Blender has to link hdf5 itself if desired. |
December 8, 2013, 08:01 (GMT) |
Fix for Alembic find script to link all the necessary allembic libs. |
December 8, 2013, 08:01 (GMT) |
Use a #define to mark old point cache code in the RNA. Even thoug it will have to be kept enabled for now, this makes it easier to identify the old stuff and remove it later. |
December 8, 2013, 08:01 (GMT) |
Moved old point cache RNA from rna_object_force to rna_pointcache. This will make it easier to replace it with the new implementation while avoiding name conflicts. |
|
|
|


Master Commits
MiikaH:n Sivut a.k.a. MiikaHweb | 2003-2021