Blender Git Loki

In preparation for removing the point cache memory frame storage, set
all point caches to 'unbaked' to enfore rebaking.

Removed the PTC_test_archive function, was just some initial testing
code for alembic API.

Removed ptcache list code in readfile.c. This was intended to provide
some forward compatibility, but it's just not feasible to support this
nicely, so better accept breaking forward compat and don't touch this
DNA any more (keep code simple).

Removed ListBase code for multiple point caches from blenkernel and RNA.

Changed point cache writefile function to only write the point cache
struct, ptcaches lists are deprecated.

Removed the ptcache ListBase linking in readfile. Only the single
point_cache pointers are used now, which conveniently are already the
active point caches, so we can just ignore the lists (setting them to
the same pointer though to avoid potential problems with forward
compatibility).

Point cache index 1 in smoke domain is not used.

Revert "Comment on smoke caches is deprecated."
Thought this was referring to the ptcaches ListBase, but actually means
the point_cache array, so i still important.

This reverts commit 8f48294bd00b61bae93f375dc0a730cac7fb594e.

Adapted bpath point cache traversal macro to single PointCache pointers.

Comment on smoke caches is deprecated.

Made all point cache DNA members relating to cache lists DNA_DEPRECATED.
This includes the ListBase ptcaches members in all the users as well as
the next/prev pointers in the PointCache struct.

Moved the new point cache settings DNA data into the existing PointCache
struct. Adding new DNA data alongside the existing structs leads to
a lot of name conflicts, better try to stay within existing structs and
gradually extend/replace struct members.

Correct time sampling for Blender frames: The cycle time is 1.0 by
default, this needs to be 1/fps. The default frame sampling can be added
by writers, Alembic will only create this sampling once internally.

Take the global G.is_break flag into account for stopping the point
cache export job.

Disabled UNDO for the export operator, it doesn't make sense anyway.
Cache files are external outside the undo system and can not simply be
restored.

Set the G.is_rendering flag and lock editor drawing during export.
This is a terribly hack which is necessary to prevent draw methods
accessing the same data that is written by the export job.

Keep a reference to the stop, do_update and progress variables from
the export job in the exporter class, so the exporter can communicate
back to the main thread. This could be done nicer, but is ok for now.

Make sure the output directory for cache files exists before trying
to create the Alembic OArchive, otherwise HDF5 throws an exception.

New settings struct for poin caches (unused still).

Added a new DNA file DNA_pointcache_types.h for all point cache related
structs and enum values. DNA_object_force.h was use previously, this is
a badly named file with a hodgepodge of different areas.