Blender Git Loki

Merge branch 'master' into wiggly-widgets

Conflicts:
source/blender/blenloader/intern/versioning_270.c
source/blender/editors/include/ED_view3d.h

Combined storage of render and realtime data in caches.

Caches now create 2 new roots below the main 'top' object of Alembic:
root and root_render for realtime and render data respectively.
This makes it easy to switch the whole archive to either of the modes
during baking and for constructing dupli caches.

Alternatively individual objects could store hires versions of their
data. This would also be more efficient if the cache contains many
simple objects which don't have 2 different variants. However, such
design decisions are difficult to make at this point and the
implementation can be modifier later.

From review: b_rlay.passes.find_by_type(type, view)

This is working, but the code that requires this is not working yet. So it will not be committed together

Make sure the cache baking job cancels properly when interrupted in the
first stage.

Allow cache libraries to store both render and realtime (viewport) data.

This is the default now. It should make workflow a lot more foolproof
and convenient, since having only one of these modes active at a time
very easily leads to broken renders and confusing situations.

The problem is mostly due to the complicated way the depsgraph layer
feature is used to handle duplicator visibility. The duplicator is
declared as a child of its group's objects (even though no real
dependency exists!), so that a visible duplicator triggers updates of
invisible group objects, making instances of hidden groups possible.

However, dupli caches have to disable this dependency in order to avoid
unnecessary costly updates in hidden layers which are overridden by
cached data anyway. At the point where these dependencies are created
the evaluation context is unknown though, which means we cannot
distinguish between render and realtime evaluation for the purpose of
cache reading ...

Removed deprecated code.

Make sure mesh data is really stored and reconstructed from the cache
entirely.

Also the code is much better structured now, moving toward a more
standardized sample/schema class definition.

added a new vertex coordinate based constraint build method, as alternative to meshisland centroid based constraint building

rename bucket_hash -> bucket_index

Make ghash_copy do actual exact copy (same number of buckets).

Simplifies copy code (no need to re-compute compute hash of entries).

fix for error clearning a gset, was removing the gset flag

pass flag directly to internal ghash_new function

minor edits

rename ghash_buckets_shrink -> ghash_buckets_contract

keep (expand/contract), (grow/shrink) separate

Shrink: Early return in case we are not allowed to shrink.

remove unneeded ghash strink calls & dont force expand/shrink to be inline

Split expand/shrink into own functions

Revert "From review: Cycles: use scene.views instead of render_result.views"

This reverts commit 43e10a6853e4659429d000225b93862708414b31.

render_result.views is already taken into account whether the views are enabled,
if they can be rendered (e.g., Stereo mode shows only left and right) and when
multiview is off, it has a "" fallback view.

Leaving it as it was for the time being.

minor edits

WIP