There are a number of features that use a kind of "internal linking" in nodes:
1. muting
2. delete + reconnect (restore link to/from node after delete)
3. the new detach operator (same as 2, but don't delete the node)
The desired behavior in all cases is the same: find a sensible mapping of inputs-to-outputs of a node. In the case of muting these links are displayed in red on the node itself. For the other operators they are used to relink connections, such that one gets the best possible ongoing link between previous up- and downstream nodes.
Muting previously used a complicated callback system to ensure consistent behavior in the editor as well as execution in compositor, shader cpu/gpu and texture nodes. This has been greatly simplified by moving the muting step into the node tree localization functions. Any muted node is now bypassed using the generalized nodeInternalRelink function and then removed from the local tree. This way the internal execution system doesn't have to deal with muted nodes at all, as if they are non-existent.
The same function is also used by the delete_reconnect and the new links_detach operators (which work directly in the editor node tree). Detaching nodes is currently keymapped as a translation variant (macro operator): pressing ALTKEY + moving node first detaches and then continues with regular transform operator. The default key is ALT+DKEY though, instead ALT+GKEY, since the latter is already used for the ungroup operator.
* made bmesh_structure.h function names more consistant.
* remove unused code in bmesh_structure.c
* removed 'Edge Flip' operator (missing from bmesh but looked into trunk feature and dont think its worth keeping).
* tagged some BMESH_TODO's
* better type safety for BM_elem_flag_* and BM_elem_index_* functions.
* add BMElem type to be used where BMFace/Edge/Vert/Loop can be interchanged.
* fix bug in select manifold, found when making functions more strict.
* added BM_elem_flag_set, BMO_elem_flag_set. to avoid 'if(...) enable(); else disable();' all over the place.
* added bmesh_operator_api_inline.c, the header file was getting messy.
* New armature_mat_bone_to_pose, which is just the reverse of armature_mat_pose_to_bone (currently used by armature evaluation code only, but might be used by constraints space conversion code too, see note below).
* Found another place where another (a bit erroneous) pose_to_bone code existed (apply_targetless_ik(), in transform_conversion.c, used by "auto ik" option), replaced it by a call to armature_mat_pose_to_bone.
Notes:
* All those changes leave many #if 0 commented code, I will remove those in a few weeks.
* There is at least one other place where generic armature_mat_(pose_to_bone/bone_to_pose) functions should be usable: the space conversion function of constraints (constraint_mat_convertspace(), in BKE's constraint.c), but here I have some problems (old code is also somewhat erroneous, but using new one makes old rigs using bone space constraints wrong, e.g. sintel one, and it’s just impossible to make conversion code...). So I'll wait and investigate more for this one.
`_set` suffix was used in two ways (confusing)
* to set a flag to be enabled.
* to set a value passed as an argument.
now use enable/disable rather then set/clear for functions which change flags.
also remove BME_weld.c, the file didnt contain much code and the current extrude works well
These changes are to make the bmesh api more consistent and easier to learn, grouping similar functions which is convenient for autocomplete.
This uses similar convention to RNA.
* use face/loop/edge/vert as a prefix for functions.
* use 'elem' as a prefix too for functions that can take any type with a BMHeader.
* changed from camel case to underscore separated (like RNA).
Issue was caused by transformation tools weren't taking metastrip clips into
account, so strip inside of merastrips are placedon different from metastrip side
of time cursor, extension happened in unexcpected way.
Curve points of tracks curves now can be selected for X and Y channels
separately and can be moved along Y axis of curve viewer, points currently
can't change frame they belong to. This allows to smooth spikes caused by
unwanted marker jump.
Also fixed some mistakes in cancel transform in cases when locked tracks
were being trying to be moved; fixed incorrect calculation of marker speed
for curve drawing.
and length of the meta strip properly, only its contents. Also overlap check was
done too soon giving false positives, should be after meta strip update.
WARNING: This commits modifies how translated unconnected child bones with *no Inherit Rotation option* are positionned. This means that if you open a posed/animated armature using such (corner-case) setup, you'll have to adjust manually the locations of such bones: now, disabling Inherit Rotation/Scale will no more move the bone, only affecting its rotation/scale.
Many thanks to Bassam Kurdali (slikdigit) for his advices and tests of the patch!
-----
Dev notes : the pchan_to_pose_mat() func was added to BKE_armature.h, which computes two matrices to get the pose transformations (pchan) of the bone directly in pose (i.e. armature object) space. The first matrix is the rotation/scaling parts, the second one is for location.
That new function is used by (hence deduplicating and simplifying their code):
* The pose evaluation code (where_is_pose_bone()).
* The interactive transformation code (add_pose_transdata(), in transform_conversion.c).
* The snap to cursor/grid code (through armature_loc_pose_to_bone()/armature_mat_pose_to_bone()).
