The issue with wrapping is that it requires correct interpolation of the border pixels. Since interpolation is done at the far left end of the node tree in buffer/image/etc read operations, the wrapping
setting can not be used directly in those operations (otherwise in-line translate operations would cause conflicts). To make wrapping work correctly we need to add a buffer in front of the translate
operation, which can then be interpolated correctly based on wrapping. The WrapOperation becomes a variant of ReadBufferOperation, which uses its wrapping setting to determine the correct "extend" mode
for interpolation of the buffer.
This merges all mix operations into a COM_MixBaseOperation
(naming could be better, but this way it corresponds to what's
going on with math operations.
Same was done with RenderLayers operations.
Overall this gives 20% of bf_compositor library compilation
time decrease. And it was rather annoying to have tens of
files with just a single-line constructors anyway.
TODO:
- All Convert operations could also be merged into a single file,
but that would require adding some ConvertBaseOperation to
reduce code duplication (ideally). Leaving it unchanged for now.
- Some operations' headers seems to be wrongly including MixOperation
header, they need to include NodeOperation instead it seems.
This is rather harmless, but would be nice to doublecheck on this
eventually.
The limit for dragging a gesture on an area corner was smaller then the area hot-spot,
so you could click on the right-most side of the bottom-left corner, drag left - pass the gesture threshold and still be in the same area. so a motion intended as a join would register as a split.
Happened more with high DPI values. fix by ensuring the drag limit is always higher then the hotspot.
- rotating the view is faster.
- arrow keys work (was only wasd-rf before)
- when stationary wheel or +/- will set z direction (so mouse wheel always sets the fly axis back to z)
The channels define code in ANIM_init_channel_typeinfo_data() was not synchronized with the
definition of channel types (eAnim_ChannelType) in ED_anim_api.h.
give a result more similar to the Compatible falloff option. The scale is x2
though to keep the perceived scatter radius roughly the same while changing the
sharpness. Difference with compatible will be mainly on non-flat geometry.
Track margin checks needed some tweaks to deal better with the fact
that normalized values for the same pixel values might be different
across X and Y axis.
Also, non-centered patters are expected to be handling better now.
* 'Show Debug' now enabled for all newly created drivers. For most users, it is
useful to be able to see this to help figure out what's going on
* Removed failed experiment of creating new drivers with Generator FModifiers. I
had hoped that this would make it easier to create drivers that doubled or
halved the input values, but that has proved to not be the case, and instead
made harder for most users to set things up (as they'd have to remove these
first).
Now, when adding drivers from the UI, these get created with two keyframes (at
(0,0) and (1,1) for a 1-1 mapping), which can be easily tweaked normally.
However, for backwards compatability of scripts (notably rigify, and perhaps
some others out there), when creating drivers from scripts, they will still get
created with Generator FModifiers for now. We can review this situation again
for 2.7, but for now it seems ok.
of keyframe verts
Previously, every time you toggled the selection of all keyframes (using AKEY),
the active curve would get deselected and deactivated. However, this was a pain
when trying to tweak the shape of a particular curve, as doing this would cause
that curve to either fade into the background or into the jumble of other
curves.
It is now possible to use "Apply Scale" for Empties. While Empties don't exactly
have any Object data attached to them which can be used for supporting "true"
apply scale (i.e. with non-uniform scaling), they do have a drawsize value which
controls how large the empties are drawn (before scaling). This works by taking
the scale factor on the most-scaled axis, and combines this with the existing
empty drawsize to maintain the correct dimensions on that axis at least.
Core Assumptions:
1) Most scaled empties have uniform scaling anyways (i.e. most empties used for
bone shapes)
2) On balance, preserving non-uniform scaling of empties after apply scale is
not as important as not being able to do it at all
