Age | Commit message (Collapse) | Author |
|
|
|
|
|
This functions doesn't seem to be used even from a commented code.
|
|
Some lazy developer named Sergey typed zero instead of actual
mapping flag in unit mapping function in the original f-curve
normalization commit.
|
|
It turned out this was leading to accidental deleting in some cases when the
info message was missed by users. Fixes T37801.
|
|
|
|
This was broken in a recent bugfix for zero length hairs.
|
|
|
|
that gives an empty mesh.
|
|
correct.
|
|
in some cases.
The smoke modifier now ignores the modifier evaluation for generated texture
coordinates, which would previously cause the undeformed mesh to be cached for
flow objects. Dynamic paint has a similar exception, and other physics systems
avoid it by not being a constructive modifier.
|
|
Other error cases are already handled elsewhere
|
|
Added ui_handle_afterfunc_add_operator so a button can queue an operator
to run without executing it immediately.
|
|
The poly_order and mode properties were missing update and range, now they
match the UI code.
Reviewed By: brecht
Differential Revision: http://developer.blender.org/D116
|
|
* Support for symmetry in lasso masking
* Optimize away symmetry multiplication of gravity vector if no gravity
active
* Move flip_v3_v3 to paint_utils (used in masking as well)
* Use OpenMP for mask flood fill too.
|
|
|
|
|
|
This was actually a regression after color management re-implementation, need
to copy settings from saved image buffer to an original one since they might
be modified during save.
Also noticed image format planes detection didn't work properly from an image
buffer. Made it so save operator works fine now, but also marked a TODO in
BKE_imbuf_to_image_format() which needs to be investigated further.
|
|
when filepath is not writeable.
|
|
Simply check for non-embossed TEX buttons inside uiList items, and align their text like labels do.
|
|
sculpted object if last sculpt stroke ends outside the mesh.
Code here used the last true location which is invalid in that case,
just reuse the average stroke location instead.
|
|
Also fix previous commit original coordinate change for proxy calculation
(broke all non-anchored brushes -facepalm-) and cache the gravity support
result.
|
|
This must be an ancient bug, almost as old as anchored brushes
themselves. Code did not do sphere node intersection against original
bounding boxes if an anchored brush required it. In practice this was
not easy to see because vertices were displaced inside brush radius most
of the time. However in the provided report file, displacement was far
away from the brush sphere, making the issue apparent.
Also added original check in proxy combination code (might cause issues
with dyntopo but quick test did not show any) and did some style int ->
bool style changes.
|
|
Turns out we didn't properly free node proxies that gravity did use and
strokes ended up accessing previous data. Also deactivated gravity when
a smooth brush is used. It is kind of distracting to get the mesh
gravitated while smoothing.
|
|
|
|
|
|
constrained)
Currently we don't properly support snapping + axis-constraints, but
there are cases where its handy to project-snap for eg, and constraint
to an axis so re-enable this.
|
|
They're very verbose and while useful for debugging leaks - a unique
string is normally enough to track down the problem.
|
|
This is now always enabled, when you want to paint on a individual faces you
can use face selection masking instead.
Fixes T37855.
|
|
|
|
Summary:
Issue here most probably is that the start point in ray-casting is too
far away from the mesh. As a result the triangle intersection code can
sometimes miss the ray intersection. To solve this, we project the ray
segment to the boundary of the root node.
Reviewers: brecht, sergey, campbellbarton
Reviewed By: brecht
Maniphest Tasks: T37177
Differential Revision: http://developer.blender.org/D115
|
|
|
|
|
|
|
|
|
|
render.
|
|
upgrade to 1.4.
|
|
Python was indirectly causing redraw tags during drawing, which interfered
with the ARegion.drawrct, now ignore these during draw.
|
|
This was broken since BMesh.
|
|
|
|
It applies color management on a pixel in a way, based on number of
channels of this pixel.
Simplifies partial update code a bit.
|
|
|
|
It was only used by opengl render and in fact it needed just to
set DISPLAY_BUFFER_INVALID flag for the image buffer.
In theory it wouldn't make any change to opengl render speed
(because this change just moved rect_from_float from color
management code to image save code). And could not see any speed
changes on my laptop.
|
|
Was using a different font size with selection and drawing,
But scaling both so they mostly matched.
This made selection shift noticeably when done on an interface zoomed in/out.
|
|
|
|
|
|
The issue wasn't new in fact, just partial buffer update
function was still broken. Not sure how those issues were
not noticed for so long.
Perhaps partial_buffer_update_rect() could be simplified,
but i've got some more upcoming changes here which might
run into conflicts if i'll do simplification now.
|
|
|
|
Were still cases where numbers could be clipped incorrectly and icon
size was accounted for twice when left clipping (error in own recent commit).
|
|
|