Added provision for saved corrupt blend files - caused by a startup.blend
addressing > 16 GB space, which was read in 32 bits.
Now an invalid screen will get removed immediate after read. Might give
a memory-not-in-memlist print, but that's quite safe.
Allowing any-key modifiers to work is cool, but this needs more work.
My previous fix made overlapping key presses (where you release key 1
just after pressing key 2) to not register as valid shortcuts.
A previous bugfix disabled the dynamic paint modifier for orco texture
coordinate evaluation of the modifier stack. However the MOD_APPLY_USECACHE
flag is not a good way to check if the modifier is evaluated for orcos.
Instead I've added a MOD_APPLY_ORCO flag. Also removed a bunch of
applyModifierEM callbacks, none of them served a purpose except for the
subsurf modifier.
Root of the issue was fixed by Brecht in svn rev56441.
This change only prevents crash of files created in
blender before that fix, and also gives more useable
information about what's wrong (apparently, assert
here was doing nothing).
Was a regression in rev53509, whic hallowed manual edit of properties
which weren't supposed to be editing manually.
Added the same effects length update as in trnaslation code, so now
updating strip frames from py/interface will keep things consistent.
compressed files we were not passing the full wide char path to zlib, so not
all file names worked. Now we use gzopen_w available in new zlib versions.
Patch by Tamito Kajiyama, I added an extra check for the zlib version so it
keeps compiling with older versions for now.
For platform maintainers:
Part of this commit are zlib 1.2.8 libraries for windows 32 bit. We still
need update libraries for windows 64 bit and mingw. There's a readme.txt
and build.bat included with instructions on how to build.
image, no images need to be refreshed then but the cursors needs to be redrawn
still. Otherwise it gets stuck, especially annoying in unexpected cases where the
mask has reached the maximum for all pixels in the brush.
of editmode on the child object.
Problem was that the object custom data mask was not taken into account when
rebuilding the derivedmesh in some cases, which is needed for the derivedmesh
to contain the mapping back to the original vertices. Now this data mask is
used for any derivedmesh build that will be cached.
Also problematic was that the datamask for the active object was applied to
all objects in the scene, which caused the parent object to be recalculated
when it didn't need to be. Now this datamask is only used for the active object.
properly with blender internal is probably for after release, when we can add
the same system to gather textures from brushes, modifiers, force fields, ...
Moved the code after the masking check so we can skip the texture lookup
if the pixel is done, is a bit faster. Also hide the color wheel for these
tools, only did it for 2D paint in previous commit.
used to see if area has been covered by brush, so exclude the texture
alpha and texture alpha masks from mask comparisons. Coming next:
texture masks for 2d painting.
* 2D image painting with textures that contained alpha did not work correctly,
had been broken for a while.
* 2D image panels texture (mask) panels showed wrong buttons for texture overlay.
* Texture map mode 3D now also uses masking, like Tiled and Stencil the texture
does not move along with the brush so it works fine.
* 2D image paint View mapping did not work correct, especially noticeable with
Rake rotation.
* Masking is now disabled for the smear tool, this can't really work because
the original image is constantly changing and gave artifacts.
Animplayer treated .tif extensions as movie files, so that didn't work.
Added another hardcoded check for it, like for png tga exr jpg etc.
Why FFmpeg thinks .tif is a movie... that's for another day :)
Quicktime QTKit did same btw.
scaling it along one axis, now there's a Reset Transform button.
The Image Aspect button is now also hidden unless the texture is an image texture.
And also hide the color wheel for painting tools that don't use colors.
* 2D image painting support for masking to limit the max contribution of a stroke
to a pixel, to get it working compatible with projection painting. Not strictly
a bugfix, but the inconsistency here was annoying.
* Fix python errors in Texture Mask panel in image editor, was missing overlay
options.
* Clamp paint mask to 0..1 in case some texture exceeds it, this could give black
pixels due to integer overflow.
- highlighed buttons that were never pressed could be activated by accident when canceling search popup.
- canceling a search popup on a popup would close both.
accessing a bmesh from python would reallocate all customdata layers.
add an assert to BM_data_layer_free(), when its called unnecessarily since its reallocating all layers.
crossing themselves.
Again an old issue, the code that was limiting the max contribution to a pixel by
doing a max() operation, which is very sensitive to the spacing of the stroke dabs.
Instead we now use a formula does adds up on repeated dabs but approaches the
maximum brush value slowly and never exceeds it, which gives nice smooth results.
mask_accum = mask_accum + (sqrt(brush_max) - mask_accum) * sqrt(brush_max) * mask;
It's now default to 2D textures, and no AUTO mode at this
moment, since detecting which method is the best not so
simple.
Image drawing could manually be switched to GLSL for tests
and feedback, but for default GLSL is not so much great.
Reason of this is huge images, where operations like panning
becomes dead slow comparing GLSL vs. 2D texture.
Dynamic Paint was also being calculated during "orco" mesh generation, causing image sequence baking to use orco derived mesh instead. This likely affected vertex type surfaces too in some cases.
Was caused by svn rev53181, and it worked before because
image buffer didn't have quality set and in this case
fall back to scene settings happened.
Now or render result quality from scene settings is always
used, image buffer's settings is ignored.
The node link operator had a feature to automatically expose sockets in node groups when ctrl+shift+clicking on a node socket, which would create a node group input/output node. This was intended as a shortcut but conflicts with other features such as socket selection and viewer creation. It is also hardly necessary now that input/output nodes have an extension socket, which is much easier to use. Removed this expose functionality completely.
Labels in UI were not drawing greyed out when disabled.
Bug in post 2.66a commit to make greying out buttons draw much nicer.
User reported Addon menu was not clearly showing disabled Addons anymore.
- was reading outside memory bounds checking the 'x' point.
- inserting a point to the right of the last point would add a point to the very left instead.
resolve by using faces adjacent to the ones directly connected to the edge that sliding.
This isnt a prefect solution but it resolves the common case where an edge slides along a UV seam.
Fixed by adding a "texture_user_property" member to spacebuts' context, and using it to get the prop identifier in ui script.
Thanks to Brecht for its advice!
previous fix for 2D image painting were done, and also:
* For brushes that do masking (keeping the max contribution to a pixel limited),
the alpha from textures is now part of the mask. In many cases the logic
worked out the same and where it didn't it used to cause artifacts.
* Color interpolation for smear and soften tools now happens in premultipied
space, to avoid bleeding of RGB colors from zero alpha areas.
* Fix a few cases where byte <=> float conversion was not doing the proper
straight <=> premul conversion.
* Replace some float division by constants with multiplications, is faster.
Note: float texture painting seems to have some issues updating the OpenGL
texture, but issue was already there before this commit.
Since the bug was fixed that made blur brush _not_ blur all verts, there is no way to blur vertex colors after baking.
While this was hidden it was useful especially for baking vertex colors.
Sequencer was always trying to do GLSL color space
conversion, not respecting user settings at all.
This failed a lot when RGB curves a used in color
management settings.
Now sequencer will fallback if GLSL can not be used
and will also respect user settings (however, draw
pixels are not supported, sequencer always uses 2D
textures).
Color Pickers in Blender support precision editing with holding Shift key.
Unfortunately, on first click the picker cursor moves to the mouse position,
making the precision dragging quite useless.
Now, if you hold Shift, the picker editing will only apply the delta motion
of your mouse drags, making it much nicer for fine tuning of color grades
in Sequencer or Compositor.
old issue, the formulas here were never quite right, should all work ok now
with byte and float images.
Some differences:
* Colors with zero alpha from the background will never have an influence, so
you don't get alpha fringes when painting over such areas. This does give
hard edges when looking at the RGB channels alone, but there's no way to
avoid that and fringes at the same time, same behavior as other painting apps.
* Add/Subtract/Multiply/Lighten/Darken now leave the alpha channel unchanged
and work only the RGB channels, again same behavior as many other apps.
* Erase/Add alpha now compensates for premultiplied float images to keep the
straight RGB colors the same.
Next: fix projection painting.
previous commit was incorrect, the face flipping depended on the orientation of the curve.
fix by passing the bevel direction to the fill function so we can have a reliable front/back.
This also gives some speedup for all curve filling since we can avoid calculating the normal since its already known.
- Sequencer preview was clamping float buffers
- ACES color space wasn't correct, was noticeable when
applying display processor from linear space to display.
- Extended sRGB LUT to sRGBf from nuke-default config.
Makes sequencer behave much better in sRGB space.
Loop cut slide now works with multicut.
issue is that since edge slide now supports more configurations, cases where loop-cut-slide silently failed before,
now slide (albeit in an awkward way at times).
Fix this my adjusting the selection of after loopcut so edge slide can properly operate on it.
Issue was introduced in svn rev54374.
Now made it so image's flag from imagewrap*() argument
is checked instead of using tex->ima, which seems how
this functions were supposed to work.
Doublechecking on this change is really welcome!
Originally I was thinking of not restoring this functionality (see report for
further details). However, upon rechecking the code, it seems that there was
actually a define set up for this already, but which wasn't actually hooked up
yet. So, this commit basically exposes this option ("Sync Length") which ensures
that when exiting tweak mode on a NLA Strip, that strip instance (and not other
users of the same action) will be updated to include the newly added keyframes.
This option is not enabled by default, and shouldn't really be needed in most
("intended") workflows.
Issue was introduced by svn rev56273 and was caused by
both smokeModifier_do and smokeModifier_reset (which is
called from smokeModifier_do) do mutex lock/unlock.
This lead to quite undefined behavior caused by the same
thread released mutex twice.
Solved by not locking mutex from inside routines calling
from smokeModifier_process -- mutex is locked in parent
function anyway.
Blender's old default of 1 was too small here.
Now we don't override the default.
If it's really needed to use a smaller value here, the property should
be first exposed in the ui, but don't think this is necessary.
Fixes [#35076] Dynamic objects have periodic glitch in velocity when on an incline
Some enums' items actually are generated from data (like the render layers of compo nodes), so they should not be translated. Added a PROP_ENUM_NO_TRANSLATE new RNA flag to tag those enums (only found those for nodes, but may be more of them around).
Also fix similar issue in main list of render layers (Py UI code! :P ).
- update copying code for linestyles to copy gaps
- update copying code for freestyle line modifiers to copy mapping flags and seeds
- fix drawing code for Material modifiers after a recent refactoring
at vertices for the entire mesh, which meant that discontinuous face colors
would be lost for unrelated parts of the mesh too. Now only vertices within
the brush radius will be affected.
The problem was that vertex colors only have 8 bits of precision, and integer
division always rounds down, so after some color blending iterations everything
gets darker. Instead use integer division that behaves like round() instead of
floor() for blending operations.
- when in wireframe mode: don't snap to faces, instead snap to the closest edge/vertex.
- when not in wireframe mode: snap to the front-most element (was a problem that it could snap to an edge/vert behind the face)
- reduce the distance for selecting ruler points, was too easy to accidentally drag a ruler.
Now copying a scene will also duplicate groups that consist entirely of objects
that are duplicated with the scene. The rigid body world will then also pointers
to these new groups.
The regions of the space are stored in a different place depending if it is
active or if another space is in use. The code here was iterating over both
but it should be only one because the other might contain regions of another
space.
- build with netbsd works again.
- select uv more/less was crashing when called outside image space.
- node RNA property update was crashing when not called in node space.
nicer when the new preview draws over the old one.
The code was changed so that the render result is freed all the time because
freestyle manipulates render layers. Now it only does it when freestyle is
enabled so cycles and regular blender internal can still avoid it.
viewport render changes.
Actually was an older issue that would cause unnecessary preview render restarts
if the render result was not allocated before it was drawn. Now the render result
is being reallocated each time for freestyle which made the issue worse.
Added a mutex lock for smoke data access. The render was already working with a
copy of the volume data, so it's just a short lock to copy things and should not
block the UI much.
The design changes coming with pynodes for the node editor allow editing multiple node groups or pinning. This is great for working on different node groups without switching between them all the time, but it causes a problem for viewer nodes: these nodes all write to the same Image data by design, causing access conflicts and in some cases memory corruption. This was not a problem before pynodes because the editor would only allow 1 edited node group at any time. With the new flexibility of node editors this restriction is gone.
In order to avoid concurrent write access to the viewer image buffer and resolve the ambiguity this patch adds an "active viewer key" to the scene->nodetree (added in bNodeTree instead of Scene due to otherwise circular DNA includes). This key identifies a specific node tree/group instance, which enables the compositor to selectively enable only 1 viewer node.
The active viewer key is switched when opening/closing node groups (push/pop on the snode->treepath stack) or when selecting a viewer node. This way only the "last edited" viewer will be active.
Eventually it would be nicer if each viewer had its own buffer per node space so one could actually compare viewers without switching. But that is a major redesign of viewer nodes and images, not a quick fix for bcon4 ...
The previous fix limited overlap, but is sometimes
too conservative, and artists want way to turn off
the limiting, so added 'Allow Overlap' option to
modifier.
Casting a ray onto an editmesh was building a derivedMesh, raytree, then freeing for every ray-cast.
Noticed while using ruler+snapping in editmode.
Instead of attempting to align the MFace and edit-mesh tessfaces, just use editmesh for ray-casting.
Reverted the trunk revision 56136 and part of revision 56127 concerning the local Main
for Freestyle and temporary scene generation for stroke rendering.
The function do_merge_fullsample() in pipeline.c has access to the Scene of each Render,
so that the temporary Scene generated by Freestyle has to be kept for FSAA even after
stroke rendering has been done. By the same token, the local Main has been moved from
the BlenderStrokeRenderer class to Render. It is noted that free_all_freestyle_renders()
in pipeline.c is intended to get the temporary Scene of each Render released specifically
taking account of the FSAA case.
- disabling ruler snap wasn't setting the points depth back to its previous value before snapping.
- adding a new ruler uses the depth of the previous active ruler (fallback to view center as before)
- deleting the ruler now sets the prev/next ruler active.
- enable site-packages for bundled python distrobution, py3.2 had a problem where it would try to parse headers we didnt include, but its resolved now.
- workaround for glitch I was having on arch-linux where lib64 would be be used for the bundled python directory when it was just a symlink.
Changing actions via RNA (or apparently from the Action Editor browse menu too)
didn't perform all the necessary updates/tagging/recalc needed to have the
results of the new action get immediately applied in the 3D view. This caused
problems for exporters, as the first frames exported would be incorrect (though
this could probably be worked around by manually stepping the current frame
forward one frame then back again).
Allow separate control for tex/mask/cursor overlay. This commit implements
separate overlays for mask textures and cursor curves. The user can turn on
and off separate parts of the overlay by using the appropriate widgets.
The cursor overlay widgets are located at the tool selection panel
Also fixed alpha masks not getting correctly masked and mask texture mapping
not having the correct update callback
properties were continuously redrawing, which slowed down everything else.
The problem was integer overflow, with a short only capable of storing values
up to 32767. Note that sockets are collapsed by default since the previous
release, and that's it's not very useful to edit such complex node setups in
the properties editor, it's mainly meant for simple setups or group nodes to
present just a few sockets.
Two main things:
- Made a python operator for selection in a viewport
which will de-select everything if nothing is under
the mouse.
To do so needed to modify VIEW3D_OT_select, so invoke
sets mouse location which is later used by exec
function.
This way it's possible to select stuff from python
defined operator.
Not best-ever solution since ideally exec() shall not
do OpenGL stuff, but we've got this issue in some
other operators. We'll solve this later.
- Used a keymap from Gianmichele Mariani as a reference,
updated his keymap to latest changes in operators.
We shall match Maya keymap much better now, thanks
for the keymap dude!
Most of the places which relied on RNA_path_resolve() did so believing that if
it returned true, that it had found a valid property, and that the returned
pointer+property combination would be what the path referred to. However, it
turns out that if the property at the end of the path turns out to be a
"pointer" property (e.g. "data" for Object.data), this would automatically
become the pointer part, while the prop part would be set to null. Hence, if a
user accidentally (or otherwise) specifies a path for the single-property driver
variable type like this, then Blender would crash.
This commit introduces two convenience functions - RNA_path_resolve_property()
and RNA_path_resolve_property_full() - which mirror/wrap the existing
RNA_path_resolve() functions. The only difference though is that these include a
check to ensure that what was found from resolving the path was in fact a
property (they only return true iff this is the case), and make it explicitly
clear in the name that this is what they will do so that there's no further
confusion. It is possible to do without these wrapper functions by doing these
checks inline, but the few cases that had been patched already were pretty
hideous looking specimens. Using these just make it clearer and simpler for all.
I've also beefed up the docs on these a bit, and changed these to using bools.
Displays such information as current frame dimension,
frame number within image sequence/movie and in case
of image sequence input displays current file name of
a frame.
Not entirely happy with such approach, but was requested
a lot by artists.
Issue was cause dby ima->ibufs being stored in a ma on
undo/redo and ima->rr not. In case of multilayer image
ibufs hares pointer with render result, so current
undo/redo policy lead to a dead pointer stored in ibuf.
Made it so ima->rr also stores in the map and restores
on redo/undo keeping all the pointers fine.
rewrite to use only 2.
also use generic, reusable functions for selection flushing so each operator doesn't need to implement its own.
and merge more-less operation into the same function, just call the selection flush function with select/deselect arg.
This is a experimental collision detection function, so the API might
change in the future.
Note: The simulation needs to be stepped before this function can be
used, otherwise the rigid body world might not be valid.
Patch [#34989] Bullet Convex sweep test API
by Vilem Novak (pildanovak), thanks!
The original rationale for adding the node_type property was to allow node group operators work generically on any node group type automatically. The problem is that detecting the appropriate node group type and node tree type to use for a group depends on using a node base type. Due to the fact that RNA does not allow multiple inheritance (mixin classes) this is impossible to achieve if node types also have to use a base type such as ShaderNode, CompositingNode or TextureNode.
The idea is now to just "make it work" by limiting the node group operators to the standard tree types. For future pynodes we can implement these operators nicely in Python, which will allow pynodes to use mixin base classes or derive their own operator types and re-use the same keymapping.
After planar codecs support minimal FFmpeg was
bumped to 0.10 which was not so much nice because
it was only released only later last year.
Didn't find a way to make compatibility code local
in ffmpeg_compat, so there're some ifdefs in
audaspace and writeffmpeg.
Not entirely happy, but having a bit of ifdefs in
code better than lots of real PITA for platform
maintainers.
Wrongly replaced a "&" with "|" for a define. That made node headers in node editor
draw too dark. Also made header for default node not use alpha.
In general this drawing is not very good - it gets too transparent and dark.
Needs nicer AA function in interface, will be done later.