but they got this case! :)
Bug was that for autohandle, the previous position of handle was used for
calculating the position too, which I really don't remember why... (that
is code from 10 years ago). Problem with that approach is that extreme
changes in handles don't go correct immediately, but need to itterate a
while. Or even worse, can give NaN values, resulting in this:
http://www.blender.org/bf/ipobug.jpg
Now the handle points are fully recalculated, as it should.
When Blender tries to open a non-existing file, two error popups were
called. Very annoying.
Was simple one-liner fix; the 'read exotic' function should return a -1
when file doesn't exist. :)
Creating texture coordinates for Curve bevels didn't allocate a correct
sized memory block when both U and V directions of a bevel are circular.
This will also fix "UV orco" for such curves (like donut shapes).
Particles keep fighting with the depsgraph... time to get a depsgraph with
time dependencies in it...
Anyhoo, this fixes lack of parenting updates in hierarchies where children
have particles.
Silly: when using vector blur on a curve or text object, without having a
material assigned to it, the default material didn't get initialized OK
for vector blur, causing random streaks.
- Ztransp looked weird in Node previews, only showing the backfacing pixels
- previous change in preview.blend accidentally set camera clipping too low
for correct display of lamp preview
- refresh issue solved in preview when using Node shaders with ray-mirror
Path in blender file loader end (setup_app_data) for armature conversion
didn't properly limit to armatures only, and was even called before a
scene dependency graph was made.
Rare crash btw; you have to save a file *before* a redraw was finished
in the UI.
naming convention for Compositing:
- Render Result node -> Render Layers node (name only appears in Add menu)
- Compositor image -> Viewer Node image
I've also added a version patch (2.41 saved files only) to rename existing
"Compositor" Images.
The buttons uiBlock was stored in a Node, for retrieval of node buttons,
however that won't work when multiple windows show same Nodes.
Now implemented more correct unique Block names (as all over in UI), and
use that name to retrieve buttons with API call uiGetBlock().
- removed "Unified" button, replaced with "HD" preset for 1920x1080 output
- removed the unused "Pass" options
- removed the unused "Strands" render-layer option
Because the internal render pipe supports this already; added two more
render-layer options:
- "Sky", to enable/disable sky render in a layer (this was part of "Solid"
before, not so correct... to ensure previously saved files work, the
"Sky" option is set by default when "Solid" was set. The version patching
will do this temporally always, until we've bumped up version to 2.42
- "Edge", to enable/disable edge render in a layer. Nice for compositing.
Also in this commit: fixed warnings for exported functions for the new
Node Editor pull-down menus.
By default it is disabled (depth 0.0), so rendering is as usual.
The meaning of "depth" and "falloff" will be extensively shown in the
release log pages. Coming soon!
(Patch provided by Ed Halley)
- STUPID mistake from me in setting the waitcursor...
- icons for UI got freed, without checking if it existed (crash in end)
- call to close mainwindow didn't check if window existed
note:
I usually test the "blender -b" case, which should start blender,
initialize all, free all, and print "blender quit" to signal all is fine.
New "Dimension" button: when using TAB to cycle over buttons, the dimension
event was sent multiple times, accumulating scaling. The code was also not
prepared to handle multiple changes at one event.
Also: added object_get_boundbox(Object *ob) in BKE_object.h, so the code
now really supports most primitives in Blender.
Particles: option "even" failed when faces where all exactly same sized,
then it sometimes skipped entire faces.
Just the regular <1.0 that had to be <=1.0
Old issue... Blender saves jpg RGBA files. That's unsupported in all
programs except Blender, always a source of problems with that feature.
Now it just falls back to saving 24 bits when the "RGBA" option is set.
- Selection wasn't possible on center dots in Object mode. I keep adding
it and it keeps disappearing again. Added clear comments in code too.
- Particles: (debug still) print "build particles" happened on each redraw
when "Display percentage" for particles was set to zero.
The Scene audio.mixrate variable was zero on creating new scenes, causing
synced playback to fail.
Added initialize in kernel, and patch in buttons drawing to set the value
for wrongly saved files.
(Pixelshading: small correction for more optimal code)
now scale correctly.
Formerly, they just used to scale away from the curve into nirwana. Now,
they nicely follow the curve while scaling. Both the curve and the text
objects themselves can be scaled.
This will of course break compatibility with files that used this "bug".
Acceptable? I guess :-)
Added support for XVid and H264-codecs in codec-selection. (only work, if
ffmpeg is compiled with XVid and/or H264-support. Failure in doing so
results in an error message that codec can't be selected.)
Both are written always to AVIs since raw-h264-files created by ffmpeg
can't even be opened by itself...
Video render options are reset to sane defaults (=DVD preset) on startup now.
Don't expect quicktime-support to be very exciting, since ffmpeg can't really
multiplex quicktime files. (Tried several codecs with the ffmpeg-commandline
tool,... sigh)
Timestamp crash on Debian-Sarge version is fixed.
And more particle issues... the attempts to allow particle systems to be
part of duplicators (groups especially) isn't easy to get working!
This commit solves regular animated particle systems... they rendered on
the wrong location, using wrong object transform.
This fixes again some segfaults on error conditions.
It also gives some reasonable error-messageboxes if chooses the wrong
DV-format-options.
More to come...
Bugfixes:
- hddaudio now allocates it's decode buffer + AVCODEC_MAX_AUDIO_FRAME_SIZE
safety interval. (The former code expected all decoded audio frames to
be the same size which can hurt under some circumstances e.g. VBR files)
- writeffmpeg: some pointers where not initialized on start and after
deletion not set to null. Could segfault randomly on error conditions.
- drawseq: use startdisp and enddisp to decide, if a strip should be drawn.
(Just extend strips first or last frame beyond screen dimensions in
the previous version and watch the strip disappear)
Small fix, but results are at least less frustrating now. It uses the
"compatible euler" function from inserting key positions here, preventing
euler values to be constrainted that differ weirdly.
I've tried several other approaches to get a definite rotate constraint,
but only constraining a single axes seems to me impossible magic still...
The new (months ago) added option for "Local Constraint" always copied
and restored a PoseChannel quaternion, that disabled the required
normalizing of quaternions.
Only visible when you manually draw Ipo Curves, but it's probable that
it showed in other cases too.
Object Layer Ipos didn't work when the Ipo was moved an Object Action.
Can't get this to work though... this option has been added with a lot
of exception handling already, and for action/nla it's very nasty to
blend/mix/add layer bit values. So; better not support that.
As feedback for users, I've added a notice popup when you try to move
layer-ipos to an action.
Compositor:
Viewer Nodes inside of Groups now work too. To not frustrate interactive
speed, the following rules apply:
- Making a Group editable (or closing) doesn't signal recalculation of
composite
- clicking on an Input socket always checks changes and calulcates
- When there are Viewers inside an edited Group, Viewers in the main tree
are not executed.
Also added: a "hide unused sockets" icon in the header of Viewer nodes.
This allows cleanup of Groups, to prevent these sockets get reveiled.
Each report was about a different failure with Particles, all related to
weak handling of animation systems and the depsgraph.
Fix has 2 parts; depsgraph now signals "object recalc" to be for time
changes; this then is used to bypass particle-building (since that's baked).
Other part is better object caching while makig particles.
