* Rendering twice or more could crash layer/pass buttons.
* Compositing would crash while drawing the image.
* Rendering animations could also crash drawing the image.
* Compositing could crash
* Starting to rendering while preview render / compo was
still running could crash.
* Exiting while rendering an animation would not abort the
renderer properly, making Blender seemingly freeze.
* Fixes theoretically possible issue with setting malloc
lock with nested threads.
* Drawing previews inside nodes could crash when those nodes
were being rendered at the same time.
There's more crashes, manipulating the scene data or undo can
still crash, this commit only focuses on making sure the image
buffer and render result access is thread safe.
Implementation:
* Rather than assuming the render result does not get freed
during render, which seems to be quite difficult to do given
that e.g. the compositor is allowed to change the size of
the buffer or output different passes, the render result is
now protected with a read/write mutex.
* The read/write mutex allows multiple readers (and pixel
writers) at the same time, but only allows one writer to
manipulate the data structure.
* Added BKE_image_acquire_ibuf/BKE_image_release_ibuf to access
images being rendered, cases where this is not needed (most
code) can still use BKE_image_get_ibuf.
* The job manager now allows only one rendering job at the same
time, rather than the G.rendering check which was not reliable.
library loading was using some non-public OS X functions, which give
linking issues for me. Since OS X 10.3 standard unix dlopen() is
supported, so I just removed this code, we don't support earlier
versions anyway.
changed with the file path changes, the reason to check this first
is that it allows to edit py scripts without having to run the
build system for each change.
This (biggish) commit generalises the rotation modes functionality added for Bones, allowing Objects to use the various Euler Rotation orders, Axis-Angle, and Quaternion rotation representations.
I've also cleaned up the nomenclature of the rotation-related settings so that the naming styles are more consistent with each other. Unfortunately, this will break all files involving object or bone rotation animation made in 2.5 versions (2.4x will still get correctly converted).
General Notes:
* By default, Objects still default to using Eulers, while Bones will use Quaternions by default still.
* I've fixed all areas that I'm currently aware of to work with these changes. However, there are probably a few places where I've missed a few changes (i.e. auto-keyframing will need attention later).
* Removed the old "IPO-Keys" stuff from Transform code. I'm unlikely to restore this in the near future, and trying to fix that to include support for this commit would have been too much work.
http://wiki.blender.org/index.php/BlenderDev/Blender2.5/Unix_FHS
for scons WITH_BF_FHS enabled an alternative layout eg.
scons WITH_BF_FHS=1 BF_INSTALLDIR="/usr/local"
for CMake just run "make install" after make (CMAKE_INSTALL_PREFIX is used for the base path)
Currently only scripts use both the system and user path correctly, other areas of blender have their own path code inline with lots of ifdefs, needs to be carefully updated.
* VecLerp3f and Vec2Lerp3f - interpolate 3 2d/3d vectors from 3 weights
* AngleToLength(angle) - useful for making even width shell/walls based on the angles of the surrounding geometry from each point. (same method used in 2.4x python solidify script).
also quiet some warnings.
Added a utility function to check which transforms for an object or bone are animated, returning these as bitflags and/or optionally retrieving the relevant F-Curves too. Beware that this method may not be working correctly yet, but it shouldn't hurt anyone in the meantime :)
Also, split RNA-path building function up into a version which only creates the path up to the given struct, with the other parts being added later.
- use_radius option, off by default for 2.4x files, on by default on new curves.
- curve deform modifiers (think tentacles)
- follow path (parent mode and constraint)
- curve guides
- added back Alt+S to scale point radius
- Mat3Scale and Mat4Scale arithb.c functions to make a new uniform scale matrix.
- TODO, effectors, looks like they have no way to scale from the radius yet.
* Added a few new methods for axis-angle conversions, and used these instead of manually performing those steps elsewhere
* Axis-angles to other representations now get their axes normalised to make sure that odd scaling doesn't occur.
* Made a few more tools work with axis-angles properly
* layout.itemR now has icon_only option to show only icon
in e.g. enums buttons, for uv editor header.
* Automatic key shortcuts in menus now show the shortcut even if
operator properties don't match. Not sure this will work well
everywhere, but seems to be working ok for now.
* Open recent now show shorter filenames instead of the whole
file path.
* Tweak object Duplicate menu item.
* Split object_edit.c into multiple files:
object_add.c, object_edit.c, object_hook.c, object_relations.c,
object_select.c, object_transform.c.
* Rename files to have consistent object_ and mball_ prefix:
object_shapekey.c, object_lattice.c, object_vgroup.c, mball_edit.c.
* Added operators:
* vertex group menu and set active
* apply location, rotation, scale, visual transform (location is new)
* make local
* make vertex parent
* move to layer
* convert to curve/mesh (not finished yet)
* Many small fixes for marked issues, but still much code to be cleaned
up here...
moved CU_2D to Nurb->flag in do_versions
This made simple type checks confusing to read.
many... if( (nu->type & 7)==CU_BEZIER)
replaced with ... if(nu->type == CU_BEZIER)
made setting rna curve_2d clamp the Z values. still more RNA/UI changes to do.
* Added missing prototype from arithb
* Added define for Axis-Angle rotations. This has not yet been hooked up for actual usage yet, since there are some issues regarding evaluation which I'd like to investigate further first.
* Editing numbuttons for posechannel transforms now updates in realtime. Sending ND_POSE|ND_TRANSFORM was causing confusion for the listeners (which only check for either/or)
* Partial fix for axis drawing on bones. Now the axes are drawn at the tips of the bones again, but unfortunately the texts aren't (since they only use the object matrix, they get drawn at the origin).
* Copy Rotation constraint should now work ok with this new code again. Previously, it was the only thing that really went beserk when the typos were still uncaught.
* Fixed one other case of a potential case where typos would cause problems.
* Made changing the rotation order setting perform conversions of the current rotation to an equivalent representation in the other orders/forms. This is done at RNA level, so maybe not that great for switching representations while animating?
* Rotation order code should be more correct now. Previously was only shuffling axes, and was also doing some evil things to provided that that it shouldn't have been doing, which was causing some flipping issues.
* Built-in keyingsets for 'visual' options should now be more correct. The old code had typos, giving wrong array indices to start from.
http://www.graphicall.org/ftp/ideasman42/curve_auto_twist.png
Access next to the "3D" edit button.
details...
- open curves use the first points orientation and minimize twist for each new segment.
- cyclic curves calculate the least twist in both directions and blend between them
- AxisAngleToQuat replaced inline code.
- Notice the model on the right now has more even corners. added Vec3ToTangent to arithb.c.
* bring back 'player' libtype, after investigation with ideasman.
scons/mingw works nicely, for some reason msvc fails to link still, will look further into it.
NOTE: this needs changes to stubs.c, but need to check with ideasman_42 how to fix with cmake. Probably linking order issues, but stubs.c currently generates warnings for msvc (redefinition of funcs) and errors for mingw (same redefinitions). Removing the offending lines from stubs.c fixes that.
* further cleaning of 'player' stuff. Now only 3 libs are remaining, of which ideally the stubs lib will be fixed at some point, fading away into the dark history of not-so-nice code. The current blenderplayer part is still a little bit hackish, I'll see if I can find a better alternative, for now it works good enough.
* first working changes to get blenderplayer linking
* blenderplayer/ moved into source/ (CMakeLists.txt changed for that too)
* added externs for bprogname to gp_ghost, so that it links properly
Now there is (in order of speed):
* Nearest neighbour (very rough quality)
* Linear (medium quality)
* Quadratic (good quality)
* Cubic Catmull-rom (very good quality, crisp)
* Cubic B-spline (very good quality, smooth)
Thanks!
Story time: Once upon a time, in the green valley of fileselect, BLI_end_threads would get called on an empty threadbase, depending on the result of a previous call to readdir(). The function would then gladly decrement thread_level to -1 which would cause all kinds of fun havoc. THE END.
Made sure thread_level is only incremented and decremented when needed. The caller should never have to make sure of that, especially since it already lets you call with a null threadbase.
Please report any further hang (and how to reproduce, if possible).
* Added a 'rotOrder' parameter for constraint evaluation objects and constraint targets, which describes the rotation mode used for the matrices there.
Todos:
* Constraint targets default to using XYZ only for now. This will need be be addressed eventually.
* Copy Rotation constraint currently cannot use the new rotation order code for the target matrix. What's surprising is that even though it's just using XYZ as the old code did, it doesn't work, and yet everything else works nicely. Silly constraint! (it is almost impossible to improve this constraint further without breaking a rig out there)
* All tools where rotation order matters for armature bones have now been adjusted to use the new code
* Transform now uses the new code for bones too. However, there are some jumping issues here that I'm not too sure how to solve yet. Help fixing this is welcome.
This commit is the start of an implementation of (euler) rotation orders for Bones (later to be extended to Objects too).
Technical details and references can be found at:
http://wiki.blender.org/index.php/User:Aligorith/EulerRotationOrder
In short, I've added a new set of Euler conversion functions (EulO... and ...EulO), coexisting with the old functions for now, which can handle different rotation orders.
Changes have only been made to the basic evaluation code. However, the following places will still need modifications:
* Transform code - needs to be made to use functions which take rotation order into account instead of using XYZ only
* Rotation constraints - same story
* Other rotation editing tools for armatures also need a check up, since there might have been some missing code when I ported eulers earlier
Also, made the Outliner's horizontal scrollbar work better for keymaps view. It's still using an approximation of the width, but at least you can scroll now.
