ironing out a few glitches related to goal pinning
volatile test blend ->
http://www.wund.homepage.t-online.de/hidden/sb_col_must.blend
+ adding a few alternative collision ball calculation algos
since i did not want to blow up (waste) DNA space without need,
coded as -1.0 -0.1 -1.1 values in 'ball size' (look at tooltip popup)
added in place operations.
Vector_iadd vec1+=vec2
Vector_isub vec1-=vec2
Vector_imul vec1*=float or vec1*=mat
Vector_idiv vec1/=float
length is now writable vec.length= float
removed the need for casting python objects to Vectors pyobjects when performing vec/float arithmatic.
a PyObject for coercing has also been removed from the vector struct so a little less memory will be used also.
Benchmarked before and after this change
___________________________________
import Blender
v= Blender.Mathutils.Vector
m= Blender.Mathutils.Matrix
a= v(1,2,3)
b= v(3,2,1)
c= m()
t= Blender.sys.time()
for i in xrange(20000000):
a*b
a*10
a/10
a+b
b-a
a*c
print Blender.sys.time()-t
_______________________________________
Before 63.5sec
after 49.5
about 3 sec of that is looping
file was also happening on files made with current CVS (until next release).
Added a simple check to make it only happen on old files where there was no
radius before.
This patch adds the feature of IPO curve mirroring. It mirrors the selected keyframes in the active ipo curves over a user-nominated mirror-line.
Options are:
- Mirror over current frame
- Mirror over vertical axis (X = 0)
- Mirror over horizontal axis (Y = 0)
Functionality can be accessed from:
- Hotkey M in IPO editor
- Submenu in header of IPO editor
** This is a heavily modified version of the patch, but the idea is the same. **
This adds X and Y as hotkey for constraining IPO transforms (grab only) and also fixes the display to only show relevant information when constraining (grab and scale).
That is, a 3D curve now displays as ->->->-> instead of -|-|-|-|.
Comparing the two, I didn't find the arrow display any more confusing then the straight lines when dealing with loops and tilt variations. Sharp turns are more likely to display crossed lines though.
Others should feel free to disagree and make it better.
simple (naive) self collision
estimates a collision ball
using the spring lenght to attached neighbours
(idea stolen from old cloth modifier files, but i'm sure ZAZ and genscher won't mind)
-- usefull for untangeling static cloth like objects
-- may be used for cloth simulation with known limitations
like 'tunnel effect' if objects are moving too fast ...
demo -- volatile files
http://www.wund.homepage.t-online.de/hidden/sb_without_selfcol.avihttp://www.wund.homepage.t-online.de/hidden/sb_with_selfcoll.avi
have fun
BM
Some operations like edge loop deselect would cause selection history
to become invalid. Added a sanity check to countall() to try and catch
these when they happen.
The wait cursor was being called during editmode enter and exit for meshes.
This was a problem for several reasons. First of all, python modules like
Mesh now make use of editmode features. These methods that wrap editmode
tools may be called many times during the execution of a script
and lead to the wait cursor rapidly flickering on and off.
The other problem was that the wait cursor wasn't being called for editmode
enter and exit of all data types. This is unified now.
-New Arguments
enter_editmode() should be passed a nonzero integer or simply EM_WAITCURSOR
if the wait cursor is desired. Currently only the python API passes a '0'
to enter_editmode()
exit_editmode() has several options and they are passed in as the bitflags
EM_FREEDATA, EM_FREEUNDO and EM_WAITCURSOR. These flags are defined in
BDR_editobject.h.
countall() wasn't being called when scenes were being added/deleted.
This caused the info header to temporarily display incorrect information
about scene statistics.
Added Mesh .key .removeAllKeys() and .insertKey() for MDD support (was using NMesh just for keys before)
Since this is aparently an experemental feature in NMesh we may want to change this.
Gives local structure stability for though fabrics, thin steelplates ..
even suzanne gets pretty crash resistant that way.
If non zero it not only adds diagonals but all other "2edge-wide" springs
so it somehow replaces stiff quads for meshes too.
- A new Save image function, next to the existing Save As.
- Shortcut keys Alt+S, Alt+R, Alt+O and Alt+N for Save, Reload, Open
and New image respectively, consistent with the text editor window.
- Hide Select and UVs menus when UV editing isn't active.
- Removed duplicate and unused code, also moved some functions.
BPyRender.imageFromObjectsOrtho() returns an image
Made all image bake scripts ask before overwriting a file. as well as displaying the newly created image once its rendered.
This never worked that well, so now after unwrapping, uv charts are
rotated to fit in their smallest enclosing rectangle. This should give
the expected behavior in the most common cases.
When using new Shadowbuffer option "Auto clip", the clipping planes could
become so narrow that sampling code - using ints - could result in
overflows (sign flip). This especially for small shadow buffers with high
bias values.
This fix adds an extra check in shadow sampling for such overflows.
This is another proportional edit falloff that slightly randomises the
influence. It's not strictly random, it's blended with linear falloff so that
it's a bit smoother. The nice thing is that it works with all transforms, axis
locking etc, including the special ones like to sphere, shrink/fatten/etc. It
can be used for all sorts of things like roughening surfaces. I most recently
used it to add a bit of randomness to the folds of some cloth. I
made a short demo video here: http://mke3.net/blender/etc/prop_random-h264.mov
Also included nicer icons for the falloff types.
with regards to the noise functions in yafray based on Ken Musgrave's
original code.
I had left the 'Copyright' notice in the comments, and according to
Cyril Brulebois this is a problem. In fact, from what I understand this
makes it even impossible to use or modify the code in other software.
But since it is not a verbatim copy of the code but rather based on Musgrave's
work, he suggested I change it to explicitely state that it is in fact
based on the code from the 'Texturing & Modeling' book.
And since the yafray code is in turn based on the blender code, I better
adapt the blender code too.
This reminded me that I also have forgotten to include the copyright notice
in the mersenne twister rng code I used for the Python Noise module.
This does clearly state to include the original notice with any resdistributed
code, in modified form or not. So I added that too.
I hope that solves the problems.
previous one was ok as a quick fix during orange and showed distortion obviously
but it was kind of extreme, and I'd rather not be blamed for causing blindness
or brain tumours because of it. This pattern is nicer, and still shows
distortion and its location on the map well.
falloff and spacing. Click the button labeled "P" next to the sliders.
Works best for opacity, other options can give poor quality strokes in
some cases. Also performance needs to be improved.
Remember, works only on X11 and Mac, no Windows support for tablets yet.
If you define multiple RenderLayers with different view-layers, these
view-layers are rendered entirely, not looking at the faces that are
invisible. This allows layered composite, but only when geometry does
not intersect or overlap partially.
This new Renderlayer option "All Z values" will fill in the Z of not
visible faces (in the Scene view-layers) as well, which then allows
masking out as if these faces were set to Material option "Env".
Here's an example composite to show the effect:
Two cubes, in view-layers 1 and 2, and rendered in 2 RenderLayers:
http://www.blender.org/bf/all_z0.jpg
Same scene, but with both RenderLayers set to "All Z values":
http://www.blender.org/bf/all_z.jpg
Added to existing scn.objects
scn.objects.active (get/set the active object for the scene)
scn.objects.selected - an iterator that only uses selected objects
scn.objects.context - an iterator on objects in the user context (visible in the current 3d views layer and selected)
These are the same type as scn.objects but .add() .remove() .new() .active etc raise errors. so scn.objects.selected.add() will raise an error.
Made nested loops possible with scn.objects, metaball.elements and ob.modifiers, by initializing the iter value as NULL and creating copys of the pyobject when _getIter() is called if ->iter is not NULL.
This is how pythons xrange() works.
Previously had Vertex snapping in knife tool adjustable by the 'threshold'
value in the 'Mesh Tools' panel. This was actually a misunderstanding on
my part of a user request (oops). What was really needed was to make the
sensitivity of intersection code for vertex cutting adjustable. Vertex Snap
is now set at same tolerance as vertex selection.
feature. Old files that had "sticky" set would not show "sticky" under
the new system. Not that anyone ever actually used "sticky".
Also, these commits only add support for rotated external target objects,
not target bones inside the same armature.
making it much more useful. Requested and taunted for frequently by
Plumiferos folks. Click on "Use Rot" to take target object rotations
into account.
Good for using rotated empties, etc., as a sloped floor.
