Note that the script only works in recent builds (about r48650+)
since there is a part already written in c. Bart works on making it faster.
great Tool
I remember when Ton asked over mailing list for someone able to get in touch with Phymec followed by creepy silence lasting weeks or months… then now we have the tool almost integrated in Blender, wow! Again, thanx Campbell!
lol, “creepy silence”.
i’m building trunk this very moment just to try out the script. thanks for updating this thread, i had no idea this was still being worked on. 
ops, almost forgot to thank Phymec for his fantastic work…
i forgot to ask, is there a thread somewhere about the procedure to actually use this script? i’ve gotten as far as i can with it on my own, and now i’m a little confused.
edit: never mind, i guess i needed to re-read the first page of this thread 
i’m building trunk this very moment just to try out the script. thanks for updating this thread, i had no idea this was still being worked on.
Windows 64 by chance? 
And I am very happy this is being worked on, thanks a ton to Phymec.
nah, sorry, Linux 64-bit. 
if i knew how to build for Windows 64, i totally would. i’m using the script right now, and i think it was a bad idea to set recursion level to higher than 0… it’s taking forever to fracture a cube into little pulverized pieces…
Hopefully it will speed up a bit when ported into C/C++.
it’s in the latest build which you can get here
from the meeting notes:
“- Campbell Barton is working on recursive voronoi fracture; based on initial Phymec code - but he decided it would be more efficient to re-implement it in Python entirely as add-on. Should be finished and documented in a week.”
so I think documentation is going to appear soon
thanks, i actually got it working just from trial and error. i was half-expecting this to be something i could use without having to actually set up any game logic, but immediately after i posted i realized that wasn’t the case. no biggie.
i had some fun with it but have gotten some weirdness when fracturing more complex objects (like Suzanne - see the video). but that’s probably user error, i probably have to tweak settings some more.
We cant (or at least I) see the video because its private.
i was half-expecting this to be something i could use without having to actually set up any game logic, but immediately after i posted i realized that wasn’t the case
We can hope this will change after the GSOC Bullet integration is finished.
ah - how weird, sorry about that. i didn’t even set that option. should be fixed now.
and yeah, after playing with this today, i’ve been very curious how this would work with the new bullet integration. i’m stoked!
One of the “rules” when fracturing objects is that they should be watertight. If it has holes in it then it will mess up.
If you take a close look at the Suzanne model you will see that she has holes where the eyes are. Maybe that is what causes the weirdness in this case?
actually that’s what i meant by weirdness - i ran the fracture script on Suzanne and it created a really wonky fractured Suzanne - with no eyes, spaces between cracks, and a lot of pieces sticking out randomly all over… lookin kinda like Doomsday from Superman! but again i’m sure that was just user error, i probably needed to adjust some parameters and let it compute a little longer or something.
i used the same settings on the cube, and it has the same issues, spaces between various pieces. i ran it with different settings on another cube and it worked beautifully.
edit: OHHH, you mean that the mesh itself should be completely manifold? i see. that very well may be part of the problem then. but i probably still messed something up when fracturing it, because of what happened to the cube also…
at any rate, this is an awesome script. way to go, Phymec!
Yup. The mesh needs to be manifold before running fracturing. The spaces between pieces is probably because of the margin setting, it is set to 0.001 by default. But i’ve also had problems with getting spacing between parts even with margin set to 0, it is apparently caused by some crappy convex hull thing that is causing all sorts of problems. Campbell is looking into it.
hehe, it’s actually more like “way to go Campbell!”. Phymec has little/nothing to do with the development of this addon
according to bart there are some issues “the are problems with the existing convex hull calculation in blender”. Which is imo cause of some of the issues.
any chance you could post the .blend file for that?
ack, i tried to post it on pasteall but it’s 79 MB lol! let me know if this doesn’t let you download it - https://www.dropbox.com/s/cqsf43q6csz7nht/phymec_tools_testing.blend
hehe, it’s actually more like “way to go Campbell!”. Phymec has little/nothing to do with the development of this addon
i dunno about nothing 
but you’re right, campbell has credit due him, he seems to working hard to get this implemented.
awesome!
got it, it works, now I just gotta figure it out:eyebrowlift:
so much to learn
Thanks
Dave