[ODE] reply on mesh mesh contact

[Aras Pranckevicius]

> but having contacts  for depths computations some more could  be
considered
> for ex paper  http://www.cs.unc.edu/~geom/DEEP/icra02.pdf

I have considered DEEP (and some other things, like GJK-derived ones for
penetration estimation) a while ago. I have even thought I would implement
one, but as always, no time :(

All these approaches are NOT for arbitrary triangle models (like current
TriMesh) - they are for convex geoms. I guess it would be very good to have
a convex geom in ODE, but it requires some work and some asset processing
tools (eg. a convex-piece decompositor or something like that).


Aras Pranckevicius aka NeARAZ
http://www.gim.ktu.lt/nesnausk/nearaz/