Trimesh-trimesh collider (was Re[2]: [ODE] Triangle-box collidercontribution)

[Adam D. Moss]

Nguyen Binh wrote:
>         Oh. Guys! Let hope that is the main result why
>         test_moving_trimesh.cpp behaves unbelievable.
> 
>         Adam, did you find sometime for fixing it? If it's not so
>         hard, please pass to anyone (possibly me :) ). I can't stand
>         seeing a "working trimesh-trimesh" in ODE.

I think it's probably fairly easy (mind, apart from breaking
the collision/simulation barrier I don't know why this can't
be done automatically within ODE -- implementing that cleanly
might be another good small ODE project for someone).

If anyone wants to do it then a patch is welcome -- hopefully
it's as simple as (something like) getting the body's rotation
matrix from ODE and sticking that in the top-left of the
homogenous 4x4, getting the position vector and sticking that in
the 4th column, and then feeding that back into dGeomTriMeshDataSet()
as per the documentation.

--Adam
-- 
Adam D. Moss   . ,,^^   adam at gimp.org   http://www.foxbox.org/   co:3