[ODE] Lots of stacked boxes/breaking constraints (was Re: Inside ODE)

[Adam Moravanszky]

> i have nothing against this approach, but i'd like to somehow
> build on the existing box-box collider. the existing box-box


After you have a single contact with your method, and the boxes are just
touching but not intersecting, you can limit the further search of contacts
to the separating plane formed by the contact point and normal.  Other than
that, you have no proof of there being only one contact unless you do an
exhaustive search.

I would generally recomend giving ODE users a larger than minimal tool set
and *choice* about which tools they use.  In this case I would provide two
box collision detection options, one as-is, and one that finds all contacts.
The current code is good enough for a lot of projects (including ours) but
some people have problems with it, which is why we have this thread.

I would actually not worry about there being too much overhead in searching
for additional contact points.  In the box-box case, you will mostly have a
single contact only when a moving ball is bouncing off another.  And
bouncing contacts don't last long.  On the other hand it is unlikely to have
a situation where boxes are at rest or sliding and have only one contact
between them.  And it is this case which takes place over a longer period of
time, so it is more important when optimizing.

--
-- Adam Moravanszky
http://n.ethz.ch/student/adammo/