[ODE] Updating ODE documentation (was: hello world)

[Adam Moravanszky]

> solver, and could be easily exported through the APIs. if you need
> this mechanism, can you add it to ODE?

I could.  I have a deadline for finishing up interactions between simple
Torque engine objects and ODE objects next week.  Then I will compile a list
of stuff I don't like about ODE and try to prioritize.  On monday I have
been reassigned from doing dynamics to improving Torque's graphics for the
rest of my time, (because ODE works so damn well, we thought we'd shift my
focus somewhere it is actually needed) so any ODE hacking will happen on the
back burner.

I already have ODE objects work perfectly with the Torque client side
network interpolation and extrapolation.

Some folks on this list have mentioned that they are trying to bring Cal3D
and ODE together.  We are also trying to do this in our project -- we'd like
to have an AI controlled creature whose ODE-dynamics skeleton is animated by
applying the right forces.  (In fact, the author of Cal3D is the person
working on this, though for the time being we decided not to use Cal3D
itself.)  We have read a paper by another group ('Pseudo interactive') who
claim they did inverse dynamics for such an animation controller.  (Knowing
that there are at least as many ways to perform inverse dynamics as forward
dynamics, this information in itself doesn't help much.)  I was wondering if
this is really neccesary, or perhaps a PID controller by itself would work.
We doubt we have the resources to get into inverse dynamics.

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