[ODE] friction and restitution

[Gary R. Van Sickle]

> "1" does not mean "infinite roughness". If you use the
> Approx1 approximation, the mu tells you how much counter- 
> force the friction will apply as a scale of the object force 
> into the contact (IIRC). If you don't use Approx1, I think mu 
> is a very simple force scaler (and not really all that useable).
> 
> In fact, the code will compare mu to dInfinity to switch into 
> a more efficient "infinite friction contact" code for each joint.
> 
> I still think multiplication is a reasonable approximation, 
> with the only question being what happens when one value is 
> 0, and the other dInfinity :-)

I guess you just have to pick one that wins.  The thing is though that
nothing in reality is "infinitely smooth" or "infintely rough", hence my
statement about "first-order approximation".

-- 
Gary R. Van Sickle