[ODE] friction + collision problem

[Megan Fox]

What step size are you using?  I ended up with fairly bad instability if I tried to go smaller than about 0.01s.

-Megan Fox

> Making more tests, I've seen that using dWorldStepFast1 instead of 
> dWorldStep, everything works perfectly! (I use 10 iterations).
> But anyway, I'd like to know the reason of that strange behavior. If it 
> is a bug and I can help in anything to solve it, I'll be more than 
> willing ;)
> 
> thanks
> 
> 
> > Hi, I've already sent this question in another message, but I resend 
> > it with a proper subject ;)
> >
> > My problem is the following one:  If I set friction "mu = dInfinity", 
> > everything works fine. But if I set friction = 0 or 1 (or any other 
> > small constant), two of the boxes of my simulation enter one inside of 
> > the other (while ODE gives messages: "LCP internal error, s <= 0 
> > (s=0.0000)"), and suddently they collide with such a strength (since 
> > they are one inside the other), that both boxes run away of my 
> > simuation field.
> >
> > more details of my simulation: there are 3 boxes, two of them attached 
> > with a "ball" joint, and there is a flat ground (I allow a maximum of 
> > 4 collision points among each pair of objects)
> >
> > Should I always set friction = dInfinity? I'd like to have surfaces 
> > with lower frictions... do I have to do something special to support 
> > them?
> >
> > I've been searching in the archives of this mailing list, and I've 
> > found 2 other people experiencing the same problem than me. However, 
> > they got no answer... :(
> >
> > thanks in advance!
> > _______________________________________________
> > ODE mailing list
> > ODE at q12.org
> > http://q12.org/mailman/listinfo/ode
> >
> 
> _______________________________________________
> ODE mailing list
> ODE at q12.org
> http://q12.org/mailman/listinfo/ode
> 
>