[ODE] Implementing a new joint

[Ruud van Gaal]

> From: Nick Winters <nwinters3000 at gmail.com>
> Subject: Re: [ODE] Implementing a new joint
...
> On 1/7/06, kurimail <kurimail at arrakis.es> wrote:
> > What about setting the rotation of the body to 0,0,0 every step?
...
> can you set the inertia of the body to something really 
> massive... to make rotating really difficult?

A couple of years back, I think someone managed to store (in ODE) the
inverse of the inertia matrix, so storing 1/inertia instead of just
inertia. That enables using 0 to define infinite inertia for axes. Not
sure whether there was ever a patch or that the code would be compatible
with today's Quick/Faststep algorithms...

Ruud