[ODE] Possible solution to unstable rotating bodies problem

[gl]

> I find that damping works well enough, and has the added benefit of
> modeling air friction (in a crude way). So, unless your game is in
> space, I'd suggest adding about 0.02 of damping first and see if it
> doesn't do it for you.

Right, but Tyler's solution works in all scenarios, including when you 
choose a dampening factor that doesn't catch everything.  Having ODE 
crash-proof on this issue would be a good idea IMO.
--
gl