[ODE] Possible solution to unstable rotating bodies problem

[Adam D. Moss]

I still think that some users are missing the point that
even if it worked reliably (it doesn't), damping isn't
an adequate workaround because sometimes you just don't want
damping (or, you don't want as much damping as would be
required to avoid the exploding).

Additionally, how do you tune your damping to
*ensure* rotation-explosion-avoidance?  In an open-ended
simulation you might observe that you've magic-const-tuned
the chances of this happening to an average of once in an
hour of simulation runtime, or four hours, or a day, with
only a small amount of undesirable damping.  Then the level
designed quietly raises the ACME anvil's mass by 10%...

--Adam
-- 
Adam D. Moss   -   adam at gimp.org