[ODE] Iterative solution

[Gary R. Van Sickle]

> >Gary R. Van Sickle said:
> >
> >I don't see why you have to or even want to keep the island idea.  Well I
> >suppose for collision detection it's probably still useful, but for doing
> the
> >solving, each constraint is treated as its own "island" by definition
> (i.e.,
> >you're dealing with it in isolation, without regarding any other
> constraints).
>
> Islands are still important because each island can converge in its own
> number of steps. Sometimes an iterative solver can solve to an acceptable
> error threshold in as little as a single iteration. The very next time step,
> that same system could take many, many more iterations. There is no reason
> to bog down all constraints with extra work, when only a few are typically
> trouble makers.

Aha, good point.  BUT: instead of relying on islands, could each constraint be
provisionally removed from the loop when it meets the error criteria?  Then when
the entire system looks like it's solved, do one last run through all
provisionally-solved constraints just to make sure they're still solved?  It
seems to me that something like this would have better granularity.

--
Gary R. Van Sickle
Brewer.  Patriot.