SV: [ODE] Iterative solution

[Billy Zelsnack]

>> Imagine a stack of boxes sitting on the ground.  First you apply gravity,

>Russ Smith said:
>your argument make sense ... but i have been assuming that while an
>iterative soluting is busy iterating, only the system velocity would
>change, not the position as well. (the position would be updated after
>the new velocity had been finalized). i believe that if only the velocity
>is changed during the solution process then the update order makes no
>difference.

The velocities will still propagate, it will just take more iterations to
achieve the same quality of solution.

>can you (or someone) give a reason why position update is better? (i have
>not done any experiments on this yet).

I don't understand this either. It just seems that it would be a lot slower
because you have to do numIterations times more collision detection.

-billy