Iterative solver (again) [Was: Re: [ODE] Russ' plans for ODE?]

[Alen Ladavac]

Hi Nguyen,

Granted, there is a lot to what you say, but I still have some doubts...

>         After playing with Novodex, I found out that they use a
>         iterative solver like StepFast1(old story, every one know).

Is that what you concluded by looking at the Novodex examples and SDK, or
did they actually tell you that?

>         Of course, Novodex's solver is much more well designed, they
>         employ many optimizations like : Auto disable (But maybe much
>         more better than ODE's), contacts filter and a very robust
>         collision code.

I believe that. But I wouldn't jump to conclusions that it is enough. Could
it be that the "better designed iterative solver" is somewhat similar to
what was hinted by Richard Tonge a few years ago on this list? Our
experiments with successive approximation lead me to believe that it has
great potentials.

> So, out tiny step will be 1/600!!! I think it's too small to
> make "logical" and useful simulation.

Timestep in a simulator can never be too small. :) Smaller step will always
be more precise - unless, of course, you hit float precision in the time
variable, but we're not talking about that... ;)

Just my 2 cents,
Alen