[ODE] stepfast beta 2

[David Whittaker]

I believe this fixed them:

1) changed to dWorldID
2) commented #define TIMING
3) switched back to LCP solver... sadly, this means there's a bug in my
factoring algorithm somewhere :(

The test_buggy still likes to drive itself around a bit, but I think I
know why: it only has 3 wheels.  The back two wheels exert twice as much
force to resolve their constraints as the front one, and somehow this
gives a net force backwards.  Without rolling friction, even the tiniest
of forces will start the buggy rolling.  There are two ways I can think of
to fix it, full 6 DOF contact joints, or give the test_buggy 4 wheels
(i.e. test_crash).

Either way, the 3 fixes at the top are in the zip now so you can download
it again.  Maybe I'll take a stab at implementing rolling friction
directly in a contact joint one day this week.... whether I'll make it
very far or not, I don't know.

David

> David Whittaker wrote:
>  > I've just put a new version of the iterative solver at
>  > http://www.csworkbench.com/stepfast.zip .
>
> Yay!
>
> I've just taken a quick look.  Three problems found so far though:
>
> 1) Can't use the headers from C because dxWorld hasn't
> been defined for objects.h in the declaration of
> dWorld{G,S}etAutoEnableDepth()
>
> 2) Lots and lots of timing / statistics-printing is done
> even when -DdNODEBUG
>
> 3) test_buggy's buggy still falls through the floor pretty
> soon. :(
>
> On the plus side, disabled bodies no longer have gravity applied,
> and test_crash has started to work for me with this version.
>
> --
> Adam D. Moss   . ,,^^   adam@gimp.org   http://www.foxbox.org/   co:3
> "Half of the harm that is done in this world is due to people who
> want to feel important.  They don't mean to do harm.  But the harm does
> not interest them." -- T. S. Eliot
>
> _______________________________________________
> ODE mailing list
> ODE@q12.org
> http://q12.org/mailman/listinfo/ode