[ODE] Particle/cloth/skeleton (was: Re: Choleski factorization)

[Joakim Eriksson]

> >This is not true. The constraint system was activated as quickly as
> >any force (Bomb/bullet) hit the character so there was some large
> >forces involved sometimes. I was at GDC and heard his talk (He's
> >a very funny and good speaker by the way) and there he showed a
> >level with a bunch of characters and then he set off a big bomb.
> >In the middle and they all flew away. Looked fun.
> i wasnt at GDC so my knoweldge is limited to what i have seen 
> in the game.
> I havent fully played it but i havent noticed any explosions 
> affecting the bodies.
> Probably i had a too superficial look at the game.

Can't say I have seen that many explosions in the game either
but that could be because when you blow up a bomb in the game
your usualy a long way from the blast at that point. (Car bombs
are a good example)
But something you will have seen a lot is how bodies react to bullets.
Thats one of the nice things with that system. He would just run
the key frame animations as usual then feed the current and
old positions into the particle skeleton system then apply the 
bullet forceand the body would react correctly carrying the
correct momentum that it got from the keyframe animations.
That would be a bit harder to do using a normal velocity/force
based approach.

> >In the paper he states that he models rigidbodies using a 
> tetrahedron.
> >Its hardly not used in the games and I realy dont know why. 
> But in one
> >place I remember they had a trashcan that you could tip over and
> >shoot at. It was surprisingly fun =)
> yes but that's far from stacking etc...
> i dont remember the details of the paper but i think he described the
> way the method could be extended to generic rigid bodies without
> saying that they actually implemented a system like that.

Not sure I follow you. What he did in hitman was "generic" rigidbodies.
You can stack them all you want and do everything you can with
normal rigid bodies. I'm not sure how stable the stacking would
be but I seriously think it would be stable enough for most
(game) uses. 

A constraint system built using that tetrahdron model is very stable.
I did a test about a year ago that involved 256 bodies and 512 
body-body constraints and it ran at 20Hz on my P3 866.
However they all had the same mass and the only constraint modeled
was body-body. So there wasnt any collision constraints and those
are the hard to do in that system.

> >I have been searching for this paper but I cant seem to find it
> >anywhere. Do you known where I could get a hold of it?
> i emailed directly the author, in order to save him same time 
> i will email it to you.

Thanks!

/Joakim E. - http://www.planestate.net