Re: [Algorithms] Kinematic Collision

[Pierre T. \(GMail\) <pie...@gm...>]

(for some reasons I don't receive my posts to the list, but they still seem
to go through. Weird.)

> It can happen for many reasons; our constraint solver runs for a fixed
> number of iterations, so it can terminate without fully resolving - stacks
> of objects will sink into each other over time, and the penetration vector
> is essential for resolving this. Also we have lots of motored joints like
> pistons etc which can force objects through each other.

...but you are talking about a physics engine here. I was only talking about
an old-style "character controller", a.k.a. collide-and-slide, à la Paul
Nettle / Quake / etc. Typical "modern" physics engines certainly need
penetration vectors. Character controllers, not necessarily - you can go a
long way without.

> If you code the collision routines carefully, you should be able to pull
> out the penetration info at the same time as finding the TOI...

Can you give more details here? I think you can find "a" penetration
vector -in the direction of motion-, yes, but it is vastly different from
"the" penetration vector, a.k.a. MTD, which can be in a completely different
direction. I don't think I know a way to compute both with the same piece of
code.

- Pierre