[ODE] contact joints / attaching them

[nlin at nlin.net]

On Sat, Oct 11, 2003 at 11:47:36AM +0100, gl wrote:
> Yeah, but the amount of penetration determines the force.

Not really. A penetration is an "error" in the rigid body configuration
(if bodies are rigid, they should not ever penetrate). Since ODE is a
discrete simulator, it makes these "errors."

The Error Reduction Parameter (ERP) reduces error by pushing bodies apart.
Also, the Constraint Force Mixing allows slight errors (violations of
constraints) to be tolerated.

> And the amount
> will be different with different timesteps - ie. the timestep is relevant to
> the collision results indirectly.

Only because of ERP and CFM.

See joint.cpp, function contactGetInfo2, and notice how the penetration
depth is not used for determining the resulting velocity. It is only
used for ERP and CFM.

-Norman