[ODE] Contact point merging strategy

[Jon Watte (ODE)]

Yes, it's that "simple".

You may actually want to ignore the similarity of normals, or allow for 
a big slop, depending on your intended result; driving over a "flat" 
trimesh with a ball can be pretty bouncy unless you filter pretty 
aggressively based on contact proximity. You do want to average all the 
normals you filter out per geometric point, though.

In my experience, the simulation will run faster, and will not lose 
stability, unless you're too sloppy with the filtering (i e, merging two 
contacts that are 2 meters away from each other or something).

Cheers,

          / h+


Mark Williams wrote:
> Is there some sort of standard strategy for optimising contact points, 
> or can anybody comment on any attempts that they've made? Is it as 
> simple as finding all the contacts that are within a given tolerance 
> that have similar contact normals, then replacing them with one which is 
> the average? What implications does this have for solve speed/accuracy ?
>