I am trying to understand just how the manifolds are working with compound shapes. Is there one 4-point maximum manifold shared amongst all child shapes, or are there N manifolds, each 4 points maximum? I am trying to establish performance impact of a compound given that each new manifold has the potential of adding more rows to the solver.
Any insight would be helpful.
S
How do the manifolds work for compound shapes?
-
- Posts: 456
- Joined: Tue Dec 25, 2007 1:06 pm
Re: How do the manifolds work for compound shapes?
The second answer is the right one as far as I can understand. Don't know much about performance, but I guess that using compound shapes is slower, as long as contacts keep getting created/removed inside more than one of their child shapes.sphet wrote: Is there one 4-point maximum manifold shared amongst all child shapes, or are there N manifolds, each 4 points maximum?