Walking avatars --> collision detection?

[Eunice]

This is called 'the bounding volume' method in collision dection.

Originally posted by Veleno
It would seem to me that it could be possible to have each avatar (and the viewpoint) surrounded with invisible rectangular prisms (or octagonal for a more "round" fit) that follow them wherever they go.

If avatar A gets to close to avatar B, then avatar A will be stopped by the prism surrounding avatar B.