Roland Plüss wrote: > d - e1 - e2. The result is then dynamic with the position and still more > or less accurate. > > Your proposal fails when the objects undergo rotation during the step (unless they both are spheres). Cheers, / h+