From: Douglas Gregor (doug.gregor_at_[hidden])
Date: 2007-07-27 18:47:23
On Jul 27, 2007, at 5:35 PM, Aaron Windsor wrote:
> One thing though - the GridPositionMap concept is really a refinement
> of the PositionMap concept, since the positions it maps to are
> guaranteed to be "grid points" - in this case, non-negative integers,
> whereas the PositionMap seems to map to real numbers. Or is this too
> trivial a distinction?
That's the kind of distinction I would make in the documentation of
the algorithm itself. "The values stored in the PositionMap must be
non-negative integers," or something like that.
- Doug