On Sat, Dec 22, 2012 at 2:58 PM, Mark H Weaver <[email protected]> wrote:
> Alex Shinn <[email protected]> writes: > > > On Sat, Dec 22, 2012 at 11:42 AM, John Cowan <[email protected]> > > wrote: > > > > > I think any slow-down is better than giving the wrong answer, > > > > That's what exact arithmetic is for. > > > > Exact arithmetic can run out of memory. > > So can your proposed inexacts. In order to avoid underflow and > overflow, the number of representable values cannot be finite, because > there can be no maximum or minimum representable magnitude. Therefore > the amount of memory needed to represent your numbers is unbounded. No > matter how clever your compression method is, that fact is unavoidable. > It's not a compression technique, and the amount of memory is in practice bounded by the limitations of computation. -- Alex
_______________________________________________ Scheme-reports mailing list [email protected] http://lists.scheme-reports.org/cgi-bin/mailman/listinfo/scheme-reports
