| From: Thomas Bushnell BSG <xxxxxx@becket.net>
 | Date: Thu, 16 Jun 2005 08:27:30 -0700
 |
 | Aubrey Jaffer <xxxxxx@alum.mit.edu> writes:
 |
 | > Polar representation works because the polar coordinates of Real
 | > numbers are in one-to-one correspondence with the Real numbers.
 | > But one cannot casually add new inexact representations.
 |
 | I didn't say you could; I'm speaking about adding a new *exact*
 | representation.  You seem to keep missing that, despite my repeating
 | it now three times.

My apologies; I misinterpreted "this exactness can be easily preserved
with a straightforward implementation." to refer to both exacts and
inexacts.

This thread is the result of my neglecting to qualify (as specific to
inexacts) a comment about numerical representations.  R5RS facilitates
the incorporation of new varieties of exact numbers and their
representations.