should all floats be nonintegers?



>>>>> "RF" == Richard Fateman <fateman at eecs.berkeley.edu> writes:

RF> I suspect that "professional" numerical analysts would, by and
RF> large, say that floats represent exactly some particular number.

The mailing list for the IEEE's working group for an interval standard
agrees wholeheartedly with that assessment.

There have been a few posts suggesting that floats are intervals, all
where (politely!) flamed to oblivion with the argument that an approx-
imation very much != an interval.

-JimC (who had to be -- and was -- convinced)
-- 
James Cloos <cloos at jhcloos.com>         OpenPGP: 1024D/ED7DAEA6