Re: # in place of a digit.

[halstead@crl.dec.com]

I don't recall ever being present for a discussion of the intended semantics
of # in usages like 18#, so I hope that if somebody was involved in such
discussions, they will step forward and add some wisdom from the "founding
fathers."  Also, like Matthias, I've never seen a person actually use # in
this fashion.  But, after seeing the previous string of postings and
considering the issue a bit, it seems to me that the most logical
interpretation of 18# is that it should represent a number that, if rounded to
two significant digits, would yield 180.  That means that the number should be
in the range 175 <= n <= 185.  (I'll let other people argue about whether
either of the <= signs should be just <.)

It's true that the visual representation 18# strongly suggests an
interpretation of 18<arbitrary digit>, suggesting a range of 180 <= n < 190,
but I find the analogy with usual interpretations of floating-point numbers
more compelling: the inexact number 1.8 would be expected to be between 1.75
and 1.85, and why shouldn't 18# be a good way to write the result of
(* 1.8 100)?							-Bert