[Python-Dev] PEP239 (Rational Numbers) Reference Implementation and new issues

Tim Peters tim.one@comcast.net
Wed, 02 Oct 2002 21:21:54 -0400

[Andrew Koenig]
> Much as I like APL, I'd rather use Scheme's numeric model.

> I've heard that before, but I've also heard criticism of Scheme's
> numeric model.  "It works in Scheme" doesn't give me the warm fuzzy
> feeling that it's been tried in real life.

We've been thru this before too <wink>, but it doesn't even work in
Scheme -- the Scheme std is too permissive in what it allows conforming
implementations to get away (rationals aren't required; unbounded ints
aren't required; ints *period* aren't required; while an "exact" flag is
required, it has no portable mandatory semantics outside the (also
undefined) range of numbers needed to index vectors; etc).  Real
number-crunchers have no use for it even in a full implementation, as it
doesn't have a way to force precision-vs-space tradeoffs without extending
the language.  There's a reason the NumPy folks never bug you for Scheme
features <wink>.