General question about Python design goals

Paul Rubin http
Mon Nov 28 06:43:44 CET 2005

Jean-Paul Calderone <exarkun at> writes:
> >I can't think of a single use case for the addition (+) operator
> >working where either of the operands happens to be the number
> >0x15f1ef02d9f0c2297e37d44236d8e8ddde4a34c96a8200561de00492cb94b82 (a
> >random number I just got out of /dev/urandom).
> If you seriously believe what you just wrote, you have failed to
> understand the phrase "use case" (and possibly a lot of other
> things related to programming ;)

Heh, you must not remember the famous Pentium FDIV bug, where the
Pentium gave incorrect results for floating point division with
certain rare combinations of operands.  Intel at first refused to
acknowledge that the bug was a real problem (although it had already
been found and quietly fixed in later steppings), then refused to
replace people CPU's unless they could explain their use case where
the error could cause them a practical problem.  My school's biology
lab got its Pentiums exchanged by claiming it was using them to model
some kind of experimental drug treatment (I don't know whether the
claim was true) but other people just got told: sorry, but what you're
doing doesn't need correct FDIV results.  Intel of course eventually
had to back down and exchange everyone's cpu's after a huge public

> However (fortunately for you) I suspect you don't.  If you really
> did, you may want to pick up one of those platitude-filled XP books
> and give it a careful read.  You may find there's more there than
> you were previously aware.

I've read some of the XP stuff.  I see it mostly as a way of
explaining normal programmers' instincts to PHB's.  If the books did
some good in reducing PHB interference in some projects, that's nice;
however, while I felt that the general approach to programming running
through them was good, the specific practices and the type of
following they got bordered on cultish.

More information about the Python-list mailing list