[Python-Dev] -Wall output for the Python 2.0c1

[Tim Peters]

[Tim]
> I'll take this one [wngs putatively due to DUP_TOPX].  It looks easy
> to fix ...

[Skip Montanaro]
> Seems to me like a patch that ought to wait until after 2.0final.  It's
> definitely not a show-stopping bug...

Would have been a more interesting argument had you posted this before the
patch was checked in <wink>.

But wngs are non-negotiable with me anyway.  This example turned out to be a
good one for showing why:  the bogus wngs covered up a legitimate
"uninitialized vrbl" complaint (see Fred's later discovery and following
checkin).  I've simply got zero tolerance for wngs or for failures in the
std test suite.  My view is warped by prior experience, though:  the last 6
years, I worked on projects where the compilers were fiddled to treat
warnings as fatal errors.  So while this may have looked like a harmless
batch of wngs to you, as far as I was concerned Python couldn't even be
compiled anymore <0.7 wink>.

extremism-in-defense-of-simple-best-practice-may-or-may-not-be-
    vice-but-it's-sure-effective-over-the-long-term-ly y'rs  - tim