[Python-Dev] RE: Call for reviewer!

[Tim Peters]

[David Goodger]
> I thought the "backwards compatibility" issue might be a sticking point ;>
> And I *can* see why.
>
> So, if I were to rework the patch to remove the incompatibility, would it
> fly or still be shot down?

I'm afraid "shot down" is the answer, but it's no reflection on the quality
of your work.  Guido simply doesn't want any enhancements of any kind to
getopt to be distributed in the standard library.  He made that very clear
in a conference call with the PythonLabs team, and as the interim 2.0
release manager du jour I pass that on in his absence.

This wasn't a surprise to me, as there's a very long history of rejected
getopt patches.  There are certainly users who *want* fancier getopt
capabilities!  The problem in making them happy is threefold:  (1) most
users don't (as the lack of positive response in this thread on Python-Dev
confirms); (2) users who do want them seem unable to agree on how they
should work (witness the bifurcation in your own patch set); and, (3) Guido
actively wants to keep the core getopt simple in the absence of both demand
for, and consensus on, more than it offers already.

This needn't mean your work is dead.  It will find users if it you make it
available on the web, and even in the core Andrew Kuchling pointed out that
the Distutils folks are keen to have a fancier getopt for their own
purposes:

[Andrew]
> Note that there's Lib/distutils/fancy_getopt.py.  The docstring reads:
>
> Wrapper around the standard getopt module that provides the following
> additional features:
>  * short and long options are tied together
>  * options have help strings, so fancy_getopt could potentially
>    create a complete usage summary
>  * options set attributes of a passed-in object

So you might want to talk to Gred Ward about that too (Greg is the Distuils
Dood).

[back to David]
> ...
> BUT WAIT, THERE'S MORE! As part of the deal, you get a free
> test_getopt.py regression test module! Act now; vote +1! (Actually,
> you'll get that no matter what you vote. I'll remove the getoptdict-
> specific stuff and resubmit it if this patch is rejected.)

We don't have to ask Guido abut *that*:  a test module for getopt would be
accepted with extreme (yet intangible <wink>) gratitude.  Thank you!