[Python-Dev] PEP 3144: IP Address Manipulation Library for the Python Standard Library

["Martin v. Löwis"]

> ipaddr looks like a reasonable library from here, but AFAIK it's not
> widely used outside of google.  I don't know if it's reasonable to
> want some amount public usage before a brand-new API goes into the
> standard library, but such use is more likely to uncover API flaws or
> quirks than a PEP.

OTOH, the PEP process *is* the stronger of the two approaches, allowing
people to provide explicit opinions even if (and especially if) they
dislike the technology entirely (whereas for an external module, they
would just ignore it).

If they refuse the comment, they can't complain when it gets added
to the standard library - they can still chose to ignore it, then,
of course (just as many people ignore xml.dom).

In the specific case, I'm not worried about timing. Either 2.7 or 3.2
are still a year ahead, which should give people plenty of time to
experiment.

OTTH, I *like* people to comment strongly on the PEP, in particular
if they are authors of competing libraries. It's no surprise that they
get emotional when their hard work won't be appropriately honored in
the long run - and if they believe there is something wrong with the
technology being proposed (rather than just the words used to describe
it), they are probably right.

I said it before - this is not going to be a fast acceptance path
of a library that gets accepted just because GvR works at google.
People of competing libraries *could* write competing PEPs if they
wanted to see their library incorporated instead - or they can just
state that they don't want *this* library to be incorporated for
specific technical reasons.

Regards,
Martin