[Python-Dev] Python 1.6 timing

[David Ascher]

[I just got GvR's post on the topic, but I'll send this anyway]

BAW:
> All very cool things that could easily wait until 1.7.  After all,
> what's in a number?

Guido has promised some of those features as being in 1.6 at conferences in
the past, but I agree that string methods for example are a more major
change than I'd expect to see in a 0.0.3-delta version change.

Maybe with a deadline (as Greg suggests) we can integrate some of the
pending patches (I agree with Greg that I at least would have found the time
for a revised patch for rich comparisons if I'd had a deadline -- call me
human =).
Coercion and extended slicing also seem like relatively minor changes,
compared with changing everything to be a class or adding GC!

Regardless, just like Greg, I'd like to know what a pumpkin-holder would
mean in the Python world.

I propose that it be called the Oracle instead.  As in, whoever is Oracle
would get some training with Tim Peters and learn how to channel G__do.  As
a Python user, I'd be most comfortable with such a change if the Oracle just
took over the technical stuff (reviewing patches, CVS checkins, running
tests, corralling help for doc & code, maintaining release notes, building
installers, etc.), but that the important decisions (e.g. whether to add a
feature to the core language) would be checked with G__do first.  We could
call the position "Administrative Assistant", but somehow that doesn't have
the prestige.

A progressive schedule where Guido watches over the Oracle periodically
would probably help build trust in the new mechanism.  The Oracle would be
expected to ask Guido for his opinion with everything at the beginning, and
as a trust builds between Guido and the Oracle and the community and the
mechanism, progressively less.

--david ascher