[Python-3000] [Python-Dev] What should the focus for 2.6 be?

Josiah Carlson jcarlson at uci.edu
Thu Aug 24 04:21:22 CEST 2006

"Terry Reedy" <tjreedy at udel.edu> wrote:
> "Josiah Carlson" <jcarlson at uci.edu> wrote in message 
> news:20060823125951.1A60.JCARLSON at uci.edu...
> > The intent of my post was to say that all of us want Py3k to succeed,
> I should hope that we all do.
> > but I believe that in order for it to succeed that breakage from the 2.x
> > series should be gradual, in a similar way to how 2.x -> 2.x+1 breakage
> > has been gradual.
> Given that the rate of intentional breakage in the core language (including 
> builtins) has been very minimal, this would take a couple of decades, which 
> to my mind would be a failure.

If we could stick with a 12-18 month release schedule, using deprecation
and removal in subsequent releases, every removal could happen in 2-3
years. 2.6 could offer every feature of 3.0 (except for
backwards-incompatible syntax), warning of removal or relocation (in the
case of stdlib reorganization), 3.0 could handle all of the actual
syntax changes.

> > I believe we agree on this basic point
> To the contrary, you seem to have a basic disagreement with the plan to 
> make all the core language changes at once and to clear the decks of old 
> baggage so we can move forward with a learner language that is a bit easier 
> to learn and remember.

I disagree with the "all the changes at once", but if Guido didn't agree
with a gradual upgrade path, then the 2.6-2.9 series wouldn't even be
considered as options, and we'd be looking at 3.0 coming out after 2.5,
and there not being a 2.6 .  Since 2.6 is planned, and other 2.x
releases are at least possible (if not expected), then I must agree with
someone, as my desires haven't previously been sufficient to change
Python release expectations.

> > according to your talk and your posts here, you want Py3k alpha
> > in the next year or two, while I'm thinking that Py3k alpha should come
> > somewhere after 2.6 and probably 2.7, maybe even after 2.8 or 2.9,
> Whereas I wish it were already out and would be delighted to see it early 
> next year.  Some of the changes have already been put off for at least five 
> years and, to me, are overdue.

As a daily abuser of Python, I've not found the language to be lacking
in any area significant enough, or even having too many overlapping
features suffient to warrant such widespread language breakage.  We
disagree on this point, and that's fine, as long as Guido agrees that
2.6+ make sense, which he does, and states as much in his talk and all
relevant postings I've seen, then I don't need to drug him.

He also agrees that 3.0 should come out sooner rather than later, but
that's not going to stop me from attempting to make the case that 3.0 is
going to be generally unused until later gradual 2.6+ releases close the
gap and make the transition more natural.

But hey, I'm just a guy who writes software who is going to have to
transition and maintain it. Obviously there can't be too many of us, go
ahead and break the language, I'm sure everyone will be happy to
upgrade to 3.0, you won't even need to maintain the 2.x series, really.

 - Josiah

More information about the Python-3000 mailing list