[Python-Dev] Mercurial migration readiness

[Jesse Noller]

On Tue, Jul 6, 2010 at 7:47 PM, anatoly techtonik <techtonik at gmail.com> wrote:
> On Fri, Jul 2, 2010 at 3:34 PM, Antoine Pitrou <solipsis at pitrou.net> wrote:
>>>
>>> > After the switch, hg.python.org/cpython will be the official repo, and
>>> > code.python.org/hg will probably be closed.
>>>
>>> Why this transition is not described in PEP?
>>
>> Because it's not a transition. It's a mirror. It was put in place
>> before the hg migration plan was accepted, IIRC.
>
> Where is this migration plan then if it is not in PEP?
>
>>> How code.python.org/hg is synchronized with Subversion?
>>
>> What does your question mean exactly? It's a mirror (well, a set of
>> mirrors) and is synchronized roughly every 5 minutes.
>
> Method. Software used, which parameters are set for it, how to repeat
> the process?
>
>>> Why it is not possible to leave code.python.org/hg as is in slave mode
>>> and then realtime replication is ready just switch master/slave over?
>>
>> The two sets of repositories use different conversion tools and rules.
>> They have nothing in common (different changeset IDs, different
>> metadata, different branch/clone layout).
>
> That would be nice to hear about in more detail. As I understand there
> is no place where it is described. I already see +1 from Fred Drake
> and another +1 from Steve Holden down the thread.
>
> However, Antoine Pitrou, Dirkjan Ochtman and Jesse Noller object. They
> afraid that contributors won't survive low-level details about
> Mercurial migration. I'd say there a plenty of ways isolate them and
> at the same time satisfy "Mercurial aficionados" either on the same
> page or in different places.

No, I don't need you misrepresenting anything I've said Anatoly - I
said there's no need to maintain SVN alongside mercurial after we
convert, and doing so is silly. I maintain that once we convert, we
very happily stay converted, and drop official "other" mirrors unless
other volunteers step up to maintain them.

I have no problem with additional documentation should people wish to
volunteer to write it.

We do not work for you Anatoly.

> On Fri, Jul 2, 2010 at 4:06 PM, Stephen J. Turnbull <stephen at xemacs.org> wrote:
>>
>> There is no reason at this point to suppose the transition can't be
>> complete by the end of summer.  However, as always, the devil is in
>> the details, and one of them may be a showstopper.  We'll just have to
>> see about that.
>
> The transition can be complete in a few minutes. The question is how
> good it will be. As there are no plan, no roadmap, no status - it is
> hard to judge if it is feasible at all.

No. There is no question except in your mind. We all have a rough idea
of the status, modulo the PEPs being updated. It is also perfectly
feasible. I would love it, and offer you a christmas card if you could
drop the hyperbole and misrepresentation.

>
> Ok. Given that nobody is able/willing to say anything more - I've
> gathered all your feedback concerning current status of Mercurial
> migration on this Wave -
> https://wave.google.com/wave/waveref/googlewave.com/w+4_fnAVHwA  I
> hope you will find the time to enhance it with more info so not
> contributors proficient with Mercurial could help to speed up the
> transition.

While the summary is nice; your wave entry has nothing to do with the
mercurial transition, if you want to help, please ask someone to take
on an open task, or volunteer to write/accentuate the PEPs, or help
with documentation for post-migration workflow. Your contributions can
be effective and useful, rather than noisemaking and abrasive.

The mercurial transition will occur, barring someone directly involved
finding show-stopping reasons otherwise, with or without you. The
decision was made some time ago, and despite your recent noisemaking,
will continue on.

jesse