[Python-Dev] Preview of 3.4 rc2 (so far) is up

Nick Coghlan ncoghlan at gmail.com
Wed Feb 19 12:47:30 CET 2014


On 19 February 2014 20:09, Georg Brandl <g.brandl at gmx.net> wrote:
> Am 19.02.2014 11:04, schrieb Nick Coghlan:
>>
>> On 19 Feb 2014 14:05, "Larry Hastings" <larry at hastings.org
>> <mailto:larry at hastings.org>> wrote:
>>>
>>> The URL has changed slightly.  Please go here:
>>>>
>>>> http://midwinter.com/~larry/3.4.status/
>>>
>>> You'll notice two things:
>>> a "merge.status.html" file, which shows you the list of revisions that I've
>> cherry-picked after rc1.
>>> a tarball containing the resulting source tree.
>>> As I cherry-pick more revisions, I'll add new tarballs and update the merge
>> status.
>>>
>>>
>>> For the record, I've passed over only two requested cherry-pick revisions so far:
>>>>
>>>> http://bugs.python.org/issue20646
>>>> select and kqueue round the timeout aways from zero
>>>>
>>>> http://bugs.python.org/issue20679
>>>> improve Enum subclass behavior
>>>
>>> I haven't rejected them, I just want more review.  If you'd like to see these
>> changes get cherry-picked for 3.4.0 rc2 (and final) please review them or
>> convince someone else to contribute a review.
>>>
>>>
>>> Only thirty cherry-picked revisions so far.  Gosh, you're making my life easy,
>> guys,
>>
>> Larry, you announced your preferred release candidate management process too
>> late to rely on it entirely - you should still audit all the deferred blockers
>> and release blockers flagged for 3.4, and ask for an update on their status,
>> with a pointer to the archived python-dev post describing how to request that
>> the change be included in 3.4.0 rather than being left to 3.4.1. I know at least
>> I have been setting those on the assumption things would work the same as they
>> have in previous releases, since you hadn't said anything prior to rc1 about
>> doing things differently.
>
> To be fair this isn't really different from 3.3.0, just that I didn't require a
> specific format for issues and went through all changes manually.

That's the part that worries me - if Larry is assuming the post to
python-dev is enough to get people to change their behaviour at short
notice, I'm concerned things that should be release blockers won't end
up blocking doing so.

Cheers,
Nick.

-- 
Nick Coghlan   |   ncoghlan at gmail.com   |   Brisbane, Australia


More information about the Python-Dev mailing list