[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#683323: unblock: python-apt/0.8.7



2012/9/3 Michael Vogt <mvo@debian.org>:
> On Mon, Aug 27, 2012 at 01:24:24PM +0200, Julian Andres Klode wrote:
>> On Mon, Aug 27, 2012 at 02:10:17PM +0300, Martin-Éric Racine wrote:
>> > ke, 2012-08-08 kello 01:50 +0200, Cyril Brulebois kirjoitti:
>> >
>> > > Julian Andres Klode <jak@debian.org> (30/07/2012):
>> > > > Since the version of testing, this contains mostly bug fixes and
>> > > > many translation updates, but also (starting with 0.8.5) one new
>> > > > module (apt.auth) which is a cleaned up version of an internal
>> > > > software-properties module (and not used by any code in unstable
>> > > > AFAIK).
>> > >
>> > > Then I'm not sure we need or want this new module in wheezy…
>> >
>> > Should this new module be rolled back and a new python-apt release
>> > produced just for Wheezy, then?
>>
>> How about adding a warning to the module that it is experimental? I'd
>> like to avoid rolling back stuff we already have in newer. Otherwise,
>> we could upload a 0.8.8 with a disabled module as well, but that might
>> create problems later on with merging Ubuntu packages that depend on
>> python-apt (>= 0.8.5) and expect apt.auth to be there.
>>
>> We could also create a 0.8.4.1 release for testing, by branching off
>> unstable's 0.8.7, merging the changelog entries for 0.8.5, 0.8.6,
>> and 0.8.7 into 0.8.4.1, and removing apt.auth mentions, and disabling
>> or removing that module. But that is a bit more work. We could also
>> name 0.8.4.1 as 0.8.4+wheezy1, but given that unstable is past 0.8.4
>> already, just using 0.8.4.X instead of 0.8.4+wheezyX seems a bit
>> nicer.
>
> I created a branch at
> http://anonscm.debian.org/bzr/bzr/apt/python-apt/debian-wheezy/ that
> is based on the debian-sid branch but disables auth.py. I am happy to
> upload this to wheezy if that approach is approved by the release
> team.

Release Team:

A separate branch was created for Wheezy and a debdiff submitted. Is
there anything else missing to move this issue forward and achieve a
concensus about what will be allowed into Testing?

Regards,
Martin-Éric


Reply to: