Re: Multiarch support in dpkg — really in time for wheezy?
On Sun, Oct 30, 2011 at 3:09 AM, Raphael Hertzog wrote:
> On Sat, 29 Oct 2011, Michael Gilbert wrote:
>> You could also make a case from a terminological perspective as well.
>> Unstable is where development in Debian is supposed to happen, so it's
>> perfectly acceptable to upload unfinished/unstable changes.
> No, sorry. That's not how we are using unstable. We push there stuff
> that we believe to be mostly ready for the next stable release.
Yeah, and multiarch is intended for next stable, so it needs to get
there at some point, and the sooner the better; as stated even by
> Unfinished/unstable work is supposed to go to experimental. And in the
> case of dpkg doubly so because it affects everybody.
I did not say that every change to unstable needs to be
unfinished/unstable, but there comes a time when a maintainer has to
say enough is enough and take the risk and get it out there so
development elsewhere can continue. And again, if there are problems,
there are plenty of people ready to fix them. It's not solely
Guillem's responsibility if these patches do ultimately end up
I never said anything about it not making a stop-over in experimental anyway.
> Fortunately multiarch support is not in a poor state and I believe it to
> be of release quality (and it made a release in Ubuntu and did not
> generate a flurry of bug reports).
> Furthermore I don't think that guillem is worried of breaking unstable.
Then pushing the existing changeset should not be so excessively
dramatic. If he trusts the people that wrote it, and he doesn't have
sufficient time to review it now, then he can push it now and review
it later when he has time and fix it up to any degree of perfection he
Anyway, this response shouldn't have been necessary. Please
contemplate a little more about the writers' arguments before jumping
to incorrect conclusions about their point. t would immensely help
the quality of communications on Debian mailing lists.
Thanks and best wishes,