Re: Shortening release cycles
I believe that this is the reason Joey Hess suggested that we start adding new
and tested packages as an addition to our last stable release. Many package
maintainers release new versions of their packages for both the stable, as well
as the unstable release. XFree is a perfect example of this, and there is no
reason why we can't add these updated packages designed for stable once testing
has been done on the updated package. This idea was debated on -devel for a few
weeks before disappearing. I'm not sure exactly what happened, but I suspect
that, with a few exceptions, people WANT to see this. It wouldn't add to the
load of the developers, since they are already doing the releases of new
packages, it wouldn't take away from the quality, since the packages would be
tested. It would be some additional work for the debian-cd and possibly floppy
images, so those who work on these two would need to be involved, but even then,
it shouldn't be a major change from the previous "stable". Does anyone see a
problem with this? I sure don't.
Dave Bristel
On Tue, 28 Sep 1999, Martin Schulze wrote:
> Date: Tue, 28 Sep 1999 15:12:31 +0000
> From: Martin Schulze <joey@carelia.infodrom.north.de>
> Reply-To: Martin Schulze <joey@infodrom.north.de>
> To: Debian Development <debian-devel@lists.debian.org>
> Subject: Re: Shortening release cycles
> Resent-Date: 28 Sep 1999 16:36:09 -0000
> Resent-From: debian-devel@lists.debian.org
> Resent-cc: recipient list not shown: ;
>
> Steve Greenland wrote:
> > I liked a lot of these ideas, but:
> >
> > On 12-Sep-99, 20:22 (CDT), Martin Schulze <joey@finlandia.Infodrom.North.DE> wrote:
> > > Our current situation results in our stable release being hopelessly
> > > out-dated and the unstable release not being releaseable. That's
> > > quite bad for a lot of our fellow users and developers.
> >
> > and then wrote:
> > > No major changes are allowed to go into the distribution after two
> > > months after the last release. (e.g. release on december 1st,
> > > major changes are only allowed to happen until february 1st).
> >
> >
> > If you release every six months, this leads to up to 10 months before
> > a new release of a package makes it into stable. Suppose Xfree86 4.0
> > is released on feb 10th. That's too late for the june 1 release (which
> > froze on feb 1) so it won't be released until dec 1. People will complain.
>
> It would be up to the release manager to get convinced about exceptions.
> I'm willing to appreciate them. However, what you miss is that we are
> worst than that, slink was frozen in Nov '98 or so, that's about *at least*
> 12 months before the next release (assuming that potato will be released
> in December this year).
>
> > (Not me, as I'm not one of those who believes anything more than a month
> > old is "hopelessly outdated".)
>
> Granted. Though, we still need testing which takes time, we also need
> to give other packages time to keep up (e.g. a major perl update requires
> to recompile some 50 packages or so, X is too fuzzy for me to make assumptions).
>
> Regards,
>
> Joey
>
> --
> The only stupid question is the unasked one.
>
> Please always Cc to me when replying to me on the lists.
>
>
> --
> To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
> with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
>
Reply to: