Re: How to make Debian more attractive for users, was: Re: The number of popcon.debian.org-submissions is falling
Hello,
On 07/23/2010 02:03 AM, Jesús M. Navarro wrote:
> On Thursday 22 July 2010 23:51:10 Don Armstrong wrote:
> [...]
>> Testing's primary purpose is as a staging ground for the next release;
>> while it'd be nice to try to keep it working as a fully installable
>> version all of the time, progress to the next release is more
>> important than that.
>
> And that's exactly my point while such valuable people as Russ Allbery wants
> to challenge that notion.
Not necessarily challenge but extend it. We only get testing tested or
unstable unstabled when people are actually running those and are working
with these. And the testing should not start in testing but in unstable,
otherwise we'd not need unstable in the first place.
The paradox is that the more a package is away from essential, the fewer
accidental users it is likely to have, but this means that we need more users
to have problems with those packages identified. So, with our increase of packages,
we need more users to test those new uploads, i.e. more users working with
unstable.
I have not been around when unstable was designed. I do not want to
contradict that it probably was an acceptable concept to have it break
often 10 years ago. But it is not today IMVHO. And from my experience,
it does not break too often, indeed.
I very much like the idea of CUT, possibly somehow merged with snapshot.d.o.
I feel that the blends should help with their respective experiences for
some subsets of Debian. And we could think of asking popcon for the
number of eyeballs a software has seen for the package's acceptance.
I have not thought through what this might mean for the release
management or Don's implicit concerns that it may hamper the release
of the next stable version. Nothing overly obvious strikes me, though.
Steffen (from his stable unstable laptop)
Reply to: