Re: Let's be friends. (Non-maintainer uploads made easier)
Manoj:
> > > Unless I am missing something, this proposal actually
> > > restricts the candidates for non-maintainer uploads; I also fail to
> > > see how it makes non-maintainer uploads any easier.
Kai:
> That's also my impression.
>
> > Yes, you are missing something: the "friends" uploads are considered
> > to be basically the same as maintainer uploads; the "friends" uploads
> > are *encouraged*. Maybe a better name of "friend" would
> > be "Vice-Maintainers", or, "Replacement-Maintainers", in that, when
> > the real maintainer is away, the "friends" take on the role of the
> > real maintainer.
>
> And just how does this make non-maintainer uploads easier?
>
> I really don't see the point. The maintainer can always tell other people
> to go ahead, make a NMU (we see this happen all the time).
That's precicely one of the points: If the maintainer is away, he cannot
tell other people to go ahead.
My proposal of Co-Maintainers, or Friends, or whatever should not make
it more difficult to do a "real" NMU. It's just that when the real
maintainer is busy, away, ill, whatever, there's somebody else who
acts as though he is the temporary maintainer, and who can
fix simple bugs, do upgrades to upstream source, etc.
Yes, we now already can do non-maintainer uploads. But if am
away for one week, nobody really has the guts to do a non-maintainer
upload for me now. Even though many packages I maintain, I would not
mind a bit at all for most packages if somebody else uploads a
non-maintainer release. So, I really want a way of saying "please,
everybody, if I don't answer my mail for more than 2 days, just go
ahead, and do a nmu".
OK, I can say that now. But actually, for some packages, I would like
to select a somewhat smaller group of people who I would *like* to
make NMU's, when I'm away. (like libg++).
Really, I see the "Co-Maintainer" system much more as a way to expres
that I am not the "one-and-only" person who is allowed, or encouraged
to make uploads of a certain package. For many packages, I'm just
"one of many" people who know that package well, and are quite well
able to make uploads. So, during a perioud when I'm busy, while some
other maintainer has time on his hands, I'd actually want to say:
you just go ahead, and do the work.
> Just what, exactly, will be different under the proposed scheme?
The *encouragement*. A "Co-Maintainer" will be encouraged to make
uploads whenever the real maintainer isn't there for some time, or is
busy. At the moment, if I vanish off the face of the net, the first
few weeks nobody notices. Then an email is sent to debian-devel
"where's joost witteveen". Again a few weeks later it will then be
decided that joost witteveen has indeed vanished off the net, and
that we probably should declare that all my packages "Orphaned".
During all those weeks before that decision, people will be weary
of making NMU's of my packages, as I may return.
Under the "Co-Maintainer" system, simple bugs get fixed immediately
when I don't answer my email for a few days, by one of my "Co-Maintainers".
> (You say that the current scheme is equivalent to Friends: nobody. Seems
> to me that it rather is equivalent to Friends: debian-keyring instead.)
No. I definately don't want to make it more dificult to do Non-Maintainer
uploads by Non-Co-Maintainers. A "Non-Co-Maintainer" upload should be
just as dificult as it is now.
--
joost witteveen, joostje@debian.org
My spamfilter is so good, it correctly catches 90% of incoming spam,
*including* all email from my PhD supervisor.
--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to
debian-devel-request@lists.debian.org . Trouble?
e-mail to templin@bucknell.edu .
Reply to: