Re: sponsor rules
Hi,
Quoting Richard Braakman (dark@xs4all.nl):
> So, what do we do with developers who upload crap packages? So far
> we've glared at them until they stop or get better, and it seems to
> work.
> (This is why I think the Maintainer field should point at the sponsor,
> with the sponsoree's name in the changelog. But not many developers
> seem to agree with me on that.)
This sponsorship thing has been bothering me for a while now.
A sponsor is fully responsible for the sponsored packages, and for the work
the actual maintainer does on the sponsored packages.
A sponsor shouldn't recieve 'credit' for a sponsored package.
Should a sponsor be 'bothered' by bugreports on a package he doesn't
maintain in reality ?
We have a lot of lazy maintainers. I'm probably one of them.
What if an 'evil' maintainer (ab)uses a sponsor to upload 'modified'
packages in the scenario we now use ?
A 'real' maintainer agreed to the social contract, to the debian guidelines,
and is (if everything went 'the right way') 'checked' to be a trustworthy,
'fitting' developer. Sponsorship works around this (i don't sponsor packages
for this reason).
In my opinion¸ every package should either have a 'debian.org' address in
the maintainer: line, or a <packagename>@packages.debian.org address; if we
really need the sponsorship construction, then we should let the actual
maintainer and the 'official' maintainer in on every message concerning the
package, and have a mechanism to point him to the responsability he didn't
have, if something goes wrong. Ofcourse having the sponsoree's line in the
changelog is a great idea; this keeps track of who actually did the work.
If sponsorship is used, construct something that allows reports/messages to
<packagename>@packages.debian.org to go to both the real maintainer and
the sponsor.
Greets,
Robert
--
Linux Generation
encrypted mail preferred. finger rvdm@debian.org for my GnuPG/PGP key.
Nine out of ten men who preferred Camels have switched back to women.
Reply to: