Bug#510415: tech-ctte: Qmail inclusion (or not) in Debian
Steve Langasek <vorlon@debian.org> writes:
> On Fri, Jan 09, 2009 at 09:33:50AM +0100, Raphael Hertzog wrote:
>> We all know how NEW processing regularly result in complaints.
Telling people "no" results in complaints. Unless we're saying that we
shouldn't ever tell people no on accepting packages, I don't think this is
an argument for moving this function from NEW processing to somewhere
else. It just moves the complaints.
>> Trying to enhance the policy to be more fair could help.
It's going to be difficult to be fair across all packages in the archive
just because the ftp team will miss things. But when they do notice
things, I think they should do something about them. I'd rather not
sacrifice what review they have time to do because we can't do it for all
packages.
If the wrong criteria are being applied, that's a separate problem, but
out of the overall volume of packages and rejects, I don't think that's a
serious problem. There is the occasional package where the maintainer
gets upset, but it's fairly rare.
>> IMO the quality issues that are not covered by an explicit policy point
>> should result in bugs being filed and not in package rejects.
>
> I generally agree with this, except that I think the ftp team should
> have the discretion to reject packages for bugs that qualify as grave or
> critical.
I'm with Steve on this. I think the ftp team review is valuable, and as a
project it takes us much more effort to deal with critically buggy
packages after they're in the archive than before they get there.
All of the teams who have to deal with critically buggy packages once
they're in the archive are chronically understaffed. If the packages
aren't accepted in the first place, fixing the bugginess is the problem of
the person who wants to upload them. After they're accepted into the
archive, in practice dealing with it often becomes the problem of a lot of
other people who have other critical tasks. Overall, I think an ftp team
reject is a fairly good tradeoff unless they're frequently wrong, and I
haven't gotten the impression that they are.
--
Russ Allbery (rra@debian.org) <http://www.eyrie.org/~eagle/>
Reply to: