[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Improving our response to "duplicate" packages in Debian

On Thu, Jun 28, 2012 at 04:42:10PM +0200, Guus Sliepen wrote:
> I believe our current way of responding to ITPs for software that
> duplicates the functionality other software that is already in Debian
> is wrong. We have a very lengthy discussion everytime such an ITP
> happen, but usually they change nothing (the submitter just goes ahead
> with packaging it) or worse, they scare away the maintainer along with
> his/her ITP.

Hi Guus, thanks a lot for this mail of yours, which I find very

> So, keep the friction low for maintainers who are actually doing
> something, and if you really feel strongly about duplicate software
> polluting Debian, concentrate your efforts at the existing packages.

I believe you hit the nail on the head for the "ITP threads" problem.

Complaining *just* because there are similar programs in the archive is
demotivating for the ITP-ers who are simply trying to contribute to
Debian in a way they are comfortable with. That should be avoided. Also
because, as others mentioned, the dichotomy between working on "leaf"
packages and working on "core" teams is not necessarily true. In fact, I
know many many people currently active in core teams who started
contributing from leaf packages, and then gradually increased their
Debian involvement. To be honest, I hardly imagine how it could be
otherwise. If those experiences are representative of more general
trends, not bashing ITP-ers of leaf packages is actually an investment
in the future of the Project (and core teams).

But then it's true that some kinds of leaf packages induce archive-wide
costs that should be monitored. For instance forks, which you mentioned,
induce maintenance costs on the security team which are very similar to
those induced by code copies that we try to avoid in the archive.

It's a trade off then, and it seems to me that the guidelines you
propose are a good compromise. They ask not to complain gratuitously,
but rather detail reasons for doing so, putting some research burden on
the complainers. That is good, more productive and less socially awkward
than the more easy option of just asking "do we really need this in the

Guus, after having collected feedback from this thread, could you please
propose a patch to devref for documenting the outcome of this


PS as "related work" on this topic, I also vaguely remember a post by
   Joey Hess discussing the drawbacks of -devel culture of tearing apart
   ITPs. I can't seem to find it right now, anyone else has a pointer to
Stefano Zacchiroli     zack@{upsilon.cc,pps.jussieu.fr,debian.org} . o .
Maître de conférences   ......   http://upsilon.cc/zack   ......   . . o
Debian Project Leader    .......   @zack on identi.ca   .......    o o o
« the first rule of tautology club is the first rule of tautology club »

Attachment: signature.asc
Description: Digital signature

Reply to: