[Ack, I intended to reply to the list the first time, not in private.
Sorry, Steve.]
On Sun, 2002-03-03 at 18:54, Steve Greenland wrote:
> Quite frankly, with very few exceptions and for limited times (e.g.
> experimental new versions of packages that I use from developers I
> "know" and are soon to be in main), I personally don't have much
> interest in packages that are not in the Debian archive system. They
> don't get near as widely tested, there's no real expectation that they
> comply with Debian Policy (the real strength of Debian, in my opinion),
> they're not part of the Debian BTS, and the developer has made little
> or no commitment to maintain the packages.
This is IMO a self-sustaining problem. People don't use non-Debian
packages because they are of low quality. Upstream therefore make crappy
Debian packages with the expectation that few people will use them,
because the Debian packages are/will be better.
A prime example of this was the "LoserJabber" LiveJournal client. Its
maintainer, Christian Surchi, was otherwise occupied. In the meantime, a
bug was found in the client that made it completely useless and unable
to access the LiveJournal server. It was fixed upstream within a day.
The Debian package languished for a long time, and in the interim myself
and another LJ user provided packages.
I agree that policy is the greatest strength of Debian. But, remember,
policy is available to non-developers too. The solution is to pressure
non-DD packages into complying with policy, not by outright rejecting
any non-Debian packages, but by working with them and making them aware
of the problems. There is no reason that non-Debian packages cannot be
of high quality; there is no reason not to expect them to be high
quality.
By providing infrastructure to make it easy for users to use other APT
sources, we therefore
- Get wider testing, since more people can use the package.
- Get better packages, as there is more testing.
- Get more policy-compliant packages, as upstream has more motivation
to work on Debian packages, and more feedback about them.
I realize this is an incredibly simple and optimistic view of the
software development process. On the other hand, it's the one I've
observed as I work on Feta. There's no magic bullet to make non-Debian
packages not suck, but end-user tools designed to ease the management of
sources.list will go a long way, IMO.
The best way, of course, is the current practice of the maintainer
working very closely with upstream. But the nature of Debian makes
maintainers scale much more slowly than upstream authors, and the nature
of time prevents maintainers from working with many upstream authors.
--
- Joe Wreschnig <piman@sacredchao.net> - http://www.sacredchao.net
"What I did was justified because I had a policy of my own... It's
okay to be different, to not conform to society."
-- Chen Kenichi, Iron Chef Chinese
Attachment:
signature.asc
Description: This is a digitally signed message part