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

Re: long term goals of debian membership



CLARIFICATION: My point was that more people will naturally mean that
abandonded packages will be picked up because of the Unix
philosophy of "scratching an itch".  Later in the thread the issue
of the "apprentcieship" came up, an idea I had championed some months
ago.  When I championed the idea, my point was that it was better than
nothing, as NM hadn't turned out anyone then.  Basically, I'm thinking of
a free-form apprenticeship, where a maintainer is strongly encouraged to
sign keys for acceptable bugfixes: no assignment, just people scratching
an itch.  Scratch a given number of itches, and you get to maintain a
package or two under sponsorship.  Do that successfully for a release
cycle, and you get DD status.  This can also be expanded to documentation,
ports, whatever.  There is one point that you should leave with: there
should be a single criterion, and the NM committee should only make sure
that everyone plays fair--NMs don't get credit for trivial bugfixes, DDs
don't renege on their obligation to sign keys, and the NM advances when
they make the grade.  No time limits (except the "one release cycle"), no
muss, no fuss.  In fact, the only requirement on this whole oddessy should
be a key to sign.


On Tue, 5 Dec 2000, Brian Mays wrote:

> galt@inconnu.isu.edu (John Galt) wrote:
> 
> > > > The thing you don't get is no matter how tough you make the
> > > > screening, circumstances change.  People get new jobs that require
> > > > them to sacrifice volunteer work for paid work, people lose
> > > > interest, any number of things.  ATM, they post a quick intent to
> > > > orphan and packages stay orphaned for three years in some cases
> > > > because there's nobody to adopt them.  I'd hope you're on -qa,
> > > > because every additional barrier to entry makes -qa's job that
> > > > much bigger.
> 
> Brian Mays <brian@debian.org> replied:
> 
> > > The problem is many in the screaming hordes at our gates trying to
> > > get in do not take the time to adopt these packages, work on them,
> > > love them, feed them, and fix them.  You do not need to have the
> > > label "Debian developer" to work on orphaned packages.  All you need
> > > to do is fix bugs.  I know, it's how I got my start.  IMHO, it is
> > > the perfect preparation for the life of a developer.
> 
> Daniel Stone <daniel@kabuki.eyep.net> wrote:
> 
> > What if they don't want them? For instance, what's the point of
> > someone whose machines are all console-only (I've been there) taking
> > up a window manager? What if you never use that package? Then the
> > quality goes _DOWN_.
> 
> Read what was said above.  My opponents in this debate are complaining
> that we need to lower the barriers to let more developers in so that
> they can adopt abandoned packages.  Yet you counter my arguments by
> saying that these potential developers don't want to adopt these
> packages.  For crying out loud!  Pick a position and stick with it.  You
> can't argue both sides.
> 
> > > When you have something significant to contribute, then we'll let
> > > you in.  If you don't have the patience to deal with that, then what
> > > kind of developer will you make when userland starts slinging bug
> > > reports your way?
> >
> > Now let's turn this back around straight at you using that argument I
> > just said. ... What if you don't care about the app, but you adopted
> > it because some tightarse changed the rules so you _had to_. L^HUser
> > bug reports come flooding in. You don't give a shit; you haven't used
> > the app in 4 years, and when you did (for .5 seconds) it seemed to
> > work OK ... *LART LART LART*
> 
> Nobody is saying that you *have* to adopt abandoned packages.  I'm
> saying that we are going to take you more seriously if you do.
> 
> Really, if all you want is a way for anyone off the street to upload
> their packages, Linux already has such a system.  It's called RedHat's
> contrib directory.  I would prefer, however, for Debian to have higher
> standards than that.  The most reasonable way to do this is to be
> careful in choosing those that we allow to contribute *directly* to our
> distribution.

Be careful of the *BSD "CVS Cabal".  Remeber that Theo De Raadt forked
OpenBSD because he was frozen out of CVS commit access for NetBSD.  BTW he
wouldn't have fit in my criterion for booting, he was extremely productive
when he got frozen out, he just had a big mouth (a guy after my own heart
:)--one that seems to have shut WRT bugtraq :P

> > > Don't forget that Debian has a reputation to uphold.  Whenever one
> > > of our developers answers a bug report (or more importantly, DOESN'T
> > > answer a bug report), it is a direct reflection on us as a group.
> > > Certainly, allowing someone who has done nothing for us and whom
> > > we don't even know to represent us is not a wise idea.  (We have
> > > enough trouble with the problems caused by our productive members.
> > > ... *grin*)
> >
> > Based on this, virtually no-one would ever get hired. What if people
> > start picking up on -devel and taking that as the official line? It's
> > just as plausible as if they started reading bug reports. If the
> > media/whoever want an official line, there's press releases and
> > official contacts. However, the likelihood of them reading _bug
> > reports_ (of all things), is _MUCH_ lower than them reading -devel,
> > where they'd see all these people who seem to be speaking for
> > Debian. You see?
> 
> What are you talking about?  Who is talking about an "official line"?
> All I'm saying is that I want to avoid, as much as possible, situations
> whereby users can complain:  "That Debian package was pretty bad, bug
> ridden and all.  I tried to contact the maintainer and I filed a bug
> report, but nobody responded."

Unfortunately this is the "official line": remember when Bruce Perens
pointed out on -legal about Corel's violation of the GPL?  Within a day it
was on slashdot as official Debian policy, despite the fact that he hadn't
been DPL for quite a while when he wrote that.  Basically, to the rest of
the world, what crops up on the mailinglists is policy regardless of who
says it.  We know differently (how many of you would say that >I< had much
say in the way things are run...), but we can't ignore the fact that it
comes across that way.

> - Brian
> 
> 
> 

-- 
Pardon me, but you have obviously mistaken me for someone who gives a
damn.
email galt@inconnu.isu.edu




Reply to: