Re: task & skills
On Sun, Dec 03, 2000 at 10:47:56PM +0100, Christian Kurz wrote:
> If you are not interest in packaging anything, when then do you want to
> apply for new maintainer? (I'm just curious to know).
He wants to help with other things, I thought that was obvious from the rest
of his message.
> > There is a lot more about debian then making packages. There is
> > documentation, internationalization and debian internal projects such as
> > dpkg, apt, the installer, boot floppies and others. These take more than
> > a single person to code up and a list ends up as the maintainer of the
> > resulting packages.
>
> Right, but most of this stuff is already done by people that maintain
> packages in debian for quite some time now.
There's no reason he can't help by doing the little things while the
experienced people do the big things.
Take a look at Advogato's certification interview page (there's a link to it
somewhere at http://www.advogato.org/). It clearly describes several levels
of someone's participation in a project: Lead Developer, Developer,
Contributor, Helper. (Admittedly, you can't apply the term "Developer" to
some/most people we know as "Debian Developers", but that's a minor issue. ;)
> For helping to improve the webpage, there's as far as I know, no need to
> have a debian.org-Account.
Yeah but having an account has a few benefits over not having any access --
you needn't bother with sending patches to a mailing list, you can commit to
CVS right away and not waste time of other people; and there's benefit over
having pserver access -- you can use the more secure ssh method.
> For PR it would be good to have one, but I don't know if the PR-Team
> currently needs help or not.
>From what I can tell, proof-readers and people who collect press information
are needed. But yeah, an account isn't strictly necessary for that, it's
just convenient.
> And the server maintaince is done by the debian admin team, which consists
> of mostly long-time developers and not new maintainers.
I agree that the very fresh maintainers shouldn't be allowed to jump in such
groups unless we have really solid proof that they're strictly needed there,
that they have the required ability, and that their intentions aren't
malicious. (Though I think Neal mentioned this just as an additional
possibility, a corner case.)
> > The thing that bothers me the _most_ is that all of the tasks that I
> > have listed take _way_ more effort than maintaining some package of
> > the packages in our archive: many are updated a few times a year, if
> > that.
>
> Well, but you forget to see the for a part of these tasks you need to be
> technical skilled and if someone is technical skilled he should have no
> problem with creating debian packages.
But not all people are skilled or even should be skilled to make packages,
simply because there are varying levels of technical competence. I don't
think someone should be forced to package anything because they want to
participate in enhancing Debian in some other way. That will just cause
those people to fetch a random crappy app from freshmeat, run dh_make on it,
get it fairly lintian clean so to pass FTP admin check, and forget about it!
Let me elaborate on the essence of this issue, on an example. Everything
is IMHO, of course, and not aimed at you personally, Christian.
("Uh-oh, he's gonna rant" :)
Some people simply don't create programs, they write documentation for them.
Aside from knowing all they need about the thing they are documenting, they
know how to use the language, they know how to explain things, they know how
to intrigue the reader, and other things that make a piece of text
worthwhile to read. (I try to have such qualities sometimes, too :)
Some people, on the other hand, create programs. They don't provide any docs
for it, perhaps maybe a quick and dirty README to get you started, and from
then on, you're on your own. Use the source, Luke! And often, the source
doesn't even have comments in it. It happens because the coder either
doesn't want to document it, or doesn't know how. Sometimes (often?) both.
Documentation, while coders might despise it, is an important part of any
program that is to be considered complete. Are we going to ask the people
who don't want or know how, to do it? Or are we going to ask the people who
want and know how, to do it? (these were rhetorical questions, no need to
answer ;o)
Anyway. My opinion is that we don't need to give such people access to the
machines, or to the FTP archive, nor do we need to add them to the same
keyring, if that's what someone might object to. But we should officially
acknowledge their participation _somehow_: calling them Debian Developers
and giving them a nice @debian.* e-mail address (just an alias, even!)
sounds good for a start.
Take a real example: people who have translated the New Maintainers' Guide
(for those who don't know what it is, it's in maint-guide* packages). They
went over about 64KB of SGML which was written in English and created a
translation to their own language. Many prospective developers who speak
German, Spanish, French, Japanese, and Russian will benefit greatly from
having such a document. These translators also keep up-to-date with all the
changes in the English document. Many of them didn't only translate this
document, they've helped translate some other things.
OTOH you have people who package a couple of smaller programs and maintain
them. The programs are aimed at some smaller audience, so the maintainers
get a low amount of user feedback, and often less bug reports. Their
packages are useful to the people who use them, of course.
It is quite clear that the amount of effort invested by those maintainers
isn't particularly larger than the amount of effort that those translators
made. The amount of users these two types of people help is also similar.
I think work of both groups should be recognized appropriately, and that
neither of them should be given more attention or privileges just because
the group size (we all know that there are more than 550 maintainers in
Debian, and not nearly as many people who do other things).
--
Digital Electronic Being Intended for Assassination and Nullification
Reply to: