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

Re: What about a user-contrib directory?



> On Fri, 17 Oct 1997, Enrique Zanardi wrote:
> 
> > > > But _all_ of our software is user-contributed!
> > > 
> > > Yes, but the process is a bit more than just making a deb and
> > > uploading it.
> > 
> > Can you ellaborate a little bit on that, please?
> 
> Yes.  You must send a message asking to become a developer, find some way
> of authenticating yourself, and then wait for an account for you to be
> created on master.  Many people could probably care less about having
> another account...  And also, most people will feel obligated to follow
> the mailing lists, participate in discussions, etc.  Personally, I like
> to, but it definitely increases the 'pain in the rear' factor.  All in
> all, a time consuming process that many would rather avoid.

I think we need to consider wheter we actually need to make it
easier to upload/create packages. As far as I know, Debian is
at this moment by far the biggest distribution, with the most
packages. If it really were the case that there are many
pieces of software left undebianised, that desperatly need
debianisation, then, yes maybe we should be thinking
about making it easier to upload .deb's.

But I feel that at the moment, the biggest problem is with packages
not following guidelines, or packages still beiing libc5,
or other minor/major flaws in them. We've already got the biggest
distribution. We want it to be the most bugfree distribution too.

By simply providing the users with many more .debs, created by people
that didn't want to follow debian-devel, and generally don't
care too much about debian, I think/fear the debian users may well
helped less, rather than more.

Note, that all .debs created 1.5 year ago, now most probably have
serious bugs in them (they are aout, "old source format", who knows,
can dpkg-1.4.19 handle them at all?). If we now make it easy for
people to upload, without asking some (minor) commitment from them,
we may well have to disappoint debian-users in 1.5 years from now.
And that is comming soon.

I'd guess that most packages that would be uploaded by non-maintainers
are packages that eighter
  a) Provide similar functionality as a package already in debian,
    but possibly different interface/whatever.
  b) Are usefull for a (very) small portion of debian-users.
    (otherwise, they would attract a real maintainer).
If in 1.5 years time the package would stop working, the users of
the "a)" type packages certainly would dislike ever choosing that
contributed package: they would have prefered it if the contrib
package never existed, then they never would have to change.
For the "b)" type packages,  you could say "well, at least the past
1.5 years were fun". But again, to be honest, I'd say we would have
provided a better service to the user if we would just had told them:
if you really really need this package, maybe you should think about
using a different distribution/OS. We cannot guarantee this package.
You know, having to switch distributions/OS, just because some
(for you) essential package suddenly stops being maintained is not
very nice.

Note also, that when the "core" distribution is smaller (as with
redhat, I believe), the need for "contrib" packages grows, and the
"contrib" packages themselves become more important. So, the 
"contrib" section from smaller distrbs. may well be better 
"maintained" by the contributers, as the packages in them are
more important than the contrib packages that are now proposed 
to be added to debian (we simply already have tho bigger/more 
important packages in our main system).


-- 
joost witteveen, joostje@debian.org
#!/usr/bin/perl -sp0777i<X+d*lMLa^*lN%0]dsXx++lMlN/dsM0<j]dsj
$/=unpack('H*',$_);$_=`echo 16dio\U$k"SK$/SM$n\EsN0p[lN*1
lK[d2%Sa2/d0$^Ixp"|dc`;s/\W//g;$_=pack('H*',/((..)*)$/)
#what's this? see http://www.dcs.ex.ac.uk/~aba/rsa/


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to
debian-devel-request@lists.debian.org . 
Trouble?  e-mail to templin@bucknell.edu .


Reply to: