contributions

[Scott Kitterman]

On Wednesday 14 May 2008 06:13, Morten Kjeldgaard wrote:
> It seems the main concern of many of the posters in this thread is
> that you may have a package you care about and would like to
> maintain, and you do not want a random contributor grabbing it in
> front of your nose.
>
> I am a big believer in letting computers solve as many problems as
> possible. That way we humans don't need to argue needlessly :-)
>
> The concern above can be solved if people subscribe to bugmail for a
> specific source package that they want to claim. If that involves
> several people, they will have to work it out among themselves. This
> can be detected by software (i.e. MoM or other automated procedures)
> and a note could be given that a given package is claimed and should
> not be touched unless otherwise agreed.
>
> My guess is that the number of claimed packages is rather small, and
> that in most cases, the last merger will be happy that someone else
> carries out the next merge.
>
> Cheers,
> Morten
>
> PS: At the moment, there is collective maintainership of all packages
> in Universe. Does this discussion in reality stem from a wish that
> Ubuntu maintainership of some packages should be possible? If so,
> that question should be dealt with politically.

I do not want to maintain packages on a dedicated basis.  Packages that I'm 
that interested in I already maintain in Debian.  It has previously been said 
by others that if the Debian maintainer of a package is someone who is active 
in Ubuntu, you really should (not must) check with them before changing a 
package.  I think this is true and I hope not subject to a lot of dispute.

Another example of the situation I'm talking about is clamav.  Currently we 
sync clamav from Debian.  For Dapper through Gutsy this was not the case.  I 
am not and do not care to be the sole maintainer of clamav, but over the 
course of a two release cycles I worked on the package and with the Debian 
maintainer and got it to a sync.  If someone else had jumped in and made 
changes to the package without coordinating with me, it would have set that 
process back.  

Maybe something would be urgent and need doing regardless.  I'm OK with that, 
but for the normal case just merging without regard for my previous work on 
the package would have made things more difficult.

I don't think we need more of a rule than people should look and see if one 
person has made a substantial contribution to a package (it'll be in 
debian/changelog) and if they have, defer to them on the package unless it's 
urgent and they aren't available.  

Scott K