Getting new packages into Ubuntu

Scott Kitterman ubuntu at kitterman.com
Mon Oct 10 15:50:17 UTC 2011


On Monday, October 10, 2011 04:42:44 PM Sebastien Bacher wrote:
> Le lundi 10 octobre 2011 à 10:23 -0400, Scott Kitterman a écrit :
> > If we try to reduce the work to the available developers by reducing
> > the scope
> > of the archive, then we are also reducing the pool of potentially
> > interested
> > developers as well.
> 
> There is still enough in the system components stack (or main) to be
> interesting, do you think a stack of unmaintained "small softwares" in
> universe is what makes people want to join the project? They could join
> the same way to help getting those in extras with the benefits that:
> - they wouldn't need to get access to the main archive, so their job
> would be easier
> - it wouldn't add new things to universe that will stay there if whoever
> has been added those lost interest and create extra work
> - it wouldn't force upstream developers to fit in an artificial 6 months
> cycle which might not work for their project
> 
> It does make sense to freeze and ship a consistent archive for the
> system components (base system, plumbers, desktop shells, default
> applications), it doesn't make sense to try to make small softwares
> (ubuntu-tweaks, simple-lightdm-manager, etc) respect our freezes, cycle,
> rely on acl to upload to the main archive, etc

I disagree.  FWIW, if this kind of system had been in place in 2007, I 
probably wouldn't be an Ubuntu developer today.

I think people join when they can work on something that interests them.  By 
shrinking the scope of what we consider for the archive, you are shrinking the 
scope of potentially interested people.  

OTOH, I don't mind being aggressive about cruft removal where stuff is not 
looked after in Debian or Ubuntu.  If you filter leaf packages at the front 
end, you filter both the packages and potential developers.  If you remove them 
when they cause problems, then you leave yourself open to new contributors.

A lot of this also revolves around the question of what Ubuntu is.  Is it 
meant to be a distribution that supports user's needs or just a core that 
randome third parties offer functionality on top of.  Historically it's aimed 
to be the former and I like it that way.

Scott K



More information about the ubuntu-devel mailing list