[Debian-med-packaging] Your NMU of pirs

Andreas Tille andreas at an3as.eu
Mon Jan 16 17:12:40 UTC 2017


Hi John,

On Sat, Jan 14, 2017 at 11:12:15PM +0100, John Paul Adrian Glaubitz wrote:
> > You should be aware that other maintainers might consider this kind of
> > NMUs (without bug, no delay, no asking for confirmation) as offensive.
> 
> Correct me if I'm wrong, but you are listed on the LowNMU page [1] which states:
> 
> > The maintainers and maintainer groups listed below declare that their packages
> > (all, or the ones listed specifically below) may be fixed and uploaded without
> > delay, at any time, as long as the NMU procedure in the Debian Developer's
> > Reference is otherwise followed. You don't need to contact the maintainers
> > beforehand, and you don't need to use a delayed upload queue. If the package
> > maintainer or maintainer group is active, it is polite to let them have a
> > stab at fixing the problem first.
> 
> If I see a maintainer on this list with the statement "all packages", I assume
> they mean it that way. If you prefer to be contacted beforehands, please put
> a comment on that page explicitly stating that.

As far as I know an NMU should at least fix a bug.  As I tried to
express I'm *personally* fine with what you did.  However, as you
noticed the conflicting uploads could have been avoided and I might add
to this page that the NMUer should try to get the changes commited to
VCS (where ACLs are set to grant any DD commit permissions).
 
> Besides, it's not really acceptable to have a package in an FTBFS on release
> architectures for such a long time, really.

If you know a solution for this kind of problem documenting it in BTS is
the traditional way to go.  The Debian Med team is quite quick to upload
packages with patches applied.

I just want to avoid two things:

   * competing uploads
   * changes missing in VCS to avoid regressions

Thanks for your fix anyway

      Andreas.

-- 
http://fam-tille.de



More information about the Debian-med-packaging mailing list