Re: No native packages?
On Sat, 2013-02-16 at 12:11:29 +0100, Stefano Zacchiroli wrote:
> On Wed, Feb 13, 2013 at 03:41:53AM +0100, Guillem Jover wrote:
> > > (Random data point: I have 14 packages with versions indicating they
> > > are NMUed native packages installed on my system. Some of them have
> > > priority standard or higher.)
> > I've a similar number on my system, but most of these appear to be
> > from people that have not been active in the project for a while?
> Precisely, hence the usefulness of NMUs. In the interim when you don't
> know if those people have "disappeared" from the project (for all sort
> of valid reasons) or not, other developers can help using NMUs. That way
> they help both the maintainers in question, relieving some of their
> duties if they are gone only temporarily, and the project as a whole
> that avoids being technically stalled for a while.
I was not attacking the NMU faith or its usefulness, otherwise we'd
obviously not have NMUs; what I was getting at in this thread, that
got trimmed in the quoted mail, is the thought that the current NMU
procedure for native packages is bogus, as argumented previously.
Even then, we do not consider long term maintenance of orphaned
packages through QA uploads a good thing, so even more reason for this
not to be done for dead native packages, which lots of maintainers
might rely on. But, let me repeat, if this needs to be done anyway,
then those changes should be clearly distinguished as not coming from
the native package developer(s), by not taking over the upstream file
releases and versionspace.
When you don't know if the native package developer(s) have disappeared,
well, you ask, and might want to consider doing the same we do with
dead non-native upstream projects, also mentioned previously, which
might include being offered to be added as co-upstream, an agreed
handover, a fork if the developer(s) do not want to step aside, a
transition of reverse dependencies to another package, or a possible
upstream adoption (after some safe timespan) if the developer(s) have
completely disappeared from the face of the earth. This did not appear
to have been done with those currently NMUed native packages, hence my