Re: Backports service becoming official
On 28/09/10 at 09:16 +0200, Stefano Zacchiroli wrote:
> On Tue, Sep 28, 2010 at 12:37:55AM +0200, Lucas Nussbaum wrote:
> > > OK, thanks for the clarification. Still, we need to decide—sort of
> > > now—whether we need to add support in reportbug for mailing backport
> > > report bugs to the bpo list or not (and that might require some time, as
> > > someone needs to do the work, coordinate with the reportbug maintainer
> > > and with the release team, to check whether there's room to have the
> > > change in testing or not). Do you think we should add such a support or
> > > not?
> > I don't think so. Reporting bugs to a mailing list sounds like a hack
> > when you can use the BTS.
> That's not the point, is it? From this discussion is clear that the long
> term solution is to use the BTS, but also that as of now the needed
> support is not in place. So you can't simply veto the alternative
> solution—unless you volunteer to add the needed support in the BTS. The
> whole point of this whole sub-thread was estimating when the support
> will be in place and, according to that, decide whether we should try to
> get the "hack" in Squeeze or not.
So, until we can use the BTS to handle bugs reporting against backported
packages, we have the following solutions:
[A] continue to ask users to (manually) report bugs to a mailing list,
and avoid the use of reportbug. backporters have to subscribe to
[B] add support in reportbug to report bugs against backported packages
to a mailing list instead of firstname.lastname@example.org. backporters have to
subscribe to that list.
[C] ask users to simply use reportbug, have bugs filed in the BTS,
possibly annoying maintainers that don't care about backports, since
the infrastructure for auto-detecting backported packages is not in
place. backporters have to subscribe to the package on the PTS.
My preference goes to [C]. It's not that hard to ignore bugs if you
don't care, and PTS-subscribing is probably a good idea if you care
about a package enough to backport it anyway.