Re: pidgin-data 2.6.3-1~bpo50+1 brakes pidgin

[Paul Harris <harris.pc@gmail.com>]

2009/10/30 Gerfried Fuchs <rhonda@deb.at>

* Paul Harris <harris.pc@gmail.com> [2009-10-30 12:02:09 CET]:

> 2009/10/30 Gerfried Fuchs <rhonda@deb.at>

> > * Maciej Sieczka <msieczka@sieczka.org> [2009-10-30 06:44:00 CET]:
> > > Just a moment ago I have upgraded to pidgin-data 2.6.3-1~bpo50+1 which
> > > uninstalled my pidgin2.6.2-1~bpo50+2. I cannot install pidgin 2.6
> > > anymore now:
> >
> >  Out of curiosity, why didn't you stop it to uninstall your pidgin
> > package?
>
> The same problem hurt me a while ago too.

 And will hurt you (as including other soo) again next time it will get
uploaded.

> Out of curiousity, why doesn't the pidgin maintainer ensure the package is
> built on all platforms before uploading it into the pool?

 Because that's not the maintainer's call as they don't have access to
the buildd network but rather something the archive software would have
to do. And it would mean that pidgin would still be stuck at 2.5.8
for lenny-backports because ia64 wasn't built since. There are several
approches to the issue, all of them have their benefits and issues.

Alexander said that "they are built out of the pool".  But you say you don't have access to the buildd network.

So, am I right to say that normally the source packages are uploaded, and buildd does the build work?  But not for backports, instead, backports is built on your own machine and then uploaded.   Right?