Package: qa.debian.org
Severity: wishlist

08:25 < bubulle> hmmm, /me wonders whether #352166 (shadow FTBFS on hppa)
shoul djust be closed or not. It seems
that there's been a binNMU this morning but I have no clue
about who did it and whether this
just says the FTBFS was a problem on the hppa buildd
08:31 < vorlon> bubulle: well, you could thwap aurel32 for circumventing the buildds, for starters
08:31 < vorlon> anyway, sarti has serious issues with basic shell stuff, so
the bug can probably just be closed  again.
08:31 < vorlon> wrong severity for the bug to be opened at, btw. :)
08:32 < bubulle> as usual..:)
08:32 < bubulle> note to self: thou shalt not open bugs before the needed
level of cafeine floods in thy veins
08:34 < bubulle> vorlon: how did you see that aurel32 was the one doing the
binNMU? I need to learn somethign here
08:38 < vorlon> bubulle: grab the .changes from incoming, gpg --verify
08:39 < vorlon> bubulle: dato also has a script that checks the uploader;
don't remember the details of it, though
08:39 < bubulle> should be an interesting addition to do on
packages.qa.debian.org
08:39 < bubulle> buxy: ^^^^
08:55 < isaac> bubulle:
08:56 < isaac> bubulle: http://people.debian.org/~adeodato/bin/who-uploads
if you want to have a look

Right, this is also discussed at
  * #347809: devscripts: Please include script indicating recent uploads (Was: Re: RFX: Gajim, a Jabber client)

I guess only one of these packages needs to implement this, and malex
has good point regarding the keyring.  So if pqdo implements it first,
remind me to close #347809 (unless someone else wants it).

Justin

tags 352200 + moreinfo
thanks

The bug report was not very clear to me. I guess, from the title, that
the request is to show in PTS pages binNMU versions of a given
package. That actually breaks one of the fundamental assumption of the
PTS, namely that of being "source-based". As binary versions can
change from architecture to architecture, it is not clear to me how
the PTS can display them. For sure we don't want a big row anywhere
listing the different versions on each architecture.

Maybe, what can be done, is to add a pop-up on version numbers for
specific releases which shows binNMU versions only for the
architectures which have some. Still, doing that would require
downloading version information for all architectures, which are
currently not available, even though nowadays they are easily
available from UDD.

Would that satisfy submitter request?

Cheers.

Quoting Stefano Zacchiroli (zack@debian.org):

If I correctly remember the original rationale for this bug, yes.

Still, That wish was not that a critical one or something really
important, so if you feel like it would be breaking some design
philosophy of the PTS, I'm fine with the bug being marked "wontfix"
(or closed....I'm never fond of bugs remaining opened for years when
marked "wontfix").