On Sat, May 07, 2005 at 02:17:34AM +0200, Adrian Bunk wrote:
> now that Frank has uploaded a NMU for this bug ("shlibs incorrect" in
> liborbit0), the breakage already caused should be fixed some time after
> this fixed package has reached sarge (to catch all possible t-p-u or
> s-p-u uploads of broken packages).
>
> Thankfully, it seems all broken binary packages are easily identifiable
> due to the unversioned liborbit0 dependency.
>
> What's the best way to handle packages with such a broken dependency on
> one or more architectures?
>
> Binary NMUs?
> By whom (needs access to all 11 architectures including the ability to
> install packages)?
> RC bugs?
Depends on the affected architectures. A list of currently affected
packages (I've found no cases where the liborbit0 depends is at
the end of a Depnds line in case someone wonders)
[EMAIL PROTECTED]:/org/packages.debian.org/files/archive$ grep-dctrl -sPackage
-n -FDepends liborbit0, testing_*Packages | sort | uniq
libgnome-vfs-common
libgnome-vfs0
liborbit-dev
orbit
sgcontrol
[EMAIL PROTECTED]:/org/packages.debian.org/files/archive$ grep-dctrl -sPackage
-n -FDepends liborbit0, unstable_*Packages | sort | uniq
bonobo
libbonobo2
libgnome-vfs-common
libgnome-vfs0
liborbit-dev
orbit
sgcontrol
For sgcontrol, only arm seems affected, for gnome-vfs all arches, same
for bonobo.
So I would say binNMU for sgcontrol and source-full uploads for the
other two.
Let's hope we don't get way more of these...
Gruesse,
--
Frank Lichtenheld <[EMAIL PROTECTED]>
www: http://www.djpig.de/
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
