[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: koji breakage for F12 builds

Am Mittwoch, den 13.05.2009, 19:17 +0200 schrieb Nicolas Chauvet:
> 2009/5/13 James Antill <james fedoraproject org>:
> > On Wed, 2009-05-13 at 14:15 +0100, Bastien Nocera wrote:
> >> On Wed, 2009-05-13 at 13:15 +0200, Christoph Wickert wrote:
> >> > I think the problem is, that some packages  notification-deamon instead
> >> > of (virtual) desktop-notification-deamon:
> > [...]
> >> Except that I don't see why those would cause problems during the build,
> >> when they're not even getting installed in the buildroots.
> >>
> >> The problem was/is that libnotify-devel requires a
> >> desktop-notification-daemon (via libnotify itself), and yum will install
> >> both packages that provide it.
> >
> >  How did you come to this conclusion?
> >  I'm not saying it's impossible, but the above seems unlikely unless yum
> > picks one on the virtual provides hit and then it gets an explicit
> > requires on the other (which isn't the same thing).
> xfce4-notifyd : 12 characters
> notification-daemon:  19 characters
> Short name wins! 

This is only partly true as Seth pointed out on this list just recently
(not sure which topic it was though): IIRC short name wins only applies
if both packages provide *exactly* the same thing, and this shouldn't be
the case here.

> xfce4-notifyd is picked unless something within the
> dependencies is hardcoded to notification-daemon.

There *are* packages explicitly requiring notification-daemon, so this
needs to be rephrased: Why is xfce4-notifyd choosen *although* (!=
"unless") some packages are hardcoding notification-daemon, which
already satisfies all requirements? 


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]