On Wed, 18 Apr 2007 15:32:53 +0200, Thorsten Leemhuis wrote: > On 18.04.2007 15:26, Hans de Goede wrote: > > [...] > > This raises 2 issues: ... which could have been posted in a less reproachful way. > > 1 Notice how the deps are only broken for the i386 version in the x86_64 tree, > > this is something outside my control. If script XXX decides to put a i386 > > gnumeric in the x86_64 tree, then the script should also make sure it puts in > > all need deps from the i386 tree It does. But Core and Extras are not merged yet, so it cannot do anything inside the Core repo. > perl.i386 was in the Core tree but got removed without proper > announcement/discussion beforehand (and even worse: that happened on the > day of the feature freeze). See > https://www.redhat.com/archives/fedora-devel-list/2007-April/msg01004.html Also note that if gaim had stayed in Core, it would now be broken in Core x86_64 due to this removal of perl.i386. [FC6 ships gaim.i386 and perl.i386 for x86_64] > > 2 Why does the script put the i386 version of gnumeric (an application) in > > the x86_64 tree at all? That just doesn't make sense. > > I assume because there is a gnumeric-devel (the multilib-magic scripts > afaik try to track in most of the devel packages, which thus tracks in > the main package normally, too). Yes. And that means, this issue is _not_ "outside packager's control", because in most of the cases the packager *could* split off a sub-package and adjust the -devel pkg requirements accordingly. However, (!) whether this is the way to go with these broken deps this time, too, or whether to exclude them from the multi-lib pushscript, needs further investigation. -- Fedora-maintainers mailing list Fedora-maintainers@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/fedora-maintainers -- Fedora-maintainers-readonly mailing list Fedora-maintainers-readonly@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/fedora-maintainers-readonly
