On Sat, Dec 13, 2008 at 02:24:18PM +0100, Robert Scheck wrote: > Hello Matej, > > On Wed, 10 Dec 2008, Matej Cepl wrote: > > Yes, there should be some mechanism how Bodhi should stop package > > from entering updates, but I guess bodhi developers will have to > > work a little bit harder than putting yet another stupid form on > > the website somewhere, which doesn't integrate with anything than > > with itself. > > Bodhi is nice, yes. But people are not using it that much. Even packages > with a broader usage width are getting less to no karma points. And there > is no reflection into Bugzilla if something not works. Because if something > not works, that is likely a bug - why are such things only kept in Bodhi > but not noted as bug report? Given that Bodhi updates bugzilla reports when updates are pushed, maybe it can be made to update them with karma and comments too? > Anyway, bodhi only refers to updates-testing. Can't we unpull packages from > updates fast once the karma gets negative? I'm still not happy, that there > was such a slow reaction for the dbus/PackageKit breakage...the interaction > from bodhi to the results on the mirrors seems to be currently ~ 48 hours > when looking to my last phpMyAdmin security update (or is therefore just > our scheme of how to handle security updates responsible?). The package for > EPEL of phpMyAdmin made it into the mirrors much faster as on Fedora. If there were PK integration, there could be lots of interesting ways to allow the user to interact with the update system, such as single-click karma reporting back to bodhi, displaying the karma on updates to the user so they can choose whether to update that package or not, setting a preference that says somthing like "wait until karma gets to +3 before defaulting the checkbox to enabled to update this package". -- fedora-devel-list mailing list fedora-devel-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/fedora-devel-list
