Am Samstag, den 29.12.2007, 01:09 +0000 schrieb Kevin Kofler: > Christoph Wickert <christoph.wickert <at> nurfuerspam.de> writes: > > You are right. I thought we already had policy for that but the wiki > > says: > > "If you feel that community testing is unnecessary for your update, you > > can choose to push it straight to the stable fedora-updates repository > > instead." > > > > IMO this is wrong, it should only be allowed for security updates. > > It should also stay allowed for critical regression fixes at the very least. > Unfortunately, sometimes updates slip through testing despite causing serious > issues. For example, an SDL update caused _any_ build against SDL-devel to fail > (because SDL-config.h was corrupted by a trivial typo in the specfile which was > missed during testing). We really want the fix to such regressions to go to > stable as quickly as possible, not to have to go to testing first. I completely agree with you. Maybe we could say that updates are allowed to bypass testing if they fix a) serious bugs b) bugs marked as "urgent" c) broken deps For c) I think only plain rebuilds should be allowed. If the broken deps are fixed by updating to a newer version the new package should be tested again, at least for one or two days. Christoph -- fedora-devel-list mailing list fedora-devel-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/fedora-devel-list
