On Tue, 2010-03-02 at 11:52 -0500, Matthias Clasen wrote: > On Tue, 2010-03-02 at 08:45 -0800, Adam Williamson wrote: > > On Tue, 2010-03-02 at 11:33 -0500, Matthias Clasen wrote: > > > > > So, I think having an automated test that basically does these steps and > > > verifies that the current set of updates can be successfully installed > > > on a GOLD install would be very useful. > > > > it wouldn't hurt, but we (QA and Richard) already know about this > > problem. The thing is, it's basically impossible to fix, for F12. > > Right, I know. And I also know how to work around it now. > My question was about making sure that we don't run into similar issues > with F13. As I said, we really need to have priority PK updates to make that possible. At present there's no way we could guarantee it, even with testing; exactly the same thing could happen again. This didn't happen when F12 first came out, we couldn't have caught it before release; it only appeared after the update set for F12 came to include packages which trigger the bug. At present something similar could happen in F13 - a hidden bug in PK which would only emerge with some update package that came out after release - and we would not be able to 'fix' it for the same reason: we can ship a fix for PackageKit as an update, but if PK can't install the initial update set, it can't install the update that fixes itself. :) We could *spot* the problem with automated testing, but we couldn't prevent it. -- Adam Williamson Fedora QA Community Monkey IRC: adamw | Fedora Talk: adamwill AT fedoraproject DOT org http://www.happyassassin.net -- test mailing list test@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe: https://admin.fedoraproject.org/mailman/listinfo/test