On Tue, 2016-03-29 at 05:58 +0000, Fedora compose checker wrote:
Missing expected images:
Workstation live i386
Cloud_base raw-xz x86_64
Cloud_base raw-xz i386
Atomic raw-xz x86_64
Kde raw-xz armhfp
Minimal raw-xz armhfp
Workstation live x86_64
Images in this compose but not Rawhide-20160323.n.0:
Workstation raw-xz armhfp
Robotics live x86_64
Robotics live i386
Cinnamon live i386
Soas live i386
Soas live x86_64
Cinnamon live x86_64
Jam_kde live i386
Jam_kde live x86_64
Images in Rawhide-20160323.n.0 but not this:
Soas raw-xz armhfp
Workstation live i386
Mate live i386
Astronomy_kde live i386
Mate live x86_64
Xfce raw-xz armhfp
Lxde raw-xz armhfp
Astronomy_kde live x86_64
Server raw-xz armhfp
Workstation live x86_64
Minimal raw-xz armhfp
Docker_base docker x86_64
Failed openQA tests: 37 of 74
I mentioned this on qa-devel, but - a lot of the failed x86_64 tests
are openQA needle match issues (i.e. we aren't hitting an obvious bug,
but rather the appearance of something changed enough that the
screenshot matching thingy openQA uses doesn't match any more). In this
case it's not entirely clear whether the changed rendering is
'intended' or not (some bits of text in anaconda look somewhat
different and it's not obvious whether this would actually be how the
anaconda devs want it to look now, or if they'd consider it a bug and
it's some kind of change in GTK+ or something else like that which
anaconda wasn't accounting for). So I'll talk to the anaconda devs and
maybe GNOME folks in the morning before deciding what to do about it
for openQA. For now let's say it seems that the 2016-03-28 Rawhide
compose is broadly workable on x86_64 at least. We probably won't re-
run the tests for this compose even if we wind up updating the needles,
so more 'accurate' results will come for a future compose.
--
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | XMPP: adamw AT happyassassin . net
http://www.happyassassin.net