[14:39] <jtaylor> hm I though adt failures of rdepends prevent proposed migration
[14:40] <jtaylor> or doesn'T that apply to main breaking universe?
[15:05] <ScottK> That shouldn't make a difference, AIUI.
[15:05] <ScottK> jtaylor: What's the example?
[15:05] <jtaylor> scipy
[15:06] <jtaylor> broken by pillow
[15:13] <ScottK> http://people.canonical.com/~ubuntu-archive/proposed-migration/log/utopic/2014-10-03/13:11:27.log
[15:13] <ScottK> Definitely looks like scipy results weren't considered in migration.
[15:15] <ScottK> scipy hadn't been run yet.  It didn't get run until today.
[15:15] <ScottK> So the question is how come pillow didn't trigger ADT for scipy.
[15:15] <ScottK> jtaylor: I'd ask pitti that question on Monday.
[15:16] <jtaylor> it did trigger it and it failed
[15:16] <jtaylor> or something triggered it at least
[15:16] <ScottK> According to the ADT hisotry, it triggered on 10/4.
[15:16] <ScottK> Pillow migrated on 10/3.
[15:17] <ScottK> So the test run that might have prevented migration never ran..
[15:17] <ScottK> https://jenkins.qa.ubuntu.com/job/utopic-adt-python-scipy/
[15:17] <jtaylor> right
[15:21] <jtaylor> I guess its to slow
[15:21] <jtaylor> an hour adt test is a little much