On 13 September 2011 13:15, Manik Surtani <manik(a)jboss.org> wrote:
I agree that the instability in the test suite removes the urgency to
get things fixed, and that is a bad thing.
Which is why I think we should group the handful of occasionally unstable tests and put
them in a separate test group so that they are not run all the time. I'll look into
this.
+100
I think it's acceptable to disable those tests which occasionally fail
and create a JIRA for each one to be re-enabled when we have time to
look into it. It also tracks what works with confidence vs. what might
be broken.
I don't always know if a failing test is "one of the usual ones", or a
new one which should ring an alarm bell, so this would be useful for
anyone but especially for me as I don't work on it everyday and have a
terrible memory.
Cheers,
Sanne