On Tue, Apr 8, 2014 at 12:33 AM, Didier Roche <didrocks@xxxxxxxxxx
<mailto:didrocks@xxxxxxxxxx>> wrote:
If you look at the dashboard, those crashers are there for nearly
a month (mentioned in previous landing emails, not actioned
though). So I doubt it's the cause. You have the same crashers
even when the tests are passing.
In the last four images, we have the test failing twice. Maybe you
can add more debug info in the test itself and let the
infrastructure running it? Those info will be always valuable if
you don't have enough right now in intermittent failure.
The problem is that we are blind, so there's no good place to put more
debug info. The only thing that helps in these cases is to get a
screenshot of the state of the screen at the moment of the exception.
If the app is open and the hang button is visible, then we check why
autopilot doesn't see it. If the hang button is not visible, then we
check the script that makes the fake call. If the app is closed, then
we check the crash. There could be something completely unexpected
happening.
I'll run it some more times to see if I find something.
