canonical-ci-engineering team mailing list archive

[Thomi Richards <thomi.richards@xxxxxxxxxxxxx>]

On Tue, Jan 7, 2014 at 2:30 PM, Chris Lee <chris.lee@xxxxxxxxxxxxx> wrote:

> TnT guys, am I missing any details or important points?


That sounds good, but I wanted to clarify something: the question we're
trying to answer is: "Does autopilot trunk introduce test regressions which
are not present in the latest image build?"

With that in mind, whatever we end up with should be identical to the daily
image build, but with newer autopilot packages.

Currently, if I run the gallery-app (as an example, nothing special about
gallery-app) autopilot test suite on my phone, I get *much* worse results
than appear on the CI dashboard - and I have no idea why. The failures
don't seem to be issues in autopilot itself, but rather some environment
setup I'm missing. So the idea here is to take the knowledge encoded in the
jenkins jobs and re-use it for our purposes.


Cheers,

-- 
Thomi Richards
thomi.richards@xxxxxxxxxxxxx