Proposal for new test-run of Mozmill tests

[Henrik Skupin]

We have now started the process to implement new Mozmill tests for the
new Add-ons Manager which has been landed with Firefox 4. Specifically
we want to focus our work on the Discovery Pane (Get Add-ons) for the
beginning.

Those tests will interact with the client and the underlying website.
All of them can be found in the WebQA section on Litmus. Due to the
different nature of those tests I would propose that we create a new
test-run, so we do not conflict with our functional test-run which is
used for BFT/FFT tests only on the client side.

I would like to get feedback from you, especially if you agree on the
proposal and also how we could name this testrun. I don't have a good
name handy yet, but it should indeed contain 'web'. Has anyone an idea?

Thanks,

--
Henrik Skupin
QA Execution Engineer
Mozilla Corporation

[Geo Mealer]

I'm not wholly sure why this calls for a distinct test run. It's definitely using a different approach, but it's still a functional test at the end of the day, right?

Does it have other environmental setup needs that distinguish it from the other functional tests or some other factor that means we wouldn't want to run them together?

Explaining more what the differences are here might help that understanding and suggest a good name.

Geo
--
Geo Mealer
gme...@mozilla.com

Thanks,

--
You received this message because you are subscribed to the Google Groups "MozMill Developers" group.
To post to this group, send email to mozmi...@googlegroups.com.
To unsubscribe from this group, send email to mozmill-dev...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/mozmill-dev?hl=en.

[AaronMT]

Agreed, it's still a functional test. Henrik what do you mean by "...
so we do not conflict with our functional test-run", what confliction?
timeouts? aren't these already gracefully handled?

On May 18, 1:59 pm, Geo Mealer <gmea...@mozilla.com> wrote:
> I'm not wholly sure why this calls for a distinct test run. It's definitely using a different approach, but it's still a functional test at the end of the day, right?
>
> Does it have other environmental setup needs that distinguish it from the other functional tests or some other factor that means we wouldn't want to run them together?
>
> Explaining more what the differences are here might help that understanding and suggest a good name.
>
> Geo
> --
> Geo Mealer

> gmea...@mozilla.com

[Henrik Skupin]

AaronMT wrote on 5/18/11 9:45 PM:

> Agreed, it's still a functional test. Henrik what do you mean by "...
> so we do not conflict with our functional test-run", what confliction?
> timeouts? aren't these already gracefully handled?

As discussed in the yesterdays Automation meeting we will go ahead with
a new testrun because the nature of the tests are different from our
existent functional tests which map 1-1 to our client tests. The new
testrun will be named 'remote' and new tests will have to be added below
'tests/remote/'. For the upcoming tests of the Discovery Pane it will
mean they are stored in 'tests/remote/testDiscoveryPane/'.

See the following bug for the current work:
https://bugzilla.mozilla.org/show_bug.cgi?id=658231