On Wed, Dec 18, 2013 at 10:29 AM, Antoine Musso <[email protected]> wrote:

> What about prefixing / suffixing the article titles with the browser
> name or a random unique token?
>

I will repeat again what I said many many times. :)

For tests to be robust, every test has to live in it's own universe. That
means that the test itself should take care of creating anything it needs
(users, pages, files...) and cleaning after itself (if needed).

In this case, I would suggest creating a page with random name (in that
case we know no other test is using it) and doing all actions there. Since
deleting pages is not easy (if possible at all), the test should just leave
the page as is after the test is done. We can work on that today in our
pairing session.

Željko
_______________________________________________
QA mailing list
[email protected]
https://lists.wikimedia.org/mailman/listinfo/qa

Reply via email to