Linkintegrity test fails randomly on jenkins.plone.org

[tisto]

http://jenkins.plone.org/job/plone-5.0-python-2.7-robot/2255/robot/report/robot_log.html#s1-s15-t3

cc @pbauer

[pbauer]

@tisto I'll fix it

[pbauer]

Fixed in 1b50bd3

[tisto]

@pbauer seems the test is still unstable: http://jenkins.plone.org/job/plone-5.0-python-2.7-robot/2275/robot/report/robot_log.html#s1-s15-t3

[pbauer]

Fixed in 6d466db

[tisto]

Test still fails randomly:

http://jenkins.plone.org/job/plone-5.0-python-2.7-robot/2301/robot/report/robot_report.html#suites?s1-s15

[pbauer]

@tisto I have no clue how that can fail. After Wait Until Page Contains Element css=.pattern-relateditems-result-select.selectable there should not be a problem clicking the same element with Click Link css=.pattern-relateditems-result-select.selectable. Ideas?

[tisto]

@pbauer the element could not be visible (then you need to use "wait until element is visible", sometimes other html elements have a higher z-index so the underlying element can not be clicked. "Wait until page contains" just makes sure the element is there, not that the page is fully loaded and the javascript code has been fully executed. It is also quite common that the underlying JS implementation itself is fragile. Selenium can't fix that then...

[pbauer]

@tisto thanks for the tip

[pbauer]

I think I finally fixed the test for good in e409071

The main problem was that css=.pattern-relateditems-result-select.selectable (linkable items) existed multiple times before entering Foo and only once afterwards. So the Wait until element is visible caught one from the initial list that disappeared before clicking due to the entered text. Now I wait until the result is exactly what I need using Wait until element is visible jquery=.select2-highlighted .pattern-relateditems-result-select.selectable:contains(Foo).
By the way: the jquery-selector is awesome. Much easier to write that xpath or css.

[mauritsvanrees]

Awesome! Thanks a lot.