This introduces smaller improvements in the test framework coming from my upstream project.
By setting -Dchromedriver.headless=false you can enforce a browser window opening during test run.
Improved the end of script detection slightly (or eased an unfixable race condition inherent to Seleniums Webdriver system)
Improved the Spec 1423 Test by replacing one end of script execution wait condition in favor of a "Dom Change" wait condition!
(which should be applied wherever possible)
Fixing and improving the internal comments!
Fixed also a small issue with Issue2906IT, where due to Chrome a hidden NBSP broke the test on some test configurations (difference between Chrome and HTML Unit how to treat the sourcecode was the cause)
I've tested out these new changes, and they work well for Open Liberty. We would appreciate this being included in the upcoming version of the TCK, assuming the review by @BalusC is resolved.
This introduces smaller improvements in the test framework coming from my upstream project. By setting -Dchromedriver.headless=false you can enforce a browser window opening during test run.
Improved the end of script detection slightly (or eased an unfixable race condition inherent to Seleniums Webdriver system) Improved the Spec 1423 Test by replacing one end of script execution wait condition in favor of a "Dom Change" wait condition! (which should be applied wherever possible)
Fixing and improving the internal comments!
Fixed also a small issue with Issue2906IT, where due to Chrome a hidden NBSP broke the test on some test configurations (difference between Chrome and HTML Unit how to treat the sourcecode was the cause)