What I'm working on, and will commit tonight, is that pages always emit a data-page-initialized tag on the <body> element; this is either initally "true" if the page has no JavaScript initialization, or is set to "true" once page initialization is complete. The latter part of that has been around for a while.
In any case, this means that I can change the Selenium waitForPageToLoad() to wait for the data-page-initialized attribute to go to "true". This will vastly improve the stability of all the tests, which often fail when they outrace the page initialization. I may have a few minutes to work on it tonight once I get back from the my client (I'm in Ireland at Aviso today). On Mon, Aug 26, 2013 at 1:26 AM, Dimitris Zenios <[email protected]>wrote: > I already posted on a previous thread regarding the CI failure on issute > TAP5-1007 but most probably nobody saw it. Re posting again here > > This is a legitimate bug. > > If your filesystem is case sensitive like most linux os are > then classLoader.getResource will return null.After that when calling > validateUrl with a null parameter it returns instead of throwing the > expected exception (IllegalStateException) of the test case. > > Most probably if the load resource method returns null it should throw an > exception right away.It doesn't need to pass to validateUrl. > > P.S:Its very weird that in masimos box all tests are running fine.Most > probably he is using a case insensitive filesystem (vfat?).In my box the > test fails like CI > > Below are details of my pc and filesystem. > Linux localhost.localdomain 3.10.6-200.fc19.x86_64 #1 SMP Mon Aug 12 > 14:49:42 UTC 2013 x86_64 x86_64 x86_64 GNU/Linux > /dev/mapper/fedora-home on /home type ext4 > (rw,relatime,seclabel,data=ordered) > > > On Mon, Aug 26, 2013 at 11:18 AM, Massimo Lusetti <[email protected] > >wrote: > > > Yep. > > That was my feeling too but that kind of solutions are lurching, usually. > > > > > > On Mon, Aug 26, 2013 at 9:02 AM, Howard Lewis Ship <[email protected]> > > wrote: > > > > > Been playing wack-a-mole with some of the Tapestry tests ... but it > > appears > > > its just a matter of adding enough waitForPageInitialized() calls. > > Mostly. > > > > > > -- > > > Howard M. Lewis Ship > > > > > > Creator of Apache Tapestry > > > > > > The source for Tapestry training, mentoring and support. Contact me to > > > learn how I can get you up and productive in Tapestry fast! > > > > > > (971) 678-5210 > > > http://howardlewisship.com > > > > > > > > > > > -- > > Massimo Lusetti > > > -- Howard M. Lewis Ship Creator of Apache Tapestry The source for Tapestry training, mentoring and support. Contact me to learn how I can get you up and productive in Tapestry fast! (971) 678-5210 http://howardlewisship.com
