This test often fails in NB-Core-Build builder.
Recently it has failed on Aug 27, two times on Aug 28 and at least two times on Aug 30.
See the attached output.
Created attachment 139496 [details]
Created attachment 139499 [details]
Sometimes it gets to the next test and fails at ReusableEditorTest
After fixing of issue 234781 I'm no longer able to reproduce any random openide.text tests failures so optimistically marking as duplicate.
If the problem persists please reopen.
*** This bug has been marked as a duplicate of bug 234781 ***
Unfortunately, it still fails, sometimes. :-(
Unfortunately I still cannot reproduce ReusableEditorTest and ReusableEditor2Test on any of my machines and VirtualBox instances but I've at least attempted to improve robustness of the CloneableEditorInitializer. I'll wait few builds and check if the continuous build would fail again and possibly reopen the issue.
I've experienced NotifyModifyTest failing once. Currently the reload threading is a bit diferent - it requires one extra pre-reload part posted into EDT and also (potentially long) loading part running outside of EDT (as requested by an issue to not use input stream from EDT). Hopefully fixed the NotifyModifyTest.
Also I've updated DocumentCannotBeClosedWhenReadLockedTest to use count down latches instead of wait/notify mechanism which was failing for me ocassionally and also reproducibly when running the test under debugger.
It looks like the last fix helped and the build no longer fails on the tests and no ReusableEditor2Test appears failing.
I would like to ask QE for review since I would like to integrate into release74.
I'd like to ask Honza L. for review. Thanks.
I think the patch can be backported to release74.
verified in trunk
a410c353a968 transplanted to e77ec40e102a
*** Bug 236081 has been marked as a duplicate of this bug. ***
*** Bug 236049 has been marked as a duplicate of this bug. ***