On Mon, 1 Nov 2021 15:34:42 GMT, Alexander Zuev <kiz...@openjdk.org> wrote:
> Looks like the problem occurs when system is slow to actually close windows, > then keyboard events are going to the wrong window. The idea of the fix is to > bring the new window to front before actual testing starts and close window > on EDT and then wait for 2 seconds for actual hardware window to disappear. Marked as reviewed by prr (Reviewer). When originally reported this test frequently timed out. So if it didn't have the same timeout in 2000 iterations perhaps the root cause of that is fixed ? Any ideas what that fix might have been ? But with that and these modifications we can put this test back into production ------------- PR: https://git.openjdk.java.net/jdk/pull/6194
