I found that if I replace the default gnome desktop environment with just the `compiz` window manager, then all tests (including this one) pass reliably. So, it's definitely something to do with the gnome desktop environment that is causing this async-iness to occur when repositioning and resizing windows. Why this happens, I still don't know, and I'm running out of ideas to try.
Bug 1600391 Comment 23 Edit History
Note: The actual edited comment in the bug view page will always show the original commenter’s name and original timestamp.
I found that if I replace the default gnome desktop environment with just the `compiz` window manager, then all tests (including this one) pass reliably. see try: https://treeherder.mozilla.org/#/jobs?repo=try&group_state=expanded&selectedJob=279931219&revision=a5c059e5668e3db07165855f67038f3c8a07866d Having `compiz` also fixes a bunch of other tests that implicitly relied on it, resulting in fewer unexpected failures. Which is good. But, as I was about to put up a patch to use `compiz` window manager full-time in bug 1601828, it was brought to attention that this is modifying the environment for sake of the test, so I have decided against replacing `gnome` with `compiz` for the time being. Which means tests will continue to fail. So, it's definitely something to do with the gnome desktop environment that is causing this async-iness to occur when repositioning and resizing windows. Why this happens, I still don't know, and I'm running out of ideas to try.