45
edits
(Add meeting notes for Screenshots section) |
(Add meeting notes to This week I learned) |
||
Line 209: | Line 209: | ||
=This week I learned= | =This week I learned= | ||
<ul> | |||
<li>[agashlin] Windows [https://devblogs.microsoft.com/oldnewthing/20160108-00/?p=92821 FILE_FLAG_DELETE_ON_CLOSE] starts deleting when the first handle closes, but doesn't actually delete it until the last handle closes; it is inaccessible in between ([https://bugzilla.mozilla.org/show_bug.cgi?id=1696772 Bug 1696772])</li> | |||
<li>[florian] `TestUtils.waitForCondition` waits first 100ms (or the specified interval) before checking the condition for the first time. This implicit timer is a cause of several intermittent test failures when the condition isn’t exactly what the test needs, and the test relies on something happening during the initial timer. Attempting to change this behavior made 20+ intermittent failures become perma-fails. I’m [https://bugzilla.mozilla.org/show_bug.cgi?id=1596165 fixing] this.</li> | |||
<li>[florian] `BrowserTestUtils.openNewForegroundTab` awaits `BrowserTestUtils.switchTab` which also has an implicit [https://searchfox.org/mozilla-central/rev/4e87b5392eafe1f1d49017e76f7317b06ec0b1d8/browser/modules/AsyncTabSwitcher.jsm#88 300ms timer]</li> | |||
<li> | |||
[florian] EventUtils.synthesizeKey can sometimes trigger 2 “keypress” events instead of one, when the event is transmitted to remote content. | |||
<ul> | |||
<li>The Event object has [https://searchfox.org/mozilla-central/rev/be906232eedb22c064b78f3806b38964c04f1fbc/dom/webidl/Event.webidl#88-100 isWaitingReplyFromRemoteContent and isReplyEventFromRemoteContent attributes] that I didn’t know existed.</li> | |||
<li>Some intermittent failures are caused by a test waiting for a synthesized keypress event, and the next test receiving it too!</li> | |||
<li>Event.preventDefault() stops the event from being sent to the content process after having been handled by a listener in the parent.</li> | |||
</ul> | |||
</li> | |||
</ul> |
edits