test_button_with_subtree seems to rely on the old behavior

Comment on attachment 9016587 [details] [diff] [review] clear.py_button_export.diff Review of attachment 9016587 [details] [diff] [review]: ----------------------------------------------------------------- Thanks Olli!

Interesting that no other browser vendor, and I assume have already implemented that behavior, has complaint about this test yet.

Hmm, this test gets run on all the browsers? The test is basically data:text/html,<button><input value=foobar></button> and one is supposed to be able to use that input element. And can in current Chrome, and in FF with the patch for bug 1089326

If it’s of any help, it looks like other vendors run test_button_with_subtree but that the test in its current state is failing for them: https://wpt.fyi/results/webdriver/tests/element_clear/clear.py?label=experimental I suppose we will get the real answer to whether or not they implement the correct behaviour when this patch lands, and wpt.fyi does another overnight test run.

Yes, that is indeed correct. Lets see how this changes when the patch landed.

Pushed by opettay@mozilla.com: https://hg.mozilla.org/integration/mozilla-inbound/rev/7029e8b4d7d0 clear.py relies on old Gecko's <button> hit testing, r=ato

Created web-platform-tests PR https://github.com/web-platform-tests/wpt/pull/14477 for changes under testing/web-platform/tests

Upstream web-platform-tests status checks passed, PR will merge once commit reaches central.

Upstream PR merged