Proposal to remove <audio> element from contain-inline-size-replaced.html

[vmpstr]

Test List

https://wpt.fyi/results/css/css-contain/contain-inline-size-replaced.html

Rationale

I'm proposing to remove element subtests from the contain-inline-size-replaced.html tests, which is a part of containment interop

The element is underspecified and is implemented in different ways in different browsers. It is sized differently and responds to containment in different ways. One of the subtests is failing on 3/3 major browsers. The other subtest is failing on 2/3 major browsers.

Additional discussion: web-platform-tests/wpt#39325

[vmpstr]

@dholbert you mentioned that you would be ok with this change, is that still the case?

@nt1m would you be able to say if WebKit is OK with change as well?

If everyone agrees, I can make the change

/cc @foolip as an fyi

[nt1m]

@rwlbuis @cathiechen What do you think?

[dholbert]

Yes, I'm still OK with the change; my thinking from the "additional discussion" linked-issue still applies.

(tl;dr: the test is assuming that browsers don't have UA stylesheet rules to apply a particular writing-mode or default width/height values on this element. In practice, browsers do actually do those things. And in terms of actual author-impacting interop concerns, the "does contain:size collapse an audio element to 0x0" question does not seem particularly important, given that contain:size is meant to be used on container elements (to contain the size of their arbitrary subtree content), and it's not super useful on leaf nodes.)

[rwlbuis]

@rwlbuis @cathiechen What do you think?

Cathie knows size containment better. FWIW, at the time audio was the only subtest I could not fix (https://bugs.webkit.org/show_bug.cgi?id=244799), so would be nice to get this corrected and passing.

[cathiechen]

Sorry for the slow reply!

Yes, this change looks good to me!
In WebKit, audio has min-width: 44px !important, so contain: inline-size is not effective in this case.