The WPT test css/css-flexbox/flexbox_direction-row-reverse.html (imported/w3c/web-platform-tests/css/css-flexbox/flexbox_direction-row-reverse.html) fails on Mac due to small differences in how the test font its rendered: Check the diff: https://wpt.fyi/analyzer?screenshot=sha1%3A49959e251d81e5e1b670007fb2b16f987ffc3cb8&screenshot=sha1%3Aa0079d4c058551db56b11d3caed6e736abffc876 This issue doesn't happen on other ports like iOS or GTK. The test imported/w3c/web-platform-tests/css/css-flexbox/flexbox_direction-row-reverse.html and css3/flexbox/csswg/flexbox_direction-row-reverse.html are in theory the same and should be de-duplicated. However, in r180652 we did a change to css3/flexbox/csswg/flexbox_direction-row-reverse.html that was never exported to WPT so we ended with two different copies of the test. The change done in r180652 is about using Ahem font instead of a normal font to the test. Ideally we should: - Replace imported/w3c/web-platform-tests/css/css-flexbox/flexbox_direction-row-reverse.html with css3/flexbox/csswg/flexbox_direction-row-reverse.html - Export the changes done in r180652 to WPT
These 3 additional ones have the same issue imported/w3c/web-platform-tests/css/css-flexbox/flexbox-align-self-vert-003.xhtml imported/w3c/web-platform-tests/css/css-flexbox/flexbox-align-self-vert-004.xhtml imported/w3c/web-platform-tests/css/css-flexbox/flexbox-align-self-vert-rtl-001.xhtml
<rdar://problem/94161403>
Sounds like a good idea. Is the WebKit bug tracker the right place to track a change to WPT? Maybe this bug is just being used to track another WPT bug?
I exported Myles' change to WPT: https://github.com/web-platform-tests/wpt/pull/35482
https://github.com/web-platform-tests/wpt/pull/35921