50327 2010-12-01 09:28:48 -0800 CSS 2.1 failure: font-family-rule-011 2022-08-20 14:39:47 -0700 1 1 1 Unclassified WebKit CSS 528+ (Nightly build) All All NEW http://test.csswg.org/suites/css21_dev/20110323/xhtml1/font-style-rule-001.xht P2 Normal --- 47141 1 simon.fraser webkit-unassigned ahmad.saleem792 ap bfulgham rniwa simon.fraser tabatkins zalan oldest_to_newest 315689 0 simon.fraser 2010-12-01 09:28:48 -0800 The font-family-rule-011 test fails. http://www.w3.org/TR/CSS21/fonts.html#font-family-prop Note the words in the spec: "The keywords 'initial' and 'default' are reserved for future use and must also be quoted when used as font names. UAs must not consider these keywords as matching the '<family-name>' type." 1892680 1 ahmad.saleem792 2022-08-20 11:59:27 -0700 https://wpt.fyi/results/css/CSS2/fonts?label=master&label=experimental&aligned&view=subtest&q=font-family-rule It is not WPT. Do we follow WPT only test for CSS2 or we do follow CSS2 original test suite? Test Link - http://test.csswg.org/suites/css21_dev/20110323/xhtml1/font-style-rule-001.xht 1892698 2 ap 2022-08-20 14:39:47 -0700 WPT is certainly not the source of truth in any way, but this test is on wpt.live: http://wpt.live/css/CSS2/fonts/font-family-rule-011.xht (note that it's "011", not "001" as above). That said, the test is fundamentally broken, because it relies on a "default" font being installed in the OS, which no one has, and Safari wouldn't use anyway (we only expose a fixed set of fonts, to reduce fingerprinting). Simon, do you remember how you observed the failure? Is this still actionable?