Bugzilla

For the record: historically, we failed these tests because nsBulletFrame was rendering these "bullet" types. Now that we've fixed that and are using generated content (bug 1542807), we still "fail" them (before this patch) because we use the built-in `-moz-bullet-font` font for them but the reference expects the `::marker` parent's font. (see https://github.com/w3c/csswg-drafts/issues/6200 for reasons behind our decision to use a built-in font for these particular `list-style-type`s)

So this patch makes the test pass in Gecko, but arguably these tests are bogus since there is no normative rendering for `list-style-type:square/circle/disc`. CSS2 says: "Glyphs are specified with disc, circle, and square. Their exact rendering depends on the user agent."
https://www.w3.org/TR/CSS22/generate.html#list-style

For the record: historically, we failed these tests because nsBulletFrame was rendering these "bullet" types. Now that we've fixed that and are using generated content (bug 1542807), we still "fail" them (before this patch) because we use the built-in `-moz-bullet-font` font for them but the reference expects the `::marker` parent's font. (see https://github.com/w3c/csswg-drafts/issues/6200 for reasons behind our decision to use a built-in font for these particular `list-style-type`s)

So this patch makes the tests pass in Gecko, but arguably these tests are bogus since there is no normative rendering for `list-style-type:square/circle/disc`. CSS2 says: "Glyphs are specified with disc, circle, and square. Their exact rendering depends on the user agent."
https://www.w3.org/TR/CSS22/generate.html#list-style