Last modified: 2007-08-21 23:28:45 UTC
Hello mediawiki team, I've found IPA characters in Konqueror (khtml - what about webkit - safari?) aren't rendered as good as in Mozilla Firefox. I looked to generated css files and saw definition: .IPA { font-family: "Chrysanthi Unicode", "Doulos SIL", Gentium, GentiumAlt, Code2000, "TITUS Cyberbit Basic", "DejaVu Sans", "Bitstream Cyberbit", "Arial Unicode MS", "Lucida Sans Unicode", "Hiragino Kaku Gothic Pro", "Matrix Unicode"; font-family /**/:inherit; } According to http://www.w3.org/TR/REC-CSS2/cascade.html#cascading-order 6.4.1 - rule 4 "if two rules have the same weight, origin and specificity, the latter specified wins." In this class, font-family of parent object overrides that nice definition of UTF8 IPA capable fonts. You can use !important keyword and change first line to font-family: "Chrysanthi Unicode", "Doulos SIL", Gentium, GentiumAlt, Code2000, "TITUS Cyberbit Basic", "DejaVu Sans", "Bitstream Cyberbit", "Arial Unicode MS", "Lucida Sans Unicode", "Hiragino Kaku Gothic Pro", "Matrix Unicode" !important; or define that list of ipa fonts as second one. Same problem is with classes Unicode, latinx, polytonic, mufi, lang(*) and several others. By the way, I used rule below for IPA UTF8 fonts. You can add this font to your list too, but I don't think it helps anybody. font-family: Junicode, FreeSans,FreeMono, FreeSerif, DejaVu Sans Mono, ClearlyU, Arev Sans;
Please tell the sysops of the appropriate wiki, e.g., at http://en.wikipedia.org/wiki/MediaWiki_talk:Common.css . Developers and sysadmins do not control this custom CSS. Incidentally, the behavior you note is intended. Internet Explorer 6 (I think not 7) is the only major browser that doesn't substitute in appropriate glyphs where they don't exist in the current font. Incidentally, it also doesn't parse comments correctly. Thus it ignores the second rule as malformed and uses the fonts specified, while more intelligent browsers are free to use their own selection rules because they correctly interpret the second rule as overriding the first. This is what's known as a "CSS hack".
It would be better specify IPA capable font for all browsers, not just for IE, then rely on magic substitution. KHTML and maybe webkit has problem with this too - not just internet explorer.