Re: [css3-fonts] unicode-range and unicode normalization from John Daggett on 2010-07-12 (www-style@w3.org from July 2010)

From: John Daggett <jdaggett@mozilla.com>
Date: Mon, 12 Jul 2010 01:06:33 -0700 (PDT)
To: fantasai <fantasai.lists@inkedblade.net>
Cc: Yuzo Fujishima <yuzo@google.com>, www-style@w3.org, www-font <www-font@w3.org>, Anne van Kesteren <annevk@opera.com>
Message-ID: <1173010099.59059.1278921993011.JavaMail.root@cm-mail03.mozilla.org>

fantasai wrote:

> > I don't think we currently require normalization anywhere else. In fact,
> > there is probably JavaScript code out there that assumes nothing is
> > normalized.
> 
> That's true, but font-matching is not quite as simple as matching
> string characters against a set of available Unicode characters.
> Each grapheme cluster needs to be rendered entirely in a single
> font. If different characters in the grapheme cluster map to
> different fonts, it needs to be clear how to resolve to a single
> font.

Precisely.  This boils down to H&aring;kon and Ha&#x30a;kon displaying
consistently, something in which Opera has a vested interest!

  http://people.mozilla.org/~jdaggett/tests/aringtest.html

Opera/Chrome display the text correctly on Windows for MS Gothic and
Trebuchet MS, FF3.6/IE9 do not.  Opera doesn't display with Osaka correctly
on OSX 10.5.

John

Received on Monday, 12 July 2010 08:07:08 UTC