W3C home > Mailing lists > Public > www-font@w3.org > July to September 2009

Re: Merits and deficiencies of EOT Lite

From: John Daggett <jdaggett@mozilla.com>
Date: Tue, 28 Jul 2009 12:37:02 -0700 (PDT)
To: karsten luecke <list@kltf.de>
Cc: www-font@w3.org
Message-ID: <10742468.76681248809822012.JavaMail.root@cm-mail02.mozilla.org>
> If I understand you correctly that this is a Uniscribe issue which affects
> 
> - other browsers that rely on Uniscribe (including Firefox)
>   in combination with
> - other web font formats too (did I get this right?)
>   then this is rather the opposite of "largely irrelevant to this
>   discussion".. For foundries, this is at least as relevant as 
>   "ouch, good to >   know".  :)

This is an orthogonal issue, which was roc's follow-on "oops" comment.

Dynamically loaded CFF fonts, currently supported in Firefox, Safari
and Opera 10 Beta 2, cannot be shaped using Uniscribe. Firefox is
affected more by this because we try to use Uniscribe by default to
get things like kerning and ligatures.  Other browsers only use
Uniscribe when needed for complex script rendering.  Because of this
bug, you'll see standard GDI rendering without Uniscribe for CFF
fonts, same as you see in Safari and Opera 10 beta which don't use
Uniscribe by default. Scripts that require shaping will fallback, as
John Hudson has noted.

Demo using a CFF font:
http://nicewebtype.com/fonts/graublau-sans-web/

> (And you robbed Mr Daggett of his core IE argument against EOT Lite.)

Sorry Karsten, this doesn't change the fact that EOT-Lite will not give
you backward compatibility with CFF fonts in IE and backward compatibility
is the main argument behind the apparent rush to support EOT-Lite.

This and other problems with Windows GDI rendering of CFF fonts
affects all browsers. We could all switch to something else just as
Safari gives users the option, not on by default, to render using
Apple's own font rasterizer. It would be far better for all involved
to fix the default platform rendering issues.
Received on Tuesday, 28 July 2009 19:37:43 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Saturday, 11 June 2011 00:14:03 GMT