W3C home > Mailing lists > Public > www-style@w3.org > November 2008

Re: CSS3 @font-face / EOT Fonts - new compromise proposal

From: Håkon Wium Lie <howcome@opera.com>
Date: Fri, 14 Nov 2008 19:54:06 +0100
Message-ID: <18717.51534.240440.600347@opera.com>
To: "Tab Atkins Jr." <jackalmage@gmail.com>
Cc: "Thomas Phinney" <tphinney@adobe.com>, "www-style@w3.org" <www-style@w3.org>

Also sprach Tab Atkins Jr.:

 > Getting *any* support for font-linking is the goal.  Getting an
 > acceptable form of support for commercial font-linking is bonus points.
 > This is why many of us on the list are so reluctant to agree to anything.
 > Well, it's one reason (the other is because we like being contrary).  We can
 > live without commercial fonts.  Commercial fonts can live without us, at
 > least for a while still.  Bringing the two camps together would be
 > wonderful, but we are not so desperate for such a meeting to agree to things
 > that are ultimately harmful to us.

I agree with this description.

That being said, I believe there's a narrow window for a successful
compromise. All browsers could possibly be convinced to support:

 1) the same-site restrictions Mozilla has proposed. That is, browsers
 would only allow a certain font to be used on pages on the same site,
 unless an HTTP header (Access-Control-Allow-Origin) says otherwise.

 2) a light-weight obfuscation/compression scheme. This could either
 be XOR-ing a few bits in a strategic place to hide the font to common
 systems (obfuscation), or introducing a targeted compression scheme.
 The compression should be based on an algorithm with no known claims.
 This could be MTX, but it currently seems safer to reuse gzip. The
 compression would be applied selectively to chunks inside the file.
 (If gzip is applied to the whole file, it is likely to be unzipped
 automatically at the HTTP level.)

 3) linking to standard TTF/OTF files (as Safari/Mozilla/Opera/Prince
 has implented)

Combined, the first two of these seem to address Adobe's requirements,
as expressed by Thomas Phinney:

 > We just want the original completely unprotected font converted to
 > a marginally-more-protected web font *by the end user we licensed
 > the font to*, prior to them sticking said font on a web server.

              Håkon Wium Lie                          CTO °þe®ª
howcome@opera.com                  http://people.opera.com/howcome
Received on Friday, 14 November 2008 18:55:27 UTC

This archive was generated by hypermail 2.3.1 : Monday, 2 May 2016 14:27:41 UTC