Re: Quick WOFF 2.0 update

You are correct John.  The @font-face src format hint provides the
mechanism for sites / web font services to simultaneously serve both WOFF
1.0 and 2.0 files, and for user agents to request the preferred format
supported by that user agent.

For a complete example, see Amiri which we are serving via Early Access (
http://google.com/fonts/earlyaccess) in both formats.

http://fonts.googleapis.com/earlyaccess/amiri.css

/*
 * Amiri (Arabic) http://www.google.com/webfonts/earlyaccess
 */
@font-face {
  font-family: 'Amiri';
  font-style: italic;
  font-weight: 400;
  src: url(//
themes.googleusercontent.com/static/fonts/earlyaccess/amiri/v2/Amiri-Slanted.eot
);
  src: url(//
themes.googleusercontent.com/static/fonts/earlyaccess/amiri/v2/Amiri-Slanted.eot?#iefix)
format('embedded-opentype'),
       url(//
themes.googleusercontent.com/static/fonts/earlyaccess/amiri/v2/Amiri-Slanted.woff2)
format('x-woff2'),
       url(//
themes.googleusercontent.com/static/fonts/earlyaccess/amiri/v2/Amiri-Slanted.woff)
format('woff'),
       url(//
themes.googleusercontent.com/static/fonts/earlyaccess/amiri/v2/Amiri-Slanted.ttf)
format('truetype');
}
:

Note, current format hint for the candidate WOFF 2.0 is "x-woff2".  As we
progress, we expect this to quickly become "woff2".

Thank you,
David



On Tue, Apr 2, 2013 at 2:03 PM, John Hudson <tiro@tiro.com> wrote:

> When the notion of WOFF 2.0 was first mooted, I believe some concerns were
> raised regarding compatibility issues and uninterrupted support of WOFF 1.0
> webfonts alongside the new compression model. I'd appreciate some comment
> on this (from David? Raph?) and explanation of how this is expected to play
> out, and what expectations this will put on both user agents and fonts.
>
> I am presuming that something like a metadata format version string will
> suffice to let user agents know which version of WOFF they are encountering
> with a given webfont. What about serving WOFFs to user agents with
> different levels of version support? What mechanism would enable a site to
> serve WOFF 2.0 fonts to one browser and 1.0 to another?
>
> I'm sure these are all resolvable issues. I'd just like to get some idea
> of the direction in which we're heading.
>
> JH
>
>