Re: .webfont Proposal

On Wed, 2009-07-08 at 14:57 -0700, John Daggett wrote:
> > Note, though, that the proposal says that mismatches still allow the
> > font to be used, but suggest that UAs may offer an unobtrusive alert
> > about the mismatch.
> 
> Ah, so this is root strings with "warn on load" rather than "refuse
> display on load".  No quite sure the user is the one who needs warning. 
> The maintenance headaches and need for web infrastructure changes are
> still there.


There is emerging precedent within W3C and
implementation practice for something very
close to "warn on load".   See section 7.3
of this:

http://www.w3.org/Submission/2008/SUBM-ccREL-20080501/

That's the ccREL submission from the folks at
Creative Commons.

I don't think that browsers should be checking
rootstrings and guessing as to their legal interpretation:
browsers should just make it easy for users to know that
there is meta-data attached and to view that meta-data.

That's why my "wrapper" proposal suggests a two-part
MIME file with one part being a binary font file, the
other being an HTML "about" document.   Such an "about"
document can use  RDFa:

http://www.w3.org/TR/xhtml-rdfa-primer/




-t