Re: Font MIME type (was RE: EOT-Lite File Format v.1.1)

On Friday, July 31, 2009, 10:47:36 PM, Vladimir wrote:

LV> On Friday, July 31, 2009 3:59 PM John Daggett wrote:

>> We should probably register a MIME type for EOT-Lite (e.g.
>> application/font-eotlite).


LV> I would like to inquire the group if you see any need to apply to
LV> IANA for a top-level MIME type registration for fonts. 

I tried to do that, about a decade ago (1998 or so if I recall correctly), and met active resistance. Not "we are not interested" type argument but "we will strenuously oppose this and take up lots of your time if you persist".

Apparently the original MIME concept is still contentious and some parties, primarily those dealing with email rather than HTTP, think that adding audio and video was a real bad idea and strongly oppose any new top level types.

On the other hand, the model top level type was added later (primarily for IGES, which uses it, and VRML, which does not :) )

LV> The same
LV> question has been raised more than once in different organizations
LV> (in ISO/IEC SC34 and SC29, in OpenType forum) and since the number
LV> of different font formats and flavors (both current and
LV> historical) used by various applications now exceed two dozens, it
LV> would make sense to consider having "font/..." registered as a
LV> top-level MIME type (similar to image/... and other MIME types).

I agree, it would make sense.

LV> I know that in the past (circa 2003/2004) W3C Timed-Text group
LV> has made an approach to do that, and it failed mostly due to the
LV> lack of support; but the same issue has recently been discussed in
LV> both ISO/IEC SC29 and SC34 subcommittees, and together with ISO
LV> and W3C on board I believe we can submit a joint application and
LV> present a solid case why the top-level "font/..." MIME type is needed.

LV> Any opinions are welcome.

I would be happy to review any such proposal. I have been on the IETF-types mailing list since well before joining W3C, so I have some background there.

I would suggest looking at the proposal for the model top level type, to see what arguments were made for it being necessary.

I would also suggest a strong 'security considerations' section which notes that fonts can be an attack vector and (for some formats) can bring down an entire OS, not just a single application.

LV> Thank you,
LV> Vlad





-- 
 Chris Lilley                    mailto:chris@w3.org
 Technical Director, Interaction Domain
 W3C Graphics Activity Lead
 Co-Chair, W3C Hypertext CG

Received on Monday, 3 August 2009 11:29:43 UTC