Re: Agenda, action items and suggested WOFF changes

Richard Fink wrote:

> At the moment, I don't see what any language about the bits would help
> clarify. I don't see a compelling need to run the risk of varying
> interpretations ...

Existing varying interpretations are exactly what language about the 
bits would help clarify.

 >... which, as we've seen with the language in the OT spec, can
> all too easily be a consequence when a technical spec tries to step outside
> technical boundaries.

The TT/OT embedding bits sought to provide a *technical* measure to 
restrict embedding of fonts in document and installation of fonts on 
recipient systems. It's probably debatable whether this constitutes 
'stepping outside technical boundaries'. In any case, I think the 
problem with regard to varying interpretations isn't a symptom of such a 
step -- there may easily be varying interpretations of technical 
specifications --: it's a symptom of unclear language in the OT spec 
and, critically for the context of web fonts, the explicit conflation fo 
web serving and document embedding in the EOT spec (indeed, the 
duplication of the font embedding bits in the EOT wrapper).

Perhaps it will help to see the kind of clarification I am proposing for 
the WOFF spec in light of the EOT precedence. If EOT had not explicitly 
asscoiated the embedding bits with web fonts, I doubt if it would seem 
important to clarify whether they relate to WOFF. I'm perfectly content 
if this clarification consists only of a statement such as this:

	User Agents MUST NOT check the level of font
	embedding permissions set in a font delivered
	using WOFF format. In general, it cannot be
	assumed that embedding permissions in the
    	font OS/2 table fsType field correspond to
	licensing of fonts for use on the web. Web
	authors are therefore expected to have made
	adequate efforts to make sure that the font
	license corresponds to the intended use.

i.e. all embedding bits would be explicitly irrelevant to WOFF in the 
context of WOFF file creation and user agent display. [The context of 
printing from a web page to a PDF, or similar processes, is a different 
matter, since this constitutes document embedding as now commonly 
understood.]

JH

Received on Friday, 14 May 2010 03:07:52 UTC