RE: css3-fonts: should not dictate usage policy with respect to origin

[John Hudson:]
> 
> Bjoern Hoehrmann wrote:
> 
> > http://lists.w3.org/Archives/Public/www-style/2011Jun/0476.html is the
> > original comment. It states the problem quite clearly and offers a so-
> > lution that's also quite clear. There is nothing about backward compa-
> > tibility in there, so you don't seem to have followed the thread very
> > well.
> 
> Since then, Glenn has gone on to make multiple arguments and suggestions,
> sometimes claiming backwards compatibility as his concern and sometimes
> forwards compatibility. His most recent argument has been that certain
> technical telecommunications specs reference early drafts of css-fonts,
> and hence should not be considered non-conformant if their implementation
> doesn't end up matching the final recommendation. Does that make sense to
> anyone?

The one reliable pattern I am able to parse is that we always go back to 
an expectation that is contrary to the w3c process. Namely that a draft 
specification establishes a normative level of conformance that future 
draft revisions must comply with so that one can implement today's draft 
and still conform at REC time. We know that expectation has no basis, w3c 
staff such as Liam have confirmed it, CSS specs explicitly state that it 
is inappropriate to even refer to these documents as something other than 
a work in progress i.e. only RECs are the basis for conformance. As such 
I think this misunderstanding is far more fundamental than css3-fonts and 
whether or how same-origin should behave for web fonts; it is not at all 
clear to me that resolving the latter to Glenn's satisfaction would in any 
way preclude more objections to future updates of other drafts these devices 
already do or will come to depend on. 

I am happy to discuss the feature on its technical merits. I think the 
overarching conformance issue is a larger, orthogonal impasse that needs 
resolution by w3c, Samsung and the industry group that has taken these 
dependencies.

Received on Wednesday, 29 June 2011 21:25:02 UTC