Re: [1.2T LC] WAI-ARIA use of @role as discussed under 'extensible metadata attributes' (ISSUE-2074)

Hi, Al-

Al Gilman wrote (on 10/9/08 9:25 PM):
> aside:  I still don't see how in one breath you can say that use with
> RDFa requires all @rel/rev values to be CURIEs and in the next breath
> say you can mix and match RDFa and microformats.

Well, it's possible that my approach is too loose, though I think it's
sound.  I'd like to explain it, so if I'm making a faulty call, someone
can let me know.

A CURIE can be a prefixed or prefixed value, so it has to allow for bare
strings to be used.

A microformat is just a bare string.

While an RDFa processor may not recognize a microformat value string, it
would not reject it because of its format... it would just not assign
meaning to it.

AFAICT, microformats don't care what other values they share the
attribute with (cf, @class, which can mix CSS class names with
microformat keywords).

I don't see where a conflict would necessarily arise unless there were
overlaps in the strings.

To SVG processors, they are all just spaced lists of strings, so SVG is
agnostic to the rules imposed by other formats.

> Rough hack at rewrite:
>> rel/rev values may be CURIEs indicating concepts identified by IRIs
>> after the manner of RDFa, or may be given definitions using the
>> techniques of microformats.
> On the other hand, as far as it affects the discussion of WAI-ARIA in
> the extensible metadata attributes paragraph, it all looks fine to me.

Okay, thanks.


Received on Friday, 10 October 2008 02:49:36 UTC