Ben Adida wrote:
Henri Sivonen wrote:
I suggest changing RDFa to use full IRIs instead of CURIEs.
I appreciate you putting a concrete proposal on the table.
I believe it is rather arbitrary and in conflict with existing uses in
HTML4. First, the use of IRIs in @rel is *quite* new, I know of no major
(or even minor) vocabulary that's deployed this way.
http://philip.html5.org/data/link-rel-rev.txt has a few @rel values
starting with "http://";. One is:
<link rel="http://schemas.google.com/g/2005#feed"; ...>
seemingly defined in
http://code.google.com/apis/gdata/docs/2.0/reference.html , though
that's Atom and not HTML. The other cases all look like people got
confused by the difference between @rel and @href. So it does seem
reasonable to say that intentional use of IRIs in @rel is unusual.
(The use of prefix-colon values also seems very rare. By far the most
common is to just make up a string ("nofollow", "apple-touch-icon",
"dropmenu1", "SQDynButton", etc).)
And finally, microformats suggest using @profile to interpret their
@rel's and @class's.
http://philip.html5.org/data/profile-values.txt indicates that pretty
much nobody (from a roughly random selection of pages across the web)
uses @profile, except for the one value that's in the default Wordpress
template. About 130 of the pages use class="vcard" but only 4 seem to
indicate it in @profile. So the Microformats suggestion to use @profile
appears to be largely irrelevant in practice.
--
Philip Taylor
pj...@cam.ac.uk
