Mark Birbeck wrote:
I have to say though, that I think all of this is going to come back
and bite us, just like the issue of not generating triples if @class
is non-prefixed. I still favour some kind of switch that can turn some
of these features on and off, since it seems to me just as legitimate
to deal with these issues at the level of some triple store, as it is
to constantly tweak with the syntax.
I personally still think that using @class at all is a huge mistake.
The (new) HTML Working Group appears to be making that mistake too;
adding semantics to class values where none previously existed. We are
introducing new functionality, and it should be coupled with a new
attribute. e.g. @role. Or, if we don't like @role, something else. I
know that we don't like having people duplicate things, but seriously...
principle of least surprise. Alternatively, use a profile value to mean
"this page wants to create RDFa triples using profile X" were X might be
"use role for this" or "use class for that". Dunno. But at least that
way it isn't by accident.
--
Shane P. McCarron Phone: +1 763 786-8160 x120
Managing Director Fax: +1 763 786-8180
ApTest Minnesota Inet: [EMAIL PROTECTED]
