Re: long-range datatyping and rdfa/microdata


On 8 Jun 2011, at 17:02, Dan Brickley wrote:
> This can be used in RDFa as so: <p>blah blah <span
> property="foaf:age">39</span> blah</p>.
> If we try to persuade publishers to put datatype="xsd:integer"
> alongside each age, ... we'll have a hard time. So is there anything
> we can do at the schema level?  Mumble mumble range mumble...

I see this mostly as a syntax problem, not a schema problem.

Realistically, I think the correct answer is to bury RDF's datatypes under a mountain of syntactic sugar. Turtle does this quite a bit and it works well. With JSON we can do it to the same degree. RDFa didn't attempt it, but rather tried to align the surface syntax closely with the abstract syntax, which I think in hindsight (20/20 etc) was a lost opportunity.

Speaking of microdata. They said: “Forget prefix mappings, we'll just use short URIs.” That's a winning formula. Two years from now, the rdf:, rdfs:, and xsd: namespace URIs with all that clutter inside will feel like they were designed by a deranged madman. We should pave the road so that a future WG can unify them as something like <> or even <>.

To recap, if someone wants to state their age in RDFa, it is:

     prefix="foaf: xsd:"
     <span property="foaf:age" datatype="xsd:integer">39</span>

But if the RDFa syntax had turned out a bit differently, and the FOAF project had chosen a different namespace URI, it could well have been (in HTML+RDFa):

     <span integer property=>39</span>

Or perhaps the RDFa WG and the FOAF project can still make it so. At any rate, it doesn't look like a model problem to me.


Received on Thursday, 9 June 2011 08:50:05 UTC