- From: Dan Brickley <danbri@danbri.org>
- Date: Wed, 7 Mar 2012 13:45:36 +0100
- To: Sandro Hawke <sandro@w3.org>
- Cc: Pat Hayes <phayes@ihmc.us>, RDF-WG Group <public-rdf-wg@w3.org>
On 7 March 2012 13:31, Sandro Hawke <sandro@w3.org> wrote: > Isn't this stuff handled just fine with subproperties? This seems much > simpler than the general n-ary or time-parameter stuff, because it is a > subproperty -- every home-phone is also a phone -- unlike to time > stuff. (It doesn't hold that every address-last-week is an > address-for-all-time.) Part of the problem here is workflow. Is it an anonymous subproperty? Or one whose URI is in the same doc as the message/doc? Or does the authoring process include creating e.g. a purl-like URI on a more stable service? Interesting Q is how it looks in RDFa. In some ways things are easier since we can mention the super-type alongside Inside some markup describing Sandro, in RDFa 1.1 we might have <span rel="foaf:phone uuid:1234123:homeFaxColor ... ...but we'd also need the info declaring the subproperty somewhere too; either inline or in the Web. I'm not comfortable with too much modeling leaking into properties-of-properties, but it seems appealing to allow at least attribute value pairs. Hard to know where to draw the line. Dan
Received on Wednesday, 7 March 2012 12:46:20 UTC