- From: Toby Inkster <tai@g5n.co.uk>
- Date: Thu, 13 Oct 2011 19:35:05 +0100
- To: Toby Inkster <tai@g5n.co.uk>
- Cc: "Michael Steidl (IPTC)" <mdirector@iptc.org>, public-rdfa-wg@w3.org
On Thu, 13 Oct 2011 17:13:48 +0100 Toby Inkster <tai@g5n.co.uk> wrote: > <p rel="rnews:creator"> > This picture was taken by > <a target="_blank" typeof="rnews:Person" > rel="rnews:homepage" href="http://www.riecks.com/" > property="rnews:name">David Riecks</a>. > </p> By the way, I should point out the fact that the snippet above takes advantage of the remarkable and somewhat intuitive situation that the following two links generate different triples: <span about=""> <a rel="" href="target" property=":title">Foo</a> <a href="target" property=":title">Foo</a> </span> In the first link sets a title for the page the links were found on. The second link sets a title for the target of the link. The presence of @rel (even empty) has a surprising effect. It all comes down to steps #6 and #7 in the RDFa processing sequence which include different techniques for establishing the subject of a triple depending on the presence of @rel/@rev. (If you think that's confusing, in earlier drafts of RDFa 1.0 it was a lot worse. The situation depended on whether the values in @rel/@rev could be expanded to valid URIs or not. See, e.g. 2008-02-21 working draft.) If we were chartered to make more backwards incompatible changes... -- Toby A Inkster <mailto:mail@tobyinkster.co.uk> <http://tobyinkster.co.uk>
Received on Thursday, 13 October 2011 18:34:27 UTC