- From: Simon Reinhardt <simon.reinhardt@koeln.de>
- Date: Tue, 10 Jun 2008 23:21:00 +0200
- To: public-lod@w3.org
Michael Smethurst wrote: > Pretty soon (honest) we'll be putting RDF views of the data live Already looking forward to it! > - if we have full RDF, should we also be using rdf-a? Yes, because ... > - if so what would be the benefits (yahoo search stuff?)? Having the RDF in a separate file is cool and a good start, but if you rely on that then you create a "shadow web", as Ian Davis put it. You're missing the context for the data. While the data can still be used by crawlers and other bots, the user can't get to it from the page they're looking at. Well, they can follow a <link> element if the browser shows them it's there. But they can't e.g. click on something they see on the page and import the data behind it into another program. With RDFa they can! You could write scripts which highlight the data and which put it into a "semantic clipboard". On Ian Davis' post on the shadow web I even pointed this out giving TV programmes as an example: http://iandavis.com/blog/2007/11/is-the-semantic-web-destined-to-be-a-shadow#comment-433 - so you guys are making me really happy with your work. :-) I hope other channels will follow your lead. > - and what would be the costs (ham-fisted html hackers wrecking the > templates springs to mind ;) )? You have to be careful that the triples which result from the RDFa always match the data in the RDF/XML file (or whatever format you use), especially when you do content negotiation. That's the downside of this redundancy. And I think I've heard microformats people complain that RDFa has a problem with CSS or can't be styled as easily as microformats but I can't really see why. But that's the microformats vs. RDFa conflict anyway, nothing to do with using RDFa in combination with other RDF formats. I don't think the costs (as in performance costs) would be higher. Regards, Simon
Received on Tuesday, 10 June 2008 21:20:28 UTC