- From: Ben Adida <ben@adida.net>
- Date: Wed, 27 Aug 2008 19:45:23 -0700
- To: Ian Hickson <ian@hixie.ch>
- CC: Dan Brickley <danbri@danbri.org>, Henri Sivonen <hsivonen@iki.fi>, "Bonner, Matt" <matt.bonner@hp.com>, 'Julian Reschke' <julian.reschke@gmx.de>, "'Tab Atkins Jr.'" <jackalmage@gmail.com>, "www-archive@w3.org" <www-archive@w3.org>, "Ralph R. Swick" <swick@w3.org>, Dan Connolly <connolly@w3.org>, WHAT-WG <whatwg@whatwg.org>
Ian Hickson wrote: > On Sun, 24 Aug 2008, Dan Brickley wrote: >> I'm sure a small group of these could be assembled to discuss an >> RDFa-for-HTML5 proposal, should one be made. > > I would be very interested in taking part in such discussions. I would > in particular like to see discussion of: > > * What the problem being addressed is. > > * What research shows that it is an important enough problem that it > should be addressed. > > * What the requirements are. So a fantastic UI tool has just been released: http://labs.mozilla.com/2008/08/introducing-ubiquity/ Take a look at the video. It's great work by Mozilla. (And I have nothing to do with it, so I take no credit.) Consider specifically the Craigslist example, where the user selects a few of the apartments and says "map these." Clearly, and as the voice-over states, the site needs embedded metadata that easily connects "what the user is pointing to" to the structured data required for mapping. It's also really useful if the resulting map, when pasted into an email or another web page, carries with it the same structured data that was there in the first place, so that the newly formed data can still be parsed and interpreted by the tool. So, that's why you want embedded metadata right in the HTML, closely linked to the human-visible rendered text. Now, why RDFa instead of, say, microformats? Consider how you would go one step further than the video, i.e. "map these apartments and color the map pins according to price." Is there a microformat that combines location and price? Not that I know of. And that's because the granularity of microformats is a *set* of data fields for a given data type. If you want to be able to mix and match concepts like "location" and "price", you need the granularity of RDFa: one field at a time. Finally, take a step back: we want this "map+price" functionality to be enabled on more than just Craigslist. With RDFa, *any* publisher can add as many data fields as they want, discovering fields that others use, following the URIs to uncover the exact meaning of a data field, and putting together the structured data that they know will be immediately interpretable by an automated tool like Ubiquity. Instead of site-specific parsers, tools like Ubiquity can just parse RDFa and get exactly what they need. This is great work by Mozilla. I think RDFa would make this application (and others like it) even more powerful. -Ben
Received on Thursday, 28 August 2008 02:46:04 UTC