RDFa

WWW2006 is barely over, and folks are writing about RDFa. This is great news. The enthusiasm and inevitable questions and concerns are very encouraging, and we hope to address these as quickly as we can.

Check out Benjamin Nowack’s careful comparison of RDFa to other metadata embedding techniques. The one downside he points out: RDFa does not currently validate in XHTML 1.0. That’s true, but we had no other choice to achieve the features we need:

• independence: you pick your vocabulary of metadata terms
• modularity: you can reuse other vocabularies
• evolvability: you can change the meaning of your vocabulary over time (yeah, RDF!)
• DRY: don’t repeat yourself, if you render the data in HTML, why repeat it in the structured form?
• in-context: the metadata is right next to the data it describes, and if you copy and paste the HTML, you get the metadata along with it.

But fear not: RDFa does not break anything in today’s browsers. Not even in XHTML strict rendering mode. That’s a big deal, and it means that our upcoming effort to standardize RDFa for XHTML1 is going to be fairly straight-forward.

One significant advantage of RDFa: you can express metadata about other documents (embedded images) and about fragments of the document (a blockquote). And another significant advantage: you can aggregate content from multiple sources, and the metadata schemas will not conflict, thanks to XML namespaces.

Meanwhile, Evan Prodromou compares RDFa to Microformats. He seems worried that the disagreements will be problematic. We don’t think so. Microformats are useful for expressing a few, common, well-defined vocabularies. RDFa is useful for letting publishers mix and match any vocabularies they choose. Both are useful.

And since RDFa is more generic than Microformats, we have a proposal for transforming Microformats to RDFa. We’re still debating this, but it’s very promising. In fact, our calendar bookmarklet uses this technique to read both iCal RDFa, and the hCal microformat.

A question came in the mail regarding the fact that the W3C validator currently fails on RDFa. This is true, and it will likely remain true until we finish the RDFa specification for XHTML1. RDFa does involve changing the XHTML schemas, and the plan is under way to do just that.

That said, it’s important to note that, for most RDFa use cases, only the about, property, and content attributes are new. It is currently under debate whether these extra attributes should indeed cause the validator to fail, and all browser implementations we have seen are perfectly happy to accept extra attributes, even in strict, XHTML mode, without any change in rendering.

In other words, prototype away. Validation will work soon. In the meantime, nothing else will break.

It’s been a busy couple of weeks for RDFa, with 4 different talks. All the slides are now posted on this site. Enjoy!

In addition, if you’ve got budding RDFa implementations, examples, or thoughts, please don’t hesitate to contact us via our individual contact information (see BlogRoll in the right sidebar). We’ll be happy to post your work on this site for everyone’s benefit.

For semantic web folks, RDFa is just another way to write RDF. And for many people, RDF should be written in N3. Okay, fair enough.

Introducing: the N3 extraction bookmarklet. On any HTML page, click the bookmarklet: you’ll get an N3 extraction of the RDFa on the page. You’ll even get this with the right MIME type, text/rdf+n3. In addition, the page content stays local, which is important for privacy. You can find this bookmarklet at our bookmarklet page.

This has only been tested on Firefox/Mozilla so far. It doesn’t work on Safari, yet, and hasn’t been tested on IE or Opera. But we’re working on it!

(At the same time, we fixed a small bug in the RDFa javascript library, regarding the display of compact URIs. This fix will automatically trickle down to all other bookmarklets when your cache expires. No functionality change, just a bug fix, so the URLs for the bookmarklets stayed the same.)

The RDFa bookmarklets are now available, hosted at W3C for stability. Apart from bug fixes, these will not change, so you can rely on them for demos, etc…

So what do these bookmarklets do? Well, they can extract calendar data using the RDF calendar vocabulary, and add Google Calendar and ical links. They can highlight all the metadata in an HTML page. They can let you blog Creative-Commons-licensed content in WordPress with automatic carry-over of the attribution information — e.g. “quote taken from Smith.”

And there’s a HOWTO for building your own bookmarklet that looks for RDFa properties you’re interested in.

I gave a lightning talk at the W3C AC Rep meeting this past Sunday. I had 3 minutes, but the response was fairly positive. You can check out the slides.

We’re starting to put together the web site for RDFa. Bookmark this URL, subscribe to the RSS, we’ll keep you posted. For now, the first thing you should do is check out the RDFa Primer.