Does the web need RDFa?

16 Oct 2008

After over four years at the W3C, RDFa (nee RDF/A, nee RDF/XHTML) is now a full recommendation, which to all intents and purposes means 'it's a standard'. Amidst the noise of the fireworks and dancing, let's have a look at exactly what RDFa allows us to do, and why it was created.

Two axis

One of the key ideas in RDFa is that a document contains two 'axis' of information. The first is information about the document itself, such as the fact that it has div's and span's, headers and sections, colours and font- sizes, and so on.

The second set of information relates to the content of the document; a page might be a recipe, and this section contains the ingredients. Another page might be a user's profile, and this section contains the person's address and phone number.

Information in the first category can be marked up with a standard based on a single attribute, called @role. This allows us to indicate that some div or span is actually playing the role of a menu or a check-box, something that is crucial in today's Ajax-driven world. (ARIA is a standard for defined terms like 'drop-box', 'menu', and so on, which makes use of @role. It is finding support in IE8, FF3, and Ajax libraries such as YUI provide ARIA support, too.)

Information in the second category though, is much trickier. A popular way to spell out what some content 'means' has been to use the HTML @class attribute. This is quite a neat solution, but the big problem is that @class has been overused. In fact, it's been over-overused, since it bears the weight of describing the purpose of some markup (menu, footer, etc.), the semantics of some markup (address, ingredient, stanza, etc.), and the behaviour of some markup (red, black background, underlined, etc.).

@role was created to take some of the strain off @class, and to allow authors to indicate the purpose of some markup (menu, footer, etc.), and