May 30, 2006
Semantic web - conflict or synthesis?
An interesting debate is taking place on the shape of semantic markup in HTML. We've had microformats for a while now, although actual use is very limited. We also have W3C's RDFa.
I quite like Microformats. I also quite like RDF. The combination looks very interesting - however a popularity competition between the two looks like an expensive and morale draining waste of time. Luckily there is some sensible discussion going on, and perhaps there really is a way to move forwards without requiring an uncomfortable dislocation.
I think its quite clear now that HTML 4, and ill-formed XHTML 1.0, are going to be around for quite some time, so a backwards-compatible approach for RDFa would be more sensible than requiring new HTML markup from XHTML 2.0. I also think that Microformats aren't going to get much further without namespaces - its already getting pretty confusing when there are several different formats within the same page.
One thing I'm not sure of, however, is page semantics. At least when you subscribe to a feed you're pretty sure what you're subscribing to. Just pointing a semantic aggregator at a site root and saying "go figure it out" doesn't seem realistic. You need to at least hint that "here's where the event listing is, and here's the news page" - unless your site is just one big page. So I'm not sure here where the advantage lies in taking the embedded approach - yes, you do get a single target URL for both the browser and the metadata aggregator, but is that sufficient?
Just to try things out, a while ago I added support for both hCard and hAtom to this site, but I really haven't been able to see much by way of benefits, unlike when I first added RSS (and later Atom). We're still, after all this time, at a very early adoption stage.