Yesterday, Elliott Rusty Harold draws his conclusions about the Semantic Web:
Digesting last week’s WWW2004 conference, I think I’ve come to two conclusions. The first conclusion is that the semantic web as envisioned at the W3C (RDF, OWL, URIs) is hype. Nobody is actually using it to accomplish anything useful. It’s of great interest to theoreticians but has little to no practical impact, and is not likely to have any for the foreseeable future.
Wow — that’s not really an encouraging statement ;-) I’m not yet decided on what I think about the SemWeb; I’m torn between deep skepticism and slowly mounting interest. Today, Elliott is praising XForms (his second conclusion). Finally, something about which I don’t have an opinion about at all. Yet :-)
“Case studies that sounded good and useful, and indeed were. However, when I actually googled my way into the projects’ web sites and looked at the code, the projects proved to be using little to no semantic web technology. A few had rdf:about attributes stuck somewhere in a mass of plain XML. Some didn’t even have that. Any semantics these applications had was based purely on element names and namespace URIs. RDF tuples were an afterthought, if indeed they were present at all.”
Erm, duh! Most of my RDF looks like XML with rdf:about sprinkled about too. That’s a feature! And while some SemWeb folks like to see their triples, I personally like mine hidden, and looking like vanilla XML (which is why I got the XML serialization tweaked to make rdf:RDF optional). Geez, I’d expect more from ERH. That’s supremely ignorant.