The Guardian takes a look at where we are with Tim Berners-Lee's semantic web. Conclusion? Not very far given the promise of the idea. So far.
In the article there seems to be some kind of aha! expressed by TBL - enough theory already!
"Berners-Lee did concede that as with the world wide web, the semantic web should "serve useful stuff". "One of the problems we've actually had with the semantic web, I only recently realised, is we haven't been doing that."
Not enough useful RDF data has been left online, he explained: "The whole value-add of the web is serendipitous re-use: when you put it out there for one person, and it gets used by who-knows-who. We want to put data out there for one purpose, then find it gets linked into all kinds of data. And that's been not happening, because we forgot 'serve useful stuff', not to mention 'make useful links'."
'More doing, Less theorizing' is TBLs message. Danny Ayers has been urging this too with some passion recently.
What is missing from the semantic web evangelists toolset are the cool demos with working code and easy-to-grokk tools showing and proving end-user value. This SPARQL scheduler gets close, but it just doesn't let the potential shine through. The microformats / live clipboard demos (using real, working code(!)) gets close. Absolutely a step in the right semantic web direction, but we need many more of these to get the excitement really going.
For things to really take off, you (the semweb heads reading this) need to thrill a critical mass of geeks and get them working on this stuff with you. And simplify. Simplicity is how the web went from small to huge in a very short time. Time for a Repeat.
Simply stated: I still haven't seen the semantic web killer Demo - you know, the one that I watch go: 'oh yeah! I really get it now!'. It's 'out there' somewhere, but in somebody's head. That's no good.