RDF, simplicity, and so on

Danny posted, I wrote a mini rant, and it's probably worthy of archiving so I can enjoy my own hubris later.

I understand description logics. I understand most of RDF's theoretical underpinnings, from the base language up to OWL-DL. (I even have a fair grasp of some of the programming language stuff that scares Danny — it scares me, too! — being a Lisper and a Lambda reader.)

I've used Racer, I've used Protégé, and I write my ontologies to work under inferencing systems.

But I mostly use Wilbur, which does no reasoning; the RDF I write and use is as simple as possible (no simpler), and works. (This does not mean I don't bear the rest of the stack in mind, but that's because I have bothered to understand it.)

No, the reason I use RDF is not for the inference — neat though it is — or the rest of the stack. It's because doing the things I'm doing now, any other way, would be more difficult. In fact, to get Greenspun on people's asses, I'd end up re-inventing RDF. I don't like that much effort.

Maybe I'm just lazy enough?

The ‘stack’ doesn't bother me. I use what I need to use — i.e., very rarely do I use rdf:Bag! — because it saves me building my own general relational framework based on URIs and literals. Relax, people!

Posted at 2005-09-19 14:27:19 by Richard • Link to RDF, simplicity, a…
Comments, trackbacks.