- From: Danny Ayers <danny@panlanka.net>
- Date: Sat, 14 Apr 2001 08:56:29 +0600
- To: "RDF-Interest" <www-rdf-interest@w3.org>
Finally I've got a straightforward way of explaining what I've been rambilng on about with this process profiling business. Consider a college library. You can make a list of the resources in the library, for instance the book catalogues, there may be academic paper archives, whatever. To this list you can add other resources such as a video recorder, an overhead projector and a photocopier. You can go further and for instance organise a timetable of when these resources will be available. I would suggest that this is the current state of most metadata initiatives. But look at the library again - there is a whole set of resources that isn't being taken into account on the list - the human resources. Katy is a genius at cataloguing, Fred know how to use the overhead projector (and can teach you how to use it yourself if you like) and Simon makes a good cup of coffee. Now if I look at a catalogue of college resources I can not only find out where to get a book, I can also find somebody to make me a cup of coffee. I'm suggesting that if the processes of the web (web services, agents, inference engines, whatever) are described in the same language as the data resources (pages, feeds etc) then these can be reasoned about in exactly the same way. The Babel of agents and engines is reduced, directory services are simplified, advertising and discovery services (Jini etc) become more interoperable. This strikes me as a way of making life a lot easier all round. --- Danny Ayers http://www.isacat.net
Received on Friday, 13 April 2001 22:59:57 UTC