Bosworth's Web of Data
Subject:   Misleading remark
Date:   2005-04-23 01:29:10
From:   danja
"Contrast this with the Semantic Web, which requires that you get a large group of people to agree on the schema of everything." - that's factually incorrect, and seriously misleading.

The Semantic Web technologies (RDF, OWL) allow anyone to create their own schemas. But unlike XML schema languages and for that matter SQL schemas there doesn't need to be extensive or agreement, certainly not on "everything". The SW languages make it possible to work with limited/partial agreement and understanding without being sloppy.

RSS 2.0 and Atom may offer useful, simple communication formats, but neither defines any model outside of the core (content-oriented) language.

You can get away with, even benefit from a lot of sloppiness when the (human) end user is interacting directly with the data - i.e. dealing with content such as that delivered by HTML or RSS 2.0. But when you start talking about machine-processable data, that sloppiness needs to be manageable - do you really want Amazon to be sloppy with your credit card details? Exactly because agreement on everything is *not* required by Semantic Web technologies, they are currently probably the most promising approach to adding machine-oriented data facilities to the Web.