On a lark, I decided to start lurking in public-web-http-desc@w3.org. I’m interested to see how this discussion evolves. Specifically, I’m most interested to see what scenarios they specifically decide not to support.

Standing up and saying “we’re not going to do X” is an important step in the development cycle. I think XSD would have been better off if this had happened more often during its infancy.

Description languages are models, and as such they are subject to the tradeoffs involved in modeling. Specifically, they must find the right balance between generality, efficiency of representation, and precision (the ability to provide high-fidelity renderings of the concepts they model).

In my thinking the ideal result would be an efficient and precise rendering of a small number of concepts specific to a small number of critical scenarios. The tendency with these things is to go super-abstract and over generalize, which leads to models that are neither efficent or precise.

I guess we’ll see what happens.