Raw

Another marvellous piece of heresy from Mark Nottingham: The Document in Document-Oriented Messaging. He analyses the modelling aspects of XML, noting the good points, then gets his teeth into XML Schema and the Infoset.

Some classic lines:



Why do Web services folks think it’s OK for end users to use XML Schema if it isn’t good enough for describing WSDL?

…

However, XML was made for document markup, not data modelling. … The result is — for data-oriented use cases — a complex data model not designed for the task at hand being described by a sub-optimal constraint language.



I think his description of the problems are pretty accurate, what I find a little worrying is that his conclusions for future work are that either profiling and subsetting XML Schema and the InfoSet; or starting fresh and/or changing horses. There is already a WG for the first, apparently. By changing horses he means using RDF and OWL. Woo-hoo! I hear you me say. But it's troublesome that he characterises it as a whole different stack.

I'm more optimistic that a merging of horses is possible, that there is common ground. If there isn't enough common ground then there's a lot more trouble ahead - the WS brigade pushing forward with an inadequate but Baroque model and RDF folks wondering why everybody's so obsessed with syntax they won't use their lovely graphs.

Oh yeah, that's *now*, isn't it…

(spotter: Andrew Newman)

PS. On a moderately related note, here's a good post on KISS and WS-* from Adam Bosworth

[Danny]