Matt Sergeant wrote: > At it's core, XML is a very elegant syntax for defining a rich dataset > of nodes
It's a syntax for defining a dataset of nodes that all conform to XML's ideas about what a dataset of nodes looks like. I'm not convinced about rich or elegant. :-) > > and you find yourself tied down to only generating valid > > XML (which few real world web pages are, even if they should be). > > My web pages come out as HTML 4.0 transitional, not XHTML, and they are > generated with XSLT. They often omit closing tags where appropriate for > HTML. I don't do any fancy post-processing. Sorry Andy, but the above is > not true at all. OK, fair enough, you can generate output that contains invalid XML markup. But you can't write a source template that includes invalid XML as literal text. At least not without escaping all the < and > characters and other such madness. However, I accept that this is construed as a feature in these kind of systems. I just happen to prefer a slightly different kind of system. A