This interesting article on XHTML 2.0 up on the IBM developerWorks site got me thinking. For me the acid test for XHTML 2.0 will be how strictly it parses content and how reliable the renderings are across the various browsers that are out there today.
I know a guy who works on a fairly popular web-browser at the layer between the abstract model that sits on top of it and he pointed out that the W3C really hasn’t had a good set of test cases for how things should work and render.
In fact, in one scenario they had to take test cases from a later specification and dumb them down to conform to the older spec, and even then there was a lot of wriggle room.