I think that's too simplistic, as it would turn <p><b>hello</b></p> into
<b>hello</b> in its own. Interesting idea, though, of a post-processor
to clean up; it would be elegant because it could simply be thrown away
when switching to XHML 2.0. And in an XSLT context it could be
implemented natively.

There's lots to think about here. I hope I'm not the only one who's
going back to his/her XSL and wondering how they can improve it...

(PS for the record, I have a hybrid policy of turning any <p> which
contains a list, eg, quote, table etc etc) into a <div>, and leaving
text-only ones alone. Horrible, eh. I think I'll stop doing it...)
Sebastian Rahtz

Information Manager, Oxford University Computing Services
13 Banbury Road, Oxford OX2 6NN. Phone +44 1865 283431

OSS Watch: JISC Open Source Advisory Service