- From: Paul Prescod <paul@prescod.net>
- Date: Tue, 01 Oct 2002 18:16:20 -0700
- To: Jonathan Borden <jonathan@openhealth.org>
- CC: www-tag@w3.org
Jonathan Borden wrote: >... > That is a different issue (I think). For me, it is the key issue. It is why I don't recommend XLink to my customers. And most of the complaints I have seen from the XHTML group have resolved around syntax. > It's one thing to say that the > _semantics_ of XLink are unacceptable for XHTML but another that "I just > don't like the way it looks" -- surely not from the very WG that has foisted > XHTML Modularization DTDs on us! :-) ... > The point is that sometimes we have to put up with ugly syntax > in order to get the desired semantics. I disagree strongly. Syntax is malleable. Semantics are malleable. The mapping between them is malleable. If we can't find a way to get decent syntax and semantics then we aren't trying hard enough. > Seriously though, consider that XHTML > Modularization DTDs are butt uggly for the very reason that such contortions > are _required_ in order to get around the problems that DTDs have with XML > Namespaces. And surely this is one of the many reasons that various schema languages were invented: so that we could have decent syntax for namespace and datatype-integrated schemas instead of something butt-ugly (see also DT4DTD). And it turns out that the XML representation of schemas was still somewhat butt-ugly so James Clark invented the RELAX compact syntax. In the last few years we've totally changed the syntax, semantics and mapping between the two in the schema realm. QED. Paul Prescod
Received on Tuesday, 1 October 2002 21:16:52 UTC