Re: MathML expression, conformance, DTD

FYI, the approach used in the SVG spec looks like
it could work for MathML.

http://www.w3.org/TR/SVGMobile12/conform.html#ConformingSVGDocuments

On Tue, 2007-10-09 at 21:46 -0500, Dan Connolly wrote:
> On Tue, 2007-10-09 at 22:57 +0100, David Carlisle wrote:
> > > I'm struggling to understand the role of DTDs in the new
> > > MathML spec.
> > 
> > sorry to make you struggle.
> > 
> > Appendix A is currently unchanged from MathMl2 apart from the addition
> > of a couple of editorial notes which I hoped would indicate that the
> > current state is unstable  In particular the second note floats the idea
> > of making a relax schema normative (and by implication making the dtd
> > not normative).
> > 
> > The WG is still discussing the detailed specification of some of the new
> > elements so we haven't generated a full dtd/schema yet. practically
> > speaking I think the best course of action to get consistent DTD and
> > schema is to author the schema in relax ng and derive dtd and xsd from
> > that. there is then a question as to which of any of these should be
> > normative, and to word the conformance section appropriately.
> 
> OK; I'll stay tuned. Thanks.

-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/