- From: William F Hammond <hammond@csc.albany.edu>
- Date: Mon, 07 Jun 2004 14:09:03 -0400
- To: W3C MathML Discussion <www-math@w3.org>

David Carlisle <davidc@nag.co.uk> writes: > It would be possible to derive an SGML rather than XML based DTD for > MathML (and so use SGML syntax such as <log> rather than <log/> ) > and then use that it the SGML based HTML DTD but > > a) I don't think any MathML system (including Mozilla's renderer) would > understand the resulting markup unless it was first converted to XML and > > b) I don't think anyone has yet made such a DTD. (I thought at one time > that such a DTD might be useful for DocBook or TEI integration but both > those languages moved towards XML so the necessity of an SGML compatible > DTD version seemed to go away) Yes. Since the vocabulary of MathML is not author friendly, I doubt if there is much justification for expending the effort to make provision for MathML in non-XML document types. On the other hand, the desire for author-friendly SGML document types providing mathematical markup and admitting translation to XML document types that include MathML is worthwhile. > . . . It may be possible to build an SGML declaration that allows > the use of SGML-notation empty elements (<br> <hr>) in some parts of > the document and XML notation empty elements (<log/> <sin/> > elsewhere) but I haven't seen anyone attempt this. One can build an SGML declaration for an SGML document type with which a defined-empty element may be marked up in any of three ways: (1) <foo>, (2) <foo></foo>, and (3) <foo/>.* This is, in fact, the case with the SGML version of "article" in gellmu's didactic production system. ( It would be sensible for HTML 4.02. :-) Didn't the OP also ask if it would be reasonable for XHTML+MathML to be mime-typed as "text/html"? Some of us once thought so and may still think so, but we lost that battle. Cheers. -- Bill * corresponding, respectively, to "\foo ", "\foo{}", and "\foo;"

Received on Monday, 7 June 2004 14:09:08 UTC