- From: William F Hammond <hammond@csc.albany.edu>
- Date: 04 Aug 2003 10:22:50 -0400
- To: www-html@w3.org
Masayasu Ishikawa <mimasa@w3.org> writes:
> I'm not aware of any concrete plan to completely abandon DTDs in XHTML 2.0.
> It has always been the Working Group's position to provide multiple
> schema implementations, including DTD. Whether it should be normative
> or not is highly debatable.
>
> That said, more and more issues make it rather impractical to use DTD -
> accommodating RDF/XML is a notable example. The RDF in XHTML Task
> Force [1] will remain stuck, unless we give up DTD or give up RDF/XML.
. . .
> [1] http://lists.w3.org/Archives/Public/public-rdf-in-xhtml-tf/
Why stuck? [1] is an email archive. Could you summarize in a few
words or is it really complicated?
Granted that DTD models do not incorporate fine-grained specifications
of the type that are important in the EDI side of XML, nonetheless I find
that DTD validation is good for locating structural markup problems, and,
therefore, is helpful to human authors.
With the document side of XML where a DTD model is insufficient, isn't
the tradition to require DTD validation _and_ a noiseless run through
some other form of validating processor?
I think that makes sense for XHTML. That is, "valid" might, for example,
mean both DTD valid and relax-ng valid.
-- Bill
Received on Monday, 4 August 2003 10:29:45 UTC