On Fri, Aug 21, 2009 at 7:42 AM, Michael Kay<> wrote:
> A better approach might be to treat DTDs in the structure of the article as
> co-equal with other schema languages, and mention in the prose that for
> historical reasons they are given a special status in the XML
> recommendation. That's a better reflection of the reality of usage.
No, I don't think that's sufficient. FDTDs are fundamentally
integrated into the design of an XML parser. They are not optional
pieces one can ignore. Nor are they just a schema language either.
They have noticeable effects on a document's infoset even in the
absence of validation. Yes, in hindsight, this was a mistake; but it's
not one we can ignore or sweep under the rug, much as we might wish
things were otherwise. :-(
--
Elliotte Rusty Harold
