Cart
[Home] [By Thread] [By Date] [Recent Entries]
K. Ari Krupnikov scripsit: > Must be my lack of SGML experience. Can you elaborate on this one? It's just nice to be able to write: <!ELEMENT h1|h2|h3|h4|h5|h6 (%flow.model;)> instead of six different element declarations. > > General issue: Should there be some way to indicate candidate roots? > > In existing DTDs, any element can be a root. > > On a tangent, I'd like to point out that the name of the root element > in the doctypedecl (production 28 in XML, "<!DOCTYPE foo...") is only > required for SGML compatibility and is superfluous in XML where tags > may not be omitted and the first encountered element is unambiguously > the root. If the proposal will break SGML compatibility anyway (the > snipped text suggests it will), it might as well drop this unnecessary > requirement and its associated validity constraint. Indeed. But note that I am *not* proposing a revision of XML DTDs, but rather the creation of a new schema language to be a superset of XML DTDs. Schemas of this kind cannot be referenced by DOCTYPE declarations, because the result would be not well-formed. Instead, they must be specified to the external validation program. This is purely my own idea and may not be the way that the committee is going. Indeed, they are in a use-case/requirement collecting phase and not yet doing syntax. -- John Cowan <jcowan@r...> http://www.reutershealth.com I amar prestar aen, han mathon ne nen, http://www.ccil.org/~cowan han mathon ne chae, a han noston ne 'wilith. --Galadriel, _LOTR:FOTR_
|
