Cart
[Home] [By Thread] [By Date] [Recent Entries]
Yes. Still, one has to account for multiple interpreters one of which is the author. It is useful to know who is authoritative or dominant in a communication. While the system need not enforce a relationship in the instance, one should provide that information when requested. How would you envision a layer or layers that then includes the principal and auxiliary types first? Then the architectural types? Next, how would DTDs have to change to support that view? What should the instance signal to associate itself to these layers? len -----Original Message----- From: Gavin Thomas Nicol [mailto:gtn@r...] On Friday 01 February 2002 07:18 pm, Joe English wrote: > Having thought more about the second question, I now believe that > the only thing that's intrinsic to an XML document is > the stuff that's manifest in the parse tree: GIs, attributes, > and content. Everything else -- the principal type, auxilliary > types, architectural forms, etc., -- are imposed by the observer. Precisely what I have been saying.... (though I would say "interpreter rather than "observer"). In compilers, you usually (not always, usually) go from source->AST->object tree with something doing the conversion from AST->object. XML is the same. The XML core DOM is roughly an AST... and we interpret that as we wish. Even humans go through the phase of parsing into "elements" then determining their "type" . The magic is is in the interpreter, not the data. ----------------------------------------------------------------- The xml-dev list is sponsored by XML.org <http://www.xml.org>, an initiative of OASIS <http://www.oasis-open.org> The list archives are at http://lists.xml.org/archives/xml-dev/ To subscribe or unsubscribe from this list use the subscription manager: <http://lists.xml.org/ob/adm.pl>
|
