OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.


Help: OASIS Mailing Lists Help | MarkMail Help



   Re: [xml-dev] Symbol Grounding and Running Code: Is XML Really Extensib

[ Lists Home | Date Index | Thread Index ]

Dare Obasanjo wrote:

> Dare Obasanjo scripsit:
> > Agreed. XML is about syntax and nothing else. People who think
> > otherwise have been misled by the XML hype of yesteryear.

If you limit yourself to XML 1.0 as defined, this is true. I think people
are generally talking about methods to define semantics for particular XML
vocabularies, which is unquestionably possible. The idea that there is some
sort of Universal Semantics for XML is akin to the idea that the Star Trek
universal translator exists.

> Every word that Tim Bray says about markup at
> http://tbray.org/ongoing/When/200x/2003/08/11/SymbolGrounding
> applies with exactly equal force to ordinary human-readable
> natural language text.

Yes, and we have ways of teaching humans what natural language text
means. Do we have ways of teaching machines what natural language text
means let alone markup tags.

There have been machine programs for natural language interpretation in use
since the 1970s. The U.S. government has devoted an unknown, but known to be
large, amount of $$$ to this problem. In any case "Babelfish" etc. are

Like I said in
http://www.kuro5hin.org/story/2003/8/8/122029/2205. To claim that one
can do this means one has a technique where


2.) I can take a vanilla W3C XML Schema processor and pass it a schema
with embedded Schematron assertions which it automatically learns how to
use to validate an input document in addition to using the W3C XML
Schema rules.


Perhaps that isn't possible (I've not given this any thought), but there
certainly are other ways in which machines can "understand" in some sense,
what text "means" in some sense. In the case of RDF, assuming the various
namespace qualified vocabularies are defined in RDF Schema or OWL, then
there certainly is a formal semantics which defines what the document means.
Indeed you might not find that meaning useful, for your own tasks, but
nonetheless, there is a meaning which may be useful to software which knows
what to do with it (e.g. RDF and OWL agents).

On the other hand if people are suggesting that RDF or OWL is intended to
define some type of universal meaning for all XML documents, that is plain
silly. The only meaning defined by RDF or OWL is specifically for RDF/XML
documents i.e. a specific dialect of XML (a syntactic subset) for which
there has been defined a formal semantics. These documents are specifically
intended to be "understood" by software programs which have been programmed
in accordance with (e.g. to "understand") this formal semantics. Such
programs can "exchange meaning" by transmitting documents encoded in the
RDF/XML syntax. Namespaces are essential to this particular technology -- 
not because there is no other way to do what XML Namespaces do, rather
because these specifications are written requiring that all XML elements and
attributes are XML Namespace qualified. Don't like it -- use another syntax,
write another formal semantics (if you care) for that syntax, and teach your
software how to understand your syntax. The fact that your software and
syntax isn't RDF and doesn't use namespaces doesn't at all affect the
software that does use RDF and is namespace aware.



News | XML in Industry | Calendar | XML Registry
Marketplace | Resources | MyXML.org | Sponsors | Privacy Statement

Copyright 2001 XML.org. This site is hosted by OASIS