[
Lists Home |
Date Index |
Thread Index
]
-----Original Message-----
> From: Pete Kirkham [mailto:pete.kirkham@baesystems.com]
> Sent: Thursday, January 24, 2002 5:47 AM
> To: gtn@rbii.com
> Cc: xml-dev@lists.xml.org
> Subject: Re: [xml-dev] There is a meaning, but it's not in the data
> alone
>
>
> gtn@rbii.com
> >The "domain model" as you put it, is roughly equivalent to a
> >vocabulary with an associated set of semantics. So long as you agree
> >on the terms, you can communicate, and that is the whole point.
>
> The domain model has no fixed language, only semantics.
...
> The AIM may be an ontological framework or an EXPRESS or a
> MOF L1 model (such as a
> UML domain model class diagram; UML fudges this a bit by
> using the same language
> for its domain models [which are abstract subsets of the
> information in a domain,
> not models of a whole domain] and for application constructs).
I'm not sure if you're saying that the conceptual model (domain model) has
to be expressed in a different language than the physical model (AIM).
I'm not arguing that the two models are the same, it's just saying that it
is possible to express both in the same syntax, assuming that syntax is
expressive enough.
And, yes, meaning is not independent from language, but the two influence
each other.
> A vocabulary is part of the mapping of an entity to a
> concrete syntax (this is also
> true for natural language, though the domain models for
> different people are often
> inconsistant).
Which argues for a domain authority, which is why we have a proliferation of
vocabulary standardization organizations in OASIS and elsewhere.
What are they using to indicate that entity x in in domain y? Namespaces?
Naughty naughty. Doctypes? I don't think so. There is no universally
accepted mechanism. Can that mechanism be contrived in XML? I think so. I
think it has to.
> If you resolve a namespace into something, then that must
> contain both a syntax and
> a mapping of that syntax to a model to be useful, or you are
> stuck in text
> processing rather than information.
And if both models are in XML format, I think all the better.
|
