[tmcl-wg] Any TMCL dependencies: OWL Maybe?

[Lars Marius Garshol]

* Mary Nishikawa
| 
| Will the TMCL depend on XML Schema Part 2, or OWL or something else?

XML Schema, part 2 specifies the system for primitive types in XML
Schema. To me it seems that we can either do like RELAX-NG and create
an open type system where applications can choose what set of types
they want to use, or we can create a closed one like the XML Schema
one.

The benefit of a closed type system is that validators/engines/query
engines etc can implement the whole system and be sure that they will
never encounter an unknown type. The downside is of course that you
will be limited to whatever is in that type system.


Now, as for OWL, OWL really has two different parts. One is the simple
structural constraints (such as, resources of class X may have Z
statements of property Y), and the other is the more ontological
"constraints" (such as, class X is equivalent to class Y, class X is
disjoint with class Z, and so on).

It is clear that it is not really practical to use the combination of
RDF Schema and OWL as a topic map constraint language. It works, if
combined with an RDF-TM mapping, but does not provide all the features
needed for a topic map constraint language.

This suggests to me that TMCL must be a constraint language completely
independent of RDFS and OWL. We may, however, be able to reuse the
more ontological part of OWL unchanged, should we want to do so. The
question, I guess, is whether we want to.

-- 
Lars Marius Garshol, Ontopian         <URL: http://www.ontopia.net >
GSM: +47 98 21 55 50                  <URL: http://www.garshol.priv.no >