[sc34wg3] Individual contribution on the U.S. N.B. position o nthe progress ion of Topic Map standards

Jan Algermissen sc34wg3@isotopicmaps.org
Fri, 02 Apr 2004 09:47:59 +0200

Jan Algermissen wrote:

> Again: the RM enables standardized means to define what a scope, occurrence
> etc. is and also arbitrary extensions of these semantics.

Let me stress one thing here:  The RM *enables* standardized means, it
does not *provide* them (syntax etc).

For TMTK (my RM implementation) I made up a syntax for defining semantical
commitments. This is of course experimental (not a *standardized* means)
but it maybe illustrates the whole RM idea a bit better.

My (partial) version of the Standard Application Model (the core set of semantics
such as name, occurrence, etc.) can be found here:


Dublin Core semantics are here:


and (partial) UML2 Deployment semantics are here:


You can also access the directory:


Even if this is far from ready, I think it is a pretty good way to convey
to another party what my semantics are.

Note that this works a lot like RDF namespaces (only that it is richer) and
that the defined properties work directly for RDF. For example, the
ITIL/CMDB[1] semantics define a property WarrantyExpiryDate which can be
used in RDF as:

http://foo.org/somePrinter http://purl.org/tm/ITIL/CMDB/WarrantyExpiryDate "20.10.2005"


[1] Configuration management for IT service support (actually a fascinating Topic Map
    use case)

Jan Algermissen                           http://www.topicmapping.com
Consultant & Programmer	                  http://www.gooseworks.org