[sc34wg3] A new idea for the Topic Maps standard

Lars Marius Garshol sc34wg3@isotopicmaps.org
05 Feb 2003 15:39:03 +0100

* Michel Biezunski
| 1) Tell these people their stuff is limited, not appropriate, that
| there is something much better called Topic Maps and that *THE* XML
| syntax is called XTM. It's the only one. Period. If people are not
| doing it, it means they don't know what's good.

I think we all agree that this attitude is counterproductive, and that
we should not assume it.
| 2) Tell them that no problem these syntaxes can be interpreted as
| topic maps and open for wider interchange. Doing that would ensure
| that Topic Maps will be able to fulfill its objective, i.e. to merge
| knowledge, furthermore regardless of which syntax it's expressed in.

I am very much in favour of this, and have already discussed with
Peter van Dijck whether to create an official XFML-to-SAM mapping to
be published on the XFML site, complete with published subjects. He is
in favour, and we will probably do this before very long.
| This has one consequence, that may help solve one problem that Lars
| was pointing at (what to do with HyTM?): Make XTM as well as HyTM
| non-normative *examples* of how topic maps can be interchanged. 

I think we are already doing this and will continue to do it. There is
a topic map model, SAM, and there are several syntaxes for topic maps.
Implementations are not required to implement any particular syntax,
but whatever syntax they do implement they must implement according to
the specification.

Isn't that all we need?

| Speaking of how the standard is structured, it may also have the
| consequence to make SAM a non-normative technical report.

That doesn't make sense, I'm afraid, since the normative XTM
specification would then rely on a non-normative technical report for
the key part of its expression. Ditto for Canonical XTM, HyTM, TMQL,
and TMCL.

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