[sc34wg3] Quietly Storming Out

[Lars Marius Garshol]

* Martin Bryan
| 
| Who's talking about "interchange between arbitrary markup systems"?
| I'm talking about interchange between systems that have compatible
| markup systems in agreed scenarios. This is just as valid as the
| "one-size-fits-all" approach of existing topic map solutions. We
| should not throw this possibility out to support the myth that any
| topic map system shoud be able to arbitrarily receive and process in
| a meaningful way any topic map.

Hear, hear! It *is* a myth, but it's been challenged too rarely. I
think Martin is right, and that is another reason why the problems
with embedded XML do not worry me too much: They are too like the
problems of receiving a topic map with an unknown ontology.

I once received a topic map in Japanese (even the IDs were in
Japanese) from Hayashi-san, a colleague of Naito-san's. I could work
out what types where in the ontology, that one of the association
types was hierarchical, and that was about it. I was then in the same
position as a piece of software receiving a topic map whose ontology
it is unfamiliar with...

(It turned out to be a medical thesaurus, but not even that was clear
to me from looking at the topic map.)

| Are you saying that we should allow XML, but not namespaced XML? I
| could live with this, but not with the PCDATA only argument. I'd
| love proper fragment control, which this won't happen until Paul's
| group or my DSDL group come up with something usable unfortunately.

I think I agree with this: the fragment control is really something we
need XML mechanisms for, plus some mechanism for connecting the XML
mechanisms into the topic map world.

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