Re: [topicmaps-comment] referring to a topic from outside a TM

[Lars Marius Garshol <larsga@garshol.priv.no>]

* H. Holger Rath
| 
| We know that TOMAS is *really* necessary and that it is not a 
| trivial thing.  But we will address it.

* Thomas Bandholtz
| 
| I would not think about a special URN  scheme first. URNs are only
| names (ok, globally uniq names, hopefully). 

I agree completely.

| We need an  addressing method that can provide on-line access to the
| addressed node.

What do you mean?

| Today most of us use URL with *fragment identifiers* (which is not
| widely implemented for XML) and not *queries* (using ? in an URI),
| so they are tied to documents.

They reference documents, but that does not mean they are tied to
them. There is a subtle difference here.
 
| I can only imagine a future for TM when they can be maintained in
| databases without needing a physical file representation. 

We have been there for years already. You could get several different
implementations of this tomorrow, should you want to.

| The PSI activities today are dominated by file-oriented thinking. I
| think this is very risky.

I disagree. (Not that this is very risky, but that PSI activities
today are file-oriented.)

Let me explain: Bernard publishes a piece of text on the web that
explains his idea of what the concept of "nothing" is. I, using a
relational database implementation of topic maps, then want to create
a topic representing "nothing". So, I create a row in the TOPIC table,
and put the URI of Bernard's definition in a column in that table.
Done!

If Steve were to do the same, in a different topic map (in the same
database, or outside it) we would then be able to merge our two topic
maps, because they use the same subject identifier. That this subject
identifier happens to point to a file is of no significance: it's just
a string from the point of view of a topic map engine.

The PSI activities today are not really file-oriented; they just look
that way if you're not really clear on how these things work. The PSI
activities are about establishing URIs that identify concepts through
subject indicators. Whether the URI of a subject indicator points to a
part of a file, to a file, or to a web service is up to the publisher
of the subject indicator to decide.

For everbody else the issue is really irrelevant. They just use the
URI as an identifier for the subject, and know that should some human
being somewhere be curious about what the subject is they can resolve
the URI and read the definition.

-- 
Lars Marius Garshol, Ontopian         <URL: http://www.ontopia.net >
ISO SC34/WG3, OASIS GeoLang TC        <URL: http://www.garshol.priv.no >