[sc34wg3] Draft Reference Model

Lars Marius Garshol sc34wg3@isotopicmaps.org
23 Nov 2002 11:20:30 +0100

* Lars Marius Garshol
| Huh? Does this mean that the
|   borders-with(norway : neighbour, sweden : neighbour)
| relationship is not a relationship according to the RM? Or do I
| misinterpret you?

* Sam Hunting
| I would say -- slightly. The relation you model using the above
| syntax would have all its implicit members expressed explicitly in
| the graph.

Sorry, but I don't follow at all. Could you show me how the above
would be modelled in the graph?
* Lars Marius Garshol
| Like the approach that we took with the SAM in Montr=E9al, and which
| I've pointed to twice already: generate blank placeholder
| nodes. This avoids the undesirable merging, but even so I'm not sure
| I am happy with it.
* Sam Hunting
| What are the subjects of these "blank" placeholder nodes?

You were there, weren't you?

Their subjects are the subjects that type the roles in the
association, whatever they may be.=20
| Not sure what you mean by "generate". Would the Syntax Processing
| Model define them?

Deserialization specification, if we are to stay with the chosen
terminology. And yes, that is the part of the standard that says they
are to be generated.=20
| This proposal seems similar to the approach I showed at Montral and
| documented at http://www.etopicality.com/RM/WF.shtml -- where I had
| "placeholders" (if you mean what I think you mean by that word) for
| everything.

I'm not sure they are comparable. They may be.
| No. HyQ "is a data content notation for use with HyTime location
| addressing facilities" (ISO 10744 first edition). More than that I
| can't say.

Read the original text again. It compares the SAM to the HyQ and XTM
syntaxes. Sounds like a bizarre thing to do without further explanation.

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