Re: [tm-pubsubj] Re: Country.xtm PSI; also proposed region.xtm PSI

[Murray Altheim <murray.altheim@sun.com>]

Lars Marius Garshol wrote:
[...]
> * Murray Altheim
> |
> | This is the first time I'd heard of this, and I'm curious as to how
> | this conclusion was arrived at.
> 
> The PMTM4 model uses PSIs to define association types for core topic
> map concepts like the topic-basename relationships so those PSIs will
> have to be in the RM-to-SAM mapping somewhere for it to be
> self-contained.
> 
> Given that it seems only reasonable that the PSIs for the constructs
> common to the reference model and the SAM model (such as "topic") go
> into the same bag.

Yes, looking at it in this light I understand.
 
[...]
> | I am speaking specifically of enumerating the components in XTM, not
> | in topic maps generally, since the application space I'm talking
> | about only XTM, speaks only XTM.
> 
> Well, XTM is ISO now. The revised topic map standard will define both
> HyTM and XTM, as well as the relationship between them. So XTM will no
> longer be a TopicMaps.Org thing at all, which makes sense, given that
> XTM is very much core.
> 
> | I would certainly want members of the ISO committee to review our
> | work (as I'm certain they would too), but it's always possible to
> | map the topic map reference model (which might not even be in XTM
> | syntax) to the PSIs using our very own mapping technology.
> 
> I am not sure exactly what you mean here, but I think the need for the
> RM and SAM models to be self-contained as specifications is what will
> require those PSIs to be specified as part of those specifications,
> rather than outside them.

I guess it just hasn't been clear that the topic map reference model
would be defined in terms of XTM (ie., in terms of XML, the relevant
application space for at least my work, as well as many others'). I'd
thought the ISO committee was really concentrating on models, and that
they would likely take the form of meta-DTDs, meta-models, etc. since
XTM is only one of a number of possible syntaxes for topic maps.
 
> | Maybe this seems more complicated than it is: I'm only talking about
> | having a PSI for each language component in XTM, one for each
> | element, one for each attribute. This would allow a named mapping
> | from XTM to, say, RDF while preserving the semantics via PSIs.
> 
> In that case we may not be talking about the same things. The elements
> and attributes in the XTM DTD are not what the core.xtm PSIs represent,
> so they are not what I am talking about.
> 
> This xtmdtd.xtm PSI set is something I haven't thought about before,
> and I am not sure what it would be used for. I have no idea how you'd
> do an XTM-RDF mapping this way.

Well, perhaps I'm overstating, though I'm not sure. I'm *in a sense* 
unconcerned with what the core.xtm PSIs represent. I merely need a 
way of re-representing XTM syntax (whatever it happens to mean) in 
another syntax by use of PSIs for the language components. Had the 
original core.xtm included PSIs for say, "base name", I'd be able to
characterize a resource in RDF as a base name. Absent all of these
PSIs, it's always a "custom" map with no resolveable hooks back into
XTM (other than what is supplied by human-readable documentation). I'd
want this (as in my example) to establish ontological relations between 
an XTM topic map and a map expressed in RDF.
 
> * Lars Marius Garshol
> |
> | Again, this is ISO territory, to be thrashed out by SC34/WG3.
> 
> * Murray Altheim
> |
> | It is insofar as the models and core work. How this is mapped out to
> | a set of PSIs, and how PSI modules (such as I described) work
> | together certainly seems within our charter.
> 
> Do you mean that the documentation and publishing of the PSIs for
> concepts in the models should follow the PubSubj TC guidelines, even
> if the documentation is published by ISO? If so, I agree fully.

Yes. That is, if the concepts in the models even use XTM syntax for
their expression. I don't see that this is clear.

> In that sense there certainly is a relationship, but the actual
> working out of the PSIs and their definitions will have to be done
> together with the model work since they are so closely connected, so I
> think it's better for the PubSubj TC to just leave this to WG3 and
> provide WG3 with a good set of guidelines to follow.

Yes, if my last supposition is true. And then I fall in line behind 
Steve Pepper in wanting something yesterday. I wonder how long it'll
be before we see these PSIs?

Murray

...........................................................................
Murray Altheim, Staff Engineer          <mailto:murray.altheim&#64;sun.com>
Java and XML Software
Sun Microsystems, 1601 Willow Rd., MS UMPK17-102, Menlo Park, CA 94025

       Ernst Martin comments in 1949, "A certain degree of noise in 
       writing is required for confidence. Without such noise, the 
       writer would not know whether the type was actually printing 
       or not, so he would lose control."