[sc34wg3] The interpretation of facets

Steve Pepper sc34wg3@isotopicmaps.org
Sat, 26 Apr 2003 22:43:59 +0200

At 08:59 26.04.2003 -0400, Michel Biezunski wrote:
>Why instead of mapping to SAM don't you consider this
>approach? Map the HyTm DTD to the TMM. Interchange could then
>take place at that level instead and you might not have to change
>the DTD at all. You would still be able to use the HyTime DTD as
>such with facets and all the rest and be able to interchange with
>a SAM-based application?

In one sense this is an interesting suggestion. The results
might indeed help throw more light on the RM. On the other hand,
I think this suggestion, coming from one of the editors of 13250,
is less than responsible.

The committee has agreed that both XTM and HyTM should be
expressed in terms of the SAM. That is what Martin has tried to
do - heroically - in N391, and for the most part he has succeeded.
There remain a few warts, including the issue of facets. In my
opinion, the editors of 13250 have an obligation to work together
to see if these remaining warts can be removed.

You, Michel, have a particular responsibility in this regard, I
would say, because you are not only an editor of 13250. In many
respects, you are "Mr. Facets" himself, don't you agree?

* Facets are only in 13250 because you fought tooth and nail to
   keep them in there, in the face of very strong opposition,
   especially from the proponents of the Whataburger model who
   thought scope did the same thing, only better (which it doesn't).

* You were also present (in Dallas) when the unanimous decision
   was taken by TopicMaps.Org to not include facets in XTM.

* Finally, you wrote the only official documentation (N277) of the
   position of the editors regarding the lack of explicit support
   for facets in XTM.*

N277 says the following:

   Facets are not mentioned in the XTM DTD.

   In the HyTime-based meta-DTD, facets are qualifiers used to
   assign a property (and a value for the property) to an
   information object. Facets are not needed in the XTM DTD
   because the XTM DTD allows a topic map author to explicitly
   regard an information object as a subject in itself (a subject
   constituting resource). In the XTM DTD, therefore, it is
   possible to associate properties and values with information
   objects by means of <association> elements.

This accords well with the position I have outlined in my two
postings recently, but it is far from sufficient. The Japanese
National Body asked for further clarification a year ago, in
N310, and specifically requested

   examples on how a topic map author might use XTM syntax to
   "explicitly regard an information object as a subject in
   itself (a subject constituting resource)" and "associate
   properties and values with information objects by means of
   <association> elements", and how this is similar to but
   somewhat different from how the facets in HyTime-based
   meta-DTD act as qualifiers to "assign a property (and value
   for the property) to an information object" should be

It is time this request was answered. If it is the case that
the information content inherent in a <facet> element can be
equivalently expressed in XTM, the method for doing so needs
to be spelled out. As you will see from my mammoth posting
earlier today, there are at least two proposals for ways in
which this might be done (mine and N391). We need to decide on
a single unified approach, otherwise we haven't done our job.

I kindly request, therefore, that as an editor of 13250 with
almost a special responsibility for facets, you help clear
things up, instead of simply pointing Martin in the direction
of the RM. As an editor of 13250, you (like SRN and Martin)
have a responsibility to address issues like this with the
utmost seriousness.


* At least I assume you wrote it, since the same explanation
   (with slightly different wording) appears in your chapter of
   Jack Park's book.

Steve Pepper, Ontopian