[sc34wg3] Structuring the topic map standards

Patrick Durusau sc34wg3@isotopicmaps.org
Fri, 31 Jan 2003 15:20:18 -0500


Lars,

I think the names given various "parts" of the work on topic maps in
prior documents can be misleading. There was (is) little or no agreement
on the details of what goes with each name (the "devil being in the
details") so I would find it helpful if you could say what you think the
various parts you agree or disagree with will contain and how they
relate to other parts.

For example, you mention the SAM, which I assume you are equating that
with the document that you have been writing which is bound to the XML
Infoset? A topic map API using the XML Infoset is an important thing,
but I don't think its place in the world of topic maps has yet been fixed.

I tried to start such a thread saying what I thought about the
principles of topic maps and a description of the topic map model using
a particular methodology. I am still working on what I think the other
bits would be and their relationships to other parts. I would prefer to
write the next parts in conversation with what others see as the various
parts of the topic map effort. That is difficult to do when people use
names that have no shared understanding. The speaker knows what is meant
quite clearly but if the listener has a different understanding, there
has been no real communication. (This is where I put the prior history
of "agreements" in this group. I am really interested, as you have 
stated, in determining what we want to do and then doing it. Nice to 
have someone besides myself who is impatient for action.)

Hope the week is ending on a high note!

Thanks!

Patrick

Lars Marius Garshol wrote:

 >* Steve Pepper
 >|
 >| In N372, the following structure is suggested:
 >|
 >|    * ISO 13250: Topic Maps Basic concepts.
 >|                 Annexes for XTM, HyTM.
 >|    * ISO xxxxx: Topic Maps Data Model (SAM)
 >|                 Annexes for deserialization of XTM (and HyTM), and
 >|                 mapping to Topic Maps Reference Model.
 >
 >This part I am unhappy with, for the following reasons:
 >
 > a) splits the definition of the concepts off from the data model that
 >    determines their structure and gives us an ISO 13250 which does
 >    not actually provide all the criteria for conformance; in fact,
 >    the real standard will be the second one,
 >
 > b) it duplicates the same content across two different standards, so
 >    that if we want to change topic maps we will have to change both
 >    standards, and anyone wanting to implement this will have to
 >    follow two different and possibly contradictory standards, and
 >
 > c) it lumps the syntaxes in with the SAM, which is not the cleanest
 >    way to organize things, given that there may be more syntaxes to
 >    come,
 >
 > d) it puts the SAM and its RM mapping into the same document, which
 >    means that the SAM can't be published before the RM is finished.
 >
 >Not only am I unhappy with it, but I also see no point in this
 >proposed structure. It actually runs counter to the advice Charles
 >gave you (the second part, the real standard, will be long, awkwardly
 >organized, and an even more horrible read than it needs to be), and
 >the only thing to recommend it is that there will be a definitive
 >introduction to the topic map standards.
 >
 >However, I think that's better achieved through a separate document
 >which actually aims to teach the use of topic maps the way the W3C
 >primer documents do. Doing it that way will give us a better structure
 >for the standards, and that the same time give us a better tutorial
 >because the tutorial will actually be written as a tutorial.
 >
 >|    * ISO xxxxx: Topic Maps Information Aggregation Model (RM)
 >|    * ISO xxxxx: Topic Maps Conformance Canonicalization syntax.
 >|    * ISO 18048: Topic Maps Query Language
 >|    * ISO 19756: Topic Maps Constraint Language
 >
 >This part I think is OK, though as stated before I would prefer a
 >multi-part 13250. I actually think the old roadmap was exactly what we
 >wanted, and that we should go back to it.
 >
 >
 >

-- 
Patrick Durusau
Director of Research and Development
Society of Biblical Literature
pdurusau@emory.edu
Co-Editor, ISO Reference Model for Topic Maps