[sc34wg3] Thoughts on the RM

Steve Pepper sc34wg3@isotopicmaps.org
Fri, 25 Apr 2003 16:00:44 +0200

At 08:33 25.04.2003 -0400, Sam Hunting wrote:
>On Fri, 25 Apr 2003, Steve Pepper wrote:
> > I am more concerned with the substance, of which I find there to be
> > very little. Even more, I am concerned with the motivating factors for
> > the substance. Until I understand what the RM is supposed to *do*,
> > and why anybody would want to do whatever that is, I see no point in
> > wordsmithing.
>To be clear, the heart of your objection is that there is no "why."
>(editorial and technical objections, if any, are subsidiary.) How would
>you like the "why" expressed in terms of a deliverable? Part of the
>standards text? Informally on this list? In terms of process (and the
>limited time up to and in London, it would be helpful to know this.

What I feel missing is a clear explication of the "what", the "why",
and the "how".

I would like to have these explained informally (but clearly) on this
list first, thereafter in London. Then, once they have been clarified,
I would like them expressed in the document text.

In the first instance, the "what" should focus on what the RM does
that 13250 doesn't already do. It would be helpful if, initially, you
could express this without resorting to new terminology, and
especially without talking about "TMA"s. Assume the SAM (or today's
13250, if you like) as a given, and provide plenty of examples of other
"world views" that could conceivably qualify as "TMA"s, e.g. RDF, the
relational model, NewsML, XFML, or anything else that you think might
make the whole thing clearer. (HTTPGET-1.1 was useful for starters,
but not really convincing, since the SAM already provides the ability
to model the subject/subject indicator and subject indicator/subject
identifier relationships. Show us something that cannot be done with
the SAM.)

The "why" should cover the user requirements or business case and show
how the "what" is of actual use. (Naturally, these should be user
requirements that are not covered by today's 13250, otherwise there
is no justification for generalizing what we already have.)

The "how" should steer clear of technical details and new jargon as
much as possible, but at the same time show very clearly what would
be involved, in practice, for an information owner and for an
implementor, in applying the concepts of the RM to achieve the "what".
This will enable us to see whether the effort involved is in
reasonable proportion to the "why".

Once the "what", the "why", and the "how" have been clarified, it will
be time to discuss the "when" ;-)

I hope this is of use. I don't expect you to manage all of this in
a single posting, or even before London, but the description of what
I and others are looking for could perhaps give some guidance as to
what is likely to be the most effective use of the Saturday afternoon

Best regards,


Steve Pepper, Ontopian