[sc34wg3] a new name for the Reference Model

Steven R. Newcomb sc34wg3@isotopicmaps.org
03 Jan 2003 16:37:43 -0600


"Mason, James David (MXM)" <masonjd@y12.doe.gov> writes:

> Steve says, "we will defeat ourselves if we direct
> public attention toward 'ISO 13250'"; does he think
> the public will be able to read the RM, whether it's
> part of ISO 13250 or some separate standard?

No, I don't.  That's why I think we should leave the
scope of 13250 exactly where it is: 2 interchange
syntaxes for Topic Maps, and not a multi-part standard.
That way, when we tell people to look at it, they get a
nice, friendly, self-contained, BRIEF introduction, and
no more.  They can also skip straight to the DTDs, as
you, I, and many others would normally do.

Later, when/if they have deeper questions, they can
discover the SAM and the RM.  13250 can make discreet
normative references to them, in order to let its
readers find answers to deeper questions.  If they want
the answers to deeper questions, we've already
persuaded them, and the existence of such depth will
only increase their comfort level and further encourage
adoption of Topic Maps by them.

> This is why I am with the RM sort of like I am with
> the base standard: I skip the text and go straight to
> the diagrams.

Well, in the case of the RM, the diagrams illustrate
only Clause 3 of the RM, and only part of that clause,
really.  Clauses 4, 5, and 6 are also very important
clauses.  In fact, Clauses 3 and 4 can be seen as
merely prefatory to Clauses 5 and 6, which are the main
events.

> After reading Steve's essay, I'm still asking the
> same questions I asked before. Are these things
> standards or technical reports? How do they fit
> together?

Just in case my answers were concealed in my usual heap
of verbiage, let me reiterate as briefly as possible:

> Standards or TRs?

Standards.

> How do they fit together?

The *deep* meaning of both of 13250's syntaxes is
partly defined by the SAM, and partly by the RM.

*******************************************************

> Some of these issues may fall out in the course of
> revision and editing, but I'm still troubled. When
> Steve says the RM is for someone who wants to
> "achieve subject location uniqueness for subjects
> that are specified by domain-specific relationship
> types", he's using language that's way out in
> philosophical territory, not something that sounds
> like the specification of "Document Description and
> Processing Languages" (with emphasis on "languages").

True, the RM does not describe a language.  Nor does it
describe a language for describing languages, as some
SC34 projects have done.  Instead, the RM describes the
requirements that must be met by the definitions of TM
Models, regardless of the language(s) in which such
Models are defined, in order to facilitate the
achievement of the subject location uniqueness
objective.

So, maybe the RM -- and therefore any general
description of the Topic Maps paradigm sufficient to
constrain the design of languages for Topic Maps -- is
outside the scope of SC34, because SC34's mission is
limited to standards for *specific* languages.  Is that
what you're saying?

With regard to "way out in philosophical territory": In
Clause 5, the RM contains a checklist of ten components
that a TM Model definition must have in order to
conform.  I don't think this list can be described as
being "way out in philosophical territory".  On the
contrary, it's hard-headed and practical, and
conformance to it is 100% verifiable.  See the
checklist at

http://www.y12.doe.gov/sgml/sc34/document/0344.htm#parid0282

Topic maps that conform to TM Models whose definitions
conform to these checklist requirements can be
aggregated in such a way that the achievement of the
subject location uniqueness objective can be
significantly facilitated.  When topic maps don't
conform to their TM Models (called "TM Applications" in
the current draft), or when their TM Models don't
conform to the RM, this facilitation is not possible.
Therefore, the Conformance Clause imposes specific
constraints on

  * TM Models ("TM Applications" in the current draft),

  * definitions of TM Models,

  * implementations of TM Models, and

  * interchangeable topic maps

See the Conformance Clause at

http://www.y12.doe.gov/sgml/sc34/document/0344.htm#parid0443

> How can you write a conformance specification to
> something like that (and if you can't specify
> conformance, you're not writing a standard)?

I don't understand why you say that the RM can't
specify conformance.  It not only can, it does.

> Graham's message, which he placed outside this
> thread, is appropriate: (to oversimplify greatly) we
> need a formula for getting from the RM to the SAM and
> back.

I think what we *really* need is to express the SAM
*accurately* in RM terms.  We're working on that.
(It's a challenge, but it can be done.)

> Steve is aware of the issue. To my question, "Does
> [the RM] specify or interpret?" he responds:
> 
> It specifies. If it only interprets, then its
> constraints are optional, and we abandon the idea
> that "Topic Maps" means reliable, predictable,
> ontology-neutral knowledge aggregation.
> 
> I agree that a standard must specify. But at this
> point I find the RM is, like this essay, heavily
> interpretive.

Maybe I don't understand what you mean by
"interpretive".  As I read it, the RM does little else
but to specify, specify, and specify some more.

-- Steve

Steven R. Newcomb, Consultant
srn@coolheads.com

Coolheads Consulting
http://www.coolheads.com

voice: +1 972 359 8160
fax:   +1 972 359 0270

1527 Northaven Drive
Allen, Texas 75002-1648 USA