[sc34wg3] a new name for the Reference Model

[Steven R. Newcomb]

Lars Marius Garshol <larsga@garshol.priv.no> writes:

[Steve Newcomb:]
> | Personally, I'd really like to change the name of
> | the SAM to "the Topic Maps Standard Model", or just
> | "the Standard Model".  This name seems stronger,
> | shorter, and more appropriate than the
> | confusingly-qualified name, "Standard Application
> | Model".  ("Application" is the wrong word to be
> | using, unless, when we say "Standard Application
> | Model", we're talking about modelling a piece of
> | software that's called the Standard Application.
> | Which we're not.)

> Or we could just say "Topic Map Data Model". That's
> what it was meant to be, and it certainly will make
> it clear that there's only one model to deal with.

I have problems with the name "Topic Maps Data Model".

The term "data model" means many
things to many people.  People frequently say, "the
relational data model", and they also frequently use
the same term to mean much more application-specific
things, like, "the SAM data model" (which has nothing
to do with our SAM, by the way -- see
http://d0db.fnal.gov/sam/doc/design/sam_entities/story.html).

The relationship between our RM and our SAM is
analogous to the relationship between 

* the relational model 

  and 

* a relational model for accounts receivable.


It's as reasonable to say: 

  "the [RM] data model" 

as it is to say 

  "the relational model".  


It's as reasonable to say 

  "the SAM data model" 

as it is to say 

  "our accounts receivable data model".  

Therefore, the name "The Topic Maps Data Model" is too
vague.  The RM could also be called a "data model", and
many people would vigorously agree with the use of that
term to characterize the RM.

I don't think any of us wants to create any kind of
competition between the RM and the SAM.  We've expelled
all SAM-ishness from the RM, so there can be a rigorous
modular distinction between the two.  I hope we will
avoid naming the RM in a way that could cause people to
think that it occupies any of the space of the SAM,
and, similarly, I hope we will avoid naming the SAM in
a way that could cause people to think that it occupies
any of the space of the RM.

The RM's name is wrong precisely because its name
makes it appear to occupy the space of the SAM.  

I think the SAM's name is wrong for another reason
(i.e., that "Application" creates horrendous
confusion), but the answer to that problem is *not* to
name the SAM in such a way as to make it appear to
occupy any of the space of the RM.  I think "Topic Maps
Data Model" would create exactly that appearance, and I
hope we won't make that mistake.

> | So what should be the new name of the RM?  I'm
> | hereby proposing "TM Modeling Principles".  We'll
> | "test drive" this name in the next iteration of the
> | RM, to see if we like it.

> I agree with the criticism others have put
> forward. This sounds like guidelines for modelling
> topic maps more than what it really is. I think
> "metamodel" is much more accurate.

At the moment, the only objection that I can think of
to Martin's "Topic Maps Metamodel" suggestion is that
"Meta-" has the unfortunate marketing connotation of
"not mainstream".  But maybe that's a good thing, if
we're trying to focus mainstream attention on the SAM.
I dunno.  Let's chew on it.

-- 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