Re: Multi Valued Interface Models?

dawn wrote:
> Let's start here. What is your definition of a data model as in the
> term "relational data model"? You can bounce off
> http://www.tincat-group.com/mewsings/2006/01/naked-model.html or choose
> your own. Once I understand your definition of this term, I will be in
> a better position to see if I can rephrase. (I'll respond to other
> aspects of your posting later, gotta run). --dawn

Well this is the heart of the issue for me Dawn - I do not believe in a "naked data model".

Just because there is a Logical Data Model and a Conceptual Data Model does not mean that by dropping the prefixes we can be left with some pure data model. I understand your argument here, but it appeals to the logic of human naming semantics, which can be a very misleading path.

A conceptual data model, a logical data model and a physical data model have practically nothing in common and I don't believe there is much correspondence to extract from them that is the "data model" component. Conceptual and logical deal in information, but physical does not, it only works on data. Conceptual utilises human interpretation upon general abstractions, logical requires formal operators on well defined domains. There is no pure semantic meaning to "data model", and the term's use is completely contextual.

On top of this (as I mentioned in a followup to the blog), the term "data model" itself is rather non-sensical. Data is nothing more than a one or a zero, a groove in a record, a bead moved down an abacus. How on earth can such an abstract nugget of nothingness ever be "modelled"?

In my eyes, the term should be "information model", and if it was we'd start to lose the confusion between layers. Consider a team developing a game's engine with a physics model. They distinguish the model and its encoding, not some "logical" physics model and the "physical" physics model. That sort of terminology would seem daft to them - we are just unfortunately used to it here.

Now if you are defining a new meaning for "data model" that corresponds to an interface that is fine - but it is certainly not a logical data model (as you acknowledge) as we use the term normally, and so comparing it to the RM is like comparing apples and pears in my book.

All best, Jim. Received on Mon Feb 13 2006 - 16:25:22 CET