Re: Relation or attribute and why

From: Bob Badour <bbadour_at_pei.sympatico.ca>
Date: Wed, 24 May 2006 17:48:00 GMT
Message-ID: <k%0dg.12429$A26.295083@ursa-nb00s0.nbnet.nb.ca>

Gene Wirchenko wrote:

> On Wed, 24 May 2006 12:33:43 GMT, "David Cressey"
> <dcressey_at_verizon.net> wrote:
>
> [snip]
>
>

>>I disagree.
>>
>>Conceptual Data Model has enough detail to describe the problem domain from
>>a data centric point of view.
>>The napkin model doesn't.  The difference between the napkin model and the
>>CDM is analysis.

>
>
> Of course. A conceptual model (and note that I do not put the
> word "data" in there) is that undetailed. That is what the analysis
> is for, and the result is a logical model.
>
> What do you disagree with?

I disagree with the part where analysis results in a logical design. Analysis results in a conceptual model, which is the starting point for a logical design.

Requirements (identified need/business case) |
V
Analysis -> a conceptual model (information)

  _________________/

|
V
Design -> a logical design (data)

  _________________/

|
V
Implementation (physical design) -> a physical layout (media encoding)

Separating design and implementation (logical and physical) as much as possible has numerous benefits including adaptability, error avoidance, automated maintenance etc. Most products and all non-relational products force one to combine them to a large extent.

Separating analysis and design as much as possible has numerous benefits too including early cost estimation, identifying buy vs. build opportunities, scope identification, early identification and reconciliation of diverse needs, identification and verification of unspoken assumptions, automating design etc.

The difference between information and data is one of the most fundamental and important differences in our field. Received on Wed May 24 2006 - 12:48:00 CDT