Re: Some Laws

Laconic2 wrote:

>
> I got into a set-to with Mr. Novoa recently about whether building
> databases
> should or should not be restricted to the "qualified". I strongly favor
> the idea that everyone who proposes to build a database would be well
> advised to
> learn how to do it right. But I even more strongly support the idea that
> anybody can make the attempt.
>

I have been trying to formulate a suitably puffed-up version of the basic statement: "People understand tables just fine." Some attempts I've had include, "The RDM is maximally comprehensible (and useful?) with no abstraction added or removed" or "Human beings intuitively understand tabular organization."

It seems perhaps you have the same intuitive feel that I do, or perhaps the same experience?

Tables are the great leveler. When you start drawing boxes and listing columns, and finding where to break out details and where to put reference information, suddenly everyone just seems to know what is going on and contributes. I've seen people of all educational backgrounds, titles, status and responsibility come together and actually produce useful work if they go straight to table design. Customer or vendor, makes no difference, technical or "domain expert", makes no difference.

My own current projects take this as axiomatic, and seek to eliminate as much labor as possible between table layout and printing pick tickets.

This does not mean that an inexperienced person will not fall into the same old traps such as the OTLT, only that the nature of the trap can be explained and the person can be helped along the way. Even the traps are comprehensible and can be warned against with reasonable arguments. This is why the most important internal doc in the shop is not the lead programmer's treatise on the next unstoppable object-oriented framework, rather it is the guide to recognizing patterns in customer data and translating them into proven table layouts.

The only consistent exception I've seen are the hard-core procedural coders, the lifers. I'm convinced that they see life in terms of long programs that stretch through time, and cannot grasp a quantum change from state to state. They cannot seem to make the leap from the classical world of step-by-step action and motion to the quantum world of discreet and precisely defined columns.

-- 
Kenneth Downs
Use first initial plus last name at last name plus literal "fam.net" to
email me