Re: A Normalization Question

Neo wrote:
> The lowest (most general) level will be applicable to all higher (more
> specific) levels.

OK, I can see how you're thinking.

But as an analogy you wouldn't write all your programs in machine code would you even though it's possible. You'd use assembler which in turn might be used by C and in turn that might be used to create some higher-level language.

You wouldn't build a computer by mining all the silicon etc. yourself, you'd get the large components ready-made. Even the people who make the printed circuit board would buy the capacitors etc. ready-made from some one else (or I assume they would).

It's useful to break down tasks into different levels, where each level hides its implementation details from the level above. Do you agree?

That's what the relational model does: it separates out the logical and the physical. So that the data analysts can concentrate totally on the business model and not be worried about the implementation details. And the DBMS writers can concentrate on the physical implementation without worrying about what specific data will go in the database.

As I said before:
It sounds like you're trying to roll two concepts of "redundancy" into one when it's more powerful and useful to have them separate.

Are you suggesting that it's better to *not* distinguish between the two forms of redundancy?

Check out this famous painting by Magritte: http://en.wikipedia.org/wiki/Image:MagrittePipe.jpg and this discussion of it:
http://en.wikipedia.org/wiki/Ren%E9_Magritte

I think it's kind of relevant to the distinction between the word "brown" itself and the colour that is referred to as "brown".

Paul. Received on Thu Jul 08 2004 - 21:37:01 CEST