Re: Relational vs. PICK/Object DBMS

From: Dawn M. Wolthuis <>
Date: Wed, 21 Apr 2004 11:43:19 -0500
Message-ID: <c668bn$ji2$>

"x" <> wrote in message news:40861c4a$ <snip>
> I would like to know which model/DBMS perform better (in terms of
> programmers productivity, system requirements, etc.) on the following
> problems (stated in terms of relational model) :
> 1)Been given the domains {A,A1,A2,...,An}, with large n and the following
> database schema:
> - for all i in 1..n, Ri(A,Ai) is a base relation
> - for all i in 1..n, A is a candidate key for the relation Ri

By using terms such as "base relation" you are locking into some relational thinking already.

I am revising this in-process message (stalled when responding last time) in light of other comments in the thread already.

If you could come up with an example business problem to address, that would be helpful -- then we would not be starting with a solution. If you are starting with a solution of what to place in what base relation, then your question is really directed to the relational world.

[If you include an attribute named Man# in your example, I'll be soooooooooooo disappointed bz that was offensive in the 70's (and I spent many "man hours" ridding my company of such a field name) and is REALLY offensive in 2004! You should see the woman pages we put out there as documentation! smiles. --dawn]

> answer the questions:

Again, if you frame these questions as a "user" of the database (e.g. either software developer or end-user), you will have a better shot at answers, I suspect. It appears to me that there is not enough information presented in your example to retrieve a pick model of the same attributes.

Thanks. --dawn Received on Wed Apr 21 2004 - 18:43:19 CEST

Original text of this message