Ed,

>Walter, it's truly amazing the number of people with years of experience with a plethora of database products, that use a number of different forms of database, that you maintain don't understand the relational model. Is it possible that they do and you are sadly mistaken?

I've graduated on the subject, so I pretent to know at least the basics of the relational model. My graduation project was going to be used for education material at the universaty, i don't know if they actually did.

The main problem, is that the theory of the relational model, can't be tracked within any database product. Sure, some properties are reqocnizable with in a product like SQL-server, but even the training courses don't give you the picture of the relational model. You've got to read documentation about the relational model itself and try to understand it troughely. I do subscribe to a local hardcode DBMS (called DB/M) magazine. And even here it's hard to find the rules of the relational model.

BTW there are some (R)DMBSs which have attempted to grow more towards the relational model. INTERBASE had some extra support, but to my knowledge they dropped it for unknown reasons.

Part of the problem is that SQL is not a correct implementation of the relational model tough it has some simularities with the relational language. Most of the relational operators cannot be found in the SQL standard, they have to be managed with subqieries, maybe intermediate results etc.

So if you know SQL, you absolutely don't know the relational model. The only person I know of here that at least knows what the relational model is about is Jim Booth.

Of course even Jim and I have our differences how the relational model is to be implemented, but that's what this discussion is all about.

If anyone is familiar with the relational model I would gladly discuss this with him/her.

Walter,