Re: Will ORM take off?

In article <3d2a8071$1_at_usenetgateway.com>, Nathan Allan <member_at_dbforums.com> wrote:
>
>I am suggesting that someone needs to create a visual modeling tool that
>can represent the ENTIRE relational model including all declarative
>constraints (that most DBMSs don't even support). This would be the last
>visual modeling tool we would ever need (right Ellison?) ;-).

Actually, it is not very hard to come up with such a language. You can for example use existential graphs (or socalled Peirce diagrams) to represent in a graphical the whole of first-order logic. And I would be surprised if there wasn't a way to extend this to higher order logics, and there have been extensions of ORM with set constructs.

If you want to do it textually you can even do it directly in first-order logic; the NOLOTs would become unary predicates, the LOTs also, and the fact types would become n-ary predicates. The objectified fact types would be a bit tricky, but even that can be solved.

It has always amazed me why the ORM people haven't paid more attention to this. It's so simple that as a database researcher I wouldn't even dare to write a paper about it, but it would be a very practical extension.

• Jan Hidders