Re: "Structured" Entity-Relationship Model?

On Tuesday, April 30, 2013 7:30:54 AM UTC-7, Wolfgang Keller wrote:

> And just two (mutually co-dependent) constraints wouldn't be enough
> to correctly implement the semantics, as far as I understand.

Don't understand.

> database design books don't seem to explain it either.

It is not clear to me what you mean by "it".

> This issue (it's "just" a 1:n relationship with n>0, right?) must be
> as old as the relational database model itself (fourty years now?).

This focus on arity of relationships is an *ER* concern not a RM concern per se.

In the RM you associate a predicate about the world with each relation. The relation holds the tuples that give a true proposition from the predicate. The constraints are consequent true propositions about the base relation variable values; equivalently, they are truths about the world. Entities and relations are not particularly helpful concepts other than in the sense of predicate logic objects and predicates respectively.

Read Hugh Darwen's free online ebook http://bookboon.com/en/an-introduction-to-relational-database-theory-ebook . Fabian Pascal has a predicate-oriented ER-reminiscent relational approach and I suggest you check out www.dbdebunk.com. Terry Halpin's ORM is oriented towards predicated base relations. But ER is neither necessary or sufficient for proper (relational) modeling.

> Does someone in this group know of a book (or an online document) that
> shows how to *correctly* implement such things?

I'm not sure what you mean by "such things". ER models (relationally) Multiple simultaneous constraints? Circularities are logically irrelevant and arise as necessary from proper design of normalized predicates.

philip Received on Thu May 02 2013 - 05:10:38 CEST