Re: The Theoretical Foundations of the Relational Model
Date: Thu, 04 Jul 2002 20:25:07 +1000
Message-ID: <3D242283.6080801_at_deepthought.com.au>
Jan.Hidders wrote:
>>what makes you think that OO models do not have n-ary relationships?
>>
>Because if they did they would be object-relational models. :-)
>
actually, that's a good response - the object "purists" should remember
that good object models are also (at least partly) relational models.
> But seriously, (althoug I'm only half kidding there) note that I said "many" and
>not "all".
>
yes, I didn't read properly.
>I'm quite aware that the UML object model supports n-ary
>relationships, but I'm not sure if that is relevant here because as far as I
>know there is no database that uses it as its primary data model and we are
>discussing this in the context of databases and not data models in general.
>
actually, even ODMG-93 compliant databases are not too bad in this
respect (althought ODBM-93 is terrible overall) - they use List<>, Set<>
etc to define not only multiplicity, but add some semantics as well
(uniquness, ordering etc). This probably grates against the
sensitivities of relational purists, but it is quite useful, and can be
handily mapped into executable class models in an OOPL.
- thomas beale