Re: Distributed key constraints

<jingleheimerschmitt_at_hotmail.com> wrote in message news:1114671340.993232.131080_at_z14g2000cwz.googlegroups.com...
>
> Roy Hann wrote:
> > Is there any commercially available dialect of SQL which implements
> > declarative distributed key constraints? (I'm betting not.) Failing
> that,
> > has anyone ever proposed syntax them?
> >
> > Roy
>
> If by distributed you mean 'in more than one database', then I'm not
> answering your question, but if by distributed you mean 'in more than
> one table' then, yes, there is one system that would allow you to
> enforce such a constraint without resorting to stored procedures or
> client-side enforcement. But it's not SQL. Dataphor by Alphora supports
> transition constraints which can be used to write an expression for
> each table involved in the constraint and enforce a key that spanned
> multiple tables.

Thanks Bryn, yes, this is what I meant. I was already aware that Dataphor could do it the way you describe, and I have rolled my own transition constraints to kinda-sorta-nearly do it in SQL. (Tedious, but surprisingly powerful!)

> If this isn't what you meant, sorry. Darwen has used the term
> distributed key to describe a key constraint involving multiple
> relation variables, but I prefer the term generalized key because it
> avoids precisely the confusion we may be seeing here. I believe he did
> propose some syntax for them in the paper in which he introduced the
> term.

I shall have to try to dig that out. (I was unaware that he originated the term.)

Roy Received on Sat Apr 30 2005 - 13:07:53 CEST