Re: Nested Relations / RVAs / NFNF

From: Marshall Spight <>
Date: Wed, 27 Oct 2004 17:36:57 GMT
Message-ID: <Z0Rfd.258396$wV.95677_at_attbi_s54>

"Kenneth Downs" <> wrote in message
> Marshall Spight wrote:
> >
> > So, on a related note, it's a little weird that we use integers as
> > keys. Integers support e.g. addition, so we are allowed to add
> > two keys together. This is nonsensical. Perhaps, if we want
> > to have surrogate keys, something different from integers
> > might be in order. That brings me to my 2a) idea, which is a
> > datatype that only supports these operations, and none other
> > 1) make me a new one
> > 2) copy
> > 3) test two such values for equality
> >
> > [this idea is not original with me.]
> >
> Right now I'm digging around in PostgreSQL, so I relate everything to that.
> They allow type extensions, and have some built-in types, including various
> surrogate ID types, that behave as you describe.

Uh, can you be more explicit? You don't mean OIDs or the serial type, do you? I never looked at OIDs very closely.

> Those who dislike surrogate keys are probably turning green at the thought.

Heck with 'em. Surrogate keys are a fact of life. (Which doesn't mean natural keys aren't "better"-- but sometimes you don't have a natural key, and you gotta have *some* key.)

Marshall Received on Wed Oct 27 2004 - 19:36:57 CEST

Original text of this message