Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> comp.databases.theory -> Re: Objects and Relations

Re: Objects and Relations

From: Keith H Duggar <duggar_at_alum.mit.edu>
Date: 14 Feb 2007 18:38:53 -0800
Message-ID: <1171507133.691165.6530@h3g2000cwc.googlegroups.com>


David BL wrote:
> JOG wrote:
> > I hope that the pennies really are falling. I am
> > seriously dubious as to the intentions of your posts.
>
> I have decided to completely change my approach. I am
> trying to completely ignore the subconscious urge to see
> things in the way I feel comfortable with, to try to see
> it from your shoes.

Good idea. Though to "master" would be more useful than to ignore. Regardless, I learned much from a similar personal battle to think and view without the entity crutch.

> I chose the Lego example because I thought maybe you were
> going to show me how it was useful to not think of the
> pieces as entities.

Being an intelligent man, you will learn more from applying an entity-free mind-set to your own work than you will from contrived examples. While doing so, pay attention to how in the end, what matters is not "entityhood" but rather values, facts, statements, and their relation to one another.

> That hasn't been the case in this example because the
> location information ends up uniquely identifying the
> Lego pieces anyway.

Perhaps you have a better example from your own work that you would like to share and discuss?

> I assure you that my intentions are sincere.

For what it's worth, I believe you. That is why I tried to you help you early on with some friendly cdt advice (from a fellow newcomer). Unfortunately you took them as flames.

> > And that's it. That's how we work everyday. If we ever
> > have to do deal with items that are indistinguishable by
> > anything but physical location (or any characteristic we
> > can't keep track of), we tag them with a surrogate
> > identifier to represent their unique nature. It's
> > just common sense really.
>
> If we indeed choose to introduce a surrogate identifier
> then I don't see why we can't call it an entity identifier.

Because like the word "exists", "entity" and hence "entity identifier" is semantically vacuous. Calling a tuple or a surrogate key an "entity" or "entity identifier" adds no meaning.

> The article in Wikipedia appears to confirm that.

Well damn, why didn't you say so in the first place? Now that we know what Wikipedia thinks there is little point in discussing it further. Right? ;-)

> Is there some other reason why you say that a key
> shouldn't (generally) be regarded as an entity identifier?
> Perhaps you could refer me to the relevant literature or a
> previous thread on cdt.
>
> > As an addendum, once implemented a surrogate key becomes
> > a natural key. I find this fascinating - it seems
> > somehow analagous to "Nature abhoring a vacuum".
>
> You lost me there. What do you mean by implemented? In
> what sense does it /become/ a natural key?

Search this group for "familiar surrogate".

Keith - Fraud 6 Received on Wed Feb 14 2007 - 20:38:53 CST

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US