Re: what are keys and surrogates?

From: rpost <rpost_at_pcwin518.campus.tue.nl>
Date: Thu, 14 Feb 2008 18:50:11 +0100
Message-ID: <68947$47b47f53$839b4533$30541_at_news1.tudelft.nl>


Bob Badour wrote:

>Gene Wirchenko wrote:
>
>> One might not have traced ones ancestry back that far. Say you
>> have gotten ten generations back. How do you record great-to-the-8th
>> grandparents? You do not have data on your great-to-the-9th
>> grandparents yet and so do not have FatherIDs and MotherIDs for your
>> great-to-the-8th grandparents.

Absolutely, but that holds for every possible data model. I don't see the relevance of your remark.

What I was trying to establish is that situations may arise in which we have to work with this information, in which no other attributes can be used; and that a data model with references (oids) is the natural way to model such a situation.

>What happens when you learn more about your great-to-the-9th
>grandparents and you fill out more and more information, and then it
>turns out one of them is the same individual as one of your previously
>known great-to-the-7th grandparents?
>
>How exactly does OID help in that situation?

Again, this may happen in any data model; I don't see the particular relevance to a model with oids.

-- 
Reinier
Received on Thu Feb 14 2008 - 18:50:11 CET

Original text of this message