Re: relational reasoning -- why two tables and not one?

From: Bob Badour <>
Date: Wed, 14 Oct 2009 17:44:18 -0300
Message-ID: <4ad63826$0$23784$>

lawpoop wrote:

> On another mailing list about a database which shall not be named, a
> poster asked about a single-table database structure. It was to track
> donations, and that thank-you letters had been sent in reply.
> A couple folks ( myself included ) thought that there should be at
> least two tables -- 'donors' and 'donations' . But the poster argued
> that no, there would never be a holiday fund drive appeal sent out to
> all donors, or a year-end statement, or anything of that sort. So a
> single table would suffice.
> I tried to argue the point that having 'donors' and 'donations' more
> accurately modeled 'reality' , and failed. Another poster, who was in
> fact in favor of two tables, argued against the 'modeling reality'
> argument, saying that theorists would it's wrong to have
> 'donation_dates' as a separate table, even though the relationship "in
> reality" is one date for many donations. If it meets the functionality
> specs, it's fine.
> ( There was talk of needing Donors as an entity later on, but there
> was no need for it in the specs now, so it's moot ).
> I couldn't think of a good argument against it, so I must be wrong.
> But my gut instinct or intuition is telling me there is some
> understanding of relational theory or something that I am not
> grasping, which would prove insightful in this discussion.
> Thoughts and comments?

I don't see any particularly compelling reason for donors and donations relations based on the meager information available from your post. However, I suspect a compelling argument can be made for the necessity of donations and replies relations. Received on Wed Oct 14 2009 - 22:44:18 CEST

Original text of this message