Re: Database design

From: Mark Johnson <102334.12_at_compuserve.com>
Date: Tue, 21 Feb 2006 23:23:24 -0800
Message-ID: <304ov1pmss2f4rnjb8n8fic06okkhqhun5_at_4ax.com>


"dawn" <dawnwolthuis_at_gmail.com> wrote:

>2D, given that each value can be accessed with its [i][j]. This
>misunderstanding crops up all the time, and I fault those with the
>mathematical definition...for this one because I think they really do
>understand the alternative use of the term, while the programmers don't
>always know the mathematical definition.

This apparently sore spot with an issue of, flatness?

The context is always a hierarchy. Those who object are well aware of this. They attempt to make a point, as it were, without addressing the point. And so one is saying one thing, another insisting it wasn't said, and saying something else.

A relation is flat in that there is no ordering. There's no sort. It is a set of instances, of entities (though they will not allow multiple senses of this term), just simply, tuples. There is nothing above. Nothing below, in this context. And those who object, know exactly what is meant by this. And furthermore, I had to ask - isn't there? If you have a relation of sentences, or plain text passages, one cannot necessarily precede any at all, but must appear in proper order. At what point would a sort on a relation be termed an ordered relation, or is the mere appearance or suggestion of such explained away as another 'thing', some 'else', something not a "relation"? in which case it might strike one as mere semantics, as the term is commonly used. And then definitions are tossed about, perhaps even without meaning. Received on Wed Feb 22 2006 - 08:23:24 CET

Original text of this message