Re: How to normalize this?

On Wed, 1 May 2013 11:20:47 +0200
Jan Hidders <hidders_at_gmail.com> wrote:

> On 2013-04-30 22:17:44 +0000, Erwin said:
>
> > It's often bothered me that normalization theory/procedure seems to
> > quietly ignore the notion of "nullability or not" of any of the
> > attributes in the "initial table design" ...
>
> Nullability doesn't really matter for normalization. From a
> redundancy point of view null is just another value.

I guess I don't know what a "redundancy point of view is", but

1. Null is not a value. Null is a mark noting that the value is missing. I don't see how a non-value becomes a value depending on one's point of view.
2. Null is not "just another" value. A nullable column by definition has an optional relationship to the key. Any nullable column can be placed in a separate table with an FK relationship of 0:1 or 0:N cardinality.

AFAIK functional dependency analysis does not entertain nulls or address missing information. The database designer should pay attention to them, though, because they indicate the potential of more than one kind of relation in a single table.

--jkl Received on Wed May 01 2013 - 22:17:59 CEST