Re: Urgent help please

"Harald Fuchs" <nospam_at_sap.com> wrote in message news:pufzjyf0xf.fsf_at_srv.protecting.net...
> In article <3f41c994$1_at_news.comindico.com.au>,
> "Alan" <alanpltse_at_yahoo.com.au> writes:
>
> > I have a table contains parent and children information,
> > this is not the actual table that I am working with but has similar
> > structure:
>
> > ID Name Parent NoOfChildren Age
> > 1 John NULL 3 50
> > 2 Mary 1 0 20
> > 3 Peter 1 0 19
> > 4 David 1 0 30
> > 5 Alan NULL 1 60
> > 6 Candy 5 0 25
> > 7 Bobby NULL 0 35
> > 8 Vincent NULL 0 18
>
> Drop NoOfChildren since it violates normalization.
>
Beg to differ. It might be logically redundant, but, afaik, it does not violate the definition of a functional dependency, nor does it violate normal forms. A relational expression value and a tuple attribute value both will be functionally dependent on the key and nothing but the key.

A table constraint would probably be beneficial in ensuring consistency between the derived value resulting from the relational expression and the tuple value, but this is an integrity issue, not a normalization issue.

Regards,

Daniel Guntermann Received on Wed Aug 20 2003 - 14:37:32 CEST