Re: Plea for help - Normalisation Problem

From: skotske <skotske_at_hotmail.com>
Date: Tue, 03 Dec 2002 23:09:55 +0000
Message-ID: <vRaH9.2806$8e2.1139_at_news-binary.blueyonder.co.uk>


André Gamache wrote:
> Hi:
>
> 1- The passage from 1NF to 2NF does not end up to the second relation
> CDs Borrowed(#LoanNo, #cdID). This relation is redondant. The primary
> key of the first relation should be a composite key (#LoanNo, #cdID).
>

Thanks for replying - do you mind if I ask a follow-up?

The first method that removes the repeating group to a new relation specifies (and I'm quoting directly from the book here)

  • remove the repeating items from the original table and form them into a new table
  • create a new attribute of the table which is the key of the original table
  • make the primary key of the new table this key value together with a qualifying value

but surely if this is the case (and the all other attibutes in the table are dependent on the qualifying value) then it'll always end up with this redundant table after 2NF and 3NF?

Otherwise is there a qualifying rule or method that tells you how to decided that the compostite key should be in the first relation rather than the second?

Or is the mistake I made is that I created the artifical cdID to replace the attibutes (Artist, Title, DateBought) that made up the qualifying value?

Thanks again for your help.....

S Received on Wed Dec 04 2002 - 00:09:55 CET

Original text of this message