Home -> Usenet -> comp.databases.theory -> Re: Does entity integrity imply entity identity?

Re: Does entity integrity imply entity identity?

From: David Portas <REMOVE_BEFORE_REPLYING_dportas_at_acm.org>
Date: Tue, 4 Aug 2009 22:48:16 +0100
Message-ID: <bqadndoEiME-MeXXnZ2dnUVZ8gWdnZ2d_at_giganews.com> 






"Bob Badour" <bbadour_at_pei.sympatico.ca> wrote in message 
news:4a7884c6$0$23751$9a566e8b_at_news.aliant.net...

>




> Rule 3 is absolutely iffy. I tend to consider Rule 4 solid. Some might 


> question the necessity of Rule 4, and I know I have some serious 

> reservations about the catalog design Codd proposed in RM/V2, but it seems 

> perverse to me to deprive those who need to manage data the best tools to 

> manage the data about the managed data.




I have trouble understanding the 13 rules as anything very profound or 
significant. They are nothing like a set of necessary and sufficient 
definitions and to me are more like informal observations on some properties 
of a relational system. Rule 12 seems especially odd. As far as I understand 
it at all, Rule 12 is basically a rule that says you can't break any of the 
other rules!



We have much better science of the relational model today than in 1985 so it 
does puzzle me why people still cite Codd's Rules as if they feel they are 
still important.


-- 
David Portas


Received on Tue Aug 04 2009 - 23:48:16 CEST












Original text of this message