Re: Does entity integrity imply entity identity?

From: Mr. Scott <do_not_reply_at_noone.com>
Date: Sun, 9 Aug 2009 21:18:37 -0400
Message-ID: <HoidnXQ9VL3w6OLXnZ2dnUVZ_hidnZ2d_at_giganews.com>


"paul c" <toledobythesea_at_oohay.ac> wrote in message news:cCpfm.40191$PH1.35653_at_edtnps82...
> Mr. Scott wrote:
>> ,,,...
>> I probably should have included D and E columns that don't allow nulls.
>>
>> (X,D,E) "X is an order placed on date D by customer E."
>> (X,A) "A is the e-mail address for order X."
>> (X,B) "B is the snail-mail address for order X."
>> (X,C) "C is the delivery phone number for order X."
>>
>> Then there would be no question as to whether an order table would be
>> needed in addition to the order_e-mail table and the order_snail-mail
>> table.
>>
>
> Okay, that's clearer. Now I get the impression that SQL doesn't support
> all the FK constraints desired, Is that so? If so, does that mean that
> SQL can't test relations for equality?

SQL doesn't support foreign keys that reference views, but I'm not so sure if that is a bad thing. I don't see what it has to do with whether SQL can test relations for equality or not. Received on Mon Aug 10 2009 - 03:18:37 CEST

Original text of this message