Re: 3 value logic. Why is SQL so special?

From: Cimode <cimode_at_hotmail.com>
Date: 16 Sep 2006 15:00:57 -0700
Message-ID: <1158444057.660487.86310_at_i42g2000cwa.googlegroups.com>


David Portas wrote:
> Cimode wrote:
> >
> > Not the point. A UNION has nothing to do with a JOIN. They are both
> > separate operators applyable on relations. Trying to prove that one is
> > derived from another is just
> > meaningless, the same way you would try to derive a DIVIDE operator
> > with a PLUS operator...This is a dead end.
> >
>
> In Codd's algebra that's true but only because it doesn't have a
> complete set of operators. JOIN is the relational AND operator and
> UNION is a special case of the corresponding OR operator. Via De
> Morgan's laws they CAN be defined in terms of each other. NAND or NOR
> plus REMOVE are the truly primitive operators according to Date and
> Darwen. Details in TTM(3).
Let's say that such additions do not me feel always confortable because it gets RM way too far from the math background that supports it..

Nevertheless, you are correct. Which is the precise reason why such definitions should not be done accordingly to SQL 's LEFT OUTER JOIN as initial comments were made. Only leads to additional confusion I was trying to clarify to an audience that has no clue about the difference between operators and hears that both LEFT OUTER JOIN and UNON a perfectly equal. Distinguishing the two operators is essential to begin use them. Thank you for helping clarifying...

Regards...

> David Portas
Received on Sun Sep 17 2006 - 00:00:57 CEST

Original text of this message