Re: more on delete from join

From: Tegiri Nenashi <tegirinenashi_at_gmail.com>
Date: Wed, 26 Aug 2009 19:09:25 -0700 (PDT)
Message-ID: <06b68751-9359-42dd-bd09-4146b010da54_at_y28g2000prd.googlegroups.com>


On Aug 26, 5:24 pm, Bob Badour <bbad..._at_pei.sympatico.ca> wrote:
> Wouldn't it be easier to just say tuples are propositions?

Hmm, what are tuples? Here is somewhat unconventional perspective. Given a predicate

R(x,y)

and two constants a and b, then

R(a,b)

is a proposition, I agree. However, given the same two (scalar) values a and b we can construct relation

T(x,y)

that has only one tuple <a,b>. Here is my observation:

R(a,b) = R(x,y) /\ T(x,y)

where the /\ is set intersection join! (Set intersection join is natural join projected onto symmetric difference of attributes). Set joins are known to be closely related to quantifiers, and quantifiers bind variables. In our case, in order to get proposition R(a,b) which is 0-ary predicate, we have to bind free variables in R(x,y)... Received on Thu Aug 27 2009 - 04:09:25 CEST

Original text of this message