Re: WWW/Internet 2009: 2nd CFP until 21 September

From: Mr. Scott <>
Date: Sat, 15 Aug 2009 12:13:02 -0400
Message-ID: <>

"paul c" <> wrote in message news:06jhm.41254$PH1.2906_at_edtnps82...
> Mr. Scott wrote:
>> "paul c" <> wrote in message
> ...;
>>> How can four values be substituted into a predicate with six
>>> place-holders?
>> I think the process is called currying. A n-ary predicate is a function
>> of n variables that maps into the domain of truth values. Each row in a
>> table is a function application that under the closed world assumption
>> maps to the positive truth value. An incomplete row (a row with nulls)
>> is a partial application that due to entity integrity still maps to the
>> positive truth value. What remains is not a predicate but a formula with
>> free variables that is known to be true.
> So instantiation requires currying and the result of instantiation
> includes free variables! Sounds like what Bob B might have been
> expecting.
> First I've ever heard that SQL depends on currying, let alone FOL and RA,
> didn't know they needed such an operation. Definitely sounds like a new
> RT operator (eg., couldn't possibly be a form of projection since that
> eliminates variables) for null fans..
> Sarcasm aside, I'm pretty darn sure all logical formulae/wffs that are
> eligible RT propositions or constraints can be expressed in relational
> extensions, given enough storage. Maybe more to the point to say that If
> they can't, then they aren't eligible (eg., otherwise relations aren't
> closed under RT ops).. If the ones that aren't RT-eligible are somehow
> eligible to be expressed in SQL tables, then those tables can't be
> expressed by Codd's relations, which I guess is just another reason why
> SQL isn't relational and it is phony to say RT applies to SQL tables and
> phony to use RT arguments when talking about SQL. I'd like to know where
> is the SQL 'Theory'.

What constitutes an "eligible" wff? Remember that Codd endorsed the use of null in the relational model. He devoted two complete chapters to them in his book 'The Relational Model for Database Management, Version 2,' contrasting the four-valued logic of RMv2 to the three-valued logic needed in RMv1. Isn't it true that in first order logic which is the basis for the relational model, Pabcdxy is an atomic well-formed formula even though x and y are variables? Doesn't Pabcdxy being true convey information about a, b, c and d, regardless of the values for x and y? Received on Sat Aug 15 2009 - 18:13:02 CEST

Original text of this message