Re: Very simple question to relational theorists.

On Friday, August 24, 2012 1:19:27 AM UTC-7, Evgeniy Grigoriev wrote:
> Philip, thank you for answer.

From your messages and papers and quotes and misquotes, you don't understand the relational model. SQL is a mess as a programming language and a mess at being relational. The OMG and O/R camps do not understand the relational model; SQL object extensions are partly not relational, partly messily relational. Your proposals are essentially these. You need to learn about the relational model before you try to unify it with something.

> It seems that we think in opposite directions. You assert that (meaningfull?) predicate is main condition that the relation exists. I assert that any formal relation by definition has a predicate - regarless to its meaning "for DBA". Who of us is right? Fact is that if we forget about "meaning" we will mean the same formality.

The formal system maps relations to relations but also predicates to predicates. You only understand the relation part. You do not understand the predicate part. You do not understand the connection between the parts. So you do not understand the formal system.

> > A formal system does not know meanings, but it produces meanings given other meanings.
> ??? "Formal system produses a meaning" ???

Is a calculus book about expressions like "f(x)" or is it about curves? If you give an expression for a curve then calculus is about getting an expression for another curve: of the slope along the first or of the area under the first. Put a meaning in, get a meaning out. The manipulation in between is per the "form" of expressions instead of their meanings: thus "formal" systems.

> What is "meaning of opertion" here? they are formal ones.

In language semantics we tend to use "meaning" for things mapped to. (It is better not to use that vague term but to name things: values, predicates, extensions, propositions, assertions, operators etc.)

There is a correspondence between relation operators and logic operators (connectives & quantifiers): JOIN with AND, UNION with OR, PROJECT with EXISTS, MINUS with AND NOT, RESTRICT and EXTEND with AND =. This gives the logic "meanings of the relation operators". This also gives the relational "meanings" of the logic operators.

> > That query meaning is in terms of the meanings of relations...
> > and meanings of the relation operators.

One maps between a relation variable name and its predicate and between a relation operator and its logic operator. Thus one maps between a relation expression and a predicate.

> >Codd's relation operators are designed specifically so that if the value of a relation is the set of all tuples that make its predicate true then a query result is the set of tuples that make its predicate true.
> Yes. It's formal rule. Fact is that this trueness doesn't depend on meaning.

You do not understand the important implications of this fundamental rule. It is about how the formal system produces not just relations from relations but predicates from predicates. Everything in the relational model follows from this.

The rule mentions the predicates ("meanings") of relation expressions that are built from given predicates ("meanings") of relation variable names and the above logic operators ("meanings") of relation operators. Thus that every relation expression corresponds to ("means") a predicate expression. And that every predicate expression corresponds to ("means") a relation expression. The rule says that if the value of a relation variable is set to the tuples satisfying its predicate then a relation expression returns the tuples satisfying its predicate. The relation variable predicates are how you set and read a relation variable. An expression's predicate is how you understand a query or update.

If you are not using relations with predicates this way then you are not being relational. If you don't understand this then you don't understand the relational model.

See my message (dense but precise) at:
  From: com..._at_hotmail.com
  Newsgroups: comp.databases.theory
  Subject: Re: Informal Survey #1 -- joins on foreign keys   Date: Sat, 5 Nov 2011 17:07:26 -0700 (PDT)   https://groups.google.com/d/msg/comp.databases.theory/Hjj_lxB8pXE/aoEDtm__To0J Maier's book is free online:
http://www.e-booksdirectory.com/listing.php?category=296 Hugh Darwen's bookboon.com free online book is also linked there.

More specific comments later about your operators.

philip Received on Fri Aug 31 2012 - 09:15:34 CEST