Re: Tarski school influence on Database Theory
Date: Fri, 25 Sep 2015 00:40:23 -0700 (PDT)
Message-ID: <dd5bcfab-6beb-4b52-91b4-945d904d4fe4_at_googlegroups.com>
> Besides inventing relational algebra, Codd also initiated and championed query safety, integrity, normal forms and other issues involving application of relational algebra to database management, ie the relational model. What really would have helped the database field but that *didn't* happen is if people had paid more attention to what Codd was saying about being relational.
(i) Relational Algebra.
(ii)
It is not true that Codd invented the "First normal form". Codd added "First
Regarding the integrity, I can say that this issue has been thoroughly
researched, before Codd worked on the topic. Here I will mention just the
integrity, which refers to the keys. Theory and software producers, did a
lot of things. For example, the theory and software manufacturers have
introduced the primary key, foreign key, invalid key, symbolic key and many
other details related to the keys and integrity.
>
> philip
I agree that Codd invented relational algebra.
Integrity and normal forms.
Regarding the normal forms, I must say that Codd did not invent the "First
normal form." As a novice programmer, I had the opportunity to witness
conflicts among senior project managers, that are on this topic. Project
managers from one group have used variable length of the records, and
project managers from another group have used records that have a fixed
length (that is, they were working with the first normal form). In both
cases, I had the chance to see a very good, very complex and very
interesting solutions.
So the idea of "First normal form" was performed and analyzed in detail
before Codd. All the advantages and disadvantages of "First Normal Form"
were well analyzed in very complex cases. Note that variable length of
records and entities, we can not apply to relations.
The third thing I would like to point out is that there were some good
ideas, which due to not understanding are not accepted.
What is true, it is that Codd has introduced functional dependencies as
mathematical theory. So Codd has raised this story about the keys on
scientific level.
However, it seems to me that your sentence "Codd also initiated and
championed ... integrity ..." is not true.
(iii)
The main reason for this my post is the following your statement:: "...
involving application of relation algebra to database management, ie the
relational model."
In my opinion there is a big difference between relational algebra and relational model. I'll try to explain this difference. As I presented it in my discussion with Jan, the most important models in human activities are mathematical models. Roughly, this is about the following triplets:
Man ------- Model ------ Real world( real objects )
For example in architecture, an architect (man) made the model (it is an architectural drawing of a building) and on the basis of this plan, we build a real building. So according to the above-mentioned triplets, we have:
Architect -------Plan------Building
In other words, we have an architect with his solutions, ideas and thoughts.
Then we have a drawing that is an architectural plan that was developed in
mathematical notation. It's a model.
And finally we have a real object in the real world, that is, the
corresponding building.
In the architectural drawing (model) we have simple mathematics, mostly we apply geometry.
In the the relational model, we have a very complex mathematics. The relational model is completely done by Gottlob Frege, 120 years ago. (You can see Frege's "relations" in my post of September 24, 2013, in thread "Sensible and NonsenSQL Aspects of the NoSQL Hoopla".)
Relational algebra as opposed to relational model, mainly dealing with derived data (various queries, views, reports, etc.)
On the other hand, the relational model entails much more; semantics, logic,
language etc. Frege's discovery of semantics and a number of other important
things, determines the above-mentioned triplet, completely. The relational
model is just part of Frege's theory.
Frege's discoveries fully describe and linking the following large areas:
thoughts --- model (language, meaning, logic...) --- real world; into one
whole. However, E. Codd never mentioned Frege's.
Instead, he attributed predicate calculus to his countryman B. Russell. For
example, in his work of RM / T, section 3, E. Codd writes the following: "A
database are structured will then consist of two parts: a regular part
consist of a collection of time-varying relations of assorted degree (this
is sometimes called extension) and an irregular part consist of predicate
logic formulas that are relatively stable over time (this is sometimes
called the intention, although it may not be what the logicians Russel and
Whitehead originally intended by this word). "
The work of B. Russell that is related to predicate calculus is
insignificant.
When it comes to B. Russell and his relationship to the the work of G.
Frege, in my opinion, a major scientific fraud was made by B. Russell. This
is about one of the most important scientific work in the history of
civilization, because the scientific work of G. Frege, has a great
significance on mathematics, philosophy, logic, semantics, theory of
language, theory of mind and a number of other areas including the theory of
databases .
However, if you know for any contribution that was done by B. Russell in
predicate calculus, let me know.
(iv)
This thread began with the topic of binary relations. As you know, the Codd
tried to solve the problem of binary relations in his "paper" RM / T. Binary
relations are a fundamental concept, because they are atomic data
structures. Therefore binary structures represent the basic elements which
are used to build data structures.
It was clear that the binary structures must have the following format:
(Key, OneAtribut). Here, the key has to be simple. Otherwise the binary
structure would not look like as the binary structure, if the key has n
elements.
Codd is completely solved the RM / T, and he did it immediately in his
paper, without any proof and without any theory. He introduced a simple key.
It is a surrogate key. This Codd's key is "invisible", which is nonsense.
Codd also introduced "P-relations", they are just binary relations, with one attribute. That is what we need -there . Codd also introduced "P-relations", they are just binary relations, with one attribute. There is nothing new in this "P-relations," we know how "binary structures" should look like. That is what we need. But, these are structures, for which we must have proofs of many important things. And also we need procedures, that will bring any data structures in the form of binary structure.
Codd did not do any proof, procedure or scientific theory related to the construction of the binary structures.
Regarding that Codd's paper about binary structures do not belong to the domain of science, it is surprising that the work RM / T is accepted and published.
In my papers I showed how to construct the atomic structures and how to
build a proper theory about the atomic structures. This new theory is very
different from Codd's relational model.
Codd's followers Date and Darwen have also tried to solve the problem of
bimarnih relations. They went to the "end". They did it in the same way as
Codd - without proof and without any theory. They presented the binary
relations in the following form:
Relvar R is in 6NF if and only if it satisfies no nontrivial JDs at all
Is this a theory? No, this is not theory.
What is this? The authors gave a name for the relvar.
What is that name? This name is "6NF".
What does this name denote? 6NF is the name which denotes a relvar that
satisfies no nontrivial JDs.
So, the authors of 6NF have not given a procedure that leads a relvar to 6NF.
But this procedure is only thing that is important here.
The second thing I want to point out, is that 6NF is not normal form.
Formally, it is a normal form. But in fact, it is a fundamental element for
the construction of a data structures.
Therefore, 6NF is failed attempt about most fundamental things. We need to
say clearly, 6NF has no the scientific value, it's just a name.
I write about binary relations presented by Codd, Date and Darwen for the
following reasons:
- Their work on binary structures is not a science.
- They tried to take the place in advance, someone who would scientifically and really solve this problem.
The authors of "Anchor Modeling" put in the title of their work, the following text:
"Anchor Modeling An agile modeling technique using the Sixth normal form for structurally and temporally evolving data"
As I have explained, "6NF" is just a name. Therefore Anchor Modeling can be
understood as "a technique", which uses the following name: "6NF". I want to
point out the extent of oddities, which may be obtained by crossing these
"theories".
Note that the work "Anchor Modeling" won first prize in the ER, 2009.
(v)
Codd is the first, who constructed data model that is fully mathematical
model. It is the most important value of Codd's work and therefore Codd's
papers have lasting value. However I think that my comments in this post are
important.
Vladimir Odrljin Received on Fri Sep 25 2015 - 09:40:23 CEST