Issues with the logical consistency of The Third Manifesto

A while ago I read certain parts of the first edition of "Foundation for Future Database Systems: The Third Manifesto" by C,J. Date and Hugh Darwen. Later I also read parts of the second edition. This page might only make sense if you have read this book.

Date and Darwen presented the maxim: All logical difference are big differences and it's corollary All logical mistakes are big mistakes as central to their work on the Third Manifesto.

This maxim was the key point I learned from this book.

The Third Manifesto proceeded to identify what is refered to as the Two Great Blunders

• Equating relvars and classes
• Mixing pointers and relations (or more specifically allowing database relvars to contain object IDs)

However, in my humble opinion, the argumentation used to substantiate the claim that these so-called blunders are truly to be viewed as blunders is somewhat weak. This opinion is based on the following maxim: logical conclusions should only be drawn from logical premises.

I wrote a paper entitled A critical reading of the Third Manifesto which addresses some issues I percieve with the content of this book. This paper is available in PDF format.

Later I wrote an amendment of this article presenting a perspective based on Codd's position on the subject matter under discussion. As the objective reader can establish, Codd's position is incompatible with Date and Darwen on at least two points:

• Codd sees a specfic need for Object Identifiers (which he calls surrogates)
• Codd specifically states that the relational model for databases requires that all domains be simple domains. Codd's position is thus in opposition of that of The Third Manifesto which claims that the structure of domains is orthogonal to the relational model. It is because Date and Darwen claim that the structure of domains is orthogonal to support of the relational model, that this issue cannot be dismissed by asserting that there is confusion about the exact definition of a simple domain. Since it is obvious that the phrase simple used by Codd is incompatible, with domains of arbitrary complexity allowed by domains as defined by Date and Darwen in The Third Manifesto.

It is interesting to note that the position of Date and Darwen seems to be in flux. Consider for instance the position of Date articulated in Persitence Not Orthogonal To Type. Notice that much of the argumentation Date presents against ODMG styled Persitence Not Orthogonal To Type is also applicable to complex domains as introduced by The Third Manifesto?

The Dutch magazine DBM published a version of my critique of The Third Manifesto in their september 2004 and october 2004 editions.

Here is the initial response by Hugh Darwen to these articles as it is available on the size of DBM site and published in the DBM magazine. A more elaborate response is to also available on the TTM web site.

Since, I feel that Date and Darwen did not address my concerns in their response, I proceeded to write the article entitled A Gentle Introduction to Orthogonals. This example driven article proceeds to provide more evidence that Date and Darwen have erred in their assessment of the alledged great blunders.

The concept of an orhogonal as presented in this paper will show that complex domains can be accommodated without violation of the First Normal Form constraint. This fact confirms that from the perspective of data modelling, complex domains as introduced by Date and Darwen in The Third Manifesto represent a needless complexity, since using orthogonals, complex domains are shown to be nothing more than a generalisation of the concept of an identifying relationship.

In addition orthogonals allow for an intuitive and convincing argument in favour of Object Identifiers (a.k.a. surrogates ) which is well aligned with Codd's position in Extending the Database Relational Model to Capture More Meaning

After the appearance of the third edition of The Third Manifesto Databases, Types and the Relational model wrote an article entitled The Third Manifesto revisted. This article draws the conclusion that Date and Darwen seem to be regressing relative to Codd from certain perspectives.