[Foundation for Object Relational Databases: The Third Manifesto ] is a book by some of the progenitors of relational databases that presents further extension of the principles. It notably addresses shortcomings of:
SQL;
"Object oriented" database systems.
A severe problem with SQL systems is that they haven't strongly followed Codd's relational rules . Consider:
View Updating Rule
No commonly available RDBMS provides any ready way of automatically handling updatable views, which severely injures their usability. PostgreSQL was a pioneer in providing ways of building update rules, but it makes no attempt to handle this automatically (though this may change in version 8.2...).
Physical Data Independence
Oracle and MySQL are particularly bad in this regard; you need to be able to predict data usage and generation patterns beforehand in order to be able to competently select from EXTENT policies, table types, and tablespaces.
One of Knuth's rules is that " Premature optimization is the root of all sorts of evils." Typically, SQL systems do not merely tempt you to do so, but pretty much require that you prematurely optimize the system in order to achieve acceptable performance
Integrity Independence
Most SQL databases provide only rudimentary capabilities for defining integrity constraints. Some implementors, such as MySQL AB , have a history of downright ridiculing attempts to have the DBMS manage integrity.
In addition, locking semantics vary substantially from system to system, throwing in the user's face issues of whether the system is using row locking, page locking, table locking, and the like.
Unfortunately, the book is a difficult read, as it is not quick to present, in a readable form, a prescription for what the authors would commend as the Right Thing. They concentrate on theory, and commit a vast amount of material to pointing out the shortcomings of SQL and so-called "object relational" database systems. The database language they describe, [ Tutorial D], has not been fully implemented in a form so as to make it fully usable. And in that they do intend it to be an alternative to SQL, it would be rather valuable to present it in a more " positive" fashion. Reportedly, Fabian Pascal's Practical Issues in Database Management: A Reference for the Thinking Practitioner provides more direct prescriptions for how to work around the relational weaknesses of SQL.
An early version of their paper that presented in the ACM SIGMOD journal may be found as [The Third Manifesto] published in PostScript form.
The CQL database system is seeking to provide an implementation of the specification described in the [Third Manifesto]. Several other systems are also trying to implement parts of [ Tutorial D].
http://www.hughdarwen.freeola.com/TheThirdManifesto.web/Missing-info-without-nulls.pdf
Windows-based implementation of D4, similar to Tutorial D
The Opus Programming Language Web Page
Windows-based language that is a cross between C and 3rd Manifesto "D"
Duro - relational database library
A GPL-licensed project implementing a library supporting parts of D for C programs, storing data using Berkeley DB