Read it from CiteSeer: http://citeseer.ist.psu.edu/schewe98readings.html

216 pages
------

''Thanks for recommending FundamentalsOfObjectOrientedDatabases, I was wondering if I should buy that one; does it address some of the critiques of OODBs (such as Date's), or is it purely "OODBs are perfect and here's what you should know"? I'll buy it on your recommendation in either case. -- DougMerritt (at 216 pages, I'm sorry I was wrong in my assumption that I could buy a bound copy!)

FundamentalsOfObjectOrientedDatabases is free to download from CiteSeer. Of course no free on-line book is really free, the aprehension effort far outweighing the monetary value for all but the most expensive book. No, it doesn't address specifically Date's critiques, one reason being is that it is an edited collection of earlier articles -- that's why it provides a good overview of the important topics. Among the co-authors there's for example Catriel Beeri, one of the most important contributors to the development of the relational theory so any presumption of anti-relational bias goes out the window. I can only vouch that it has very useful insights even for practitioners, even as it is theoretically heavy. The first two chapters dealing with the ObjectIdentity problems once and for all, and even as they do not address directly Celko's position (DatabaseIsRepresenterOfEntities), the competent reader will know better than to buy Celko's story with the "tables model entities therefore we allow duplicates".

Date's critiques of OO database models (such as in TheThirdManifesto) are hardly addressed specifically anywhere in the scientific literature, even to this day. And there's maybe something to discuss about that: first most researchers are probably aware of them but they consider they're not worth addressing (beside the point). It's not only them, but apparently their reviewers as well. Some have it (according to Jan Hidders in comp.database.theory)  that the relationship between date and researchers in the field is one of mutual neglect, which is understandable to some degree. Again, an astute reader can make the parallel between say the theoretical framework presented in FundamentalsOfObjectOrientedDatabases and the critiques presented in TheThirdManifesto and see that the critiques do not apply. Date is certainly lacking in the documentation department: not one citation in the second edition of the ThirdManifesto is relevant for either the theory behind OO or OODBMS. Undoubtedly he would be more than displeased with an outsider knocking down the relational model by criticizing the account given by say JoeCelko (in his SQLxxx series of books), but in turn he doesn't see the perfect simmetry when all he quotes from the OO literature are the perfect OO counterparts to Joe Celko. --CostinCozianu

Aha, very interesting. -- Doug
----
See also: ObjectOrientedDatabase