Fundamentals of relational database management systems. (Q870667)

This is another database textbook, this time from Indian authors S. Sumathi and S. Esakkirajan. It covers the mainstream of topics that are worldwide included in database courses, including the relational model and its OO/OR extensions, database design, SQL as well as Oracle's PL/SQL, query optimization, concurrency control and recovery, data warehousing, security, distributed database management, as well as a number of more recent topics that represent new challenges for databases. While the selection of topics seems nice at first glance, the book suffers from a number of flaws. First and foremost, it does not point the reader to most relevant or even recent literature; indeed, there is no citation newer than 1999, as if the area had been frozen before the turn of the century. Moreover, many references are outdated (since, for example, newer versions of a book have appeared), and recent research is simply ignored, surprisingly even for such topics like XML and genome database. Next, a number of choices regarding the content are difficult to explain. For example, why bother with Oracle 8, where Version 10 has been around for quite some time already? In Chapter 3, there is a confusing distinction between ``set operations'' and ``database operations'' in the relational algebra; since relations are sets (although this follows only implicitly from the fact that rows are not ordered, p. 68), all operations on relations are set operations. In Chapter 7, the authors introduce the notion of a ``partial'' transaction, so far unknown in the literature. The section on data warehouses in Chapter 10 does not provide a single example of a dimension with several levels. This list could be continued. In general, most of the chapters kill the reader with the high number of terms that are (too) briefly introduced and not explained in depth. Finally, the book suffers from a number of syntactical flaws. The English needs polishing in many places, the use of capital and small letters appears stochastic, and there are quite a few typos (e.g., ``objected-oriented'' already on p. XIX. The selection of ``pioneers'' is pretty random. (Fagin is definitely one of them, but what about Vardi, Papadimitriou, Ullman, Maier, Silberschatz, to name just a few?) In conclusion, my feeling is that the authors have tried to cover too much material; the book would benefit considerably from restricting the material and treating the result with appropriate depth.