Using Teradata as an exemplar, Inmon describes the advantages and implications of a parallel processor whose architecture is particularly suited to the special characteristics of DSS processing. The book begins with a basic discussion of the difference between DSSs and operational (that is, application- or transaction-oriented) systems. It contrasts processing and technical requirements as well as the content and structure of data, and cites obvious and well-known reasons why operational systems form a poor basis for DSSs. This contrast becomes a recurring, and repetitious, theme throughout the book. The treatment is not inaccurate, but it is often naive, and few references to the extensive literature on this subject are provided.

Subsequent chapters cover a variety of topics loosely organized around the DSS development process. The author discusses data modeling, logical design, physical design, design review, development methodologies, and organizational responsibilities, all within the context of database machines and DSS. The treatment of data modeling is nonstandard and at too high a level to be more than an introduction. Subjects covered in logical design include granularity, sampling, summarization, and dual levels of granularity. Physical design topics include denormalization techniques, selective redundancy, selective merging, partitioning, and extracting. The chapter on design review provides nothing new in the way of process, but does contain a good list of questions and issues that need to be addressed during review. This may be the single most useful part of the book.

The chapter on development methodologies contains a good discussion of their shortfalls and then goes on to say that a DSS methodology will be described. The book provides only a set of figures suggesting a methodology, however, with no description.

The chapter on organizational responsibilities discusses the impact of DSS database technology on systems management, data administration, database administration, the information center, and DSS analysts, suggesting in each case a broader or deeper role than these units traditionally play. None of this is inaccurate, though the discussion is cursory.

Few references are provided, and even fewer are cited in the text. Most of the references were published prior to 1983, with none from the journal literature on either database technology or DSS. The glossary contains useful definitions, though many of the terms listed never appear in the text. Some acronyms appear in the text without definition and are not listed in the glossary (a notable example is “EIS/AI”). Teradata itself is described in chapter 4 but is alluded to, without explanation, on several occasions in the first three chapters.

The book contains numerous syntax errors (some egregious) and a few typographical errors. The writing style, like the content, is basic and often uneven. A number of sidebars on various topics are scattered throughout the text, seemingly with little relevance to the topic at hand; I found them distracting. The book contains no outright content errors, though the author makes occasional overstatements. The chapter summaries are generally good.

The author describes the book as appropriate for “managers, designers, developers, data administrators, data base administrators, systems architects, and students of computer science.” Because the treatment is general, it will be of use only to those whose knowledge of computing is fairly elementary, or to managers with an indirect interest in database machines or decision support systems.