The average reader encountering a book entitled “Computer hardware” would expect a discussion of this topic. This volume does not even come close. A better title would be Analyses of, and recommendations for, US government policies on computer hardware development. Similarly, the chapter called “An Overview of Technology” is a taxonomy, not an overview. “Development Trends” is on resource needs and available support for research and development. “Technology Management” discusses just one of many possible approaches to managing technology projects. The other chapter titles are equally misleading.

Much of the book is a waste of paper. Chapter 5 is unedited vendor sales literature, reset in a uniform typeface. McGraw-Hill should be ashamed of presenting this blatant advertising to paying customers as editorial material. Later, we get six pages of Texas Instruments promotional literature on artificial intelligence. Chapter 8 is a 75–page, rule-by-rule “expert system knowledge base” on computer hardware elements. The expert system is trivial from an AI viewpoint, hyped beyond belief in the text, and too long for its stated purpose of helping nontechnical people learn about expert systems; in content, it is simply a reformatted version of chapter 7. Many wide-ranging reports are included in their entirety, apparently to inflate the page count. They cover material that is irrelevant to computer hardware, such as the research programs of the National Institute of Diabetes and Digestive and Kidney Diseases and available research opportunities in arctic systems science and global tectonics.

Much of the text is out of date. Most of it was apparently written in 1985 and 1986. Discussions of East-West tensions reflect Cold War thought patterns and were outdated when the book went to press. Some material is even older, such as the reference to “further technical discussions to take place in May 1984” (p. 36); more recent material includes a mention of “the new Bush administration” (p. 188).

That is the bad news. Let us tear out over 130 wasted pages, lamenting the trees sacrificed to print them and their contribution to the price tag. Let us give the book and its chapters less overblown, more honestly descriptive, titles. What remains is far from worthless.

This book offers a taxonomy of the technologies that underlie computer hardware, with internationally used definitions and diagrams of their relationships. The discussion of US policies in technology development, control, and transfer is excellent, if slightly dated. A similarly thorough discussion of trends in development needs, funding, and support offers well-thought-out recommendations in this area. The author also gives reasons for having a supercomputing policy and recommendations for such a policy.

Fleer offers a novel but worthwhile approach to managing technology integration projects, based on exploding the end product into a tree of underlying technologies and managing its branches. He then gives a breakdown of all computer technology into such a multilevel tree. The top-level branches are systems hardware development technology, digital computer system utilization technology, logic and high-speed memory assembly technology, storage technology, digital computer display and peripheral technology, and hybrid computer technology. An example of a leaf entry is 1.4.3.4.1.1.b, which indicates laser beam positioning and servo techniques for optical disk read/write transducer assemblies. This breakdown is invaluable for policymakers who need both the big picture and the supporting detail.

My first reaction was that this book was a hastily edited collection of outdated reports and vendor hype. While much of the book fits that description, it is far more than that. It contains enough meat to be important reading for anyone interested in US government technology policy--what it is, what it should be, and why. The reader will suffer a bit on the way to enlightenment, but the result will be worth the trouble.