At first sight, this work is an elegant, concise survey of progress in research on aspects of distributed algorithms. The preface gently warns the reader that “the purpose of this book is not to present an overview of the entire field of distributed algorithms, but instead it treats several subjects in depth.”

After a closer look, though, the reader cannot help but be disappointed. The book has no real structure. It is divided into five chapters, titled “Introduction,” “Synchronization in ABD [Asynchronous Bounded Delay] Networks,” “Assertional Verification,” “Distributed Infimum Approximation,” and “Garbage Collection.” The overview of topics presented in the introduction would have been a wonderful opportunity to stratify and catalogue different aspects of the problems and solutions, but instead it offers a seemingly random arrangement of topics. The chapter titles similarly reflect a somewhat ad hoc grouping of topics. My attention was drawn to the chapter on assertional verification, where I hoped to find an insightful, fundamental treatment of the subject. It is of course not entirely fair to blame an author for not excelling or inspiring, but the book does seem to come so close that it is almost a disappointment that it does not quite deliver.

As a small example, the conclusion of chapter 3 reasons that “All proofs [discussed in the chapter] are formalizable. A protocol skeleton can be refined to a complete protocol, allowing the programmer to tune the protocol to his needs.” Everyone who has designed an algorithm or a protocol and given an informal or semiformal proof of its soundness goes through this devious thought process: all my proofs are by definition trivially true and could be formalized, given enough time. The misleading part is that precisely that step is the most valuable and the hardest to get right. (Who has the patience to prove something that they already “know” is true?) Probably the most important lesson we must learn from formal design methods is that shortcuts such as “could be formalized” or “could be refined” hide the cobwebs in our reasoning. No wonder I am disappointed that Tel does not help to puncture this bubble. From the rest of the book, it is clear that he could easily have done so. So, at best, my recommendation is mixed, in the mild hope that Tel will write another book soon: the real one.