O. Kagan 2007-12-23

0 of 0 people found the following review helpful:

From my first examination of "The Pattern on the Stone..." I expected it to be a beginner's guide to understanding how computers work (read: not a how-to book, or manual, this is mostly "big-picture"/"theory" book). Through a lucid organization and several examples per chapter, Daniel Hillis indeed explains "the simple ideas that make computers work."

At 153 pages, this is not an all-inclusive journey into the micro-history and deep inner-workings of today's machines; instead Hillis's method was to pick several important points (Boolean Logic, the idea of Bits, algorithms, heuristics, to name a few) and give the computing novice a clear understanding and a general landscape of where the concepts stand in the engineer's world.

While Hillis does much (both in the text, and the diagrams) to make the concepts (and by extension, the examples) nice for the non-mathematical mind, a few parts may still be too left-brained. The quality that redeems the un-inspiring diagrams and harder portions is that the book's organizational fitness still allows the reader to achieve the goal of the book; one can read on only halfway "getting" the example and still come away with a grasp of the overall ideas.

The overall ideas in "The Pattern on the Stone" are part of what make the book fascinating; Mr. Hillis does not shy away from giving the reader his perspective (re: parallel processing, for example), on the topics covered. First published in 1998, Hillis's visions of the future are not uninteresting today, some have even occurred. Moreover, the nine (almost ten) years that have passed have seen many changes in computing, I am willing to recommend "The Pattern on the Stone" because, due to its strong premise and the foresight of the author, it has not aged much remaining relevant to those looking for an introduction to the world of computing.


p.s. - If you are looking for a companion book to "The Pattern on the Stone," that goes more into the history of computing as well as some of the ideas covered, read Howard Rheingold's "Tools for Thought: The History and Future of Mind-Expanding Technology" (ISBN: 0262681153). Note: It is available for free online.

wiredweird 2007-12-15

2 of 2 people found the following review helpful:

Unfortunately, I'm a nerd, so I'm not in this book's target audience.

Hillis presents the big ideas of computing. This starts with a little about the ones and zeroes that underlie all of digital computing, and the fundamental idea of a "state machine". Next, Hillis describes the basics of programming languages. In order to keep this at the conceptual level, he deliberately avoids jargoneering, and mentions only an extremely simple language used in education. This builds up to a discussion of Turing's "Universal" machine. That's a mathematical abstraction, but it's what lets computer scientists talk about all possible computations, not just what one machine can do. Ideas don't get much bigger!

That completes Hillis's discussion of the machine itself. After that, discussion moves on to what goes on inside the machine, the algorithms and other techniques that get the answers. Along the way, Hillis mentions decidability - the computer scientist's statement that some problems simply can not be solved by computation, a form of Goedel's theorem. Without using tech-talk ("NP completeness"), he also describes problems that are exactly soluble in theory but not in practice. Their exact solution requires more millions of years than the universe has existed. Instead of hard mathematical rules for perfect solutions, Hillis describes heuristic rules of thumb for good-enough solutions. Discussion becomes more philosophical after this point, with musings on how to harness multiple computers to solve problems faster, and speculations on what might follow the current epoch of computing technology. Think about it - the technology is barely fifty years old, and has achieved incredible feats. What do the next fifty years bring? Or five hundred?

Hillis never descends to "magical" descriptions of what goes on in computers, but never loses his sense of wonder, either. I think that's the right balance, They are man-made machines, after all. Understanding them doesn't eliminate the awe however, at least not for me, it only sharpens it.

-- wiredweird

Bruce Kimball 2006-07-01

1 of 3 people found the following review helpful:

A cover quote from Kirkus Reviews says that this 1998 Basic Books publication is for parents who need a "straightforward answer" to their childrens' question, "how do computers really work." Naming the book "The Pattern on the Stone" isn't a very straightforward start, nor is the "magic in the stone" as the title of the preface. This is typical of the book's ongoing flirtation with things mystical and archetypal, although the title actually refers to the physical architecture (pattern) of silicon-based integrated circuits (stone).

Luckily, the book abandons these poetic ambitions and makes a serious and fairly successful effort to live up to that Kirkus quote (who is Kirkus - the captain of the Enterprisus?). I would recommend this book if you want to step back, prioritize some of the main ideas that make up computer science, and actually feel like you have gained some insight into these concepts. It's a little book - barely 150 pages, simply formatted with 9 chapters that focus on 9 key ideas. That would have to be the case if it's going to do what it says.

I found the beginning a little shaky and had a hard time following the first obligatory chapter on Boolean "Nuts and Bolts". It definitely gets better, with the keenest examples being Chapter 4, "How Universal are Turing Machines?" and Chapter 5, "Algorithms and Heuristics". Few books on this level attempt to explain the Turing machine, one of the truly fundamental ideas behind computers. It also takes on randomness, sorting algorithms, and parallel processing, to varying degrees of success. In spite of Kirkus, I agree with reviewers who say it's actually best to already have an acquaintance with computer science in order to get the most out of this book. It fits better on the shelf next to volumes on microprocessor architecture than next to Mother Goose.

The author is W. Daniel Hillis, a "Disney fellow." On the back cover he is described as "one of the hottest computer scientists today." For you doubters, there's a miniature full-body photo of him imprisoned by a web of colorful tinker-toys, one-upping the old Wrox full-facials. Rick Pracher, the cover designer, gets only one star from me, though he does a good job of visually expressing the ambiguity in the book's intended audience.

A specialist in artificial intelligence, the author does have an underlying philosophy, explicitly expressed in the last section, titled "evolving a thinking machine." He sees humans as fundamentally computer-like, which he says is more a compliment to the machines than an insult to the people. Luckily, you don't have to buy into this perspective to get a lot out of the book.

James Sack 2004-04-18

6 of 6 people found the following review helpful:

I am not a computer expert but I have been immersed in computer books for the past couple of years in order to jump-start a new career. I can say without hesitation that in my reading experience Daniel Hillis has no peer in writing about computers. I suspect he has no peer in his field of research as well, but I'm not qualified to comment on that. This book is a joy to read. So rare that a great technical mind can communicate as brilliantly as this.

Daniel Lamb 2003-05-04

4 of 4 people found the following review helpful:

I really enjoyed this book. Mr. Hillis has a way of writing that is both educational and entertaining. I had taken this book out of the library at least a half-dozen times, so many in fact that my wife suggested I just buy the darned thing (I had a enough late charges to pay for it at least once!) It is easy to read, engaging, and really does get to the heart of what a computer is and how it works in a manner that even my eight year old daughter could work with. Good work Mr. Hillis!