Books March/April 1999

by Sue Petersen



Questioning The Millennium
by Stephen Jay Gould
Random House Publishing, 1997
ISBN #0-609-60076-1
192 pages, $17.95 USA, $24.95 Canada, hardcover

As I write these words, it is 396 days and a few hours until the dawn of Y2K. When you read this, it will be considerably fewer. Like all of us, Stephen Jay Gould has been staring down at this approaching milestone for all the days of his life, and he's got some thoughts to share upon the occasion.

Gould deliberately avoids prediction and psychology, instead he chooses to explore calendrics, astronomy, and history. He asks what is the millennium, when is the millennium, and why are we so fascinated by such 'special' transitions in such an arbitrary system of counting?

Heaven knows, if our code isn't ready by now, it's too late for anything but triage and damage control. Nevertheless, it's interesting to think about how we got here, and indeed, where exactly here might be. Gould is a poet and a philosopher, as well as a scientist. His prose sings, and never more so than in this book.


The Invisible Computer,
Why Good Products Can Fail,
The Personal Computer Is So Complex,
And Information Appliances Are The Solution

by Donald A. Norman
The MIT Press, 1998
ISBN #0-262-14065-9
340 pages, $25.00 USA, hardback, alk. paper

Don Norman is one of those people I've learned to buy on sight, so it was a no-brainer to grab his latest book when I spotted it sitting by the checkout counter on my last book store run. I wasn't disappointed, exactly, but I will say it took me longer to 'get into' it than it has any of his others. If that happens to you, be patient and give it some time. His thesis is interesting, and while I don't buy all of it, I do think it's worth paying attention to.

Norman talks about too many topics to cover in depth, so I'll just mention the ones that struck me as I read. He describes the lifecycle a new technology goes through as it matures, and the kind of companies that prosper at each point in the cycle. Along the way, he explains why the 'best' products don't always win in the marketplace, and his examples go all the way back to Thomas Edison's phonograph. Norman believes that successful products in a mature market must be task-driven, not feature-driven. He explains what this shift will require of computers and the implications it will have for the world in general and for our industry in particular. His eventual goal is the information appliance--a small tool that is specially tailored for its particular task. Instead of the monolithic PC on your desk, you'll have a separate appliance for each thing you wish to accomplish... I'm not totally convinced, I have nightmares about designing the interfaces between all of this stuff, but it's an intriguing idea.

As I write this, the Microsoft antitrust case is in the news daily and I found another part of Norman's thesis quite compelling. He talks about the difference between substitutable and non-substitutable economic goods. The first type allows markets to be shared and competition to flourish, the second leads to inevitable domination by a single player. Government regulation can interfere with this process, but nothing else can. Most applications software is substitutable, when I need a word processor it is useful for me to have several different companies competing for my dollar. Operating systems are non-substitutable. Like the continent-wide electrical grid or universal phone service, we must agree on basic standards for the back-bone layer so that substitutable goods (lamps, hair dryers, spreadsheets) can flourish in the market that the infrastructure creates. This has very strong implications for the antitrust battles now being fought out in the courts.


Tales From The Tech Line
edited by David Pogue
Berkeley Books, 1998
ISBN #0-425-16363-6
127 pages, $10.00 USA, $14.00 Canada, paperback

This hilarious little book is a compendium of computer mishaps that the author has garnered from email, internet postings and interviews. He credits the original source wherever possible, but I suspect that many of these stories are destined to become urban legends, if they haven't already reached that lofty status. Read it and giggle, but between the giggles, remember... These hapless people are not stupid. As Don Norman so succinctly points out elsewhere--they are confused, confusing and misled. They are the victims of nonsense, not its perpetrators.


Visual Explanations : Images and Quantities, Evidence and Narrative
by Edward R. Tufte
Graphics Press, 1997
ISBN #0-9613921-2-6
156 pages, $45.00, hardback

Tufte is another author I buy on sight. Visual Explanations isn't as coherent as his first book, The Visual Display of Quantitative Information, but it's still a must-read. Tufte says that Visual Explanations is about "pictures of verbs." Here, he's interested in the best, most effective ways to present information about "motion, process, mechanism, cause and effect." As always, Tufte discusses ways we can increase the honesty and information density of such displays. In the first chapter, he gives a wonderful critique of a supercomputer model of a severe thunderstorm, and demonstrates how to improve the information content of the display. His revised chart is worth framing, both for its sheer beauty and as a wonderful example of data display.

One of the most fascinating chapters in the book is titled "Visual and Statistical Thinking: Displays of Evidence for Making Decisions". Here, he explains how the 1854 cholera epidemic in London was stopped when Dr. Snow used data analysis and display to demonstrate that the Broad Street pump was contaminated by an unknown substance. (The cholera bacterium was not discovered until 1886, but Snow's work proved for the first time that cholera is transmitted by contaminated water.) Tufte contrasts this use of data analysis to the tragic decision to launch the space shuttle Challenger. The link between low temperatures and O-ring damage was there in the data, but their charts did not present the evidence in a way that would highlight it.

Tufte believes that charts and visual displays must be much more than passive pictures. The display of data should be an active process, where theories are formed, tested, proved or discarded. His books belong on the shelf of every programmer and programming manager.


The Psychology of Computer Programming, Silver Anniversary Edition
by Gerald M. Weinberg
Dorset House Publishing, 1998
ISBN #0-932633-42-0
384 pages, $50.45 (includes $5.50 for shipping UPS), paperback

I won't even pretend to be objective about this book. Jerry Weinberg is a mentor and friend, as well as a colleague and one of the best thinkers and writers our profession has produced. I've been following the progress of this new edition for a couple of years. I had the opportunity to comment on the manuscript in progress, as did many of his other students, but I was busy and never did get around to reading it. So I came to it fresh when it recently dropped upon my doorstep.

Psychology was written in 1969, published in 1971, and has been in print continuously since then, a record no other computing book can boast. In this new edition, Jerry looks at where we were 30 years ago, where we are now and where we might be in the future. Instead of changing the original text, he's added new comments to each chapter. This allows the reader to compare and contrast his thinking over the decades, showcasing the errors and omissions as well as the threads that bore fruit.

The technology in Psychology is obviously outdated, but I found it fascinating nonetheless to note the 'current' ideas that were to become so influential later on. Psychology was written before Fred Brook's Mythical Man-Month, but Jerry makes a passing reference to Brooks' thoughts on Chief Programmer teams. It was written a year after Dijkstra's "Go To Statement Considered Harmful" ignited a firestorm of controversy but before Structured Programming had swept the field, and Jerry makes several comments about readable code that demonstrate the ferment of ideas that Structured Programming arose out of.

Most of the 'people-stuff' in Psychology has stood the test of time better than the technology. Jerry's comments are as relevant today as they were 25 years ago.

"...good programmers are made, not born; therefor we should turn our attention to the manufacturing, or training process." (page 176)

This one issue--communication--has been at the core of Jerry's work for decades. Unknown to him at the time, Psychology was to form the outline of his life's work. He's spent 30 years expanding on Psychology and the roots of his more recent work are evident here. It isn't my all-time favorite Weinberg book, Becoming a Technical Leader probably has that honor. But Psychology is valuable as history in a field that is all too ready to repeat the errors of its past. Read Psychology as a picture of where we've been, where we are now, and where we need to go next. Read it as an index to the thinking of one of the most influential figures in our field.

Copyright © by Sue Petersen. All Rights Reserved.