The Dream Machine by Mitchell Waldrop: a short review

The Dream Machine, which is nominally a biography of J.C.R. Licklider, is actually an overview of the history of computing from M.I.T.’s Whirlwind effort through the beginnings of true personal computing in Silicon Valley; much of the book concerns ARPA and ARPAnet. Lick’s biography is embedded in the story, but its purpose is to center the discussion. The predominant focus of the book is on the efforts of Licklider’s colleagues, and it often strays far from his life story.

This is a terrific book. The writing is lucid, the research–though predominantly from secondary sources–is excellent. If you plan to read one book about the ARPA computing effort, this should be that book.


This short review was originally published on LibraryThing.

I’m Feeling Lucky by Douglas Edwards: a short review

This is a better book than I anticipated. Edwards–one of Google’s earliest hires–was obviously fascinated by Google’s founders, and the culture of the company they created. We watch as they repeatedly reorganize the leadership structure–an important concern for a middle manager–and as the author learns how he can contribute to the company. It’s an interesting, nitty-gritty view of the office (and its politics) from a privileged seat. This is well worth your time.

Google has resemblances to Carnegie Steel. Like Carnegie, Google is closely controlled, respects statistics, and is consciously disruptive. New technology is constantly put in place; failed projects are scrapped and forgotten. The leadership worries a lot about competitors, and embraces change as a competitive tool. Small edges are constantly devised and implemented, while big, industry-changing innovations are rolled out with astonishing regularity. Also: Like Andrew Carnegie, Sergey Brin and Larry Page are kinda preachy, and seem blind to some of the impacts and pitfalls of their colossus.

Andrew Carnegie eventually retired, and worked hard at giving away his fortune. His successors–JP Morgan allies–rebuilt the company into another model. It seems probable that Google will meet a similar fate, and that worries me far more than the casual arrogance of the company founders.


This short review was originally published on LibraryThing.

Computing in the Middle Ages by Severo Ornstein: a review

This was definitely not the book I expected, but is well worth reading. Ornstein has things to say, and knows how to say them.

The author was involved in computing from the mid-fifties to the early eighties, and played fairly important roles in the SAGE, TX-2, and Linc projects, all of which are key to understanding how computing developed. He also was heavily involved in BBN’s pioneering Arpanet efforts, and moved on to Xerox PARC in its prime, where he helped design the first laser printer. So he had a first-hand view of the development of electronic computing in the period between the pioneering efforts and the beginnings of microcomputing. This is a different, quite personal, account of what his computing projects were like, and his assessment of the issues as they looked to the participants during the period.

So there’s little new here, but there’s a level of detail about specific efforts, and about the personnel involved, that the journalists and historians who’ve tackled the topics lack. There’s also a quite-deliberate recasting of the context, which is Ornstein’s excuse for writing the book; he thinks the more formal histories impose more design (or perhaps a destiny) on the efforts than was actually there. Interesting stuff, with wry humor.

The chapters have odd, amusingly victorian, titles. For instance, Chapter 5: “A piano enters the lab and comes up against TX-2. DEC is formed and there is an error on Page 217. Fourier is proven sound and we land on an aircraft carrier.” The book might be worth reading just for those.


This review was originally published on LibraryThing.

Electronic Computers by Saul Rosen: a review

A journal article, not a book; available here.

This is easily the best short survey of the early history of computing I’ve seen, and is well worth a read. It’s an excellent 30 page essay describing electronic computing history through the late 1960s, with most significant projects and companies briefly sketched and their contributions–and failures–described. The essay is organized by technological era (vacuum tube, transistor, early ICs), with each era’s discussion organized by company or project. Some effort is made to put each project into historical and technical context.

The author mostly ignores early electo-mechanical computing projects, and almost completely ignores efforts outside the United States. This significant shortcoming is acknowledged in the introductory section.

While the emphasis is on describing projects, the author provides quite a bit of analysis. Details, of course, are sacrificed for brevity’s sake, and for focus. A fascinating, and well-done, survey.


This review is also posted on LibraryThing.

Dealers of Lightning by Michael Hiltzik: a review

The overarching question is “Did Xerox let the PARC technologies escape?” And the answer is “Of course.” Of course.

Hiltzik argues–successfully, I think–that the question oversimplifies the reality, in several dimensions: Xerox did use some of the Palo Alto Research Center creations, Xerox didn’t really have the ability/agility to implement others, and that clashing cultures made some gains difficult. He also explores the strengths and weaknesses of Bob Taylor’s management practices at some length (an interesting thing, actually, as I’m also reading Katie Hafner & Matthew Lyon’s Where Wizards Stay Up Late, which also features Taylor as a key player). Finally, he points out, Xerox continued to fund those rebels it supposedly didn’t listen to.

Well-written & researched. Each chapter is thematic, and mixes contextual explanation with word portraits of the key players. The book is very much focused on PARC’s computer and system projects, but aware of external events occurring more or less simultaneously with them and how those interacted. It acknowledges, but doesn’t really explore, the non-computer activities occurring in the center.

All in all, one of the better histories of early computing.


This review was originally published on LibraryThing.

Bright Boys, the making of information technology by Thomas J Green: a review

This was painful to read; I was hoping for something much better.

First we must deal with the title. Despite a subtitle which implies a book about the development of computing, it would be quite as reasonable (and as inaccurate) to describe the contents as an exploration of the birth pains of the Air Force as it transitioned from the Army Air Corps. For the most part the Bright Boys of the title remain shadowy figures working on unspecified problems. In contrast, the principal officers of the emerging Air Force are well drawn, have clear objectives, and argue well-described, sometimes conflicting, opinions.

The author seems to have no sense of proportion. Assuming he really intended a book about the MIT Whirlwind computer and its developers, he’s presented far too much detail–roughly half the text–about the development of the Air Force. There’s surprisingly little about the personalities of the Whirlwind participants or the details of their technical accomplishment. He’s far too dismissive of the clearly-legitimate concerns of the Navy’s research arm about the costs and objectives of the Whirlwind project. Moreover, much of the material seems misdirected; while he asserts the importance of Whirlwind’s technical mastery as the principal source of nearly all subsequent developments in information technology, he offers little to support that assertion. The result is highly disappointing.

More problems: The book’s organization makes it difficult to follow the chronology of both the overall narrative and the individual stories the author tells within the larger tale. There’s an enormous amount of redundancy, as Green’s general narrative style is to circle endlessly around his topic. While the author usually writes clearly, occasional paragraphs are absolutely incoherent. And while it’s obvious that Green’s read an enormous amount of material while preparing his manuscript, it’s far less clear that he mastered that reading. Finally, he dismisses the generally-acknowledged contributions of other early computing projects as irrelevant or misguided, excepting only the ENIAC project.

This is a fanboy’s history of the Whirlwind project. Although there’s some valuable information here, I find it impossible to recommend to anyone but a specialist reader.


This review was originally published on LibraryThing.

Grace Hopper and the Invention of the Information Age by Kurt Beyer: a review

This book is not, in any meaningful sense, a biography of Grace Murray Hopper. There’s a perfunctory sketch of the first 36 years of her (pre-Navy) life, and some mention of mid-life depression and alcoholic binges, but otherwise the book is fully devoted to describing her career in computing, her impact on the industry, and (to some extent) the development of both hardware and software in places outside her immediate purview. For all practical purposes this book ends with the standardization of COBOL; Hopper’s subsequent career is only lightly touched, and her late-in-life celebrity is briefly described in the first chapter but not really discussed.

On the other hand the book covers Hopper’s essential achievements extremely well, and is perhaps the best survey of early computing technology (and its associated community) I’ve seen. Like most early-computing books to date, it’s not intended to be a comprehensive survey; within its intended scope, it’s truly excellent.

DO NOT buy this book from MIT Press as an e-book (but see my update below). Their e-book definition is limited to online reading; you log into their website and use their software to read the book. This software isn’t terrible, but it’s quite frustrating: It can’t remember your last-page-read, it doesn’t remember your font settings, and (at least for this user) doesn’t fit well on the computer screen. While it’s possible to annotate the book, and to make highlights in the text, the software seems to prevent copy/paste activities, thus unnecessarily complicating note-taking. And then there’s the preposterous notion that storing books in many web locations is a reasonable way to manage my library. All in all, a bad experience; I rather wish I’d purchased the book in hardcover. But I’ll not do that after investing $20 in the electronic copy, however unsatisfactory I find it.

Update 22dec10: MIT Press, I’ve discovered, also permits offline reading using a client called iOffline, which was built using Adobe Air and more or less duplicates the online reading experience, including most of the limitations mentioned above. I think this is a slight improvement, but I’m still annoyed with the situation.


This review was originally published on LibraryThing.

Fire in the Valley by Paul Freiberger and Michael Swaine: a review

Another pleasant reread of a personal computing history book I originally read in the 1980s.

The authors–both of whom edited computer publications as the stories developed–tell the story of the beginnings of the PC revolution from the perspective of Silicon Valley. Their version heavily overlaps Stephen Levy’s Hackers, which was published a few months later, but it’s a very different tale in style and substance.

For one thing, this is a less literary effort. It’s also differently focused, as these guys care more about technical details than Levy does. And the largely west-coast perspective lets this book examine relationships in ways Hackers’ structure didn’t permit.

The book consists of many short sections, organized into eight thematic (and roughly chronological) chapters. While the sections are related, they’re essentially independent. It’s pretty common to find more than one version of a story/encounter within the book, often in widely separated places. This in no way harms the narrative; it’s just a quirk of the book’s organization.

Because the book’s nearly three decades old, some of the context seems a little odd. In particular, a pervasive fear of IBM dates the book–not to say the fears weren’t real, but we now know IBM had a significantly different impact than the PC industry expected. Similarly, there’s essentially no recognition of the immense power Microsoft would come to yield in the industry, and (of course) no clue about Apple’s long stagnation, and resurrection. And the Internet has no presence in this book whatever.

A very worthwhile effort. If you’re interested in this era’s history, you should read both this book and Levy’s; their differences and their similarities are both instructive.


This review was originally published on LibraryThing.

Hackers by Steven Levy: a review

I originally read Hackers when it was new, as I’d already been reading Levy’s computer explorations elsewhere. Rereading it was great fun.

Steven Levy pretty much framed the way we remember the early personal computing days with this book. This is a collection of stories–three main ones, and a host of smaller tales within the large ones. The main stories cover the MIT hackers of the 60s and 70s, the Bay Area folks revolving around the Homebrew club who perfected the modern personal computer, and Sierra Online’s birth as an important gaming company. The smaller stories are mainly about individuals and events. Over the course of the book the emphasis changes from truly personal hacking to the ways hackers interacted with the development of their obsession into an important industry. This journey into business defines the book.

Then the first appendix circles back to MIT, and Richard Stallman, and the nature and culture of hacking. Two other appendices, written for the tenth and twenty-fifth anniversary editions, do some updating–and muse, once again, about the nature, costs, and benefits of the changes the main book sketches.

Levy managed to identify many the key figures in the computing revolution while that revolution was occurring, and they make appearances in the story. But most of the book’s characters are minor actors on the world’s stage, and could be seen as representative types were they not so clearly individuals. We see, again and again, youngsters (almost always boys) mastering machines (or code, or both), then trying to master themselves. The outcomes vary enormously.

Levy might have written a different book–perhaps Berkeley rather than MIT, ARPANET instead of Homebrew, Peachtree might replace Sierra–and reached similar conclusions. I knew midwestern PDP-7 hackers every bit as obsessed as any described here, and they’d have made fine fodder for Levy’s musings. And he’d have found hackers in IBM, had he looked. But Levy’s story works, and we’ve largely internalized it.

Terrific book. Well worth your time, if you’ve not read it.

Late note 8/22/10: Freiberger and Swaine’s Fire in the Valley approximates the alternative book I describe a couple paragraphs up, at least in part. Both books are worth reading; in fact, I’d call them both essential to understanding this industry’s roots.


This review was originally published on LibraryThing.