Those of us who've worked in the computer industry sometimes find it hard to explain to non-techie loved ones what it is we do at work, why the hours are so long, why software doesn't do what one would it like to do, or why programmers like Chinese food, among other imponderables. When my own words fail, I resort to pointing out a book, such as Tracy Kidder's Soul of a New Machine, Paschal Zachary's Showstopper, Steve Levy's Hackers, or Michael Swaine and Paul Freiberger's Fire in the Valley. Each of these books is a vivid snapshot of a particular time and place, but also depicts situations that are generalizable to the rest of the industry, and from which we can often draw lessons.
One problem with most of these books, as useful as they are, is that they are usually written by a journalist or outside observer, rather than by those in the eye of the storm. Jerry Kaplan's Startup: A Silicon Valley Adventure (Houghton-Mifflin, 1995. ISBN 0-395-71133-9) is different: Kaplan was principal founder and CEO of GO Corporation, a highly visible pen-computing startup that in the early '90s enjoyed center stage as the leading player in an emerging industry, one that would revitalize a stagnant sector of the U.S. economy.
Kaplan's book is surprisingly well-written, and chronicles the history of GO, from initial flash of insight (a 1987 epiphany during a flight in Mitch Kapor's private jet) to the initial $1.5 million in funding (a commitment secured after a 10-minute pitch to a blue-chip venture-capital firm, without a formal business plan, financial projections or a technical VP) to the key initial product demo (the machine was assembled for the first time the day before, with no guarantee that it would work a second time) to early adulation by the press, to the insatiable thirst for capital ($75 million over five years, burning $2 million a month at its peak) that ultimately led to GO's demise and the undignified picking apart of its bones at public auction. Along the way, many paper fortunes vanished in the stroke of a pen, while a few lucky GO employees were able to cash in, in one fleeting moment by jumping ship to a derivative startup called EO, which was acquired by AT&T and dissolved shortly thereafter. (These employees were not necessarily the most hard-working or deserving of the 200+ employees who devoted countless burn-out hours furthering Kaplan's vision of pen computing.)
The book lives up to its jacket blurb about being a ``kiss-and-tell'' story about the birth, life, and death of GO. Kaplan depicts most of the movers and shakers in the personal computer industry, not always in flattering terms. Kaplan's story has the drama and timing of a page-turner, and is well-stocked with rapacious villains, pompous buffoons, back-stabbing martinets, and bullgoose-looney venture capitalists (Bill Gates, John Sculley, Alain Rossman, John Doerr). IBM is shown as a club-footed giant beset by a panoply of bumbling bureaucrats, beady-eyed wannabees, and the occasional power player (Jim Cannavino). After reading this book, you'll understand why OS/2 will never succeed, and why Microsoft rules the roost.
Of course, the narrative is guided (and warped) by Kaplan's personal agenda-as with any autobiographer, he cannot escape the all-too-human desire to portray one's own role in the best possible light.
Kaplan prefers to set the scene with a few well-chosen sentences, and let the characters portray themselves in their own words.
For example, here is a vignette about arriving, in mid-1988, at Apple's headquarters for a meeting with John Sculley:
Jean-Louis Gassee [Apple's VP of technology]...instantaneously appeared as if out of thin air. He must have been watching me eye the place. He leaned toward me and said: ``Obscene gross margins.''These almost comic-book characterizations add to the overall sense of drama, and seem true in essence even if they may lack verbatim fidelity. The author relies on his own recollection, of course, but bolsters his memory with a diary that was recorded verbally on tape three times a week during seven years of GO's existence. When transcribed, there were a thousand pages of notes, corroborated by 50 MB of archived e-mail and internal company documents and memos. A story this complex has many angles and players, and draconian filtering is a necessity.``What?''
``This splendor is the happy result of Apple's obscene gross margins. And do you know why?'' He spoke with a smooth French accent: ``Because we control the whole platform, both the hardware and the software. The rest of the PC manufacturers are going to the dogs, eating each other for lunch as they get squeezed by Microsoft on the software side and Intel on the hardware side.'' He gestured out the window, to the world outside Apple. ``Would you believe they want us to license our software to others?'' He narrowed his eyes at me, as though I should know precisely what fools he might be referring to. ``But we don't have any competition. I say keep the prices high and profits will roll in.''
Kaplan meticulously stays on the safe side of legal propriety. For example, he himself does not outright state that Microsoft stole GO's ideas and concepts and, through numerous Machiavellian machinations that bordered on the unethical, was able to get PenPoint ISVs to abandon GO in favor of Windows/Pens. Instead, he leaves it to other characters in the book:
[Robert] whispered through clenched teeth: ``They ripped off our stuff!''Kaplan also describes meeting with the FTC about Microsoft's trade practices and shortly thereafter getting a phone call from Bill Gates:``I'm just glad we showed Lloyd [the Microsoft engineer in charge of Windows/Pens] our ideas so long ago, so he could only copy our earlier stuff!''
``You got [******], plain and simple,'' said Michael Baum of Pensoft. ``They just figure you don't have the wherewithal to go after them.''
``[Gates said] I just hope you understand what you could be getting into. You know our look-and-feel dispute with Apple?'' Now Gates spoke with restrained emphasis, ``I've spent over four million dollars so far on that one lawsuit!''... We both knew what he was talking about. There was no way a small startup company like GO could possibly afford the cost of legal action against Microsoft.More than once, the journey undertaken by GO's staff is described as a ``joyride to hell.'' Although the book is a jolly good read, Kaplan's story is limited by some key blind spots, demonstrating that even a center-stage seat does not provide omniscience or self-awareness. One of Kaplan's goals in this book is to write ``a cautionary tale about what can happen to a young company when its timing is wrong, its technology too speculative, and its market not yet ready.'' He doesn't quite succeed, missing or soft-pedaling some key aspects of the story of GO.
One reason is that Kaplan's book is mostly devoid of technical content. Of course, this is largely because it is written for a mass audience; but part of the reason is that Kaplan, who has impressive technical credentials (doctorate in computer science, principal author of Lotus Agenda), gave up hands-on development early on, and seems to have spent most of his time on jet planes, speaking at conferences, seeking out deep-pockets investors and making deals with large-volume customers. Essential tasks, to be sure, but all the product demos, investment capital, and purchase orders in the world won't help if your product is unfinished and/or unworkable.
In early 1988, soon after GO embarked on its development, I was lucky enough to visit the company in a converted loft and see the early prototype, whose descendant would be unveiled to the general public three years later. I later spent a week at its first developer camp for its PenPoint operating system, and wrote applications for a PenPoint ISV. Along with many other software developers, I was impressed by PenPoint's innovative design and elegant architecture. There are aspects of Carr's creation, such as PenPoint's embedded-document architecture, that are only now being copied or finding their way into mainstream software systems (OLE, OpenDoc, Taligent).
Despite my respect for the talents of GO's technical staff, I came to the realization, a year or two before GO's doors were to finally shut, that PenPoint would not succeed. After seeing a prototype pen application that took four skilled engineers four months to build in PenPoint, I was able to knock off the same application in three weeks of part-time work, writing for Microsoft's competing Windows/Pens environment, using pedestrian tools like QuickC and Visual Basic. Although the Windows/Pens environment is a kludge-layered-upon-kludge and its convoluted architecture makes San Francisco's famous Lombard Street seem as straight as the Route 280 expressway, Windows' longevity in the marketplace gave Windows developers order-of-magnitude advantage in time-to-market (never mind that the market was largely nonexistent). The PenPoint system was like an elegant castle in the air, lacking the infrastructure and foundation of tools, documentation, support, or a sufficiently large native population of developers to support this structure (nevermind the bugs, or the four-minute boot time). Kaplan does not mention this, nor does he dwell on other shortcomings of GO's pen system, such as that the machine was too heavy and expensive, and the handwriting recognition was not adequate for its intended market (although, for that matter, neither was anybody else's).
Although by necessity there is much of GO's story that ended up on the cutting-room floor, Kaplan's book contains some interesting, and heretofore secret, tidbits. I was surprised to find out that Intel invested several million in GO, yet succeeded in keeping this investment a secret. Likewise, IBM almost blindsided GO by negotiating with Apple to joint-market the Newton. No matter how much you think you know about this industry, you'll learn something new from this book.