A new book about a certain software company that can't seem to keep a low profile these days promises to tell all, but does it?
Barbarians Led by Bill Gates, by former Microsoft developer Marlin Eller and co-author and journalist Jennifer Edstrom, delves into the strategies and blunders that have made Microsoft what it is today. Based on Eller's personal recollections as well as interviews with numerous current and former Microsoft employees, the book provides windows into everything from the software giant's efforts to videotape a competitor's product demonstration to its handling of antitrust regulators.
Eller, who worked at Microsoft from 1982 to 1995, served as a lead developer for Windows' graphics, and later for the company's Pen Windows project. He also was involved in its first Internet efforts. Edstrom is the daughter of Pam Edstrom, the public relations guru of Microsoft's top dog, Gates.
The book sets the scene for Windows' birth during 1982's Comdex in Las Vegas: After stopping by VisiCorp's booth and watching a demonstration of its graphical interface, VisiOn, Gates fears his competitor in the applications space will use the product to jump into the operating systems business.
Six months later, after studying Xerox PARC's Star System, which featured icons such as folders and in-baskets, developers Dan McCabe and Rao Remala assembled a prototype that mimicked VisiOn.
"[The Windows team's] only guidance from Gates was [an order] to squash VisiOn," the book says. "As for technical direction, the Windows team was left to their own devices."
Eller also recounts a time early in his career when he inadvertently insulted Gates and his technical ability.
As Eller tried to fill in the background of a round digital clock with color, using a flood-filled algorithm, it didn't work. He was able, however, to write some code that enabled the BASIC code to correctly flood-fill the clock.
Pulling Gates over to view the initial problem, Eller said: "Who was the jerk who wrote this brain-dead piece of shit?" Without waiting for a response, Eller demonstrated how he had fixed the problem.
After Gates noted that the new algorithm was "nice" and returned to his office, a colleague informed Eller it was Gates who was the "jerk" who initially had written the "brain-dead" algorithm.
More titillating, though, is the book's revelation that dumb luck and good timing, rather than any master plan, ultimately led to Microsoft's divorce from computing giant IBM.
"Everyone close to the process knew what a fluke the Windows 3.0 success had been," the book maintains. "Until the first million copies shipped, the Windows group had always hung randomly off of someone's organizational chart."
As the book tells it, even though Gates initially wanted Microsoft's NT team to develop a system that would run OS/2 applications, developers made sure the OS/2 could be jettisoned if the platform did not succeed.
As relations between Big Blue and Microsoft faltered, the developers proposed eliminating support for IBM, and Gates agreed. Microsoft then released Windows 3.0 in 1990, even though at the time it was publicly supporting Big Blue's OS/2 operating system.
Microsoft's competitors believed that Windows, which shipped two million copies in its first six months, was the strategic work of nefarious brilliance, initiated by Gates.
But according to the book, representatives from Big Blue got their first look at the NT interface in early 1991, while giving a presentation to IBM's rank and file about Microsoft's new 32-bit Windows API set. It was painfully obvious that nothing from OS/2 PM had been preserved, and as a result, the book states, IBM and Microsoft abruptly parted company.
Eller also includes an anecdote that could provide some insight into charges that Microsoft has been overly aggressive in the strategies it employs to beat out its competition.
He describes the efforts of GO, a developer of the Penpoint operating system, to demonstrate its product to the Boston Computer Society in January 1991, and says that Microsoft's Wink Thorne joined the society a day before the demo was set to take place. Thorne allegedly attended the demo, videotaped the presentation, and returned with tape in tow to Redmond, where Eller's Pen Windows group screened it and dissected it for intelligence.
"[Eller's] group spent the remainder of the month making sure they could demo everything the GO system promised," the book says. "They even added some features."
As for Microsoft's Internet efforts, the book's authors claim that the software giant was still "oblivious to the potential of the Internet" as recently as 1993.
Apparently Nathan Myhrvold, Microsoft's chief technology officer, outlined in a memo ways in which Microsoft could expect to leverage Windows, by making it the "ubiquitous platform for online services." According to the book, Myhrvold explained how Microsoft could expect its online revenues to easily surpass its operating system revenues.
At the time, Gates had approved funding for a project, then code-named "Marvel," which was designed as a proprietary system that ultimately would be a clone of America Online. Marvel eventually became the Microsoft Network, which has faced stiff competition from the moment it hit the Internet scene.