Programming Windows: Marvel (Premium)

In the 1990s, Microsoft experienced a level of growth that was unprecedented in personal computing or any other industry. It was all tied to the success of Windows, an operating environment that many, including IBM, once Microsoft’s biggest partner, had initially written off. But where Windows had entered the 1990s with just a few million users, by the time the 1990s ended, it accounted for several hundred million users and over 90 percent usage and market share.

Microsoft’s success in the early 1990s wasn’t just about Windows. Instead, the firm parlayed its success as the gatekeeper to personal computing by establishing a family of productivity applications and then systematically destroying the competition, leaving companies like Ashton-Tate, Borland, Lotus, WordPerfect, and many others in its wake. It expanded into networking and workgroup computing, ending Novell’s leadership in those markets. It pushed NT into workstations and servers, defanging Sun Microsystems and other makers of high-end Unix systems.

To these companies and other erstwhile competitors, Microsoft was a voracious cancer, eating and destroying anyone or any company that it saw as a threat. It was as friendless in the industry that it had championed as it was alone atop of it.

And it broke the law, repeatedly, while establishing and maintaining its dominance. Microsoft became infamous for pretending to be interested in partnering with companies with innovative products so that it could find out more details and then bald-facedly copy what they were doing. So, the software giant had increasingly found itself on the receiving end of legal and regulatory challenges in the early 1990s.

The two biggest challenges were an Apple lawsuit that alleged that Microsoft had stolen the Mac user interface for Windows and the first of what would be several Federal Trade Commission (FTC) antitrust probes, which ultimately ended in a deadlock. We’ll be looking at these issues, in particular Microsoft’s many antitrust problems, later in this series. For now, it’s just important to know that with these threats eventually cast aside, Bill Gates and Microsoft were filled with the sense that they could never be defeated. So, the firm steamrolled ahead, with its eye on the 1994 release of Windows 4.0. Which would ship, finally, in 1995 as Windows 95.

But Windows 95 wasn’t the only product that Gates and Microsoft couldn’t ship on time in the mid-1990s. In 1993, Gates and Microsoft chief software scientist Nathan Myhrvold began work on a book about the future of the industry called The Road Ahead that was set for publication in 1994. The book touted Microsoft’s take on the “information superhighway,” which the two had renamed the “Information Highway” for some reason. But the book, like Microsoft, almost completely ignored the Internet, which was as that time the hottest topic in the rest of the world.

“I found Gates’ observations on the Internet weakest of all,” Seattle Times reviewer Paul Andrews wrote of The Road Ahead. “The World Wide Web receives just four index citations and is treated as a functional appendage of the Internet (rather than its driving force), and both come off as a subset of the Information Highway, a term Gates uses with abandon despite its disfavor among digerati.”

“For all his business savvy, Mr. Gates has been caught flat-footed by [the] sudden emergence of the Internet,” The New York Times review likewise noted in its own review of the book.

Gates would quickly address The Road Ahead’s obvious shortcomings in a major revision a year later that added 20,000 additional words and changed the focus to the Internet. But what is of more interest, perhaps, was how Gates and Microsoft had missed the obvious in the first place. All around them, the industry had been changing. And Microsoft, a company that had never ignored an opportunity or challenger was inexplicably missing the boat.

How on earth did this happen?

Bill Gates later said that the Internet first came to his attention during one of his famous annual week-long sabbaticals, in April 1993, that he called “Think Weeks.” During these solitary weeks, he would read and research, write lengthy strategy memos for Microsoft, and study what the competition was up to, uninterrupted by the daily requirements of corporate life.

Gates wasn’t impressed. The Internet was decentralized and could not be controlled by any one company. Worse, it was free, so there was no business model, as he saw it. What he focused on, was somewhat obsessed with, was yet another market that Microsoft could dominate by leveraging the growing market power of Windows: Online services. Companies like America Online (AOL), Compuserve, and Prodigy were raking in the bucks and growing quickly. So he told a key product manager, Russ Siegelman, to come up with a Microsoft strategy for online services.

After a buyout was rejected by AOL, Siegelman returned to Gates with a plan: Microsoft’s online service, codenamed Marvel, should be bundled with Windows 4.0 (later renamed Windows 95). Less than a month after his Think Week, Gates signed off on the strategy.

(Fun aside. When news leaked that Microsoft was working on a project named “Marvel,” Marvel Comics sent the software giant a legal threat explaining that the name was trademarked. Since Marvel was just a codename, however, Microsoft ignored the threat and never responded.)

Marvel needed to come together quickly because Windows 4.0 was originally scheduled to ship in mid-1994, a little over a year after it was green-lighted. So after experimenting briefly with just copying Prodigy, the quickly-growing Marvel team settled on some key tenets for the new online service. It would focus on consumers, not technical users. It would be based on the X-25 network type used by other online services, and not on TCP/IP, like the Internet. And because it would ship with Windows 95, Marvel would simply be integrated directly into the Windows 95 user interface: Users would access Marvel’s various services using the same Explorer windows they used to browse their PCs’ hard drives.

Around this same time, Gates was still convinced that the future of the PC was multimedia and interactivity, and he had directed Microsoft to purchase a Seattle-based multimedia company called The Daily Planet. A group from that company formed the basis of a new team that would build tools that Microsoft and others could use to create interactive content for Marvel. It was called Blackbird, and it would work like Microsoft Publisher, but for the online service.

Marvel needed a name, too, of course. So the team came up with ONVO, which stood for … well, nothing. The team just liked that it started with “on.” But when they presented this idea to Steve Ballmer, he literally threw them out of his office. “Fuck no,” he said when told of the name. Unable to come up with anything better, the team started simply referring to it as “the Microsoft Network.” That name stuck. As did MSN, as Microsoft often shortened it.

By late 1993, it was clear to some of Gates’ underlings that the Internet was important and would, in fact, usher in radical change to the personal computing industry. Rob Glaser, who would later found ProgressiveNetwork (later RealNetworks), was then researching Microsoft’s planned partnerships with media giants like Time Warner, and he was among the first to sound the alarm. The more he researched this topic, the more convinced he became that Microsoft should recast the Microsoft Network as an Internet service.

“I became convinced that Microsoft was building the last minicomputer,” Glaser later told Paul Andrews, the author of Overdrive: Bill Gates and the Race to Control Cyberspace. “The Microsoft Network was based on the notion that your competitors … were on-line services like America Online. But the reality was that the Internet was going to be a fundamental paradigm shift, [and] that you needed to think about your strategies fundamentally differently.”

Glaser’s ideas were never implemented, but it wasn’t solely because of tunnel-vision. Russ Siegelman suffered a life-threatening brain aneurism and the Microsoft Network team didn’t want to make major changes with him out for several months. Glaser explained the Internet to the team and implored them to move the Microsoft Network to this platform after it shipped in Windows 4.0. But few on the team were convinced. And a Glaser demo, using the early web browser Mosaic to view a live image of a coffee pot in England, fell flat with most. The team elected to simply keep the project on track and deliver what they had promised.

Defeated, Glaser handed all of the Internet materials he had collected to Bill Gates’ technical assistant. A man named Steven Sinofsky.

“There were only a few people at Microsoft at the time who were true Internet believers,” Glaser later said. “Steve [Sinofsky] was one.”

As hinted by Glaser’s Yoda-like recollection, there was another. James Allard, a dumpy-looking coder who would later undergo a personal transformation and rebranding as J Allard, also saw the potential of the Internet to disrupt Microsoft and the entire industry. And as 1993 turned into 1994, both Allard and Sinofsky would work from within to get Gates and Microsoft to focus on what they saw as an existential threat to the software giant.