In a 12 April 1988 PC Magazine article, "OS/2: A New Beginning for PC Applications," Charles Petzold restated what had become by then the party line amongst the personal computer prognosticators: "Microsoft expects OS/2 to establish the foundations of PC operating systems for the next decade."


Robert Hummel begged to differ. Only a few pages later in his PC Tutor column, he made one of the most spot-on predictions to grace the magazine.

Years from now, when programmers sit around and wax nostalgic, someone is sure to ask, "Remember OS/2?" Everyone will chuckle. Despite the hype and fanfare, I believe OS/2 is going to be short-lived. Rather than getting an improved DOS, we've gotten a new, completely incompatible operating system.

Or as reader Patrick Anderson stated in the 11 October 1988 Letters section,

All the gushing over OS/2 is amazing. It shows how far out of touch the gurus in Redmond and the magazine editors in New York are with real PC users.

But along with Robert Hummel, Ray Duncan was keeping in touch. He'd previously predicted that it'd take ten years for "OS/2's successors to eclipse MS-DOS." But in two October 1990 issues, and then in the 15 January 1991 issue, he drastically collapsed that time frame. Writing in his 16 October 1990 Power Programming column, Duncan observed that

Somewhere along the tortuous path from the original implementation of OS/2, thing went badly awry. A system designed to provide users and programmers with a painless migration path from DOS was transformed into a system designed to sell hardware and compete with Unix.

Two weeks later, Duncan counted up an installed based of 45 million DOS users, and short of an outright catastrophe, predicted 100 million DOS users by 1995. Microsoft, he advised,

should reconcile itself to the marketplace's resistance to the size and complexity of OS/2, and commit itself wholeheartedly to making DOS everything that it can be—regardless of the impact this might have on Microsoft's Joint Development Agreement with IBM or on OS/2 sales.

In fact, he was handing out advice that had already been taken. Microsoft had indeed fully committed itself to "integrating Windows into DOS," and would soon abandon OS/2 in favor of the massive installed base of DOS and Windows applications.

The momentous event—the dissolution of the Joint Development Agreement between IBM and Microsoft—happened that year. As with the hiring of David Cutler in 1988 to design Windows NT, it took a while for the news to leak out, and then everyone was so committed to the established storyline that it took even more time for the news to sink in.


In the meantime, DOS powerhouses like WordPerfect and Lotus invested heavily in OS/2. They were caught flatfooted when Windows took off like a rocket and never recovered. IBM acquired Lotus and it slowly faded away. In the worst deal of the decade, Novell bought WordPerfect and then sold it a few years later to Corel for pennies on the dollar.

Rumors of the "great divorce" between IBM and Microsoft had circulated the previously year, finally prompting coordinated press releases from the two companies in September 1990. The statements "reaffirmed their relationship" and extended the licensing arrangements for DOS, Windows, and OS/2.

"Semantic content: zero" was how Ray Duncan summed up the substance of these press releases. Authoring two separate articles in the 15 January 1991 issue, he again cut to the heart of the matter:

Although IBM and Microsoft agreed to cross-license everything, they committed themselves to nothing in the way of marketing the cross-licensed products. I suspect that Microsoft took a hard look at the startling success of Windows 3.0, compared it with the dismal penetration of the desktop market by OS/2 after three years (less than 2 percent by the most optimistic estimates), and decided to cut its losses.

Two weeks later, John Dvorak observed that "there has been much chitchat about a falling-out between IBM and Microsoft with denials all around, and more and more evidence indicates that the two are going in opposite directions." But then in the 30 April 1991 issue, Dvorak hedged his bets once again to pooh-pooh a report from January of that year.

The biggest fiasco in the industry was the obituary written in Wall Street Journal recently when OS/2 was pronounced dead. Microsoft was supposedly going to drop the product and concentrate on Windows. After all the facts were straightened out it seemed that nothing changed except there was even more talk of a portable OS/2.

Well, the Wall Street Journal got it exactly right. Microsoft handed OS/2 development back to IBM and concentrated its efforts on Windows and the Win32 API. This guaranteed that compliant programs written for DOS-based Windows would also run on NT, thus staving off the drought of applications that had plagued OS/2 from the start.


With that (mostly) "painless migration path from DOS" now in place, the fate of OS/2 was sealed. Five years later, in the PBS documentary Triumph of the Nerds, Steve Ballmer recalled the moment when everything went sideways.

We were in a major negotiation in early 1990, right before the Windows launch. We wanted to have IBM on stage with us to launch Windows 3.0 but they wouldn't do the kind of deal that would allow us to profit. It would allow them essentially to take over Windows from us, and we walked away from the deal.

After a decade of tumultuous growth, the weirdest marriage in American corporate history was over. And yet the true believers still couldn't believe that digital Mom and Dad were really getting divorced.

Related posts

The future that wasn't (introduction)
The future that wasn't (1/7)
The future that wasn't (2/7)
The future that wasn't (3/7)
The future that wasn't (4/7)
The future that wasn't (6/7)
The future that wasn't (7/7)

The accidental standard
The grandfathers of DOS

Labels: computers, economics, tech history, technology