Public Discussions > General Discussion

Quora: Why did IBM's OS/2 project lose to Microsoft, given that IBM had much mor

(1/2) > >>

Martin Iturbide:
Hi. I found the need to repost this from Quora right here:

Why did IBM's OS/2 project lose to Microsoft, given that IBM had much more resources than Microsoft at that time?

Reponse from "Dave Whittle".

Great question. I'm the founder of Team OS/2 and IBM's first OS/2 Evangelist, so I lived through the answer to your question for a decade. There's just no easy answer - it's like asking what makes any given startup a success or failure - but I'm happy to share the way I see it.

First a few facts - from memory - you might find relevant that support your question.

- IBM spent more than a billion dollars developing and marketing OS/2.

- It was the most advanced small systems operating system of its time - the most secure, the best architected, and the most powerful - without question (unless perhaps you were a journalist defending your decision to give the nod to Windows, perhaps in anticipation of the legendary envelopes of cash that landed mysteriously on Microsoft-friendly reviewers' desks). One example, it had pre-emptive multi-tasking (now a staple in multi-core systems and operating systems) when Windows 3.1 was still running on top of DOS and context-switching was the norm for any other desktop OS. It would run multiple DOS, Windows, or OS/2 apps smoothly. It was reliable and almost never crashed - something DOS and Windows was prone to do regularly. Yet Microsoft slammed OS/2 in the press (and got the media to echo their whining) because it needed 4MB (MB! not GB) of RAM - "too much memory" - and could be crashed by Ballmer at trade shows using specially written code on a diskette.

- IBM's Personal Software Products (PSP) - the division I worked for - had more employees and was better funded than all of Microsoft in the early '90s. IBM was the dominant force in that relationship, much to Gates chagrin, but was nonetheless weak-willed in using its power aggressively. I constantly heard "we should take the high road" in discussions about dealing with Microsoft. I value ethics as much as anyone, but IBM's Business Conduct Guidelines assumed perpetual dominant status in the industry and because they handcuffed executives and employees, they were thus inadequate to deal with foul play on the part of an underdog business partner. I once heard from reliable sources that Gates had called executives at IBM to complain about my violation of IBM's Business Conduct Guidelines, without specifics. I'm pretty sure what he was talking about was how I got quoted in PC Week as saying that "Bill Gates's gift to the industry is a win/lose mentality." Wow, I was so guilty as charged.

So, with those facts established, from my perspective, here are the following lessons to be learned from IBM's failure to establish OS/2 as the "operating system of the future" - as Bill Gates once called it:

LESSON ONE. As a company, if you are going to adopt and insist on compliance with strict Business Conduct Guidelines (as IBM and many other company's did - similar to Google's 'don't be evil' mantra), be aware of your strategic vulnerability to a company (such as Microsoft) willing to use other company's scruples as both shield and weapon in their war against you. In the words of a wise man, you need to be "wise as a serpent, and harmless as a dove." IBM had the harmless down pat - but they were unable to outsmart the serpent.

Gates was brilliant in negotiating deals to take advantage of other companies' blind spots - including their ignorance of Microsoft's willingness to bend the ethical constraints honored by other companies. For example, Novell entered into a contract with Microsoft that allowed MS to include Novell's networking code in Windows Version 3.1 (a consumer OS where Microsoft was strong and Novell wanted to make inroads). Ever wonder why the first version of Windows NT was Version 3.1? Now you know. Windows first major business OS literally stole its ability to co-exist in Novell networks based on a contract specifically designed to keep Microsoft from using it in a business OS (where Novell was strong and Microsoft was weak). When Noorda flew to Redmond to try to avoid having to sue and work things out with Gates, after making Ray wait for seemingly forever, Bill's response was "So sue me." Gates knew that the courts were too slow. No wonder when later, the media asked Noorda why he didn't just have a heart-to-heart with Gates, his reply was "To have a heart-to-heart, you have to have two hearts." Sure enough, within years, Microsoft had literally stolen leadership in the networking market from Novell on the basis of stolen rights to use code.

Everyone also knows that Microsoft encouraged other vendors to develop first for OS/2 and then for Windows. Some have called this the "head fake" that destroyed the ISVs (Independent Software Vendors - Lotus, WordPerfect, et al) and allowed Microsoft and their Windows apps to take the lead and establish dominance in the field of application software where they were followers, not leaders.

What is forgotten is that Microsoft also developed for OS/2. What is not well-known is that Microsoft again sabotaged whatever they shipped for OS/2. So running Word or Excel on OS/2 was a miserable experience, especially compared to running Word or Excel on Windows. I know - I used and tested all apps for Windows and OS/2 available at that time rather extensively. There's no question in my mind that Microsoft's OS/2-app crappiness was deliberate on their part. It was as if someone tried to turn a Tesla into a Prius by developing microcode that would run on either car.

And not all of the evidence of Microsoft's lack of scruples came out during their trial for violation of anti-trust laws. There was little or nothing, for example, on their online character assassination campaigns or manipulation of the media. As the target of one of their campaigns, I can tell you that they didn't play nice. They once concocted a scheme to cancel my cable service as if I were moving to Redmond. Strangely, they were recruiting me at the same time they were trashing my reputation online and it wasn't a case of the right hand not knowing what the left hand was doing. I got a phone call a week after the cancellation asking if I had gotten the message. What message? “We know where you live. Oh, by the way, how many children did you say you had, Dave?”
Get the picture?

Worst of all, IBM offered no support. In fact, they asked me to document everything I had ever said about Microsoft and Bill Gates online. I gave up and resigned in the fall of 1995 after being told to stop writing my book, while hearing that IBM was defunding OS/2 but would continue to publicly declare ongoing support for it, and putting two and two together. I had dinner with Steve Ballmer at an InfoWorld party, and I thought it interesting how interested he was in who might be remaining at IBM carrying the torch for the transformational online marketing ideas I had become known for. Yeah, I probably walked away from a fortune in declining to join Microsoft, but I couldn’t find the necessary levels of cynicism to join a company that was dragging my good name through the mud while telling me it was “nothing personal. Just business.”

In summary, IBM was foolish to continue to play nice with Microsoft and take the high road and treat them as a Business Partner under the Business Conduct Guidelines even after it was clear that Microsoft was out to destroy OS/2 and IBM's leadership in the PC industry.

LESSON TWO. 1) Strategic brilliance in exploiting the resources you have to deploy against the resources your competitor has, 2) smart marketing execution, and 3) cunning media relations ALL trump engineering genius. Microsoft had the former. IBM had the latter.

LESSON THREE. If you want to establish a desktop OS as the standard during the coming of age of the Internet, you had better understand how to get the media on your side. They established perceptions, and perception became reality.

Microsoft played many in the media like Itzhak Perlman plays a Stradivarius. And the media created many of the myths - inconsistent with reality - that persist to this day and can be seen in many of the answers to this question.

Microsoft succeeded in perpetuating myths like:

-  "OS/2 was clumsy and IBM's programmers were incompetent while Microsoft's programmers were geniuses." Totally backwards. OS/2 was a billion-dollar miracle of software engineering. IBM created rock-solid, reliable, flexible, elegant, mission-critical operating systems that businesses relied on. OS/2 was in most ATMs for well over a decade. Can you imagine using Windows 3.1 or Windows 95/98 in an ATM? Hahaha. IBM was the company that was #1 in Forbes for attracting the best and the brightest - especially engineers and scientists including Nobel Prize winners - back then. Not Microsoft. Case in point: the web (HTML) is modeled after IBM technology, not Microsoft's. Yet Microsoft would talk to the media, and IBM was pretty insular. So Microsoft's twisted version of reality won the day.

- "IBM is proprietary but Microsoft is open." How the media bought this lie is a mystery. Truth: Both were / are proprietary. When it came to enterprise marketing, Microsoft copied a lot from IBM - just years later. Only in hindsight is it obvious that the difference (in the consumer tech space at least) is that Microsoft produced unreliable proprietary crap marketed well, and IBM produced reliable proprietary quality marketed poorly.

- "IBM doesn't care about consumers or the little guy." Again, totally backwards. IBM supported its products. Microsoft did not and still does not. IBM did the same thing then that Apple does now - create a solid infrastructure of well-supported quality and insist that others play by their well-designed rules. They then supported their products in order to constantly improve them.

- On the other hand, Microsoft created an opportunistic wild, wild west of anything-goes shoddiness that they crammed down the market's throat using their ruthless disregard for anything but their own best interests. Microsoft Explorer anyone? We've just grown accustomed to believing that software companies don't need to support their products because that's what Microsoft pioneered.

- "IBM couldn't market its way out of a paper bag." IBM didn't get to be the biggest company in tech throughout the '60s, '70s, and '80s without knowing how to market. Their marketing prowess in B2B was rightfully legendary.

Where they failed was in media relations and in countering Microsoft's ruthless perfidy in establishing a monopoly for Windows, largely because they were under the constraints of defending themselves against the anti-trust lawsuit brought by the U.S. Justice Department in 1969 that lasted for 13 years and had a major impact on IBM culture and policy.

The IBM culture, as a result of that lawsuit, was shaped by IBM’s “Business Conduct Guidelines” that were heavy on being seen as ethical and fair and light on being as competitive as possible within ethical constraints. There was even a provision forbidding “disparaging competitors,” that was often used against IBMers who spoke up about unethical competitors or business partners. There were provisions that forbade any IBMer who wasn’t at the level of Director or above from speaking to the press at all.

So these factors had even more to do with IBM's failure to establish OS/2 than its marketing decline of the '90s and its poor showing in transitioning from B2B to B2C marketing for OS/2. In other words, IBM's marketing would have been good enough if Microsoft had played by the same ethical rules - honoring the law and their agreements and playing fair - as most other companies of that day.

LESSON FOUR. That same genius that worked to get you established and make you successful in the first place is inadequate to defend your position at the top. IBM was the Google of the '60s and '70s, but by the '90s was often rightfully compared to an elephant trying to dance. Proprietary (but elegantly designed) systems lost to cheaper, better marketed systems. What made Microsoft and Windows successful is now working against them as Windows becomes increasingly irrelevant to a new generation growing up with awesome computing and networking power they can hold in their hands.

LESSON FIVE. When creating an infrastructure to support an OS or platform, treat application developers as if they were kingmakers. Because they are. IBM saw developers (ISVs) as both business partners AND customers. Microsoft saw them for what they were - critical partners.
Today, App Stores have assumed the role that ISV relations used to have, which renders this point a bit less important, but the point of using an app store to empower and support developers for your platform is still of critical importance if you want to have any chance of your platform’s success.

LESSON SIX: Never under-estimate the willingness of the market to support an underdog and adopt cheap but easy technology. Windows was the cheaper and easier path for most people. Never mind that it was the low-quality path. A market is just like water - always flowing downhill following the easiest path it can find.

There is a lot of truth (and some myth) in the other responses as well. In the end, the history books are written by the conquerors and not the vanquished, so I appreciate your question and the opportunity it provided me to share the perspective of one of the vanquished.

Dave Whittle

Neil Waldhauer:
I love this article. I want to remind readers about Lou Gertzner's book "Who Says Elephant Can't Dance" for another view of IBM's OS/2 project. In this book, Gertzner talks about the decision to end the OS/2 project.

Gertzner turned IBM around by making major changes. The change he is most proud of is the cancelling of OS/2. The decision seems to have been made before Warp 3 was released. This thread continues with remarks about marketing OS/2, but if OS/2 was already cancelled, it's understandable that the marketing to end users would not have a high priority.

Dave Yeo:
It's all complex. It's a good article but misses some points.
Bad timing was one, expensive memory screwed OS/2 in some ways. My brother, a gamer, got sold on the Warp ads about how wonderful it was for gaming, Bought Warp v3 and installed it with all options on his 486DLC with 4MB's of ram. I remember trying to test it, all it seemed to do was swap, not a good experience.
Unluckily 4MB's of ram was standard and another 4MB's was expensive, if your motherboard had spare slots for it, or really expensive if you had to replace what memory you had.
My brother gave me his stack of Warp V3 floppies and I installed it, first as a simple DOS replacement, still using BBS's under DOS and Trumpet Winsock to dial in to use Netscape on Windows, it was a redbox install. And slowly moved to native applications, loved WebExplorer. But with 4MB's of ram, I had to tune it and it was a long time before I used the WPS, after I upped my ram to 8MB.

I remember very well the battle between OS/2 Warp and Win95 (and subsequent).
Win95 was very bad in terms of technology, but it wins for several reasons:

OS/2 did not offer:
- A good office set (Word, Exell, Scanning Soft), (IBM proposal was very poor)
- Good games (new ones released only for Win)
- Proper hardware support (drivers for sound and video cards)

This story showed us that it is not enough to make an excellent OS. For its success, you still need to release a huge number of applications at least to begin with.

Roderick Klein:
There is indeed a very very long list of reasons.

OS/4 mentions not a decent Office Package. Warp 3 came with IBM works, it was a hell of lot more then Wordpad. I mean at the OS front OS/2 was hit by Microsoft. But look wat happened to Lotus 1-2-3-> Excel. Wordperfect which was well the defacto DOS word processor. Word and Excel pushed these guys out of the market as well.

IBM did pay vendors to publish games for OS/2. Such as Maxis that published Simcity 2000 for OS/2 and some other game vendors.  But indeed this was to little at to late.

With people getting Windows 95 at home according to one analysis I read a few years ago. People at large companies started to complain why they should use OS/2 at work if they had Windows at home...

As for the driver support that is true. What I think what has also not helped is that CD drives had different standards. Not all CD ROM's where supported by OS/2. With Windows 95 you could load a DOS device driver. Making it easier....



[0] Message Index

[#] Next page

Go to full version