IBM, Forests, and Trees

Sponsored by


IBM may someday make a decision that a new OS/2 client version is not necessary. What should be our reaction, if that were to happen?

When we say, "IBM makes a decision," who do we mean by "IBM?" We mean upper management, "the suits," the men and women who make more money in a year than I have made in my entire lifetime. They have the power to make decisions about whole product lines, market strategies, and customer bases -- what to keep and what to throw away. They have a perspective based on interests and priorities that is quite different from yours and mine. Just what is that perspective, and how does it affect their decisionmaking? What are IBM's "strategic directions" that affect their viewpoint on OS/2 and OS/2 desktop users?

First of all, let's examine an old adage, "sometimes you can't see the forest for the trees." In other words, sometimes the details of a particular situation are so cumbersome and our focus is so narrowly on those details, that we miss the bigger picture. It is like looking so intensely for something that is right in front of you, that you fail to see it. Most of us have had that experience, that "brainlock," from time to time. It's a matter of focus and perspective. The eye sees all, but the brain sees through filters and sometimes misses important information.

The IBM'ers who make these high-level decisions have a different problem: they can't see the trees for the forest. It is like being the rider on an airplane, looking down at a large, forested area. You can see the big picture, but picking out one individual tree is difficult or maybe even impossible. The IBM brass who make high-level decisions can see their strategic direction, they can see where they want to be in terms of market control or market share, and they can see what they want to focus on in terms of profit margins, technologies, and market segments. But they can't see the perspective of the individual user, the solitary tree in the forest. From the perspective of a well-informed individual user, OS/2 is the solution to all or most of the nightmarish computer problems. From the perspective of IBM upper management, OS/2 does not solve IBM's nightmarish marketplace problems.


A computer user wants something that's easy to use, reliable, flexible, and has plenty of people to support it -- to help with adjustments, upgrades, fixes, and such. The typical user would be overjoyed to have OS/2 on their PC, provided they could have their choice of applications, utilities, and games. PLUS, they would require having friends and co-workers who could sit down and walk them through certain procedures such as installing or removing an application, finding a file, converting something from one format to another, etc. From the user's perspective, the computer ends at the screen. They don't know or care where or how their data is stored. The fact that this data is probably on a fileserver somewhere makes no difference to them. They are focused intensely on the "tree" or screen in front of them.


IBM corporate priorities are totally different. IBM's entire business can be summed up in one phrase: "managing complexity in the most cost-effective, reliable manner." Complexity is the key ingredient to IBM's current computer strategies. In the "old days" of mainframe machines, the complexity was contained within a single box called the mainframe. Data, access permissions, security, and all other resource management tools were right there in the "box". Complexity was relatively easy to manage.

Personal computers have introduced a new level of increased complexity. The basic rule of thumb is this: for N individual PCs on a network, the complexity rises as N-squared. Management of complexity becomes a nightmare when data, resource access, security, and other individualized pieces of information are scattered about among hundreds, thousands, or even millions of individual computers. This is the "forest" that IBM sees, and their entire focus is on managing the forest, not the care and feeding of each individual tree.


The solution IBM has derived to the problem of complexity management is THIN CLIENTS. This is the only configuration of system resources that allows for the distributed processing capabilities, the flexibility, and the ease of growth of the PC-based systems, while still maintaining control of the issue of complexity. Managing complexity via individual PCs is extremely labor-intensive and therefore costly. Corporations want to reduce labor expenses (by layoffs, by pension reduction, by outsourcing, by any means possible) because providing shareholder value is now more important to them than providing employee comfort zones such as long-term employment, medical care, pensions, and career-growth opportunities. Cost-cutting of the information infrastructure is far more dependent on complexity reduction than it is on raw computing power, and this is where IBM sees its opportunity.

IBM is having significant difficulties in getting this message across to corporate America. Corporate information management has been pushed so far down the line that lower-level managers make computing decisions based on "trees" instead of "forests." Microsoft has many of these big companies dancing on a string, or running on a constant treadmill of so-called "upgrades" that merely add complexity and increased long-term costs with each turn of the screw. It is far more complicated an issue than just the reliability of the individual desktop machine, just as a forest's ecological infrastructure is far more than merely a bunch of trees that are physically close.

IBM cannot succeed by offering a different flavor of the same poison, so to speak. As a result, any product that does not support the "forest management" philosophy at IBM will raise a lot of red flags. IBM sells Windows because they have to, not because they want to. There is no concurrent marketplace force that can dictate to IBM to sell full thick-client implementations of OS/2 to the public. Thick-client versions of OS/2 do not provide the complexity reduction that IBM wants and needs in order to provide the complexity management architecture that corporations desperately need but are often not wise enough to implement.

The issue here is credibility. How can IBM provide credibility for thin-client implementations if they push thick-client solutions? Thin clients already have to fight the FUD from Microsoft, including Microsoft's bogus and poorly-implemented version of "thinness" that is merely designed to give a bad name to thin clients, not to solve any problems. If IBM cannot provide a clear, unified model to its customers, they cannot convince them of the essential importance of the complexity-localization issue. Selling tree seedlings is not IBM's business; selling forest management is.

Frankly, arguing that IBM would be "stupid" not to make a new OS/2 client version is missing the point. If IBM was in the business of consumer products (trees), then a new fat-client OS/2 would be extremely appropriate right now. But that is not where IBM is focusing its energies. IBM will probably never again make the mistake of trying to be "all things to everyone" and therefore losing focus. IBM's decisions are not always the best ones for individual OS/2 users, but they probably are the best ones for IBM and its target customer base of huge forests of PCs that need a far better scheme of management than the "mainstream" currently provides.


Where do we go from here? Obviously, OS/2 is still the best overall choice for the PC power user who knows how to shop carefully and can download and install their own updates and device drivers. In terms of ease-of-use, ability to run a wide spectrum of applications, and of course reliability, OS/2 provides the best combination of these characteristics. If you run OS/2, keep using it and enjoying it. If you haven't tried OS/2 lately, consider it -- it is still a great product and will always be a cut above the mediocrity of Microsoft.

Where will the PC market be in ten years? That is hard to say. IBM's goal of complexity management allows for anything to be on the client side -- PDAs, Linux boxes, OS/2 boxes, Windoze, Webtv, whatever. It is hard to imagine that in ten years, everyone would be using the same kind of platform to do all their work that they use today. It is also hard to imagine that OS/2 will not have a place on the desktop, since it has aggressively embraced open standards such as the Internet. My recommendation is to KEEP PLANTING TREES, and LET IBM WORRY ABOUT THE FOREST. Keep using and promoting OS/2 on the desktop, thick client and all. IBM has other things to worry about, like Y2K, complexity management, and overcoming MS FUDmeisters. Keep building and growing the set of OS/2 capabilities and products. OS/2 is not "going away" from the desktop; IBM is going away from the desktop.

At the same time, keep an eye on Linux. IBM is beginning to respond to the marketplace demand for Linux, a "pull" that OS/2 supporters have not been able to generate for their favorite product. Even longtime Microsoft supporters are beginning to see an opportunity to "shear the sheep" (squeeze more money out of their clients) by jumping on the Linux bandwagon, since Microsoft has failed to deliver on its promises of reliability and complexity management. The promise of an open platform -- and therefore at last a truly level playing field -- is bringing a lot of former MS allies out of the woodwork, allies who were stabbed in the back and abandoned by Microsoft. At some point, the sheer weight of money, people, and other resources being allocated to Linux may eventually make it a more attractive platform than OS/2. But that time is not yet here. Keep using what works best for you. If you believe in the OS/2 desktop, consider participating in Operation Rock Soup, VOICE, POSSI, or attending Warpstock or other OS/2-oriented activities.

IBM makes decisions based on its own priorities. I have made a decision to keep using OS/2. Now it's your turn to make a decision. Are you a forest, or a tree?


[ Top | Home Page | Faq ]

Copyright © 1999, Tom Nadeau.
All Rights Reserved.