If OS/2 Had Lived

  • Smaller Small Medium Big Bigger
  • Default Helvetica Segoe Georgia Times
Why was OS/2 a better operating system than Windows 3.0? Why did it fail to win its fair share of the marketplace?

Who cares anymore? Well, maybe the old fogies, stalwart IBM fans, or the enemies of Bill Gates! Especially since, last week, IBM reaffirmed the formal dates when it will drop support of OS/2.

Though nearly a decade has passed since IBM released new software for OS/2, IBM has continued to maintain support and supply fixes for known bugs. Good old IBM!

Yet, according to recent announcements, IBM will continue to deliver fixes only until December 31, 2005. Then, in December of 2006, IBM will remove fix packs from its Web site, stop providing defect support, and close the cover on OS/2.

Entering the IBM Technological Graveyard

At that moment, OS/2 will enter the netherworld of the IBM Technological Graveyard (ITG) in which the bones of old operating systems and computer hardware lie bleaching in an iridescent glow of flickering computer room light, somewhat like the mythological elephant burial grounds.

Probably the best description of the ITG was written by Richard Mirth in the historical novel The Silverlake Tapes.

Imagine a place where all these various architectures come together--a place haunted by the ghosts of aborted jobs and the tailings of library members left too long in the QTEMP libraries, where the flapping and screeching of unidentified indicators leap from storage dumps and the howling of phantom user error messages echoes plaintively because they can never be reproduced. This is a place where our feeble understanding of software and machinery skitters on thin ice. This universe is vast and confusing. For not only are the "great" IBM machine models and architectures hidden here--the System 360s, the System/3s, and so forth--but here as well are every machine model prototyped and every line of code debugged by the industry giant. From the greatest water-cooled mainframe to the PCjr Chiclets keyboard, if IBM's geniuses have conceived it, then it exists here. Not as rotting junk, but as hot, running, functioning machines and software, still vibrant, still alive, still dreaming of success. And this is that confusing place where we've been propelled: the IBM Technological Graveyard, or ITG.

Mirth's humorous description of the ITG reminds us all that the importance of technology is always temporal: Yesterday's great innovation is tomorrow's obsolete piece of junk, except of course in the ITG.

Nevertheless, the demise of OS/2 causes one to consider what might have happened had it lived longer and thrived. Would our networks be more secure? Would the rate of software obsolescence be greater or smaller? Would we have such a successful e-business infrastructure?

The Requirements for e-Biz

Certain things are required for the success of an e-commerce or e-business infrastructure, including, but not limited to, the following :

The standardization around some form of user interface or platform

  • A generally accepted, redundant networking architecture
  • High-speed telecommunications
  • An acceptable security mechanism
  • A mechanism for connecting to back-end computing systems

Of course, today we think we have most of these things in the form of browsers, the Internet, broadband routers, firewalls, proxy servers, and WebSphere or .NET applications. But back in the early 1990s, the Internet as a medium for e-business communications was an idea whose time had not yet come. Customers that needed to connect their information infrastructures to other companies--to begin the processes of establishing the new technology called Electronic Data Interchange (EDI)--were forced to build alliances with the largest equipment/service providers simply to establish a basis of internal standards.

Meanwhile, the race in computer technology had become a race to develop a multitasking PC operating system that could take advantage of the inexpensive Intel microchips. The larger concept of e-commerce--or "e-bidnez," coined by IBM's Lou Gerstner years later--seemed like a pipe dream at the end of a very, very long length of pipe. In other words, companies were in the very first phases of e-commerce development--what IBM now identifies as "Phase One of Deployment" in its On Demand schematics.

Standards, Standards, Standards: Just Pick One!

Just as importantly, the late '80s world of technology was rapidly filling with hardware and software products that could barely communicate with themselves, much less with one another. The problem was not that there were no standards but that there were too many.

It was into this realm that IBM tried to introduce OS/2 in 1987. Microsoft was supposed to be the coauthor, but some IBM engineers later complained that Microsoft only provided its "dumbest" personnel for the task. Microsoft too wanted a multitasking operating system, these engineers speculated. What could they learn from IBM?

In "A Short History of OS/2," David Both (president and founder of Millennium Technology, Inc., who at the time was involved with OS/2 at IBM) says: "By late 1990, Microsoft had intensified its disagreements with IBM to the point where IBM decided that it would have to take some overt action to ensure that OS/2 development continued at a reasonable pace. IBM, therefore, took over complete development responsibility for OS/2 1.x, even though it was in its dying days, and OS/2 2.00. Microsoft would continue development on Windows and OS/2 3.00. Shortly after this split, Microsoft renamed OS/2 V3 to Windows NT."

Nevertheless, for IBM, personal computing was still just a growing niche market that was somewhat separated from its mainframe business. IBM had severely underestimated the size of the PC market (originally expecting to sell half a million, tops), and to IBM, PCs were still just appliances. And though they were proving to be powerful motivators for personnel and corporations, they were not really "network appliances." And networking the world to IBM mainframes was what IBM's grand business plan was all about. If IBM could build an infrastructure of networked machines, it could lead the industry in business-to-business commerce.

But IBM had so many products--most of which could not communicate--that it decided it would need an overweening architecture of standards to pull them all together. That architecture had a name: Systems Application Architecture, or SAA.

SAA, as we all know, was abandoned as an IBM initiative after the company suffered its worst economic setbacks in history. Microsoft went on to make a harvest from the work with IBM and others and released Windows NT as a new product line.

After the Divorce

IBM went on to release OS/2 Warp and Warp Server.

Warp Server, released in early 1996, was a landmark product that combined the power and functionality of Warp 3 with the network server capabilities of IBM's LAN Server 4.0 product. Warp Server included many features that would cost extra with other server operating systems. In fact, OS/2 Warp Server delivered one of the first integrated platforms for the emerging application server environment as well as a complete set of traditional file and print services.

Warp 4 was released in September of 1996. It had a significant facelift for the OS/2 Workplace Shell, and there were many new features, including, most importantly, the availability to use Java. IBM called Warp 4 the "Universal Client" because of its unparalleled network connectivity. Customers could connect to anything, including LAN Server, Warp Server, Windows NT Server, Novell NetWare, NetWare Directory Services, PCLAN Program, IPX-SPX, LANtastic for DOS or OS/2, Warp Connect, Windows NT Workstation, Windows 95, Windows for Workgroups, TCP/IP (including DHCP, DDNS, FTP, TFTP, Telnet, SLIP, PPP, SMTP, and SNMP), SNA, and NetBIOS. And because Java was built into it, there was no need for additional software to run Java applications locally or right from the World Wide Web.

E-Business Without OS/2

The point is that, step by step, IBM had gone on to invest in OS/2 to deliver the services and the security elements that were required for the development of e-commerce, despite the lack of economic success of OS/2. OS/2 stayed way ahead of its time, even though it could not gain the popularity at the desktop that Microsoft had won. For this, IBM deserves plaudits for advancing technology that many would ultimately use but few would buy.

Today, the loss of OS/2 from the IT infrastructure will not greatly diminish the growth or vitality of the industrywide e-business initiative. Instead, it merely represents one more element of IBM's final exit from the PC arena. Still, as the vision of e-business continues to grow, much of it will be based upon the legacy that branched off of the original IBM/Microsoft development process for OS/2. Microsoft may have won the PC operating systems wars, but OS/2 won our hearts.

And in the minds of many technical experts, had OS/2 lived a longer, more prosperous life, many of the security issues currently plaguing Microsoft e-commerce implementations would have been more quickly and quietly resolved by IBM. They base this assertion on IBM's experience with building secure architectures--something that Microsoft has had to learn the hard way.

There was a rumor that IBM might release the code of OS/2 to the open source community. Unfortunately, because of the IBM/Microsoft history of development, IBM quickly laid that speculation to rest.

So it will be that on December 31, 2005, OS/2 will start its year-long journey to the IBM Technological Graveyard. No doubt it will find existence there much more sustaining.

Thomas M. Stockwell is Editor in Chief of MC Press, LP.