Less than a century ago, when the Sears catalog was still the dominant symbol of American consumerism and planned obsolescence was not yet a product feature, certain possessions were treated with reverence. These items either ensured a family's survival or were not easily replaced by virtue of cost or scarcity. A carpenter's tools, a sewing machine, a hunting rifle, fine linens, a musical instrument; these items were passed on to succeeding generations, respected, sustaining, giving service in support of life.

The context has changed, but in many respects that tradition has resurfaced in the S/36. The thing simply refuses to take its rightful place on the scrap-heap of yesterday's technology. It is handed down from business to business, unpretentiously doing what it was designed to do-giving service. The S/36 is just too useful to be retired.

IBM, of course, holds a slightly less romantic view. It has been trying to drive a stake through the heart of the S/36 for years. But successive AS/400 announcements and migration incentive programs have delivered ineffective blows, and the beast survives. Over 90 percent of the original S/36 crop has yet to be plowed under.

But IBM is warming up the tractor. This time, it's an AS/400 that is object- code compatible with the S/36. It runs SSP and all S/36 applications. The system will serve as poster child for the new 64-bit Power PC chip that will stoke IBM's 1995 issue of AS/400s and promises a two to five-fold performance improvement.

Alas, what the new AS/400 delivers in performance, it temporarily lacks in range. The new AS/400 Advanced System/36 will not be able to run OS/400 until customers acquire an upgrade in 1995. Until then, the limitations of SSP will define the margins of the system.

Even after upgrading, both operating systems will remain segregated so that S/36 applications will not benefit from OS/400 features.

This new offering faces the same problem confronted by past efforts to pry 200,000 technological tree-huggers from their favorite technology. IBM is attempting to market the system to an audience that doesn't need a new computer. Perhaps more accurately, S/36 customers appear supremely happy with their systems. Like a man still driving a classic '65 Mustang, it is hard to sell him a new one just because the new ones go faster.

The profile of current S/36 customers is dominated by low-tech, low-budget shops supporting static businesses with fixed software requirements. Leading edge technology or, for that matter any computing trend to emerge after the dawn of the S/36, has little practical significance for these businesses. Relational databases, networking, communications, client/server, and object- oriented programming, while desirable, dance outside the practical limits set by the penurious economics of a S/36 ownership.

John Capaccio works for Lumber Jack Enterprises which provides student housing and food services at Humbolt State University in California. His S/36 has been running for over 13 years. It is still not at capacity, it is not objectionably slow, and it has proven to be nearly bomb-proof with only one hardware failure in 4,745 days of operation. John can not be described as wildly motivated to scrap his S/36.

Christine Mulligan manages a S/36 that supports the six administrators of the San Mateo County Medical Association. The county butts up against San Francisco Bay on the east side, the Pacific Ocean on the west, and two counties on either end. It is not growing. Consequently, the number of physicians in the county remains relatively constant, as do the association's computer support needs. Mulligan acquired the S/36 in 1991. Prior to that, the office limped along with word processors. "Now," she quips "we have DisplayWrite, the word processing software from hell." Mulligan looked at migrating to an AS/400 but found it too costly for her organization. If the price were as low as an elephant's toe, she would be interested in the new offering, otherwise she plans on enduring DisplayWrite with monkish stoicism.

Levina Ranch is growing, but not in size. It grows fruit: grapes, apples, prunes, and cherries. Ric Potter recently bought an AS/400 F model, but is still running the S/36 that has been humming along for eleven years with only one three-day failure. Potter migrated primarily because, like a man who has just won eleven consecutive spins of the roulette wheel, he figured his luck was about to change. "There are times that are critical for us to be operational," said Potter. "I don't feel safe on the S/36 anymore." Why? Ironically, because it's too reliable. "We've already gotten four years more use from the disk drives than they were designed to give."

Potter sees no real use for the AS/400 Advanced System/36 that runs SSP. "Our applications all run transparently in the S/36 environment on the AS/400," he says.

However, there are some compelling reasons to switch, particularly for those S/36 customers who do have capacity considerations and especially for those who no longer feel "safe" on the S/36. By transplanting the S/36's brains into an AS/400 body, IBM extends the life of the S/36 and, paradoxically, prolongs its marathon pursuit of the S/36 customer. IBM's goal, after all, is not simply to migrate S/36 applications from one physical box to another, but to persuade users to embrace OS/400 and a new generation of applications and technologies. By surrounding SSP with state-of-the-art hardware, IBM insulates customers from the need to migrate to a new operating system. Moving SSP to an AS/400 gives S/36 customers another decade of trouble-free computing on hardware that is even more fail-safe than the enduring S/36.

For the capacity conscious, the Advanced S/36 will offer an estimated 100- megabyte memory capacity and up to five gigabytes of disk space. Additionally, the near-term possibility of running both SSP and OS/400 on a single machine suggests a less painful and more leisurely migration path for those who want to take advantage of the AS/400s advanced connectivity features and programming tools.

Then, too, there is a sense of finality about this offering. IBM has been extending the olive branch to S/36 customers for over six years. Arguably, it has done a clumsy job of it, first it beat S/36 customers over the head with the AS/400, then it ignored them when they refused to respond. But this new system is likely to be the best deal S/36 customers are going to get. The enormous flexibility of the AS/400's system architecture permits a native SSP implementation. It is to that architecture that IBM is committed, and it will not spend additional development dollars to reinvent "Son-of-S/36." Technology is a train that does not wait for the undecided. It's best to catch it before the S/36 shrugs off its manufactured coil.

Some years ago, I heard about a light bulb that apparently refused to recognize its own planned obsolescence. The annoying thing had been burning continuously since the early 1920s. Located in a firehouse somewhere on the East Coast, it had become a local curiosity. Doubtless, it no longer generated a great deal of light, and the fixture was aged and tarnished. I'll bet every couple of years someone looked at that old lightbulb and thought it would be a good idea to get a newer, brighter, better bulb. But the thing just kept on shining; steady, dependable, persistent. After a while, like the S/36, no one could find a genuinely good reason to replace it.

Victor Rozek has 17 years of experience in the data processing industry, including seven years with IBM in Operations Management and Systems

Also Read

IBM i Shops Pivot to Battle COVID-19

Energy, Vibe, and Atmosphere

Murphy’s Security Law