Feb 1 ’05

z/Bottom Line: One for the Ages

by Editor in z/Journal

Who would have thought it would come to this? The mainframe has survived an onslaught of misery, challenge and all-out conversion war to rise to a prominent place in today’s business. The common knowledge of the ’90s that prophesized it was only a matter of time until the last of the big iron was unplugged has been replaced by a new group of devotees in awe of the mainframe’s continued abilities to deliver. Ironically, these companies are championed by the executives who didn’t listen to the “wisdom” and kept using the technology that had more than returned their investments.

Survive it did, and in a grand scale that no one could have guessed. But unless immediate action is taken on a wide scale to reverse the current course, the mainframe will die a certain and natural death.

The mainframe community focused on all the important survival strategies except one. IBM buckled down and leveraged manufacturing economies to dramatically drive down the price per MIPS. They also found ways to open the mainframe’s proprietary doors to the world of open systems. Customers found ways to make their mainframe investments continue to be the benchmark that caused all other technology competitors to grow green with envy. But all parties missed one important point in the quest for mainframe survival: It’s all about reproduction.

It’s a basic tenet of natural history. If a race doesn’t reproduce within its breed, it becomes extinct. And the mainframe race appears to have given up on the reproduction system. You can see it as you look at the monochromatic color spectrum across the mainframe landscape—it’s nearly all gray.

Gray, as in the color of the hair of all the people who are responsible for the continued growth, development, deployment, and management of the mainframe. We have an entire breed who is aging quickly; in fact, many will be departing the profession one way or another very soon. There is no formal call for cross-breeding. Something must be done. We didn’t come this far to be forced out by natural selection!

Development teams at IBM are still crafting the very guts of the operating systems and many of the important technologies such as CICS, DB2 and others, in nearly the same Assembler language that has been used for the last 40 plus years. How many schools today are offering courses in Assembler language? What about IBM “University”? Mainframe Assembler language is about to go the way of Aramaic. Any thoughts about ensuring that z/OS will still be able to be written, or even read?

Attend any of the industry conferences around the globe, including SHARE, CMG, the IBM zSeries Conference and WAVV, and you see a remarkably singular demographic. Nearly all gray-haired and nearly all men.

Take a hard look at your entire support staff, the programmers writing and modifying the very important thousands of lines of COBOL code and the systems people whose job it is to ensure the continued operation of the critical: hospital systems, international banking centers, scientific research facilities, national defense systems, air traffic control systems—you get the picture. Chances are you’ll see the same demographic in your own backyard.

I’ve had many conversations with people about this issue, and universally the response boils down to the same concept, “I’ll be long retired by the time that becomes a problem—so I’m not worried about it.”

But someone needs to be! It’s not overly dramatic to realize the critical nature of the world’s infrastructure is still, and will continue to be, reliant on mainframe technology. And 15 years ago, the all-out assault to convert the mainframe to “client/server” showed how infeasible that proved to be. So that can’t be an answer.

This is a DEFCON 1 alert; it’s time for the board of directors of the world’s corporations to map out a plan to solve this. It’s time for IBM to ensure that the inner- breeding of 35-plus-year veterans are transmogrifying the new college grads to be able to continue to write and maintain the code that is still powering the world today. CEOs need to wake up to the fact that although they may retire in the next 10 to 15 years, so will most of the mainframe workforce they depend on today—and there’s no one to fill their shoes! They owe it to their shareholders to have a more forward view than that.

CIOs must implement a plan that provides for cross- training and a development plan of new talent. This includes systems programmers, application programmers, even operations support. IBM needs to communicate the same type of skill transfer effort to its customer base to ensure the investors of billions of dollars in mainframe technology that its investment will endure. Heeding our own pontifications, at illustro, we’ve just begun training one of our 24-year-old Web developers in the fine art of mainframe Assembler language programming. These are the types of steps that must be taken.

Or, we could wait until we all stop reading this magazine. It will be someone else’s problem then.

That’s z/Bottom Line.