Operating Systems

IBM mainframes still matter for much the same reason Willie Sutton robbed banks—because “that’s where the money is.” That may be an anachronism, however. If Willie Sutton were alive today, would he still be robbing banks?

Here’s another more pertinent question: How has IBM—after all, a typewriter company from 1933 to 1991—fared in staying relevant in today’s information-obsessed world? Calculus is a branch of mathematics ideal for analyzing this question. It’s concerned with both summation of small factors to determine the whole (integral) and the measure of the instantaneous rate of change (differential). So, let’s examine the calculus of z/OS.

Integral Analysis

In the timeline of data processing, the IBM mainframe represents an unlimited series; they’ve been there since the start. 

Incremental improvements over the wide variety of mainframe technologies throughout the years have served IBM well. The reliability, scalability and security of the mainframe are ideal for today’s high-volume transactions driven by the Internet and mobile devices. In addition, IBM provides a wide variety of access points to the mainframe (i.e., stored procedures, DDF, MQ, CICS Transaction Gateway, etc.). The plumbing is in place.

This leads to an obvious question: If the IBM mainframe is truly so well-positioned, where’s the customer validation? IBM has often pointed to the growth of Millions of Service Units (MSUs) year-over-year as a health indicator. But in many cases that doesn’t accurately reflect industry contraction; it’s more a measure of the big getting bigger. Instead, where are the new mainframe organizations, the new mainframe entrepreneurs, the new mainframe start-ups? The new zBC12 machine addresses some of these issues, but is the bar of entry still set too high?

Differential Analysis

To IBM’s credit, they’ve been diligent about evolving their own technologies. While CICS, IMS and DB2 are generations old, they’ve continued to grow; one coming back to the technology after years away might even find much of it unrecognizable.

IBM has been much less successful when they chase technologies outside their realm. Java is slowly settling into a legacy language but z/OS continues to do Java poorly (poorly enough that IBM doesn’t even charge you for the processing). Again, the new COBOL takes steps but doesn’t resolve the problem. Linux on System z is an attractive (and very green) alternative, but has it reached any level of critical mass? And will their contortion-worthy implementation of Hadoop resolve its basic philosophical differences?

Historically, the IBM mainframe has been incredibly insular with unique programming languages and development environments and a cloistered programming society. While this has served IBM well, it has led to a troubling phenomenon—the aging of the programming talent pool to the point of putting businesses at risk. In this regard, IBM might have the wind in their favor. The current generation of programmers isn’t nearly as wedded to languages and platforms as their predecessors.

Does this all matter? Yes. IBM still exerts a significant gravitational pull in the universe of enterprise information systems. And while they continue to move away from being a hardware company, the mainframe remains a notable exception. But while IBM is big enough to plow its own path in some areas, it’s no longer so big that it can ignore the rest of the industry. Is the IBM mainframe at risk? Of course. Everything is. Just look up Lehman Brothers or Pontiac. 

IBM needs to continue to lead when possible, follow when appropriate and consider moving farther out of their comfort zone (into areas such as open source and massive open online courses, for instance). They need a strategy to better attract today’s software entrepreneurs. In doing so, they will be in the best possible position to weather the one thing that can be guaranteed moving forward: change.