360 Henge: How the Mainframe Shapes Our Future
5/1/2019 8:39:49 AM |
By Reg Harbeck
Every solstice and equinox, people flock together two hours’ drive west and slightly south of London, England, to observe the alignment of the shadows in one of the earliest known round complexes of monoliths, considered by some to be an agrarian season computer.
Whether or not it predates other hand-held calculating devices such as the abacus, Stonehenge has certainly withstood the test of time, and it continues to cast a long historical shadow, even in the summer. And while smaller calculating and time measurement devices may have fallen prey to modern electronics, we can continue to study and learn from the steady, predator-resistant presents of this lithical complex to human history.
This is a monumental reminder that, in life and business, there’s more than one kind of complexity that we deal with. Sometimes, like Stonehenge, it’s a deliberately planned and designed response to thoroughly investigated requirements; but sometimes it’s more like a growing ball of detritus rolling down the street and picking up everything in its path. That’s been true throughout history, and it can be just as true of IT.
Likewise throughout history, people have often taken an opportunity to step back from the daily grind and notice what works well versus what merely exhausts entropic potential (or exergy). The result of this attention is sometimes referred to as wisdom, and it is the values of such wisdom that are essential to the building of lasting value.
Which brings us to a common statement coming from the mouths of those who have paid attention and are noticing when someone is reinventing the wheel: “Been there, done that.” What is perhaps more poignant about such observations is that the reinvention often fails full circumspection of prior art or wisdom, and is often just intended to operate like a single facet or dimension compared to more deliberately complex existing systems.
Such commodity knock-offs are often knock-on consequences of complaints about the ponderousness of existing systems and the requirements to remain compliant with them. So we see the ubiquity of swelling consumer bloatware that seems to cancel all the implicit advantages that Moore’s Law is supposed to be bringing, as people build an inverted pyramid up from a point of functionality.
At such times, it’s good to be able to take a breath and look for systems of breadth that truly rock, no matter how many stones are thrown in their direction. Which brings us full circle to the system where the compass rose to its true 360-degree potential.
Leading up to the April 7, 1964 announcement of IBM’s System/360, many lessons were learned, much research was done, and a great deal of strategy was planned at vast expense to create the ultimate business computer. There was roughly two decades’ prior art of developing electronic computing, preceded by five-plus decades of electrical and electronic calculation and tabulation, preceded by millennia of mechanical or physical devices using everything from shadows and light to gears, water, and springs.
Foundational Mainframe Layers
Today, approaching three generations that the IBM System/360-descended mainframe has been around, with the advent of the z14 and its pervasive encryption, we are perhaps more crypto-mainframers than ever before, as we almost invisibly run the world economy. Most people—even in IT—stumbling upon a running z14 would be as stumped about its function and relevance as they probably are about Stonehenge. It’s time for us to help their level of awareness rise from their long daze about IT that actually works as the touchstone of business computing.
So, let’s take a step back from the current knee in the ongoing IT adoption curvy street and look at what has shown itself to work well enough that we don’t have to keep dismantling and rebuilding from the ground up.
While innovation seems to be an inevitable part of the ongoing story of IT, much of the best of IT has come to a point of reliability where the innovation can build on top of it rather than changing foundational layers. One of the most important moments in IT history was the announcement of the System/360 with the promise that anything written for it would run on all sizes of the machine, and all future versions of it—a promise that IBM has substantially kept, even while staying far ahead of all other platforms with new innovations that enhanced the business functionality of the platform without forcing massive rewrites of the existing applications.
It’s easy to forget that, until System/360, every time a new platform came out, all the applications had to be rewritten to work on the new architecture. That was one of the draws of COBOL (celebrating its 60th birthday this year). You could just recompile a program, rather than having to rewrite it in architecture-specific assembler.
Not that other platforms have made or kept such a promise—good luck running MS-DOS or Apple II applications on anything but an emulator these days. But then, they’re not IT bedrock to the world economy, either.
An Unparalleled Platform
Over the first decades, the IBM mainframe (and initially some of its early competitors, now mostly faded into the landscape of history or Intel-based emulators) took us from punched cards and tractor feeds to virtual terminals. Many of the foundational concepts were set firmly into place, creating a secure, virtualized, backed up, performance-managed platform that had no parallel. As long as one platform showed itself to be indispensably reliable for the key functions of the world economy, other platforms could play the games of commodity and risk and consumer pricing.
Fortunately, other platforms have often taken the opportunity to learn and emulate some aspects of the IBM mainframe, even if not to the same level of quality and exactitude for reasons of cost and capacity. But just as often, they’ve let the mainframe continue to be the backend system of record for the most important, sensitive, reliable and available data and processing—which is fine, as long as this fact isn’t forgotten.
Today we find ourselves back in yesterday’s future, 55 years after the System/360 was announced, with the world economy more heavily reliant on it than ever. With hundreds of billions of lines of COBOL running the world economy on this platform that simply works, there is no logical or affordable prospect of moving to another platform, even if one existed that had the same qualities of service. And so we can confidently look hundreds or even thousands of years into the future and see that, just like the wheel, lever and hammer, which have become irreplaceable parts of how the world works, so the IBM mainframe may someday be recognized as the henge on which the world economy hinges. And that’s a foundation worth coming around to.