In June 1969, I was a few months into my first programming job and assigned to combine a handful of labor accounting programs written in COBOL (Common Business-Oriented Programming Language) for my employer, a subsidiary of a calculator company called Wang Laboratories. Wang would later become a big player in the 1980s, but at the time desktop computers were still being invented by some colleagues a few offices away. All the news of the computer industry was delivered monthly in Datamation magazine. A bit of a contrast from today's Information Technology industry rate of change.
My computer was an IBM 360-65, the top of the line mainframe ancestor of today's IBM Z-series mainframes. It had tapes, and disks, and telecommunications, and printers. The 360-65 had about 1/400th of the computing power of today's iPhone 3G, but it got the job done for the big applications then. It also had IBM's mainframe operating system and a dozen IBM software products such as the COBOL compiler. Cost over $10 million. When you bought or rented the machine, it came bundled with IBM services including IBM's systems engineers to care and feed the IBM software (and typically, the customer's software).
In 1969, there were few choices in software development. The largest companies had IBM programmers turn business processes into application programs, with the goal of IBM reselling these to recoup some of the investment. There were a few small custom programming shops, but most software was developed in-house. Very slowly and at great cost. Business programmers were uncommon, there were no tools for application modeling or even code maintenance (think punch cards), systems analysis was not systematized, and computer time was very expensive.
IBM's decision to unbundle software and services, under the real threat of an antitrust breakup of the company, created the inflection point that inexorably led to today's panoply of computer options. I commend the linked article by an IBM insider on how that decision was made. The import of IBM's decision was simple but had widespread effect: customers no longer had to buy IBM software or services to use on their IBM hardware. Within a year, John Cullinane was working in our data center on Culprit, the first product of Cullinet Corp, which went on to create one of the great mainframe databases, IDMS. Charles Wang created a company, Computer Associates, around a magnetic tape library utility. The industry flourished in the 1970s with software that ran on standard IBM mainframe hardware and operating software. Application, development software tools, and system software competition for IBM products became an industry itself. Gene Amdahl went the other way, creating plug-compatible mainframe hardware that ran IBM operating software and all the other third-party and user application software.
The IBM unbundling decision led to the common understanding of the model consisting of proprietary hardware architecture surrounded by a ecosystem of companies that provided add-on system software, applications, and services. In the 1970s, that model served to jump-start the minicomputer industry with the rapid growth of Digital Equipment, Data General, and Wang Laboratories, followed by Prime and Sun Microsystems, to name just a few.
In the early 1980s, IBM repeated the model by selecting a tiny microprocessor compiler company, Microsoft, to deliver a basic operating system for IBM's highly proprietary Personal Computer, which used Intel's microprocessor. Today, of course, Microsoft and Intel have joined IBM on the list of Dow Jones Industrials. And IBM isn't even in the PC business any more.
I could go on in more detail and reminiscences, but I've made my point: IBM created the robust global computer industry of today by unleashing the forces of competition with its 1969 unbundling decision. What IBM thought at the time was a defensive decision to protect its core mainframe profits actually created a much larger business that allowed IBM to grow larger and more profitable than it would have been as a 1960s-style monopolist.
So, Happy Birthday, independent software and services industry. I've enjoyed the ride.