Prologue

Why the Open Industry Standard Matters

THE PHRASE “open industry standard” isn’t used much today, but from late 1983 until the early ’90s, it was a central theme of the personal computer (PC) industry. Then it became a permanent part of the culture, and, gradually, almost everyone forgot about it. But whether we realize it or not, many of the conveniences we take for granted today exist because of the impact of the PC open industry standard on computers and technology. If it hadn’t existed or if it had ended prematurely, we would be doing things a lot differently.

Take, for example, the automobile industry. In the beginning, cars worked differently. New buyers had to be taught how to drive a particular car—and if they changed brands, they had to learn a different way to drive. Then most automakers began to standardize the way they built cars. Steering was done by a wheel in front of the left front seat; the “gas,” or accelerator, was under the right foot; the brake pedal to the left; and the clutch pedal to the left of the brake.

With standardization, once a person learned how to drive any brand of car, he could change brands and not have to learn again. Other features became standardized: all cars began to use gasoline so pumps along the road featured the same type of fuel; spaced their tires the same width, so roads could be built; and used similar materials, so that suppliers could gear up for higher volumes and reduce costs.

Since a buyer could switch brands easily, competition became more intense and prices came down. The result was that growth in the auto industry exploded and technology advanced rapidly.

Before the PC industry standard, the computer industry was similar to the early automobile industry, but once standardization began, the effects went much deeper. Industry-standard PCs worked the same, their keyboards and screens looked the same, and their diskettes loaded and unloaded the same. PCs even used the same “engine,” which in a PC was the microprocessor and memory. There was also something new called “software,” which was necessary for PCs to be able to do useful work. All industry-standard PCs used the same software. While cars had add-on options, PCs had add-in boards and peripherals, such as modems and printers.

This meant that the companies that produced the parts, software, add-in boards, and peripherals could focus all their resources on a single version instead of five or ten different ones as they had done in earlier computer markets. In addition, when they came up with a new capability or innovation, they had to implement it only once for it to work on all industry-standard PCs.

By the late 1980s, all PC technologies began to advance more rapidly because of this increased focus and the increased competition. The focus led to more efficiency, while the competition provided the incentive and motivation. The third ingredient was capital, which was more available because of the explosive growth of the market.

Over the next two decades, processor performance and memory capacity grew more than a thousandfold. Screens went from bulky, heavy, and power-hungry monochrome cathode ray tubes to thin, light, and low-power color liquid crystal displays (LCDs). Data storage expanded from 40-million-byte disk drives to 4 terabytes—a 100 thousandfold increase.

The magnitude of these advances is hard to comprehend today. But while technology would have advanced without the industry standard, it wouldn’t have advanced at the same staggering rate.

The iPhones and iPads so ubiquitous today could not have been created in 1980 or 1990—they weren’t even possible in the first few years of the twenty-first century. Different companies had similar ideas and made repeated attempts, but the technologies to make them easy to use weren’t invented yet and the chips weren’t powerful enough. The tiny computers we hold in our hands today are the result of a very long series of advances and innovations that began with Intel’s invention of the microprocessor during the 1970s and finally reached critical mass in the mid-2000s. Without the increased rate of technological advance that came as a result of the open industry standard, we would still be waiting for the technology necessary to create the amazing devices we now take for granted.

The Beginning

The twenty-first-century Information Age came to life and took its first steps at a particular moment in time, over a unique eight-year period that still shapes the direction of technology today, especially personal technology rushing to fill every space of our lives. The ever-more mobile, ever-more powerful, ever-more personal information tools we hold in our hands trace their origins directly to the battles won and lost in the 1980s for the future of personal computing and technology.

If you walk into a computer store today and select any PC software package, you expect that program to run correctly on any brand of PC—whether it’s a PC from Hewlett-Packard, Compaq, Dell, Sony, Lenovo, Asus, or any other company. But if you had walked into a computer store in 1981, every brand of computer required a different, specially adapted software package, much like what exists today with different video game consoles.

What happened between 1981 and 1989 that fundamentally changed the PC industry? Most would say IBM entering the PC market caused the change, and that certainly was an important part of the story. IBM was the largest computer company by far at that time, and its brand was as strong as Apple’s is today. Although IBM’s PC established the open architecture around which the industry standard developed, the truth is, IBM never intended for other PCs to be able to run the same software. The computer giant didn’t want standardization; IBM’s executives liked the status quo. It had become the world’s largest and most powerful computer company playing by the old rules—and it was willing to do almost anything to keep it that way.

IBM’s executives liked the status quo. It had become the world’s largest and most powerful computer company playing by the old rules.

What about Intel and Microsoft? These companies stood to gain the most from an industry standard that would require Intel’s microprocessors and Microsoft’s operating system. But neither of them was in a position to define and develop a standard involving every part of a personal computer. The creation of a true industry standard that would grow and evolve required a PC manufacturer other than IBM that wanted a standard to exist. However, established computer companies, such as Hewlett-Packard (HP), Digital Equipment Corporation (DEC), and Texas Instruments (TI), initially introduced proprietary PCs and were also content with the old rules. When they finally did figure out where the industry was headed, they were too late getting into the game—and someone else had already taken the lead.

Compaq Steps Up

That leader turned out to be Compaq. We had the need—and will—to achieve total compatibility. From the beginning, the company’s founders were absolutely certain that we had to be able to run IBM PC software straight out of the shrink-wrapped box; otherwise, we would have no software at all for our portable PC. And we soon figured out that we would need the whole industry with us if we were ever going to get out from under IBM’s shadow.

Making our portable run all IBM PC software meant we had to find and fix every single incompatibility—one at a time. It was very tedious and time consuming, which perhaps explains why no other company ever achieved the same degree of compatibility with IBM as Compaq. We didn’t realize it at the time, but we were developing an incredibly important and valuable technological capability. We learned how to design PCs to include key innovations while maintaining complete backward compatibility; that is, the ability to run all existing software and use all existing add-in boards and peripherals.

Compaq quickly gained the reputation for being the most IBM-compatible of all the PC brands and eventually for even being more backward compatible than IBM. Then in late 1983, in a transaction known only to a few people, Compaq licensed our own version of Microsoft’s Disk Operating System (MS-DOS) back to Microsoft for it to sell to our competitors. In doing so, Compaq set the industry on a course toward more consistent compatibility with an industry standard that gradually became less dependent on IBM.

As the movement toward standardized software and add-in boards gained momentum and acceptance during the mid-’80s, IBM decided to abandon the architecture that had started it all and changed to one that was much more tightly controlled and protected. It didn’t want competitors to sell personal computers that could run IBM PC software without buying a license that carried significant royalties. At the time, that seemed like a smart move, one that industry observers had anticipated for a long time. Given IBM’s dominant leadership position in the PC market, its strategy should have succeeded. But what IBM didn’t count on was not all of its competitors were willing to surrender without a fight.

Anyone looking at Compaq during our first few years could not have guessed that we would emerge to lead the PC industry in its revolt against IBM’s move toward proprietary control. We were a tiny start-up that built only portable computers—hardly a threat to giant IBM. Even when we did move into desktops, our market share was almost immeasurable at first.

But in a sequence of decisions and events that essentially amounted to a “perfect storm,” Compaq quickly moved up to the third position behind IBM and Apple and developed a reputation as a technology leader and innovator, just in time to resist IBM’s bold move. IBM surely believed that no company—not even Compaq—could block its success. What IBM never expected—and what was arguably Compaq’s most important decision ever—was that we would recruit and lead all the other PC companies to join us in the battle. If it was IBM versus Compaq, IBM would almost certainly win, but if it was IBM versus the entire PC industry, that would be a real battle.

How was Compaq able to move so quickly, from a start-up wannabe with a four-page business plan to an industry leader capable of defeating IBM?

In the end, IBM didn’t succeed in taking proprietary control of the PC industry. The perfect storm that had prepared Compaq to lead the clone army continued to blow in our favor as the press gave very positive coverage to the “Gang of Nine”¹ announcement about our industry coalition. The successful introduction of the Extended Industry Standard Architecture (EISA) bus was followed a year later by products that proved EISA’s performance superiority over IBM’s Micro Channel. The industry coalition completely debunked IBM’s claim that it had no choice but to sacrifice compatibility in order to achieve increased system performance.

As a result, there was a sense that IBM had tried to scam the industry. Its reputation was damaged not only in PCs, but in the rest of the company’s business as well. A few years later, IBM announced that it was discontinuing production of the PS/2, and some years after that, IBM left the PC business completely.

How was Compaq able to move so quickly from a start-up wannabe with a four-page business plan to an industry leader capable of defeating IBM? The complete story includes: extraordinary people, a unique company culture, bold and often unexpected strategy, a powerful decision-making process, excellent execution, and quite a bit of good luck. Intermingled throughout our story were a handful of extraordinary decisions, each one a turning point that resulted in Compaq taking advantage of both our competitors’ mistakes and the opportunities then emerging during the formation of this new personal computer industry.

To really understand why the open industry standard developed, it is helpful to go back before IBM introduced its PC and follow the sequence of events that developed. It actually happened in stages over a period of years.

Every Computer Required Different Software

Before IBM entered the PC business in 1981, every PC required a different version of every software program. A Radio Shack store would carry a proprietary version of VisiCalc, the first spreadsheet program, which ran only on the TRS-80. A ComputerLand store would carry a version of VisiCalc that would run only on the Apple II, and, in addition, a different version for the Commodore 64. When VisiCalc became the most popular PC program, any PC without its own version of VisiCalc was unlikely to find shelf space in a computer store and was, therefore, likely to fail.

IBM quickly became the market leader with the introduction of its Personal Computer.

VisiCorp, the company that developed and sold VisiCalc, had to adapt its program to each new PC. Limited software development resources meant that not all PC brands would have VisiCalc at the same time; one or two brands would have it first, and then one or two more would follow, and so on over a period of months. That situation existed for every software application company and every PC company. It was a vicious cycle: The PC brands that got the most customers tended to get the popular programs first, and the PC brands that got the popular programs first tended to have the most customers. A fierce battle developed between PC companies to get software companies to adapt popular programs to their PCs first, or at least put them early in the pipeline.

This became a very big problem for start-ups. A major computer brand, such as HP or DEC, could get a good place in line for one of its new PCs, but a new and unknown computer company like Compaq was completely left out in the cold. The situation seriously threatened to cut off the flow of new, innovative PCs from start-up companies. Then along came IBM, the largest computer company in the world. At first IBM’s entrance to the PC market made the problem worse, because almost all software and peripheral companies immediately moved IBM to the top of every development priority list. But before long, it became apparent that IBM had enabled an unexpected—and unintended—solution to the problem.

When IBM first introduced its PC, the strength of the IBM brand attracted almost every software and peripheral company. Its PC sales took off to a great degree because all the popular software and peripherals quickly became available for it. It soon became the market leader, and the software and peripheral companies continued to first introduce their new products to run on the IBM PC. Although there were still a dozen other PC brands to which they would adapt their products over the following months, start-up PC companies would never get them.

Then Compaq and a few other companies discovered the solution IBM unintentionally helped create—if software companies wouldn’t adapt their programs to a new company’s PCs, then new companies would adapt their PCs to run the popular software that already existed. In IBM’s rush to get its PC to market, it had not taken the time to incorporate any major barriers to cloning, which is when a company makes its PC run the software written for another company’s PC. Compaq and other companies set out to create PCs that would run IBM PC software; from that, the IBM PC-compatible market was born. We had effectively eliminated, or at least lowered, the barriers to entry for new and existing computer companies. At first, some established computer companies were skeptical, but as Compaq’s sales grew at a staggering rate in 1983, it wasn’t long before every company but Apple was jumping on the PC-compatible bandwagon.

Just five years after IBM introduced its PC, IBM and a few dozen PC compatibles accounted for more than 75 percent of the U.S. market. Apple’s market share fell below 15 percent and the PC industry had been completely reconstructed around the open industry standard. But in spite of its early success, there were a few more challenges that had to be overcome before the future of the open industry standard was assured.

____________

1 AST Research, Compaq, Epson America, HP, NEC, Olivetti, Tandy, Wyse, and Zenith.