Out of a Beige Box, a New Era

The list of inventions that have truly changed the world is short. But by almost any standard, the roster should include the little machine introduced 20 years ago this Sunday by the world’s biggest computer company.

The IBM personal computer, unveiled Aug. 12, 1981, was not the first PC on the market. Nor was it the cheapest, the most powerful or the most technically advanced. In the strictest sense of the word, it was not even an “invention”: Although IBM engineers devised its overall architecture, they relied on components, including the microprocessor, memory chip and floppy disk, that had been developed by others over a period of years.

But IBM’s introduction of a so-called microcomputer under its own brand name rocketed what had been a niche market into the big time. Corporate purchasing agents had been wary of buying PCs from the leading maker, Apple Computer, which had been in existence for scarcely five years. But as the saying went, “You couldn’t be fired for buying IBM.” Within a year, IBM dominated an industry that had grown to more than 100 manufacturers shipping 2.4 million machines annually.

IBM “had the size, the resources and the experience to set up the infrastructure to deliver millions of these,” Kenneth H. Olsen, founder of Digital Equipment Corp., a leading maker of medium-sized minicomputers, said later. “After they had done it, it became easier for others to enter the market.”

By selling PCs through Sears and other consumer retailers, moreover, IBM signaled the machine’s elevation from a plaything for nerds into an appliance for every home, like a refrigerator. Another lasting, if unintentional, legacy was the creation of one of the world’s largest monopolies and personal fortunes: Microsoft, which provided IBM with some of the machine’s fundamental software and went on with co-founders Bill Gates and Paul Allen to ride the PC boom to billions of dollars in riches.

Technology’s Virtues, Drawbacks

To computer historians, IBM’s introduction of its microcomputer marks the start of the PC age, an era in which the beige-colored box with a video screen perched on top came to exemplify technology’s virtues and drawbacks in equal measure.

In naming the computer its “machine of the year” for 1982, Time Magazine managed to be both hyperbolic and conservative in its prognosis. Time reported that most Americans expected PCs to be “as commonplace as televisions” in the near future (today only about half the households with TVs also own PCs, and the ratio might never change much). But it underestimated the growth of the overall market by noting that experts were forecasting that 80 million PCs would be in use by 2000. In fact, there are an estimated 500 million PCs in use worldwide today; last year, new sales alone totaled 131 million.

The PC’s evolution has mirrored the march of technological change: Today’s $1,000 desktop machine runs on a chip with roughly 300 times the power of the original. It holds at least 2,000 times as much memory and nearly 120,000 times the data-storage capacity--all for a price that has fallen, in constant dollars, by more than 80%.

Nevertheless, many users curse the PC as a device that, despite 20 years of development, still doesn’t work right.

In part, that is because the PC is expected to perform so many disparate functions that its software programs often interfere with one another, producing its notorious propensity to freeze and crash (often blamed on glitches embedded in the standard operating software from Microsoft).

Moreover, the question of whether the PC has been an unalloyed economic blessing remains wide open. Although corporate purchases of PCs have been justified for decades by the expectation of sharply increased worker productivity, there are signs that computerization might have the opposite effect in some industries, partly because it generates billions of dollars in hidden costs.

“It’s very costly to maintain this stuff. There’s a long learning curve. They break down all the time,” said economist and business essayist Jeff G. Madrick. Much of what computers enable people to accomplish doesn’t generate real economic gains, he contends.

“Legal briefs are a lot longer than before, but the practice of law isn’t arguably better,” he said. “Writers write faster, but certainly novels aren’t any better.”

Madrick does believe, however, that computers might have helped industry manage a social change that already was occurring--the fragmentation of the consumer market because of buyers’ demands for more choice. Without small, fast computers, manufacturers might be hard-pressed to keep straight the vast variety of detergents, automobiles and, indeed, PCs themselves that inundate the marketplace. “The computer has been a great tool for keeping up,” he said.

Meanwhile, some experts speculate that as a centerpiece of daily life, the PC already might have reached its high-water mark. PC sales in the U.S. are falling this year for the first time.

Industry analysts expect the PC’s 55% household penetration in the U.S. to grow slowly, if at all, in coming years. “There are fewer and fewer new buyers,” said Roger Kay, director of client computing at the market research firm IDC. “People who want and can afford PCs already have them.”

Donald Norman, a former Apple and Hewlett-Packard executive who advises companies on ergonomic design, said, “We’re beginning to see the transition” to more specialized computing devices. In the future, he says, more homes might be equipped with a central computer managing a network of household appliances, including information and entertainment devices. But it will do most of the work behind the scenes, like the computers that regulate the engines of modern automobiles. “You won’t need to touch it,” Norman said.

That would be an improvement on today’s PC. “Almost all users use their computers for only a few options,” wrote Thomas K. Landauer, an expert in human-computer interaction at the University of Colorado, in his 1995 book “The Trouble With Computers.” “Nevertheless, their machines and minds are loaded up with a vast junk pile of options, commands, facilities, doodads and buttons--most of them superfluous . . . the situation cries out for simplicity, but the cry goes unheard. Instead, we get ever greater complexity.”

Long before IBM stepped into the fray, the personal computer had been coming on strong. The first one worthy of the name had been invented in 1973 at Xerox Corp.’s Palo Alto Research Center. Engineers at PARC foresaw that the price of computing resources--then so scarce that most users had to access large-scale computers via “time sharing”--was falling so fast that within 10 years, individual computers would be cheap enough to place at the disposal of individual users.

Alan Kay, a PARC computer scientist who coined the term “personal computer,” illustrated his vision by fashioning a cardboard mock-up of a notebook-size keyboard-and-screen device, which he called a “Dynabook.” The first working prototype was not quite that small. Dubbed the “Alto,” it was a squarish box on casters that just fit under a desk and featured a vertically oriented 8 1/2-by-11-inch screen displaying black characters on a white background.

During the next few years, more than 1,000 Altos were built for PARC. The computers were linked together by a local area network called Ethernet and individually to laser printers--both of them PARC inventions. The lab wrote reams of software for the unique machine, including the first “what you see is what you get” word processor, which displayed typefaces identical to those that would print out on a page, and a graphical screen display that became the precursor to that of the Apple Macintosh and Microsoft Windows.

Outside PARC’s walls, a separate hobbyist market was emerging, inspired by the 1975 launch of the Altair 8800 by a tiny Albuquerque electronics firm. The Altair, which was named for the Starship Enterprise’s destination on an episode of “Star Trek,” was a squarish blue box with 36 red lights and toggle switches on its front panel. For a mail-order payment of $397, buyers received an empty case and a bag of electronic geegaws needing to be carefully soldered together. These included an Intel 8080 microprocessor, one of the earliest in a revolutionary line of silicon chips that eventually would beget the Pentium.

The Altair did not do much at first beyond accepting primitive commands that set the lights blinking in a programmed pattern. But it was a huge hit and soon faced a boatload of competitors. These included machines from Apple Computer, which had been started by Silicon Valley schoolmates Steve Jobs and Steve Wozniak, and Radio Shack, whose TRS-80 package, including a keyboard and video display, sold for $599 through its 3,000 stores nationwide.

Bursting Out of Its Hobbyist Pigeonhole

In 1975, a computer engineer named Dick Heiser opened the first independent computer store in the world on Pico Boulevard in West L.A., calling it simply The Computer Store. “It was strictly a hobbyist thing,” recalled Heiser, who had been awarded an exclusive contract to sell Altairs. “One Altair dealer in the Bay Area named his store Computer Kids, though I think he regretted that later.”

It would not be long before the personal computer burst out of its hobbyist pigeonhole. The catalyst was VisiCalc. The first commercial spreadsheet, VisiCalc was a sort of turbo-charted calculating program that allowed users to input numerical values and manipulate them by row or column, creating (among other things) complex statistical analyses.

VisiCalc was recognized instantly as a powerful business tool. When a version was released in 1979 for the new Apple II, it single-handedly turned that machine into a must-have business appliance.

By early 1981, the microcomputer market was full of competitors, including such major companies as Hewlett-Packard. In April, Xerox introduced the Star, the product of a five-year effort to build the PARC Alto and its associated technologies into a commercial line.

The Star featured an Alto-style vertical screen, with graphics, and a suite of office software that included a word processor, document filing system and even a way to print scientific formulas. All were so cleverly integrated that the system could still draw thunderous applause from a standing-room audience in 1998, when several of its original designers cobbled together a working model from bits and pieces of decommissioned Stars for a historical demonstration at the PARC auditorium.

But the Star was aimed at Xerox’s traditional corporate market. Not only was each individual computer priced at more than $16,000, they were designed as pieces of a system, to be networked together and to central printing stations, for a total installation price of close to $250,000. Unfortunately for Xerox, this was a market that was about to vanish forever, done in by a most unexpected executioner.

IBM’s plans to enter the personal computer market had been the most closely guarded secret in the industry. The company had all but missed out on the last computer revolution--the minicomputer, a medium-sized machine that had deeply eroded IBM’s franchise in giant mainframes. Although most observers expected Big Blue to do something, even insiders at the vast company doubted it could get a personal computer into the market in less than three years.

“Everyone always talked about how getting IBM to move was like getting elephants to dance,” Heiser recalled.

‘Executing Programs and Feeling Good’

What no one anticipated was that IBM would short-circuit its own bureaucracy. Working out of a branch office in Boca Raton, Fla., a small team assembled the product from standardized components made by other companies and available off the shelf. From Microsoft, then a tiny partnership in Bellevue, Wash., it licensed several programming languages and the most crucial software of all, an operating system--the program that allowed the machine’s microprocessor and memory to communicate with its floppy disk drives. (Microsoft won that contract after IBM’s first choice, a California company named Digital Research, balked at Big Blue’s restrictive terms--thus joining the historic roll of missed business opportunities.)

Compared with the technologically sophisticated Xerox Star, the IBM PC was like an artifact from the Stone Age. Rather than the Star’s crisp graphical display, the IBM machine featured a bilious green monochrome screen with a text-based interface. Users had to type in commands in a specialized language and learn to interpret its responses. In its earliest ads for the machine, featuring Charlie Chaplin’s “little tramp” character, IBM sought to make a virtue out of the laborious training process: “There’s no reason why you can’t be executing programs and feeling good with the results within your first week,” read one.

But the IBM PC had several advantages over the Star. The latter was designed around a “closed” architecture, requiring Xerox engineers to be called for any change in its software. By contrast, IBM’s architecture was “open,” meaning that anyone in the world could write software for the PC. The resulting proliferation of thousands of independently marketed programs contributed mightily to its explosive success.

Another factor was cost. At $3,000 for the standard version, the IBM PC was cheap enough to place on every executive’s desk. (More elaborate models sold for as much as $6,000.) At a time when typewriters were strictly secretaries’ tools and copy machines were hulking contraptions that got their own rooms, this deceptively trivial change fundamentally altered the dynamics of office work.

The PC ushered in a period of headlong innovation and ferocious competition. The period of six years or so following the IBM launch “was the golden age,” said John Dvorak, a long-term observer of the industry who today is a columnist for PC Magazine.

Within a few weeks of the IBM PC launch, a small team of executives at Texas Instruments quit to found Compaq Computer. “We knew that if there was any time to start a company, this was it,” recalled Rod Canion, who became the new company’s first chief executive.

The Compaq group’s idea was to build a portable PC better than anyone else and to make it absolutely compatible with the IBM so that software written for IBM’s model would run flawlessly on the Compaq. That meant using the same electronic components and writing a basic input-output system, or BIOS--the program that kicks in with the “on” button and serves as a PC’s starter engine--that duplicated IBM’s without infringing its copyright.

The first fully compatible clone, the Compaq portable debuted in November 1982, housed in a suitcase-size case so heavy it was popularly known as a “luggable.” (Compaq included Microsoft’s operating system, for the software company’s deal with IBM allowed Gates to license his software to other manufacturers, a key to the burgeoning Microsoft empire.) Compaq’s 1983 sales of $111 million were the largest first-year sales of any company in history. By 1987, sales were $1 billion a year, and it ranked as the largest PC company in the world until recently, when it was unseated by Dell Computer.

For Some, It Was Boom and Bust

Other companies experienced a boom-and-bust life span that would be familiar to any veteran of the recent dot-com crash. In San Diego, Andrew F. Kay, whose Non-Linear Systems had been manufacturing digital voltmeters since 1953, saw the PC revolution approaching even before IBM’s launch. After learning that Radio Shack had sold 120,000 computers in 1980, he started developing his own.

The Kaypro computer, which also came in a luggage-style case, reached the market in March 1982 for $1,795. That year, Non-Linear, which had never sold more than about $3.5 million worth of electronic instruments in any 12-month span, sold $72 million worth of computers. The next year, it went public as Kaypro.

“We went from $72 million to $120 million in another year, and I was set to double again in 1985,” recalled Kay, 82, who still manufactures desktop computers out of a facility in Solana Beach. “It never happened. Cheaper computers started coming in from Japan. Cheap, cheap, cheap--that was all anybody wanted.”

In 1990, Kaypro went bankrupt, taking Kay’s last $10 million with it. “What I should have done, instead of putting my entire personal fortune into it, was let the company get smaller,” he said recently.

Not that IBM found the PC to be the key to unalloyed success, either. “IBM could have maintained its leadership, but it gave up on its own standard,” recalled Dvorak. Irked by competition from the clones, IBM in 1987 dropped the PC architecture in favor of a substitute that it hoped to keep under control.

The decision was a blunder. IBM’s original design, clunky as it was in many respects, had become the de facto standard. Major manufacturers, given the choice of keeping the PC design or trying to clone IBM’s alternative, stuck with the PC. To this day, many believe that saved the computer world from domination by one company.

“People complain about Microsoft stifling innovation,” said Compaq’s Canion, who is today chairman of Questia, an online library of academic books and journals. “But if IBM had won, things would have been much worse.”

Instead, IBM became just another PC maker, with worldwide market share today of scarcely 7%. Its design, however, did come to dominate the computer world: Apple Computer, which once shared a 39% market share with Tandy and developed its own proprietary computer architecture, now accounts for a mere 3% of the worldwide consumer market. Virtually all the other machines out there claim as their familial ancestor that very first IBM PC.

*

MORE INSIDE

Computing milestone: The PC has a past richer and an impact greater than most 20-year-olds. Section T