The Soul of a New Machine

by Tracy Kidder

  What sorts of cables and connectors should they use in Eagle? Should the machine turn on with a key or a button? Good arguments for both existed. The decision mattered a great deal to the Field Service Department. West would deal with that.

  There was Software to worry about, and Manufacturing — would Manufacturing agree that they could build this machine in large quantities? There was a meeting of the Product Board coming up some time in June. There, for the first time, West would finally present Eagle to de Castro. West started preparing for that day months in advance. He was proselytizing for Eagle

  throughout the company now.

  The Eclipse Group was missing deadlines, but the whole program was moving along fairly well these days. All those

  lessons on new recruits, on software — were paying off, it seemed. but just around the time when West had stopped thinking about his risky decision to use those fancy new chips called PALs, he got word that the sole supplier might be about to declare bankruptcy.

  Something had gone wrong, clearly. They could not get all the PALs they wanted. Never mind getting the thousands they would need to mass-produce Eagles; they scarcely had enough to repair the ones in the prototypes. Rasala took to keeping a list of how many PALs they had on hand. For months, they teetered on the verge of running out. "We could lose the whole thing right there," said

  West.

  Data General was working on the problem through other channels. In the meantime, there wasn't much that West could do about it, except worry. So he did. The younger members of the team did not know at that time how serious the problem was. No one told them, of course.

  These nights, when West got home, he headed for his living room and settled into the corner of an old beige couch, always the same corner. His pants looked baggy on him now, bunched up around his waist, as if pleated. He lay back into the cushion, staring straight up at the ceiling, and sweeping his hair back with one hand, brought a cigarette up to his lips with the other. From this position he drew portraits of members of the team, ones to whom he had scarcely ever said good morning. He knew their hobbies. He knew this one's strengths, that one's weaknesses. It was startling. Sometimes he bragged about them, even Peck. "The guy's good."

  You could never reproduce a project such as this one, even if you wanted to. Wasn't that obvious? West brought up his cigarette and inhaled deeply. "Ummmmmmmh. Yeah, you can repeat it." West stubbed out his cigarette, lit another, and went back to looking at whatever it was he saw in the ceiling. "The postpartum depression on this project is gonna be phenomenal. These guys don't realize how dependent they are on that thing to create their identities. That's why we gotta get the new things in place."

  Before midsummer, while Eagle remained in the lab, still flunking many diagnostics, and while the number of PALs on Rasala's list kept shrinking, West had dreamed up the vague outlines of about half a dozen new computers and had decided, in consultation with his lieutenants, which new machines would best suit which members of the team. "Sure," said West. "What do you think I do all night while I'm staring at the ceiling? I don't just think about boats and sailing away."

  GOING TO THE FAIR

  Each June, in order to promote the further rise of information processing, the computer industry puts on a fair, called the National Computer Conference, and known, of course, as the NCC. This year it would run in New York City for three days. The Eclipse Group was going there for one, in their very own bus. Alsing had made all the arrangements, through Carl Carman. Good old Alsing.

  At roughly six in the morning, Hardy Boys and Microkids drove in one by one, to the parking lot behind Building 14A/B. The sun was up, the day already warm, and for once they did not go inside. Instead they milled around the door of their bus. They were washed and combed and dressed for town, and the smell of after-shaving lotion hung in the air. Rasala and Ken Holberger, however, after shuffling around with the crowd for a moment, slipped away. They went straight to the lab and gathered around Gollum. "It ran EMORT for seven hours last night," said Rasala. "Pretty good." It looked as though they might take up debugging stations. In a moment, though, one of the Microkids stuck his head through the doorway and said, "Hey, the bus is leaving." Rasala wrote Do Not Disturb on two sheets of paper and placed them on Coke's and Gollum's consoles. He hurried from the lab. Swift partings often are best.

  Several didn't make the trip. Rasala tried to talk West into coming; Holberger told West he could wear a disguise and that way have a good time without any of the team knowing it. West refused. He sent along some of the refreshments, through Alsing, with instructions not to say where they came from. In an hour or so, West would be in his office, planning for the coming meeting of the Product Board, surrounded by complete silence for a change.

  Most of those wearing jackets and ties took seats in the front two-thirds of the bus, while those in shirt sleeves occupied the back. Sudden bursts of laughter from the back of the coach, followed by short mysterious silences. Alsing swaying up the aisle, dispensing from a box full of soft drinks and donuts. Holberger and Jim Guyer in a technical dispute about stereo equipment. Steve Wallach holding forth on the subject of industrial recruiters, predicting there'd be headhunters roaming the floors of the conference, in search of young engineers, whom they'd take back to "hospitality suites," ply with liquor and caviar, and talk into defecting to other computer companies. The bus driver calling out over his shoulder, "Keep it down back there!" and his injunction being answered by a roar of laughter from the rear of the bus. It took you back. Doesn't everyone remember awaking to the realization that there would be no school today, that an entire summer lay ahead, that the class was going on a trip? Is there anything more gorgeous than a holiday outing in June?

  In no time at all, it seemed, the bus had achieved Manhattan. Rasala glanced out the window at his hometown without expression. Wallach, who, like Rasala, had grown up in Brooklyn, spoke with the air of a proprietor of local history, about vanished Ebbets Field and the old Brooklyn Dodgers. Bob Beauchamp, who had seen New York City only once, on a trip with his high-school class from Missouri, gazed out his window all the way down from Harlem to the New York Coliseum. Computers and the hosts of equipment that attend them filled the spacious floors. The crowd was thick and wearing badges.

  One badge went by that read Applications Analyst. Was this someone who tried to figure out what to do with all this stuff? "Hey, check this out over here," said Beauchamp. "There's bubble memory over here," said Jon Blau. One knot of Microkids and Hardy Boys headed off across the first floor. Rasala, Alsing, Holberger and Wallach went searching for DEC's booth.

  When they found it, they went right to the VAX 11/780 that was on display. They knelt down. Wallach had just opened up the door of smoked Plexiglas behind which stood the boards of the CPU, when a saleswoman from DEC came over and peered at their badges. " 'Data General, Engineer.'" She made a face and shooed them away. Laughing, they went over to Data General's turf.

  It was at one of these fairs, eleven years before, that Data General had launched itself upon the industry, raising its placard higher than any other company's. This had apparently become a tradition, for Data General's sign was once again this year the tallest in the hall, taller even than IBM's — raised, indeed, to such a' height that it threatened to become unnoticeable. Maybe its altitude was a function of increasing competition. If so, and the trend continued, someday Data General would have to set up its booth outdoors.

  The company always brought something catchy to the fair. One year, Wallach remembered, Data General was introducing a microprocessor, a computer on a chip, so they hired a belly dancer, who plied her trade at the company's booth with the chip lodged in her navel. This year Dr. Gideon Ariel, a practitioner of sports medicine, was the featured attraction. Dr. Ariel had devised techniques for using a computer to improve the performance of athletes and was working with members of past Olympic teams. Data General had given him an Eclipse. An inspired act of largess, which had gotten Data General
a nice mention on NBC's "Today" show. At the moment, a TV screen was playing back that bit of television footage. "Yeah, sure," whispered Alsing. "We gave him an Eclipse with only two-K of memory and no peripherals." A computer thus equipped would be quite useless, that was the joke. It probably wasn't true, but the other engineers grinned. An old-timer, Alsing liked to think of Data General putting something over on national TV.

  Dr. Ariel was up on the stage now, in sneakers and perspiring, demonstrating how his Data General system could capture in detail — in pictures of a sticklike body — a jogger's movements. He was offering vivid, computerized proof that the best way to jog is on the balls of your feet. The engineers turned away.

  "Oh," said Alsing. "There's my machine."

  He went over to the C/150, a recent model of Eclipse. It was the one for which he had written some of the microcode last summer, on his microporch. He played with some of the switches on the front of the machine.

  Holberger and Rasala, meantime, wandered over to the M/600. This was their last machine. Rasala took up a position beside the cabinet that contained the CPU, leaning against it with his arm draped over the top. He wore a faint smile. He looked as cool as a hot-rodder at the wheel, with his arm around the obligatory girlfriend.

  Alsing approached. "Say. I'm a customer, and there's something I don't understand about this computer. What's this switch for?"

  "It works," replied Holberger. "But it's not useful for anything."

  The community of Eagle became, for the day, unglued. I looked around and none of them were in sight except for Wallach, but he was the right companion in this spot.

  A grinning fellow approached, offering his hand to Wallach. For a moment I entertained the hope that this was a headhunter about to bestow hospitality upon us. But he was just the chief engineer of another minicomputer company. He and Wallach chatted for a while, and as the man was walking away, Wallach told me, too loudly for comfort: "They've got a thirty-two-bit computer. A 1963 design."

  You could have sensed that Wallach knew computers just from the way he carried himself. Approaching a display, he looked as though he were sniffing the air. A snort signified a kludge; a shrug, good but ordinary work; a judicious nod, rarely bestowed, that here at last was evidence that some people knew what they were doing. IBM required a lingering inspection. You realized first of all that IBM's large and many-sided booth was not made of ersatz materials but real oak. "You'll notice," Wallach added, "they have just the right number of black people." Women, too.

  IBM seemed to be making no attempt at flashiness. They had no catchy display, just men and women in white shirts describing the machines on exhibit. These computers were mainly members of IBM's new 4300 line, which the company had announced not many months before. In fact, at the moment, IBM didn't have to talk people into buying this new kind of machine. IBM's problem just then was how to produce enough of these computers to meet the orders rolling in. At the moment, an estimated three years' worth of back orders had piled up, a piece of good luck for the many companies that, like pilot fish, follow IBM around —the so- called plug-compatible manufacturers, which produce processors and various other equipment that can be hooked up without modification to computing systems built mainly around IBM computers. In some sense, all companies in the computer industry are orchids on IBM's tree. Everyone has to consider IBM's pricing and strive for compatibility with IBM equipment.

  It seemed curious: a company suffering from too much demand. Not that in the long run IBM was going to suffer more than discomfort from this backlog, but the collapse of many small, promising computer companies had begun with a similar problem. They'd announce a new product and then for one reason or another they wouldn't be able to produce it in sufficient quantities to meet their obligations. They'd asphyxiate on their own success. But a small company had to court disaster. It had to grow like a weed just to survive.

  From IBM Wallach led me to some of its traditional competitors. First, to Sperry Univac, the descendent of the first real computer company, which might have become the IBM of the industry had it not blown its early lead; Sperry had a big new machine and a slide show all about it. Burroughs had erected a small theater, with chairs set up to face a rank of computers in big white cases that looked like nothing so much as dishwashers and refrigerators; a recording of trumpets playing fanfares ushered you into this homely spectacle. The main thing memorable at National Cash Register was the pair of golden-haired women, identical twins, posing at its booth. "The bipolar blondes," said Jon Blau, when we ran into him later. It was a fairly witty remark, drawn from the language of semiconductor technology, but one of those that loses flavor during the time it takes to explain it.

  Wallach took a look at all the 32-bit superminis at the fair, the machines with which Eagle, if it ever got out the door, would presumably compete. These machines were hot items now. Some estimates had it that total sales and back orders for DEC's VAX were approaching a thousand, which would translate into something like half a billion dollars in revenue, just for that one new machine.

  A certain broad division between companies was apparent. The established and successful, with many wares to show, had sectioned off pieces of the central floors, setting up theaters and playhouses. The smaller, newer, less diversified had small booths along the walls. They were the little jewelry and camera stores west of Broadway. Wallach gave them a window-shopper's tour. Many of these little outfits were selling "pin-compatible memories," ones that could be added to DEC'S and Data General's machines more cheaply than the extra boards offered by the big manufacturers. "We're suing some of these guys," said Wallach.

  We glanced at many systems for producing graphics, one of the latest segments of the industry to experience a boom. You can draw pictures with these machines. Some of the most popular programs for graphics systems created charts of the type dear to executives with ambitious plans they want to sell to their bosses — the kind of charts that carve some commodity such as revenues into various parts, as you might divide a pie. Computers were cranking out many sorts of pie charts: multicolored pie charts, pie charts that rotated, pie charts in three dimensions. We wandered past array processors performing tricks in FORTRAN, past tape drives, Winchester disks, printers and consoles in sleek- looking packages, and the latest "modems" for hooking up computers to telephones.

  "I notice more Japanese here than in other years," said Wallach. This was a big issue. Already the Japanese had made substantial inroads into the American market for integrated circuits; and you knew that segment of the industry was worried, because they were clamoring for protective tariffs. Data General, on the other hand, had bought half of a Japanese minicomputer company — no flies on Data General.

  We wandered on, casting passing glances at trade magazines, which were selling advertising and subscriptions; past software houses, which specialize in creating those crucial, ever more expensive user programs that had set the stage for Eagle; past systems houses and OEMs peddling turnkey systems —buy from us and you won't have to worry; just turn the key and your computer system works. A great deal of the overall industry, it seemed, consisted of computer companies taking in each other's wash.

  Here and there we saw products of the future, such as bubble memory. After a while, however, most of what was on display began to look the same. From every booth came the insistent clamor, that here, at last, was the piece of computing gear that was the fastest, the most reliable, the easiest to use, the "most intelligent," simply the greatest in the universe. The spectacle began to overwhelm me — not so much the variety and versatility of machines themselves, but the volume of gear, the size of the crowd, the sheer number of firms. Wallach said that two of the best-named companies were not on hand: Itty Bitty Machines (another "IBM") and Parasitic Engineering. But I saw many other names, passing by. Among others, I saw Centronics, Nortronics, Key Tronic, Tektronix and also General Robotics. There were Northern Telecom and Infoton and Centurion, which had a fellow dressed as a Roman sol
dier standing by its booth. There were Colorgraphics and Summagraphics; Altergo and C. Itoh; and Ball.

  "Hey, wait a minute. What's Ball doing here? Aren't they the mason jar people?"

  "Yeah, but they also make disk drives."

  Also: the Society for Computer Simulation, and Randomex, and Edge Technology, and Van San, which sold "Quietizers." There were Datum, Data Pro and Data I/O, Tri Data, Epic Data, Facit Data, Control Data, Decision Data, Data General and Data Specialties. And we didn't have time even to glance at the wares of Itek, Pertec, Mostek, Wavetek, Intertek, Ramtek ... Ah, Ramtek.

  "In 'seventy-three," said Wallach, "there were two floors, and now there are four floors and it's just as crowded."

  Norbert Wiener coined the term cybernetics in order to describe the study of "control and communication in the animal and the machine." In 1947 he wrote that because of the development of the "ultra-rapid computing machine,... the average human being of mediocre attainments or less" might end up having "nothing to sell that is worth anyone's money to buy." Although Wiener clearly intended this as a plea for humane control over the development and application of computers, many people who have written about these machines' effects on society have quoted Wiener's statement as though it were a claim of fact; and some, particularly the computer's boosters, have held the remark up to ridicule — "See, it hasn't happened."

  Since Wiener, practically every kind of commentator on modern society, from cartoonists to academic sociologists, has taken a whack at the sociology of computers. A general feeling has held throughout: that these machines constitute something special, set apart from all the others that have come before. Maybe it has been a kind of chronocentrism, a conviction that the new machines of your own age must rank as the most stupendous or the scariest ever; but whatever the source, computers have acquired great mystique. Almost every commentator has assured the public that the computer is bringing on a revolution. By the 1970s it should have been clear that revolution was the wrong word. And it should not have been surprising to anyone that in many cases the technology had served as a prop to the status quo. The enchantment seemed enduring, nevertheless. So did many of the old arguments.