Idea Man

by Paul Allen

  A short time later, we licensed BASIC to NCR for $175,000. Even with half the proceeds going to Ed Roberts, that single fee would pay five or six programmers for a year. Though we still relied on MITS’s marketing team to help with OEM sales, our business was growing fast. Over time, as Microsoft became the language development company for the personal computer industry, its partners divided the labor to play to their strengths. Bill focused on legal and contract issues, drummed up new business, and navigated our license sales. Whenever I pointed him to a new microcomputer, he’d be all over it to try to sell our software.

  It was my task to guide our programmers as we crafted BASIC for a mushrooming OEM clientele, adding features and fixing bugs. I did a good deal of the grunt work myself and made sure that our products got delivered and properly implemented and upgraded. But my most vital charge was to chart our future. Where Bill eyed tomorrow’s markets, I looked to a more distant horizon. What would our customers want six months or a year from now, and what did we need to do to get it to them before anyone else?

  Our primary task was to adapt BASIC, our bread-and-butter product, for a new group of microprocessors now competing with the Intel 8080. This work had kicked off the previous fall, when MITS announced an Altair for Motorola’s new 8-bit chip. It gathered steam with the release of the 6502 chip from MOS Technology at an unheard-of $25. For each new microprocessor, I created a new set of development tools on the PDP-10, while Ric helped with the BASIC interpreter rewrites. Though the work was a grind, I was glad to extend our company’s reach. Bill and I aimed to provide all language software for every microcomputer on the market. Those BASIC adaptations would be the bedrock of our revenue for years to come.

  I also kept pushing to extend our work into other programming languages. In August we hired Steve Wood to develop an 8080 FORTRAN compiler to broaden our clientele among scientists and engineers. When our FORTRAN-80 was released the following April, we were no longer just “the BASIC company.” We had a second product line.

  * * *

  BY OCTOBER 1976, Micro-Soft had outgrown our living room. We ordered swivel chairs and desks from a discount house and moved to our first real headquarters, a leased suite of four offices just off Central Avenue. The decor was modest and the offices small, but they had spectacular views of the thunderstorms scudding across the desert valley.

  As business pressures weighed on Ed Roberts, our interests grew less complementary. To maximize Micro-Soft’s revenue, we needed to distribute our software as widely as possible; in Ed’s perfect world, we’d sell it to Altair owners exclusively. Our relationship became strained when he jammed me to release a version of BASIC that had fallen behind schedule.

  “But it’s got bugs in it,” I told him. “You can’t ship something with so many problems. The customers are going to kill us.” We might deliver late, but I always wanted our products to be top quality before they were released. I’d seen what had happened to MITS with its off-brand memory cards, and how a lemon could permanently tarnish a company’s reputation. After a shouting match, Ed gave in. In November I resigned from MITS and moved full-time to Microsoft, the trade name we had registered with the State of New Mexico around that time. With Bill about to quit college for good and Chris Larson back on board, we leased another four offices early in 1977. We were at full strength now, our forces joined, ready to rock ’n’ roll.

  CHAPTER 8

  PARTNERS

  Bill’s intensity was nonstop, and I couldn’t keep working with him and living with him, too. It was time to leave the Portals. I rented a rambling three-bedroom house in the suburbs with Ric and Marc, while Bill and Chris Larson took an apartment near the air force base. When Bill asked me for a walk and talk one day, I knew something was up. We’d gone a block when he cut to the chase: “I’ve done most of the work on BASIC, and I gave up a lot to leave Harvard,” he said. “I deserve more than 60 percent.”

  “How much more?”

  “I was thinking 64–36.”

  Again, I had that moment of surprise. But I’m a stubbornly logical person, and I tried to consider Bill’s argument objectively. His intellectual horsepower had been critical to BASIC, and he would be central to our success moving forward. That much was obvious. But how to calculate the value of my Big Idea—the mating of a high-level language with a microprocessor—or my persistence in bringing Bill to see it? What were my development tools worth to the “property” of the partnership? Or my stewardship of our product line, or my day-to-day brainstorming with our programmers? I might have haggled and offered Bill two points instead of four, but my heart wasn’t in it. So I agreed. At least now we can put this to bed, I thought.

  Our formal partnership agreement, signed on February 3, 1977, had two other provisions of note. Paragraph 8 allowed an exemption from business duties for “a partner who is a full-time student,” a clause geared to the possibility that Bill might go back for his degree. And in the event of “irreconcilable differences,” paragraph 12 stated that Bill could demand that I withdraw from the partnership.

  Later, after our relationship changed, I wondered how Bill had arrived at the numbers he’d proposed that day. I tried to put myself in his shoes and reconstruct his thinking, and I concluded that it was just this simple: What’s the most I can get? I think Bill knew that I would balk at a two-to-one split, and that 64 percent was as far as he could go. He might have argued that the numbers reflected our contributions, but they also exposed the differences between the son of a librarian and the son of a lawyer. I’d been taught that a deal was a deal and your word was your bond. Bill was more flexible. In my experience, he believed that agreements were open to renegotiation until they were signed and sealed. There’s a degree of elasticity in any business dealing, a range for what might seem fair, and Bill pushed within that range as hard and as far as he could.

  THE MOMENTUM IN personal computers shifted for good in 1977, away from self-made trailblazers like MITS and IMSAI and toward big brand-name companies. Three second-generation machines, the “1977 Trinity,” were released over a span of six months: the Apple II, the Commodore PET, and Tandy’s TRS-80. All were fully assembled, out-of-the-box computers with built-in keyboards. The Commodore and Tandy threw in integrated monochrome monitors and cassette recorders, and were bargain-priced at around $600. The higher-end Apple II, at $1,298 and up, was easily expandable and offered color graphics capability. (By contrast, an assembled Altair 8800b—with no monitor, keyboard, memory, or data-storage device—sold for $1,070.)

  The newcomers were plastic builds with ersatz space-age stylings, and none of them bowled me over. The PET had a horrible “Chiclet” keyboard; the TRS-80 was awkward to expand. The Apple II had a better design, but it was pricey and came without a monitor. None of the three were packaged at the start with floppy disk drives. But despite their flaws, they offered more computer for less money than anything before them, and they all sold well. It wouldn’t be long, I thought, before turnkey machines—better turnkey machines—were everywhere. And as computer hardware kept getting smaller and faster and cheaper, better software would be needed to create a compelling package.

  Even as I tracked the computer flavor of the month, I never stopped thinking about the advances to come and how people would use them. Here’s an excerpt from my column in Personal Computing, circa January 1977, four years before the Osborne I became the first portable computer: “I expect the personal computer to become the kind of thing that people carry with them, a companion that takes notes, does accounting, gives reminders, handles a thousand personal tasks.”

  A few months later, in an interview with Microcomputer Interface, I took my train of thought further. Fifteen years before the World Wide Web, I imagined a computerized society that was far-flung yet intimately linked:

  For machines like the [Commodore] Pet that aren’t connected to any central network, I don’t see much of a future. I don’t see the housewife really learning to program in BASIC. What I do see is a home t
erminal that’s connected to a centralized network by phone lines, fiber optics or some other communication system. With that system you can perhaps put your car up for sale or look for a house in a different city or check out the price of asparagus at the nearest grocery market or check the price of a stock. …

  The technology wasn’t nearly there yet, and I didn’t use the phrase in so many words, but that was my first public intimation of what I’d later call the Wired World.

  THE 1977 TRINITY effectively doomed the Altair. Ed Roberts was a man of huge vision but weak execution; he’d set off the revolution but couldn’t keep his company in front. Most of all, Ed lacked the relentless price-cutting mindset you needed in a company selling to a mass market. But even had he done everything right, MITS’s days were numbered from the start. As microcomputers became more functional, hobbyists were finding ways to use them at their jobs as stockbrokers, research scientists, and engineers. And as the market became more lucrative, big companies swooped in, just as Texas Instruments had with handheld calculators. There was no way for MITS to match Apple’s innovation or Tandy’s economy of scale and its RadioShack distribution network. Ed got overwhelmed and burned out.

  For Bill and me and our compatriots at the young Microsoft, on the other hand, this should have been the best of times. The makers of the Trinity knew that they couldn’t develop their own BASIC quickly enough, certainly not one as good as ours. Apple tried to get by with a homegrown version, but customers complained that it lacked floating-point math. The company wound up licensing our 12K BASIC interpreter, burned it into read-only memory (ROM), and branded it Applesoft. (Whereas RAM is reusable and can accommodate any number of programs, ROM is a fixed-memory chip on a computer’s motherboard.) Just two years out of the gate, Microsoft was establishing the industry standard for microprocessor languages. In a news report that spring on Albuquerque’s KOB-TV, a local expert marveled that it was “getting to the point where software will cost as much as the machine.” Our company seemed poised to prosper. Until, that is, we were threatened with the loss of our keystone product.

  As MITS’s market share eroded, and Ed Roberts became desperate to retain a competitive edge, he began killing deals to license our software to Altair rivals. He maintained that he was under no blanket obligation to sublicense our source code to competitors. That was a big problem for Microsoft, because he perceived just about every microcomputer company as a competitor. The issue became acute after Ed decided to sell MITS to Pertec, a Southern California manufacturer that believed it was buying all rights to BASIC. In April 1977, as the deal with Pertec was closing, Ed canceled two of our third-party sales. After we threatened to terminate the MITS contract, MITS/Pertec filed for arbitration. They got a judge to bar any new BASIC sales while the decision was pending. We were frozen out of our main source of revenue.

  In June, with our cash flow drying up, Bill and I prepped for our testimony with our attorney, Paull Mines. After each session, we’d race down to the garage to see who could get back to our office first without a traffic ticket. My Monza was quick but no match for Bill’s Porsche. I had to be crafty to beat him, taking shortcuts through alleyways and parking lots, and even then I’d be hard-pressed to win one time out of four. One day I left a session to get a document from Bill’s car. At the end of the afternoon, Bill and I looked at each other and nodded, our signal that the race was on. I tore out of the garage and won easily, though I thought it was strange that I hadn’t seen the Porsche the whole time. Then I felt in my pockets and found Bill’s car keys. After I returned to the law office, Bill shook his head in disgust. “That’s the only way you beat me,” he said.

  The hearing was grueling. It was disquieting to see Pertec line up their three lawyers against our one, and tough to watch Ed testify against us. The arbitrator’s leanings seemed to shift from day to day. It gradually dawned on us that he had little understanding of software, a mostly untested arena for litigation. After ten days of testimony, he took the case under advisement with no telling when we’d get a final ruling.

  A lot rode on the outcome. If Pertec won the rights to our source code, it would mean the end of Microsoft BASIC as we knew it, along with the lion’s share of our business. We’d have to write a new BASIC from scratch or move on to a different kind of software. Our future would be imperiled.

  Meanwhile, the restraining order was starving us. As summer turned to fall, we struggled to make payroll and to cover our rent and time-share expenses. It reached the point where we had to borrow $7,000 from Bob Greenberg, a Harvard classmate of Bill’s whom we’d hired to write a BASIC for a new Texas Instruments chip. We owed our lawyer and could be liable for the costs of the hearing if we lost. The other side figured they could use their deeper pockets to outwait us, and I began to wonder if they might be right. “Look, we’re just about out of money,” I told Bill over dinner one night. “I think we should consider settling.”

  And Bill said, “You’ve got to trust me on this one. I talked to my dad, and he thinks we’ve got a good chance to win.” When pressed, Paull Mines took the same position. So I swallowed hard and hung in there, and it was the best advice Bill never took from me. In November, seven long months after the process began, the arbitrator handed down a twelve-page decision. The special clause in our contract, those papers that Ed never checked, had made all the difference. “The testimony was undisputed that MITS never really embarked on what could be considered best efforts in marketing the source code,” the decision read. By vetoing our sublicense sales, MITS/Pertec “materially breached its best efforts obligation. … I find this an act of corporate piracy not permitted by either the language or any rational interpretation of the Contract.”

  The ruling was a total victory for Microsoft. Our contract with MITS was terminated, with Pertec held accountable for all unpaid royalties. Most crucially, Bill and I recovered all rights to our BASIC interpreter and could now sell it to whomever we pleased—and better yet, keep all the revenue. Our one big roadblock was gone.

  Not long after that, Ed Roberts quit Pertec after the company rejected his design for a laptop computer. “We’re not convinced that people need personal computers on their desks,” an executive told him, “but we’re sure as hell convinced that they don’t need them in their laps.” (Like me, Ed could be too early with a promising idea.) In 1978, the TRS-80 shipped 100,000 units; the Commodore PET, 25,000; the Apple II, 20,000; the IMSAI, 5,000; the Altair, 3,000. Soon after, Pertec discontinued the Altair brand in favor of its own label. It closed the Albuquerque plant in 1980 and moved production to California.

  As always, Ed found something else to do. He bought a vegetable farm in Georgia, entered medical school in his midforties, and became a country doctor. Though he lived a rich life after MITS, he felt bitter to be left out of the history books. “We created an industry,” he said in Triumph of the Nerds, a 1996 documentary, “and I think that goes completely unnoticed.” He was half right. Ed did indeed create the first truly commercial personal computer, the first widely affordable general-purpose machine. He spearheaded every aspect of microcomputer marketing, from publications and conventions to a retail dealer network. His imagination was boundless. The Altair even debuted a digital camera interface back in 1976.

  And he gave two college dropouts the opportunity of their young lives.

  But Ed was wrong about being forgotten. When he passed away in April 2010, his obituary made the front page of the New York Times. (The photo showed him leaning over the Altair 8800 in his doctor’s whites.) I’d made up with Ed a long time before, and he and Bill had gotten past their differences. When Bill flew out to see him a few days before he died, Ed was talking about the latest nanotechnology and how he might work with it. He was looking ahead all the way to the end.

  AFTER WINNING IN arbitration, we repaid Bob Greenberg with interest and closed our BASIC deals for the Commodore PET and the TRS-80. The flat-fee price tags on those licenses were whatever the market would bear, because n
o one knew for sure what the personal computer would become. I flew to Fort Worth to meet Bill for a demo at Tandy, which had taken a lot of heat for using Tiny BASIC in an initial manufacturing run of the TRS-80. We were ushered into the sprawling office of John Roach, a tall, plain-spoken Texan who was en route to becoming Tandy’s chairman and CEO. He was not a man to be trifled with.

  “Now,” Roach said in a thick drawl, “can you boys really deliver a BASIC interpreter that works for our computer?”

  “We believe we can,” Bill said, and then he rattled off some of our software’s outstanding features.

  Roach nodded and said, “And how much is this going to cost me?”

  And Bill said as coolly as he could, “Fifty thousand dollars.”

  “That,” Roach said flatly, “is the biggest pile of horseshit I ever heard.”

  Bill and I looked at each other, wide-eyed. We’d heard a lot of bargaining tacks, but this was something new. Afterward Bill said to me, “Well, maybe I did ask for a lot, but horseshit?”

  We didn’t back down, though, and eventually got our price. I always marveled at Bill’s bravado; he’d come on superconfident, and people like John Roach never knew how much we feared losing the deal. Our near-disaster in arbitration was one more lesson for us. Going forward, we would aim for maximum market share in any sector we entered. You could never have too many customers.

  In any event, Tandy did well by our arrangement. With our BASIC inside, the TRS-80 became the hottest-selling computer in the world until Roach got outflanked by Apple. But we were used to the shifting landscape of the hardware market by then. Machines came and went; good software lived on.