Hard Drive: Bill Gates and the Making of the Microsoft Empire (12 page)

In June of 1975, the MITS Mobile Caravan was out on the West Coa
s
t for the National Computer Conference. One of its stops was Rickey’s Hyatt House in Palo Alto. By then, a number of Homebrew members had ordered an Altair. When the MITS- mobile arrived, about two hundred people crowded into the Hyatt’s Edwards Room to look at the Altair. They discovered an Altair with features not available on the machines they had ordered. It was connected to a teletype and a paper-tape reader, and was running the 4K version of BASIC developed by Gates and Allen. At this point, no one at Homebrew who had ordered an Altair had received a copy of BASIC, though they had sent MITS their money for the program. According to one account, someone from Homebrew picked up the punched paper tape containing BASIC lying on the floor near the Altair. Someone else later ran o£F copies of the tape, and at the next Homebrew meeting a large box of tapes were passed out to anyone who wanted one. Gradually, then exponentially, BASIC spread from computer club to computer club like a virus. And no one was paying for it.

When Gates learned what was going on, he was beside himself. No wonder he and Allen were receiving so little money from their royalty agreement with MITS, he thought. One day he stormed into Roberts’ office and threw one of his fits that many around MITS had become accustomed to. “I vividly remember the conversation,” recalled Roberts, “him coming into my office that first summer and screaming and yelling at the top of his lungs that everyone was stealing his software, and he was never going to make any money, and he wasn’t going to do another thing unless we put him on the payroll.”

Roberts said he put Gates on the company’s payroll for about a year, and paid him about $10 an hour. Gates, however, later claimed he never worked for MITS. Technically, he was right. According to David Bunnell, the salary Gates received for the hours he worked enhancing and selling BASIC was actually an advance against royalties. He was never on the MITS staff.

Gates eventually became convinced that BASIC wasn’t selling very well because so many people had obtained copies without paying for them. At one point, frustrated and demoralized, Gates offered to sell Roberts all rights and ownership to BASIC for about $6,500. In hindsight, it would have been the bonehead deal of the century. “Clearly, it would have been a bad decision on Bill’s part, because there might not be a Microsoft today,” said Eddie Curry. But Roberts decided not to take Gates up on the offer. He later told Curry that he liked both Gates and Allen and didn’t want to take advantage of them because they were so young. In truth, Roberts decided it made better business sense to continue paying royalties to the two, and reap whatever benefits came from the enhancements they were making with BASIC. Had Allen quit as MITS’ software director, which he almost certainly would have done if he and Gates no longer had a financial interest in BASIC, then there would have been no one to make the badly needed upgrades in the language. “In retrospect, said Curry, “it worked out in a way that I don’t think Ed is very happy about today, but at the time was the right decision based on what everybody knew.”

Gates decided the best thing he could do to stop people from stealing his software was to strike back publicly at the thieves. He asked Bunnell to publish a letter in the Altair newsletter,
Computer Notes.
Entitled “An Open Letter to Hobbyists, Gates noted that the most important thing inhibiting computer hobbyists was the lack of good software.

Almost a year ago, Paul Allen and myself, expecting the hobby market to expand, hired Monte Davidoff and developed Altair BASIC. Though the initial work took only two months, the three of us have spent most of the last year documenting, improving, and adding features to BASIC. Now we have 4K, 8K, EXTENDED, ROM and DISK BASIC. The value of the computer time we have used exceeds $40,000.”

Gates noted that while feedback from enthusiasts was strong, he
’
d noticed two things: “1) Most of these “users” never bought BASIC (less than 10 percent of all Altair owners have bought BASIC), and 2) The amount of royalties we have received from sales to hobbyists makes the time spent on Altair BASIC worth less than $2 an hour.”

He then accused hobbyists of stealing software programs. “Is this fair? One thing you don’t do by stealing software is get back at MITS for some problem you may have had. MITS doesn’t make money selling software. The royalties paid to us, the manual, the tape and the overhead make it a break-even-operation. One thing you can do is prevent good software from being written. Who can afford to do professional work for nothing? . . . The fact is, no one besides us has invested a lot of money in hobby software ... but there is very little incentive to make this software available to hobbyists.”

He went on to add that those who resell BASIC software “give hobbyists a bad name, and should be kicked out of any club meeting they show up at.

“I wo
uld appreciate letters from any
one who wants to pay up, or h
as a suggestion or comment.
Nothing would please me more than being able to hire ten programmers, and deluge the hobby market with good software.” The letter was published on February 3, 1976. Bunnell not only printed the diatribe in the Altair newsletter, but he made sure it ran in most of the major industry publications, including the newsletter of the Homebrew Computer Club.

Gates’ letter caused quite a stir. The Southern California Computer Society, which had been visited by the MITS-mobile in early 1975 and by now had several thousand members, threatened to sue him. “They were upset that Gates had called them thieves,” said Bunnell. “They were not all thieves . . . just most of them.” Only a handful of people who possessed pirated copies of BASIC sent Gates money as he had asked them to do in his letter. Some fired off their own angry letters in return. What was the difference between making copies of BASIC and taping music off the air rather than buying the recording artist’s music, some wanted to know? Others argued the altruistic position that BASIC belonged in the public domain, an argument that had some merit since Gates and Allen had created BASIC using the PDP-10 at Harvard, a computer funded by the Defense Advanced Research Projects Agency. In other words, these people argued, the computer time they had used was paid for with taxpayers’ money.

Regardless of what philosophical rationalizations hobbyists used to justify why they had not paid for BASIC, they did have one legitimate beef: MITS was partly to blame by engaging in a pricing policy that all but assured Altair owners would do anything possible to avoid paying for BASIC. No one wanted the memory boards MITS was turning out. But everyone wanted BASIC. So early on the company had decided to charge $500 for BASIC alone, about a, hundred dollars more than the Altair itself cost. But for only $150, a customer could get a memory board plus BASIC. Of course, word quickly got around that most of the boards didn’t work, and MITS knew this, which is why it priced BASIC out of reach of many hobbyists. MITS was forcing customers who wanted BASIC to buy the atrocious circuit boards.

Gates did not stop the antipiracy campaign with his letter. In late March, he spoke out on the piracy issue at the World Altair Computer Convention in Albuquerque in his first industry speech. The convention, the first ever for microcomputers, was the brainchild of Roberts and Bunnell. By then, MITS had annual sales of well over a million dollars and Roberts wanted to showcase the company with a convention that would bring together key people in the industry. Bunnell organized and promoted the convention, which was held March 26-28, 1976, in a hotel near the Albuquerque airport. Several hundred people came, including a few
uninvited
competitors who crashed the party. One
of these competitors was Proce
ssor Technology, a startup hardware firm producing 4K memory boards that could be used with the Altair. The company had wanted to set up a display

booth at the convention, but Roberts, who had become increasingly paranoid about competition, vetoed their request. The boards produced by Processor Technology not only worked, but were selling very well, much better than the faulty circuit boards designed by MITS. The only reason customers bought a MITS board at all was to get a copy of BASIC for the $150 package price.

When Roberts told Processor Technology founder Bob Marsh he could not have a display booth at the MITS-sponsored Altair convention, Marsh rented the hotel’s penthouse suite and posted hand-written signs in the hotel lobby directing convention goers to his company’s suite. Later, Roberts published his own diatribe in the Altair newsletter, describing Processor Technology and other companies who dared produce memory boards for his computer as “parasites.” In the end, he only damaged himself in the industry further.

The three-day convention, which included guided tours of MITS, was more of a conference than a trade show. There was no hardware on display, except in the penthouse suite. Most sessions were held in a large room in the hotel where attendees listened to speeches and talked about microcomputers. When Gates spoke, his speech was another belligerent attack on those hobbyists who he said were “ripping off” his software. At the time, Gates was fairly unknown in the industry. He was better known as the author of the scathing letter in
Computer Notes
than as the author of BASIC. Now twenty years old, he looked more like fourteen. His hair was uncombed and hung helter- skelter over his eyebrows and ears, and his thick, oversized glasses accentuated his childlike appearance. His high-pitched voice underscored his youthfulness. But Gates did have a certain charisma. His words crackled with the authority of someone much older and wiser. After his talk, people crowded around him to ask questions. Although he came off as brash and arrogant to many, Gates did have his supporters. “I was very much in sympathy with his attitude,” recalled Winkless, the
Personal

Computing
magazine editor. “How do you get your investment back?”

Roberts later asked his friend Eddie Curry to talk with Gates and persuade him to write a second letter, in the hopes of undoing some of the public relations damage. Roberts had been furious with Gates over the first letter because it had been written on the MITS stationery. “It looked like we were accusing all our customers of being crooks. ... I was very, very upset,” Roberts said.

Gates acknowledged that he had made a mistake by using MITS letterhead, and he agreed to work with Curry on a suitable second letter. “In retrospect, it really didn’t help anything to accuse people of being thieves,” said Curry. “All it did was to work against Bill. He didn’t feel he had done some terrible thing, but in the calm light of day he understood it was not the most prudent thing to have done.”

Gates’subsequent effort, entitled “A Second and Final Letter,” ran in the Altair newsletter in April:

“Since sending out my “Open Letter to Hobbyists” of February 3rd, I have had innumerable replies and an opportunity to speak directly with hobbyists, editors and MITS employees at MITS’ World Altair Computer Convention, March 26-28,” Gates wrote. “I was surprised at the wide coverage given the letter, and I hope it means that serious consideration is being given to the issue of the future of software development and distribution for the hobbyists.
...”

Gates then went on to say, “Unfortunately, some of the controversy raised by my letter focused upon me personally and even more inappropriately upon MITS. I am not a MITS employee and perhaps no one at MITS agrees with me absolutely, but I believe all were glad to see the issue I raised discussed. The three negative letters I received objected to the fact that

use were not obtained legitimately and not to issue a blanket indictment of computer hobbyists. On the contrary, I find the majority are intelligent and honest individuals who share my concern for the future of software development. .. . Perhaps the present dilemma has resulted from a failure by many to realize that neither Microsoft nor anyone else can develop extensive software without a reasonable return on the huge investment in time that is necessary.”

Gates ended by saying he considered the pirating matter closed. He predicted BASIC would become the foundation for the development of new and exciting application programs for microcomputers.

The BASIC that Gates and Allen had written in those eight frantic weeks at Harvard a year earlier had now spread all over the country, thanks in large measure to the very actions of hobbyists Gates had so bitterly denounced. BASIC
had
become a
de facto
standard in the young microcomputer industry. When new computer companies joined the revolution and needed a BASIC
language
, they came to Albuquerque and did business with Gates and Microsoft. And they came with pockets stuffed with money.

Part of what made Microsoft so successful during the company’s infancy was the team of programmers that Gates and Allen began to assemble in the spring of 1976. They became known as the Micro
ki
ds—high-IQ insomniacs who wanted to join the personal computer crusade, kids with a passion for computers who would drive themselves to the limits of their ability and endurance, pushing the outside of the software envelope.

Chris Larson, the first of the programmers, was still in high school and could only work summers. Richard Weiland came and went for a couple years before he left for good. Marc McDonald, however, was hired as Microsoft’s first permanent programmer, and thus became the first of the Microkids. McDonald, who arrived in Albuquerque in April, was a 1974 graduate of the Lakeside computer room. He and Gates had known each other since their days at C-Cubed trying to crash the PDP-10 computer system. Since Microsoft did not yet have its own offices, McDonald worked on a computer terminal in the apartment he shared with Allen, or he used one of the terminals at MITS. Not long after McDonald joined the team, Wei- land returned. He had left several months earlier after developing BASIC for the 6800 chip. Weiland, who moved in with Allen and McDonald, took on the role of Microsoft’s general manager. He would eventually be offered part of the company by Gates and Allen as an inducement to stay, but instead he decided to go to business school at Stanford University.

Two more programmers arrived in the fall of 1976: Steve Wood and Albert Chu. Wood had grown up in Seattle, but went to public schools and did not know Gates or Allen. He had been finishing his master’s degree at Stanford University in electrical engineering and looking for a job when he noticed a Microsoft recruitment poster in the placement office of the Stanford Linear Accelerator Center, where he worked as a research assistant. The center was one of the first facilities in the country to acquire a
microcomputer
lab; it had also become the regular meeting place for the Homebrew Computer Club.

Unlike the others already working for Microsoft, Wood was married. He was a year older than Allen and Weiland. After Wood interviewed for the job and returned home with an offer, he and his wife threw their belongings in a small U-Haul trailer and headed for Albuquerque. Microsoft had just rented its first offices on the eighth floor of a bank building near the airport. The address, Two Park Central Tower, would be Microsoft’s home for the next two and a half years.

Although Microsoft now had offices, it did not have its own computers. Instead, the company contracted to timeshare with the Albuquerque city schools, which used a DEC PDP-10 system. Microsoft’s programmers worked at so-called dumb terminals without a printer and each day someone had to go down to the school
administration
building and pick up computer printouts.

Gates and Allen had decided that FORTRAN was the next high-level language to develop as Microsoft expanded its product line; when Wood and Chu arrived they immediately went to work writing FORTRAN code for the 8080 microchip. At the time, FORTRAN was probably the second most popular computer language after BASIC. This strategy of anticipating the market and being the first out with a new product was a competitive edge Microsoft would hone in the years ahead. Allen, even more than Gates, had a knack for figuring out the direction of the industry three or four years down the road. The question constantly before the young company was whether Microsoft should invest in this product or that product, this set of software features or that set of features. Gates and Allen had to determine where the market was heading, what the technology would be like. “Microsoft definitely has a vision of where it wants to go. It’s very broad compared to some companies that have a narrow focus on only one product,” said Allen later. “Part of this goes back to when Bill and I were working on new languages in Albuquerque. Every time there was a new language we thought was going to be popular, we’d see that as another possible market for our software technology.”

In late 1976, Microsoft landed its two biggest customers and most lucrative accounts to date, National Cash Register and General Electric. Both wanted BASIC. General Electric only wanted to buy the source code. But NCR needed a digital cassette BASIC that would work with its 8080 file system. That job was given to Marc McDonald, who later developed what became known as Stand-alone Disk BASIC for NCR. It was one of Microsoft’s most successful and important software products to date, although Gates, rather than McDonald, would get the credit.

Allen remained as MITS software director until November, when he quit to work fulltime for Microsoft. The company was growing fast. Revenues for its first full year were more than $100,000. This was expected to triple in the next year. Things were working out just as Allen and Gates had imagined that summer at TRW in Vancouver when they talked about one day owning a software company.

In January of 1977, two months after Allen quit MITS, Gates dropped out of Harvard, this time for good. Sam Znaimer, who had roomed with Gates when they were freshmen, recalled that Gates told him shortly before he left school that he had come to Harvard because he was looking for people smarter than he was. “He told me he had not found them,” Znaimer said. “I think Bill just got bored.” But Andy Braiterman, who was rooming with Gates at the end, said Gates realized even he had limitations. “Bill was trying to do too much. He really did have a sense that things were out of control. He didn’t have time for both his business and school, not to mention trying to keep up with poker playing. He was really worn out by the end. ... If he had stayed in school, it would have been a tremendous mistake.”

Because Braiterman had entered Harvard as a sophomore, he graduated at the same time Gates dropped out. During their final semester together, they had talked often about computers and the future. “Bill was one of the first people I ever knew who really had this concept of computers being everywhere,” said Braiterman. “He saw that as being the future. I’m not sure he really saw things quite the way they turned out, though. He saw personal computers as being omnipresent' in people’s homes, as opposed to offices. But he was clearly thinking in the right direction. He also talked to me about the concept of everyone being able to discard all their books and paper materials and access everything they wanted to know by computer—to do all communication by computers.”

Back in Seattle, Gates’ parents took the news hard that he was dropping out of Harvard to work fulltime at Microsoft.