Reluctant Ahab

As with the IBM PC, many myths surround the rise of Microsoft to high tech's position of paramount leader. The seminal myth hearkens back to the company's anointment as the supplier of the OS for the PC, the single greatest coup in the history of business. The popular story (backed up by such sources as The Pirates of Silicon Valley, an interesting and well-acted film that does a complete disservice to the cause of truth) is that IBM intended to use Digital Research's buff new 16-bit operating system, CP/M-86, for its new PC.

Kildall, through a series of misunderstandings and miscommunica-tions that to this day are the stuff of legend, refused to talk to IBM's representatives. IBM then turned to Microsoft for its OS. Despite that Microsoft had no such product, the company bamboozled IBM into agreeing to buy a nonexistent product and then turned around and scarfed up Quick and Dirty Operating System (QDOS) from a small computer company, Seattle Computer Products. QDOS was written by Tim Paterson to support an 80866 prototyping board that the company was selling to software developers.

The reality is a bit different. In 1981, the industry's biggest fish first swam up to Microsoft, not Digital Research, in search of both computer languages and an OS for the PC. At the initial meetings, Gates candidly informed IBM that Microsoft had no OS to sell. At the time, Microsoft made most of its money from the sale of languages, particularly BASIC. Microsoft was overjoyed at the chance to sell its products to IBM, but it suggested that for an OS, IBM representatives should contact Kildall and Digital Research to talk about CP/M-86. Dutifully, the Big Blue Whale traveled south to California to meet with Kildall, who didn't think the initial conference important enough to attend and allowed his wife, a vice president at the firm, to conduct the opening ceremonies. There was an argument about signing a confidentiality letter, neither group found much to like about the other, and IBM left Digital Research without even a preliminary agreement to talk about CP/M-86.

The IBM contingent then asked Gates to talk to Kildall and persuade him to be more receptive to their overtures, but even this led nowhere. IBM was "the establishment," and many programmers brought up in the 1960s and 1970s regarded the company with a certain disdain. IBM was big, bureaucratic, and its machines, although beloved by big businesses everywhere, weren't accessible to hackers and hobbyists. Kildall, in tune with the spirit of the Altair and doing a nice business with CP/M, wasn't overly impressed by IBM and saw no need to kowtow. It was only after these initial rebuffs that Microsoft stepped into the OS situation and agreed to provide one for an IBM becoming increasingly nervous about meeting its ship dates for the IBM PC. After all, if IBM couldn't ship its PC, it wouldn't need Microsoft's BASIC. Fortunately for everyone concerned, except Kildall, the serendipitous existence of QDOS made it possible for Microsoft to deliver on its promise.

6 Ads for this board appeared in BYTE magazine.

Over the years, rivers of ink have been spilled bemoaning Kildall's rotten luck and the cruelty of an unfair world, but most of the hand-wringing seems misplaced. Kildall had been placed in the unique position of having had the largest of large blue whales swim up to his door, beach itself in his office, roll over on its belly, and point to the spot where the harpoon should be placed, and he had refused the shot. A fair person can hardly blame Gates for stepping up to the prow for his own throw at the great beast, and unlike Kildall, Gates's aim was true. The contract he negotiated with IBM turned out to be Microsoft's first step on the road to industry supremacy.

Yet, even this happy turn of events for Microsoft was not all it seemed. It would take further blundering on the part of Digital Research before the company was truly and finally fish food.

When the PC first shipped, PC DOS was indeed the operating system of record. But this didn't mean as much as it seems. DOS wasn't preloaded on the IBM PC. The unit had no hard disk, and DOS wasn't stuffed on a chip in your PC. You booted your OS from a floppy every time you turned on the machine.7 Nor was DOS bundled into a purchase of a PC. It came in a separate box and you paid for it separately. During the initial rollout, IBM had put no extensive marketing push behind DOS; all of its emphasis was on the PC. But as the system's sales momentum built, IBM did, however, make much of the fact that no less than three OSs were available for the PC: PC DOS; the UCSD p-System, which was really a development system for programmers interested in developing "write once, run anywhere" software (no, Java wasn't the first time someone had that bright idea); and . . . CP/M-86.

CP/M-86? How did that get in there? Hadn't Kildall blown it in those legendary meetings and phone calls?

Well, not completely. As the enormity of what he had done began to sink in, Kildall took a close look at a copy of the forthcoming IBM OS and noticed that, by golly, it sure looked a lot like CP/M. And that didn't seem fair at all. There quickly ensued some legal harpoon rattling, a

7 As a salesman at Macy's "professional" computer store, I attempted many times to sell prospective customers CP/M-86 instead of IBM DOS. Having worked with CP/M in its 8-bit incarnation, I knew it was a superior choice, but the product's pricing made it almost impossible to sell.

quick visit to Boca Raton, and voilĂ ! CP/M-86 was now an officially supported IBM OS that shipped in an IBM box and was available directly from IBM.

CP/M-86 was late to market, but despite this, shortage of software support would be no problem for the still feisty DOS competitor. The press and most technical gurus regarded CP/M-86 as superior to DOS, and publishers of older CP/M software hadn't found it hard to port their applications to the new OS. For example, MicroPro, at the time the world's largest microcomputer software company, had ported WordStar, the industry's leading word processor, as well as most of its other business packages, to CP/M-86. Ashton-Tate, publisher of the best-selling dBASE II, had a CP/M-86 version of the program. Other companies produced spreadsheets, games, utilities, and other products in anticipation that CP/M-86 would quickly sweep DOS from the market.

Nothing of the sort happened. Compounding his initial errors, Kildall had made a fundamental pricing mistake with CP/M-86. Upon the introduction of the IBM PC, the cost of PC DOS had been set at $40.00 (when anyone actually paid for it; the product was heavily pirated). This decision by IBM had reset market expectations as to what an OS for a microcomputer should cost (a reality the company would find out 6 years later applied to itself during the introduction of OS/2). CP/M-86 upon its release cost $240.00, a price close to that paid by purchasers of the 8-bit CP/M. The huge disparity in price made it almost impossible to sell CP/M-86 to retail purchasers, and the OS began to wither almost immediately.

Years later Kildall would claim that IBM had decided on the price difference between the two operating systems. There is good reason to question this statement. At CP/M East8 in the autumn of 1983, the last

8 This show was the first I attended as a MicroPro employee. I spent most of my time demoing InfoStar for CP/M-86 and was one of the people who tracked Gary Kildall down to discuss the pricing issue destroying CP/M-86. This event became legendary among MicroPro employees for what became known inside the company as the "Schmuck 'n' Shark" riot. MicroPro rented out the New England Aquarium for an evening and handed out about 700 tickets for a surf-and-turf dinner with an open bar. Approximately 3,000 people crashed the event, and a few drunken revelers had to be forcibly restrained from doffing their clothes and diving into the shark tank for a swim. A radical contingent from MicroPro was in favor of allowing the partygoers to jump in with the sharks and watching what happened, but the more conservative faction prevailed.

major trade show ever held dedicated to promoting Kildall's brainchild, a group of people9 from various companies publishing CP/M-86 software cornered Kildall on the busy show floor to discuss pricing and the OS's future. In the impromptu discussion that followed, Kildall was repeatedly implored to adjust CP/M-86's price so that it could compete with PC DOS and warned that failure to do so would kill the product. Kildall was polite, pleasant, and adamant that CP/M-86 was "priced just right." "The market understands the difference between a toy OS and a professional product," he proclaimed before disappearing into the show crowd.

CP/M-86 was effectively defunct by the end of 1984.

A despondent Digital Research would try to make a comeback with GEM, a Macintosh look-alike shell for DOS that enjoyed a brief measure of success before it was crushed by a litigious Apple. In 1987, Digital Research obtained a more solid measure of revenge when it released DR DOS, a "clone" of MS-DOS (though who was the actual clone is a legitimate matter of dispute). Though no major PC vendor ever picked up the product, for a couple of years Digital Research did a brisk business selling DR DOS to second- and third-tier manufacturers while simultaneously giving Microsoft and Gates minor fits.

The fun came to an end when Microsoft struck back by placing messages in beta versions of Windows 3.1 that warned users of possible "problems" that might occur if you used DR DOS with Windows.10 This was all nonsense; DR DOS worked fine with Windows 3.1 and public pressure eventually forced Microsoft to back away from this unsavory tactic, but in the interim a great deal of marketing damage had been done.

More significant were the changes Microsoft made in its licensing agreements that made it difficult to buy MS-DOS without also purchasing Windows and tied discounts to exclusive purchases of Microsoft products. These were tough tactics, and they would come back to haunt Microsoft during its defense against the U.S. government's charges of predatory and monopolistic business practices. But even if Microsoft had been a kinder, gentler opponent, unless a major player such as IBM

9 I was one of those individuals.

10 Wendy Goldman Rohm, The Microsoft File: The Secret Case Against Bill Gates (New York: Times Business Books, 1998). I was a DR DOS user and personally experienced this situation.

had intervened, DR DOS could never have amounted to more than a minor presence in a market moving inexorably to a GUI model of computing a la the Macintosh.

Attack of the Clones

The second great myth surrounding Microsoft's rise to power is that the original DOS contract with IBM immediately provided the company with a massive and unfair advantage over its competition. Again, the truth is somewhat different. Over time, Microsoft's DOS contract did prove to be a cash cow of legendary proportions, but it took idiocy of monumental proportions on the part of IBM, Apple, and other industry players to transform Microsoft's good deal for its quick-and-dirty DOS into the industry's shiniest gold mine.

From a financial standpoint, the original DOS deal put a nice bit of up-front cash in Microsoft's pocket and provided the company with a lucrative revenue stream from royalties on sales of PC DOS. But far more significant was that the contract gave Microsoft the right to resell DOS to other companies, something the company promptly began to do under the rubric of MS-DOS.

This, however, didn't turn out to be as lucrative a business as Microsoft had initially thought. Many of the first "clones" of the IBM PC weren't true clones; rather, they tried to improve on the PC's design. These machines, from companies such as DEC, Otrona,11 RadioShack, Victor, Texas Instruments, Hyperion, and many others that have vanished into obscurity, were collectively known as "MS-DOS clones." Some offered better hard disk support, sported different keyboard layouts, and provided better graphics capabilities, an area in which the original PC was considered weak.

It was the issue of graphics support and compatibility that proved lethal to the MS-DOS machines. Developers for the IBM PC had discovered something about MS-/PC DOS early on: It displayed graphics slowly. To solve the problem, software developers quickly learned to bypass the OS and directly access IBM's graphics hardware to improve screen performance.

11 I had use of an Otrona, a "light" portable (around 20 pounds) for about a year in the early 1980s. The unit was retrofitted with a compatibility board that let it run most IBM PC software.

The appearance of the MS-DOS clones presented software developers with a dilemma. Should they build customized versions of their software to support these new computers, most of which didn't possess substantial market share? Or should they hedge their bets and use MS-DOS to handle screen updating? Most hedged, and buyers of MS-DOS clones soon got used to watching their software work veeery slooowly on their systems while IBM PC users enjoyed word processors and spreadsheets that seemed to snap to attention. Interest in the MS-DOS clones, which had been high, was soon replaced by skepticism, and then derision. No one wanted an MS-DOS clone; everyone wanted an IBM PC or a true PC compatible, one able to run IBM PC software out of the box. The MS-DOS clone market quickly collapsed, and Microsoft's advantage seemed less significant than it had been.

But fortunately for Gates and company, IBM had unleashed the Silicon Beast. As quickly as the MS-DOS clones withered from the market, they were replaced by hordes of IBM compatibles able to work with PC displays and graphics without any need for machine-specific customization. All the new generation of clones required to go to market was an MS-DOS license. Microsoft's IBM deal started to turn golden indeed.

Microsoft's good fortune was compounded by a decade of fumbling stupidity on the part of IBM as it sought a replacement for MS-DOS. First to flop was TopView, a clunky, multitasking, character-based pseudo-OS released in 1985 just as Apple's Macintosh was educating the market on the benefits of a GUI. Next to fail was something called CP-DOS (one of its many names), an abortive attempt to create an OS that took full advantage of the IBM AT's 80286 chip. Along the way, IBM continued to break Kildall's heart with flirtations over different versions of CP/M-86 that never lead to consummation. OS/2 has earned its own inglorious chapter in this book. In the mid-1990s, after its storied divorce from Microsoft, IBM even attempted to sell its own version of DOS in the retail and original equipment manufacturer (OEM) markets and did as well with it as it had with OS/2.

But IBM would never succeed in developing a successor to DOS. Apple would never follow its own early example with the Apple II and liberate the Macintosh OS from its sterile preserve to grow and flourish in an open environment. Digital Research would eventually fade away, unable to ever recover from its early missteps.

Over the next 20 years, Microsoft would make the most of its competitors' mistakes and stupidity as it slowly leveraged its advantage in desktop OSs into absolute control over what would prove to be high tech's most strategic terrain. Using the generically named Windows as its base, it would slowly branch out to take control of the business applications market and then move from there to a position of preeminence in web technologies such as browsers. Like its competitors, Microsoft wouldn't always be completely ethical or nice in the way it did business. But unlike them, Microsoft would consistently avoid making stupid mistakes again and again and triumph from this ability.

Ebay Cash Insider

Ebay Cash Insider

I refined my information, added notes as things came up and took out old or outdated information. Each time I discovered a new little tip or trick as I was making money with EBay, I added that information. Today, I bring home anywhere from 500 to 2000 each week with EBay. Thats not too shabby for someone who was just a year ago afraid to even purchase something from the site.

Get My Free Ebook


Responses

  • GERDA
    What is ahab in marketing?
    7 years ago

Post a comment