Justice Return Top Return
The Microsoft Decision Was Wrong!
By Jerry Pournelle, Byte.com
It is traditional to accuse anyone not bashing Microsoft of being in their pay, so let's get that out of the way now: Microsoft has never paid me a dime [1], nor do I own any of their stock, for the same reason that I don't own stock in any of the companies I write about. Now I can turn to the Microsoft decision. Judge Thomas Penfield Jackson has given his findings of fact: 98 pages worth in the single-spaced Times New Roman 12 point I printed it out in. Superficially, it seems comprehensive, and I have heard people who ought to know better say it is a brilliant analysis of the computer industry. Far from that, it is both foolish and a disaster.
Foolishness The foolishness comes from unfamiliarity with this industry. Jackson thoroughly misunderstands [2] what happened to OS/2 (paragraph 46), and gives as examples of harm to consumers, Microsoft's efforts to dominate, wait for it, the Channel Bar (see paragraphs 311 and following). On a more serious note, he is entirely ignorant of the real history of the operating system wars.
IBM failed with OS/2 because IBM did not attract third-party software developers. The honorable judge believes this was due to Microsoft's dominance of the market. Not so. It was because IBM made a number of disastrous errors.
For example, at a critical period, IBM considered the Software Developer's Kit a profit center item and sold it at about $600, while Microsoft was aggressively giving their SDK away free. At one Spring Comdex, IBM was proudly announcing you could get the OS/2 SDK at a reduced price -- about $500 -- during the show. By contrast, if you walked past the Microsoft booth and looked at all like someone who might be a programmer, Microsoft thrust a Windows SDK into your hands. I collected about 30 of them that week [3].
Perhaps that would not have been decisive, but at a crucial time in the OS/2 vs. Windows competition, IBM made another major error. Incidentally, that competition was begun by IBM, at a time when IBM was by far the larger company. The war was declared when a top IBM executive rudely walked out of a Microsoft-hosted media/financial analyst event dedicated to "OS/2 - Charting the Course For the Future." (Amusingly, the briefing charts and briefing books for that conference glorifying OS/2 -- there was almost nothing about Windows -- were generated in PowerPoint on a Mac because there was no Wintel machine capable of doing that: At that time, Apple owned the desktop publishing world.)
IBM's fatal errors were insufficient attention to CD-ROM and not recognizing the market potential of incorporating networking into the operating system. Apple had always had that built into their systems, but for Wintel machine users, the only way to make machines talk to each other was through third-party hardware and software. Novell dominated the desktop client/server world, and having a certificate as a Novell certified engineer was a sure ticket to job security. For the rest of us who only needed peer to peer file transfers, there was Artisoft LANtastic (hardware and software) and Traveling Software's LapLink. LapLink came with a four-headed blue cable for serial transfer and a yellow cable for parallel transfer. Neither was very fast, and it was often easier to transfer files by physically carrying floppies from one machine to the other. File transfer for small offices was painful.
Microsoft very clearly understood the value of peer to peer networking if it could be made easy enough for the average user. They brought out -- in cooperation with Intel -- Windows for Workgroups, which became Windows 3.11. This could be bought as software alone, but the usual way to get it was in a package with two Intel EtherExpress ISA cards. Installation of both hardware and software took about an hour (at least that's how long it took me), after which your network was established and each machine on it could see drives on all the other networked machines. That included CD-ROM drives.
Microsoft had long been a pioneer in promoting CD-ROM, and for years sponsored the only major conference dedicated to CD-ROM. In those days, Microsoft was a much smaller company than Phillips or PolyGram records, which sponsored the now forgotten CD-I system as much superior to CD-ROM. The big boys said CD-I would bury CD-ROM, and, apparently, IBM believed this since one of the big problems with OS/2 was getting it to believe in CD-ROM drives. Early editions of OS/2 made it sheer hell to install those. (In those days, 50 megabytes was a big hard drive, and many operating systems simply wouldn't handle a 600-Mbyte file system.) While OS/2 was having problems with CD-ROM, Microsoft was integrating them into their OS. The drives were expensive, and it was important to be able to access them through networks since few users had more than one CD-ROM drive.
At that time, there was a lot of competition in word processors, and the fact that Word wouldn't work in OS/2 wasn't particularly important; there were others just as good, and many considered WordPerfect better.
Then Microsoft came out with Bookshelf. This integrated a thesaurus, dictionary, "Bartlett's Quotations," and a small encyclopedia on one CD-ROM that could be accessed through Word. The spell checking was improved. Most industry observers considered Bookshelf a stunningly valuable innovation [4]. Word got an instant boost in popularity. Over time, Word became an important word processor -- and OS/2 couldn't run it, and IBM didn't do much to develop anything that competed with it. This wasn't lack of resources, it was IBM's business decision that the small computer was an entry-level system and serious users would soon outgrow these "toys" and graduate to "real machines" made by IBM. IBM didn't see Microsoft as a competitor for its core products.
In other words, the story told in this judgment is flat wrong: IBM didn't lose because of Microsoft's dominance; it lost because it didn't take the competition seriously. I am reminded of Aesop: The fox, chided for failing to catch a rabbit, said "I was only running for my dinner. The hare was running for his life." Microsoft, an unimportant company much smaller than IBM when this competition began, was running for its life. In doing so, it ended up eating IBM's dinner.
The applications development history is important, but it isn't given in the decision. It can be summed up in a remark Bill Gates made a few years ago: "In 1989, I personally went to all the applications developers and asked them to write applications for Microsoft Windows. They wouldn't do it. So I went to the Microsoft Applications Group, and they didn't have that option."
Paragraph 47 Paragraph 47 blames Apple's problems on Microsoft. Like most of this document, it ignores the real history of the industry. When the Mac first came out, had Apple gone for market share instead of quick profit, it would be the dominant platform today. Apple's business decisions dictated its small place in the market.
Has everyone forgotten that at one time the only spreadsheet program available was VisiCalc for the Apple II? That was a business dominance Apple chose not to exploit, but it's hardly Microsoft's fault that Apple let Lotus 1-2-3 steal that supremacy for PC boxes. Then for a while, Lotus dominated in that market. They lost their dominance through silly business decisions, including an insane copy protection scheme. Notice that Microsoft has only sparingly used copy protection schemes. Both Dvorak and I pointed out in the early 80s that copy protection was a quick way to sure death in the mass market. The point is that in both cases, Microsoft's competitors lost through bad decisions, not Microsoft's practices. Microsoft has never been all that clever in business. What they have done is avoid disastrous mistakes in their core product marketing. Microsoft doesn't so much win as avoid losing.
The conclusions in paragraph 47 don't follow from the premises, and the whole paragraph neglects the real history of how Apple lost its dominance. Apple had every opportunity to turn MacWrite into a winning word processor. They never did. When Microsoft Excel first came out, it ran only on the Mac. I recall the introduction at Tavern on the Green in New York's Central Park. Most of the computer press believed that Excel guaranteed the Mac dominance in business computing. Apple never exploited that edge.
This continued until well after the introduction of Windows, at a time when Apple was comparable in size to Microsoft and Apple was absolutely dominant in the field of desktop publishing and graphics processing. Apple's failure to win market share was due to Apple's bad decisions, not Microsoft's dominance.
Note, though, that Apple still exists, and Office 98 for the Mac was until recently at least as good an office suite as you could obtain on a Wintel platform. A G4 Mac, especially if coupled with a low-cost Linux box used as a communications server -- more on that later -- is more than adequate for most applications except games. Apple has lost its ownership of desktop publishing and graphics processing, but the Mac is certainly far from dead, and remains a viable alternative to the whole Wintel world -- as my readers often point out when I tell a story of a particularly painful "Micro$oft Windoze" experience.
Harming Consumers The real joke in this decision is the section on harm to consumers. Microsoft may in fact have become a de facto monopoly. It may have used some questionable practices to maintain that monopoly and, in one case -- the Stac affair -- behaved despicably [5]. However, before the awful majesty of the law can fall on a company, it must be shown to have harmed consumers. Harming competitors can result in civil lawsuits; but to make this a case of the People of the United States vs. Microsoft, there must be harm to consumers.
And there isn't any. The worst harm the document can establish is that they believe Windows 98 ought to cost about $30 less than it does. That may or may not be true, but to me, it would clearly be a disaster if the government, through courts, special masters, or a "Software Fair Practices Board" set software prices for the rest of us. If Microsoft prices their OS too high, I'll buy a Mac. Make it high enough, and Linux will take off like a rocket -- not that it's not doing that anyway. But don't let the government set prices.
The worst consumer harm the document alleges is suppression of innovation. It doesn't show any because that's all speculation and theory and there aren't any facts to support it. Oddly, this same document also charges that Microsoft is too innovative, bringing out version after version of its OS whether the new features are needed or not, and inducing applications programmers to make use of the new features whether they are wanted by consumers or not; all this in aid of increasing Microsoft revenue.
For instance, paragraph 61 says: "Microsoft has incentives to innovate aggressively despite its monopoly power. First, if there are innovations that will make Intel-compatible PC systems attractive to more consumers, and those consumers less sensitive to the price of Windows, the innovations will translate into increased profits for Microsoft." Well, duh. "Second, although Microsoft could significantly restrict its investment in innovation and still not face a viable alternative to Windows for several years, it can push the emergence of competition even further into the future by continuing to innovate aggressively." In other words, Microsoft has been too innovative -- although suppression of innovation is one of the consumer harms charged in later sections of the decision.
Many commentators loudly mourn the innovations that might have happened had Microsoft not suppressed them, but they are shy of naming them. None seem to see the innovations Microsoft has made to Windows, which now incorporates dozens of items we used to buy from third parties. These include calculators, text and programming editors, search functions, games, file viewers, audio recorders and players, networking, and, dare I say it, Web browsers. None of these are necessarily the best of their class, but most are adequate, and their inclusion does not harm consumers -- although it may well harm competitors.
Indeed, most of these findings appear to have been written by Microsoft competitors, who, having shot themselves in the feet and legs, now want government Medicare to help them recover. If they can't get that, then they want the government to shoot their competitor "to level the playing field."
I could go on, but it's pointless. Nearly every page of this decision indicates that it was written by lawyers unfamiliar with this industry or by partisan advocates who will use any stick to beat Microsoft and do not blush to berate Microsoft for suppressing innovation while being too aggressive in bringing out innovations. Contradictory findings of fact cannot both be correct.
However, American public law makes it very difficult to challenge findings of fact on appeal. This is going to have significant effects, one of which will be heavy pressure on Microsoft to settle the case and be done with it. Better to pay and get it over with than have this case using up executive time and energy.
If those terms include government regulation of the software industry -- and power to "protect consumers" was clearly one of the things the Department of Justice crowed about in their announcement of this decision -- then this is an unmitigated disaster for the rest of us. If Microsoft had had to go to Webster Hubbell to get permission to put networking in the Windows operating system, we would still be connecting parallel cables and using LapLink and the Ethernet chip set would still cost several hundred dollars for the chips alone. Imagine the consequences of going to Janet Reno for permission to distribute Internet Explorer free as opposed to charging a price "competitive with Netscape." Imagine a government agency trying to determine -- as this document purports to determine -- which browser is "better" and what "comparable" means, and how low -- not how high, but how low -- Microsoft would be allowed to price features like new versions of Internet Explorer.
A government marketing board to set prices and define "comparability" and "competitive nature" is always subject to political influence and manipulation. One reason for not giving government business power is to reduce the temptation on civil servants to sell influence. This is elementary political science.
The best this decision offers us is a small drop in the price of the operating system. Whether computer makers would pass this on to the rest of us or keep part for themselves, I leave as exercise for the reader. The cost of that theoretically lower price of our operating system is likely to be government intervention in the one industry that has sustained our long boom. The effects of bureaucratic control in the aerospace and defense industry are plain. This weekend in California, a judge decided it was unfair to lower the price of milk; this in implementation of a regulatory system that is supposed to protect milk consumers. Is that analogy too strained? I doubt it.
Fortunately, Gates is unlikely to cave in to any agreement that restricts his control of his company or his ability to demand aggressive research and development, indeed in putting new features into Windows whether we want them or not. That's his nature. He built the company by stuffing in features, often to the detriment of performance. He's unlikely to sign any agreement that stops him from doing that. The Microsoft philosophy has always been to put out innovations first, even if they are not well implemented, in the sure and certain hope that user complaints will identify problems Microsoft can fix and hardware improvements will fix the rest. You may recall my condemnation of Office 97 as "bloatware" because it needed 200 Mbytes of disc space. Today, you can't buy discs with less than 4 gigabytes, and the 200 Mbytes is trivial. Gates has always bet on Moore's Law, and he has always won, and those of us who forget it get egg on our faces.
Gates can hire the legal talent to tie this case in knots and keep anything from happening for years. By the time it gets to the Supreme Court, it will be heard by justices not yet appointed who will be called upon to decide issues so remote to what is then current in the industry as to take a historian to recall what the shouting is about. By then, Microsoft may or may not have the dominant position it has. If it does, it will be because Microsoft has kept up with the technology, taken account of competition such as Apple and Linux, and come out with products that consumers prefer to the competition. It will be because Microsoft is running scared, as it always has.
And I'll still be cursing because some new Windows 2020 feature has caused my system to crash, and I'll still be wishing there were a small, reliable, easy to use operating system -- and, who knows, maybe I'll have one, in which case Microsoft will be trying to figure out how to put out a lower-cost version they can market as almost as good. That's the way this industry works.
Notes 1) With one exception: I bought a copy of Encarta 99 at Fry's and sent in for the rebate offer that effectively reduced the price to the California sales tax. The rebate did come from Microsoft.
2) "Never ascribe to malice that which is adequately explained by incompetence" - Napoleon Bonaparte.
I have to say that the temptation to suspect malice -- in the sense of a desire to exert control over Microsoft simply for the sheer pleasure of the power involved -- is high.
3) At that same Comdex, Byte, which then did the official Comdex show awards, gave IBM OS/2 the Best of Show Award for technical excellence. As I was presenting the award, I had to find someone to accept it. It took me three hours to find an IBM executive willing to come to the awards ceremony to accept what was then the most prestigious show award in the industry. Most of their marketing people had never heard of Byte. Microsoft had been nominated for the award and Microsoft people showed up in droves and were not shy about expressing their disappointment at not winning.
4) I learned about Bookshelf from John Dvorak. At the time, we were on a panel in front of a large audience and John took full benefit of showing he had better sources than I did. Ah, well.
5) The amusing sequel to that story is that Stac's disk compression technology turned out to be technically excellent, but commercially worthless as disk drives became relentlessly larger. Microsoft eventually paid Stac in settlement far more than the company was worth at the time of the settlement. If Microsoft hadn't engaged in sharp practices and had to pay compensation, Stac would long before have been Chapter 7 bankrupt. Of course, that doesn't excuse Microsoft for bamboozling the Stac people.
Jerry Pournelle, Ph.D., is a science writer and Byte.com' senior contributing editor. Contact him at jerryp@jerrypournelle.com. Visit Jerry' Chaos Manor at www.jerrypournelle.com. Reader letters can be found at Jerry' letters page.