Bill Gates presents himself as the voice of competition. But you explain that Microsoft actually uses every means at its disposal to eliminate its competitors. Can you describe the methods you accuse this company of using more precisely?
Microsoft's public statements totally contradict the company's practices. On the one hand, Microsoft says, "our software is the best, because the public buys it." On the other hand, it uses a very sophisticated arsenal of techniques to prevent the public from choosing anything other than its products. This shows that the leaders of this company do not believe at all in the quality of its programs. When you look at Microsoft's conquests and practices, individually, this all looks very banal. It is a bit like a game of chess, where, if you only watched a few moves out of context, you would see a pawn being taken, a knight put in check, a rook advancing... nothing spectacular. But when you step back, and look at all of Microsoft's strategies in all the battles it has fought –and won– you start to see a chronicle of a monopolistic conquest. Let's look at the practices themselves. Microsoft uses several types of weapons to annihilate its competitors. The most common and best-known of these weapons are its commercial tactics, beginning with the inequitable contracts that it imposes on computer manufacturers.
What exactly are the relationships between Microsoft and the companies that sell computers, such as Compaq, Dell or Gateway?
Microsoft tries to force the large computer manufacturers to offer only its products. This practice dates back to the very beginnings of the company. In fact, up until 1995, Microsoft required that all distributors of IBM and IBM-compatible PCs pre-install its programs on every computer they sold: first this was MS-DOS, and later Windows. The type of confidential contract that it signed with these manufacturers involved a license fee "per processor". It was written in black and white that IBM or Compaq had to pay a set fee for every computer they manufactured and sold, whether or not the end user wanted to have DOS or Windows installed on the computer. We can see that it was in the distributor's best interest to install Windows rather than something else that would cost them more. As for the end user, there was no point in them asking to have DOS or Windows uninstalled, because, in any case, they had paid for it. This is called an abuse of dominant position, which is harmful to all of Microsoft's competitors. A customer who wanted to buy Digital Research's DR-DOS (which was later purchased by Novell, and finally by Caldera), for example, would still have to pay for MS-DOS.
So, the general public was "free" to choose a Microsoft product, or to pay twice for their operating system! This practice was criticized by the Justice Department, and also by the European Community. But nothing concrete came out of this. After years of legal procedures, a settlement was reached which resulted in the 1995 Consent Decree, where Microsoft agreed to... nothing at all. As we saw in chapter 1, not only did Microsoft have no fines to pay for its many years of illegal activities, but it was also able to get around these fetters by developing a new practice had exactly the same effects. Instead of forcing computer manufacturers to sign license contracts "per processor", they now had the choice between a license "per model" and a license "per copy".
In the first case, the manufacturer agreed to install Windows, that they paid very little for, on an entire product line (a given model of a computer), that the end user could not purchase without Windows. In the second case, the manufacturer only purchases licenses corresponding to the actual number of units ordered by its customers. But in this case they have to pay two to three times as much for each copy of Windows! On the surface, computer manufacturers have a choice and the Consent Decree was respected. But we should also note that the European version of this agreement (note 17) explicitly specifies that a license "by model" would be illegal if all of a manufacturer's models had to be included under this type of license. But, when it comes down to it, computer manufacturers systematically choose the first option, by model, for all of their computers sold to the general public. And end users still have no freedom of choice, exactly like before.
Even worse: if you wish to purchase a new computer, and you already have Windows (this is a typical situation for most companies who are renewing their computer equipment), these large manufacturers will not let you purchase new computers without Windows.
Do you mean that a user cannot buy a PC today without Windows?
This is actually impossible; unless you find a small computer store that will put together a computer for you piece by piece. But if that happens, you had better know how to open the hood in case you have any problems. What this means is that if you buy your computer in a large chain store, there is no way you can avoid Windows. You have the same lack of choice as a company if you are looking for a distributor to sign a maintenance contract where your computers will be repaired in less than 48 hours. A student at the Center for American Public Policy and Politics of UCLA, David Chun, made a survey in June 1998. He contacted the twelve leading computer manufacturers, including Gateway, Dell, Micron, IBM, Packard Bell, Hewlett-Packard, Toshiba, NEC, Sony, Unicent, Umax and Quantex and asked them the same simple questions:
Chun's conclusions are very clear: "Of the twelve OEMs (original equipment manufacturers) contacted, none would sell a computer without Microsoft Windows, and none would offer a discount if I returned Windows and asked for a refund"! (note 18) Most of these manufacturers explained to David Chun that "their contract with Microsoft required that they sell Windows with each computer". The funniest of all was that IBM, who created OS/2, an operating system which is one of Windows' competitors, requires that its customers purchase their computers with a license for Windows... even if they make it very clear that they only want to use IBM's OS/2!
Why don't computer manufacturers, who are powerful companies, rebel against Microsoft's "diktats"?
It is important to point out that, in recent years, computer manufacturers have not had very high profit margins. Most of the profits made in the computer business is made on software, because there is almost no manufacturing cost. Once a program has been made, with costs that may be relatively large, it can be duplicated on a CD-Rom for less than a dollar per copy, or even be downloaded over the Internet, and, in this case, it is the user who pays to download it. Hardware has fixed costs, and these costs lower only a small amount as the volume of sales increases. Profit margins on hardware sales are very small because of this. For this type of market, competitors fight it out with daggers drawn: manufacturers struggle to save a few dollars here and there on their price.
It is therefore obvious that no computer manufacturer would take the risk to be the only one to pay its licenses for Windows, Office or any other software, even ten or twenty dollars more than its competitors. Because Microsoft could very easily, in retaliation, refuse to sell "per model" licenses at reduced cost to manufacturers who are trying to be too independent. A manufacturer who would have to pay the "per copy" license on more than half its sales would probably go bankrupt!
These huge computer companies that look so strong, are really built on shaky foundations. They are merely puppets, and their strings are pulled by the real bosses in this industry: Microsoft, and in a lesser manner, Intel.
This is also why computer manufacturers accept to pay the huge cost for technical support by telephone, that should be the Microsoft's responsibility, as we have seen in chapter 2. But Microsoft does not twist the arms of just hardware manufacturers: it has a very good reputation for using its coercive power and aggressive marketing strategies on all the links of the distribution chain, down to the end user.
Can you give some examples of these aggressive marketing techniques?
Microsoft sometimes takes its customers hostage by forcing them to buy a product they don't want in order to get a program they do want. In Japan, for example, Microsoft's Excel spreadsheet program is very popular, but Japanese people do not like its word processor, Word, which is not very well adapted to Kanji characters. The problem was that, up until recently, Excel was only sold in its Office suite, which also includes Word. This means that Japanese users who wanted to purchase Excel also had to buy Word, even if they just threw it away (note 19). Microsoft only stopped this after the Japanese Federal Trade Commission began investigating these practices.
Another questionable Microsoft tactic is to use the "sheriff" of the profession –the Business Software Alliance (BSA)– to put pressure on companies who make illegal copies of software. Notice that I do not like and do not use the word "pirated", because of its connotations of blood-thirsty thieves that have nothing to do with the world of software. But listen to this incredible story: one fine day in 1995, representatives of the BSA discovered a certain number of irregularities at the Uruguayan telephone company Antel. This company had installed many more copies of Novell and Microsoft programs that they had officially purchased licenses for. Companies must buy one license for each computer, but they rarely do this, especially in developing countries. Lawyers from the BSA filed suit for $100,000, in damages. So far, this is a normal story. But in 1997, the suit was dropped, because of an "out-of-court settlement" between Microsoft (who was not the only plaintiff represented by the BSA) and Antel, according to the following agreement: the BSA would forget everything... as long as Antel replaced all its software (including the Novell programs; and Novell was one of the plaintiffs) by Microsoft products! This story was presented by Rachel Burstein in the January 1998 issue of the Mother Jones magazine. (note 20) Representatives from Microsoft and the BSA quickly denied the entire story. But if it was not true, then why did Lotus and Novell publicly announce that they would no longer use the BSA's "services" in Latin America? Burstein's article highlights something that is far from an isolated incident: the BSA is so closely tied to Microsoft that some of its competitors have left to join the Software Publishers Association. This organization, which is more influential than the BSA, has recently strongly criticized Microsoft (even though Microsoft is also a member of this association) (note 21).
Others in the computer industry also suspect Microsoft of threatening retaliation against some of its partners that have tried to be too independent. It seems that some of the players in the hardware business –especially card manufacturers like Xircom, but also other chip and computer manufacturers– have refused to give the technical specifications for their hardware to Linux developers, who would like to make sure that this operating system be compatible with as many hardware components as possible. This means that these companies are deliberately cutting themselves off from 8 million Linux users. Is this in their interest? Obviously not. So, the only logical explanation is that if they cooperated with Linux developers, they might have troubles with Microsoft. It would be very easy for Microsoft to no longer send the latest beta version of Windows or Windows NT to companies that release products compatible with Linux. None of this would be said explicitly, of course. But they would just answer: "Oh, you haven't received the latest Windows beta? The package must have gotten lost." After a few days, the company would understand, change two or three specifications of its product without telling Linux developers, and make sure that Microsoft is aware of it. And then, suddenly, as if by miracle, they would receive the beta software they had been waiting for the next day. You can find lots of stories like this on the Internet. But, obviously, none of this can be proven.
We have seen that Microsoft's commercial practices are, to you, both like steamrollers and nuclear deterrents. What about offensives that it has made against its direct competitors?
Microsoft's corporate culture is based on eliminating competitors. The most insidious and efficient way to kill off competing products is to use the "network effect" (which deals with the interoperability between different applications), described earlier, to export its Windows monopoly to all the other sectors of the computer industry. When a software publisher owns both the operating system (Windows) and the applications (Word, Excel, Explorer), it is technically possible for it to modify the operating system to ensure that competing products are either unstable or unusable, at the same time as it improves the performance of its own programs.
How does this work? We have already seen that Microsoft's programs can modify components of the computer's operating system to adapt it to their needs. For example, let's say you wish to install a copy of Microsoft Word. Not only does the program get copied to your hard disk, but the installation procedure also modifies some components of the operating system, especially shared libraries (DLLs), whose program code is used by several programs. These modifications would be impossible if the vital components of the system were locked. And, in a reasonable world, it should only be possible to touch these components very rarely, during major updates, or if problems need to be corrected. But in the world of Microsoft, any program can modify a few of Windows' DLLs. This is supposed to be done just to adapt Windows to its needs; but there are also, undoubtedly, other dubious reasons for this (note 22).
It is easy, for example, for these applications to verify the type of operating system, and to slow down or stop working if they find a competitor's product. This is exactly the nature of the lawsuit filed by Caldera against Microsoft (note 23). Caldera is the company that purchased the rights for the DR-DOS operating system, a cousin of the original QDOS, from Novell. According to Caldera, when Microsoft released Windows 3.1 (and this problem continues under Windows 95), a software layer which runs on top of DOS-type systems to make them more user-friendly, the program runs a test to find which operating system the computer is using. If the computer responds DR-DOS, Windows 3.1 refuses to run. It requires that MS-DOS be present! I hope that these accusations can soon be verified, now that the courts have ordered Microsoft to provide Caldera's lawyers with the source code for Windows 95 (note 24).
It is easy to imagine a similar but opposite maneuver between the operating system and other applications. I have noticed that if I install the Internet Explorer browser on a PC which already has Netscape Navigator installed on it, Netscape's program starts running much more slowly. The only plausible hypothesis is that when Internet Explorer is installed, it modifies some components of Windows, which, because of this modification, prevent Netscape Navigator from running correctly. I must stress the use of the word possible, because it is very difficult to prove this sort of action: the source code of all these programs is protected by copyright. In any case, it seems that this practice has been around for quite sometime: it was quite well known, in the 1980s, that the very popular spreadsheet program Lotus 1-2-3 ran very slowly under DOS. There was even a joke among computer specialists back then: "DOS is not done until Lotus 1-2-3 is undone". So, it is technically possible for Microsoft to sabotage its competitors' products on the sly when they run on its platform. And since a PC program that is not perfectly compatible with Windows is a dead product, this technique can be incredibly efficient. These are practices that remind one of the insidious properties of ivy, that is said to poison the roots of nearby plants. Try and put a thyme plant next to some ivy, and it will die in a few weeks. This is practically always the fate that awaits software publishers who think they can release programs that are better than Microsoft's and have them run under Windows.
And once Microsoft has prevented its competitors' programs from working, all it has to do is take over the market with its own software?
Exactly. Microsoft is even so powerful that all it needs to do is to announce a new product that competes with an existing program, or a new version of one of its programs, without releasing it immediately. This practice of announcing new programs that do not exist is very common in the computer industry, and is called vaporware. Microsoft is the champion of vaporware. Often this is just a bluff, the high-tech equivalent of "pie in the sky". This is a way of promising a chicken tomorrow instead of an egg today. And all you end up with is an egg tomorrow instead of a chicken today. Consumers decide to not purchase inexpensive products which have been tested and proven, because Microsoft promises them that tomorrow they will release an extraordinary program that does much better. Unfortunately, when your name is Microsoft, you get so much free publicity, and you sell to uneducated consumers, vaporware can prevent competitors from getting market share they might be able to achieve, and also gives Microsoft the time it needs to prepare its attacks.
The first well-known example of vaporware was in 1988 when Microsoft launched Word 3.0. Bill Gates presented a demo of a prototype that was full of bugs, and even crashed his computer. But once this demo was made, people stopped buying its competitor WordPerfect. This practice is still continuing: in the spring of 1998, Microsoft explained that it would soon release a beta version of Windows NT 5.0 for workstations and servers, that would contain all the advanced functions of Novell's server software, together with many additional functions. But, as of December 1998, this "miracle" was nowhere to be seen. This strategy is very clear: Microsoft is inciting customers to purchase NT 4.0, and wait for the upcoming improvements that will be made for the version 5.0, instead of choosing products which are available immediately, such as Novell's NDS, or Hewlett Packard or Sun servers, for example.
Microsoft's other controversial practice consists in bundling new programs with products where they already have a quasi-monopoly...
Yes, Microsoft uses its coercive power to impose its new programs, that are often of a lower quality than others existing on the market, together with products where they have a monopoly. Example: Windows 98 contains the PIM program Outlook that has made publishers of similar products furious. How can you convince users to purchase a competitor's product, when they already own Outlook, that is bundled with Windows 98, and that they have therefore already paid for with the operating system?
The most flagrant example of this practice, which is at the heart of the Justice Department's antitrust suit, is Microsoft's practices involving Internet Explorer. Even Bill Gates admitted that Internet Explorer started out as a very poor program (note 25). Christian Wildfeuer, a Microsoft executive quoted by Time magazine, recognized this in an e-mail message of February 1997: "It seems clear that it will be very hard to increase market share on the merits of Internet Explorer 4 alone. It will be more important to leverage the Operating System asset to make people use IE instead of Navigator."
To impose Internet Explorer, exactly like it previously tried to impose its MSN online service, Microsoft asked computer manufacturers to pre-install this program on their computers, at the same time as they installed Windows 95. Then, as Microsoft slowly shifted its priorities toward conquering the Internet, the next step was to integrate Internet Explorer into Windows 98: to plant its roots into the code of the operating system, so it would be very difficult to eliminate. Microsoft's spin, to the medias and the general public, was that Internet Explorer was a "plus" given by Microsoft. This clever move allowed Microsoft to impose Internet Explorer and take away market share from Netscape Navigator. At the same time it tried to look good to its customers, since it was giving them this browser as a "gift".
But why, as a consumer, should I refuse a free program, that simplifies my life as well, since I do not have Netscape Navigator?
Because, by doing so, you are helping Microsoft build a monopoly over information transmission, and this is against your best interests. This is a sort of Trojan horse which, in the long term, may completely eliminate your possibility of making a choice. But is it really a gift? Consumers must pay $109 to upgrade from Windows 95 to Windows 98 (or $209 for those who do not yet have Windows 95). But, when you buy a new computer, you pay for Windows with the computer. Microsoft now pretends that Internet Explorer is an integral part of Windows 98. So, saying that Internet Explorer, that is bundled with (or, more correctly, included with) Windows, is free, while the rest is your real cost, is merely creative accounting.
Netscape Navigator, which was renamed Netscape Communicator, is not only free, but it is now an open source program. This means that its source code has been put into the public domain (see chapter 5).
Let me show you how the idea of price is artificial in the world of Microsoft. There is no better example than that of the Windows NT operating system. Microsoft sells two versions of Windows NT: one, NT Workstation, is used on the client computer, the one that receives information from the server, and is sold for $319. The other, NT Server, is used by the machine that distributes data, the server, and is sold for $809. NT Server also contains additional programs used for servers such as Internet Information Server, that Microsoft claims are "free". But let us look more closely at what these CD-Roms actually contain. If you remove everything that is artificially "free", both programs have exactly the same code... except for a few bits. The only difference is an entry in the registry that contains the word "Workstation" or "Server", plus one other tiny bit, very well hidden, to make it difficult for people who wish to save $490 by changing the workstation program into the server program.
The conclusion is that if Microsoft is telling the truth, that is, if all the programs included in the NT Server package (IIS, etc.) are free, then these two tiny keys which take up a few bits cost $490. Once you realize this, it is hard to get rid of the terrible feeling of having been not only taken in but outright swindled (note 26).
But lets's go back to the browsers. It is important to understand that when you accept to use Internet Explorer (even if it's free), you are the one giving Microsoft an incredible gift. You will be statistically considered to be an additional user of Internet Explorer. And in the war for standards, market shares are worth their weight in gold, even if they do not bring immediate profits. In fact, once Internet Explorer has reached 80 or 90% of the Internet browser market, this will give Microsoft additional leverage so it can extend its control over the Internet. Microsoft can then first modify its programs slightly so that the server that works best with Internet Explorer just happens to be Microsoft Windows NT, using the Internet Information Server program. And according to some Net surfers, this may already be the case: IIS replies more quickly to a request from Internet Explorer than from other browsers.
Microsoft can then contact all the major content providers on the Internet –companies such as Time Warner and Disney– and tell them that they absolutely must use Windows NT servers. As of June 1998, Microsoft already had 22% of this market, versus 9% for Netscape, and 49% for the open source program Apache, according to a study by Netcraft. If Microsoft can control the browser, server and computer market, it will then be very easy for them to "Microsoften" the communication protocols used between these different elements of the chain. This, as we have seen, would lead to almost total control over information.
Can the effects of these practices, that you consider dangerous, be seen already?
People who find my argument excessive or caricatural need only look at what is happening on the Internet: some Web sites that belong to Microsoft are already off-limits to people using other browsers, such as Netscape Navigator, Lynx, OmniWeb or Opera. Example: Netscape Navigator users who tried to connect to the Internet Gaming Zone, one of the best online strategy game sites, shortly after Microsoft purchased it in 1996, could see a message stating that the site was unable to work with Netscape 3.0, and suggested that users download Internet Explorer instead. Things changed after users complained: now the page also has a link to download Netscape 4, but, alas, you still cannot play if you are not running Windows. Maybe this is why this site has been recently renamed the "MSN Gaming Zone"! (note 27)
Technical reasons cannot justify this type of sectarianism practiced by Microsoft sites: the Internet Gaming Zone, for example, worked fine with Netscape before Microsoft purchased the site, and it obviously required that they invest time and money to change the site so it would no longer work with Netscape.
Let me give you another example that occurred in France: the French government uses a program called Nabucco for its financial management. The program was designed so any remote client can access it using the open Telnet protocol, one of those thousand-and-one free components that are the foundation of the Internet. This means that a Macintosh, a Sun workstation or a PC running Linux can all be used, with a bit of effort, as workstations with this program. Unfortunately, the new version of Nabucco is going to be "integrated" into Windows, in such a way that only people using PCs under Windows can connect to it. Nothing required the French government to make a choice that is definitely not in its best interest.
This voluntary incompatibility of Microsoft's sites with its competitors' browsers (which, it is important to point out, are used by at least 45 million netsurfers!) leads one to think that Microsoft's Web activity exists more to reinforce its domination in the software industry than to build a new business. In any case, with 55% of the browser market, Microsoft has convinced a growing number of independent content providers to optimize their sites for Internet Explorer, and in some cases to make it impossible for other browsers to use it.
This is the case with the British store Tesco's Internet Superstore (note 28), which, under the pretext that it only uses "the most modern Internet technologies", forbids access to its site to browsers which are not compatible with Microsoft ActiveX or BVScript. What a foolish policy! What it comes down to is that Tesco has invested money to cut itself off from half of its potential customers. It would be absurd to open a new supermarket that would only accept, for example, customers wearing brown shoes!
I would like to make a more general remark here about the "terrorism" of statistics and market shares. When a program is included with Windows, each Windows user is automatically considered as a user of that program. These statistics are somewhat misleading. Look at my situation: since I have already purchased four computers with Windows pre-installed, computer industry statistics count me as four Windows users. Whereas the first thing I did as soon as I got each of these computers was uninstall Windows and replace it with NextStep or Linux!
With a bank account of $14 billion in cash, it seems that Microsoft can just buy whatever it wants, or whatever bothers it?
Yes, as soon as Microsoft spots something interesting on its radar screen, the company does everything necessary to ensure that this product does not get in its way. Whatever its motivations may be, Microsoft acquires, or invests in, about thirty technological companies every year around the world. These investments are in fields as varied as operating systems, application software, servers, multimedia and Internet standards, Java technology, electronic banking, new interactive electronic medias, video games, network access and cable. There is a Web site that has an impressive list of Microsoft's purchases (note 29), and you can also consult the "official" version, that is obviously less explicit, at Microsoft's site (note 30).
Microsoft has already eliminated small potential competitors by taking over their technology. There are cases where Microsoft has purely and simply stolen products from small competitors, by shamelessly copying their technology to integrate it into its solutions. This is what happened, for example, with Stac, the company that designed the hard disk compression software Stacker, who managed to prove this was the case and won its suit against Microsoft (note 31). Another example is TV Host, whose TV Guide was, according to its vicepresident Mike Jeffress, simply copied and integrated into Windows 98 (note 32).
In other situations, Microsoft actually purchases licenses to use technologies where it has no expertise, or sometimes it purchases the companies that have invented these technologies. These operations are called buy-ins. In this manner, Microsoft purchased Web TV, and overnight became involved in set-top box Internet access. Usually, when Microsoft buys technology from a smaller company, this company is just pushed out of the market. Two of Microsoft's recent "partners" have filed suit against them for this type of practice. The first is Spyglass, whose Mosaic program was used as the foundation for Internet Explorer. In the license agreement they originally signed, Spyglass was to receive a small percentage of royalties for Internet Explorer. But when Microsoft decided to give the browser away for free, and stopped paying royalties to Spyglass, they filed suit (note 33).
The most significant case is probably that of RealNetworks, the company that introduced the popular RealAudio standard for streaming audio over the Web. Since Microsoft's goal is to control all the standards used on the Internet, it tried to purchase RealNetworks. But its CEO, Rob Glaser, only wanted to sell 10% of its capital, in exchange for a license on part of its technology. This was a very bad idea: Microsoft used this knowledge to develop NetShow, a competitive product, that it is now distributing "for free". Microsoft then refused to purchase licenses for the new versions of RealAudio and RealVideo. And, according to Rob Glaser, Microsoft also modified its software to prevent RealAudio from operating correctly (note 34). And now that NetShow is around, Microsoft sold its interest in RealNetworks.
A similar complaint, alleging that Microsoft "tweaked" the code of Windows and Internet Explorer, to hamper the functionality of a competing product, is the one voiced by Avadis Tevanian of Apple during the Microsoft trial: QuickTime, a well-known multimedia and streaming technology developed by Apple, would, in some cases, not be able to function properly due to new "features" added in recent releases of Internet Explorer.
There are other situations where Microsoft did not purchase a specific technology, but actually just purchased market share that it was not able to conquer on its own: the 9 million subscribers to the free e-mail service Hotmail discovered, one day, that they had become part of the MSN Internet Start site. Demographics are strategic in the Internet portal war.
There are other operations, called buy-outs, that are purely negative techniques whose goal is to kill the technology or the product that is purchased, so that a competitor's platform no longer has an edge. One key example is the war between Microsoft and Java. Java is an open language developed by Sun Microsystems, that resolves the problems of differences in platforms. Using Java, people will no longer need a Windows computer to run interesting programs over the Internet. But for Java to be an efficient competitor, this language needs to have a software offer, such as word processors and spreadsheets, so there is a good development environment for companies who wish to create Java products. An innovative small company, Cooper & Peters, had designed an office suite called EyeOpener, that could have become the equivalent of Microsoft Office for any computer running Java, including of course computers that do not run Windows (note 35). Well, Microsoft bought out EyeOpener, shortly after the program was presented, officially to "accelerate the penetration of Microsoft's Java libraries". But now, this product seems to have disappeared. It is hard to not think that the only reason to purchase this company was to eliminate a subversive program.
When Microsoft cannot buy out products or companies, they buy out grey matter. This is what happened with Borland's team of developers, who had designed an excellent compiler for the Delphi programming language. Was Delphi a thorn in Microsoft's side, since its products are of a lesser quality? Well, Microsoft made a golden bridge for these programers to get them to come work on its own products. This is not formally illegal, but it is against standard practices in the profession (note 36). Since then, Borland has given up the minefield where Microsoft declared war, changed its name to Inprise, and is looking for another market niche.
These practices are certainly brutal, but why are they reprehensible? Aren't they really just a standard application of the basic rules of business: kill your competition, as long as it is still weak?
Buy-in and buy-out maneuvers are probably not illegal, except when there is blatant copying of patented intellectual property. But when a key player, who is as rich as Microsoft, does this at such a level, this represents a risk for the creativity of the industry. The only guarantee for progress is that innovation may exist unfettered. Especially when the universal technological censor has Microsoft's appalling level of scientific knowledge. Take Windows CE developers' conferences: you see the sad spectacle of serfs who are looking for crumbs on the table of their lord. They all basically ask: "What can I do that will not interest Microsoft right away?" Microsoft only needs to say, for example: "it would be a good idea to include handwriting-recognition in our next operating system; no one else should touch this domain." When it functions like this, it doesn't even need vaporware anymore.
The only "good" innovations are those that serve Microsoft's interests. But are we sure, to paraphrase something that was often said in the recent past about General Motors, that what is good for Microsoft is good for the world? Larry Ellison, the CEO of Oracle and Bill Gates' sworn enemy, describes this phenomenon as "a fight between Microsoft and Humanity... and Humanity is being left at the gate!"
Doesn't the acceptance of these practices come from the fact that the standardization around the "Wintel" platform is a driving force for the computer industry? And that thousands of companies are living off the PC market that is held by Microsoft? And don't all users, whether professionals or general public, benefit more from this than they suffer?
This argument, which is often presented by Microsoft's defenders, is quite simply ridiculous. Each time you want to defend a monopoly, you say it is better to have a bad standard than no standard at all. This is a very superficial analysis of the situation: as the standard holder's hegemony becomes established, the disadvantages largely outweigh any advantages that may have existed at the beginning. Because the company with a monopoly can then just maintain its technology and kill off any innovation. The false Microsoft standard, which is only the omnipresence of a brand that sells a varied line of products, runs the risk of actually inducing a much slower evolution of the industry than what would occur if there was no standard at all.
What is important is that it would be better to have a real standard –which must, to be called a standard, be open, documented and capable of guaranteeing the interoperability of different components– than a false standard, that is closed and modified every few minutes according to the whims of its exclusive owner. One of the best jokes told in Silicon Valley is a good example of this: how many Microsoft engineers does it take to change a light bulb? Zero, Bill Gates just has to declare that darkness has become a standard!
In any case, I find it very difficult to accept the idea that it is the Wintel standard that has opened the doors to this spectacular explosion of innovation. Don't forget that in the middle of the 1980s companies such as Amiga and Atari made powerful multimedia computers that were very innovative, while MS-DOS users were still typing "dir/w" on alphanumeric screens. And, at this time, the only relationship between a mouse and a computer was the chance of seeing the former nibble on the latter's power cable. During this period, the computer press explained that multimedia was a useless gadget, and something that was definitely not needed on computers in companies. Thanks to Microsoft, multimedia took off ten years late. Not to mention the Intel chips, with their terrible segmentation mechanism that worked with a no better Microsoft Basic limited to 64 kilobytes, that made it such that millions of computer users only used one tenth of the memory they had spent large amounts of money for, and this for years. No, the real cause of the recent boom in this industry is quite obviously the huge success of the Internet and the World Wide Web that Microsoft ignored for a long time. Books such as Barbarians Led by Bill Gates , written by two former Microsoft programers, explain this very well (note 37). Microsoft had nothing to do with the implementation of the open standards used for the World Wide Web that allowed for its spectacular evolution. On the contrary, these same standards used for the Web, together with the user-friendly interface that Web browsers present, are the very things that could allow users to avoid paying the "Windows tax". This would be one of the benefits of a universal platform, such as Java, since this language is complete enough that a program written in Java can run on any type of computer, with any type of operating system that has a Java Virtual Machine.
This is why Microsoft is trying both to "Microsoften" Java, which, by the way, has led to a lawsuit by Sun, and trying to control Internet standards. (At the time of publication of this book, Sun had just won a preliminary injunction against Microsoft, see note 38) Unfortunately, this company may succeed, because what gives the Internet its strength is also its weakness; this network, that operates using open standards, is not really controlled by anyone. This means that there is no one large player, no huge company that can effectively defend it from hegemonic appetites.
And what about governments?
Governments do not seem to have understood the real stakes of this fight. Even worse, they apparently do not realize that they have an important role to play in the future of a technology which is so promising for all of us, but could also bring many dangers if it is corrupted and used by special interests. I truly believe that the real reaction can only come from the general public.