If we follow your entire argument, it seems as though there is no way out, Do consumers still have any choice? Are there any alternative solutions, especially in Europe where people feel guilty about being technologically "behind"?
Solutions do exist, and they are very interesting, economically, technologically and socially, but they are not very well known. Before presenting some of these solutions, let me just point out that this being "behind" may actually turn out to be an advantage for Europe. Are we under-equipped? This is actually good: it means that we have the possibility to make important choices, and we still have time to avoid some of the traps that certain companies have set for us. "Missing a train" is actually quite fortunate, if that train derails! We can still choose to give our children access to efficient open source computer systems. In addition, this choice can help create jobs. Such a choice was recently made in Mexico, where 140,000 schools in the country will be equipped with computers running Linux.
To understand exactly what these alternatives are, you need to have a precise idea of the different possibilities available when you purchase a computer. There are four major elements in a computer system: hardware, the operating system, applications, and standards used for file formats and communication protocols. Unlike what many people believe, these four elements are actually quite independent. This means that the choice of one component does not necessarily determine your choices for all the others. Let us look more closely at this.
Even Microsoft Word now allows you to save files in HTML, the language used on the World Wide Web. With a few mouse clicks, you can send your colleagues files in HTML that anyone can read, and not in Klingon version x.y, which requires that this person has also decided to pay the Microsoft tax.
Public institutions are required to supply public information in a format that is accessible to all. Why should one have to buy a given proprietary program to read a legal text or a tender offer? Let me repeat, these formats do exist, and can present information correctly, and are even accessible to people who do use Word. In spite of letters of complaint from many scientists (including myself) one frequently finds on official European Union sites –such as http://www.cordis.lu– documents that can only be read by the most recent version of Microsoft Word. This is totally illogical, and not in anyone's best interest.
This point is probably the most unknown and yet the most important aspect of this issue, because it is precisely through the elimination of open standards and the introduction of proprietary standards that Microsoft is trying to replace the many open source and often free applications used at the heart of the Internet by its own products.
You, like many scholars, believe in open source software. What exactly is open source software?
Open source software is software supplied with its source code, which is the computer code used to create it, as well as all information needed to maintain it. You need to know that software programs, just like symphonies, are written in a sort of computer score, called source code. These programs are then, as for compact disks, distributed in the form of an "executable" version –a recording– of this score. Commercial publishers such as Microsoft sell executable versions of Windows, but are very careful to not reveal their source code. This comparison is not really perfect, however: if you buy a CD containing a recording of some music by Bach, it is possible to recreate the score after listening to the music, whereas with the complex software used today, this is almost impossible.
Computer programmers all around the world have designed open source software with an idea of sharing, and these programs are considered part of the world's collective property. This means they can be freely modified and redistributed, as long as they remain collective property. Anyone can make improvements to an open source program, as long as their changes are added to the package. But this does not mean that these programs are "in the public domain": to prevent that open source programs be taken over by unscrupulous companies, who would then sell them without their source code, specific licenses such as the GNU Public License (GPL), or the Berkeley Public License were created. These licenses establish the intellectual property of the program, along with the conditions for their distribution: these rules are the exact opposite of those existing for Windows. Everyone has the right to modify the program and distribute it in any way they wish, as long as they too include the source code, and distribute it under the same conditions.
There is a certain confusion sometimes around this type of software. Two terms are used to describe it: open source software and free software. Both are the same, but open source stresses the fact that the source code is available, and free that the software is freely available. In this case, don't think free as in free beer, but rather free speech. On top of that, open source software is not necessarily free. And free software is not necessarily open sourced.
Let me give you some examples to clear this up: most programs called freeware (and not free software) that are found on many Web sites are free, but they are certainly not open source software, because the source code necessary to look under the hood and take the engine apart is not provided. In this category of freeware you will find such examples as Adobe's Acrobat Reader, and Internet Explorer (that is, if we pretend the creative accounting that we previously presented does not exist).
Linux, however, is free, open source software, because its source code is given to you freely, and if you download it from the Web it does not cost anything. But you can also buy commercial "distributions" of Linux (such as SlackWare, RedHat, SuSE, OpenLinux etc.). In this case you pay the added value that corresponds to its commercialization –pressing CDRoms, distribution to stores– and most often additional services: installation, technical support or development of specific solutions. Once you have purchased a copy, however, there is no BSA to check on you to make sure that you have paid for as many licenses as the number of copies you have installed. My Linux CD, that cost a few dollars, has already been used legally to install the software on more than twenty computers.
What is the advantage of this open source software?
Open source software has many advantages for users. It generally runs better and is more stable than its commercial equivalents, because the possibility of having access to the source code makes it much easier to correct errors and to develop specific solutions. It is available for limited cost or nothing at all, and can be freely and legally copied on an unlimited number of computers. Open source software is generally designed more economically, and can run on older machines such as 486 PCs, and sometimes even on 386s, that the Windows-Intel cartel claims are obsolete. Unlike Windows or Mac OS, operating systems like Linux are generally resistant to the type of errors that beginners may make, because only authorized users can access the system's vital components. Even better: several users can share one computer, and each one has their own work space on the computer's hard disk and their own specific permissions.
Finally, and above all, access to the source code of these programs is the only true guarantee of permanence and independence of users from software publishers. Open source software is part of a larger idea of open source computing that ensures the interoperability of products among each other by the publication of the technical interfaces for each program. An open source program is a bit like a good car engine that is easy to work on, that you can repair or soup up if you need to. On the contrary, Microsoft's engine is double-locked, and only Microsoft "mechanics" can get under the hood. What is more, even if this engine does run correctly, you will have to replace it entirely at a high cost each time the Microsoft "gas station" changes the formula of its gas!
Don't forget that reconstructing the score from the music is almost impossible with software, which means that if a commercial company loses its accidentally score, that was jealously kept in a safe, during a natural catastrophe (an earthquake, an explosion) or a "virtual" catastrophe (a "blue screen of death" a bit more serious than usual), no new music will come out of this company! However, open source software is copied millions of times around the world, and each time with its score: there is nothing more resistant to catastrophes, be they natural, virtual or commercial ones such as buy outs.
But how did this community of volunteer developers get together? And what are the main programs that they have created?
The idea of open source software is very old. But if I had to choose a specific event that marked a turning point in its history, it would, without doubt, be Richard Stallman's initiative in the 1980s. Stallman was a researcher at the Massachusetts Institute of Technology (MIT) in Boston, and he had written Emacs, one of the most sophisticated text editors available, that is very popular in academia. Stallman is one of the founders of the Free Software Foundation, whose purpose was to produce an entirely free operating system called GNU, which is a recursive acronym for GNU is Not Unix (note 47). This foundation also created the GNU Public License, giving legal protection to this model of software distribution. GNU –whose logo is a gnu– set up the foundations that made it possible for a young Finn to create the Linux operating system, which is currently the most well known open source program.
The story of the birth of Linux is a good example of the spectacular role the Internet has played since the beginning of the decade. It has been used as a space for collective design, as a conduit for transmitting software, and as a tool for providing technical support. Let's go back to 1991: Linus Torvalds, a student at the University of Helsinki, had just received what was then the cutting edge of personal computers from his parents –a 386. But the two operating systems he could run on this computer, DOS and Windows 3, had neither the level of performance nor the level of security of the more expensive Unix that he used on the powerful computers at his university. So, this brilliant and enthusiastic twenty-one year old computer programmer used the GNU tools to write Linux, a compound made up of the words Linus and Unix (note 48).
As his project progressed, Linus Torvalds made his source code available online, so it could be finished and touched up by the best programmers in the world. This free program, which is available today for PCs, Macintoshes and other types of computers, quickly earned a strong reputation: Linux has flown on the American space shuttle, was used to create the special effects for the movie Titanic, and handles mail distribution for the US Post Office. It is also used for a great deal of industrial projects (note 49), and it is even used by the US Navy in San Diego, where it seems to work much better than Windows NT, that, as we have seen, crashed on boats that were at sea. Linux already counts 8 million users, and this figure doubles every year.
Before Linux, most computer scientists were convinced that writing a quality program required a private centralized approach. They thought, to use Eric Raymond's expression (note 50), that to design a program as complex as an operating system the only model that worked was the "cathedral", with an authoritarian architecture and a small group of docile programmers. But along came a Finnish kid who organized a bunch of computer hippies over the Internet, and proved that the wild confused "bazaar" model was superior.
The exemplary story of Linux should not, however, hide the vast amount of open source software available. The Apache Web server (50% of the world market), Netscape Navigator, the Sendmail e-mail management program, and the Internet Bind name server, to name just the best known such programs, are all leaders in their respective niches. More generally, open source software is used for the very foundation of the Internet. Without these programs, the network would not have seen the growth that it has experienced in recent years. And if these programs were eliminated, it would simply stop working.
Open source software seems to have staked out its place in the United States and in the north of Europe. What about elsewhere?
All of Europe, and France especially, have made important contributions to the development of open source software, even though the people active in these movements are not very present in the media. A few examples will show that France has played a key role in open source software: the National Research Institute for Computers and Automation (INRIA) has developed more than 20 open source programs; the ext2 file system now used in all versions of Linux was developed by a professor at the Paris VI University; the Linux "Posix thread" library was written by a brilliant researcher from the INRIA in his free time; the multilingual extensions for the Apache Web server were written by a student at the Ecole normale supérieure; and some video card drivers were written by other French university professors.
If there is one place that France and the other European Union countries are really behind, it is in officially recognizing the magnitude of this phenomenon. France is only now beginning to open to these alternative ideas. A small group of French and French Canadian researchers and professors have created the French Language Linux Users' Association, the AFUL, that I am a member of (note 51).
The low cost and total control over open source software seem to make it an ideal solution for schools. But many people seem to be proposing that Windows NT be used as the network infrastructure for schools and universities...
We have already seen that solutions based on the Windows-Intel platform, expensive both for software and hardware, do not really offer sufficient levels of security or performance, and I am not only talking about Windows 95 and 98, but especially Windows NT. Unfortunately, Microsoft's powerful marketing force has managed to convince many decision-makers to ignore these negative aspects of its programs. But many international analysts, such as the Gartner Group, the Standish Group or the Aberdeen Group (note 52) are starting to publish studies which openly contradict Microsoft's arguments, and present the tremendous latent risks and costs which could result from replacing Unix servers by Windows NT servers (note 53).
Let us look at the example of an e-mail server: the Sendmail program, a free, open source program, has been used on the Internet for decades, but Microsoft's solution, Exchange Server to manage e-mail, costs $5,000 (for fifty clients), and is not efficient with large numbers of clients. This means that if you have five million users and not fifty, server performance plunges, which can be seen, paradoxically, with Hotmail. This Web-based service offers free e-mail addresses to more than 30 million users was purchased by Microsoft in December 1997. Well, Microsoft wanted to require that Hotmail use Windows NT to run their servers. Up until then, they were using a mixture of Sun Solaris and FreeBSD as their operating system, and Apache 1.2.1 as the server software. But it turned out that managing the then 9 million users was much too difficult for Windows NT and Hotmail had to reinstall Solaris (note 54)!
Another small example: Windows NT is not entirely compatible with Internet standards used for remote maintenance. Let me emphasize that Unix workstations (whether they are using Linux, SCO, Solaris, IRIX, AIX, HPUX or any other flavor of Unix), include the necessary tools to manage a server without being physically in front of its screen, using protocols such as telnet and rsh, and extremely flexible command languages. The server can even be used, if it is powerful enough, and does not have a high workload at a given time, for remote execution of graphical applications, using a mouse, as if you were sitting in front of that same computer screen. Nothing like this can be done with NT: the only tools provided with the server are the program that allows it to be used a Web server, and the tools required to share files and printers. This is why it is so easy to replace a Windows NT server by a Linux server using the open source program Samba.
At the École normale supérieure, where I teach, we keep records of our computers' "uptime", which is the time between two successive restarts of the computer. We have three or four workstations that have been running continuously for more than one year, which is perfectly normal under Unix. However, under Windows NT, any time you make even the simplest change in network configuration you need to restart the program, and crashes are relatively common.
Up until now, you have been talking about Unix in general, which is traditionally considered to be a high level system, and therefore expensive... which means it is not accessible to public administrations.
Correction: a long time ago, Unix was in fact reserved only for those who could afford its high costs, and, unfortunately, certain ridiculous pricing policies, such as one database company that sold the Unix version of its program at five times the price of the Windows NT version, contributed to maintaining this situation. But the recent explosion of open source software has radically changed this: Linux, FreeBSD, and all the other free versions of Unix do not cost one cent, or more correctly they cost the price of a CD. The advantages of a Unix system are now accessible to users with any budget.
Let us examine some of the criticism that teachers have made about Windows-Intel solutions: hardware that is more than two years old cannot be used with recent operating systems; two successive versions of the same program are not always interoperable; and systems regularly suffer virus attacks and system performance decreases, because anyone can "mess with" the hard disk.
Well, using solutions based on open source software can eliminate most of these disadvantages. Costs are minimal since the programs are almost free, no matter how many copies are used, and the hardware used can be much older. These solutions are reliable, stable, and are very resistant to intentional or accidental attacks, be they by human beings or viruses.
Using and maintaining these operating systems is also easier. And their applications can be easily adapted to correspond more specifically to pedagogical needs.
Even better: access to the source code of these programs and the knowledge of their interfaces represents an extraordinary educational resource. This is what has been discovered in the many schools and universities that use this type of software.
While Linux may have developed a reputation because of its performance, it seems that it is still complicated to install, and even to use. Is it, therefore, not really accessible to the general public yet?
Linux is probably not yet ready to be used by the general public, and if I went around saying "Linux will solve all your problems", I would be probably be just as dishonest as the salespeople who tell you "Windows is the only real solution". Right now, the Linux operating system looks a bit like a Ferrari engine in an old unattractive car, like an old VW bus; whereas Windows looks like a shiny streamlined car with a motor that explodes every one hundred miles hidden inside.
So, the choice really depends on the type of person using the computer. What do you recommend?
Personally, I like to break computer users down into three different groups: experts, organized individuals and individuals. Let me give you my recommendations according to each group.
Experts: these are computer users who have a high level of computer skills, such as people in large companies, research centers or universities. These people care little about the body of their car: they want a good engine, and they want to be able to soup up the body on their own, taking advantage of the availability of the source code. For this category, the choice is clear: Linux or FreeBSD (or any other free version of Unix, or even a commercial version of Unix).
At the opposite end of the spectrum are individuals. These are users who are on their own. These are your basic computer users, who don't know much about computers, who are sometimes not even sure that they need a computer, but are interested in trying one out anyway. They'll use their computer to write letters, write resumes, or play "Doom" or "Quake". For these people, who usually purchase their computers in chain stores, Linux, in its current state, is not really an ideal solution, no more than WinTel computers, as they will discover the first time their computer crashes, or the first time they want to uninstall a baseball program. People who really want a computer just to play games should seriously consider buying a video game console (PlayStation, Nintendo or Sega), which, for a fraction of the cost of a PC, guarantees that they will have no problems.
But for those who really want a computer, I think that, until Linux's "body" is finalized, the best solution would be to buy an Apple computer: even if the engine is not the same as Unix, it is much more polished than what is found under the hood in WinTel PCs, and the interface is very user-friendly. The only problem, which used to be the higher cost of Macintoshes (note 55), now seems to be overcome with Apple's more economic recent models such as the iMac.
In between these two categories, you find what I call organized individuals. These are people who have the same lack of computer skills as the general public, but can get help from the institutions that they depend on. This is the case for people such as doctors, lawyers, graphic artists, etc. who have strong professional associations that can (even if they do not do this) provide their members with efficient support and turn-key solutions. In this case, whether the user chooses Windows, Mac, Linux or something else, they will find that the applications they need to use work correctly (or do not work correctly, if they make the wrong choice), without having to worry about installing or uninstalling anything.
This is especially the case for public administrations, and, above all, schools. Just like companies, they want a high level of stability and reliability, but unlike individuals, they can benefit from the high-level computer skills available in universities at a low cost for the government. This could be an opportunity to use efficient economical Linux-based solutions and to revive old computers with systems such as NewDeal. Instead of spending a fortune for the high-performance computers required by Windows, or in paying long-distance telephone technical support for proprietary solutions, the budget allocated to computer equipment for schools could be used for more useful things, such as supporting the installation of nonproprietary networks, developing specific solutions, etc.
It would actually be quite simple to put computers in our schools and spend almost nothing, using old 386 and 486 computers as Internet terminals.
Finally, there is a strong movement right now working to improve the user-friendliness of Linux. The first stable version of the KDE interface was released in July 1998 ( http:// www.kde.org) and the GNUstep project is moving ahead quickly. GNUstep is a free implementation of OpenStep, an extraordinary commercial system that I often use today. It was created by NeXT, the second company founded by Steve Jobs, the famous founder of Apple, who recently returned to that company, that recently bought out NeXT. OpenStep was renamed Rhapsody, and this was more recently included in the MacOs X. The major advantage of OpenStep is that userfriendly applications can be developed rapidly, while maintaining the advantages of having a real Unix system under the hood. Its availability as a free program, such as GNUstep, could be the missing piece of the puzzle that will give Linux, FreeBSD and all the other free versions of Unix a body of the same quality as their motor.
Would this also solve the problem of installation?
This is a different question: Windows, for example, is a system that is not as easy to install as Microsoft claims. An example of this would be the many problems that have been encountered by Windows 98 users (note 56). But these difficulties are usually hidden by the fact that the computer manufacturers "preinstall" Windows, which means they handle the complex job of configuring the computer for you. But no computer manufacturers preinstall Linux. So anyone who wishes to compare Windows and Linux is very disadvantaged by the fact that the installation process is visible when using Linux, but not with Windows.
In addition, since the arrival of Windows 98, there is no more free space available to install Linux as before, and you must be an experienced user to install Linux without erasing Windows. Let me try to explain this simply: you must know that a hard disk can be subdivided into several areas called partitions, and that previous versions of Windows were incapable of using partitions larger than 2 gigabytes; this is why your new PC arrived with one 2 Gigabyte partition, containing Windows pre-installed, and some other empty partitions that you could easily use to store your data or to install another operating system like Linux without too much difficulty. Today, Windows 98 can use very large partitions, so your 6 gigabyte disk is now one large Windows partition, and to install another OS you need to reduce the size of the Windows partition, which is no mean feat, even if some programs, such as FIPS, can do this for you.
It is exactly for this reason that I developed the idea of preparing a Linux demo CD-Rom, which would get over this disadvantage: this way, it would be possible to distribute a CD-Rom so Linux could be run without installing it on your hard disk, simply and easily, in the same way that you run a Windows application. Students of the École nationale supérieure have already gotten over most of the technical problems, and I am sure this project will soon be completed. In this manner you will be able to compare both systems, and see which one you like best, and then you can put pressure on computer manufacturers so they preinstall Linux together with, or instead, Windows.
How much economic influence does open source software have? Does it play a significant role in the computer industry?
Some serious studies have been made that give precise evaluations of the economic and strategic advantages that a company may have when using open source solutions (note 57). There are also several examples of European companies who have implemented these solutions successfully, and later exported them to the United States (note 58). Several large American banks depend on Linux to handle millions of dollars. Linux is used to manage warehouses for L'Oreal, to control gas station pumps and cash registers made by Schlumberger, to oversee the operation of elevators made by Fujitech, to run the corporate network for Ikea, and network computers for Corel. Well-known corporations such as Mercedes-Benz, Sony, Philips, Alcatel and Cisco use Linux. Some companies chose this open source operating system as an Internet/Intranet solution to replace a commercial form of Unix or Windows NT, while others use it to manage their corporate networks by replacing Novell Netware or Windows NT, and still others use it for their computer servers, application servers or office workstations.
An increasing number of industrial companies use Linux at the very heart of their products, such as elevators, Internet kiosks or robots. This is the case for the French company Lectra Systems, who is the world leader in computerized textile cutting machines. Lectra has been offering industrial solutions based on Linux since 1995, because, according to Pierre Ficheux, the company's systems development manager, tests have shown it to be "stable, efficient and inexpensive".
Companies using Linux are not, however, always willing to tell the whole world about it. Since this open source software is essentially free, many managers are afraid of it, and do not take the time to think about the real advantages it may bring them. In many cases, top management of these companies does not even know that they are using open source software. It is often the computer technicians themselves, who have ambitious tasks and limited budgets, who choose Linux, sometimes without their top managers' awareness. If you have a tight budget and are required to use a "Microsoft style" solution, you may solve your problem by using Linux servers equipped with Samba: your boss will only see Windows NT, but you will save a few hundred dollars on each computer!
But this is the way things go in corporations: if something doesn't work, you will be better off covering yourself by having chosen Microsoft. The person in charge of special effects for Titanic knew that he was risking his job by selecting Linux, but the result has shown that he made the right choice (note 59).
Fortunately, this attitude is changing. Two things occurred in 1998, that most of the general public has not seen, but that may be signs of an upcoming revolution. First, to resist Microsoft's attacks, Netscape turned its flagship product, the Netscape Navigator browser, into open source software and its name was changed to Mozilla. Secondly, IBM, worried about the increasing market share obtained by Windows NT in the Web server market, decided to adopt and promote the Apache Web server: this is a high quality open source program that already has 50% of its market. In other words, two of the stars of the computer business –a Silicon Valley start-up, and the giant of the business– have approved the open source software model by their strategic choices.
It is doubtful that Linux will one day overtake Windows. But it is possible that the model for developing and distributing open source software is the one that will win out in the future. Because no company, not even Microsoft, is rich enough to fight against the assembled talents of the best programmers in the world. Especially if the result of their work is then endorsed by the industry's leaders.
But what is the motivation that drives these programmers? Is it possible to make money with open source software?
The very reason that open source software is of such high quality is that the driving force behind its development is not the money that can be made from selling it, but the desire to write programs that will be used by the largest possible number of people. Also, making the source code available means that the programmer's pride is at stake, that he will be judged by his peers. And what better personal satisfaction for a computer programmer than to have helped write a program that is appreciated, used, updated and improved for ten years by thousands of programmers and millions of users, all because it is a good program.
At the same time, the skills acquired while writing open source software have indisputable commercial value. What better experience to put on your resume, if you are looking for a skilled job as a computer programmer, than having made a significant contribution to an open source program that is respected by everyone in the industry? You see, open source software developers can actually earn a lot of money, even if they do not directly sell their programs.
An operating system can only gain ground if the environment is favorable. Microsoft's strength is partly due to the fact that there are many computer consultants that support its standards, along with tens of thousands of software publishers that create applications for Windows. Isn't this a serious hurdle for Linux?
Obviously, Linux does not yet have the industrial and commercial support that the Windows-Intel standard has. But let me start by refuting the specious argument that Linux is not credible, because there is nothing to guarantee its permanence. The history of the computer industry has shown that the permanence of a hardware or software standard is never guaranteed in any way: it depends neither on the size of its installed base, nor on the wealth of the company that promotes it, nor even on the technical excellence of the product! You can ask the early subscribers to the Microsoft Network about this. The good side of Linux is that, at least, it is the users that are controlling things.
Linux distribution is becoming more organized: companies such as Red Hat and Caldera are selling the program in stores, on CD-Roms, together with manuals and basic utilities. It will cost you from $30 to $40 for an unlimited number of computers. Compare this to the several hundred dollars for each computer using Microsoft's equivalent.
It is also wrong to say that there are no applications that run under Linux. Thanks to software publishers such as Corel and Netscape, who have decided to release programs for Linux, there are now a large number of applications available. And many other software publishers will follow. In addition, there are many sites on the Internet where you can download free and commercial software for Linux, that cover most professional needs: word processors, spreadsheets, groupware, Web browsers, database software, etc. (note 60) There are even free graphics programs such as GIMP, which is almost as good as the famous Adobe Photoshop.
Office suites such as Star Division's Star Office, or Applix' ApplixWare, are on the same level as Microsoft Office. While some specific needs are perhaps only answered by Windows applications, it is also possible to have them run over a network with Linux machines emulating Windows, using open source programs such as DOSEmu or Wabi, or commercial programs such as Ntrigue or WinCenter. If you need programs like this, you can also just keep one WinTel PC to run them.
As for technical support and commercial maintenance for Linux software, this is certainly at its beginning, but certainly does exist. Companies such as Pick System or the French company Alcove, have customers such as French research institutions (CNRS, CEA) as well as many large corporations (L'Oreal, Philips, Alcatel). In addition to commercial technical support provided by distribution companies, and maintenance support provided by computer consulting firms, Linux users can always count on the legendary solidarity of the community of users, through Internet newsgroups such as comp.os.linux.hardware or comp.os.linux.setup. Linux also received a prize from Infoworld magazine for the quality of its technical support. One cannot say the same thing for Microsoft.
What attitude have different governments adopted with respect to open source software?
I am always amazed by the refusal of our political leaders to face the facts on these issues. For some of them, computers are just a technical subject that is in fashion these days, but not very interesting, as if all that counted was selecting the brand for your word processor! And those politicians who have understood what is really at stake in the information society are often merely hoodwinked by the propaganda of software companies.
The first thing to consider is that governments could save a great deal of money by selecting open source software. To have a better idea of what is at stake financially, look at the case of a university near Paris that recently purchased 15 personal computers to install Linux. No one knows exactly how much computer manufacturers pay for Windows 95, but it is probably around 500 Francs (less that $100). If the vendor makes no profit off this (which I doubt), this means that the university had to spend 15 times 500 Francs, or 7,500 Francs for a product that it did not want. Basically, this was a gift of 7,500 Francs made by the French government to Microsoft. If we extrapolate this example to all the similar establishments in Europe, and to all the different public administrations, we can see that governments are wasting millions of dollars to subsidize an American company, whose European subsidiaries are no more than distributors, and, on top of that, a company that is far from needing this money to survive.
This would be a unique opportunity to obtain a bit more freedom from the technological monopoly that is held by unscrupulous multinational corporations, and to give our companies and schools a very important strategic advantage. What we need to do is to convince computer vendors, such as HP, Dell, IBM, Gateway, Compaq etc. to stop acting as Microsoft's tax collectors, and start offering users PCs where this tax does not have to be paid. In France, after a very long effort from universities, this is starting to come true: at the time of publication of this book, HP and Dell have begun offering such an alternative (to universities only, unfortunately), and Dell is offering a similar possibility in the US.
How should Europe respond to these questions?
It seems to me important that Europe quickly develop an active and independent policy concerning computers, and information processing in general. It certainly has the technical means to do this, because Europe has skills at the same level or even better than those found in the United States. European computer research centers are at the forefront of formal software verification development methods, that have made it possible to realize many important projects, the latest of which was the second launch of the Ariane 5 rocket.
What is lacking right now is real political will: a manifestation of this could be the creation of a European agency for open source software and open standards. This could be made up of scientists who share the ambition to help the cooperative efforts of the Internet community, and build a high quality open platform for interoperable computer systems. The European Union could use such an agency to help existing cooperative developments around open source software and open standards. This would only require a few million dollars (which is nothing compared to the European budget) and would make it possible to quickly finalize strategic projects such as the GNUstep interface presented earlier (note 61), to favor the development of an infrastructure for information exchanges inside the European Union, and to create a high level network that could be a catalyst for the creation of modern, free, open and dynamic computer systems.
Only an initiative of this sort would make it possible for Europe to organize its computer industry, that is, its fate, while helping develop value-added jobs in these sectors. It makes me sick when I hear Microsoft giving us lessons such as: "crack down more on software piracy, that's what will create more jobs in the computer industry"! (note 62) Since when has the development of Microsoft software –that is entirely done in the United States– created any real jobs in Europe? And before giving lessons about piracy, it would be a good idea for Microsoft to start reimbursing Windows licenses that are imposed on so many users who don't want them.
The choice of a free, open system would have the effect of eliminating the information tax that is collected by Microsoft, making our companies more competitive and helping reduce unemployment. Because these tens of millions of dollars that only end up in the pockets of Microsoft or its fellow software companies could be spent on productive activities, such as financing maintenance contracts with local computer service companies, who could adapt hardware to fit the specific needs of companies. This could create a real opportunity for expansion, fostering growth and helping create skilled jobs for engineers who would be responsible for the quality of their products, and not only for sales people who are trying to sell products they have no control over, and whose profits end up returning to Redmond.
Let me say clearly that there is no conflict of interest between Europe and the United States on this subject: what is at stake is open computer standards, and the risk of seeing a monopoly extend its control over all of the links in the information chain; these stakes are the same for everyone, no matter where the monopolistic company is located. This is a challenge that concerns all of mankind.
Unfortunately, it seems clear from current political discourse that politicians will not be the ones to change this. This collective awareness will only come from public opinion, from the computer community, and from citizens all over the world. From these millions of computer users who have been marginalized, manipulated and held in contempt up until now by Microsoft. Perhaps one day they will be fed up with being treated like cash cows and guinea pigs, who are considered just good enough to keep quiet and pay for software that doesn't work correctly. And this would then be the serfs' revolt.
Would you go as far as saying that our passiveness in dealing with Microsoft is a hidden, tacit societal choice?
Without doubt. The Linux adventure, for example, brings to mind three words that you are familiar with: liberty, equality, fraternity. But the Microsoft adventure, as you have probably understood by now, presents a society that, in my eyes, would be closer to servitude, inequity, feudalism. Computers are now giving us the possibility of drastically changing the way we live our daily lives. But with this opportunity, there is no such thing as a free lunch: the only way you will find anything in this information society is if you bring it yourself. If we continue letting Microsoft build it, it may end up looking like a nightmare. It is therefore up to us to choose whether this revolution should lead to a technological dark age dominated by a handful of feudal lords who take control of the means of creating and transmitting information to collect taxes each time we communicate with each other. Or, if we would rather live an open and modern world, built around democracy and decentralization, where the free flow of information allows us to benefit from the huge potential of borderless cooperation and shared knowledge.