Date: Tue, 20 Oct 1998 21:56:26 +0000 From: "Karsten M. Self" <kmself@ix.netcom.com> To: lwn@lwn.net Subject: Translation of french MS Open Letter This is a multi-part message in MIME format. --------------15C46E76815747DA1E093859 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit This is pretty severly Babelfished. I've fixed a couple of eggregious manglings, but not much. -- Karsten M. Self (kmself@ix.netcom.com) What part of "Gestalt" don't you understand? Welchen Teil von "Gestalt" verstehen Sie nicht? web: http://www.netcom.com/~kmself SAS/Linux: http://www.netcom.com/~kmself/SAS/SAS4Linux.html 2:51pm up 34 days, 14:40, 5 users, load average: 1.07, 1.17, 1.25 --------------15C46E76815747DA1E093859 Content-Type: text/plain; charset=iso-8859-1; name="open_letter.txt" Content-Transfer-Encoding: 8bit Content-Disposition: inline; filename="open_letter.txt" Un petit Français FUD October 19, 1998 Madam, Sir, You know it, I have just been pleased to take the General Management of Microsoft France following promotion in the United States of Jean-Philippe Courtois. In a few weeks I discovered company powerful and innovating, but such a responsible and deeply anxious to listen to and satisfy his customers and his partners. Like every undertaking, Microsoft has competitors and detractors, nevertheless I was strongly surprised by the nature and the violence of certain recent attacks, which describe a company which in common does not have anything with that which I côtoie every day. It thus appeared significant to to me that Microsoft publishes a Letter Open explaining the reality of its position compared to the most unjust criticisms whose we are the object. The purpose of this Opened Letter, of which you will find a specimen herewith, does not claim to present Microsoft under one day ideal, nor to convince that we are a company without defect, but is simply to clarify certain points. Sincerely yours Marc Thistle General manager OPEN LETTER... to our thousands of customers, our thousands of partners, shown to be quite ingenuous victims. Microsoft is in France the preferred firm of the frameworks of company (1). The success of Microsoft however causes on behalf of certain observers of the sour comments, and this phenomenon tends to develop at the time of the lawsuit whose company is the object in the United States. It is from now on current to see relaying rumours which take part in the construction of a myth, according to which Microsoft would be a mainly hegemonic company with goal. Microsoft has without any doubt its share of responsibility, and we do not dispute that one can criticize the company, which appears to us to raise of the right of the contradiction most elementary. But we dispute the means used sometimes to do it. The theses most frequently advanced are not supported by studies, the references and quotations are not allotted and the judgements, often peremptory, make only take again nonfounded rumours which circulate on the Internet. This would undoubtedly be only little of thing, if criticism had not recently reached these same with which we owe our success. In front of so much of contempt, it is the respect which very undertaken owes to its customers and to partners who pushes us today to react. Microsoft never invented anything This vision of the things illustrates a fundamental confusion in the interpretation of the concepts of invention and technological innovation. As for very undertaken industrial, the role of Microsoft consists in transforming an invention produces some innovating. Discovery or invention represents capital exploitable only if it can to translate into innovation technological, i.e. if it can be integrated in a product useful, which satisfies a new need, at an acceptable price by the market. In all the fields, thousands of inventions remain forever unexploited, the creativity of the inventors corresponding unfortunately only seldom to the criteria of the technological innovation. One of the forces of Microsoft is that its founder, Bill Gates, impassioned data-processing technology, knew to recognize inventions which had a technological potential of innovation. It then took the risk to invest humanly and financially to make evolve/move what was at the beginning only brilliant ideas, worms of the products strongly innovating. Thus, for example, Microsoft never disputed that the language BASIC was invented by Kemmeny and Kurtz. For as much, this language could not have been largely exploited if Bill Gates had not developed out of assembler an interpreter compact ultra BASIC (less than 4 KB), and had then carried it on several machines and processors. This work of development and adaptation for several platforms made it possible at Microsoft to lay off in the form of royalties this interpreter BASIC with many manufacturers of the time like Commodore or Atari. APPLE, created two years after Microsoft, profited upon the departure of the interpreted BASIC of Microsoft in its APPLE II. In the same way, a little later Bill Gates saw as Tim Patterson, inventor of the stock of code Q-DOS, a data processing specialist of great talent. Tim Patterson was at the time developer independent and single employee of Seattle Computer. He was recruited by Microsoft, as well as other developers thereafter. >From the basic code the end product, several months of work were necessary, during which Tim Patterson and of many Microsoft developers made evolve/move the stock of code of Q-DOS, to make it compatible with the standards of quality of the time. The step was identical in the world of the applications and on more recent technologies. To innovate, a company of software as Microsoft must know to identify collaborators with strong potential, to recruit them and preserve them, in their giving the means of developing their intelligence and their creativity with the service of the company. Indeed, the raw material of a company of software is exclusively made up of grey matter: the performance of Microsoft is thus due to its capacity to recruit good collaborators Microsoft thus recruits 4000 people per annum, including one great number of developers. However, the demand for intellectual capital is immense in the software industry, in particular in the United States, where is installed the majority of the large companies of the sector. To face the shortage of data processing specialists, the American Congress besides has just voted a law authorizing the immigration of 300.000 foreign data processing specialists. The good developers are the subject thus of strong requests on behalf of the companies. It results from the law of the market that those pass sometimes from a company to the other, with what constitutes their value, i.e. their capacity to invent concepts or to develop products. Yes, for example, Windows NT was conceived by Dave Cutler, former employee of the firm DIGITAL and one of the creators of system VMS. This assertion is in conformity with reality, but it does not constitute therefore a noisy revelation. Microsoft recruited Dave Cutler in 1988. He was at the time at DIGITAL. The company entrusted to him as from 1989 the piloting of the Windows NT project. The first version of this software came out 4 years later in 1993, thanks to the talent of Dave Cutler, but also to the work of several hundreds of Microsoft developers. The repurchase of start-ups innovating by other companies forms part of the dynamics of the data-processing market: the fast evolution of technology caused a policy of acquisitions on behalf of all the large companies of the sector, and not only of Microsoft. A company, in particular in the sector of new information technologies, must always have a length to answer the fast trends in the market in advance. The company not having always in-house the products making it possible to satisfy the request at the good time, the acquisition of start-ups enables him to have a greater reactivity and to continue to innovate vis-a-vis with competition. Sun Microsystems, one of the principal competitors of Microsoft, creator of the language Java and the environment systems around Java, thus repurchased a great number of companies during last years, among which NetDynamics, Diba, Encore, IMP, Long View Technology, LightHouse Design Reads, Chorus, CRAY and still recently IPLANET. IBM, another competitor of Microsoft, have also carried out many acquisitions, not only of innovative start-ups, but also of well established companies, lay out of share of market consequent, like Lotus (for 3 billion dollars American in 1995) or Tivoli (for 1 billion dollars American in 1996). Microsoft launched out only tardily (there is less than 5 years) in the repurchase of some start-ups, having hitherto privileged the recruitment of new collaborators. It is indeed the fast trends in the market as from 1995, corresponding to the great years of the development of the Internet, which caused the repurchase of several companies by Microsoft. These small companies, often non quoted in Nasdaq, used developers of quality on incipient technologies. These technologies thus could be integrated in the Microsoft platforms, and thus could know a great diffusion. Consequently, during these years 96-97, Microsoft acquired outside a share of code more significant than in the previous years, the proportion of the in-house developed code remaining however largely majority (more than 90 %). The start-ups which were repurchased were doubly grateful: in the financial plan, since their founders and developers became very rich day at the following day; in the commercial plan, since their creations were largely diffused and used through Microsoft products. Microsoft is one of the first 5 world companies, all confused sectors, as regards investments under development research and. Microsoft devotes 17% of its revenues, that is to say this year 3 billion dollars American, in investments devoted to the research and the development. One can deny only the private companies, and Microsoft in particular, make a contribution essential with the development of data-processing science. Besides one finds a comparable situation in other sectors, like that of pharmacy. Thus, they are teams of search gathering on the whole 400 people, gathering personalities among the most considered sector (such as Jim Gray, the creator of the transactional CICS at IBM or Rick Rashid, the inventor of the micronucleuies), and reinforced by more than 13.000 developers and testers, who build each day the success of Microsoft. These teams of search, established in the United States but also in Europe, in Cambridge, work on generic technologies which have the role to be then integrated in the form of components in the systems and Microsoft applications. Among these principal research orientations, one can quote: the techniques of artificial intelligence, intended to simplify the interface enters the man and the machine; architectures networks and servers distributed to allow the rise to power of the servers containing PC; techniques of programming and software tests, to allow to develop and to put at the point more quickly and in a more reliable way software. The Microsoft products are of poor quality Microsoft technology aims at satisfying the needs for the greatest number of users, in which the majority are not specialists, and for which data processing is a tool whose handling must be simple. We will thus evoke here only the topics which directly interest the users of our technology, reserving for other places the polemic of specialists. Microsoft software can certainly contain bugs, like all the software. The bugs are in particular related on the complexity and the number of different configurations on which the products must be carried out. In addition, the developers are men and not machines, and can thus be fallible. This general observation does not prevent Microsoft working in-depth on the methods of design of its software, and from devoting significant investments to the procedures of tests, in order to reduce to the maximum the bugs. We do not dispute the fact that we must make even more efforts in this direction. The apparent amplification of this phenomenon is very related to the fact that our products are diffused very massively, and that, as a corollary, the number of cases of figure of use of our software, in particular in demanding professional environments, is very high. Microsoft indeed made the choice difficult to ensure the interworking of its products with multiple data-processing platforms, to answer at the request of its customers. One of the missions of our technical support is, inter alia, to go up with the head office of Microsoft the problems encountered by the users, in order to improve the products. Microsoft provides also free corrective measures during the lifespan of its products. These corrective measures bring at the same time new functionalities called " minor ", and they are placed at the disposal on the Web site of Microsoft (2). The Microsoft software wants to be standard products: they contain a great number of functionalities to be able to satisfy very varied uses. Another frequent criticism consists in saying that our software is encumbered useless functionalities and that only 5 % of the functions are really used in a current way. Popular wisdom would challenge the argument in the name of "which can more can less", but we will not be satisfied any. The functions integrated in the Microsoft software are it after a thorough analysis of the needs of the consumers, thanks to the compilation of information which arrives at the customer services of all the Microsoft subsidiary companies, and to the tested methods which are "Usability Labs". In Usability Labs, one asks users neophytes to take in hand the new software on PC. The testers analyze the behavior of the users - who are not helped - and film them. The users are then questioned on the difficulties which they encountered, the use that they would make of such or such function; and analyzes it their answers gives place to recommendations in the development of the products. The force of the standard software lies in the fact that it must be sufficiently rich to be appropriate for broad and various populations. Thus, let us take the example of a population of 100 users, exerting 10 different trades within a company, and using each one only 20 % of the functionalities of the word processing software Word. It is probable that by cumulating the use of these 100 users, they are 90 % of the functionalities of Word which are really used. Microsoft is however conscious that a given user who would not use that partially the possibilities of Word can wish to have a simpler interface, in the respect of a common standard: that formed part of the schedule of conditions of several of our future products. Lastly, let us not forget that the standard software has another advantage: its low unit price for the user, taking into account volumes of diffusion reached. The Microsoft software is certainly consumers of material resources, but the accelerated evolution of the technology of the PC (processors, memory, hard disk, inputs/outputs etc...) strongly relativizes this state of things. Our software has indeed a significant size, which is however not higher than that of the majority of the products of competition. In addition, the increase in the software size is directly related to the improvement of their performances and services rendered to the users. The passage to the graphic interface was thus a factor of consumption of significant material resources. At the beginning of the Nineties, at the time of the passage of MS-CDos with Windows, all the users remember to have made this observation. It was necessary several years so that the graphic interface re-enters in the m?urs, but nobody would like to return today to a word processing to cryptic command in character mode. Most of the place occupied by the software is also related to the systems of on line help and self-tuition (which make it possible to assist the users and to reduce the costs of support). However, the user is not obliged to preserve these files on the hard disk. Lastly, the increase in the storage capacity of the hard disks was much faster than the increase in the weight of the software: thus, the word processing software de traitement de texte Word occupy 10% of disk hard of a PC it there have 10 year (2 MB on a disk hard of 20 MB), whereas it occupy more only 1,2% of disk hard of a PC of today (approximately 50 MB on a disk hard of 4 GB). Microsoft plans to launch very significant programs of communication to take control of the markets of education and training, and would use the children as prescribers of its technologies Present for 10 years in the sector of Education, Microsoft there has always paid a detailed attention and falls under the dynamics of the development of new information technologies at the school. To this end we developed two programs: Granulates the Multi-media one and Competences 2000. Granulates the Multi-media one: thanks to this program, assembled in partnership with Hewlett-Packard, a certain number of primary schools (18) were seen equipped with multi-media PC and software, including software not Microsoft; their teachers followed a cycle of formation to the control of Internet and local area network which was placed at their disposal. They also took part in a collective?uvre of study of the teaching use of microprocessing and the multi-media one in the schools, whose results were made public. Competences 2000 is a training program to data-processing technologies intended for the world of education, higher education, to the professionals of data processing and the applicants for work. It is the fruit of a partnership innovating between the industry of information technologies (of which Microsoft), professionals of the formation and teaching, the public sector and local communities. It aims at creating a true dynamics aiming at reducing, in France, the shortage of technical skills, shortage for a long time highlighted by the analysts and the professionals of data processing. After a few months, Competences 2000 meets a sharp success on the ground, which shows that this program answers waiting. Microsoft brings through Graine of Multi-media and Competences 2000 a contribution, undoubtedly modest, with the awakening of the importance of the data-processing tool in the sectors of education and training. We are not the only ones with launching this type of initiatives, and it is a good thing, because it is about a major stake requiring the contest of all the actors of the market. The goal of Microsoft would not be to make good software, but to make profit, thanks to contestable commercial methods, involving the constitution of a monopoly. The first objective of a company is to serve its customers, by manufacturing products and services of quality. The associated objective is to serve its shareholders, the concept of profit, and we will not deny it, being at the base of the operation of the market economy. With regard to Microsoft, the quality of our software in all cases was recognized, at the point to satisfy today the needs for tens of million users in the world. This known as, it is true that Microsoft is a powerful company financially, and this for several reasons. The strategy of Microsoft consists in developing good software, and ensuring marketing of it on a very large scale, thanks to partners and intermediaries which make it possible to ensure this volume. The profits of Microsoft (30 % of clear margin today) are related to several factors: * the volume of the sales: for Microsoft, as for all the software publishers, the margin is directly related to the volume of the sales, itself related on the quality of the product and the system effectiveness of distribution. * size of the market: this one is in constant increase. The market of the PC grows approximately 15 % per annum in the world, and 90 million PC will be sold in 1999. The sales of Microsoft being related to the size of the market of the PC, Microsoft benefits from this evolution. * the manufacturing cost of the products: the evolution of the sales to the companies tends towards a progressive dematerialization of the products, with the signing of licence agreements. * the manufacturing cost of the products: the evolution of the sales to the companies tends towards a progressive dematerialization of the products, with the signing of licence agreements. It follows a reduction proportional in the sale of the products out of boxes, whose logistics and manufacturing cost is not negligible. business management: Microsoft is a managed company in a very rigorous way and with a permanent preoccupation with an effectiveness, and this with all the more of motivation that a great majority of his/her collaborators are shareholders. The marketing strategy of Microsoft implies a relation between partnership and the developers and the manufacturers, in order to offer powerful products to the end-users. The success of an operating system rests on the constitution of an infrastructure of competences and complementary products, in hardware and software. When an operating software is developed, integrating the new functionalities corresponding to the great technological developments of the moment, it is necessary that as much as possible applications can use this new system. Versions beta of the software are diffused with the developers, to enable them to develop new applications, before the output of the operating system. The retailers and other tests diffusing the system are those which will establish the link between new technology and the end-users. It is thus significant that they have a perfect command of the functionalities of the new operating system, to be able to advise employment to the end-users of it. Microsoft thus devotes a significant budget to this technical and commercial training of its retailers and partners These commercial methods, resting on an indirect distribution 100% of the product traditional, and are practised in the same way by the other industrialists proposing of the operating systems. However, the original situation of Microsoft on the market of the operating systems also explains its success. Indeed, whereas Microsoft limits itself to the edition of software, its competitors on the market of the operating systems (Sun with Solaris, and IBM with OS/2) are at the same time manufacturers of hardware. They thus have logically more difficulties in market products near other manufacturers of which they are the candidates. The sale of a PC with Windows concerns a necessary integration upstream, and not of the dependent sale. It is illegal to subordinate the sale of a product to the concomitant purchase of another product. But in the case present, it is an integration upstream of a finished product, and not of dependent sale. The operating system forms integral part of the final configuration which is proposed to a customer at the end of the chain, as well as the hard disk or the memory. The obligation of the manufacturer is to provide to his customers a product in operating condition, being able to give satisfaction as of the first use. Let us recall here that the mission of Microsoft is to put microprocessing at the range of all. For the end-user, the integration of a maximum of functionalities on the PC is a factor of simplification of startup and use, which this user seeks, in the final analysis. There are several channels of sale of PC, and the users have the choice between international manufacturers, national manufacturers, and assemblers. The latter, which represent approximately 30 % of the French market, sell PC by offering the choice of the operating system. They also sell machines without operating system, on which can be installed software like Linux. This channel will thus interest a population more tested in data processing. On the other hand, the large majority of the customers, including in company, ask to profit from PC under Windows, according to manufacturers' themselves. This choice of the users is closely related to the great availability of applications and peripherals designed for Windows: it is enough to visit the microprocessing ray of any store to realize it. it be good also to stress that the PC be the platform material on which the choice of operating system d' exploitation be the more vast. No other computer proposes this choice. Macintosh are delivered with the MacOS system of APPLE, and one can install Linux there. The stations Unix RISC of the manufacturers are delivered with Unix the specific version of these manufacturers. AS/400 from IBM are delivered with OS/400. On a workstation of the type PC, there are the choice between MS-CDos, Windows 9X, Workstation Windows NT, Linux, Sun Solaris, SCO Unix, Unixware, Interactive Unix, IBM OS/2 Warp, and even Rhapsody (adapted version of MacOS for PC). The position of Microsoft on the market of the office automation application and operating systems for PC is certainly significant, but it does not exempt to us to innovate permanently to meet the needs of our customers. Indeed, no constraining rule prevents the market from adopting other technologies that those of Microsoft. If Microsoft were given a secure income, the company could impose a raising of prices of its products. However the price of Windows be remain stable during five last year, whereas the service that it offer have be very largely wide (integration of Internet, support of Universal Serial Drink, tool for car-maintenance, three-dimensional chart 3d more fast, good support of play, etc). As for the prices of the application softwares, they dropped considerably since the last years, and more quickly still than the products of competition (-47 % since 1991 for Microsoft Word, against -18 % for Corel WordPerfect over the same period). Lastly, if Microsoft were really in a situation privileged compared to its competitors, the company could without danger cease innovating. However the investment of Microsoft under development research and progresses constantly, in absolute value like expressed as a percentage of the turnover, to answer permanently the new challenges of the market. Actually, the cycle of evolution of technology is so fast, that one would need only 3 to 4 years so that Microsoft disappears from the market if the company did not innovate more. Microsoft would be able to control the world chain of information and the communication, thus endangering democracy Can only one company, in fact Microsoft, control the world chain of information and the communication, and endanger the democracy by taking the control of the Internet? It is useful, once and for all, to reestablish with reality. The control of the Internet is a myth which concerns science fiction. Microsoft is a made up company men and women, impassioned technologies, which are responsible and respectful citizens laws before being Microsoft employees. It is thus unreasonable to make believe that Microsoft could make an attempt on the democracy and the personal freedoms. In addition, the Internet is the most open and democratic system which exists. The takeover of the Internet is completely impossible, and even less by only one company. Thus, and even in imagine that Microsoft provide all the standard use by the network - what be obviously not the case, Microsoft support a list very broad of standard of market define by some organization independent - that allow not more with company to control the Internet, that that have allow with France Telecom to control the telephone conversation téléphoniques. The Internet constitutes today the most competing market and that which develops most quickly with the world. As the number of users increases, more and more services will develop, and be accessible at a less and less raised cost, thanks to possibilities of geared down connections. Many economic operators are implied in the operation of the mother of all networks: software publishers, service companies, computer manufacturers, operators of telecommunication, routers, providers of access, etc. Microsoft positions firstly and mainly on the trade of the manufacture of the software, the objective of the company being to adapt to all the standards, protocols and operating platforms used by the Internet to be able to meet the needs of its customers. The developers can thus conceive various applications, which make it possible to the users as well as possible to profit from the potentialities of the Internet. The only alternative to Microsoft is Linux Linux is a movement created in 1991 by a Finnish student, Linus Torvald. On the basis of adapted a Unix core, developers communicating by Internet ensure the creation of the complementary functions and the various versions of Linux. Linus Torvald synchronizes and validates the unit. The fundamental rule is that each developer gives its source code, so that this one can be improved, and is the subject of later developments. The product is diffused under specific licence of use, accompanied by its source code. Each one can thus modify, improve or copy the software, and redistribute it in its turn, free or not. Commercial versions are also diffused. The units installed is estimated at 6 million licences in the world. Linux has without any doubt its place at the side of the operating system of Microsoft, but also of those of IBM, Sun, Oracle, and Novell. The use of software free concern however of a choice philosophical more than economic, and make some control of operating system d' exploitation a business of specialist senior. It appears that Linux does not fulfill today the requirements of the majority of the companies, and even less of general public. Linux present of the limits which are a barrier to its diffusion on a large scale, in particular in the companies and near general public. In the plan of the stability of the system, Linux poses problems of total coordination and one strongly feels the absence of a head of project. Linus Torvald left the university last year to join a Californian company. The development of Linux since slowed down considerably. In the same way, the maintenance of each functionality of Linux depends on the mobilization of the teams. Thus, certain functionalities have not known updating for two years. The installation of Linux is delicate. So that the performance of the system is optimal, each version must be regulated on each computer with each installation, by a qualified data processing specialist. Its setting in?uvre and its administration are thus not with the range of a data processing specialist accustomed to the user-friendly systems, and even less than one basic user. The use of Linux is complex, the software functioning in general with commands in text mode. Lastly, the application softwares functioning under Linux generally have limited functionalities. The word processings seldom have functions now current on PC or Macintosh: on-the-fly spelling correction, graphic data entry of tables, integration of graphs coming from applications. The majority of the word processing softwares under Linux are still connected at " Microsoft Write " which goes back to 1985. The advantages related on the exemption from payment and the transparency of Linux are nonrelevant criteria for the majority of the users. The exemption from payment of Linux is a not very decisive advantage. Indeed, the cost of an operating system is marginal compared to the different costs of the company. The price of the system is only one of the elements of the price of the data-processing chain in the companies. The installation of a traditional operating system costs several times the price of the system itself, very as much as the realization of a applicatif program and its maintenance. Thus, by installing Linux, one saves the cost of the operating system, but one increases that of the installation, and one takes risks for the maintenance of the applications and the system itself. So some consider that the availability of source code sources of Linux be a guarantee absolute of independence live have screw of software publisher de logiciels, we perceive badly which be the interest for a company or a private individual to have access with source of its operating system... On the other hand, the diffusion of the sources of the systems is very useful for the students and the researchers, to include/understand the operation of the system and possibly to modify it. Linux will thus remain probably a long time still a good subject of study for data processing specialists, rather than a system dedicated to a significant diffusion. Conclusion We hope that this document will have contributed to give to the reader a better comprehension of the position of Microsoft in the debates whose we are the object: once again, we are conscious that the confidence of our customers and partners is deserved, and criticisms whose we can be the object must cause all our attention as from the moment when they are objective. The teams of Microsoft France are in any case at your disposal to clarify and specify, if necessary, certain subjects tackled in this document, but more especially to daily meet your needs. For all additional information: Corinne Got - BDDP Corporate for Microsoft France tel. 01 49 09 26 50 or 26 54 Fx: 01 49 09 27 09 e-mail: cgot-camard@bddp-corporate.fr Notes: 1 As testifies some its first place with the classification to the public image of the companies published by the Expansion in August 1998. 2 Information on these corrective measures is provided in several forms: in the topicality produced sent by monthly e-mail free subscribing to the Microsoft customers to this information, in the Microsoft publications, and also in the many technical seminars organized by Microsoft for the customers, partners and developers. Information in addition is largely relayed by the data-processing press. --------------15C46E76815747DA1E093859--