Open Sources 2.0/Open Source: Competition and Evolution/Libre Software in Europe

From WikiContent

Jump to: navigation, search
Open Sources 2.0

Jesus M. Gonzalez-BarahonaGregorio Robles

The libre (free, open source) software[1] community is probably one of the more global and internationalized. Therefore, it may be a little artificial to try to separate the European share of it in the hope of finding peculiar characteristics. But at the same time, Europe is so diverse, so full of national, cultural, and linguistic boundaries, that it may be difficult to find common patterns in this already diverse libre software world. However, our feeling is that in between these two facts, there is plenty of room for writing about what is happening in the European libre software scene. While preparing the material for this chapter, we have come to the idea that, in fact, this is not such a global world, nor does its European fraction lack common patterns despite its diversity. With this focus in mind, we have looked for both the peculiarities and the commonalities. We have walked through the enormous amount of data concerning what is happening in the vibrant European libre software scene with the aim of offering the reader the more relevant and revealing trends and facts, providing a vision of a complex and diverse, but also uniform, landscape. And, for sure, a very personal one.

In fact, although libre software can be considered to come from the U.S., the spread of the Internet (and before that, the Usenet) in Europe made it possible to develop a fragmented European libre software community as early as the late 1980s. With time, common trends that could be called "European" seem to be emerging from this framentation. However, even without strong pan-European relationships, and maybe due to the common sociocultural background, some trends are found now and again in different parts of the continent, producing a collage with many common patterns.

In this chapter, we provide some snapshots of that collage. Instead of focusing only on issues that can be truly called European (because they involve participants from many areas of the continent), we have tried to show the diversity of initiatives and environments, so the reader can have at least an idea of the details of a very complex landscape. We have also devoted some efforts to identifying common patterns and Europe-wide initiatives, especially when we find they are a signal of an emerging common trend. Intentionally, many examples which are European by nature or birth, but have evolved into global projects, are not included, or are mentioned briefly, since they are now more global than European and therefore make little sense here. There is, however, a clear intent to show the main European contributions to the libre software world, in terms of development, use, and promotion.

All in all, the set of case examples, and the issues presented on these pages, are just a (hopefully representative) showcase of what is happening in the European world of libre software, obviously filtered by our personal biases and backgrounds. For sure, the selection by any other observer of libre software in Europe would be different. We just hope that the reader will find our selection illustrative enough.

Contents

Brief Summary of an Already Long History

Before showing the current landscape of libre software in Europe, it seems necessary to provide some historical background. From issues like the European involvement and impetus in projects such as Linux (the kernel) or the KDE project, which influenced greatly the shape of the currently available libre software, to very specific use cases in European companies, which are mainly a consequence of the global importance of the phenomenon, there is a whole rainbow of milestones which will contribute to the understanding of the present situation.

The evolution of libre software in Europe during the early days was parallel (as it was in other parts of the world) to the penetration of the Internet (and before it, the Usenet). Therefore, it is not strange that areas which had an early and deep exposure to these Nets—such as the Nordic countries, the Netherlands, and the United Kingdom—also had the first cases of involvement in global libre software activities. However, it is important to also consider some linguistic issues. For instance, those countries where the English language (which is clearly the lingua franca in the global libre software community) has more penetration (either as the mother tongue or the second language) commonly had an earlier involvement in libre software. With the passage of time, maybe since the mid-1990s, specific dynamics started to show strength in other regions, with more or less of a relationship to the global evolution. Most European countries (Germany, France, and Spain are clear examples) have grown their own communities and libre software fabric in partial isolation, to the point that many initiatives that are very well known in one country are almost unknown in the others and outside Europe, despite their valuable contributions. In many cases, the flux of information among Europeans of different areas is still carried through global (or, for that matter, American) events and initiatives. Just as a case example, we still find out about relevant news in other European countries through the American web news service, Slashdot, despite being reasonably well linked to the libre software communities in several European countries.

Considering this situation, the contributions of Europe to the history of libre software are extensive. We can consider, for example, events such as the birth of many applications and projects (consider Linux, the kernel, by Linus Torvalds, Finnish; Python, by Guido van Rossum, Dutch; MySQL, by Michael [Monty] Widenius, Swedish; PHP, by Rasmus Lerdorf, Danish; KDE, by Matthias Ettrich, German; and many more); or the foundation and development of some of the first companies with a business model based on developing or distributing libre software (such as MySQL AB, Trolltech, or SuSE); some of the first studies and initiatives denoting attention by public administrations to the libre software phenomenon (such as those by the European Commission); and some of the first research projects considering libre software as a matter of study (such as those performed by the FLOSS project).

Most of those contributions will be mentioned and presented in some detail within this chapter. Instead of following a timeline approach, we have preferred to group matters according to the different topics involved, each one in its own section: developers, community, companies, public administrations, legal initiatives, licenses, education, and research. Of course, this implies a certain degree of artificial delimiter, since many issues in fact belong to more than one of those sections. But we hope that what is lost with respect to that precision is gained in readability and comprehension of the situation as a whole.

The Development Community

Since the emergence of libre software as a concept, European developers have contributed considerably to its growth. Only recently have we had evidence of how large this contribution is: the WIDI survey in 2001,[2] the FLOSS study in 2002,[3] and the FLOSS-US study in 2003[4] showed that large numbers of developers declare themselves as nationals of a European country. In the case of FLOSS-US, from a sample of almost 1,500 developers, more than 900 (or about 60%) declared themselves to be living in Europe (including Russia), compared with about 405 in North America. In the FLOSS study, more than 70% of the developers were living in Europe (14% in North America). WIDI reported developers were 54% European and about 35% North American. Of course, all these studies could be biased (and at least they are with respect to language, since they all were done in English), since the respondents were self-selected in nature and the studies were not focused on geographic characterization, which was more like a side product of other characterizations. But they were conclusive about the European contribution to libre software development being quite high and, as an aggregate, incomparable to any other region worldwide.

It is also interesting to notice the distribution of libre software developers within Europe, since not all countries are equally represented. In this respect, results from studies are variable, although some common patterns can be inferred. For instance, France and Germany seem to be the countries with a greater population of developers in absolute numbers, closely followed by the United Kingdom and Italy. According to the FLOSS study, France has the highest, with 15% of the total of respondents, followed by Germany (12%), Italy (8%), the UK, the Netherlands, and Spain (all at about 6.5%), and according to the FLOSS-US study the first country is Germany (25%), followed by the UK, France, and Russia (all close to 4%), Spain, the Netherlands, and Italy (about 3% each). WIDI results give 20% for Germany, close to 5% for France and the UK, and about 3% for the Netherlands.

These absolute figures are hardly surprising, since they match (with some exceptions) the countries with a greater population and GDP (which in general shows a certain correlation with the number of developers all around the world). But some countries are clearly overrepresented. Among them, the Netherlands is a clear case. Sweden, Spain, Poland, Switzerland, Finland, the Czech Republic, Denmark, and Austria also seem to have more developers than their population or GDP would suggest. In fact, the Czech Republic and Austria are two relatively small countries that have some of the highest figures for libre software developers per capita.

A much more specific case study is the Debian project, where developers have the option of indicating their residence. Of those who registered, the figures are in line with the previously described results. Seven of the top eleven countries in November 2004 were European: Germany (151), the United Kingdom (80), France (57), Spain (37), Italy (34), the Netherlands (30), and Sweden (28). The total number of European developers is well over 400, which compares to 364 in North America (U.S. and Canada), the second-highest region in this survey.

In summary, Europe clearly has a high concentration of developers, possibly higher than any other region in the world. Within Europe, the Western countries have higher concentrations (though the Eastern countries of the Czech Republic and, to some extent, Poland, buck this trend). An area extending from the UK to Italy seems to include most of the European developers (with countries such as France, the Benelux, Germany, Austria, Switzerland, and the Czech Republic). The Nordic countries also have high concentrations (with respect to population). In Southern Europe, Spain seems to be the country with the most developers.

Although some successful libre software projects could be considered European by origin, or (in some rare cases) because almost all developers are European, this identification is tricky: libre software projects are global by nature (any developer from any part of the world can, in principle, join in). Therefore, we will not try to identify strictly European projects.

On the contrary, infrastructure to host libre software projects can easily be assigned a nationality (although usually it is open to the world). In this respect, Europe is well behind the U.S., since SourceForge and other well-known hosting sites are physically located there. But some hosting facilities do also exist in Europe, usually focused on a given national or linguistic community. As an example, we can cite BerlioOS , the largest one (http://berlios.de; in Germany, with more than 2,400 projects and almost 12,000 developers registered), Software-libre.org (http://software-libre.org; Spain, 66 projects and 234 developers), and Gna! ( http://gna.org; France, 360 projects and almost 2,000 developers). All of these figures (checked during March 2005) turn pale in comparison to SourceForge. This is one of several cases where, despite the importance of the European developer community, its infrastructure is not at the same level.

The Organization of the Community

The libre software community in Europe, despite being healthy and full of life, is also fragmented, reflecting the cultural, national, and linguistic diversity we have in the continent, but also lacking the strength and power it could have if it were more coordinated. However, more and more links (usually at the personal level) are being established, and something similar to a really European libre software community seems to be emerging. Although it is still some distance from speaking as loudly as it could—for instance, behind the European Union institutions, we are seeing more and more transnational initiatives—as with any other community formed around the Internet, it has its own virtual and real-world meeting points, such as news sites, conferences, and associations. However, as we will show in the next paragraphs, most are closely linked to relatively small geographic areas, and can hardly be identified as being "for Europeans."

Many news sites are strongly related to libre software. Maybe the most transnational one is The Register (http://theregister.co.uk) (probably because it runs in English), which of course, is the most popular in the UK. Although it carries news on many IT-related subjects, libre software receives more than reasonable coverage. Similar things can be said for Heise Online ( http://www.heise.de), but applied to the German-speaking community. Other sites worth mentioning are Barrapunto (http://barrapunto.com;Spanish), LinuxFr (http://linuxfr.org; French), Linux.pl ( http://linux.pl ; Polish), gildot (http://gildot.org; Portuguese), and Svenska Linuxföreningen and Gnuheter ( http://gnuheter.com ; Swedish, both read also by Norwegian and Danish speakers). Some of them are built as Slashdot-like sites, others are not, but each has its own flavor. Most of them are helping to foster a sense of community and are even assisting in the foundation or consolidation of more formal organizations related to libre software. There have been some attempts to establish a "European Slashdot," but without success (for now). Maybe that is simply not possible where there is such cultural and linguistic diversity.

If we consider real-world meeting places, we can find some that come close to being truly European. Probably the clearer examples are FOSDEM[5] (Free Open Source Developers Meeting), which is especially oriented to libre software developers and has for several years been pulling together a good quantity of European hackers (also attracting some from other regions), LinuxTag,[6] which is more oriented toward users and companies, although also attractive to developers, and LSM/RMLL[7] (Libre Software Meeting/Rencontres Mondiales du Logiciel Libre), for developers specifically devoted to libre software. Several projects also have European meetings: GUADEC (GNOME User and Developer European Conference; http://guadec.org/), YAPC::Europe ( http://yapceurope.org ; for the Perl community), EuroBSDCon ( http://eurobsdcon.org ), ApacheCon Europe,[8] OOoCOn (http://marketing.openoffice.org/conference) OpenOffice.org Conference, not specifically European, but always held in Europe), and many others. These are usually held every year in a different location.

In addition, we have the national events: the aforementioned LinuxTag in Germany, LSM/RMLL and Solutions Linux ( http://solutionslinux.fr ; a more commercial event) in France, Linux Forum (http://linuxforum.dk ) in Denmark, Congreso Hispalinux (http://congreso.hispalinux.es) in Spain, Linuxwochen (http://linuxwochen.at) in Austria, and so on. In some countries, there are no national events, but smaller meetings are usually organized by Linux User Groups (LUGs) or libre software associations.

With respect to associations, diversity is also the rule. In addition to more or less local LUGs (spread through all of Europe), there are some organizations which either represent groups of local associations, or have a wider membership, usually at the national level. Some of them are: APRIL (Association pour la Promotion et la Recherche en Informatique Libre; http://april.org) and AFUL (Association Francophone des Utilisateurs de Linux et des Logiciels Libres; http://aful.org) in France, NUUG (Norwegian Unix User Group; http://nuug.no) in Norway, DKUUG ( http://www.dkuug.dk ) in Denmark, Atviras kodas Lietuvai (Open Source for Lithuania; http://akl.lt) in Lithuania, SSLUG (Skåne Sjælland Linux User Group; http://sslug.dk) in Sweden and Denmark, ANSOL (Associaç o Nacional para o Software Livre; http://ansol.org) in Portugal, PLUTO ( http://pluto.it ) in Italy, AFFS (Association for Free Software; http://affs.org.uk) in the UK, PLUG (Polish Linux User Group; http://www.linux.org.pl) in Poland, and Hispalinux ( http://hispalinux.es ) in Spain. At the European level, there is also the FSFE (Free Software Foundation Europe; http://fsfeurope.org), and the corresponding FSF-related associations in some countries, such as FSF France ( http://fsffrance.org ), Verein zur Förderung Freier Software ( http://ffs.or.at) in Austria, and Associazione Software Libero (http://softwarelibero.it) in Italy.[9] Each of these associations has its own history. Some of them were formed as Unix associations (now more than 20 years old) and have evolved into Linux and libre software associations with time. Some others were formed specifically as LUGS, mostly in the mid-1990s. Still others are devoted to libre software in general, again mostly founded in the late 1990s. Membership, activities, involvement of companies, etc., vary a lot from one to another.

On many occasions, there have been discussions suggesting the convenience of a European umbrella association devoted to libre software, which could speak on behalf (and be some kind of federation) all of these organizations. But so far, none has crystallized.

In addition to these "community" associations, there are also some others representing corporate interests. Among them, probably the most notable is Open Forum Europe (http://www.openforumeurope.org), which has backed some actions related to libre software promotion in Brussels, but performs most of its activity in the United Kingdom.[10]

Libre Software in the Private Sector

Since the early 1990s, it became clear to some European entrepreneurs that libre software was interesting for business. Approximately at the same time in other parts of the world, companies started to use libre software, and a new market niche for companies providing services based upon libre software emerged. During the mid- 1990s, many small companies started to offer services based on the then-new Linux-based operating system (MandrakeSoft in France and SuSE in Germany are well- known cases that will be discussed later in more detail, but they were not unique). Some other companies focused on the libre software infrastructure of the Internet or on specific products (such as Trolltech AS in Norway and MySQL AB in Sweden). By the late 1990s, interest in libre software was strong enough to maintain several mid-size companies devoted (completely or in part) to generic consultancy, support, and development of libre software, such as Alcove in France, Andago in Spain, and ID Pro AG in Germany.[11] Later, starting in the early 2000s, large companies began to show interest in libre software, especially in the secondary sector (these were intensive users of software, although that was not their main line of business—such as telecommunications, automotive, aerospace, and banking). In 2005, the list of large European companies with a significant use of libre software is too long to include here.

A good exponent of the interest of European industry in libre software is the ITEA Report on Open Source Software,[12] published in January 2004, which is aimed at elucidating the libre software world (from legal and economic, to development and quality issues), and uncovering business opportunities and issues to be resolved in relation to it. The following sentences, taken verbatim from the report, may provide some insight on the opinion of those drafting it (and maybe on the view of the companies for which they work):

Depending on the business they are in, companies are likely to have different reasons for using OSS. In some cases, OSS can help them lower the cost of the service, system or product that they offer. In others, their contribution to OSS can help to establish new standards worldwide. By carefully applying license conditions it is certainly possible to derive considerable benefits from OSS, while minimizing the risks. Open Source Software is not the "magic bullet" to solve Europe's software development competitiveness. However, OSS is an important new development and an interesting option for software-intensive systems.

Different sources provide different figures for the market share of libre software in Europe, but they are consistent in indicating its continuous growth. The market for libre software in Germany in 2003 was estimated, according to a study by Soreon Research,[13] to be in the region of EUR 131 million (with a projection of more than EUR 300 million for 2007). The manufacturing industry was the one with the highest penetration, with 18% of the companies making significant use of libre software. The structure of the market was heavily based on support services (EUR 81 million) and training (EUR 27 million), and direct sales of software accounted for only EUR 10 million. However, a year later IDC[14] estimated that USD 98 million would be spent on IT services for libre software in the whole of Western Europe, and predicted that USD 228 million would be spent by 2008 (for the same region).

To provide an overview of the landscape of European libre software companies, we have selected a short list of them, probably the better-known ones. This list is, of course, not exhaustive, but will hopefully be illustrative enough to show the European contribution to the world of libre software business models:

  • SuSE[15] was one of the first companies providing services for Linux-based distributions, almost since it was founded in 1992 in Germany, as S.u.S.E., by Hubert Mantel, Burchard Steinbild, Roland Dyroff, and Thomas Fehr. They started by selling floppies of Slackware Linux partially translated into German. Later, they decided to build their own distribution, which was released in 1996 and quickly becoming the most-used GNU/Linux distribution in Germany. Based on the success of this distribution, the company grew, with a business model that also included support, training, and consultancy for libre software. It had a workforce of more than 500 and had one of the more well-known GNU/Linux distributions in the world when it was acquired by Novell in 2003, for USD 210 million. SuSE has also been known for its contributions to the libre software community, supporting or directly contributing to many projects, from KDE to ALSA to the Linux kernel itself.
  • MandrakeSoft (http://mandrakesoft.com) is a French company with major operations in the United States and other countries. It was funded in 1998 by Gaël Duval, Frédéric Bastok, and Jacques Le Marois, after the success of the first release of Mandrakelinux some months before. One of its innovations was to let people download the entire distribution from the Net, which assisted its spread around the world. Agreements with the Pearson Technology Group (then Macmillan Software) and other distributors helped too and, by 2000, had made it one of the companies with a larger market share in the libre software segment in the United States and France, having gone from three to about a hundred employees in just two years. In 2004, the company had revenues of more than EUR 5 million. Currently, its business model seems to be mainly linked to its distribution, although it also provides consultancy and support services for libre software in general. MandrakeSoft has recently acquired Conectiva, one of the other companies producing a major GNU/Linux distribution, and has changed its name to Mandriva ( http://www.mandriva.com ).
  • Open CASCADE S.A. ( http://opencascade.com ), now in the AREVA group, is a French company providing services around the Open CASCADE system (a set of 3D modeling components and libraries) and SALOMÉ (a platform for the integration of numerical simulations), both distributed as libre software. The company is rooted in Matra Datavision, also French, which was a major player in the CAD/CAM market from the early 1980s to the mid-1990s. One of its lead products was CAS.CADE (Computer Aided Software for Computer Aided Design and Engineering), released in 1993. In 1999, after the decision to change the focus of the company from products to services, Matra Datavision decided to release CAS.CADE as libre software, under the name Open CASCADE. One year later, the company of the same name was segregated and acquired by Principia (in the AREVA group). Open CASCADE S.A. is now focused on providing customized development, training, consulting, and other services, with a team of about 80 developers. The company is also fostering the building of communities around its products, channeled through a specific site ( http://opencascade.org ).
  • Trolltech AS ( http://trolltech.com ) is a Norwegian company well known for producing Qt, an essential component for KDE and many other systems. It was founded in 1994 by Haavard Nord and Eirik Chambe-Eng, with the aim of building cross-platform C++ GUI tools. Now a company with more than 90 employees, it pioneered a dual-licensing model for Qt (and other products). In the beginning, this software was gratis but not free. However, after the launch of the GNOME project (promoted by the FSF, among others, and backed by some companies worried about the dependence of KDE on the proprietary Qt), Trolltech decided to distribute it under QPL (a libre software license). Finally, Trolltech moved to the GPL (for the X11-based version of Qt), and gave those not willing to comply with the GPL the option of purchasing proprietary licenses.
  • MySQL AB (http://mysql.com) is the company that owns the code for the MySQL database server. It was founded in Sweden in 1995 by David Axmark, Allan Larsson, and Michael Widenius (the founders of the MySQL project) and has since opened offices in many countries. Its business model is based on selling support and services for MySQL, and selling licenses to those unwilling to fulfill the conditions of the GPL (that is, dual licensing MySQL). This is why my SQL AB has been careful to maintain the ownership of the code, by having all the developers of MySQL as employees of the company. MySQL AB has run on venture capital since 2001, had revenues of about EUR 15 million in 2004, and about 160 employees.

To complete the vision of this landscape, let us introduce some examples of large European companies involved in libre software. The examples were chosen at random from our experience and therefore are not representative, but again, hopefully illustrative of what is happening out there:

  • ObjectWeb ( http://objectweb.org ) is a consortium created in 1999 in France by Bull (http://bull.com), France Telecom R&D (http://rd.francetelecom.com), and INRIA (http://inria.com) to develop libre software middleware, ranging from specific software frameworks and protocol implementations (such as CORBA) to integrated platforms. In 2002, it evolved into an international and independent nonprofit organization open to companies, institutions, and individuals. The software developed by ObjectWeb includes more than 40 products in the application platforms, workflow engines, IDE plug-ins, and software engineering domains. ObjectWeb uses GPL and LGPL licenses, and has managed to create a large community of companies providing services around those products.
  • Based in Spain, Telefonica Investigacion y Desarrollo (TID (http://www.tid.es) a subsidiary of Telefonica ( http://telefonica.com), one of the largest telecommunication companies in the world), launched in late 2004 the Morfeo Project (http://morfeo-project.org), in collaboration with some other Spanish companies, universities, and public administrations. Morfeo is a framework for distributing and developing as libre software some products that TID either has produced or needs, mainly in the field of platform software (middleware, workflow, communications, etc.). It has already released products such as CORBA systems, and is trying to build a community of developers. In the long term, this could be a first step toward a strategy based on libre software for some of the activities of TID.
  • Ericsson (http://ericsson.com), based in Sweden and not especially well known for its contributions to libre software, has distributed several products as libre software. Ericsson's implementation of the Erlang programming language[16] was released in 2000, and has an active developer community. Erlang provides facilities for concurrency, distribution, robustness, and soft real-time processing. Another contribution is TIPC (http://tipc.sourceforge.net), a protocol for intracluster communication implemented as a loadable module for Linux.
  • Nokia (http://nokia.com), based in Finland, and also not normally known as a libre software producer, has distributed some software under NOKOS (the Nokia Open Source License, an OSI-recognized open source license). But recently another event related to the use of libre software by Nokia hit the news: the availability of a Python environment for some of its products (http://www.forum.nokia.com/python) using the Symbian OS, in what could be a strategy of letting libre software developers in the Python community build applications for Nokia devices.
  • Symbian ( http://symbian.com ), with headquarters in the UK and owned by Nokia, Siemens, Ericsson, and others, is releasing one of its products for Symbian OS, the Open Programming Language (OPL; http://opl-dev.sourceforge.net), as libre software, using the LGPL. OPL is a BASIC-like language used in Symbian OS phones for rapid prototype development.

These examples show how, despite their general strategy of being more or less oriented toward proprietary software, many European companies are experimenting with libre software models and, in some cases, are considering new lines of business based upon them. In fact, similar cases of exploration of the libre software world can be found in almost any medium to large-size company heavily involved in the software business.

Public Administrations and Libre Software

Before the year 2000, libre software was almost completely off the political radar in Europe. But since then, and with widely varying intensities, it has entered the political agenda in many European cities, regions, and countries. In some cases, it is considered as one possible choice for public administrations in their role as intensive users of services based in software. In some others, it is deeply linked to efforts to promote the information society. In this respect, some public administrations are actively proposing libre software as a viable alternative for citizens and companies in their area of influence. Finally, some legislative bodies have also considered law proposals which deal specifically with libre software.

Initiatives are happening at all levels: European Union institutions, national governments, regional administrations, and municipalities. However, the situation varies a lot from country to country and from region to region. All in all, our feeling is that we have reached a point in Europe where it is strange to find institutions that have not at least considered the use of libre software. In fact, several studies signal the public sector as one of the driving forces behind libre software in Europe for the coming years. In the rest of this section, we will take a look at some of these initiatives.

Actions by the European Commission

The European Commission (the institution in the European Union most similar to an "executive branch") has promoted several actions related to libre software. By 1999 an informal group of experts, the European Working Group on Libre Software,[17] was meeting in Brussels at the request of some officials of the DG-INFO (Directorate General on Information Society). The main interest of these meetings was to explore specific opportunities for Europe in the field of libre software, to provide the Commission with some input about its impact in the European IT sector. The most widely known output of the group was the report titled "Free Software/Open Source: Information Society Opportunities for Europe?" (http://eu.conecta.it/paper.pdf ) which was presented in Brussels in March 2000. This was probably the first public activity of the Commission in relation to libre software and represented a kind of turning point which led to many other actions by European institutions. The report identified some signals that evolved later into trends. It also recommended both the consideration of libre software solutions in public administrations, and activities to inform the European industry about new possibilities. In addition, the report presented a complete landscape of the libre software world, from technical, legal, and economic points of view.

Since the days of that group, libre software has been present in several actions funded by the European Commission, which has had a policy of researching and publishing information about it, without promoting it explicitly. To describe just a few of those actions, we will concentrate on some initiatives promoted by the IDA (now IDABC) program and by the IST program (in the context of the fifth and sixth R&D Framework Programs).

IDA (Interchange of Data between Administrations) was a program of the European Union, started in 1999 and aimed at the funding, development, and coordination of pan-European services for public administrations. Since 2004, it has been continued by the IDABC program. IDA and IDABC have performed many activities related to libre software, which are referred to in the IDA Open Source Observatory ( http://europa.eu.int/idabc/en/chapter/452). Among them, the following can be highlighted:

  • "European Interoperability Framework for pan-European eGovernment Services" ( http://europa.eu.int/idabc/en/document/3761 ; 2004). This reference document on interoperability was written after an extensive consultation process. IDABC considers it the highest-ranking module for the implementation of e-government in Europe. It includes several references to libre software, among which the following can be highlighted:

Open Source Software (OSS) tends to use and help define open standards and publicly available specifications. OSS products are, by their nature, publicly available specifications, and the availability of their source code promotes open, democratic debate around the specifications, making them both more robust and interoperable. As such, OSS corresponds to the objectives of this Framework and should be assessed and considered favourably alongside proprietary alternatives.

  • "IDA OSS Migration Guidelines," ( http://europa.eu.int/idabc/en/document/2623#migration ; November 2003). One of the best guides for the migration to libre software. Specially targeted at public administrations, many of its analyses and recommendations are, however, valuable for any party considering moving from proprietary to libre solutions. It includes a detailed methodology for estimating the convenience of the migration, and for putting it into practice. It also provides complete descriptions of some typical scenarios, and configurations for the usual cases (email, desktop, server, etc.).
  • "Pooling Open Source Software, Feasibility Study," ( http://europa.eu.int/idabc/en/document/2623#feasibility ) June 2002. A study on the opportunities for sharing libre software among public administrations, from technical, legal, functional, and financial points of view.
  • "Study into the Use of OSS in the Public Sector," ( http://europa.eu.int/idabc/en/document/2623#study ) June 2001. One of the first reports on the use of libre software in public administrations. Includes some general information on libre software, and details on libre software solutions (about 100 examples showing specific systems that could be useful in the public context). It also analyzes the deployment of libre software in Europe at the time of the report, and presents some interesting conclusions.
  • Organization of meetings and symposiums for sharing experiences on the use of libre software in public administrations in different countries. These have been helpful for coordinating actions and establishing links among the promoters of different initiatives.
  • The IDA Open Source Observatory. This is worth mentioning, as it provides a good compilation of information about libre software, its situation in Europe, and many issues especially relevant to public administrations.

Within the IST (Information Society Technologies; http://www.cordis.lu/ist) research program, the Commission has funded (and is funding) several projects related to libre software. Many of them are aimed at the production of libre software in a given domain, such as AGNULA (http://www.agnula.org; libre software distributions specialized in audio and video). A detailed listing of those projects is available in the area devoted to libre software in the Information Society Thematic Portal ( http://europa.eu.int/information_society/activities/opensource/european_activities ). Some others are devoted to researching libre software as a matter of study, with the aim of improving general knowledge about it:

  • FLOSS ( http://www.infonomics.nl/FLOSS) was the first academic research of the libre software phenomenon as a whole, looking at it from many different points of view. It included studies on the developers themselves, based on a survey and on the analysis of author information in source code, focusing on sociological data about them. It was the first to provide some insight about why developers participate in libre software projects, what professional profile they have, what amount of time they devote to libre software, and where they come from. The study was also successful in the introduction of the name FLOSS (an acronym for "free, libre, open source software"), which has since been used in many other cases, especially within the research community.
  • AMOS (http://www.clip.dia.fi.upm.es/~amos/AMOS) was a project to research the feasibility of building a system capable of categorizing and allowing searches among libre software package descriptions. This is especially useful for developers looking for code to reuse in their systems.
  • COSPA (Consortium for Open Source in the Public Administration; http://cospa-project.org), started in 2003, aimed to analyze the effects of the introduction of libre software and open standards in European public administrations.
  • FLOSS-POLS (Free/Libre/Open Source Software: Policy Support; http://www.flosspols.org) is a project started in 2004 as a follow-up to FLOSS, and includes research tracks on government policy toward libre software, gender issues, and the efficiency of libre software development methods for collaborative problem solving. FLOSS-POLS will also deal with libre software in e-government, and will look for feedback from governments in relation to policies about libre software.
  • CALIBRE is an action to coordinate the research on libre software in Europe, and to help transfer its results to industry. It started in mid-2004, and has already organized several conferences with a special focus on showing the results of libre software engineering, or on the use of libre software within European industry.

In the following years, it is expected that more and more projects related to the study of libre software will be approved in future calls, in what seems to be a growing interest by the European Commission research work programs to understand how libre software works from several points of view.

National Initiatives

At the national level, the situation is different from country to country. And even among those who have started some kind of action related to libre software, approaches are diverse. However, some common patterns can be identified. It is unusual to find a national government that has not issued studies and recommendations for the use of libre software in public administrations. There is also a certain consensus on some matters that have been proposed again and again, such as consideration of libre software for public acquisitions, adherence to open standards and interoperability, the need for inspection, and the importance of retaining proprietary rights on software. In any case, the following brief descriptions of the state of affairs in several countries should show both consensus and diversity of approaches:[18]


France
ADAE (Agence pour le Développement de l'Administration Électronique, Agency for the Development of the Electronic Administration; http://www.adae.gouv.fr/index.php3), formerly ATICA, maintains a good deal of information related to libre software in public administrations, and organizes activities related to that topic. It has also published several reports of special interest. Among them, it is worth mentioning the "Guide de choix et d'usage des licences de logiciels libres pour les administrations"[19] ("Guide to Choosing and Using Free Software Licenses for Government and Public Sector Entities), a complete guide to the legal implications of using libre software licenses, either for external software (obtained with or without cost) or for software produced by the administration itself (recommending the GPL in this case). There are also some recommendations for the promotion of libre software in the framework of e-administration programs, and cases of large-scale deployment, mainly of OpenOffice.org, in French public administrations.

Also in France, two of the first proposals of laws related to the use of libre software in public administrations were produced. In 1999, Laffitte, Trégouet, and Cabanel drafted in the French Senate the 2000-117 law project, aimed at enforcing the use of libre software in public administrations in those domains where technical solutions were already available (considering a whole set of exceptions and temporary measures to facilitate the transition period). It also considered the creation of a Libre Software Agency, funded by the government, which would help public administrations in the deployment of libre software technologies. In 2000, another law was proposed by Jean-Yves Le Déaut, Christian Paul, and Pierre Cohen. It was similar in objectives and rationale, but was not compulsory about the use of libre software in public administrations, but more focused on the availability of source code for applications and on the principle of "right to compatibility of software," which aims to guarantee the interoperability principle, common in European legislation. Although neither of these projects was approved, both have influenced later law initiatives in many other countries.

United Kingdom
There have been several studies and pilot experiences, which led in 2002 to the publication by the OGC[20] (the Office of Government Commerce) of a formal policy on the use of libre software, "Open Source Software: Use within UK Government" (http://www.ogc.gov.uk/oss/OSS-policy.pdf) which mandates not only the consideration of libre software in procurements, but also that decisions must be made considering "value for money" (a policy that has been widely copied in many other countries, especially in the developing world). At the same time, it establishes a policy of avoiding lock-in by proprietary software providers, supporting open standards and specifications, and exploring the use of libre software licenses for dissemination of research and development funded with public money. A document on implementing this policy was also published. "Guidance on Implementing OSS" ( http://www.ogc.gov.uk/embedded_object.asp?docid=2498 ) provides details on how and when to consider libre software, and includes a detailed study of proprietary software lock-in practices and how to avoid them.

Some trials have also been conducted, with an interesting final report about them, "Open Source Software Trials in Government: Final Report" ( http://www.ogc.gov.uk/embedded_object.asp?docid=1002367 ; published in late 2004), which states the viability of libre software solutions, with different perceived levels of maturity depending on the area of implementation.

In March 2005, the Open Source Academy was announced ( http://www.egovmonitor.com/node/319 ). Funded by the UK government, supported by several municipalities and other institutions (including OpenForum Europe and Open Source Consortium), and with the help of the private sector, it is aimed at the promotion of the use of libre software in local government.

Germany
The German federal government produced in 2002 one of the first official documents dealing with the use of libre software in public administrations, "Open Source Software in der Bundesverwaltung"[21] ("Open Source Software in the Federal Administration"), by the KBSt (Coordination and Advisory Agency of the Federal Government for Information Technology). Later, in 2003, it produced the "Migration Guidec", one of the more complete documents about how to migrate to libre software, including detailed technical information about possible paths for migration in several domains. The German government has also funded some libre software developments, or improvements to existing systems, that were critical for its IT strategy. In 2005, it announced the Open Source Software Competence Centre,[22] a web site aimed at spreading best practices regarding the use of libre software in the public sector.
Italy
In February 2004, the Italian government issued rules regarding the use and acquisition of libre software in public administrations, "L'Open source nella pubblica amministrazione."[23] hey were formally made public after the release of an official report, "Indagine conoscitiva sul software a codice sorgente aperto nella Pubblica Amministrazione,"[24] which presents some rather interesting conclusions on the characteristics of the use of software by public administrations. The rules set the criteria to consider when acquiring software (which include interoperability, nondependence on a single provider or on proprietary technologies, availability of code for inspection, etc.), and specifically included libre software as a possible choice.
Finland
A working group promoted by the Finnish Ministry of Finance produced in 2003 the report "Recommendation on the Openness of the Code and Interfaces of State Information Systems,"[25] which (among other interesting recommendations) proposes the consideration of libre software for the custom developments funded by the public administration and for the acquisition of software.
Denmark
The strategy of the Danish government with respect to libre software is exposed in the report "Danish Software Strategy,"[26] officially adopted in 2003. In summary, the approach is based on the principle of obtaining the maximum value for money, irrespective of the type of software (which is also the reason several detailed studies on total cost of ownership in libre and proprietary software scenarios are being carried out), but also not forgetting the importance of promoting competition, interoperability, and flexibility. Some preliminary studies performed prior to this report showed potential major savings could be made through the use of libre software.[27]
Sweden
In 2003, Statskontoret, the Swedish Agency for Public Management, published "Free and Open Source Software, a Feasibility Study,"[28] a complete and detailed study of libre software, including cases in Swedish public administrations, with very positive conclusions.
Spain
Many of the interesting developments related to libre software in Spanish public administrations have been achieved at the regional level. However, there are also some interesting actions by the national government. One of the most revealing is the inclusion of measures related to libre software in the document "Criterios de seguridad, normalización y conservación de las aplicaciones utilizadas para el ejercicio de potestades"[29] ("Criteria of Security, Standardization and Conservation for Applications Used in the Exercise of Authority"), edited by the Consejo Superior de Informática (Higher Advisory Board on Informatics), an interministry body of the Spanish administration. It details the issues to consider for all the applications used by the public administration and recommends specifically the use of libre software whenever technically feasible. It also recommends requiring the availability of source code for programs acquired by the administration, the use of open formats, and the use of libre software applications to access some kinds of data.
The Netherlands
The OSOSS program ( http://www.ososs.nl ) is aimed at encouraging the use of libre software and open standards in public administrations. OSOSS is a program of ICTU, the national organization for IT in the public sector, founded by the Ministry of the Interior. In the context of this program, libre software is promoted as a full-fledged option. OSOSS is basically an informative advisory body, supporting policy makers in exploring the relationship between libre software and public administrations.
Norway
The issue of libre software in the public administration has been dealt with by the Norwegian Board of Technology (a public, independent think tank on technology) in its report, "Software Policy for the Future"[30] (November 2004). It recognizes the potential interest of libre software, and recommends a policy similar to that of Denmark: pilot programs and careful case-by-case studies.

Other Initiatives in the Public Sector

There are many other initiatives in the public sector. Among them, we have selected a short list which we have found especially meaningful as illustrative examples of the whole landscape:

  • Extremadura (Spain) is a small, with a population of about a million, and relatively cash-poor region that has defined a strategy based on libre software to catch up on information society issues. The main principles of this strategy are connectivity and IT literacy for all citizens. One of the key projects for implementing it is gnuLinEx (http://linex.org), a GNU/Linux distribution originally targeted for primary and secondary education (deployed in tens of thousands of computers in all public schools), but which is now also used in the public administration and offered to SMEs and individuals. One of the latest initiatives announced by Extremadura (jointly with Lambdaux (http://lambdaux.com), a Spanish libre software company) is the CompatibleLinux (http://compatiblelinux.org) catalog, an analysis of the hardware available in the market with respect to its compatibility with GNU/Linux distributions. This initiative has also led to the AENOR[31] compatibility certificate, which can be specified by public administrations and companies seeking to purchase hardware for use with GNU/Linux distributions.
  • The French police (Gendarmerie Nationale) started a plan in 2004 to switch to OpenOffice.org in all its desktop machines[32] (about 80,000 PCs). They expect to complete the switch by the end of 2005. They estimate savings at about EUR 2 million.
  • The Kolab Groupware Project (http://kolab.org) was initiated in 2003 as a spinoff of the Kroupware contract (http://kroupware.org), which was funded by the German Bundesamt für Sicherheit in der Informationstechnik (BSI, Federal Agency for IT Security) and won by a consortium of three companies: erfrakon, Intevation, and Klarälvdalens Datakonsult. Kolab is today a libre software system that allows for the interaction among mixed groupware environments: KDE, Outlook, and web-based tools. This is one clear case of the promotion of a new libre software project by a public administration (in this case, because it was interested in overcoming this missing functionality in the libre software world).
  • The city of Munich (Germany) started in May 2003 a plan to migrate to GNU/Linux and libre software (LiMux)[33] most of its desktop machines (some 14,000 PCs). This initiative, which started as a political one (including in the process a delay to get attention to the proposed European directive on software patents, during the summer of 2004), is backed by detailed studies and has had a lot of media attention. Despite this attention, the project had not been completed at the time of this writing. However, it seems to have started a trend followed by some other European cities (although there are also earlier cases, such as the city of Florence, which passed a law in 2001 mandating the use of libre software when feasible[34]).
  • Rijkswaterstaat[35] (Directorate for Public Works and Water Management, the Netherlands) has been using the Geoservices system, heavily based on libre software, since 2003. Rijkswaterstaat has the responsibility of maintaining dikes, roads, bridges, and canals, and uses Geoservices for web-based access to geo-information obtained from many different sources.
  • The Junta de Andalucia (regional government of Andalucia, the most populated Spanish region) has instigated two of the few laws related to libre software that are actually in force. The first one was the "Decree of measures to push the knowledge society,"[36] which deals with (among other issues) the use of libre software in education. It fosters the use of libre software in public schools (not mandating it exclusively) and mandates that computers purchased for that use be compatible with libre operating systems. The second law is an order approved on February 21, 2005,[37] which mandates the distribution as libre software of any program owned by the Junta de Andalucía. This order basically amounts to releasing a large quantity of code to the libre software community, doing the same for new programs built on behalf of the Junta. To the knowledge of the authors of this chapter, this approach is completely novel and marks the beginning of a new path in the promotion of libre software by public administrations.
  • Bergen, the second largest city in Norway reported a strategy, already deployed in large part, of using Linux on servers[38] (including the servers of the network for schools). The experience seems to allow for cost cuts both in hardware and in software, and includes the wide use of libre software instead of proprietary solutions for many services.
  • The city of Vienna (Austria) announced in early 2005 a plan offering its departments migration to OpenOffice and GNU/Linux on the desktop.[39] The migration plan is voluntary, linked to lower costs charged by the city's IT department, and is currently targeted for about 7,500 desktops. For this solution, a Debian-based distribution (Wienux) has been created.

Legal Issues

Legal issues are still largely undecided for libre software worldwide. However, some of these issues are specific to Europe. Among them, we have selected two cases: the European Union directives (affecting most of Europe) that have (or may have, depending on approval) a negative impact on libre software; and the concerns about the validity of libre software licenses within European jurisdictions.

EU Directives with Negative Impact

For sure, not all legislative initiatives in Europe in the field of software have a neutral impact on libre software. In some other cases, important laws have been passed (or are in the process of approval) that cause serious problems by producing an environment hostile to libre software. Two of the most relevant cases are:


Directive on software patents[40]
Although software patents may affect any kind of software, the libre software community is especially concerned about the problems it poses for the freedom of innovation. The directive on software patents (actually, "Directive About Patentability of Computer Implemented Inventions") was proposed by the European Commission in February 2002. If approved as such, it would mean the introduction of software patents in Europe very similar to those in the United States.

Early in the process, groups all over Europe started to explain the problems this directive would cause to European software developers (be they individuals or companies) and users. In part thanks to the awareness caused by this campaign, led to a great extent by libre software activists, the European Parliament passed in September 2003 a set of amendments which, together, would amount to the invalidation of software patents in Europe. Meanwhile, the European Council of Ministers (representatives of EU national governments) has approved a text even more radical than the proposal of the Commission, allowing more clearly for software patents, in a rather strange meeting in March 2005 ( http://kwiki.ffii.org/Cons050307En ). Things are in quite a mess at the time of this writing. The Parliament asked the Commission to withdraw its proposal, what was refused, entering into the "second reading" stage, which will lead to a new vote in Parliament, probably during the summer of 2005. All in all, this directive proposal is having one of the more complex, strange, and time-consuming paths ever seen in Brussels. There is a strong perception in the libre software community in Europe that the introduction of software patents would be a strong barrier to the development of libre software. On the contrary, if Europe were to remain free of software patents, libre software development would benefit from much more legal certainty, a friendlier environment, and a more level playing field.

The European Union Copyright Directive (EUCD)[41]
This was approved in 2001. It is in many respects similar to the Digital Millenium Copyright Act (passed in the U.S.). It poses risks for libre software: the impossibility of distributing programs for handling certain file formats (for contents subject to the EUCD), and of interoperation with certain systems. This directive can make it illegal to produce libre software programs for handling DVDs, for instance. The libre software community is concerned about this problem, but is not mobilized to the same extent as in the case of software patents.

Although there are other legal initiatives in Europe hostile to libre software, these two are the more well-known ones. In particular, there is now discussion about DRM systems and the legislation surrounding them, which could develop into very dangerous laws making it impossible for a libre software system to handle content such as e-books, movies, and music. Only the future will say whether the legal environment that was, until the late 1990s, basically neutral to libre software will evolve into a fairly hostile one.

Libre Software Licenses in Europe

Most (in fact, almost all) libre software licenses were formed in the United States, in a jurisdiction alien to European countries. Therefore, for many years now, there have been concerns about the validity of those licenses in European jurisdictions. This has spawned many efforts in different directions: either to assess the validity of, or to translate and localize such licenses, having versions valid in every European country. Until now, the former approach has had more impetus, as is shown, for example, in the aforementioned "Pooling Open Source Software" study, which includes a detailed review of the validity of the GPL and other libre software licenses, concluding that it is valid for practical purposes. The latter approach is tricky, since it could contribute to the fragmentation of the libre software world and to endless problems in cases of international collaboration (so common in libre software projects). That is why many people from the legal community are considering the proposal of international regulations, which would complement intellectual property treaties with the consideration of libre software, given clearer international support for these licenses.[42] However, two of the clearer cases of the validity of libre software licenses in Europe happened in Germany. On April 14, 2004, a German court granted a preliminary injunction to stop distribution (by a company called Sitecom; http://www.sitecom.com) of a router that included code (Net-filter/IPtables) licensed under the GPL, yet failed to comply with its provisions, because Sitecom did not distribute the source code. This preliminary injunction was confirmed on July 23, 2004,[43] along with a significant judgment, after which Sitecom started to provide the source code on its web site. The second case was also a preliminary injuction ( http://yro.slashdot.org/article.pl?sid=05/04/14/2024258 ), also for the use of Net-filter/IPtables code in some firewall products distributed by Fortinet.( http://www.fortinet.com ). Both cases have lent weight to the GPL worldwide, but particularly in the German jurisdiction (and in other European jurisdictions of similar tradition).

Libre Software in Education

One of the fields where libre software has entered with most impetus in Europe is education. This does not mean that libre software is mainstream in European educational institutions, but that there are several very clear examples that seem to have been successful and that are currently being considered in many other realms. For several reasons—the specific advantages of libre software in the education field, the importance of localization, the lack of suitable tools for many educational tasks, the funding problem so ubiquitous in education, and the readiness of large parts of the educational community to accept and embrace its assumptions and philosophy—this field seems to be especially receptive to libre software.

For illustrating this rich landscape of experiences, we have selected four examples that have come to our attention:

  • SkoleLinux ( http://skolelinux.org) is a successful case of a grass-roots effort to bring libre software to the education world. It was formed in 2001 as a project for developing software systems for schools in Norway. It was originally aimed at the localization of a GNU/Linux distribution for that country (mainly by translating it into Norwegian written languages), to improve the installation and maintenance so that it would be suitable for the needs of schools (including distributed administration), and to promote the introduction of the product in Norwegian schools. In this respect, it has been successful, being used in many schools in Norway and other countries, with a healthy community of developers and users around it. The project has been funded by a loan from the SLX Debian Lab Foundation, which pays for three employees and has strong relationships with the Debian project.
  • gnuLinEX ( http://gnulinex.org) is promoted by Junta de Extremadura, the regional government of Extremadura, Spain. It is a part of a larger project (already mentioned in the section about public administrations). gnuLinEX is a Debian-based distribution, completely localized, which is currently deployed in the whole public education system of the region (about 66,000 computers in 2004, mainly in schools) and is now being considered for other kinds of environments. Teaching materials and specific applications for education are also being developed (usually under libre software or libre documentation licenses). A complete strategy encompassing training, support, development, and dissemination within the society of Extremadura is also being put into practice. The project started in 2002 and is evolving into a complete strategy for the promotion of the use of information technologies based on libre software. gnuLinEX was the first of a series of education-oriented libre software distributions that have been deployed in many other Spanish regions.
  • AbulEdu (groupe Éducation de l'ABUL; http://abuledu.org) is a French project oriented toward the use of libre software in schools. Its best known product is a GNU/Linux distribution, completely in French, developed mainly by volunteers. It includes many educational software products, and is designed to be simple to maintain in the environment usually found in classrooms. It is currently in use in many schools all over France.
  • SIGOSSEE (Special Interest Group on Open Source Software for Education in Europe; http://ossite.org), co-funded by the European Commission, has been established to investigate, inform, and advise about libre software in education. It is a kind of umbrella project providing a common space for many working groups, organizing many conferences, workshops, and seminars, and acting as a framework for relationships with other projects (such as JOIN, devoted more specifically to libre software learning management systems). This is a good case of a mixture of grass-roots and government-promoted efforts and has been successful in disseminating the advantages of libre software for educational organizations all around Europe.

Another interesting development related to education has been observed over the last two years: the appearance of studies specifically oriented toward explaining the libre software phenomenon, usually from many different angles, including technical, economic, legal, and sociological. We are not referring here to technical courses about software systems which happen to be libre, but to studies about libre software itself, which are usually aimed at developing an understanding of the complex interactions between technology, development processes, business models, licensing schemas, volunteer motivation, etc., which are inherent in libre software. Those would be needed, for example, to drive the libre software strategy of a company. At the time of this writing, we know about some master's-level programs which point in that direction: those delivered by Universidad Oberta de Catalunya[44] (Spain, started in 2003), and Universitá di Bologna[45] (Italy, started in 2004). More programs are due to start in 2005. And an informal group of universities, the MoLOS group (Master on Libre, Open Source Software), is designing a curriculum suitable for being taught as a master's study in the context of the new European Higher Education Space.

Research on Libre Software

An active research community is concerned with libre software in Europe. From sociologists and economists to software engineers, the interest in studying and understanding this phenomenon is on the increase.

One of the first projects specifically devoted to analyzing the libre software world was the aforementioned FLOSS Survey and Study (led by Rishab Ghosh, University of Maastrich, Netherlands, and finished in 2002). It opened several lines of research, from authorship of libre software code to motivations of libre software developers. Some other pioneering works were performed by Stefan Koch ( http://wwwai.wu-wien.ac.at/~koch/uni.html ) in Wirtschaftsuniversität Wien (Austria), who in 2000 was already studying the GNOME project from a quantitative point of view, and by the group to which the authors of this chapter belong, at Universidad Rey Juan Carlos (Spain), who were studying Debian at the same time.

Also in 2001, one of the first research workshops on libre software engineering took place, the Workshop on Open Source Software Engineering (http://opensource.ucc.ie/icse2001), organized by Joseph Feller and Brian Fitzgerald (both then at University College Cork, which hosts an active group on libre software engineering ( http://opensource.ucc.ie ); Brian is now at the University of Limerick, also in Ireland) and Andre van der Hoek, and continued every year since. It is interesting to note that in that many of the papers presented at the workshop were by European groups, even though it was held in Toronto.

Since those early day's many research groups have joined this field in Europe. Just to name a few of those researching libre software as a matter of study, we can mention (in no particular order):

  • The FLOSS group at MERIT, University of Maastrich (http://www.infonomics.nl/FLOSS; the Netherlands; focus on the economics of libre software and the motivations of developers)
  • The Software Engineering Group at Aristotle University of Thessaloniki ( http://sweng.csd.auth.gr; Greece; strong emphasis on studying the development processes)
  • The Open Source group at University College Cork (http://opensource.ucc.ie; Ireland, research on libre software processes)
  • The Software Engineering team at the Department of CSIS, University of Limerick (http://www.csis.ul.ie; Ireland; focus on processes, organization, and coordination)
  • The Libre Software Engineering group at the University Rey Juan Carlos (http://libresoft.dat.escet.urjc.es; Spain; focused on the quantitative and qualitative analysis of projects)
  • The Distributed Software Engineering Group at University of Lincoln ( http://facs.lincoln.ac.uk/Research/Distributed ; the UK; focused on the relationship of libre software development and agile methods, and on its distributed component)
  • The Software Engineering Group at Politecnico di Torino (Italy; focus on evolution and maintenance)
  • The Center for Applied Software Engineering at Free University of Bolzano-Bozen (Italy; research on metrics applied to libre software and its relationship to agile methods)
  • The team at the Institute of Computing Science at Poznan University of Technology (http://www.cs.put.poznan.pl; Poland; focus on data mining of publicly available information)
  • The Open Source Group at the University of Szeged (http://www.inf.u-szeged.hu/opensource; Hungary; publishing on quality and complexity metrics)
  • The Science and Technology Policy Research team at SPRU, University of Sussex (http://www.sussex.ac.uk/spru/1-4-9.html; the UK; focus on the economics of libre software development)
  • The Open Source research team at Technical University of Berlin ( http://ig.cs.tu-berlin.de/forschung/OpenSource; Germany; research on economics and politics of libre software)

There are, of course, many more research groups, and not finding one here implies nothing but my poor knowledge (please, forgive me if you are one of those not named). In particular, note that only groups, and not individuals, have been mentioned.

Some of these groups are partners in the CALIBRE ( http://calibre.ie ) coordinated action (already mentioned in the section about public administrations), funded by the European Commission and aimed at coordinating some of the research on libre software being performed in Europe and transmitting its results to industry.

Although it is difficult to tell, we think that European research on the libre software phenomenon is at a very high level, and when compared to similar efforts in other parts of the world (mainly in the United States), it may be more focused on understanding how libre software projects work (whereas in other cases, the understanding is more a side effect of analyzing software development in general).

The Future Is Hard to Read....

We have tried to show how libre software is flourishing all over Europe. Of course, there are many differences throughout an area where diversity is the rule, but also many coincidences. For now, Europe is an important pillar of the libre software world, and is maintaining an equal leadership of it. We have a large share of developers; there are companies producing, maintaining, using, and providing services; and our public administrations seem to be aware of the libre software phenomenon.

However, experiences are fragmented. Few companies are based on libre software in Europe as a whole, although an increasing number are working at the national and regional levels. The developer community, and the libre software community in general, is in fact a collection of loosely linked national or linguistic communities, with very little coordination among them. We do not have common news sites, and there are very few umbrella organizations, or even meeting events, recognized throughout Europe. The initiatives of the public administrations may be a bit more coordinated, but even those are wildly different from country to country. Maybe all this is just a consequence of the fragmentation of Europe—or maybe it is a first step toward a real European space of libre software. Whatever the reason, for now the real impact of European initiatives in the libre software world is far lower than the relative importance of libre software in Europe. In a few specific cases (such as the campaign against the directive on software patents, which is not carried only by libre software activists), we are starting to see coordinated movements that show the real strength of libre software in Europe.

In this context, we still have to wait to see whether Europe will capitalize on its current leadership in libre software penetration or, on the contrary, will lose this position in favor of other regions with a clearer and more active policies of promotion. The coming years will tell but, for now, we have the potential to be the first economic area to experiment with the benefits of large-scale deployment of libre software, creating a whole new industry around it, and promoting not only companies, but also the individual developers who are making this a real possibility.

In case libre software provides real advantages in terms of innovation, competence, and social benefits, Europe is well placed for advancing in that direction. Are those opportunities not worth exploring? Can we risk losing our advantageous position in what could be the next revolution in the information society?

Notes

  1. In this chapter, we will use the term libre software to refer both to free software (according to the Free Software Foundation definition) and open source software (according to the Open Source Initiative), except where making distinctions makes sense. Libre is a term well understood in romance languages, such as Spanish, French, Portuguese, and Italian, and is understandable for speakers of many others. It lacks the ambiguity of "free" in English ("libre" means only "free" as in "free speech") and is used by some people especially in Europe (although the term is rooted in the early U.S. free software community; see http://sinetgy.org/jgb/articulos/libre-software-origin/libre-software-origin.html for details). In this respect, it is important to notice that although the communities, the motivations, and the rationales behind "free" and "open source" are different, the software to which they refer is basically (although not exactly) the same.
  2. Gregorio Robles, Hendrik Scheider, Ingo Tretkowski, and Niels Weber, "Who Is Doing It?"; http://widi.berlios.de/paper/study.html.
  3. Rishab Aiyer Ghosh, Ruediger Glott, Bernhard Krieger, and Gregorio Robles, "FLOSS Final Report, Survey of Developers"; http://www.infonomics.nl/FLOSS/report/index.htm.
  4. "Paul A. David, Andrew Waterman, and Seema Arora, "FLOSS-US The Free/Libre/Open Source Software Survey for 2003"; http://www.stanford.edu/group/floss-us/.
  5. http://fosdem.org, which takes place in Brussels, Belgium, usually in late February.
  6. http://linuxtag.org, which takes place in Karlsruhe, Germany, usually in June or July.
  7. http://rencontresmondiales.org, which takes place somewhere in France, usually in July.
  8. h ttp://apachecon.com/2005/EU for the 2005 edition.
  9. A more complete list (which also includes some organizations that are not associations of the libre software community, but includes information from almost any country in Europe) is available at the Open Source Observatory, http://europa.eu.int/idabc/en/document/1631/471.
  10. Open Forum Europe was widely criticized by libre software advocates when its president signed a public statement in favor of software patents, in 2003 (see http://swpat.ffii.org/letters/ofeu034 and http://www.kuro5hin.org/story/2003/5/6/8355/78133).
  11. Some of these companies stopped business during the early 2000s. This was the case of ID Pro AG and Alcove (which still maintains a web site in early 2005, http://alcove.com. Andago, http://andago.com, is an example of those still remaining in the market.
  12. ITEA, http://www.itea-office.org, is a joint effort by many European companies to stimulate precompetitive research and development, specially in the field of information technologies, including in its partnership companies like Alcatel, Bosch, Bull, Daimler-Chrysler, Italtel, Nokia, and Siemens. See http://www.itea-office.org/newsroom/publications/Open_Source_Software.htm.
  13. Soreon Research "The Market for Open Source Software in Germany" (July 2003).
  14. IDC, "Services around Linux, Open Source, and Free Software—Western European Market Forecast" (June 2004).
  15. SuSEwas bought by Novell, but still maintains its original url: http://suse.com.
  16. Open-source Erlang, http://erlang.org.
  17. Although the group is no longer active, http://eu.conecta.it hosts some information about it. There is also an open mailing list, freesw, that is still used for announcements and discussions related to libre software in Europe.
  18. Some of this information was obtained from the European Information Society Thematic Portal, http://europa.eu.int/information_society/activities/opensource/cases, and from the Open Source Observatory, http://europa.eu.int/idabc/en/document/1677/471.
  19. http://www.adae.gouv.fr/upload/documents/free_software_guide.pdf, linked from http://www.adae.gouv.fr/article.php3?id_article=172, which includes a translation into English.
  20. h ttp://www.ogc.gov.uk/. Information related to libre software is at http://www.ogc.gov.uk/index.asp?docid=2190.
  21. h ttp://www.kbst.bund.de/dokumente/Publikation/,-300432/dok.htm. A long summary in English is also available at http://www.kbst.bund.de/Anlage302856/KBSt-Brief+-+English+Version.pdf.
  22. OSS-Kompetenzzentrum, http://www.kbst.bund.de/oss-cc.
  23. http://www.governo.it/GovernoInforma/Dossier/open_source/index.html.
  24. http://www.governo.it/GovernoInforma/Dossier/open_source/open%20software%20PA.pdf.
  25. http://www.vm.fi/tiedostot/pdf/en/65051.pdf, referred in http://www.vm.fi/vm/liston/page.lsp?r=65052&l=en&menu=2678.
  26. http://www.oio.dk/files/Softwarestrategi_-_Engelsk.pdf (available as a part of the Offentlig Information Online, http://www.oio.dk/software).
  27. http://www.tekno.dk/subpage.php3?article=969&survey=14&language=uk&front=1.
  28. http://www.statskontoret.se/upload/Publikationer/2003/200308A.pdf (in English).
  29. http://www.csi.map.es/csi/criterios/pdf/criterios.pdf. A summary available at http://www.csi.map.es/csi/pg5c10.htm.
  30. http://www.teknologiradet.no/files/endelig_rapport_programvarepolitikk_0066_20041109.pdf (full report, in Norwegian); http://www.teknologiradet.no/files/english_summary_041223_copy.pdf (executive summary, in English).
  31. AENOR is the Spanish standardization organization; http://www.aenor.es.
  32. http://www.solutionslinux.fr/document_conferencier/420c7d6295f27.pdf.
  33. http://www.muenchen.de/Rathaus/referate/dir/limux/89256/ (note by the city of Munich) http://news.zdnet.co.uk/software/applications/ 0,39020384,39171380,00.htm (note in ZDNet).
  34. http://www.softwarelibero.it/portale/legislazione/mozione_comune_firenze2.shtml.
  35. http://www.rijkswaterstaat.nl; report on the experience available at http://europa.eu.int/idabc/en/document/3934/470>.
  36. Decree 72/2003, March 18, BOJA of March 21; http://andaluciajunta.es/SP/AJ/CDA/Secciones/Boja/AJ-BojaPagina/2004/10/AJ-verPagina-2004-10/0,20748,bi%253D696836605883,00.html.
  37. Published in the BOJA of March 10, 2005; http://www.andaluciajunta.es/SP/AJ/CDA/Secciones/Boja/AJ-BojaPagina/2005/03/AJ-verPagina-2005-03/0,22557,bi%253D699234368885,00.html.
  38. http://www.linuxforum.dk/2005/program/slides/LinuxIBergen/Linux_i_Bergen_-_Tuftedal.ppt (slides, in Norwegian), http://europa.eu.int/idabc/en/document/3471/469 (note in IDABC), http://news.zdnet.co.uk/software/linuxunix/0,39020390,39173557,00.htm (note in ZDNet).
  39. ht tp://news.zdnet.co.uk/software/linuxunix/0,39020390,39185440,00.htm (note in ZDNet).
  40. There are many web sites with information on the directive on software patents. Probably the most complete about software patents in general (including information about the directive itself), and the European campaigns against them, is the site of the FFII group, http://swpat.ffii.org.
  41. There is information about how EUCD affects libre software in the FSF Europe web site, http://www.fsfeurope.org/projects/eucd/eucd.en.html.
  42. One of such proposed regulations is the Free Software Act, http://www.fsc.cc/node/view/69.
  43. http://www.heise.de/newsticker/meldung/49377 and http://yro.slashdot.org/article.pl?sid=04/07/23/1558219; see also http://www.oii.ox.ac.uk/resources/feedback/OIIFB_GPL2_20040903.pdf for a translation into English of the Court decision.
  44. http://www.uoc.edu/masters/softwarelibre/esp/index.html.
  45. http://www.unibo.it/Portale/Master/Master+Universitari/2004-2005/Tecnologia+del+Software.htm.
Personal tools