1. Objectives of the project

In recent years, open source software has developed as a novel form of collaborative production. Relative to its origin as a collaboration between individual non-professionals, it has seen tremendous success, both in terms of the commercial and technical strengths of the produced software itself, but also as a model of organisation: open source software is arguably one of the best examples of adaptive self-organising, collaborative networked organisations that exists today. Indeed, this success has been enough for open source to be considered as useful alternative for e-government applications, in particular within this Strategic Objective of the IST work-programme.

Several gaps still remain in our understanding of the functioning and purposes of open source software. The FLOSS project (funded by IST/FP5) resulted in the single largest knowledge base on open source usage and development worldwide, and filled some of these gaps, at least in our understanding of the economic and development models behind open source. Not enough knowledge exists, however, in the critical area of policy formation: the impact of policy choices on open source, and the effect of open source on available policy options and government actions is not well known. The support action builds on the FLOSS project to fill in important gaps in the understanding of open source with a focus on specific gaps in the policy application domain.

The primary objective of this project is to support and facilitate the objectives of the work-programme. The EU is already a leader in the usage and development of open source software, and the FLOSS project made it also a leader in research in this field. The FLOSS-POLS project aims to maintain this lead in a competitive global environment, especially in the domain of open source policy. FLOSS-POLS will work on three specific tracks:

1. evaluating government policy on open source
2. understanding gender issues in open source
3. modelling open source as a system for collaborative problem-solving.

While all areas will have a focus on studying the impact of policy and providing policy recommendations, they are designed to measurably further the understanding of open source by addressing specific, critical topics of relevance to the effective functioning of e-government and emerging networked organizations. Measurable objectives have been identified in terms of specific deliverable numbers rather than milestones, since deliverables are scheduled on specified dates.

Track 1 – Evaluating government policy on open source
This track is framed as a response to the question: What is the practical effect of government policy on open source software, and how does open source software affect government? The track aims to build a community of use and expertise through the organizing of workshops and working groups, starting with the support of a major conference on open standards, libre/open source software and e-government (D30). The workshops (D22, D23) and conference will bring together academics, libre software community participants, industry and government representatives to build a framework for policy discussions relating to the effective application of open standards and open source software.

They will be supported by, and provide feedback for, three specific empirical surveys and studies. The first is on interoperability and open standards, and will aim to document both the attitudes as well as actual implementation of these in the public sector in order to provide concrete policy alternatives (D4). The second will clearly identify user needs in open source e-government through a comprehensive survey of open source attitudes and usage in European governments (D3). The third study will examine the open source community as a “costless” skills development environment – the FLOSS survey showed that a vast majority of developers see their participation as a way of developing skills that are valued on the job market. This study will, through the coordinated surveying of developers and employers, quantify this (D7, D9) and determine policy implications (D10)

Track 2 – Understanding gender issues in open source
Several surveys show that women make up under 5% of open source developers, while they account for as much as 25% of proprietary software developers. The FLOSS survey showed that the few women who do participate, however, are much more active than the average male open source developer, so this gender bias may not be inherent. However little is known about the reasons for this bias, or about what can be done to change this situation. In an environment where the open source community is an increasingly valuable part of the modern information society, a training system and even a support structure for e-government activities, clearly it is critical that features that prevent the full involvement of women must be understood and countered. This should be an important aspect of European policy, especially in the context of a wider adoption of open source within SMEs, governments and by the public at large. Through empirical surveys, participant observation and rigorous ethnographic study, this track will provide the world’s first comprehensive study of gender in open source and develop policies to maintain the EU’s leading role in this field (D12, D16).

Track 3 – Modelling open source as a system for collaborative problem-solving
Through the innovative use of AI agent-based simulation modeling, this track will answer the question: Is collaborative exchange an efficient means of production and if so, under what conditions does it relate to proprietary exchange? This track will evaluate the adaptive self-organising abilities of collaborative networks such as open source developer communities by using model agents that are “cooperative” (information-sharing) or “selfish” (information-selling), and study their effectiveness in jointly solving problems in a networked environment under various operating conditions. Cost-benefit ratios for individual participants as well as the efficiency of the system as a whole are measured, significantly furthering our understanding of knowledge flows in networks and providing valuable inputs to the development of policies for the optimal governing of such networked organizations (D21).

Verification and measurement
No specific work package is designated for evaluation and assessment, as the workplan has been divided into several small work packages that include assessment and monitoring aspects, and the Gantt Chart identifies the delivery dates for 30 deliverables and periods of work for 16 work packages, providing sufficient detail for milestone measurement and progress assessment for a 24 month project.

2. List of Participants

*CO = Coordinator
 CR = Contractor

** Normally insert “month 1 (start of project)” and “month n (end of project)”
These columns are needed for possible later contract revisions caused by joining/leaving participants

3. Relevance to the objectives of the IST Priority

The project directly supports the goals of IST Priority Objective 2.3.1.9, “Networked businesses and governments”. The approach of having three parallel but integrated tracks allows FLOSSPOLS to address simultaneously several focal points of this IST Priority objective.

Track 1 supports the focus on “Open, secure, interoperable and re-configurable e-government” and enables the identification of best practises, attitudes and user needs at various levels of government in relation to the deployment “as much as possible [of] open source software solutions for all aspects of inter- and intra-government operations”. The survey and resulting policy recommendations (work packages WP1, WP2) will fill in the considerable gaps in the understanding of government authorities’ attitudes towards open source systems, map their current or planned usage and identify key problems faced and benefits realised. The e-government workshop and conference (WP13, WP15) will provide much needed interaction between user, developer, research and policy constituencies, while also feeding back into the understanding of government needs and the development of policy recommendations.

The interoperability and standardisation policy study in Track 1 (WP1) will also support the focus on open e-government through the identification of the role that open standards play in government deployment of ICT systems, and the costs and benefits as seen by them. In addition, the study will support the focus on “technologies for interoperability” by identifying constraints and bottlenecks in interoperability and recommending policies that could facilitate the effective deployment of open standards. This will also support the goals of IST Priority Objective 2.3.1.7 (“promoting standards and open reference architectures”) and 2.3.1.8 (“Active contribution to world-wide standards setting, in particular open standards”).

The study of the open source/free software community as a “costless” skills development platform (WP6) will provide general support to the goals of Objective 2.3.1.9, in particular relation to the focus on “Managing knowledge to support innovation” and “IST as driver for small business and government re-organisation”. In the context of the E-Europe 2005 action plan adopted last year, with the goal of making Europe a leader in enjoying an ubiquitous, accessible information society – with “Interactive public services, accessible for all, and offered on multiple platforms” – an critical element in reaching these goals is widespread training for all in using information society technologies. The open source/free software community has been shown (e.g. by the FLOSS surveys) to be above all a learning (or training) environment, which is valued by participants as well as by employers. What is most interesting is that this training environment is virtually “costless”, since no explicit monetary costs are born by either individuals, companies, universities or society as a whole in the support of such environments.

As such this training environment is proving to be a potential key driver for government and industry, especially small businesses. SMEs in particular cannot afford high training costs, and can greatly benefit from environments that provide training to their labour force or lead to a better-trained labour supply. As training is provided and received on a voluntary basis, quantifying, preferably in monetary terms, the method and value of such training is crucial for our understanding of Europe’s preparedness for the new accessible, ubiquitous information society envisioned by E-Europe 2005. The FLOSSPOLS project aims to take a significant step towards this understanding.

Track 2 of the FLOSSPOLS project envisions the first comprehensive study that will provide insights into the role of gender in open source software development communities. The fact that women make up such a small part of open source developers (well below 5%) in relation to their share of proprietary software developers (20%) is a cause for concern. The FLOSS survey of developers showed that the few women who responded had, on average, a much higher degree of contribution than the average male respondent (see figure 1), and were well within the “high-contribution” group overall. This clearly shows that when the conditions are right, women can participate very actively, and could greatly boost the open source phenomenon.

Figure 1: Comparison between men and women responding to the FLOSS developer survey

Based on 2198 male and 25 female respondents, so these differences are not statistically significant. See http://floss.infonomics.nl/report/

The IST 2003-2004 work programme (Annex 1, part 2 “Cross-cutting issues”) aims, “where possible, to reinforce and increase the place and role of women in science and research both from the perspective of equal opportunities and gender relevance of the topics covered”. Track 2 of the FLOSSPOLS project clearly supports these aims, within the context of understanding, reinforcing and increasing the role of women in the development and usage of open source software. Given that open source software is seen as a priority for e-government (Objective 2.3.1.9), it should be obvious that issues involving half the participant and user base for e-government should be addressed. The methodology of track 2 – combining empirical survey data with rigorous ethnographic study – provides a strong base of facts to support our understanding of the issues facing women in their participation in this important, highly productive aspect of the information society. It also leads to an understanding of the processes and conditions common in open source developer communities that inhibit women’s participation, and recommends ways to get around these features so that women can take full advantage of those attributes – collaboration, sharing knowledge – that are the most valuable in the open source model. Thus, Track 2 ensures that in the deployment of e-government for “all aspects of inter- and intra-government operations including electronic democracy systems, interaction with citizens and businesses, governmental process re-engineering and knowledge management” (Objective 2.3.1.9), women are not left behind, and indeed ensures that their participation is strengthened. Track 2 also addresses global EU goals for Structuring the European Research Area, specifically the key Objective of Women in Science (ERA/4 – Science and Society, Objective 4.3.5) “To boost gender equality in research, through stimulating the participation of women in science and technological development; and fostering the integration of the gender dimension throughout European research” and Objective 4.3.2 on Ethics (“ensure that rapidly advancing progress in science is in harmony with fundamental ethical principles”) through work package WP3, Deliverable D5.

Track 3 directly supports the first focus of Objective 2.3.1.9, “multi-disciplinary research into complex adaptive and self-organising systems and modelling, representing, tracking and measuring distributed work and knowledge flows”. Open source development communities represent one of the most widely used and productive self-organising productive networks that exist today. Previous research (FLOSS, LICKS) has provided detailed measurement and tracking methods applicable to participants in these communities. It is possible now to build simulation models of agents reflecting different motivations and types of behaviour, to test the widely held belief that self-organising networks of collaborative (information-sharing) agents can be more efficient than structured networks of “selfish” (information-selling) agents. Academic understanding, the behaviour of practitioners, emerging business models and even – as the Objectives of the IST Priority show – EU policy depend on assumptions about collaborative networks. Track 3 will use and customise existing tools to model the interaction of collaborative and “selfish” agents under different sets of assumed motivations. An evaluation of the resulting network structure and agent behaviour, as well as the cost-to-benefit scores for participants at the individual and aggregate levels will further understanding. Of special relevance to policy formation will be the analysis of how collaborative networks respond as increasing numbers of “selfish” agents are introduced into the system. This is highly relevant not only to business models for collaborative, networked businesses which must avoid losses from “free-riding” but also to policy makers who must determine when (and whether) governments should enter to provide support if collaborative (e.g. open source) networks are unsustainable in the face of one-sided exploitation.

4. Potential impact

The IST/FP5 FLOSS project had tremendous impact. Since the release of the final report in June 2002, there have been over 5,000 visits a month to the FLOSS web site and report. “FLOSS”, created as a project acronym, has now become a widely accepted generic term bridging the confusing gap between the terms Free (or Libre) Software and Open Source, and has even been recommended as a generic term by people like Richard Stallman, founder of the Free Software Foundation. A search on Google shows over 72,000 websites referring to the FLOSS project (1) and several mirror sites duplicating the final report in its entirety (albeit without permission). The project findings were covered by most major industry publications worldwide and press reports in several languages.

The FLOSSPOLS project is expected to have no less an impact within the community of developers, users, industry and those who study the libre software phenomenon, since the issues addressed – especially of gender and skills development – have rarely been examined before, and certainly not in terms of the comprehensive studies resulting in empirical data. The FLOSSPOLS project is also expected to have a broad impact on the government and policy community, partly because of the critical issues addressed by the project, but also because of its work plan placing dissemination and feedback at the heart of the project goals, beginning with a major conference and continuing with a series of workshops.

European added-value and relations to international research activities
The FLOSSPOLS project aims to support Europe-wide goals, and clearly needs to be conducted at a European level. Although some studies of e-government and open source have been conducted at national or regional levels, there is a major added value in a Europe-wide study as conducted by FLOSSPOLS; the other study areas are essentially trans-national in nature. Open source is directly relevant to the European Research Area (ERA) not only due to Europe’s lead in free software development and deployment, but also as the model for collaborative research and development best suited for ERA as it can be rapidly adapted and disseminated. FLOSSPOLS studies on skills development and gender are especially important to meet the goals of eInclusion and the aim of eEurope2005 to ensure that Europe is the “most advanced knowledge based economy by 2010”. The study of interoperability needs and how libre software addresses them is of special relevance to IDA’s plan to deliver a European eGovernment Interoperability Framework (eGIF) by the end of 2003, as also to the broader goals of the eGovernment priorities.

The FLOSS workshop on “Advancing the Research Agenda on Free/Open Source Software” (October 2002) jointly supported by the European Commission and the US National Science Foundation clearly saw the imperative need for research in gender studies, skills development and standardisation. The FLOSSPOLS project is involved with US and other international research initiatives in this field, while aiming to maintain Europe’s lead in this area.

4.1 Contribution to standards
Track 1 will contribute significantly to the understanding especially from a user point of view of interoperability and standards in software, in particular in the area of e-government. This is through the interoperability study (workpackage WP1) and egovernment survey (WP2).

4.2 Contribution to policy development
The project is by nature one of policy support, and will strongly contribute to policy development at the European as well as national and local levels by providing empirical evidence and analysis for the topics it covers: open source, open standards and interoperability in government; training, skills development and employment related to participation in the open source development process; gender issues related to software and participation; as well as broader issues on collaborative versus competitive and proprietary development of services.

(1) Search term: +FLOSS software OR open OR source OR libre OR free -dental -site:www.infonomics.nl -site:floss1.infonomics.nl