At around €85 billion, the European software market is about one third of the global market. Yet despite much strength in software development, especially in vertical application sectors, Europe is a net importer of software, with a significant trade imbalance.
At this time, in a field which continues to evolve very rapidly, there are very significant market developments just around the corner. Only a year or two ago, we were talking of Service-Oriented Architectures and Software As A Service. Tomorrow we can expect the Internet of Services, the Internet of Things, and, putting these together, the Future Internet.
Recently, Commissioner Reding raised the need to develop a European Software Strategy, and appealed to industry for input. This paper provides an overview of the issues raised by industry, highlighting areas of agreement and areas of disagreement among the respondents.
It is no surprise that everyone agrees on the need to continue to develop eSkills in Europe. Everyone agrees that measures should be taken to reduce market fragmentation and that public procurement should be used to accelerate innovation.
There is also general agreement on the need for interoperability of software-based systems and on the need for effective protection of intellectual property. But there is considerable disagreement on how these should be achieved.
Importantly, many of the issues raised cannot - or should not - be addressed in isolation. Not only are the topics themselves intertwined: they would require action, or at least cooperation, by agencies other than DG Information Society & Media, and some would require engagement with national agencies in the Member States.
Table of Contents
The need to develop a European Software Strategy was raised by Commissioner Reding in her speech "Towards a European Software Strategy" at the Truffle 100 event (19 November 2007), and reiterated at the Microsoft Innovation Day (4 December 2007). The Commissioner appealed directly to industry on this matter. Subsequently, a number of key organisations in the software sector have sent to Commissioner Reding their views on the issues that should influence such a strategy, and have suggested elements that might form part of a strategy.
This present document provides an overview of the issues raised by industry, in the responses to Commissioner Reding.
The annexes précis the specific recommendations of each respondent.
2. THE EUROPEAN SOFTWARE & SERVICES INDUSTRY STATUS
The importance of software goes far beyond what can be measured through its market size - which is already estimated to be more than €85Bn in Europe, representing over one third of the global market1.
Software innovation is systemically linked to innovation in other sectors. The use of software in these other sectors is key to their competitiveness and a major enabler of the post-industrial society and economy. On the other hand, innovators in software technologies are mainly driven by opportunities or needs in application domains. Innovation and application are thus two interdependent components in a system, which depend and feed on each other. Conventional models of innovation seen as a one-way flow of R&D results towards new products, processes and services are no longer effective.
This systemic aspect has to be kept in mind when devising and implementing a software strategy, as it implies that a simple view of innovation in terms of cause and effect is hard, if not impossible, to maintain. Interactions between different factors mean that a software strategy will consist of different measures for priming and improving virtuous cycles between innovation in software and innovation in other sectors.
Indeed, the European Software and Services industry has a strong presence in ‘embedded software’ in industrial systems - in the automotive and aerospace sectors, for instance - and in business systems for enterprises.
However, despite Europe’s strength in vertical markets, despite world-wide recognition of Europe’s intellectual contribution to the subject, as evidenced by its academic teaching, research performance, and technological innovation, and despite the 70% of open source software producers that live in Europe, Europe is a net importer of software. Four of the biggest five suppliers of software in the European market are US-based companies.
1 Estimate for 2008 by the European Information Technology Observatory, 2007
3. A SOFTWARE STRATEGY FOR EUROPE: THE ISSUES
Unsurprisingly, the various contributions sent to Commissioner Reding overlapped considerably in scope, but given the specific interests of the organisations they did not have a common focus and the views that they represented were sometimes in conflict.
In this section, the major themes identified by the respondents are set out. Points of agreement and disagreement among the respondents are highlighted, along with some issues that have not, so far, been addressed. Where there are differences, they may be better appreciated by references to the annexes.
Note that the subject of research was barely raised by respondents, apart from the European Software Association. This is not an indication that they think it not necessary: it is implicit in all the responses that national and European-level research and technology development programmes are fundamental to innovation.
Also note that some issues were raised by only one or two respondents and were not addressed at all by others. As far as possible, potential opposition to the views that were expressed have been anticipated, but some of these issues might yet prove to be more contentious than they seem here.
Many of the issues raised would, if they are to be pursued, require action, or at least cooperation, by agencies other than DG Information Society & Media. Also, given the principle of subsidiarity, a number of them would require engagement with national agencies in the Member States.
There is common agreement that there is a need to maintain and develop the European skill-base in software and its applications. The need is not just for IT specialists, but for software engineers skilled in applying software and services technology in all other application domains and industrial sectors. Moreover, a general public that is comfortable with IT, enthusiastic about using it at work and at home, and excited about its future prospects will offer an environment in which the youth will be drawn to it as a profession and so supply the future scientists and engineers.
The various respondents focus on different parts of this spectrum of skills, and suggest different mechanisms to support the maintenance and development of skills. However, all the suggestions are compatible and complementary. They include:
Despite the variety of mechanisms suggested, respondents did not raise any specific concerns about the structure of the educational system. While some (e.g. SAP and NESSI) wish that the work-force be better educated in the techniques of the Future Internet, and that ‘an ICT education oriented towards the Internet, distributed computing and the needs of tomorrow’ be encouraged, it is not clear whether this should be considered an application of existing underlying principles of computer science and software engineering, or whether the emergence of Software-As-Service, Service Oriented Architectures, the Internet of Services, and the Internet of Things, suggests a fundamental overhaul of ‘the basics’.
Moreover, given the likelihood of yet more rapid changes in future, the question arises as to how to continually renew and revise educational content. One basis for the establishment of the European Technology Platforms and the Joint Technology Initiatives is that these would be ways for industry to work with the educational establishment to achieve more dynamic, responsive education and training. Respondents have not indicated whether they consider that these initiatives are working in this way, or whether they could be improved.
There is also a proposal that the European Institute of Innovation and Technology establish a Knowledge & Innovation Centre in ICT to establish, among other aims, a better rapport between academe and industry in both research and education, and to drive academic excellence in software education. While this would appear to support the wishes of respondents, it is not clear that this will be sufficient and how ‘gearing’ might be achieved with industrial and national educational activities.
There is common agreement that standards are good for interoperability, and there is implicit agreement that interoperability is the main value of standards. Even when the focus of a standard might be on functionality of a product, the value of the standard typically lies in the interchangeability of products that satisfy the same function - which is one of our forms of interoperability.
There is also agreement that the traditional ‘formal’ standardisation processes of the official national and international standardisation bodies have been too slow for the rapidly changing world of ICT. This is evident in the emergence of ‘informal’ standardisation bodies outside the official standardisation system, such as W3C - the World Wide Web Consortium. (The European Commission was a founding member.) Whether a single organisation or a closed group can establish a proprietary ‘standard’ is a matter of semantic debate, but as far as the user is concerned, a dominant proprietary platform can be a de facto standard - indeed, ‘the standard’.
The slowness of the formal standardisation process has been recognised for some time, but even recently a study commissioned by DG-Enterprise reported that
“the present approach to standardisation is still ill-suited to the needs of ICT with its fast changing landscape”2. The European standardisation body, CEN, has responded to these concerns by introducing a range of quicker, easier processes by which interested parties can introduce their ideas for standardisation into the ‘official’ process, with a view to early adoption and later formalisation, even from standardisation proposals arising from projects within RTD programmes. Nevertheless, industry appears to continue to be more interested in the informal process than the formal process.
Despite this apparent agreement, there is some tension in the views of industry, with some organisations more or less content with the status quo and even critical of a perceived emphasis in Europe on formal standardisation and attempts to bring de facto standardisation into a formal framework. Others feel that Europe is not sufficiently represented on informal standards bodies and that the Commission could usefully stimulate the establishment of alliances and communities to develop and maintain standards while European industry could play a greater part in existing bodies that set or influence standards.
A further significant difference between industrial organisations lies in the importance that they believe should or should not be attached to open (but not necessarily formal) standards. (See below concerning the software market.)
3.3. Open Source Software
Although many individual European experts and communities are prominent and highly regarded in the worldwide Open Source community, the commercial exploitation of open source software, whether or not some of that software might be free as well as open source, has generally accrued to organisations outside Europe (such as ‘Red Hat Linux’, even though Linux was to a great extent a European development).
Even the respondents that promote OSS recognise that its take up, by the private sector as well as public, is inhibited by what OFE and OBOOE call a lack of ‘market confidence’ arising from concerns such as availability of support, skill levels, and understanding of license terms. Although not explicitly identified by respondents, other concerns might include continuity of supply, maintenance, and even liability. OFE and OBOOE also make a case that the application of RAND3 terms for IPR, “particularly when applied to Software Interoperability Standards”, discriminates against OSS.
“The specific policy needs for ICT standardisation” (ENTR/05/59)
3 RAND: ‘Reasonable And Non-Discriminatory’
Some respondents suggest measures to facilitate OSS exploitation including measures to address the inhibitions to adoption of OSS and giving preference to OSS licensing of IST R&D. Others, especially CompTIA, argue that any measures that favour OSS (or any business model at all) would distort the market. (It is implicit that this would be detrimental.)
Respondents generally recognise the importance and desirability of interoperability, including:
However, there is not a common view on the best way to achieve interoperability. This issue is inter-related with that of Intellectual Property, and the differences of view are addressed below in the context of ‘The Software Market’.
3.5. Intellectual Property Protection
Respondents agree on the importance of Intellectual Property Protection, for SMEs as much as for large companies, in that it:
And as CompTIA observes, even the Open Source model depends on Intellectual Property protection for its success:
“Open-source software companies ... rely on copyright to ensure that their innovations are freely distributed and not expropriated and made subject to more restrictive terms by their competitors.”
Existing and proposed measures to tackle counterfeiting and piracy4 are therefore welcomed. Suggestions for additional measures to help with the protection of Intellectual Property include:
For example, the draft Communication on an Industrial Property Rights Strategy for Europe: http://ec.europa.eu/internal_market/indprop/docs/rights/communication_en.pdf
There are also significant differences of opinion about whether the present patenting system in Europe should be changed. These issues are inter-related with those concerning interoperability, and they are addressed under the topic of the Software Market, below.
3.6. The Software Market
Although expressed in different ways, often indirectly through attitudes to interoperability and intellectual property, and to the use of Open Source Software and the use of standards, there is one fundamental schism in the community. The issue is whether the market should be left to operate under more or less the existing rules (albeit with perhaps more rigorous policing of the Intellectual Property system) or whether additional measures that would affect the operation of the market should be taken to avoid ‘lock-in’ of users to particular suppliers, and to make the market more open than it is at present.
The argument for a ‘free market’ is too simplistic: the creation of Intellectual Property protection arrangements was an intervention in a formerly more free market. Anti-cartel and anti-trust mechanisms are more recent measures designed to ensure a more competitive market than might naturally arise.
The question now is whether in the interests of both interoperability for software users, and a more open market for software suppliers, additional measures might be taken. Suggestions include:
There are also suggestions for direct alterations to the protection of rights in software and computer-implemented inventions so as to avoid Intellectual Property rights being used as a barrier to interoperation, and therefore a barrier to investment in interoperable software. In particular, the use of ‘hidden patent thickets’ by major companies is of great concern to those who might find themselves blocked from the market, or worse, after considerable investment in
An exemplar could be the Spanish initiative “Educar para Crear” (“Education for Creation”): www.educarparacrear.org
interoperable products. This is of concern to SMEs, unable to fight litigation on equal terms. It is also of concern to the Open Source movement.
‘Patent thickets’ or ‘patent fences’ are common in, for instance, the pharmaceutical sector, and they are becoming more common in chip design in the computing sector. In each of these sectors it is generally assumed that the companies involved are competing on an equal footing and are equally able to protect their own IP, to investigate the existing IPR of others, and to fight any legal battles. Respondents who suggest changes to resist such developments in the software market do not indicate why this sector is or should be special.
Respondents did not overtly consider the option of facilitating more transparency of patent protection, so that there are ‘no surprises’ for investors (cf. the fundamental tenet of efficient markets that there should be an effective flow of information).
While there has been discussion within the software community about the possibility of mandating Open Source in public procurement of software, none of the respondents in the present exercise propose such a measure. Some respondents have considered the possibility, but they consider it should be a matter of choice (e.g. ESA) or a matter for the market (e.g. CompTIA).
3.7. Small and Medium-sized Enterprises
Strictly, this is a topic that should be considered as a ‘Software Market’ issue, since if action is to be taken to help European SMEs, then such action could be considered to be an intervention in the market. However, throughout the world efforts are made to encourage the start-up, growth and full engagement of (high-tech) SMEs in the economic eco-system.
Suggestions for measures to achieve this within a European software strategy include:
6 This topic was the subject of specific ‘topical seminar’ run by the Cistrana ERA-Net in January, 2008: “How to increase SME involvement in R&D calls”: www.cistrana.org
There appears to be general agreement on most of these measures - larger companies generally recognise the value that they gain from the presence of SMEs in the gene pool of the economy. However, it is likely that an exception is the idea that there should be special RTD programmes for SMEs that would ring fence resources and make them unavailable to others - creating a so-called SME-ghetto. This is likely to be contentious.
3.8. Reduction of fragmentation
In a number of different ways, and in different contexts, industry agrees that measures should be taken to integrate European efforts more effectively.
The initiative to establish a European Research Area continues to enhance the integration, begun in the Framework Programmes, of formerly national R&D communities into a European R&D community. However, many SMEs and individual practitioners outside the research community are more nationally-oriented. Even their trade associations are nationally-oriented (witness the nine organisations that collaborated on one of the responses to Commissioner Reding). One suggestion (from the NTAs) is therefore to:
– establish a European network for practitioners to create a more efficient market in expertise and a higher profile for European software suppliers.
Such a network might facilitate cross-border recruitment, co-operation, trading, and sharing of experience7.
Note that one objection to such a network is likely to be that it would be better for Europeans to participate in international networks, rather than create a ‘European ghetto’.
The European Technology Platforms and Joint Technology Initiatives are well-regarded as mechanisms that help industry, academe and the public sector to come together in a coherent way to integrate research with product development and application, and European-level efforts with national efforts. Suggestions to build on these successes include:
While we talk of the software market in Europe, the public sector - a major ‘buyer’ of software in the internal European market - is generally a collection of national customers, often with little communication even within Member States.
7 For example, experience in “how to access local distribution partners in a given country”.
The ICT Policy Support Programme within the Competitiveness and Innovation Programme supports and accelerates the Open Method of Co-ordination between Member States to create a genuinely European market - although only embryonic at this time. Respondents did not make specific suggestions in this line, but they did express a wish to develop the internal market, and to make more effective use of public procurement, suggesting that ‘facilitated OMC’ might be a route to explore to reduce fragmentation at least in public segments of the market. Indeed, in different ways, through references to benchmarking and sharing of best practice, several respondents alluded to the potential for the OMC to contribute to formation of a real single market. But that might not arise naturally: it would probably require stimulation and facilitation by either the Commission or some grouping in the community.
3.9. Public procurement
The prospects of public procurement being used to create ‘lead markets’, in which public agencies become early adopters of new technology, has been discussed for some years. In 2006, a Working Group of the National IST Research Directors Forum drafted procedures that would comply with state aid rules of the World Trade Organisation.
All industrial parties are enthusiastic about the idea. One suggestion is that the CIP could be adapted to incorporate such a ‘lead’ role to achieve pan-European evaluation of new technologies, and accelerate the innovation process even further.
3.10. Financing software innovation
The financing of European software innovation, and particularly the taking of ideas generated in research from the lab through start-up company to the global market has been a subject of some concern for almost twenty years. Comparisons have frequently been made with the dynamic investment environment of the United States. Various attempts have been made to couple ideas to the market better, ranging from requiring research projects to submit a business plan along with their proposal, through encouraging investment forums in Europe (even establishing a European Investment Forum) to the JTIs of today with their industry ‘pull’.
Respondents to Commissioner Reding now make two specific suggestions:
have available greater leverage overall. (This is also in line with the earlier topic of reducing fragmentation in Europe.)
3.11. The Future Internet
‘Software as a service’ is emerging as a new paradigm for software provision, with the potential for more flexible, evolvable, services. In principle, it offers great potential for users to be able to build and evolve their systems more flexibly and for new suppliers to readily join a pool of service suppliers and compete on an equal footing with installed suppliers. The concept is now evolving into ‘The Internet of Services’.
In parallel, ‘things’ are becoming ‘smart things’. These are not just products on supermarket shelves with RFID tags: they are the components of the aeroplanes that we fly in, the smart surface of the roads that we drive on, the clothes that we wear, the equipment that we operate, the pills that we take, even our own ‘wet components’. They will communicate with each other and many of them will be the trillions of sensors and actuators with which the Internet of Services will interact with the real world. This will be the ‘Internet of Things’. Together, the Internet of Services and the Internet of Things will be our Future Internet.
With the exception of SAP, most of the respondents do not make specific reference to the emergence of the Future Internet, and do not make any strategic recommendations regarding its development and exploitation. Nor do they propose measures intended to affect its social and economic impact. However, SAP makes strong recommendations that would put the Future Internet at the centre of policy-making and strategic thinking in Europe. In particular, it is recommended that a Joint Technology Initiative be established for the Future Internet (see also ‘Reduction of Fragmentation’ above).
Despite the positive benefits offered by the Future Internet, there is a cloud on the horizon. A recent phenomenon is the emergence of proprietary ‘service infrastructures’ in which a ‘stack’ of infrastructural services, running on a computing ‘cloud’, is made available to the user as a basis on which to build their higher level services. The advantage to the user is that the component services are guaranteed to interoperate, and that they have been validated through field testing by many other users, so issues such as trust and availability can be assumed to be managed well, especially when the suppliers have high quality brand names like Google, eBay and iTunes. However, these are all US-based companies that are setting the pace and the standards. Their rapid development alone represents a barrier for European companies trying to enter the market unless a ‘level playing field’ for service development, deployment and access is achieved - as the ISTAG advocates in its recent report on the “Web-based Service Industry”.
The industrial respondents to Commissioner Reding did not highlight this concern, perhaps because it is such a recent phenomenon. But such service infrastructures are set to play a major role in the Future Internet and it is a topic that should at least be considered in any European Software Strategy.
4. ANNEXES: SUMMARY OF CONTRIBUTIONS
This paper has drawn on contributions sent to Commissioner Reding by:
The recommendations of each of these contributions is summarised in the following annexes.
The ESA makes recommendations in three areas:
A Knowledge Economy
ESA makes a number of recommendations aimed to raise the eSkill level of all sectors of European society, and to address the shortage of skilled professionals:
The ESA recommendations in this area are designed primarily (though not entirely) to establish a ‘level playing field’ for Independent Software Vendors and particularly SMEs:
Research & Development
ESA highlights as R&D topics:
SAP calls upon:
Specifically, SAP recommends:
Finally, SAP recommends that European R & D Instruments be aligned toward the Future Internet, including:
Increase the efficiency of the European Community’s Research & Development by:
Reinforce the role of European industry in standardization for software.
Establish regional European excellence in the software industry by:
OFE and OBOOE make recommendations in five areas:
Support to the Market via Public Sector Procurement of Open Standards
Policy Strategies that Implicitly or Explicitly Favour Proprietary Software
– mandate royalty free IPR for software interoperability.
Collaboration and Integration Models within R&D
The actions suggested to achieve better focus on OSS in R&D include:
Market Initiatives that clarify Free/Open Source Adoption
8 As Commissioner Kroes said in her speech of 10 June 2008: “For all future IT developments and procurement procedures, the Commission shall promote the use of products that support open, well-documented standards. Interoperability is a critical issue for the Commission, and usage of well-established open standards is a key factor to achieve and endorse it.”
The actions suggested to achieve better market adoption of Free/Open Source include:
Develop a funding stream for market based initiatives and best practice schemes
Market Initiatives that discourage Proprietary Lock-In
One action suggested to discourage proprietary lock-in is to:
– conduct a market trial of the Certified Open programme within the i2010/ post-i2010 framework.
CompTIA makes recommendations in five areas:
Public Policy Recommendations to Promote e-Skills:
Public Policy Recommendations for Standards and Interoperability:
Public Policy Recommendations for Intellectual Property Protection: