REPORT OF AN INDUSTRY EXPERT GROUP Open Source Software Work Group March 2009
V3.0
2
The OSS workgroup has been very active and probably like all other workgroups, we felt that our ideas refinement and passionate debates were only limited by the deadline set for this work.
Our group included members from various sectors and line of thoughts, The group included a non-profit centre of competence for OSS, industry representation with partial or high OSS degree for their revenue model, as well as organisations that base their entire revenue on the proprietary mode. As a consequence some of the views and proposed actions are not backed by all. When this occurred, the views particular to some group members have been identified as such and can be easily located in the report. Appendix 2 provides a full list of participants. Appendixes 3 to 8 also include separate statements or contributions made by some group members to clarify their position regarding some or all the report content.
Open Source Software has become ubiquitous (see § 2). As several recent reports show, OSS is now playing a significant role in the Software economy in general and is a powerful enabler for ICT as a whole. Therefore a number of OSS specific actions can be proposed to be part of the European Software initiative and could contribute to growth in Europe, jobs creation and improvement of the European Software imbalance.
Although OSS is very successfully disseminating across virtually all businesses and organisations, Market Confidence is not equally high in all areas (see § 3.1). If this could be improved, the impact of OSS would be significantly larger. Despite the importance of OSS among European based communities and developers, another issue is the relative fragmentation and lower financial strength of the European OSS actors compared to others, among which the USA. Some important trends are currently taking place in the Open Source Software economy (see § 3.2) : Coming from a purely proprietary approach, large traditional vendors increasingly incorporate OSS into their software model, typically starting with nonrevenue relevant components, resulting in a “mixed model” approach for these vendors. New business models choose freely from proprietary, mixed or OSS software models, with a perceived tendency for mixed models to migrate towards higher OSS components. A “mixed” approach is also found among many user organisations also underlined the increase of company funded OSS, the continuous “up-stack” movement of OSS from “infrastructure” to “application” layers, and the significant impact of OSS on “software commoditisation” and “adoption of “de Jure” standards.
We identified several barriers which may limit the economic impact of Open Source Software in Europe (see § 3.3). The fragmentation of the OSS space in Europe, some “technical” barriers in relation with IPR, quality and security aspects, the relative weak presence of OSS in education, fairness in procurement, some issues related to deployment and integration, and a number of important IPR, licensing and standards related barriers such as exclusions from standards implementation and unsubstantiated use of IPR threats,
In response to the identified barriers we have proposed a number of practical actions (see 3.5).
Actions having a direct effect on the ICT sector such as ‘European Digital Independence” and promotion of initiatives targeted to commoditize software products of interest to European industries.
Although we do not propose straightforward “mandating” of Open Source we suggest a number of measures related to IPR, interoperability and standards, among which IPR sanity checks, the voluntary Licences of Right regime, protection of OSS implementation of Standards against abusive exercise of IPR, promotion of open source reference implementation of critical standards, the use of open formats for public administration, and recognition of consortia-led standards.
We also considered actions aiming at addressing the OSS space European fragmentation and improving the lack of market confidence such as : the “European OSS forge, promotion of best practises, use of voluntary labels, the European OSS test bed, the support to European Software as a Service platforms based on OSS. The strengthening of OSS organisations in Europe could benefit from tax incentives similar to what some member states have under the concept of Research foundations.
To address the education barrier we propose to include OSS in school curriculum and to promote initiatives such as Computer Driving Licence based on Open Source Software.
Finally, inspired by our knowledge of the key success factors of the Silicon Valley we identified a number of barriers and actions which can impact OSS but also software in general.
Table of Contents
EXECUTIVE SUMMARY
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 led to a position paper which was presented and debated on January 20th during a meeting with Industry and European Commission representatives.
As a follow-on it was proposed by the Commission to organise seven workgroups, each in charge of refining the position papers views on a particular aspect;
Workgroup n°7 was in charge of Open Source Software (OSS) and produced the present report.
1.2. Report structure
Paragraph 2 gives an over view of the present status of OSS (in Europe and outside).
As requested by the commission the main part of the report explores the following aspects : Issues, Trends, Barriers, Benefits and Actions all grouped in paragraph 3 hereafter.
.
7
2. STATE OF THE EUROPEAN OPEN SOURCE SOFTWARE SECTOR
The following are extracts from the press release which followed the UNUMERIT report commissioned by DG Enterprise and Industry issued in 2006 (the term FLOSS (Free/Libre Open Source Software) is equivalent to OSS) :
The following two studies and Gartner and Forrester clearly show that open source has become ubiquitous:
"Eighty-five percent of companies are already using open-source software, with most of the remaining 15 percent expecting to do so within the next year, according to analysts at Gartner."
See an executive summary of the Gartner report in Appendix 1
http://news.zdnet.co.uk/software/0,1000000121,39554840,00.htm
“Survey shows that Open source components are now ubiquitous. Users are well aware that commercial vendors are massively bringing Open Source into all enterprises, without even asking their customers, changing significantly from a complete commercial build to a mixed orchestration of Open Source and commercially licensed software.”
http://biz.yahoo.com/prnews/081201/ukm003.html?.v=101
8
2.2. Short reminder on OSS
Open Source Software (OSS), also known as Free Software, Libre Software, FOSS, or FLOSS, is a software model that was defined by MIT scientist and Mac Arthur grant winner Richard M. Stallman in the mid 80s.1 and then 1998 Open Source Initiative (OSI) started a marketing program for Free Software under the term of Open Source. After maturing in the scientific, university, and individual entrepreneurial environments, OSS has meanwhile established itself in the mainstream of the commercial software industry and has become a commercially and technologically viable alternative or complement to traditional models based on the commercialization of licensees of proprietary products and services in some areas.
2.3. Type of actors in OSS
Although any grouping has an arbitrary element, it is possible to group the actors in this field according to certain criteria, some actors will fit more than one category:
Commercial OSS user
According to Gartner research, 85% of all companies recently surveyed are using OSS, the remaining 15% are planning to do so within the next year. Therefore, all companies are commercial users of OSS. Software development and integration companies are typically among the most active users of software, which is also true for OSS.
Commercial OSS developer
Commercial OSS developers base their revenue stream upon the development of software, either as an on-demand service, or in order to provide secondary means of revenue generation, e.g. see Commercial OSS distributor.
Among OSS developer are also “Hybrid OSS and proprietary” (ex: SAP, IBM, Sun, Novell) and also new actors such as Alfresco, SugarCRM, JasperSoft, Pentaho, Compiere, Talend…. Their products are Open Source but they are the only one to own the copyright. Therefore they can have a dual licensing model (free and not free2).
1
See http://fsfeurope.org/projects/wipo/fser for reference.
2
More precisely the product catalog of mixed model vendors combines proprietary, open and proprietary-built-on-open products; dual licensing (where the licensing rights are held by the company, allowing for the same code to be distributed under several alternative licenses, one of which is proprietary) and open with certified binaries (the source code is available under an open license, but binary code is available exclusively to commercial customers).
Commercial OSS distributor
Commercial OSS distributors build their revenue stream entirely or in part upon the distribution of OSS. The commercial differentiator is typically connected to some additional value offering, which can be in the form of a proprietary add-on or component, or in fitness for purpose, warranties, or other forms.
Commercial OSS integrator
Commercial OSS integrators provide a service that is similar to a commercial OSS distributor, but the service is provided specifically for a certain customer and includes additional development and customisation for the specific needs, as well as consulting and training services.
Non-commercial users, developers, distributors
The OSS ecosystem also includes non-commercial parties, which either derive their revenue from an unrelated third-sector (e.g. scientists) and participate in the OSS ecosystem with their specific knowledge from these areas, either as an activity to support their own professional activity, or as a means of creative past-time activity. There is a considerable exchange between these parties and the commercial part of the ecosystem, resulting in new business approaches and developments.
Of course frontiers between categories are blurry, actors can belong to several depending on their product and can move from category to category.
2.4. Business models based on Open Source
Business models are largely orthogonal to the issue of software or development models. There are more than 300 different combinations3 of the three factors, with new approaches and combinations emerging over time.
Generally speaking OSS vendors cannot tie their revenue models to direct licensing cost, but much of the same effect can be emulated through contractual constructions, trademarks and/or certification. There is a large overlap of business models between OSS and proprietary software, with some business models tied to pure OSS, pure proprietary or mixed model approaches.
http://blogs.the451group.com/opensource/2009/02/23/on-open-source-business-strategies-again/
2.5. Europe OSS structural imbalance
Ideally suited for a model of open innovation and collaboration, OSS has evolved faster in Europe than anywhere else in the world, possibly because its approach is well suited to an environment of diverse participants of varying sizes as it is predominantly found in Europe. But while much of the innovation and development is European in origin, and European experts and contributors are prominent and highly regarded in the OS community worldwide, Europe has yet to bring forth an Open Source champion of comparable size to those operating out of the US.
Reasons for this structural imbalance can at least in part be found in the transformative process that several of the large US IT companies have entered in the late 90s when they began to develop their own OSS strategy. This development brought about very large mixed-model companies that often become visible as champions of OSS, most of which are based in the US.
In addition, most OSS consortia – the non-profit organisations managing OSS development and marketing – appear to be based in the United States and funded by US IT companies. At a second glance, it becomes apparent that a strong European ecosystem of organisations exists, but a lack of strategic focus on these organisations by European players is causing an imbalance of mind-share in favour of the US.
Europe must address this imbalance.
In order to maximise the benefits from OSS in Europe for European developers, users and entrepreneurs, our strategic focus needs to be on the better exploitation of OSS in Europe, the strengthening of the European ecosystem around OSS, and capacity building initiatives.
A pragmatic approach, taking into account legal constraints and market reality is key to Europe’s competitiveness.
2.6. OSS is a global phenomenon
The dynamics and OSS ecosystem are global in structure, with commercial success reaped on all levels: local, regional, national, and international. Smaller companies provide employment, competency and taxable revenue for European countries even if projects like Linux, Eclipse and OpenOffice.org thrive mainly based on the code contributions financed primarily by larger US-based vendors.
In some cases, e.g. OpenOffice.org with the primary development based in Hamburg, Germany, that financing contributes directly to European GDP.
The high level of interconnectivity is a primary strength and benefit of OSS, and should be taken into account in all policy setting activities. The focus should be on the creation and support of local champions that will have a global impact.
3. THE ROLE OF OPEN SOURCE SOFTWARE WITHIN THE EUROPEAN SOFTWARE STRATEGY :
Market confidence is not equally high in all sectors due to lack of information and misinformation about availability of support, skill levels, understanding of licence terms, and liability. OSS allows users to approach integration by means of the vendor of the software, a third-party integrator, or their in-house IT department. Wrong choice of integration path can lead to higher cost, which is often falsely attributed to the OSS software model, resulting in misconceptions.
3.1.2. European software companies often get acquired by larger US-based companies
Despite the difficulties that European software start-ups might have, there are indeed European software vendors including open source software vendors with the power to acquire other software vendors (in the US, Europe or elsewhere). However, the acquisitions of European software companies like StarDivision, NetBeans, MySQL AB, Innotek (VirtualBox) and SuSE by US-based vendors show that successful European software vendors often get acquired by US-based companies. Partly due to historic reasons, there are far more US-based IT companies with a large buying power than there are European IT companies who can afford to acquire other software vendors, which leads to the effect outlined in § 2.3 above. European IT companies also have fewer alternative growth strategies than their US counterparts due to smaller/more risk adverse venture capital community and fewer IPO opportunities. This is not an issue per se, but if the lack of Europe-based software companies is seen as an issue, it is important to understand the dynamics of the market.
On should question what are the consequences of this trend. It could impair the so called “Europe Digital Independence” and also impact jobs.
Moreover, there may be situation where a particular piece of software plays a key role in some economic activity or may create security related concerns under certain circumstances. In such situation, promoting open source alternate solutions may be used to modify the situation of actors and shift the competition “playground”.
Regarding jobs we believe that ultimately what matters is where are jobs located. If European Software or OSS companies are acquired by non European firms but if the corresponding jobs stay in Europe (and if the threat on “Europe Digital Independence” is minimal) then the consequences on Europe are limited. Conversely we know that European Software companies can, just as well, while having their HQ in Europe, decide to grow or move R&D facilities to non European countries.
12
3.2. Trends
3.2.1. Role of the mixed models
Some companies combine Open Source with proprietary models to optimize development cost on the one hand and maximise differentiation on the other hand. Some vendors use closed source elements to differentiate themselves from other Open Source vendors based on the same software ("value adding"). Vendors based on a proprietary software model use Open Source in order to improve standards support and to share development cost, in particular on non-differentiating components.
3.2.2. The near future will be diverse
Open Source Software represents a software model defined by a high level of user control over the software in combination with often unequalled freedoms to study and innovate upon the software, allowing for rapid incremental innovation. These benefits have become so associated with the software model that OSS is often misunderstood as a new development or business model.
There is a wide variety of development and business models built upon OSS, ranging from traditional approaches, such as custom development or COTS to service based approaches and SAAS. Most of the large players have incorporated OSS into their strategy, resulting in a mixed model approach (see §
3.2.1 above). Considering that almost none of these companies had significant OSS offerings only 10 years ago, it is possible to observe a clear trend towards OSS.
At the current point in time it is impossible to predict when and if that trend will further expand or come to an end. New companies enter the market with models spanning the entire range of proprietary models over mixed models to OSS models. Only time will show which models will be most successful in Europe.
Additionally, as the Information Economy Report 2007-2008 of the United Nations Conference on Trade and Development (UNCTAD) highlights, the ICT sector itself is a remarkable source of innovation and economic growth, but there is strong evidence that ICT-enabled innovation in other sectors has an even larger impact on the overall economic situation.
Key factors identified by UNCTAD as part of this process are the commoditisation of software, along with open innovation approaches, based on OSS. While OSS itself has remarkable innovation benefits, it appears necessary to include ICT-enabled sectors into the assessment to understand OSS's full potential for European innovation and economic growth.
One usually look at OSS models in the pure Software space. But this is changing. In addition to software related business, more and more companies offer products and services which are not software but rely on software and on Open Source Software in particular. Google is probably the most visible example. Software is not its primary source of revenue, but Google both uses OSS produced by others and releases its software as Open Source.
3.2.3. Company funded OSS support is gaining momentum
Having evolved from it’s original academia origin to a viable option for research and development, OSS has attracted more and more companies to fund and drive communities. The drivers for this are multiple (establish standards, share maintenance costs, gain visibility). This will definitely accelerate, primarily in the area of building eco systems, as we see in initiatives as e.g. Eclipse, Android, OW2, Limo
As existing projects demonstrate real returns on investment, companies will allocate more of their research and development spending in open source communities.
In addition, the "support for profitable business models" and "the prevention of vendor lock-in/the promotion of competitive choices” are also key drivers for why companies are funding and driving OSS communities.
3.2.4. OSS users: Combining software models based on requirements
Users of OSS include companies, administrations, public institutions, schools and universities, SOHO enterprises, end users. Depending upon factors such as availability of OSS solutions, the individual capabilities of the software and needs of the users, and pre-existing infrastructures, solutions often integrate both proprietary and OSS components.
It should be added, that the trend today is that OSS solutions get used in number of companies without the upper management being aware of this shift in the companies' “procurement strategy”.
ECIS adds the following :
This is due to engineers which manage the companies' IT systems just download, test and - if it works - install different OSS solutions to complement the already installed IT base. This is possible due to the free access to the OSS without having to go through cumbersome and time consuming tendering processes. MySQL is an example of such a trend.
End of ECIS addition.
3.2.5. Growth of the “Vendor” model
Inspired by Red Hat, the OSS the “Vendor” model is being adopted by numerous companies in all domains (BI, ETL, CRM, ECM, ERP,…) .
3.2.6. OSS contributes to Software Commoditisation
Commoditisation trend : generic software follows an evolutionary trend toward commoditisation (due to intense competition that level functionalities and added value) and open source is generally a key factor of this evolution. Even domains with very hard constraints such as telecommunication systems or embedded systems requiring expertise and know-how is now subject to this commoditisation.
Commoditisation benefits users and integrators and challenges pure software vendors to innovate rapidly in order to differentiate their products from commodity versions: the opposition of these types of actors on the OSS subject is understandable.
Examples of sectors in which OSS is competitive:
3.2.7. Open source model is moving from “infrastructure” to “application” layers
Mixed models are understood as combination of Open Source and proprietary software under a proprietary umbrella.
Open Source Software is however seen as making strong inroads in terms of customer adoption in areas previously thought as strongholds of proprietary software: business intelligence, high performance relational databases, ETL (ex: Talent), ECM (ex: Al Fresco), etc.
This movement “up the stack” is expected to continue in the future
3.2.8. OSS accelerates dissemination of de jure standards
Open source is a way to promote standards either de facto or de jure. A lot of de facto standards disseminated through open source implementation (early IETF standards, W3C, …). Open source implementation can be a way to accelerate discussion and dissemination of de jure standards. An open source implementation can help consolidate a standard by pointing the inconsistencies or lack of precisions of some specifications (it’s especially true for interoperability standards).
3.2.9. OSS contribute too improve the quality and security of software
Distribution of the source code of a piece of software and the right to improve and republish it, contribute to it incremental improvement ; when a large community of contributors is motivated and well organized, the resulting code may be considered the state of the art with respect to some functionality, the quality and dependability of the software, its intrinsic security. For some specific functions such as cryptography and e-voting algorithm, access to the source may be mandatory.
3.3. Barriers
3.3.1. Need for OSS European Digital Entrepreneurship
The very first step to capture the current Open Source trends is to have a coherent strategy and also a global vision. At the moment this is not always the case. Most of the proprietary software vendors do have a strategy, a roadmap and a vision (functional and technical).
Open source is often perceived mainly as an alternative to proprietary software. Hence the vision is reduced to an implementation roadmap, trying to compete with closed source software in terms of features to implement. One should think of OSS as a global phenomena to define a strategy in terms of impact and role of OSS in the service economy. Instead of being perceived as “running after a proprietary solution” or as an alternative, Europe should try to lead and to push an innovative technical vision implemented in Open Source. As this implementation will be in Open Source, all IT players, promoting or not, using or not using Open Source, will then be able to use it as they want, without any business discrimination.
The Commission is not the owner of this vision. The various actors, creators, integrators, users, are. For it to grow it requires what we could call an “Open Source Digital Entrepreneurship” attitude, meaning that the various communities and actors could maybe better share a common roadmap of who does what for which goal in the three aspects (creators, integrators, users).
3.3.2. OSS European space fragmentation
Note: the following applies to European OSS communities not OSS vendors
The Open Source software communities do not have the critical mass and are not enough organized to cooperate and share issues, infrastructure, etc… In Europe there are a number of different Open Source communities or consortium. They all suffer the same issues: Lack of money, lack of reliable infrastructure, lack of European visibility, fragmentation of OSS foundations between countries.
Furthermore relationships between Communities and enterprises and among Communities are not always effective yet. In that respect the US show a better example. There is an understanding by US entities that supporting such entities is useful “ecosystem maintenance” for their commercial environment. Europe has a very healthy ecosystem of organisations, some of which are larger than their US counterparts, but there tends to be very little strategic understanding in European players that a focus on collaboration with these players would initiate a positive feedback cycle for the European area.