Matt Germonprez, J. Allen, Brian Warner, Jamie Hill, Glenn McClements
Throughout the 1990s, corporations involved in the design of software services viewed open source participation with skepticism. However, over the past 15 years, corporate participation in open source communities has expanded rapidly. It is no longer unusual for Fortune 500 companies to have full-time staff dedicated to leveraging open source for corporate profit. Corporate participation in open source communities has raised interesting and complicated questions about how the requirements of volunteer and for-profit contributors evolve in collaborative environments.
The Linux community offers a classic example of an evolution toward corporate participation in open source. Linux began as a student project. Its earliest community members were largely veteran Minix hackers, and because the operating system did not evolve from an existing code stream there was little to no corporate participation. The code base and community practices were simply not mature enough for integration into commercial products. As the community matured, corporations began to realize its value and started contributing toward the advancement of both corporate and open source initiatives. Linux distributions (e.g., RedHat, Ubuntu, SuSe) evolved to ease installation and ongoing management. Server vendors (e.g., IBM and Hewlett-Packard) joined to drive stability and performance enhancements into the kernel. Such involvements improved both the development economics and community maturity, reinforcing the motivations for corporations to participate , leading to new forms of industry engagement with open source communities.
Corporate participation in some open source communities has evolved into platform leadership roles. Corporations such as Rackspace, Cloudera, and Citrix have become the hubs of the open source ecosystems (OpenStack, Hadoop, and Xen respectively). In each, the success and evolution of the project is dependent on both volunteer and corporate contributions, leading to active open source communities supported by a broad range of contributors. As participants share source code, new design features and methodologies collectively emerge, contributing to the sustainability of the community, often resulting in movements reaching beyond the scope of the original member base . Within these movements, corporate competitors can coexist in open source communities, sometimes vying to incorporate specific strategic views into open source trajectories. Such coexistence can be somewhat unplanned.
In the 2010s, an analog to these unplanned partnerships emerged as a new open source business model, shifting away from the evolutionary processes that define classic volunteer and corporate participation in open source communities . An open source "community of competitors" model emerged: Market rivals began to intentionally coordinate activities for mutual benefit in precise, market-focused, non-differentiating engagements. This new model raises the question: How can open source communities be deliberately designed to effectively support a community of competitors?
An open source community of competitors is intentionally designed to enable partnerships, advance technology, and seek opportunities in ways that might be otherwise shrouded in corporate boardroom negotiations and nondisclosure agreements, or that might never occur at all. It is becoming "common practice, even amongst competitors, to work together to develop 'the basics' of software platforms so organizations can spend time differentiating their products based on usability and features, and not waste time developing standard platforms. In the automotive industry, for example, BMW, Jaguar, Volvo, Nissan, GM, Honda, and others are all working together through the GENIVI Alliance to develop a standard in-vehicle infotainment platform they can all use. Same holds true for financial services, in which J. P. Morgan, Bank of America, NYSE Technologies, and others collaborate on the OpenMAMA project, not to mention OSEHRA for electronic health records and Polarsys in aerospace" .
Here, we take a closer look at deliberately designed communities of competitors to highlight the characteristics that differentiate them from more traditional corporate participation in open source communities. While a handful of communities of competitors exist including OpenDaylight and Linaro, we highlight one specific community of competitors, OpenMAMA, to identify the structures, decisions, opportunities, and membership characteristics that define this emerging model.
OpenMAMA began as MAMA (Middleware Agnostic Messaging API), a closed-source API of NYSE Technologies. Originating in 2002 from Wombat Consulting and used by more than 150 of the world's largest investment banks, brokerages, and hedge funds, MAMA was the de facto industry standard for high-speed market data messaging, a technologically agnostic API that allowed organizations to standardize distribution of market data over multiple middlewares. The standardized distribution of MAMA aimed to reduce development times, lower costs, reduce complexity, and broaden industry support of the closed-source API.
In 2011, NYSE Technologies moved from the standardized, closed-sourced distribution of MAMA to an open source model. The move enabled external developers the opportunity to change and contribute to the MAMA design base in an effort to build a robust community of developers, sponsors, customers, and competitors, and make OpenMAMA more versatile and useful throughout the financial sector. The goal of OpenMAMA was to publish market data from multiple sources and multiple vendors in a standardized format and onto a standard, open platform. Terry Roche, chief operating officer at NYSE Technologies, explains that "NYSE Technologies is creating a new capital markets Community Platform. With a vibrant customer community already established around MAMA, OpenMAMA enables capital markets participants to take part in the Community Platform, which is a material step toward real partnership within the industry. This will fundamentally change the existing service-delivery dynamics. The Community Platform is designed to be a home for industry participants to leverage NYSE Technologies assets along with assets contributed by the community to create and go to market with rich, cloud-based, secure, value-added applications and services. We bring the community the benefits of significant cost reduction, collaboration, and dynamic global delivery to change the face of the delivery of capital markets content and services for the benefit of the capital markets industry and its customers." By having an open market of contributors participate in OpenMAMA, the capital market industry can benefit from a more robust and stable messaging API that works equally well on all competitive platforms, allowing for custom solutions to be leveraged into corporate-specific needs.
On the surface, the open source release goals of OpenMAMA are the same as more traditional open source communities. The release was intended to foster new initiatives, leading to new capabilities not envisioned in MAMA. The OpenMAMA community was open to all in an effort to expand the participant base that applies the technology in new forms. These goals are expected in all open source communities: building on prior work with the intention of fostering new innovation and new applications. In order to unpack what distinguishes open source communities like OpenMAMA and to understand its deliberately designed blueprint as a new open source business model, we identify unique characteristics in the design of a community of competitors.
Traditional open source communities foster the emergence of a commons, with the expectation that members can use the communal technologies to create economic value. The advent of a commons and its technology can have an impact on the corporate allocation of resources and commitment to an open source community. Without a commons, collective activity is difficult and leveraging open source artifacts for economic value can be problematic, resulting in unpredictable member involvement .
Communities of competitors have a distinctive design trajectory, not relying at first on the emergence of value in the commons or the technology. In a community of competitors, value exists a priori in the community around an existing open source code stream (OpenMAMA). The technology is mature, having been market-tested as closed source technology (MAMA), and is well recognized and supported in its respective community (capital markets). The open sourced technology is a value-rich, non-differentiating technology around which corporations do not compete, allowing all to realize immediate value from its existence and improvement. The common value of technology that already exists provides affordances that help articulate and highlight communal motivations for being involved .
Communities of competitors forego the complexities of developing a commons, as a commons exists from the existing code stream, relieving corporations of the initial problem of finding enough technology value to justify involvement. Communities of competitors have reduced start-up overhead and enjoy shallower design trajectories than do traditional open source communities because the key components of the commons and the technologies exist a priori from a formerly closed-source community. However, this design can come at a cost, as the members that define the commons did not evolve organically in the joint creation of the technology, but may instead be direct competitors in highly competitive industries.
Open source communities are composed in large part of a collection of corporate enemies participating as partners in a collaborative manner. Suspicion and competition may very well pervade the general relationship between participants, but common ground may be found around the distribution of non-differentiating, value-added technologies of the community. In open source communities, a level playing field is built around vendor-neutral technology, not around corporate strategy. Participation is intended to be merit-based and company-agnostic, although corporate participation has naturally changed the dynamics of many open source communities.
A community of competitors is deliberately designed with clear member identification, through executive-level steering committees where select members have direct access to strategic decision-making associated with the community. The steering committee is assembled with representation from corporate competitors in an effort to reflect broad and negotiated industry support and meaning for the open sourced technology. In a community of competitors, rival corporations intentionally take seats around the same technological table in an effort to innovate within the community commons and advance the non-differentiating, open source technology in joint partnerships. Importantly, the designed structure of such a community immediately involves a broad base of expert and competitive members improving conformance in the design of the communal, open source technology . There is no need to promote the emergence of community partnerships, as such roles are baked into the community during its primary design, fostering communal commitment and loyalty to the open source project, even among rivals.
Purposefully designing corporate participants into community roles can appear heavy-handed and controlling, giving a sense that the open community is simply a wolf in sheep's clothing. As community members do ultimately leverage and differentiate the communal technology for corporate reasons, impartial brokerage plays a key role in managing the communal design process, providing stability and legitimacy to what may appear as a corporate venture, not a communal engagement.
Open source communities ostensibly have strong decorum and respect between members, dealing honestly and openly with all. For many open source projects, openness is an inherent component, evolving over the life of the community as membership also evolves. Neutral non-profit organizations, such as the Linux Foundation or the Eclipse Foundation, are associations that serve to protect, not define, their respective communities' core technologies as well as the advancement of the commons through technical, social, and educational services.
In communities of competitors, the role of foundations can still be to protect the commons and technology, but it can also serve as a neutral home for neutral technology. As distrust may naturally run strong in a competitive, closed market, neutrality and honesty must be represented clearly in a community of competitors, so that any open source technology (OpenMAMA) does not appear as a corporate venture (NYSE Technologies). Neutral foundations foster critical mass when designing the steering committee, building awareness and reach, and providing the licensing support necessary to move a commons and technology from the closed environment of one organization to the open space of a community. Foundations also provide a safe haven for competitors to overcome their initial mutual suspicion by learning how to work together for mutual benefit.
Neutral foundations provide a "halo effect," tempering public skepticism, representing the project on neutral ground, and demonstrating the open source community as inclusive to all. While pure "objective neutrality" may be an unrealistic design goal , honest brokerage provides an offset to communities of competitors that may otherwise be considered nothing more than a corporate enterprise under the guise of open source.
"Collaboration among diverse groups of specialists has become the de facto way of working in the 21st century" . Corporate participation in open source communities has evolved tremendously since the 1990s, and communities of competitors exemplify this remarkable evolution. Open source participation is now a viable business model where traditionally competitive corporations publicly engage in the design and development of non-differentiating, market-specific technologies. Open source communities can work in hypercompetitive, hyper-capitalist industries such as capital markets, illustrating an odd but fruitful partnership that does not epitomize the traditional and grassroots ideals from which open source software was originally born .
Members of communities of competitors are principally in agreement to come together in the advancement of both shared technologies and communal engagements. Competition is set aside as members realize they have the same technical problems, and that solving them once in a standardized way benefits all. Competition does manifest, but primarily in secondary markets, outside the deliberately designed open source community of competitors, as each member is then free to leverage the shared and communal technology in corporate-specific, profit-generating ways.
Open source communities can be deliberately designed to include competing vendors and customers. Work begins from a common base of non-differentiating technology (the technology and the commons). Staunch competitors are brought together (communities of frenemies) under the auspices of a common, neutral institutional structure (honest brokerage), creating joint interest in the widespread adoption and success of the open source project. The focus is to design the community of competitors as quickly and clearly as possible and to remove the professional skepticism that comes with competitors working collaboratively. A community of competitors follows blueprints of action to define shared technologies, community commons, and the expression of responsibilities and structures. The blueprints are focused on creating clear and open signals for potential new members and removing points of confusion that may impede community advancement. In a community of competitors, minor misunderstandings can quickly become major conflicts and the community is designed to avoid such problems.
In open source communities of competitors, new works are built from the work of others; contributions are to a historical trajectory, developed from and through an open source community. Corporations must learn to manage in a world of open, interactive designs where engagements are toward communal technologies, freeing up corporate innovation capacity to differentiate and add commercial value to open source products. Communities of competitors have reformed conventions of how open source communities are structured, where corporate opportunism is kept at bay and focused coordination can form around the non-differentiated communal technology in efforts to advance the goals of all involved .
This project has been funded through the National Science Foundation VOSS - IOS Grant: 1122642.
4. Yeaton, T. Why you need open source for health exchange success. Wired Innovation Insights. 2012; http://www.wired.com/insights/2012/12/why-you-need-open-source-for-health-exchange-success/
10. Kelty, C. There is no free software. Journal of Peer Production, Issue 3, (2013); http://peerproduction.net/issues/issue-3-free-software-epistemics/debate/there-is-no-free-software/
Matt Germonprez is the Mutual of Omaha Associate Professor of Information Systems and director of the Open Source Lab at the College of Information Science & Technology, University of Nebraska Omaha. His interests include open source communities and technologies, and advancing methodological and theoretical contributions in these areas.
Jonathan P. (J. P.) Allen is an associate professor of information systems at the School of Management, University of San Francisco. His interests include open technology, digital business, sociotechnical analysis, and the promotion of environmentally sustainable economic opportunity through technology.
Brian Warner is senior client services manager in collaborative projects at the Linux Foundation. The Linux Foundation was formed to protect, promote, and advance the Linux kernel. It uses its collective experience to provide a neutral home for open source projects in collaboration with the world's top technology companies.
Jamie Hill is a director within the Enterprise Software group at NYSE Technologies, the commercial technology unit of NYSE Euronext. In this role, he is the product manager for products and services derived from the OpenMAMA open source project, including the NYSE Technologies OpenMAMA Enterprise Edition, certification program, and support services
Glenn McClements joined Wombat Financial Software in 2006. In 2008 Wombat was sold to NYSE Euronext and he is now head of R&D for NYSE Technologies, developing new technologies and products for NYSE and its customers. He is also the maintainer of the OpenMAMA project, hosted by the Linux Foundation.
Copyright held by authors. Publication rights licensed to ACM $15.00
The Digital Library is published by the Association for Computing Machinery. Copyright © 2013 ACM, Inc.