Some say that open source software development is anti-capitalist. Georg von Krogh believes that a combination of open source and proprietary development is best

Open source software development is now a cultural as well as an economic phenomenon. Instead of the traditional model of software production, where companies hide the details of code for fear that their value will be eroded, open source software is generated by the collaboration of thousands of volunteers who agree to share the resulting product for free. The Internet has enabled these worldwide communities to work together creating, most famously, products such as Linux, an operating system so ubiquitous it runs a fifth of the world’s servers. So successful has the model been that the open source model of innovation is now being copied outside the information technology industry. Biotechnology researchers are working together voluntarily to advance genetic research.

Unconventional wisdom
Open source challenges all the received wisdom about innovation. Intellectual property law – such as copyright and patents – has grown up in order to safeguard the material rewards of invention and, by extension, to create incentives to further research. For this reason, some believe that open source simply cannot last. Indeed, the sustainability of open source as a tool of innovation is of interest to many parties: to collaborators, to potential users, and to those considering offering ancillary products and services. Most companies do not want to lock themselves into technologies or, even worse, a model of innovating that has no future.

It’s good to share
Innovation must benefit the innovator and the customer, and it must be institutionalized in order to survive. So what does open source do for its legion of voluntary collaborators? Well, for starters, many are simply not voluntary. According to one recent study, no less than 40% of contributors are paid to do so. This may be because companies find open source an efficient way to innovate. They do not have to “reinvent the wheel”. In traditional software development, there are many obstacles to the reuse of code, while our research shows that in some open source projects, for every new line of code a further 25,000 may be second-hand. Open source is a way of directly solving problems. It can be much cheaper than traditional models of innovation, and it is a flexible way of developing exactly what clients want. Some companies join in open source development because they get independence from big software companies and avoid large licence fees. But that still leaves the three-fifths of collaborators who seem to be doing it just for the love of it. For many, sharing advanced and effective solutions to technical problems gives them status in the community of software developers.

Consumer benefits
Most of the world’s biggest companies now use some form of open source software. They and other software customers choose their products based on a range of criteria, such as stable product supply, customer support and product quality. Some users reckon open source software is likely to be around longer than some software suppliers. When suppliers go out of business, users are left with a piece of code, and nobody to maintain it. Open source products rarely rely on the support of only one company. Rather, they rely on the continuous support of many software contributors and other users. Some open source software products have outlived generations of contributors, while others have not. What matters is the extent to which there is a general interest in the software offered. As software and hardware companies increase their commitment to open source, more and more customer services, including distribution and customization, consulting, education and training, are being offered to support existing products. However, as customers start to look for increasingly sophisticated solutions, they may need services in addition to those currently offered, such as additional software, financing of systems, hosting and application, opening up new business opportunities. Corporate customers are rightly sensitive about the quality of software products. The cost of a security breach, a server malfunction, or a database collapse is immense and can even undermine the company’s existence. Many that choose open source cite higher quality. How are these quality levels achieved? Open source software projects shift the balance of power back to users by securing quality through direct and immediate user feedback. This feedback is constant throughout the product’s life cycle. Open source is often thought to be anarchic, and to lack the processes typical in most software corporations. In fact, open source projects secure product quality through a regimented contributor talent selection and retention as rigorous as any company. In most open source software projects we have examined, only “core-developers” have the access to change any part of the software’s “official version”, while more peripheral contributors do the testing and inspection, and give user feedback. For example, in Freenet, a file-sharing software, just 1% of contributors did more than half of the development work. To become a “core-developer” contributors have to demonstrate considerable interest in and knowledge of the product as well as programming skills. Moreover, “core-developer” status is kept only by those who make significant and lasting contributions to the product.

Genius is doing it twice
Most companies strive to institutionalize good practices. Over the past four years, the open source movement has engaged in a day-to-day process of institutionalization that enhances its ability to sustain itself. It protects itself by securing public access to products; if appropriated by an individual or company, the underlying project may be jeopardized. Open source projects deploy various means of securing public access to products, including hosting projects on public infrastructures such as Sourceforge, and officially registering project names. Several projects have also set up project governing councils to oversee compliance with the licence, make critical decisions on software design, cooperate with firms or other institutions, manage the project’s image and so on. Many, such as the Apache web server project, have turned into professional institutions that coordinate the efforts of contributors and manage the strategic development of the product portfolio.

Enter the lawyers
While open source is doing well on these three criteria, there are some immediate legal challenges. The most worrying is the SCO case. SCO claims it owns rights to UNIX, an operating code, and that this has been copied by Linux. Whatever the merits of SCO’s case, the legal uncertainty has made some companies wary of using open source software in case they could be held liable for using proprietary software in combination with open source. And, whatever the outcome of the case, it is accepted that from now on open source distributors and integrators will have to verify the origin and intellectual property rights of the code within their products. Many companies and projects already employ methods to do this. Some projects require contributors to provide legal documentation proving ownership of the code they contribute and verify authenticity by using digital signatures. A second worry is that software patents are on the rise in the United States and are being considered in many other countries. In the mid-term, open source communities are concerned about how such patents could restrict the scope of innovation across their projects. Because many open source projects face a scarcity of resources, it is unlikely that they can fight out lengthy and costly legal battles on their own.

A middle way
Open source is unlikely to overtake the traditional software industry where innovators are able to appropriate the financial returns stemming from their products. But it would also be irrational to discard the opportunities offered by open source innovation. Due to the economics of software, companies should adopt what I call compound innovation – a strategy to initiate, participate in and develop open source software while protecting other products through patents and licences, thereby securing a revenue stream. Companies to watch include MySQL and Sleepycat which make databases and data management, or Trolltech, which produces tools for development and applications. These firms do not simply sell services to support “free” open source products but embrace a dual licensing model that allows for compound innovation. Dual licensing permits customers to use software at no charge under the condition that if they use the product in conjunction with an application they redistribute, the code must be made available and freely distributed. However, if a customer wants to use a product in conjunction with software it does not want to release to open source, then it must buy a commercial licence. The companies that use a dual licensing model stand behind their open source products with certified code, a plan for developing and testing the product, and various levels of indemnity that the vendor will defend the user in the event of an infringement lawsuit. It seems compound innovation may offer customers the best of both worlds – the freedom of open source along with the benefits of a commercially supported product. With compound innovation, both software companies and open source may thrive, as will their customers.

Georg von Krogh
Georg von Krogh is a professor of management at the University of St Gallen, Switzerland, and a director at the university’s Institute of Management. His work focuses on innovation and competitive strategy. He has published 17 books and numerous articles, and his book Enabling Knowledge Creation won the Association of American Publishers’ Award for the Best Professional Business Book.