This past week Microsoft's Craig Mundie delivered an important message about the future of computer technology. We agree with Mr. Mundie on several points. For twenty years, innovation has driven the computer industry to a vital position in the global economy. As an industry, we have ignited economic boom and and fought through economic downturn because of the power of innovation. We disagree, however, with Mr. Mundie’s conclusions. The "Commercial Software Model" he describes is a false promise. True innovation, in which technology providers and customers are equal winners, will come from the Open Source Way.
The bedrock of the computer industry is computer science; without research and development there is no innovation. The very enterprise of science is based fundamentally on sharing and cooperation. Without open access to information, scientific research is impossible and innovation will wither away. Microsoft, of course, claims to understand this. Mr. Mundie goes to great lengths to point out that Microsoft's "new" Commercial Software Model will be based on a "Shared Source Philosophy." But Mr. Mundie pleads that this is not just a case of Microsoft attempting to embrace and extend Open Source:
"Many people will attempt to say that Shared Source is Microsoft's failed attempt at being an Open Source Company. This could not be a more incorrect statement. Shared Source is not Open Source. We recognize that OSS has some benefits, such as the fostering of community, improved feedback and augmented debugging. We are always looking for ways to improve our products and make our customers more successful, and to that end we have incorporated these positive OSS elements in Shared Source." (Craig Mundie, May 3, 2001)Microsoft's notion of sharing, however, hardly touches what's central to the cooperative spirit of Open Source. The very essence of community implies equal participation in the creative process, and shared access to the rewards of creation. That's what makes the scientific community a community, and that's what makes Open Source developers a community. Leading technologies like Sendmail, Apache, Linux and Samba benefit from a world wide team of developers motivated by the benefits of collaboration:
Mr. Mundie fears what he calls the drawbacks of Open Source, such as the possibility of unhealthy forking of a code base. He alludes to the early days of Unix when a number of competing variants arose. In fact the forking in Unix occurred because it wasn't open enough. Competition arose between proprietary variants like Solaris and SCO Unix. The Open Source Unix, BSD, has enjoyed a consistent development effort with a high degree of interoperability between its variants. This interoperability has been sustained because the development process is open, and because it's easier to contribute incrementally to the common code base than to start a new development branch.
It's worth noting that Sun Java vulnerable to forking by keeping it proprietary, and the Microsoft has guarded against the same risk by making C# an IEEE open standard. The moral of the story: Open Source is the cure for code forking, not the cause of it.
Mr. Mundie also implies that Open Source licenses like the GNU General Public License (GPL) disrespect intellectual property and threaten to force companies to throw away their intellectual property. In fact Open Source developers have the utmost respect for intellectual property; they simply choose to exercise their intellectual property rights to create a common good. A fierce respect for individual rights is essential to inspiring innovation. Furthermore, the GPL is only one of several Open Source licenses, and the GPL-licensed software has been successful deployed in conjunction with commercial software by many companies. If Microsoft suffers from fear, uncertainty, and doubt on how to co-exist with Open Source licenses, we're confident that other companies can show them the way.
Nowhere is the contrast between the Shared Source model and the Open Source model greater than in the contrast between the Microsoft Developer's Network (MSDN) and OSDN:
Any platform succeeds in large part through the efforts of independent software vendors who create a compelling selection of programs for customers. Microsoft understands this kind of software diversity as a competitive advantage, and indeed software diversity was a key part of the Microsoft advantage in its early competition with Apple. The purpose of a development network is, in large part, to foster development producing this kind of diversity.
In fact OSDN is tremendous success at facilitating and accelerating this kind of innovation. The benefits apply to all platforms. Roughly one in five development projects on Soureforge is for a Microsoft operating system, totalling nearly 4000 projects in all.
OSDN doesn't offer these services as an act of charity. OSDN acts to drive innovation from which all participants in the community will benefit.
Mr. Mundie also challenges the business model he ascribes to Open Source. He professes alarm at the notion of giving some technology away to leverage others. This is asking
"developers to give away for free the very thing they create that is of greatest value in the hope that somehow they'll make money selling something else. In effect, it puts at risk the continued vitality of the independent software sector." (Mundie, 2001)
Ironically, this could well be a description of Microsoft's practice of giving away Internet Explorer and Internet Information Server, a practice Microsoft has refined to the point that courts consider it monopolistic and predatory. Indeed, the last sentence in Mr. Mundie's statement reads like part of Judge Jackson's ruling.
In fact Open Source business models resemble other business models in the technology sector. Whether the business model is service, solutions, or value-added products, success will come to those who are perceived as leaders in innovation.
It's easy to confuse control with leadership. Proprietary technology locked inside a closed standard provides a comforting degree of control. This kind of control, however, discourages participation and raises barriers to innovation. No company can lead innovation alone. Even Microsoft, for all its size, and for all the impressive brainpower it has recruited to Microsoft Research, needs a community of active participants to realize the full potential of innovation. The move towards a Shared Source model shows that Microsoft is beginning to recognize this need.
Microsoft needs to move fully down the path of shared source to a genuinely Open Source model. Then it will discover that what appears to be the risk of relinquishing control is in fact the opportunity to assume leadership. We expect Open Source to lead innovation going forward. We hope Microsoft will follow.