The Essence of Open Source

This week I am giving a keynote address at the Red Hat Summit in New Orleans.  When giving keynotes or other talks these days, I often make a point of explaining what I feel is, quite simply, the key to open source initiatives and why open source is an integral part of IBM's strategy.  I do this because there is quite a bit of FUD surrounding the issue.

Some say the primary value of open source is to provide access to the source code of previously proprietary software, source code that had been available only in binary form.  Others spend a lot of time talking about open source software licenses to bemused audiences, and in particular railing against the GPL, which is the license used by Linux.  Open source licenses are an important topic for those involved in reading, modifying or re-distributing open source software.  But to the vast majority of programmers, let alone the vast majority of people who simply use computers at work or home, open source licensing is a pretty arcane subject, more of a red-herring brought up by competing IT vendors than anything they really need to obsess over.

Finally, people sometimes debate how free open source software really is, referring to the actual cost of the software, even though open source advocates go to great pains to point out that when they say free they are not talking about price.  The Free Software Foundation points out on their web site that Free software is more a matter of liberty than price; think "free" as in "free speech".

To me, open source is all about collaborative innovation, that is, working with smart people all over the world as a community to solve important problems.  In the world of research, where I have worked from the time I was a graduate student in physics at the University of Chicago in the '60s, this collaborative innovation is nothing new.  Collaborating with colleagues in your discipline is how you make progress, whether that discipline is physics, medicine, computer sciences or law.  That is why there are open professional journals, where it is an honor to have your papers published, usually after having been refereed by a group of experts, and where it is an even greater honor to have your colleagues read your paper, use what you say in their own research, and give you credit by citing your paper.  In fact, Tim Berners-Lee invented the World Wide Web for the express purpose of facilitating information-sharing in the research community.  The ability to work collaboratively with a community has changed dramatically for the better since the advent of the Internet and World Wide Web.   

Now, when you collaborate with your colleagues, they have to be able to read and understand what you say, whether you use a natural language like English, or mathematical notation, or tables of numbers.  Likewise, if the collaboration involves software, then you would expect to be able to read, modify and generally share the source code of the software on which you are jointly working.  Thus, in my opinion, open source software is just a by-product of, or rather a necessary precondition for, collaborative innovation involving software.  Nothing more, nothing less.

In the end, the real value of an open source software initiative is the quality of the community that is participating in the initiative, and what they are then able to design and build.  If you consider Linux, Apache, Grid and other large, successful open source initiatives, they have attracted a huge following of top programmers, computer scientists and other technical experts around the world, all of whom are collaborating in developing and supporting different aspects of the shared IT infrastructure that the Internet and open standards have made possible.  These are such important and complex problems, that only through the kind of collaboration made possible by open source software can we hope to attain the needed levels of innovation and progress.