The Cathedral and the Bazaar by Eric Raymond

This book is available from Amazon.com

The title essay examines the question of how to make "open source" software projects work. In the background is the incredible success of Linux. In the foreground is the development of a mail utility with the author as the leader. Based on his experience, he tries to explain how it is possible to manage such a project -- with self-selected volunteers all over the world identifying and fixing bugs and contributing new code. The contrast is between software which is developed like a cathedral by a group working in isolation and only releasing the code when it is "totally finished"; vs. the open "great babbling bazaar" which is how open source development is described. He tries to understand what makes an effective manager in such an environment, and how to keep the volunteer team motivated and on target.

In the additional essays that appear in this book, Raymond takes a closer look at motivation, and comes to the conclusion that open source coders operate as in "gift economy" as opposed to an "exchange economy." The classic anthropological case of a gift economy was the Kwakuitl from the vicinity of what is now Vancouver. There social status was determined not by what you possessed but by what you could give away or deliberately sacrifice/destroy. This type of behavior/motivation is characteristic in cases of abundance -- whether for an entire society or a class within a society. The conspicuous consumption of Veblen's "leisure class" is an instance of this general phenomenon. It is easy to see it in operation in the exorbitant expenditures of Hollywood celebrities on parties and weddings, and the public charitable contributions of the wealthy. Raymond's application of this concept to the realm of open source coders is both unexpected and convincing.

He examines the behavior patterns of this set of people with the eye of an anthropologist, presuming that the truth lies in what they do rather than in the reasons they publicly state for their actions. He uncovers interesting contradictions between public statements and actual motivation, and makes a strong case for their close adherence to a rigid set of unwritten rules. As a key player in the very society that he is describing, he proudly takes on the very unanthropological role of helping these people better understand what makes them tick, as well as helping would-be leaders better understand how to lead in such an environment.

In other words, enabled by the Internet, a self-selecting group of people has evolved which operates with the motivations of a gift culture/economy. This culture crosses all geographic barriers and all social barriers, where membership has nothing to do with wealth or class in the traditional sense. Marx would have been dumbfounded.

As if that were not enough, Raymond goes on to make convincing arguments that two well-established "laws" of human behavior do not apply in this case.

According to Garrett Hardin's "tragedy of the commons," without law and supervision, a village of peasants will turn their common -- where all are free to graze their livestock -- into a mudhole. While they all might be aware that cooperation is necessary, without enforcement they each greedily try to grab as much as they can for their own livestock, which destroys the resource for all of them. Following such a model, one would expect that a team of software engineers could not stay together voluntarily for any extended period of time, that greed would lead to its inevitable and rapid collapse. But, as Raymond points out, "using software does not decrease its value. Indeed, widespread use of open-source software tends to increase its value, as users fold in their own fixes and features (code patches). In this inverse commons, the grass grows taller when it's grazed on." (p. 151)

Likewise, "Brook's Law" from the book The Mythical Man-Month by Fred Brooks, predicts that "as your number of programmers (N) rises, work performed scales as N but complexity and vulnerability to bugs rises as N squared, which tracks the number of communications paths (and potential code interfaces) between developers' code bases. Brooks's Law predicts that a project with thousands of contributors ought to be a flaky, unstable mess. Somehow the Linux community had beaten the N-squared effect and produced an operating system of astonishingly high quality." (p. 199)

Raymond's examination of what makes that possible leads to the inevitable conclusion that not only is open source development possible, but rather that it seems to be the only economically viable way to develop a large and important piece of software that will affect many people.

At the beginning of the book, it seemed like Linux was the anomaly -- a special case that might never reoccur. By the end, it appears that Microsoft is the anomaly -- that it is extremely difficult, extremely expensive, and almost impossible to build a program as immense as the Windows operating system without the help of a vast community of volunteers. With the Internet as an enabler, with Linux, Apache, and other projects as clear examples, and with Raymond's analysis of how it all works, open source seems to be the only logical way to go. The cathedral is not an alternative to the bazaar. Rather it is an historical artifact, an outmoded method of operation, left over from pre-Internet days.

Then with clear-headed balance, Raymond, rather than simply proclaiming victory, considers the limitations of open source, and when and how a balance of proprietary and open approaches is necessary.

All in all, this book helps open our eyes to an important new force that is changing the high-tech business world today. And at the same time, it leads us to re-evaluate what had seemed like fundamental concepts of human nature and destiny. He starts by asking intriguing questions about how software is developed, and winds up providing valuable insight into the question of "what is man?"

 
This site is published by Samizdat Express,     seltzer@seltzerbooks.com    privacy statement

	Webseltzerbooks.com