The Idea Dude

CONNECTING THE DOTS ONE AT A TIME

Friday, June 01, 2007

Are you building cathedrals or bazaars?

I just finished a 48 hour programming stint with about 6 hours sleep in total. It's not the first time nor will it be the last. It is just one of the occupational hazards of software development. We have one happy customer and one exhausted developer... me. You know you're past your physical limits when you go shopping in the local supermarket and you can't name half the fruit you sit. Yeah, it's a little scary.

Earlier, sitting on the subway train, I wondered what I could have done to make this week a smoother path. We had another 2 developers on tap but their contributions, though important, was more peripheral than core, tactical rather than strategic. The time for mass collaboration is not now, it will be soon. The answer is very simple, in the beginning, you want one vision, one system philosophy, one architect. Once you've defined this, you can get in all the craftsmen to do their thing. But in the real world, you seldom have mass collaboration of architects, designers, etc. Why? you need a single consistent vision regardless of whether it is right or wrong.

Nicholas Carr writes about this in his article called "on the ignorance of crowds". He refers to the model Eric Raymond's paper called "The Cathedral and the Bazaar". Eric examined why open source is such a power mass collaboration paradigm. To the cathedral architects, it should not succeed but it does.

But here's the real truth. Most, if not all, open source projects started with one individual or very small group of contributors. Once it is past a certain point, the successful ones are taken to another level by their peers through mass collaboration. The cathedral is designed in relative isolation, the bazaar is built by peers. But you could probably not have switched roles either way, it would have been an utter disaster. Carr says it best, peer production is refining the old, rather than inventing the new, it's an optimization model vs an invention model. If you can give each a place in your business, you have a great recipe for creating something special and yet let your success be defined by the world rather than the limitations of self. Companies struggle to do both, Microsoft builds cathedrals, Red Hat builds bazaars.

The success of a restaurant is usually due to the genius of one chef. As they say, too many cooks spoil the broth. Once that broth is defined, then the conditions for mass collaboration are right. Good menus have a theme, a philosophy, a vision.

BTW: The post with the best blog title of the day goes to Jules, Call me at Low Tide.

0 Comments:

Post a Comment

<< Home