The Cathedral and the Bazaar
Musings on Linux and Open Source by an Accidental Revolutionary
The big idea
There are two ways to build something: quietly, by a small group of experts, released only when it's finished — or loudly, in public, half-broken, with everyone allowed to poke at it. Raymond's claim is that for software the noisy way wins, because the fastest debugger ever invented is a lot of other people's attention.
Why it matters
The cathedral is the old model: careful, hierarchical, source code handed down complete. The bazaar was Linux — Linus Torvalds shipping kernel snapshots sometimes more than once a day, to a crowd of strangers, and somehow getting a coherent operating system out of it. Raymond thought that shouldn't work, so he tested it. He took over a small mail-fetching program, rewrote it as fetchmail, and ran the bazaar method on purpose: release early, release often, treat every user as a potential co-developer. It worked, and the essay is his write-up of why. The line that stuck is what he named Linus's Law — given enough eyeballs, all bugs are shallow. The point isn't that crowds are smart; it's that a bug which is baffling to the person who wrote it is often obvious to someone who didn't. Different people carry different blind spots. The essay had real-world teeth: when Netscape announced in January 1998 that it would open-source the Navigator source code — the decision that eventually produced Mozilla and Firefox — its executives publicly credited Raymond's paper as part of the reasoning, and brought him in to consult. Worth holding the caveat too: Heartbleed sat in OpenSSL, one of the most-deployed pieces of code on earth, for about two years before anyone noticed. Eyeballs only help if they're actually looking.
Use it today
Take the thing you're keeping private until it's good — the draft, the design doc, the half-working prototype — and show it to one person now, at sixty percent. Say plainly that it's rough and ask what's wrong with it. You'll spend ten minutes feeling exposed and save yourself a week of building the wrong thing carefully.
Drop it in conversation
“The cathedral versus the bazaar: build it in private and release it perfect, or build it in public and let everyone find your bugs for you. The second one feels worse and works better.”
In real lifeA cook who tastes the sauce and adjusts as she goes ends up with a better dinner than one who follows the recipe silently and serves it hoping. A student who shows a messy outline to a classmate in week one finds the flaw in the argument; the one who polishes the whole essay alone discovers it the night before it's due. Teams that demo an ugly working version every Friday catch the misunderstanding in week two instead of month four. The pattern is the same each time: the cost of showing rough work is a moment of embarrassment, and the cost of hiding it is all the effort spent going the wrong direction.