open source advice: 3 + 3

I’ve never before actually answered a blog meme (unless you count this.) But sogrady tagged me with one that I thought was interesting. Here goes: three things I would tell every open source business to do; three things I would tell them not to do. (In an attempted to do more than just repeat what others have said, these may be a little more abstract and a little less useful, but I hope they are fun.)

To Do:

• do emphasize volunteer QA: one serious, concrete piece of advice, first. Most of your users want to help you make your software better, and QA is a pretty low-barrier way to let them do that. So take advantage- invest in public bug tracking systems; invest in people whose job it is to help volunteers use those systems; invest in ways to get regular builds to volunteers so that they can test and verify on the latest and greatest. User-driven QA is very, very difficult for proprietary software to do, and should be a great differentiator (both in quality and in cost) for open source.
• do something revolutionary: software is more fun and more dangerous (to the other guy) when it is really revolutionary. Not merely innovative- ‘we want to do something new’ but revolutionary – ‘we want to actively overthrow the old.’ We are here because we revolutionized the method of software production, but to continue to grow, we have to start revolutionizing other things too. Be the Wii, not the PS3. If you really want to catch them by surprise, be a revolutionary using open source, but not in software. The people outside software (well, except for poor Encyclopedia Britannica) don’t yet realize what is coming for them.
• do it with joy: if you’re in open source, you’ve given away most of the weapons past software companies used to build a lead and then coast – you can’t get lockin and you can’t get a monopoly. So unless you get bought out and retire, you’re going to be on this treadmill for a long time- someone will always be at your heels. So have fun, or the treadmill will burn you out very fast.

To Not Do:

• don’t be afraid to embrace freedom: open source businesses tend to be allergic to the word ‘free’. That is a mistake. Say ‘freedom’ a lot. Love freedom. Embrace freedom. Your community likes freedom. It differentiates you from the proprietary competition, and if you embrace it wholeheartedly (not just this weak ‘openness’ stuff) it will differentiate you from most of your open source competition too.
• don’t let your community’s fears drive your feature choices: if you’re a real open source company, you’ll have very direct contact with your customers. This is normally a great thing, but when you make a decision they don’t like, particularly if it scares them, you’ll hear about it loud and long and clear. If you let that negative feedback drive your decision making, you’ll never grow beyond the needs of those people. You must not be afraid to piss them off when you truly believe that a design decision is for the broader good. Remember, the pissed off people scream- the happy people just go on with their lives. So you can’t just say ‘more people screamed than thanked us’- that isn’t a useful metric.
• don’t stick around if this isn’t for you: This is just not for all of us. Open source is people, in all their messy glory, interacting primarily as partners and friends rather than as hierarchical worker bees. Some people take to this like a fish to water; I dare say that most people should. But not everyone will. If you don’t – if you think this should just be about a copyright license – get out now, and spare both yourself and others the irritation of dealing with you in a situation you’re just not suited for.

Can’t think of anyone I want to tag right now; maybe in the morning.