Open Source Quality Process

How come there isn't one? All the OpenSource tools seem to have this problem - once a strong developer moves on, they melt into a puddle of bullshit little hacks. MacQuality is wonderful, but it's also proprietary. How come we can't find some bazaar style means to do this for ourselves?

Lack of training and/or experience? Historically QA and development have been separate entities. OpenSource seems to have an abundance of developers and lack of QA people. XP's push for developers to learn QA skills may help this. Probably though, it won't be enough.

No, I don't think it will be. Where's the kudos in pointing out other people's bugs? At best, you'll be ignored. At worst, flamed because you didn't provide a patch.

At worst, flamed because you provided a patch.

You and I must have run across the same guy. But on another package, I submitted a simple but major restructure, and the maintainer gave me kudos and put it right in.

I think this depends a lot on the experience level (strength, if you will) of the maintainer and contributors. I'm experienced, and if I see an easy way to make a major improvement in the package, I'll do it. Even a minor improvement, if the bug bugs me. It's a matter of esthetics with me. But if you only have inexperienced contributors getting their feet wet, then you need a strong maintainer to keep order. I don't think it's a weakness with open source. That's the human condition. We just need to take it into account in managing and supporting an open source project.

At worst, some people only make their personal projects open source because it is "cool." They will flame any attempt to interfere, suggesting that if you don't like the code, you should go away and write it yourself. Then they flame you for forking "their" project. This isn't characteristic of open source. It's characteristic of people. It's useful to mention because many people like EricRaymond only focus on the other extreme--at best. But nothing is bivalent, everything is Gaussian. It's harmful to invent a general process that ignores these differences, but then again, it's harmful to invent a general process.