Is it worth starting a new open source project or joining in?

Question

There are a few various open source projects that I'm not particularly happy with, for example, I feel that various aspects of sound on Linux is simply appalling when compared to Windows. I know this is because sound card manufacturers make their hardware for Windows, and it's probably their fault that support is so bad. But I'm sure we can do better, right?

For example, I find that pulseaudio is extremely buggy. I tried to get pulseaudio working over my network (so I can share my speakers) however, this didn't work due to a whole raft of "known bugs" that have been around for years.

Also, another gripe is that Skype blames pulseaudio for Skype's lousy support.

My question is, is it worth starting a new Linux sound project (and potentially adding to the chaos), or helping pulseaudio by submitting patches?

Answer 1

The time to start a new project is when you see an unaddressed problem and know how to fix it...

When your domain is Linux audio, there is a lot to know, before you can claim to "know how to fix it". The people responsible for ALSA, PulseAudio, and other projects have thought about this problem.

Odds are very good that there are smart and capable people already thinking about the problems that you perceive. You are almost certainly better off helping them that starting over again.

Also note that there is a chance that the real problem is one of configuration and deployment, not underling code. And that might be a simpler problem to contribute to.

Answer 2

Getting involved with the discussion in and around an open-source project is a great way to make a contribution - and your patches will be considered more seriously if you're contributing positively in any way you can. Joining an existing project means there are other people already interested, and with so much time invested already, they may be more resistant to moving to a new project. Take advantage of the existing interest and join something already in progress!

Answer 3

purely my opinion: Open source relies heavily on community support (for me the keyword is "community"), so starting a project similar to what already exists is really dividing the community, rather then bringing it together. So in short, dive in and dedicate your knowledge to the existing project. I am sure everyone will be glad to have an educated and skillful person on board.

Answer 4

I'd opt for helping out with pulseaudio, since that way you can build off their existing infrastructure rather than chipping away at it with competition. No sense in reinventing the wheel (or the wheel manufacturing process).

For FOSS, I would only suggest starting a new project if the existing software projects were so totally broken you don't think there's any hope for them. (Or as an educational experience, of course)

Answer 5

I would submit patches. Otherwise, you risk inventing the wheel and adding to the chaos.

Joel wrote about this indirectly.

Answer 6

It would be silly to preemptively fork the project when you don't know if you have to. Submit your patches. If they aren't being responsively handled (or the whole design needs to be redone) then think about possibly forking.

Answer 7

Please, don't start Yet Another Linux Audio System. Read Linux audio: it's a mess and Welcome To The Jungle before you code anything.