Have you noticed the kind of market segments that succesful free and open source projects are aiming at? I have, and I've got this problem with it.
I came to think about this because I just looked at this old survey about who the people who write free and open source software are and what their motivations are. While the presentation itself is interesting, it was one of the additional slides in the end that captured the essence of my problem. It says:
How likely would you be to contribute to a free/open source software project that delivers more value primarily to average users than to you or your peer group?
I think an honest answer from any open source contributor would likely be "Not bloody likely". The reason for this is that open source projects are traditionally initiated by a programmer because he needs the software, or - as it's often put - he feels an itch that needs to be scratched. He's not very likely to scratch someone else's itch.
Steven Pemberton writes about the similar problem with the usability issues with open source software. The programmers create programmer-oriented user interfaces that work fine for them, but are hard to use for the general public.
Still, I have found nothing in the survey that says that people work on open source projects because they want to scratch a personal itch. The intellectual stimulation and the community around the projects seems to be the driving force. If that is so, it really should be just as enjoyable for these programmers to work on any project that is as stimulating, even if they feel no need to use the actual end product. But I guess people join open source projects in the first place because they use the software and feel they want to work on it as well, and that is the reason that successful open source projects either solve problems that programmers normally have (compilers, IDEs, generic library code) or belong to a broad mainstream category (web browsers, operating systems, web servers, file sharing software, media players).
The above may sound very negative about open source software, but really, I love open source software. I use it extensively. One reason for that is that I am a programmer, so I have a wide range of open source software available to help me with my job, while other software is just useful (e.g. Firefox). The problem I have experienced is that when I have needed software that is not a programming or mainstream tool, I have not found good free (as in beer) software. That market is still owned by commercial software. (Note to self for the software company business plan: Don't make software for programmers or the general public.) The kind of software that I have been looking for is for music and audio production, video editing, scriptwriting and other media related areas. One might think that there would be quite a few open source projects in these areas, considering how popular media software in general is, but note what audience the numerous media related open source projects out there are for: the media consumers, not the producers. I guess this just mirrors the poor ratio of media producers to media consumers. There are some mainstream tools that can be used by both consumers and producers, e.g. video encoders, and those work well, but the media producers have to resort to commercial software.
Oh well, I guess this is my personal itch. One day, I'll find the time to scratch it. I wonder if I'll write a video editing suite or a MOD tracker that actually has a sensible user interface... Or perhaps I'll just stick to making games.
For the purists out there: Sorry for the mixed terminology above regarding free and open source software. I know they are not the same thing.