Thanks for the comments.

Yes it’s a struggle.

Free software is a competitive market that values creativity and excellence in tools and applications. New and fast applications.

RoR is successfull not because of what it does (web database) or what it is (a LAMP) but because of how its applications are designed, produced and presented.

The model is to develop one application. Possibly not yet another web framework if you want to avoid a crowd of projects. Preferrably something big, like porting a large library or implementing big ideas.

I could never have reached the first milestone in Allegra without free software like Python, GNU/Linux and the Debian project. Full stop.

You discount your own increased productivity. We have, for free, all the tools and API available since GNU/Linux broke out of the Internet. And they are of much better quality when it comes to reliability and intelligence. They just do the right thing and perform better. And they keep improving.

Application development with the free software tool chain is so productive that it is as spread well beyond the OS barrier. I’m using Eclipse on Windows to test on Linux with Firefox an CPython application.

My experience is that free software performs better than any other OS in most application. It’s been tested by UNIX security experts: GNU tools just have less buffer overflow.

Not because there is only a miracle at work (the public good). Because there is also a huge markets for competitive application developers that can share their capital investment with their final customer.

If it’s application is worth it, it will yield more. One fin, a thousand device hackers and how many businesses depending on Linux? A lot. With businesses like IBM.

I’m in for a dream and the eventual benefits of commercial licences. Fancy drinks in ball room full of nerds, maybe books to publish. And sexy women aroused as I talk about recursion :-)

Regards,

I think what’s key here is that even people who write free software use it. I think this would fall under the guise of “adoption” as mentioned in the article. The price of GCC or Python, for instance, is far less than for VC++, CodeWarrior, VB, etc. If it weren’t for open source, many people wouldn’t be able to afford the entry cost of programming so the issue would be moot.

That being said, I do think there is a problem for writers of free software (as can be seen from the many OSS projects asking for money).

However, many, if not all, people who write free software probably fall into one of several large categories:

a) People who write software to deal with some personal need and then release the source. They were going to do the work regardless of pay, and even if they kept the source closed, probably weren’t going to make money from it anyway.

b) People who are paid to write software and their companies (or clients, if they are contractors) allow them to release the source. They are getting paid regardless.

c) People who are major players in large and popular projects (i.e. Linux kernel). They can be the beneficiaries of grants, fellowships, etc. Also, these projects often boost the person’s perceived worth tremendously which can land them lucrative jobs (often being paid to continue the work they previously did for free).

I think the last group is where the lack of a business model hurts the most. There are many large projects that just aren’t quite large enough or don’t have enough commercial appeal to draw money, but are large enough to consume much of a developer’s time. It’s a bit like college sports: only a few become professional players and reap the rewards of their work and the rest are left to struggle. Nevertheless, I think there is a large overlap with the first group (a) and most people would do it even if they knew upfront that there would be no financial gain.

I’m not actually disagreeing with your point, but your other examples are all paid for from public funds :

“good schools, laws well applied, roads that go somewhere, bridges that cross and safe harbors.”

What is lacking is a clear business model for those developing free software. Advertising revenue works well for a few (google), but those that rely on consultancy as a spin-off of free software *often* struggle. Are there alternative models ?