> but back to chandler, Rosenberg sets out thinking that the project has the
> advantage of being free of the conditions that are usually blamed for
> failed software projects:
>
> -- it's generously funded by Mitch Kapor and, later, investors and grants
>
> -- an abundance of big brains with plenty of experience under their belt
>
> -- no hostile enemies out to sabotage them
>
> it is still a huge failure -- in terms of taking forever to release, making
> bad decisions, the volunteer spirit of the community flags not just in
> terms of contribution but in terms of finding bugs in the first place
> (something like 2% of all bug finds come from the vaunted foss community)
>
> 6 years and millions of dollars later, chandler is not at all ready for
> prime time.
Chandler is not typical; I visited their offices to give a talk once, and it was a lot more like a lavishly funded non-profit foundation than a typical startup. Way different than most free software projects, which are mostly underfunded and often run by volunteer work only.
Regardless, this example illustrates why free software is important: Chandler is basically dead, but their software is still available to anyone. Same is true for a couple of different dead free software companies my friends worked for: the code lives on, and in their case it's useful code, used by many. If they had been writing proprietary software, a cancelled project or a failed company would mean no more access to the code for anyone, ever.
If I leave my current job, I will no longer have access to the code I've written there. My labor has been alienated. Working on free software, programmers always have access to the results of their labor, as does everyone else.
In other words, free software is about control, and power. Whether it produces *better* software (sometimes, maybe) is far less important.