Re: economic efficiency of free software

[email protected] wrote:
> Yes, it is obvious that you have not done development on your site in 
> a couple of years because it lacks the features that are now expected 
> in a community based site.  For example, it lacks user preferences, 
> filtering, mailing lists or web-based forum, or a rating system. 
> There is a site that offers most of these features and a few others, 
> Freshmeat.

I agree, although there are many improvements still possible in
"forum" technology, such as integrating web and email better, better
readers etc.

> The bug database system you mention would be adding additional friction 
> to developers.  It can be difficult to get them to accept bug reports 
> when emailing them directly.  You want them to go to your site and 
> check a db for bugs concerning their code?  Not likely.

It depends on the project.  GCC and Mozilla developers love their bug
databases for example.  For GCC, it works partly because the database
is well integrated with a mailing list, so the traditional methods of
working still work.  Then the database becomes a useful tool, rather
than a burden.

For smaller projects, I imagine a bug db offering will only be used if
it is _very_ easy to use and eases the burden of dealing with users
and their bug reports.  One way is if a bug db gates nicely to and
from a mailing list, and is good at supporting volunteers who help
with bug filtering and basic technical support, freeing the author(s)
of a small project to work on the project itself.

In theory anyone can set up and customise Bugzilla, RT etc.  But lots
of people use Yahoo Groups despite Mailman being easy to set up,
simply because it's convenient.  So a very convenient and _good_
bug/tracking db whose structure reflects a structure which helps small
project developers do what they want might be used.

If you managed to get payments integrated so that people who submit
bug reports or feature requests, and/or developers _receive_ a small
payment for those actions, while people can _sponsor_ bugs or features
which are already in the db (which is different from doing the work of
creating them), or just as good (perhaps better) sponsor projects as a
whole, or declare which general areas the money goes to: that would be
an added incentive for developers and/or reporters and/or intermediate
support volunteers to use the system.  But it must still be a very
good and helpful system.

Btw, it would make sense for developers or other development
stakeholders to be in control of how sponsor funds are actually
assigned in detail - with good tools to relieve the adminstrative
burden of calculating those assignments.  It's essential for those
people to retain control of the direction of their project, and not be
subject to that gnawing feeling they are working for tiny short term
goals they are doing just for the money.

-- Jamie