Ron,

> So now the hard part.

I started working on this in 1986. Brad Cox tackled these issues in his
"superdistribution" work. At the time I asked around if people would put
up with ads to be able to get software for free and the reaction was so
negative that I didn't expect our world to become what it is now.

> How do you determine the value of a contribution?  

By 1997 I was working on the idea of bulding a metering system into the
virtual memory code. So a contribution (which could be any objects, like
a drawing of a cat, and not just code) would be rewarded a portion of
the user's monthly software bill proportional to how much it needed to
be kept in RAM compared to all other objects.

Distribution would be for free and there would be no cost to having lots
of software on your disk that you didn't actually use.

As a software stack grows, people are less and less willing to pay for
the older fundamental parts even though these still provide a lot of
actual value. So someone might think of forking $90 for just the Trumpet
WinSocket TCP/IP stack to access the Internet in 1995 but would think
even $1 totally absurd for just a TCP/IP stack in 2005 when it was such
a tiny part of an OS. This race to the bottom happens even without no
cost Open Source competitors, though that does speed things up.

My scheme was meant to avoid this trend as well as make each individual
contribution values (so you can get money for a single font and not have
to bundle it with 200 more before you can make any money with it).

> How do you determine the value of the software to a company
> to determine what they should pay? 

The companies wouldn't pay anything at all, but could just build stuff
on top of other people's works. The end users (which might also be these
companies, but that is a different story) pay, and part goes to the
company for its efforts and part to all other authors.

-- Jecel