Subject: Re: A software marketplace
From: "Benjamin J. Tilly " <>
Date: Thu, 09 Jan 2003 23:17:00 +0500

Russell Nelson <>wrote:
> Click on this URL to check the validity of
> Benjamin J. Tilly  writes:
>  > Unfortunately I have trouble believing that the Street
>  > Performer's Protocol is appropriate as a general model
>  > for free software.  The problems that I see are manifold:
>  > 
>  >  o The definition of a "done project" is hard to nail
>  >    down.
>  >  o A well-defined project may not, upon implementation,
>  >    be the right project.
>  >  o Think through the incentives for the developer very,
>  >    very carefully.
>  >  o Unfortunate social dynamics may develop as people
>  >    fight over who deserves rewards, and as people who
>  >    are looking for a reward come into conflict with
>  >    people who are doing things for personal reasons.
> All of these problems are inescapable when you pay
> people to write software.  Since that is what the
> Public Software Fund does, we'll have to deal with them.
> Not a one of them is particular to the SPP.

I am not so sure that they are inescapable.  I will
agree that they are general, and any alternatives
which escape them bear their own problems.

For instance when you pay people to _have written_
software (eg through awards) then the first two
consequences become largely irrelevant, and the third
is lessened.  However you now get issues such as
identifying who to pay.  And if only a few big names
get awards, potential authors may not feel much

This happens now, but the winners of such awards tend
to be the "usual suspects" such as Jeremy Allison,
Larry Wall, and Richard Stallman.  It is hard to say
how much incentive there is in general to try to win
these awards.  However it would be interesting to see
what you could accomplish with, say, a well-publicized
university scholarship in CS for someone "who has
contributed to free software".  (I have no problem
thinking of plenty of college-bound highschool kids who
would be worthy winners - and more who would try to win
something like that.)

It is also possible to have a grant committee kind of
set-up which pays people to _work on_ software.  Again
the specifics of determining if goals have been met
are improved.  But motivating people to donate becomes
an issue - the flip side of the goal issue.

Currently The Perl Foundation has been pursuing this
approach with qualified success.  Grants have long
been successfully used in other areas of life, such as
for medical research.

However a fundamental problem that I see with any
donation model is that there is a huge discrepancy
between the amount of money that most people have to
spend, and what is needed to provide a real incentive
for well-paid professionals.  A small minority of
people have the resources to bridge that gap.  All but
the smallest of companies have the necessary resources
but are not at liberty to be free with them.  Thus I
doubt that we will ever have a large fraction of open
source development being funded financially by
charitable donations.

>  > in which case what I pay has to cover that person's taxes,
> It still has to cover their taxes.  [...]

Oops.  Misunderstanding on my part.

The last time that I worked for a non-profit was many
years ago, in a different country (with different tax
laws), and besides which, I was paid under the table.
Hence my basic ignorance...


Get your free email from

Powered by Outblaze