Subject: Re: Open letter to those who believe in a right to free software
From: "Stephen J. Turnbull" <>
Date: Thu, 4 Nov 1999 20:33:19 +0900 (JST)

>>>>> "rms" == Richard Stallman <> writes:

    This is false.  A model is a tool; there are no requirements on a
    model that do not derive from the modeler's purpose.  Since you don't
    understand mine, and claim to feel no need to understand the model for
    the light it could shed on my purposes,

    rms> I thought I understood your purpose, but perhaps I don't.

    rms> Learn about your purpose by studying your model would be
    rms> doing it the hard way.  I have too much work to do to do it
    rms> that way.  Would you like to say more about what the purpose
    rms> is and what it isn't?

This is copied from mail to a third party sent very recently; although
you were CC'd, you may not have seen it yet.  I think it's reasonably
on-topic for FSB, so I'm reproducing it.

    SJT> You [3rd party] have implied that you think I'm trying to
    SJT> explain why people work on free software and how to run FSBs.
    SJT> The first is false on an individual level; I merely will
    SJT> assemble a list of plausible motivations and treat the
    SJT> distribution of those in the population statistically.[1]
    SJT> (Note that Richard is wrong about his small subpopulation
    SJT> hypothesis; it is quite possible for the behavior of small
    SJT> subpopulations to affect the whole in economic models.  There
    SJT> is now a huge literature on evolutionary dynamics in game
    SJT> theory, which may or may not be satisfactory but clearly
    SJT> demonstrates that we economists are sensitive to this point.)

    SJT> It is correct to say that I expect application to FSB
    SJT> management, but it is indirect.  The usefulness to FSBs will
    SJT> derive not from the specific behavior adopted by FSBs in the
    SJT> model, which will be too abstract and smooth to be realistic,
    SJT> but rather from the improved understanding of how the
    SJT> environment will change under the influence of FSBs and
    SJT> respond to them.

    SJT> I also wish to quantify benefits of free software which do
    SJT> not appear in current analyses of the industry, to better
    SJT> inform policy-makers whose moral principles admit such
    SJT> analyses as relevant.  I hope to make this quantification
    SJT> much more comprehensive than current models admit.

[1]  I would also like to note here that this is mathematical statistics,
not practical statistics.  Of course RMS is correct that a practical
random sample would almost surely not pick up the few advocates like
him, but that does not prevent us from looking at models in which
invisible subpopulations grow to dominate the population.

And the word "clearly" in the parenthetical remark is perhaps
misleading.  If you know that literature, it would be clear, because
the theoretically preferred (static) equilibria are those that recur
with probability one even though the system passes infinitely often
through (static) equilibrium states where the (preferred) equilibrium
actions occur with probability zero.

My apologies to Ben Tilly if I'm using inadmissible infinities ;-)

University of Tsukuba                Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Institute of Policy and Planning Sciences       Tel/fax: +81 (298) 53-5091
What are those two straight lines for?  "Free software rules."