Subject: Re: Hal's new white paper - Free Software/Public Sector
From: "Stephen J. Turnbull" <stephen@xemacs.org>
Date: Tue, 23 Mar 2004 14:27:35 +0900

>>>>> "David" == David N Welton <davidw@dedasys.com> writes:

        A consumer
        located at x [0, 1] gets a net utility from buying the closed
        source software of

                                Uc = v - tx - p,

        Similarly, the
        consumer's net utility from adopting OSS is

                               Uo = v - t(1 - x).

    David> Why should free software necessarily be the inverse of the
    David> proprietary software?

First, you need to understand the metaphor---see Jerry Dwyer's post.

If you do understand the metaphor but are still wondering, then
there's a technical detail.  It's a convenient simplification.  You
could have "z" be the degree to which your practices are adapted to
proprietary software and "y" be the degree for free software, with "t"
being the cost of adaptation.  Then with "v" being the maximum value
of perfectly adapted practices and software:

Uc = v - tx - p,    Uo = v - ty

The decision rule is "Buy OSS iff Uo > Uc", ie

0 < Uo - Uc = [v - ty] - [v - tz - p] = t(z - y) + p

It turns out that in this kind of analysis, both z and y appear _only_
in the form (z - y).  So s/(z - y)/x/g simplifies all the formulae
with no cost in realism.

Now you have to do something about the "hanging" z and y in the
definitions of Uc and Uo, and s/z/x/, s/y/1 - x/ is a convenient one.
It's not like you should care, since nobody (least of all the authors
of the article :-) knows how to measure x, y, or z.

This trick is sufficiently general that theoretical analysis rarely
bothers with more complex models.

N.B. The model _is_ unrealistic, but that's because it assumes that
"degree of adaptation" is the negative of a distance function.  Fixing
that is _very_ hard, because most of the economically interesting
ideas can be captured by defining "distance from perfect adaptation"
:= "cost of adaptation", and in most cases it's much easier to measure
the cost function than the production function.

The only thing I've seen that looks likely to be much improvement is
"fitness landscapes" (cf. Stuart Kaufmann, _The Origins of Order_, for
example).  Everything else I've tried, or seen tried in the
literature, reduces to measuring cost when you apply it to reality.


-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
               Ask not how you can "do" free software business;
              ask what your business can "do for" free software.