Subject: Re: Open letter to those who believe in a right to free software
From: Ben_Tilly@trepp.com
Date: Mon, 25 Oct 1999 09:32:17 -0400


> >>>>> "Crispin" == Crispin Cowan <crispin@cse.ogi.edu> writes:
>
> "Stephen J. Turnbull" wrote:
>
>     >> >>>>> "Crispin" == Crispin Cowan <crispin@cse.ogi.edu> writes:
[...]
>     >> Because there would be less software, of lower-than-necessary
>     >> quality.  This matters a lot to non-programmers.
>
>     Crispin> I don't believe that this claim is sufficiently
>     Crispin> supported, either.
>
> It's not.  It is the considered judgement of an expert in the field
> who has not yet done, let alone presented, a formal analysis.
>
An expert in ***WHICH*** field praytell?

Economics.  NOT software development.

It is the considered opinion of many experts in software
development, including many on this list, that software development
is not as simple as many other fields.  In particular the availability
of cheap and reliable components significantly reduces the
marginal cost of producing new software.  Since free software is
producing those components, free software is reducing the
marginal cost of producing software.  Since, unlike with proprietary
RAD tools, the companies producing it do not gain from forcing an
upgrade, the cost of maintaining that software goes down as well.

Given that the vast majority of software created today is custom
software (as evidenced by the fact that the vast majority of
programmers are working on software that is not, and never will be
for sale), the shift in the economics of what you can sell software
for could well be virtually irrelevant.

In fact  much of this custom software can benefit from sharing the
code with other companies (most of whom are not your direct
competitors), and so free software can encourage the release of
software from a higher proportion of professional programmers
than release software today.

Additionally it is the considered judgement of many experts in
software development that proprietary code tends to be of a
substantially lower quality than equivalent free software.  This
is both an observed fact and something that people have tried
to explain.  Therefore claiming that free software (which is very
different from shareware) will be of, "lower-than-necessary"
quality is something that needs quite a bit of support.

Have you taken these basic facts of software development into
account in your opinion?

> Since you aren't in a position to evaluate my credentials, you have to
> decide whether you want to believe your heart, or somebody who wishes
> the world were the way you wish it were, but on the basis of his
> education and experience doesn't believe it is that way.
>
Your credentials as an economist?  No.  But AFAIK you have not
claimed credials as a programmer, and you appear to not see
that software has some unusual properties as an economic good.

>     Crispin>   I'm not going to state categorically
>     Crispin> that it is false, but I don't take it as obviously true.
>
> It's not obviously true.
>
I am glad you recognize that.

>     Crispin> Naturally, my claim describing a "could be" world is VERY
>     Crispin> difficult to support, so I won't bother :-)
>
> That's my job.  That's what I plan to do, although I expect to come up
> with the opposite answer to yours.  If I get the answer you expect,
> then there will be joy in Mudville.
>
The chance of your getting the answer we expect depends in
large part upon your recognizing the factors peculiar to
software development that we have recognized.

> At that point, you can suggest that we start firing integrals at each
> other at 20 paces, or you can retire from the field in disgrace.  :-)

My training is as a professional mathematician.  Did you want to
compare Riemann integrals, Riemann-Stieltjes integrals, Lesbegue
integrals, Generalized Riemann integrals, or some more exotic
type of integral?

I rather think that my experience as a mathematician is more
relevant than your experience as an economist when it comes to
firing integrals off.  I likewise think that you should not dismiss the
judgement of programmers out of hand when it comes to
identifying what factors matter for software development.

Ben