Subject: Re: Economic incentives to produce software in a free software regime
From: "Stephen J. Turnbull" <>
Date: Tue, 26 Oct 1999 20:01:15 +0900 (JST)

>>>>> "Ben" == Ben Tilly <> writes:

    Ben> Would you please reiterate the explanation of why bundling is
    Ben> inefficient when the actual cost to the hardware vendor of so
    Ben> bundling is nil?  (It is nil because the hardware vendor is
    Ben> aquiring the software for the same cost that the consumer
    Ben> could aquire it - and it comes with no strings attached.)

Oops.  You're right on the production side.  I can claim that the mfr
has to support the software to some customers who would otherwise
forego the service, but that's really weak.  I can imagine such costs
being substantial---but only if the mfr has effectively decided to
shoot himself in the foot; they should be avoidable with a little

Ie, if the mfr locks himself in with promises of upgrades of the
software or something like that.

    Ben> I fail to see with free software how anyone is forgoing
    Ben> monopoly profits.  The software would not exist in its
    Ben> current form without being free - and so no profit has been
    Ben> forgone!

I'm afraid I'm missing your point, but here goes anyway.

Ask Peter Deutsch or any of the proprietary X Window System vendors how 
to extract monopoly profits from free (or nearly so, or will be so)

The argument that something "would not exist in its current form" is
unanswerable, of course.  It's purely definitional.

Some software (any Linux kernel that would be worth monopolizing)
truly would not exist in its current form if at some stage Linus had
tried to "take it private."  Emacs is another matter; fortunately, the
FSF is bound by the assignments and its charter, and RMS is IMO
utterly trustworthy in this.  But any single-author work can be taken
private at any time, taking all patches of less the 16 lines, I think
it is, or assigned to the single author, with it.

You can argue that this is violating the trust of contributors and
users, and in most copyleft cases it would be.  But for non-copyleft
licenses and semi-free licenses, the authors explicitly indicate their
intention to make non-free versions of the product if it suits their
business plans.


The main point of my argument, though, is on the consumer side.  Some
users just don't want no DOS.  The right to be free NOT to pay for DOS 
on your 80x86 box is important.  Let the people who want it pay for
it.  And it is an inhibiting factor; although throwing away MS-DOS and 
installing Linux (even at much higher prices than anyone will ever pay 
;-) is surely economically sensible, throwing away MS-DOS and
installing DR-DOS at some small price ($75 or so) may not be worth it
if you don't need the additional functionality.  Even though given the 
choice of MS-DOS for $25 and DR-DOS for $75 you might have opted to
pay the extra $50.

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."