Subject: Back to business [was: the Be thread]
From: "Stephen J. Turnbull" <>
Date: Wed, 24 Nov 1999 13:16:08 +0900 (JST)

My comments are ordered in decreasing order (YMMV ;) of interest to

>>>>> "DJ" == DJ Delorie <> writes:

    DJ> Is there any way to create a value-add for free software
    DJ> that's scalable and self-sustaining?  Otherwise, growth is
    DJ> linear vs the resources available.

Good question.  GCC and the Linux kernel are crucial test cases, I
think.  They _have_ scaled and are self-sustaining.  (I don't see a
reason to question that.)  However, they are well past the stage where
Joe Random Hacker can contribute much, I suspect; can the process
generate superlinear growth when the required personal investment in
compiler-writing or kernel-hacking skills starts to become large?  I
believe so.  (But I don't have any qualifications to judge either
whether J. R. Hacker can contribute or whether the growth can be
superlinear; opinions? wild-ass guesses?)

The reason these cases are so important is that by restricting the
pool of available programming talent to highly skilled developers, we
will more rapidly run into McConnell's Barrier[1]:  you can't easily
recruit more volunteer programmers, but have to finance either more
effort from the current crop or new recruits.  I have some (rather
incoherent) thoughts about how bazaar processes can vault over that
barrier, but I have no suggestions on how to generate successful FSBs
from them yet.

    DJ> I hope that someday someone will post something to fsb about
    DJ> MAKING MONEY and we'll need to stitch all these parts together
    DJ> into coherent business strategies.

The way I justify my participation in this forum to myself is that the
economist's job is to show how to avoid "_not_ making money while
`doing all the running you can just to stay in one place'."  (Ie, the
economist's notion of "perfect competition.")  By exposing the
underlying assumptions, I hope that it will become clearer what kind
of strategies can actually make real money (because they are founded
on real violations of the model, and create value for customers) and
what kind of strategies depend on the assumption that the business can
outrun everybody else on "native talent" (which is probably false,
even if you think you know you're that good :^).

    DJ> Proprietary software makes money for its owner without the
    DJ> owner needing to continually replenish its value.

It also creates value for customers without anybody needing to
replenish that value.  In other words, the original act of creation
continues to create value---far more value than it creates profit,
even under monopoly---as new customers arrive.  And to get repeat
customers, you must create new value; the old value doesn't wear out.
(Even if it becomes obsolete:  customers won't return for more copies
if your product hasn't changed.)

Most business strategies in software development involve targeting
people who you aren't sure exist, and making them pay you for value
you must invest in creating without any commitment of payment from
them.  Proprietary rights in software is "merely" a matter of taking a
longer view, and therefore justifying higher investment than otherwise
would occur.[2][3]

Can similar effects be generated by harnessing the energy generated by
business models to free software development and distribution?  The
free software analogue to "taking the long view" by creating
proprietary rights might be "taking the long view" by creating
reviewer networks with "more eyes" and projects with better
modularization.  (Not that people here don't already know this; but
the analogy is possibly useful in talking to business people and
economists, and maybe for generating more ideas.)

    >> Puhleeze, DJ, take it to gnu.misc.discuss or some other low
    >> signal-to-noise forum.

    DJ> Yeah, yeah, yeah, I know.  I apologize for stirring the mud,
    DJ> but I was just answering a question.

Well, I understand where you're coming from, but if RMS will admit
that software covered by BSD-style licenses is free, I don't think it
serves the discussion to claim otherwise.

I think the distinction I made, that the free software community has
as its _goal_ perpetual freedom of software, but that it uses the same
legal instruments (free software licenses) as the open software
community, is more useful than the one you made.  Your distinction
seems like an attempt to redefine "free software" as "copyleft

I see nothing wrong with arguing the latter is "purer" or "better"
than the former (although I currently disagree), but redefining the
term seems to me to obscure the real issue, which is about communities
and their cohesion, not about licenses.

[1]  cf Steve McConnell's lead editorial in the July/August issue of
IEEE Software, specifically references to an "economic shell game."

[2]  It is true that firms like Microsoft have become successful
primarily through "rent-seeking," that is, wedging themselves into the
market in such a way as to become monopolists.  Thus they earn
apparently excessive returns to investments that are much smaller than 
their profits.  But that doesn't go on forever.

[3]  There are three arguments against that position that I know of.
One is that some kinds of property (eg, software patents) create so
much friction that they destroy more value than they generate by
creating incentives.  The second is that in community there is
strength; by sharing freely we create more value than by trade.  And
the third is RMS's argument from ethics.

All are valid to some extent; the free software community has already
demonstrated that they are sufficient for a "subsistence" level of IT,
and gone far beyond mere "subsistence" by generating self-sustaining
growth, a remarkable achievement.  But there are external (to the free
software community) benefits to open source/free software, which
business support might help us to generate more efficiently.

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."