Subject: "Reasonable Profits"
From: "D. V. Henkel-Wallace" <>
Date: Mon, 31 Aug 1998 14:25:52 -0700

(This is slightly obsolete as it was trapped for a while in a DNS failure).

I must admit to being concerned or surprised by the discussion of capping
profits and such.  It isn't that I think they're off topic (FSB is
definately the right place to discuss business models and it's part of the
reason I read it).  It's more the nature of the discussion.

Let me posit two classes of business model.

The first is the "small time operator" or "tinker" or consultant model.
This model says "generate enough cash to cover one (or X) person's living

The second is the traditional corporate model.  This model says "revenues
should grow non-linearly with the labour expended."

Many people do quite well with the consultant model.  Some of the
consulting businesses are in fact quite big.  But still revenues generally
grow roughly linearly with the number of programmers.  This model is I
think pretty well understood in the free software community.

The corporate model has a different structure, and many fewer adherents in
the free software community.  In fact I think only Cygnus and Red Hat have
any pretentions to executing on this approach.

The problem with the "pay me enough to make me my costs and X rate of
return" is that it's fundamentally the labour theory of value, and there's
more to value than can be encompassed by that bankrupt 18th-century
approach.  Maybe it works for consultants (though even they tend to
be arbitraging their knowledge vs the customers' gaps) but it doesn't work
on large scale risk/investment projects.  I think it also confuses the
gratis and libre properties of the software.  It's really a gratis software
issue, and I think gratis (qua gratis) software is by definition
antithetical to anything but the consultant approach unless the value
recovery lies elsewhere (in which case you wouldn't be an FSB anyway)

The second model is the tough one.  We founded Cygnus always intending
_not_ to implement the "bunch of consultants" model but for the first few
years we were in reality basically nothing but a consulting company (working
only on libre software of course).  And I doubt the labour theory of value can
fit red hat's channel approach.  How could you implement it?

I wonder about the opposite problem.  How does one build sufficient brand
equity with libre (or gratis) software?   That's one way to spike the
exploitation problem -- but it's tough to swing on the consultant model.