Subject: Re: Open letter to those who believe in a right to free software
Date: Mon, 25 Oct 1999 09:58:05 -0400

> >>>>> "rms" == Richard Stallman <> writes:
>     rms> But even more important is that proprietary software is based
>     rms> on telling people, "It is wrong to share copies with your
>     rms> neighbor."  That undermines the good will that makes the
>     rms> difference between a dog-eat-dog jungle and a society.
> Sharing copies of programs with your neighbors is a way of decreasing
> the share of development costs borne by people you know at the expense
> of increasing the cost share of people you don't know.[1]
In today's world the people you know includes everyone with access
to ftp who reads FreshMeat.  If your software is any good it includes
many friends of friends, such as customers of Red Hat.

That is close enough to "everyone who uses free software" for
government work.

Oh, you were talking about proprietary software?  Well the rest of us
are not.

> I don't know if that's wrong, but I am not comfortable with it.  I can
> say, somewhat inconsistently, that it's not neighborly in the Internet
> society where everywhere is here and everyone is my neighbor, even if
> I don't know them.  But the real problem is that it's just plain
> arbitrary.  Why should one's list of friends (or whether there are any
> at all) determine the price one pays for software?
Because we don't know a better way to distribute the costs.  I am a
friendly fellow.  I have taught a large number of people how to use
software better.  I only teach people that I know who show some
interest in learning.  Is that unfair?  What would you suggest that I
do differently

> True, on the Internet, you can distribute it to anonymous neighbors;
> then the phrasing is a little different.  Sharing programs is way of
> decreasing the share borne by the late adopters at the expense of the
> early adopters; or a way of decreasing the share borne by those who
> don't need handholding at the expense of those who do; or a way of
> decreasing the share borne by those for whom the stock version is
> satisfactory at the expense of those who need customization.
You are implicitly assuming that the product is developed at full
cost and then distributed.  This is rarely true.  Typically the first
release has relatively few features, and the bulk of the cost is
borne in later development and maintainance.  Therefore it is
really more of a case of distributing the cost of development
and testing more evenly through the early adopters than the late.
Additionally free software allows the distribution of those costs
through a much wider audience than would otherwise be the

There is some truth to the saying that the software will cost more
for those who require handholding.  But that is always true.  It
costs money to have your hand held.  In fact one way to make a
profit making free software is to rely on its being free resulting in
a wider distribution, and then recoup your costs off of people
who need handholding and are willing to pay a premium to get
it from the top experts on that software.

> The original phrasing (acquaintance vs. stranger) is best, though,
> because it is not contaminated by the accident that the higher price
> paid is spuriously correlated with more services received.  (To see
> that it is spurious, simply suppose that it happens that the value of
> the software itself is inversely proportional to the value of the
> ancillary services.)
> Footnotes:
> [1]  In a hypothetical competitive software market.  In the real
> world, it may decrease costs to all users if the vendor is forced by
> competition from unlicensed copies to lower price in order to get
> anybody to pay.  It can go the other way, though, if the paying
> customers have inelastic demand.

If you think that the software world tends to look like a
competitive  market then you have some serious misconceptions
about software!  The proprietary software market is based on a
series of small monopolies, and each product that you use results
in some degree of vendor lock-in.  This may be observed from the
fact that the marginal cost of producing more copies of software
is not reflected in the price.  Rather it reflects the development
costs that would be borne by any potential competitors.  (See
Microsoft's profit margins for evidence...)