Subject: Re: Open letter to those who believe in a right to free software
Date: Thu, 21 Oct 1999 10:59:27 -0400

(Oops, forgot to set CCs the first time.)
> To whom it may concern:
> Regarding my motives:
> Please understand: I intend to spend as much of the rest of my life as
> possible living in the wonderful world of free software that Richard
> Stallman envisioned many years ago in the GNU Manifesto, and that the
> participants on this list have done so much to create and enhance over
> the years.  But I am a professional economist; I am ethically required
> to consider the needs and wants of people who don't think like I do.
But why do you have that ethical requirement?  What does it MEAN
for you to have that ethical requirement?

> I know some of you don't much like the way I think.  But I am going to
> write and try to publish professional economics articles describing
> the social costs and benefits, in conventional economic terms, of
> legally establishing various intellectual property rights.  If you

An example might make you think.  I heard about this list through
Karsten Self, who I know from the InfoWorld forums.  Their forum
software has...issues.  Many forum regulars have been more than
willing to offer assistance (hey, we use the stuff) up to and including
multiple offers to rebuild the whole site.  We (pretty much
incidentally) managed to (without help) analyze how it worked.

Yesterday, through a hole that Karsten noticed completely by
accident, it became possible for me to see the code behind their
software.  A proposed fix to the existing suite of problems took me
under 20 minutes to produce.

These issues cost IWE several hours of employee time per month,
reduce the number of visitors to their site, and have caused a
reasonable amount of frustration for the forum revenues.  Their
revenue is not affected by our having access to the source.  But
they didn't write the software, and so they could not legally
redistribute it.  (Note, however, that they had multiple offers to
completely rebuild it from scratch.)

Would you care to give me a cost-benefit analysis, in the
conventional economic terms, of existing concepts of intellectual
property rights.

> don't do a much better job of explaining why "users' rights" are valid
> and "inventors' rights" are not, odds are good that they will look a
> lot like Shapiro and Varian's _Information Rules_.  Ie, I'll just take
> the current regime as given, and maybe propose some cosmetic changes.
> I suspect that the results are unlikely to justify more.
If you want an economic cost-benefit analysis, I would suggest
that you try Eric Raymond over at

> If you like that, I'll just shut up and go do it.  (I will continue to
> think about the issue, but as Richard said, I just don't seem to grasp
> the idea---it's not likely I'll have an epiphany without help.)
> But I really want to know where those "users' rights" come from, if
> they're not written by the vendor into the license contract.  I'd like
> to believe in them, but I don't.  I can't honestly base policy
> recommendations on them unless I am convinced of their plausibility.
> I can't analyze models based on the assumption they are not violated,
> because nobody will pay attention to analysis presuming them unless
> they are made plausible.
OK, I have been working on the following for several days.  It is
probably not ready for prime-time, and still have to think about a
permanent home for it, but perhaps it will clarify your thinking
about economic value versus rights, how they are connected,
and how they differ.

> Will you help me?
Does the following help?

Connecting Open Source with Free Software

In this essay I seek to compare and contrast the Open Source
movement with the Free Software movement.  In particular I will
identify what views they hold in common, and explain what I see as
the divide between them.  It is not my direct intent to convert
people to either position, but by clarifying the issues I hope to
make it easier for people to develop an informed opinion.

The first question for most people is why there is any difference in
the first place.  After all all aren't both groups talking about the
same thing?  The answer to that is that while both groups are
advocating the same type of software, they do so for different
reasons.  Furthermore each group feels that the rhetoric used by the
other group is detrimental to their own goals.  This - combined with
some strong personalities and a few nasty exchanges in the past -
has created a significant divide over the last two years.

The position of the Open Source movement is that software for which
the source code is freely available, modifiable, and redistributable
is software that can enjoy a very effective software development
model.  This model has tremendous benefits which are simply not
possible in proprietary software.  The success of this type of
software goes against what classical software engineering principles
deem possible, and constitute a very strong value proposition for
this kind of software.  This argument is generally attributed to
Eric Raymond's classic essay, "The Cathedral and the Bazaar".  You
may find it, along with much more information about the Open Source
movement, at

However, a decade before Eric tried his first free operating system,
Richard Stallman set out to build so that people need not rely on
proprietary software.  He named this project GNU (for GNU's Not
Unix), and this event traditionally marks the beginning of the Free
Software movement.

In a nutshell, Richard's argument for free software is that it is
ethically wrong to make software proprietary.  Free access to
source code is (or should be) your right.  Denying that right is
morally wrong, and anyone who  does so commits an injustice.  His
reasoning for this position may be found in the philosophical
section of the GNU website, freely (of course) available at

This is strong language, and many people would like to pass Richard
Stallman off as just another zealot.  But while he may be a zealot,
he is far from being, "just another zealot."  He is a very good
programmer, and his actions have had a tremendous impact.  His
sincerity is evident in how he proceeded.  His first action was to
quit his job in 1983 so that he could create a completely free
system, unencumbered by any limitations on the freedom he values.
A partial list of what he created includes:

  - the Emacs editor
  - the GCC compiler
  - the layout of an entire operating system
  - the Free Software Foundation (to assist him in this work)
  - too many lesser utilities to enumerate
  - many influential essays, starting with "The GNU Manifesto"
  - legal licenses (GPL and LGPL) to defend this work

By 1991 this new operating system was almost complete.  All that it
lacked was the kernel that mediates between different programs, the
physical hardware, and the signals coming from and going to the
outside world.  But unfortunately Richard was suffering from severe
tendonitis, and his supporters were bogged down in implementing the
chosen design.  In the meantime a college student in Finland, the
now-legendary Linus Torvalds, wrote a far simpler kernel as a way of
learning the advanced memory management features of Intel's 386 chip.
Free software supporters welded the GNU system on top, and a totally
free operating system was born.  This system is usually called
Linux, however many believe that GNU/Linux is a more appropriate
name in recognition of the contribution of the GNU project.

The obvious divide between the groups is their motivation.  Both
want programs whose source code can be freely downloaded, modified,
and redistributed.  But they want them for apparently unrelated
reasons.  The Open Source people want good software, and believe
that the best long-term guarantee of this is freely available,
modifiable, and redistributable source code.  The Free Software
movement is not as immediately concerned with the quality of software
as they are with stopping the injustice of being denied access to
source code.  This difference in motivation lies behind most of the
arguments between the camps.  For instance Open Source advocates
complain that Free Software's "ideological tub-thumping" scares
people off.  Conversely Free Software supporters criticize the way
that the Open Source movement, "...shuns the ideas of freedom,
community, and principle."  The disconnect is clear.

Yet the two positions are fundamentally connected.  To understand
that connection, you must understand what it means to have a right.
A freedom is a right if that freedom has (or should have) value to
you, and if that freedom is furthermore something that you cannot
validly be denied.  Think carefully about this definition.  If there
is a valid reason for denying you a freedom, then no injustice is
done by such a denial.  Conversely if you are denied something that
is worthless, then no harm is done and therefore the result cannot be
more than an annoyance.  However if you are denied a freedom of value
to you, and there is no valid reason to deny you that freedom, then
you have been wronged.  And in this case, denying that freedom
constitutes an injustice, and therefore you must have had a right to
that freedom.

With this understood, the connection between the two groups is that
both groups are recognizing the substantial value arising from the
same freedoms.  The freedoms in question, as explained earlier,
are the freedoms of having source code available, modifiable, and
redistributable.  The Free Software position is that there is no
justification to deny these freedoms.  Open Source advocates may or
may not accept that companies have valid reasons for denying access,
but hold that informed consumers should recognize what they lose with
this denial, and not fall for the trap.  So Free Software developers
work for a world in which consumers can realize the benefits of Open
Source development models while conversely consumers who accept and
understand the Open Source viewpoint are resistant to the lure from
proprietary software.

With this connection understood, key insights from each movement can
be stated in the language of the other.  Richard Stallman has been
stating for years the tremendous value in letting consumers have full
access to the source.  Eric Raymond's insight is that there are
effective development models through which the collective effect of
individuals exercising their individual rights to source code will
produce tremendous economic value.  Both sides point to the same
factors as being important such as: fix bugs for yourself; choose who
to purchase support from; not be dependent on a single source for
upgrades; learn how to program from the best possible source code;
not re-invent wheels already invented by others.  As you continue
through the arguments that each side gives, you find identical value
propositions echoed.

Can we judge the worth of this value?  Not really, but a gross order
of economic magnitude can be found by comparing Red Hat to Caldera.
Caldera was founded as a spin-off from Novell.  Caldera started with
millions of dollars of investment and was given technologies that had
been purchased for millions more.  By contrast, Red Hat started with
one programmer and one salesman, and was funded using personal
credit-cards.  (No bank was willing to loan a start-up money for the
"insane" idea of selling free software.)  However Red Hat also
recognized from the start a basic principle that Caldera did not at
first appreciate.  That principle is that free access to source code
has real value to consumers, and there are large economic
opportunities out there for a company willing to provide that value
when others do not.  Furthermore Red Hat recognized from the start
that the GNU Public License, invented by Stallman to provide
comprehensive protection of the public's rights to source code, was
also far and away the best protection available for the value they
were trying to deliver.  This is not a coincidence given that they
are trying to deliver the freedoms that Stallman calls your rights.
Therefore Red Hat guarantees that the software it writes will be
available under the GPL.

So on the one hand we have millions of dollars in capital and
products, on the other we have two people funded by credit cards and
selling the same basic system.  However the second company has the
advantage that it sells a guarantee of freedom, while the first is
selling a proprietary product.  Effectively there was no contest.
While Caldera has survived and done reasonably well, Red Hat has done
far better.  In recognition of this Caldera today, along with many
others is now borrowing techniques from Red Hat.

It is easy to be cynical about Red Hat's motives.  This cynicism is
justified.  Red Hat is not in business because of a sense of what is
right or wrong.  It is in business to make money.  Its management,
from Bob Young on down, believes that the surest way to make money
is to provide a better value.  And the better value that they try to
provide is in the form of freedoms to source code.

It is easy for late-comers to Linux to underestimate the extent of
this commitment.  This mistake is unlikely to be made by anyone who
watched and understood the entire KDE affair.  KDE is a free project
that provides Linux with a top-notch desktop environment.  But, while
KDE may be free, it was built on a non-free library named QT.
Because that made an important component of the system non-free, Red
Hat could not provide its guarantee if it shipped KDE.  Yet KDE was
very popular, and a large portion of Red Hat's users did not
appreciate the licensing issue.  Red Hat's commercial commercial
competitors wasted little time in taking advantage of this by
shipping KDE.  Many criticized Red Hat for being obstinate, and a
significant portion of its customers took their business elsewhere.
As if this were not enough, Red Hat believed that this challenge
could make it go out of business, and so it sank a significant
portion of its revenue into the immature a fully-free alternative
(Gnome).  And remember, this investment was made with full knowledge
that if Gnome succeeded then their competitors would simply use it.
This was widely called a colossal blunder.

But was this a blunder?

Today Gnome is a usable product, the developers of QT have changed
the license of their next version to be (barely) open source, and
Red Hat is more influential than ever.  In fact the real value to of
free source code such that this poorly funded company, while making
traditionally unthinkable decisions, managed to so dominate its
market that there has been public concern over its power and control!
Now I call that value!

Another lesson from this example is that some key Free Software goals
are furthered by the Open Source philosophy.  Examine Red Hat's
actions again.  Even though they are driven by a theory about how to
make a profit, the actions themselves are ones that any Free Software
supporter should like.  Certainly a Free Software supporter would
prefer having Red Hat entirely accepted the Free Software position.
But regardless, the behaviour is desirable and should be encouraged.
(Similarly a key benefit that is cited for the GPL is that it creates
reasons to release more free software.)

Clearly the Open Source approach is a more pragmatic way to
encourage the release of free software.  However there is hope that
the philosophical views of the Free Software movement may eventually
dominate because of the groundwork that Open Source is laying down.
There is a lesson from history that will help in explaining why there
is this hope.

Frequently in studying history we are confronted with difficult
truths.  One of those truths is the fact that what one group thinks
of as simple moral principles, tend not to be recognized as such by
others when it is expedient to deny what others might think obvious.
In particular perceptions about economic reality frequently
circumscribe our thinking about ethical principles.

While many examples exist, a classic extreme is the case of slavery.
Slavery is the institution of making one human being the property
of another.  Clearly there is great value in not being owned by
another!  Today we further recognize as a basic truth that there is
no valid reason that one person should so own any other, and so we
recognize a basic human right to not be owned by another.  Yet this
"peculiar institution" was generally accepted for millennia in
virtually all cultures for the simple reason that there was no other
[known] way for the politically influential portions of the
population (who not coincidentally were most likely to have others
as "property") to enjoy their creature comforts.  It was, simply
put, extremely expedient to recognize slavery as "the way of the
world" and leave it at that.

However, in the 1800's the advancement of technology brought some of
these comforts within the reach of many without having to resort to
slavery.  In many case it actually became cheaper to obtain these
comforts without the expense of a slave who needed to be fed and
clothed.  Therefore slavery entered a general decline.
(Unfortunately this did not happen everywhere.  The best known
exception was the economic dynamics created by the cotton gin in the
southern US.)  With this general decline it became less expedient
for people to accept slavery.  As a result people ceased to accept
once widely accepted justifications for it, such as the theory that
there were more powerful races and weaker ones, and it is inevitable
that the more powerful ones will command the services of the weaker
ones.  (What was lost was not the general acceptance of distinctions
between the races, but rather that it was inevitable that the more
powerful races would want or need to command the services of the
lesser ones.)  Once the justifications that had been offered came to
be regarded as invalid, the barriers to recognizing freedom as a
basic human right were lowered, and the universal assertion of that
right was assured.

In a gentle echo of that issue, the single most common justification
offered today for proprietary software is that it is necessary.  It
is believed necessary because, without it, people cannot make money
from software, and if people cannot make money then programmers will
not be paid to program, and programs that people want will not be
written.  Therefore, no matter what the value of free software once
produced, it is expedient to allow proprietary software to exist.
This justification for proprietary software is widely accepted as
valid.  People who accept as valid this (or any other) argument for
proprietary software have rejected the view that access to source
code is an ethical right.

However, an opportunity is arising to overthrow this view.  The
belief of Open Source advocates, based on examples like Red Hat
(which is neither the only nor the first company to do well based
on the value of free software), believe that it is possible to have
generally available free software providing consumers with the
programs that they want, have programmers gainfully employed, and
additionally have profitable companies based on this free software.
This is not just an abstract theory.  More and more companies are
today attempting to make a profit using the proposed models.  So far
they appear to be succeeding.  And I maintain that if they fully
succeed, then the single argument that most dissuades people from
accepting the Free Software position will lose its sting.

Therefore I believe that, whether or not the participants in the
Open Software movement explicitly agree with the Free Software
ideology, the success of the Open Software movement is creating a
potential opportunity for a general acceptance of the Free Software
belief system.  At this point there is still legitimate disagreement
as to whether this can or will succeed, and many who think that free
software models can succeed in some areas have doubts as to how
generally applicable the success will be.  Over the next several
years we are likely to see unambiguous answers to these questions.

In summary, I have attempted to show the following:

  - Both the Open Source and Free Software movements identify the
    same value arising from the same software freedoms
  - The values identified motivate both towards similar types of
  - Both movements agree that this value is very large
  - Preliminary economic evidence suggests they are correct
  - The Free Software movement further asserts that there is no valid
    reason for limiting those freedoms
  - This second assertion properly transforms the discussion from an
    issue of economic value to an issue of ethical and moral
  - It is easier to convince people of the Open Source theories
    because you avoid the need for this second assertion
  - A common reason for opposition to acceptance of this assertion is
    based on implicit economic preconceptions
  - Commercial attempts to apply Open Source economic models is
    providing a real-world test of this economic preconceptions
  - Depending upon the outcome of these tests, the Free Software
    viewpoint could become much more widely accepted in the future

It is my sincere hope that the above has clarified your understanding
of both the Open Source and the Free Software movements.  It has not
been my intent to advocate either position, or even make it clear
what my own position is.  Rather I have tried to make it clear what
they hold in common, and what, beyond mere rhetoric, specifically
divides them.

Ben Tilly