Those of us in the free/libre and open source software (FLOSS)
community know the routine by now. Despite the fact that "free software"
and "open source" refer to the same software and the same communities,
supporters of "free software" like the FSF
would have us advocate for FLOSS by talking about users' rights to use,
modify, share, and cooperate; open source supporters like the Open Source
Initiative would have us advocate for software by talking about how
securing these rights produces software with "better quality, higher
reliability, more flexibility [and] lower cost."
One reason I tend to stay away from "open source" claims in my
own advocacy is that I'm worried by the way that these arguments rely on
a set of often dubious empirical claims of superiority. Free software,
on the other hand, can be seen as statement of principles. Regardless of
whether we say "free software" or "open source," I've found that a focus
on principled statements is both more robust against counter-arguments
and does a better job of describing the motivations of most contributors.
Principles can be thought of like opinions. They may or not be
compelling but are neither right or wrong outside of a particular
ethical framework. Most people won't demand evidence for someone's
commitment to nonviolence or an adherence to the Golden Rule. What would you need to prove?
Principles are based on a type of Utopianism; they are a statement of
how we think things should be.
On the other hand, open source's argument that openness leads to
better software or a better software development methodology can be
measured, tested, and declared right or wrong. A FLOSS program might be
better or more reliable than proprietary software. Or it might be
worse. The open source methodology might be lower cost for a consumer
or more profitable for a producer. Or it might not. There are plenty of
FLOSS success stories. There are many more failures.
The problem for open source advocates is that while FLOSS is
often better than proprietary software, this is not always the case. I
was using FLOSS in the early 1990s when GNU/Linux was indisputably less
featureful and buggier than its proprietary competitors. On the business
side, we learned in the Dot Com boom and bust that, despite Eric
Raymond's assurances, building a successful FLOSS project
turned out to be harder than a COPYING file and a tarball on a
webserver: Netscape is essentially gone; VA --- the single largest Dot Com IPO --- is a shadow
of its former self; LinuxCare became a proprietary
If, as open source advocates would argue, the reason we're here
is to build software more efficiently or at greater profit, we must also
advocate for proprietary development methodologies in areas where
evidence seems to show that they are more effective. Where are these
advocates? Where are the open source advocates applauding LinuxCare for
saving themselves by abandoning FLOSS. Don Marti has observed
that this doesn't seem to be what is going on:
- Do people really spend their weekends helping annoying new
people install free software because it has a more efficient development
methodology? Of course not. If it were only about efficiency, hobbyists
would volunteer to replace the old ballasts in companies' fluorescent
Of course, Marti is right. The reason that hundreds of thousands have
spent their time assisting FLOSS efforts has less to do with a passion
for efficiency and more to do with a set of implicit principles.
Humans are driven to imagine worlds that they would want to live
in. For a growing group of people, that's a world where software can be
used, shared, and collaborated without restrictions or discrimination.
We may think of this in ethical terms, in terms of an attitude toward
innovation, or as a set of political or economic positions. But we
should realize that these are, ultimately, principled stands.
And if we are taking principled positions, it is in the long-term
interests of both our cause and our credibility to frame our arguments
and our advocacy in those terms. We can use empirical evidence to help
bolster our arguments but we should be careful to not confuse these
empirical claims with the principles themselves. They can, and sometimes
will, be proven wrong.
By honestly highlighting our principles and not shying away from
explicit Utopianism, we can return to questions of efficiency as
means toward achieving our principled ends. Approached from this
angle, we need not seek to explain why FLOSS is better than proprietary
software --- which it may or may not be at any given point in time and
for any given project --- and can instead ask how we can make it better.
Humans are creative, innovative problem solvers. We set goals and
devise social structures and technologies to achieve them. The fact that we
have created socio-technical means of creating better software through
free ways in so many areas is a reflection of this ingenuity applied
toward principles at the heart of FLOSS. We would be well served to
remember that this is how FLOSS will win, not why.
The thing that struck me when I first read the GNU Manifesto many many
years ago is that:
- Yep, I buy it. Proprietary software does suck.
- It doesn't matter what I think. It won't take many people to buy
into this to create a serious challenge to proprietary software.
Business people like to factor out philosophical calculations, which is
why, on the surface, the OSS-style arguments are appealing. However,
the problem is that you can't factor out the philosophical
considerations of others in a fully-informed business decision. For
example, if one wants to open a kosher deli, one would be well-advised
not to serve bbq pork sandwiches, no matter how tasty they might be.
We do have to understand that people will be motivated by many of the
OSS-style arguments though as well. While some of the arguments of
efficiency made in the heady early days turned out to be overblown, it's
getting tougher and tougher to argue that all of the FLOSS success
stories are flukes. There has to be something to the methodology.
That said, purely as a mechanical process, OSS vs proprietary may
be a wash. The performance difference seems to be the people-powered
edge that the free software philosophy provides.