Nettime mailing list archives

<nettime> Six Limitations to the Current Open Source Development Methodo
Felix Stalder on Thu, 14 Aug 2003 18:52:27 +0200 (CEST)

[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]

<nettime> Six Limitations to the Current Open Source Development Methodology

Six Limitations to the Current Open Source Development Methodology

The "Open Source Approach" to develop informational goods has been
spectacularly successful, particularly in the area for which it was
developed, software. Also beyond software, there are important, successfull
Open Source projects such as the free Encyclopedia, Wikipedia; collaborative
sites writing/publishing projects such as koro5hin.org; and the Distributed
Proofreading Project, attached to the Gutenberg Project.

However, particularly outside the software domain, the Open Source projects
remain relatively marginal. Why? Some of it can be explained by the relative
newness of the approach. It takes time for new ideas to take hold and to be
transferred successfully from one context to another. But this is only part
of the story. The other part is that the current development model is based
on a number of specific, yet unacknowledged conditions that limit its
applicability to more diverse contexts, say the music distribution or drug

The boundaries to the open production model as it has been established in the
last decade are set by six conditions characterizing virtually all of the
success stories of what Benkler called "commons-based peer production." The
following list is a conceptual abstraction, a kind of ideal-type. The actual
configuration and relative importance of each condition varies from project
to project, but taken together they indicate the boundaries of the current
model. In this elaboration, I draw from examples of free and open source
software, but it would be simple to illustrate these limitation based on open
content projects.

1) Producers are not sellers

The majority professional, i.e. highly-skilled, programmers do not draw their
economic livelihood from directly selling the code they write. Many work for
organizations that use software but do not sell it, for example as system
administrators. For them the efficient solution of particular problems is of
interest, and if that solution can be found and maintained by collaborating
with others, the sharing of code is not an issue. For others employed in
private sector companies, for example at IBM, the development of free
software is the basis for selling services based on that code. The fact that
some people can use that code without purchasing the services is more than
off-set by being able to base the service on the collective creativity of the
developer community at large. From IBM's point of view, the costs of
participating in open software development can be regarded as 'capital
investment' necessary for the selling of the resulting product: services.

For members of academia (faculty and students) writing code, but not selling
(often explicitly prohibited), contributes to their professional goals, be it
as part of their education, be it as part of their professional
reputation-building. For them, sharing of code is not only part of their
professional advancement, but an integral part of the professional culture
that sustains them also economically,. in form of salaries for the faculty
and stipends for the (graduate) students.

Last but not least are all those who use their professional skills outside
 the professional setting, for example at home on evenings and weekends.
 Having already secured their financial stability, they can now pursue other
 interests using the same skill set.

2) Limited capital investment

Particularly the last, and very important group of people, whose who work
outside the institutional framework on projects based on their own
idiosyncratic interests, can only exist due to the fact that the means of
production are extraordinarily inexpensive and accessible. Materially, all
that is needed is a standard computer (often even a substandard one would
already suffice) and a fast, reliable connection to the communication forums
of the community. Of course, the computer and the network rely on a level of
infrastructure that cannot be taken for granted in large parts of the world,
but for most people in the centers of development, they are within relatively
easy reach.

Once this access to be means of communication is secured, the skills
 necessary to participate in the development of code can also be acquired
collaboratively, free of charge. The number of self-taught programmers is
significant. Since no expensive diplomas are necessary to become active, the
financial hurdle is, indeed, extraordinarily low.

3) High number of potential contributors

Programming knowledge is becoming relatively common knowledge, no longer
restricted to an engineering elite, but widely distributed throughout
society. Of course, truly great programmers are rare as truly great artists
are, but average professional knowledge is widely available. This has a
quantitative and a qualitative dimensions. Quantitatively, the number of able
programmers is in the millions, and rising. Qualitatively, the range of
people capable programmers is also unusually wide, not the least because the
material hurdles are so low and the learning can take place outside of
institutions with entry exams and tuition fees. This large and diversified
pool of talents makes it possible to create the critical mass of contributors
out of only a fraction of population.

4) Modularized Production

A large software program consists of many smaller code segments (libraries,
plug-ins etc.)This makes it possible to break down the production process
into many small steps which can be carried out by distributed contributors.
If the act of integration is relatively straight forward, it allows to keep
the amount of work that each has to contribute highly flexible and also make
use of smaller contributions (bug reports, patches). Furthermore, the
modularity of the production process allows a high number of people to work
in  parallel without creating significant interferences.

5) Producers Are Users

According to Eric S. Raymond, a good open source projects starts with a
programmer scratching his own itch and finding out in the process that there
are many others with the same problem. Wanting to use a program is a great
motivation of contributing to developing it. Often, it's much more efficient
that waiting, hoping that someone will write and sell a program that will
address one's particular need.

6) No Liability

Last, but not least, software has no product liability. Paragraph 11 of the
GPL states, similar to most other licenses, that "the copyright holders
and/or other parties provide the program 'as is' without warranty of any
kind, either expressed or implied, including, but not limited to, the implied
warranties of merchantability and fitness for a particular purpose" (GPL,
v2). The absence of liability makes it possible to produce a program without
having to assign clear ownership, or other markers allowing to determine
liability, to it.

The space delimited by these condition is large and still not fully explored.
We can expect that the current open production model will find additional
niches in which it can thrive. Few could have predicted the success of
Wikipedia only three years ago, even though Open Source Software had already
been very successful at the time. However, it is also clear that many
information goods fall outside of this space. Not always are the means of
production inexpensive and readily available or the production process
modular. Sometimes,  the number of potential producers is small, more often
than not are the producers not the users of their own products, and, in many
cases, product liability is desirable.

This does not mean that the "open source model" cannot apply to, say, the
production of literary works, music, or medical drugs. What it means,
however, is that to make it viable, another round of social innovation is
required. This is slowly happening. The growth of "Open Access Journals" or
discussions around "compulsory licensing" are good, though very early




#  distributed via <nettime>: no commercial use without permission
#  <nettime> is a moderated mailing list for net criticism,
#  collaborative text filtering and cultural politics of the nets
#  more info: majordomo {AT} bbs.thing.net and "info nettime-l" in the msg body
#  archive: http://www.nettime.org contact: nettime {AT} bbs.thing.net