OLPC, many networks at Winter Camp

Andy Oram
March 4, 2009

Work and play

I mentioned yesterday that I’m at a conference with dual goals: making progress on a particular project and sharing general lessons about how to build human networks. Yesterday I spent six hours on the first goal, which was aimed at one small piece of innovation (an IRC client for FLOSS Manuals).

The scene felt like a Hollywood scriptwriter’s fantasy of life in Amsterdam. Eight of us went to a studio on the west side of the city, rented by an artist who produces designs for FLOSS Manuals. It had the requisite high ceilings, blank walls, obscure art journals and exhibition proceedings. (The movie-set ambiance was disrupted a bit by the presence of two books I had edited, and Beautiful Code.) We gathered around laptops and wrestled with database schemas or CSS for hours.

Today we took a step back to discuss our goals and FLOSS Manuals’ unique book sprint way of creating documentation. We’re holding one in Boston on March 21-22, under the auspices of the Free Software Foundation, and we can use volunteers who would like to teach GUI users how to be effective with the Bash command line.

Swimming in networks

Across the liberal disciplines, researchers are fascinated with how much of biology, nature, and society can be described in terms of networks. Networks are hard to control, impossible to predict, and fantastically productive. But as Winter Camp organizer Geert Lovink pointed out in a speech I read yesterday, The Principle of Notworking (sic), networks are more likely to slow down initiatives than to speed them up. Central authorities are much more efficient. They just don’t achieve as good results.

One of Lovink’s revelations was his interpretation of why the dot-com boom of the 1990s crashed. What he called Commerce, Inc. was trying to coerce the Internet into a centralized distribution model. We all now understand the fundamental dynamic of the Internet as a network. This means it consists of autonomous people who come and go as they please and generate new forms of interaction as they go along.

I’d like to take this idea a step further and predict that the social-networking follow-up to Commerce, Inc. is likely to come to a bad end too. When there is no central control point, it’s hard to extract a profit from activity. If there is no central locus for production, there’s no ethical justification for extracting a profit.

Understanding network theory a bit better helped me answer the question I had when I arrived: why are there so many of them? It seemed like many of the groups of artistic and political activists at Winter Camp announce overlapping goals, and I wondered whether any networks should merge or outsource some of their goals to more specialized organizations. But one can’t question why people come together, and why they choose one network over another.

Even though smaller networks must continually educate themselves and repeat all the mistakes that are familiar to other network organizers, they end up stronger and able to build new networks. What we are doing at Winter Camp is the same process I’m going through in so many other organizations: lifting the perspectives of the participants to see beyond their own goals to the goals of their networks, and beyond the goals of their networks to the ambitious social goals we tend to share here.

One of the key aspects of networks we grapple with in the world of activists is how easily nodes (people) slip away. Just a few days ago I put up an analysis of citizen participation labeling this lack of commitment as a problem. But Lovink treats it as a strength (hence the pun “Notworking”). Each network has to constantly justify its existence to its members. If it stagnates or develops oppressive forms of interaction, members will be quick to find a more congenial alternative.

The conference seems to be working, rather than notworking. Members of different networks mingle at seven in the morning when the breakfast room opens and at ten at night when the evening entertainment ends. Between sessions they gather outside for cigarettes and more conversation. (A lot more people, Americans as well as Europeans, are smoking here than at other conferences I attend. I figure it’s because they’re artists; they don’t think logically.)

The impact of OLPC and Sugar

One Laptop Per Child has been getting some bad press, not all of it from the usual authorities who fear its potential to raise a global generation of free-thinking, capable children. The organization has definitely run into problems with costs, income, and therefore funding, as well as carefully considered criticism from a lot of people in technical communities. But I think a lot of the controversy comes from the original vision splitting up and becoming more diffuse.

Instead of a simple, clear story—a single, unified system—the project has split the software component from the hardware component. This seems eminently sensible. The software component (Sugar) has broadened its base and been ported to many platforms, while the hardware component, under pressure from client governments, has been adapted to run Windows as well as Sugar.

Furthermore, as highlighted in a Wired Magazine article, the technical innovations of OLPC has forced mainstream computer vendors to reconsider their bloatware and to scale general-purpose computers down to mimic the mobile devices that are fantastically popular around the world. (Many of them achieve these goals by running Linux.) OLPC seems to be taking brickbats for experiencing this competition instead of receiving credit for changing the playing field.

I don’t know whether the universal communication device of the late twenty-first century will look like OLPC’s XO (or even be called XO). But the story isn’t about the hardware in any case. The really interesting things will come be after devices of some sort get into the hands of communities and the grassroots communication and experimentation are well underway.