[ox-en] keimform.de: ox4 Notes II: Open Hardware Challenges and Ambitions
- From: Christian Siefkes <christian siefkes.net>
- Date: Wed, 15 Apr 2009 20:24:15 +0200
This post continues my coverage of the Fourth Oekonux Conference
*Johan Söderberg* talked about the Czech open hardware project Ronja
<http://fourth.oekonux-conference.org/program/events/22.en.html>. RONJA was
developed to provide a cheap, easily producible alternative to Wi-Fi,
allowing wireless data transmission between computers. Amazingly, Ronjas
use visible light for data transmission, but they are quite fast (10
MBit/s) and allow reliable point-to-point data transmissions, except in
case of fog.
The goal of the Ronja project <http://ronja.twibright.com/> was not only to
build affordable open hardware for data transmission, but also to allow the
creation of anonymous, censorship-prone networks that can't be controlled
by companies or the state. All design information has been published under
the GNU Free Documentation License. The Ronja hardware was sufficiently
successful to be employed not only by private people, but also by
But the success also led to tensions in the community which ultimately lead
to the decline of the project. Somebody created a commercial fork of Ronja,
the *Crusader.* The Crusader was mass-produced and intended for corporate
users; it used a laser instead of a normal light source, which made it
faster (100 MBit/s instead of 10 MBit/s) but also more difficult and
dangerous to built (you can become blind if you operate a laser the wrong
way). The rest of the community didn't want to switch to this design, in
order to keep the design simple and safe for normal users/builders. Their
decision was understandable, but it also meant that the regular Ronja
hardware ceased to be competitive to normal Wi-Fi, which became faster and
cheaper over time.
Another problem was that the copyleft principle doesn't really work for
hardware (as already mentioned in regard to Jacco Lammers in the first part
of my report
Copyright (and thus copyleft) only governs information, not the hardware
built according to this information. So commercial forks such as the
Crusader weren't forced to contribute their improved designs back to the
community. This annoyed the original Ronja maintainer who was unhappy about
companies making money on his project without giving back. In order to make
money himself, he decided to switch to a "I will free the design after I've
got enough donations" model for new improvements. He did get donations, but
this switch meant that he had to work on new developments in secret, giving
up the open, community-based style of development. This in turn annoyed the
community members who had enjoyed giving their time and ideas, instead of
their money, for Ronja development.
Because of these developments, the project has now largely become inactive.
There are interesting lessons here, I think, about how commercial and
noncommercial interests can clash, and about how the lack of suitable
licensing model for open hardware projects can harm them. I wouldn't say
that profit-driven and noncommercial participants cannot form successful
alliances (free software has shown that they can), but cases such as Ronja
make it clear that they don't get together easily. Regarding the licensing
issue, the best scenario would probably be if somebody came up with a
"clever hack" that makes the copyleft/share-alike principle work for
hardware (just like the GNU GPL was the "clever hack" that invented this
principle for software). Another option would be to give up the
expectations of "everybody should give back their improvements" and switch
to liberal BSD-like licenses (which don't contain a copyleft clause).
After this sobering talk about the decline of an open hardware project,
*Marcin Jakubowski* gave a very enthusiastic talk about how he and his
group are "Building the World's First, Replicable, Open Source Global
The goal of the Open Source Ecology
<http://openfarmtech.org/index.php?title=Main_Page> (OSE) project is to
design a "global village" that allows small groups of people to live
entirely independent from capitalism, producing their own energy, food,
housing, transportation and so on. The village shall be "open source" (open
design), and sufficiently documented so that others can reproduce it. As is
reasonable, they focus on first designing and producing the means of
production that shall later allow them (and everybody who decides to
reproduce their designs) to produce the goods they need for living.
Their Global Village Construction Set
(GVCS) contains 41 pieces which they are working on, or plan to work on,
which in their view are sufficient to produce most things needed for a
simple, self-sustainable life, except for microcircuits (computer chips)
and and metal (scrap metal can be reused). The GVCS is designed for
modularity and for disassembly/easy repair.
Marcin said that the GVCS equipment for 30 people might cost about 6,000
USD (for materials) and require somewhat less than one year per person for
building; additionally, land to live on is needed. If people would buy
everything from conventional sources, the total cost would be about 830,000
USD and 2 person years (instead of 30) would be required for assembly.
Stefan Merten pointed out that, if you take labor costs into account, the
Set is not necessarily more efficient--indeed, assuming an average net
salary of 40,000 USD, 30 people would earn 1,2 million USD in a year, which
would be more than enough to buy everything from conventional sources (and
hire some specialists for assembly).
Efficiency is indeed an important point and I'm not sure that the somewhat
low-tech approach of Marcin's group will be able to compete with all the
sophisticated technology capitalism has invented. Peer production will
hardly be able to outdo capitalism if it would mean that people would have
to work *more* instead of less.
Also, the somewhat survivalist, very basic lifestyle of the OSE people
probably won't be attractive for lots of people. Again, I believe that peer
production can only win if it offers people a *better* life than
capitalism, and I don't yet see that in the OSE approach. They're thinking
too small, in my opinion--capitalist societies are extremely complex, with a
very high division of labor, and I doubt that post-capitalist societies
will be able to reach a similar (or higher) standard of living without
maintaining similar levels of complexity and division of labor. Peer
production *is* suitable for very complex societies, since it is based on
the self-selection of people to do the things they like most; and since
people's preferences vary widely, a sufficiently large group will cover a
wide range of activities. But groups of just 30 or so people won't be able
to do it, they are far too small.
Still, Marcin's project is exciting and I expect that quite a number of
goods things will come out of it. They're tackling the issue of how to
produce the means of production, which is indeed the most important
challenge, and even though it'll probably take much more time and more
people than they expect to make these designs really useful and efficient,
it's a good start.
Videos and slides of Marcin's presentation
<http://openfarmtech.org/weblog/?p=611> are available at the OSE blog.
|-------- Dr. Christian Siefkes --------- christian siefkes.net ---------
| Homepage: http://www.siefkes.net/ | Blog: http://www.keimform.de/
| Better Bayesian Analysis: | Peer Production Everywhere:
| http://bart-project.com/ | http://peerconomy.org/wiki/
|------------------------------------------ OpenPGP Key ID: 0x346452D8 --
Everyone by now presumably knows about the danger of premature optimization.
I think we should be just as worried about premature design -- deciding too
early what a program should do.
-- Paul Graham, Hackers and Painters