IF YOU'VE BEEN to a computer show in recent months you might have seen
it: a shiny silver drinks can with a ring-pull logo and the words "opencola"
on the side. Inside is a fizzy drink that tastes very much like Coca-Cola.
Or is it Pepsi?
There's something else written on the can, though, which sets the drink
apart. It says "check out the source at opencola.com". Go to that Web
address and you'll see something that's not available on Coca-Cola's
website, or Pepsi's--the recipe for cola. For the first time ever, you
can make the real thing in your own home.
OpenCola is the
world's first "open source" consumer product. By calling it open source,
its manufacturer is saying that instructions for making it are freely
available. Anybody can make the drink, and anyone can modify and improve
on the recipe as long as they, too, release their recipe into the public
domain. As a way of doing business it's rather unusual--the Coca-Cola
Company doesn't make a habit of giving away precious commercial secrets.
But that's the point.
OpenCola is the most prominent sign yet that a long-running battle
between rival philosophies in software development has spilt over into
the rest of the world. What started as a technical debate over the best
way to debug computer programs is developing into a political battle
over the ownership of knowledge and how it is used, between those who
put their faith in the free circulation of ideas and those who prefer
to designate them "intellectual property". No one knows what the outcome
will be. But in a world of growing opposition to corporate power, restrictive
intellectual property rights and globalisation, open source is emerging
as a possible alternative, a potentially potent means of fighting back.
And you're helping to test its value right now.
The open source movement originated in 1984 when computer scientist
Richard Stallman quit his job at MIT and set up the Free Software Foundation.
His aim was to create high-quality software that was freely available
to everybody. Stallman's beef was with commercial companies that smother
their software with patents and copyrights and keep the source code--the
original program, written in a computer language such as C++--a closely
guarded secret. Stallman saw this as damaging. It generated poor-quality,
bug-ridden software. And worse, it choked off the free flow of ideas.
Stallman fretted that if computer scientists could no longer learn from
one another's code, the art of programming would stagnate (New Scientist,
12 December 1998, p 42).
Stallman's move resonated round the computer science community and
now there are thousands of similar projects. The star of the movement
is Linux, an operating system created by Finnish student Linus Torvalds
in the early 1990s and installed on around 18 million computers worldwide.
What sets open source software apart from commercial software is the
fact that it's free, in both the political and the economic sense. If
you want to use a commercial product such as Windows XP or Mac OS X
you have to pay a fee and agree to abide by a licence that stops you
from modifying or sharing the software. But if you want to run Linux
or another open source package, you can do so without paying a penny--although
several companies will sell you the software bundled with support services.
You can also modify the software in any way you choose, copy it and
share it without restrictions. This freedom acts as an open invitation--some
say challenge--to its users to make improvements. As a result, thousands
of volunteers are constantly working on Linux, adding new features and
winkling out bugs. Their contributions are reviewed by a panel and the
best ones are added to Linux. For programmers, the kudos of a successful
contribution is its own reward. The result is a stable, powerful system
that adapts rapidly to technological change. Linux is so successful
that even IBM installs it on the computers it sells.
To maintain this benign state of affairs, open source software is covered
by a special legal instrument called the General Public License. Instead
of restricting how the software can be used, as a standard software
license does, the GPL--often known as a "copyleft"--grants as much freedom
as possible (see http://www.fsf.org/licenses/gpl.html). Software released under
the GPL (or a similar copyleft licence) can be copied, modified and
distributed by anyone, as long as they, too, release it under a copyleft.
That restriction is crucial, because it prevents the material from being
co-opted into later proprietary products. It also makes open source
software different from programs that are merely distributed free of
charge. In FSF's words, the GPL "makes it free and guarantees it remains
Open source has proved a very successful way of writing software. But
it has also come to embody a political stand--one that values freedom
of expression, mistrusts corporate power, and is uncomfortable with
private ownership of knowledge. It's "a broadly libertarian view of
the proper relationship between individuals and institutions", according
to open source guru Eric Raymond.
But it's not just software companies that lock knowledge away and release
it only to those prepared to pay. Every time you buy a CD, a book, a
copy of New Scientist, even a can of Coca-Cola, you're forking out for
access to someone else's intellectual property. Your money buys you
the right to listen to, read or consume the contents, but not to rework
them, or make copies and redistribute them. No surprise, then, that
people within the open source movement have asked whether their methods
would work on other products. As yet no one's sure--but plenty of people
are trying it.
Take OpenCola. Although originally intended as a promotional tool to
explain open source software, the drink has taken on a life of its own.
The Toronto-based OpenCola company has become better known for the drink
than the software it was supposed to promote. Laird Brown, the company's
senior strategist, attributes its success to a widespread mistrust of
big corporations and the "proprietary nature of almost everything".
A website selling the stuff has shifted 150,000 cans. Politically minded
students in the US have started mixing up the recipe for parties.
OpenCola is a happy accident and poses no real threat to Coke or Pepsi,
but elsewhere people are deliberately using the open source model to
challenge entrenched interests. One popular target is the music industry.
At the forefront of the attack is the Electronic Frontier Foundation,
a San Francisco group set up to defend civil liberties in the digital
society. In April of last year, the EFF published a model copyleft called
the Open Audio License (OAL). The idea is to let musicians take advantage
of digital music's properties--ease of copying and distribution--rather
than fighting against them. Musicians who release music under an OAL
consent to their work being freely copied, performed, reworked and reissued,
as long as these new products are released under the same licence. They
can then rely on "viral distribution" to get heard. "If the people like
the music, they will support the artist to ensure the artist can continue
to make music," says Robin Gross of the EFF.
It's a little early to judge whether the OAL will capture imaginations
in the same way as OpenCola. But it's already clear that some of the
strengths of open source software simply don't apply to music. In computing,
the open source method lets users improve software by eliminating errors
and inefficient bits of code, but it's not obvious how that might happen
with music. In fact, the music is not really "open source" at all. The
files posted on the OAL music website http://www.openmusicregistry.org/ so far are all MP3s and
Ogg Vorbises--formats which allow you to listen but not to modify.
It's also not clear why any mainstream artists would ever choose to
release music under an OAL. Many bands objected to the way Napster members
circulated their music behind their backs, so why would they now allow
unrestricted distribution, or consent to strangers fiddling round with
their music? Sure enough, you're unlikely to have heard of any of the
20 bands that have posted music on the registry. It's hard to avoid
the conclusion that Open Audio amounts to little more than an opportunity
for obscure artists to put themselves in the shop window.
The problems with open music, however, haven't put people off trying
open source methods elsewhere. Encyclopedias, for example, look like
fertile ground. Like software, they're collaborative and modular, need
regular upgrading, and improve with peer review. But the first attempt,
a free online reference called Nupedia, hasn't exactly taken off. Two
years on, only 25 of its target 60,000 articles have been completed.
"At the current rate it will never be a large encyclopedia," says editor-in-chief
Larry Sanger. The main problem is that the experts Sanger wants to recruit
to write articles have little incentive to participate. They don't score
academic brownie points in the same way software engineers do for upgrading
Linux, and Nupedia can't pay them.
It's a problem that's inherent to most open source products: how do
you get people to chip in? Sanger says he's exploring ways to make money
out of Nupedia while preserving the freedom of its content. Banner adverts
are a possibility. But his best hope is that academics start citing
Nupedia articles so authors can earn academic credit.
There's another possibility: trust the collective goodwill of the open
source community. A year ago, frustrated by the treacle-like progress
of Nupedia, Sanger started another encyclopedia named Wikipedia (the
name is taken from open source Web software called WikiWiki that allows
pages to be edited by anyone on the Web). It's a lot less formal than
Nupedia: anyone can write or edit an article on any topic, which probably
explains the entries on beer and Star Trek. But it also explains its
success. Wikipedia already contains 19,000 articles and is acquiring
several thousand more each month. "People like the idea that knowledge
can and should be freely distributed and developed," says Sanger. Over
time, he reckons, thousands of dabblers should gradually fix any errors
and fill in any gaps in the articles until Wikipedia evolves into an
authoritative encyclopedia with hundreds of thousands of entries.
Another experiment that's proved its worth is the OpenLaw project at
the Berkman Center for Internet and Society at Harvard Law School. Berkman
lawyers specialise in cyberlaw--hacking, copyright, encryption and so
on--and the centre has strong ties with the EFF and the open source
software community. In 1998 faculty member Lawrence Lessig, now at Stanford
Law School, was asked by online publisher Eldritch Press to mount a
legal challenge to US copyright law. Eldritch takes books whose copyright
has expired and publishes them on the Web, but new legislation to extend
copyright from 50 to 70 years after the author's death was cutting off
its supply of new material. Lessig invited law students at Harvard and
elsewhere to help craft legal arguments challenging the new law on an
online forum, which evolved into OpenLaw.
Normal law firms write arguments the way commercial software companies
write code. Lawyers discuss a case behind closed doors, and although
their final product is released in court, the discussions or "source
code" that produced it remain secret. In contrast, OpenLaw crafts its
arguments in public and releases them under a copyleft. "We deliberately
used free software as a model," says Wendy Selzer, who took over OpenLaw
when Lessig moved to Stanford. Around 50 legal scholars now work on
Eldritch's case, and OpenLaw has taken other cases, too.
"The gains are much the same as for software," Selzer says. "Hundreds
of people scrutinise the 'code' for bugs, and make suggestions how to
fix it. And people will take underdeveloped parts of the argument, work
on them, then patch them in." Armed with arguments crafted in this way,
OpenLaw has taken Eldritch's case--deemed unwinnable at the outset--right
through the system and is now seeking a hearing in the Supreme Court.
There are drawbacks, though. The arguments are in the public domain
right from the start, so OpenLaw can't spring a surprise in court. For
the same reason, it can't take on cases where confidentiality is important.
But where there's a strong public interest element, open sourcing has
big advantages. Citizens' rights groups, for example, have taken parts
of OpenLaw's legal arguments and used them elsewhere. "People use them
on letters to Congress, or put them on flyers," Selzer says.
The open content movement is still at an early stage and it's hard
to predict how far it will spread. "I'm not sure there are other areas
where open source would work," says Sanger. "If there were, we might
have started it ourselves." Eric Raymond has also expressed doubts.
In his much-quoted 1997 essay, The Cathedral and the Bazaar, he warned
against applying open source methods to other products. "Music and most
books are not like software, because they don't generally need to be
debugged or maintained," he wrote. Without that need, the products gain
little from others' scrutiny and reworking, so there's little benefit
in open sourcing. "I do not want to weaken the winning argument for
open sourcing software by tying it to a potential loser," he wrote.
But Raymond's views have now shifted subtly. "I'm more willing to admit
that I might talk about areas other than software someday," he told
New Scientist. "But not now." The right time will be once open source
software has won the battle of ideas, he says. He expects that to happen
And so the experiment goes on. As a contribution to it, New Scientist
has agreed to issue this article under a copyleft. That means you can
copy it, redistribute it, reprint it in whole or in part, and generally
play around with it as long as you, too, release your version under
a copyleft and abide by the other terms and conditions in the licence.
We also ask that you inform us of any use you make of the article, by
One reason for doing so is that by releasing it under a copyleft, we
can print the recipe for OpenCola without violating its copyleft. If
nothing else, that demonstrates the power of the copyleft to spread
itself. But there's another reason, too: to see what happens. To my
knowledge this is the first magazine article published under a copyleft.
Who knows what the outcome will be? Perhaps the article will disappear
without a trace. Perhaps it will be photocopied, redistributed, re-edited,
rewritten, cut and pasted onto websites, handbills and articles all
over the world. I don't know--but that's the point. It's not up to me
any more. The decision belongs to all of us.
For a selection of copylefts, see http://www.eff.org/IP/Open_licenses/open_alternatives.html
The Cathedral and the Bazaar by Eric Raymond is available at http://tuxedo.org/~esr/writings/cathedral-bazaar/
THE INFORMATION IN THIS ARTICLE IS FREE. It may be copied, distributed
and/or modified under the conditions set down in the Design Science
License published by Michael Stutz at http://dsl.org/copyleft/dsl.txt