from the watch-out-SOPA dept
In the 2005 "Grokster" decision, the Supreme Court ruled unanimously that file sharing networks could be held liable for copyright infringement if they take "affirmative steps" to encourage infringement. Grokster closed down as a result, and the recording industry pretty much assumed it had won that battle.
But as a fascinating analysis by Rebecca Giblin of what happened afterwards points out, against the industry's expectations, P2P filesharing flourished:
By 2007, two years after the US Supreme Court decided Grokster, there were more individual P2P applications available than there had ever been before. The average number of users sharing files on file sharing networks at any one time was nudging ten million and it was estimated that P2P traffic had grown to comprise up to 90 percent of global internet traffic. At that point content owners tacitly admitted defeat, largely abandoning their long-time strategy of suing key P2P software providers and diverting enforcement resources to alternatives like graduated response or "three strikes" laws.
So what happened? In her article, Giblin suggests pre-P2P laws were rooted in the physical world, and this led to four false assumptions being made about P2P software development that allowed it to thrive despite the Grokster decision.
The first [assumption] is that everybody is bound by physical world rules. Assuming this rule had universal application, various secondary liability principles evolved to make knowledge and control pre-requisites to liability. But software has no such constraint. Programmers can write software that will do things that are simply not possible or feasible in the physical world. So once the Napster litigation made P2P programmers aware of the rules about knowledge and control, they simply coded Napster's successors to eliminate them – something no provider of a physical world distribution technology ever managed to do.
Napster's fatal weakness was its centralized directory servers: shutting them down meant the entire system collapsed. True P2P software has no central directories, and therefore no one point of failure. The Grokster decision tried to address that issue:
In response, the US Supreme Court in Grokster created a brand new legal doctrine, called inducement, that did not rely on either knowledge or control. That rule was aimed at capturing "bad actors" - those P2P providers who aimed to profit from their users' infringement and whose nefarious intent was demonstrated by "smoking guns" in their marketing and other communications. But the inducement law failed to appreciate some of the other differences that make the software world special and thus led directly to the explosion in the number of P2P technologies.
That's because three other physical-world assumptions were made:
One is that it is expensive to create distribution technologies that are capable of vast amounts of infringement. Of course in the physical world, the creation of such technologies, like printing presses, photocopiers, and VCRs required large investment. Research and development, mass-manufacturing, marketing and delivery all require massive amounts of cash. Thus, the law came to assume that the creation of such technologies was expensive.
Of course, free software offers hundreds of thousands of counterexamples to all of those assumptions: major applications are written in people's bedrooms (hello, Linus), "just for fun" (Linus again), with the aim of sharing them as widely as possible (hello, Richard Stallman.)
As a result of this deep misunderstanding, the Grokster decision led not to the disappearance of P2P programs, but to an unprecedented flowering of open source implementations:
That led directly to the next assumption – that distribution technologies are developed for profit. After all, nobody would be investing those massive sums without some prospect of a return.
Finally comes the fourth assumption: that rational developers of distribution technologies won't share their secrets with consumers or competitors. Since they needed to recoup those massive investments, they had no interest at all in giving them away.
When the US Supreme Court created its new law holding P2P providers liable where they "fostered" third party infringement, as evidenced by such things as business models, marketing and internal communications, the result was an enormous number of programmers choosing to create new applications without any of those liability attracting elements. In the absence of any evidence that they had set out to foster infringement, they could not be liable for inducement, and having coded out of knowledge and control they could not be held liable under the pre-P2P law either.
As Giblin concludes:
The mismatch between the law's physical world assumptions and the realities of the software world meant that the law created to respond to the challenges of P2P file sharing led to the opposite of the desired result: a massive increase in the availability of P2P file sharing software. The failure of the law to recognise the unique characteristics of software and software development meant the abandonment of the litigation campaign against P2P providers was only a matter of time.
This is an important reminder that framing legislation affecting the world of software is not like traditional physical-world law-making. Problems are precisely what drives forward open source development by offering an interesting "itch" that needs scratching, so new laws merely provide an impetus to the creation of more sophisticated ways of circumventing them. It's a lesson the supporters of SOPA would do well to learn.