Over at Public Knowledge, Sherwin Siy is lecturing Ranking Member Coble about the relationship between file-sharing and identity theft. Mr. Siy claims that his recent editorial shows that Legislators Just Don’t Get P2P:

“Coble … implicitly equat[es the risk of identity theft] with both copyright infringement and p2p software generally. This is a real mistake, and a real cause for concern if that’s the attitude policymakers take when approaching online copyright infringement.”

Mr. Siy both misstates the content of this editorial and ignores evidence strongly suggesting that risks of identity theft are associated with the piracy-driven business models of some distributors of file-sharing programs. As in Grokster, it is once again Public Knowledge that Just Doesn’t Get P2P.

First, it is not a “mistake” to infer a link between inadvertent sharing of sensitive files and the prevalence of copyright infringement. USPTO has explained why the available evidence strongly suggests that inadvertent sharing of sensitive files may be a side-effect of distributor’s seemingly deliberate efforts to trick program users into inadvertently sharing infringing files. Even more information is available here.

Indeed, it is Mr. Siy’s alternative explanation for persistent inadvertent sharing—“bad software” or “dumb software design”—that is far more difficult to square with the available data. So difficult, in fact, as to suggest an ostrich-like desire to cling to a treasured belief in the face of contrary evidence.

For example, consider the “search wizard” deployed in the program LimeWire. A search wizard is a “feature” triggered when a file-sharing program is installed on a computer that has never before housed any version of that program. These search wizards exhibit a nasty behavior: They tend to “recommend” that new users with large collections of audio files in their “My Music” folder should share their “My Documents” folder recursively—and thus become candidates for both identity theft and an infringement lawsuit.

The mere fact that such a “feature” was deployed in any file-sharing program is puzzling: Anyone sufficiently familiar with Internet and networking technologies to develop a file-sharing program ought to foresee that bad things will happen if a program advises its newest users to share their “My Documents” folder recursively.

Moreover, after some developers proved that they did seem to be unusually dense, computer scientists and Congress connected the dots for them. In mid 2002, Usability and Privacy: A Study of KaZaA Peer-to-Peer File Sharing identified KaZaA’s search wizard as one of two features that were causing its users to share files inadvertently. In 2003, committees in the House and Senate held hearings that confronted distributors with Usability and the causes and consequences of inadvertent sharing. In response, the distributors of LimeWire and other programs devised a Code of Conduct that supposedly addressed the concerns raised in Usability and the hearings. One Code provision stated that LimeWire must be designed “to reasonably prevent the inadvertent [sharing] of the contents of the user’s … principle data repository.”

Nevertheless, in early 2004—after Usability, the congressional hearings, and the Code—LimeWire deployed a search wizard that tended to recommend that new users with large collections of media files should share their “My Documents” folder and all of its subfolders. The results were predictable and predicted: Within a few months, The Boston Globe was asking LimeWire why users of its program were inadvertently sharing classified military data. LimeWire’s Chief Operating Officer explained that its search-wizard was probably to blame.

At this point—after specifically identifying the search wizard as a threat to both military security and its civilian users—one would expect that LimeWire would have promptly removed it. Wrong: LimeWire kept deploying its search wizard for nearly three more years and removed it only under renewed pressure from the House Committee on Oversight and Government Reform.

With all due respect, I fail to see why a public-interest-driven analysis of copyright and technology policy would assume that the conduct described above necessarily reflects mere “dumb software design.” As USPTO noted, given time and repeated warnings, dumb designs should get corrected; by contrast, efforts to dupe program users into sharing files inadvertently should just proliferate and metastasize.

Indeed, there is something odd about Public Knowledge lecturing Congress about inadvertent sharing. As the above example shows, Congress can take due credit for repeatedly calling attention to this problem and for illuminating its causes and consequences. None of the supposedly tech-savvy public-interest groups that focused on file-sharing can do so. But for Congress, it would be much easier to believe that inadvertent sharing could be an accidental side-effect of “dumb software design.”

Second, Mr. Siy has also grossly mischaracterized Ranking Member Coble’s editorial: It neither equates all “p2p software generally” nor states that all uses of file-sharing programs are infringing. Siy’s claim that it does so “implicitly” is illogical pedantry: He concludes that Ranking Member Coble thinks that all uses of all file-sharing programs are infringing because he praised Ohio University (OU) for adopting educational, procedural, and technological measures that reduced the number of takedown notices it received by 90%.

Neither fact nor logic supports that a claim that praise for OU’s success at deterring infringement implies a belief that all file-sharing is infringing. Factually, OU has not, as Mr. Siy claims, “banned all p2p applications”: Indeed, OU states that such applications have lawful uses and makes exceptions for students who need to make legitimate use of one of more of them.

Logically, Mr. Siy’s “implication” seems akin to the fallacy of division. Ohio University’s acts were driven only partly by concerns about infringement: While OU did ban file-sharing programs used predominately or almost exclusively for infringement, concerns about network congestion also led it to ban use of programs rarely used for infringement. So Mr. Siy is trying to argue that praise for the success of OU’s efforts to constrain infringement “implicitly” praises the non-infringement-related aspects of OU’s efforts, and thus implies that all file-sharing is infringing.

But praise for the success of OU's infringement-related actions implies neither praise nor condemnation of actions related to issues like network congestion. For example, I join Ranking Member Coble in praising OU for successfully using technology to sharply reduce conflicts between copyright holders and its students. Nevertheless, I have no opinion about OU’s ban on lawfully used file-sharing programs: It surely frustrates entities using file-sharing to promote lawful ends, like authorized content distribution, but OU can surely decline to use its scarce, publicly-funded network resources to provide free hosting and distribution services to unaffiliated corporations.

Nor do other colleges and universities have to ban all “p2p” programs in order to use available technology to deter students from misusing the resources of federally funded public educational institutions to violate federal law. Sophisticated modern filtering, shaping, or blocking technologies can let institutions target infringing use more precisely.

In summary, Mr. Siy contorted evidence, facts, and logic in order to reiterate the shopworn cry they-don’t-get-p2p. In the long run, pedantry that strains to imply fundamental error from any ambiguity is not just a cheap shot, it’s also unwise: Pedants tend to impale themselves on their own swords.

For example, Mr. Siy says, “P2P is … not an infringement-only protocol…. [I]t’s a good way for developers to share large files and updates.” He thus “implicitly” claims that all “P2P” protocols—Napster, FastTrack, Gnutella 0.4, Gnutella 0.6, and many others—are all “good way[s] for developers to share large files and updates.” But not every file-sharing program or protocol is like BitTorrent, and both LimeWire and Professor Wu have explained why even BitTorrent is a lousy means to distribute “large files” that are not already very popular.

Frankly, I would have thought that Grokster itself would have deterred resort to they-don’t-get-it wailings. Before Grokster reached the Court, various members of Congress proposed many ways to try to reduce file-sharing piracy—beacons, blocks, self-help, federal civil enforcement authority. All were ridiculed: p2p-specific proposals were not technology-neutral; technology-neutral proposals were overbroad. And when Senators proposed to recognize an inducement exception to the Court’s Sony test, oh, the horror: Scores of Internet savants ran around shrieking they-don’t-get-it histrionics and prophesying the demise of innovation, Sony, radios, Silly Putty, Lego, player pianos, the New York Times etc.

Until one year later when, in Grokster, all nine Justices of the United States Supreme Court unanimously rejected such arguments and held that there was an inducement exception to the Court’s Sony test. The supposedly doomed New York Times then praised the Court’s inducement ruling.

Consequently, they-don’t-get-it histrionics are neither warranted nor wise. Because if we look at inadvertent sharing and Grokster and ask who really did "get it," the score is clear: Congress: 2, Public Knowledge: 0.

posted by Thomas Sydnor @ 2:08 PM | Enforcement & Remedies, Free Culture Movement, Internet: P2P, Search Engines..., Privacy and Security

Link to this Entry | Printer-Friendly | Email a Comment | Post a Comment(0)