You are not logged in. LOG IN NOW >

Wikileaks Has More Google Juice than Justin Bieber, but What Will Searchers See?

BY Nick Judd | Friday, December 3 2010

Over the last couple of days, Wikileaks has vaulted into the ranks of the top searched-for terms on Google, both internationally and in the U.S.

Take it as proof positive that the best way to get people to look at something is to tell them they shouldn't: Today, news broke that the Library of Congress is blocking access to Wikileaks material, which includes over 250,000 secret or classified U.S. diplomatic communications that the organization is releasing over time in coordination with major news outlets around the world, and that a Columbia University School of International and Public Affairs alumnus working at the U.S. State Department advised SIPA students not to post or comment on those documents. This alumnus, according to an email obtained by blogger Will Chen and posted on Wise Bread, said that doing so might "call into question your ability to deal with confidential information, which is part of most positions with the federal government."

Based on Google Trends, Wikileaks is far more sought-after on Google abroad than here in the U.S. New York, N.Y. is — surprise — the only American city among the 10 cities where searches for Wikileaks originate, based on that data. (And Wikileaks looks to only narrowly outpace the Biebs for popularity there.)

What people see when they search is another matter. Wikileaks.org URLs — which are still down after the domain name service provider, EveryDNS, stopped providing service — are at the top of my page rankings. This means the top result in Google for people searching for Wikileaks will be a broken link.

The organization's website was duplicated at Wikileaks.ch, hosted with a Swiss top-level domain on servers in Sweden and, according to the Washington Post, also France. But the Swiss Pirate Party has announced on its website that Wikileaks.ch was registered by their party, using EveryDNS, and is now also blocked.

More broken links. However, they've provided links to alternative URLs where Wikileaks information is available.

Additionally, another page has been set up to index, in a very Web 1.0 way, other mirrors of Wikileaks content.

With so many people turning to Google to find out about Wikileaks, there's a serious open question: What would Google do, if it was asked to block access or omit certain content from search results? And would it take different steps in the U.S. than abroad?

I've asked the Google press team those questions. Really, though, all those questions do is lead to more questions about the infrastructure of the web itself.

This whole exercise is, as Nancy Scola explained earlier today, an education in the infrastructure of the web and how nations can exert a lot more influence over the transnational Internet than many would like to think.

As Nancy Scola explained earlier today, domain name service — literally the service of taking requests from all of those people looking for what's behind wikileaks.org and pointing their browsers, via the corresponding IP addresses, to the proper servers and the content on them — is the point of weakness that is now sending searchers for Wikileaks scrambling across the Internet. Because the nameservers aren't sending people looking for Wikileaks.org — and now Wikileaks.ch — onward to the right content, it doesn't matter if they've got solid web hosting or own the domain. The only way to get to the content, unless there is some important aspect of how the Internet works that I am not properly explaining, is through the IP address, the series of numbers that only one computer connected to the Internet should have at one time. Supporters of Wikileaks are sharing IP addresses of mirrors of the site, as well as domains set up to either host copies of Wikileaks or redirect users to Wikileaks' IP address.

While the web is decentralized, there are choke points in the network: The relatively few organizations certified to be domain name registrars, where you go to buy domain names and then match them to nameservers, for example; the top-level domain providers, who control, for example, use of *.com and *.org; and, to a lesser degree, web hosts.

After the Department of Homeland Security seized domain names last month for allegedly being connected with copyright infringement, a group of people led by the Pirate Bay's Peter Sunde started a project to create a new domain name system infrastructure. Called Dot-P2P, the goal is to distribute the work of figuring out what content belongs with what domain names across a peer-to-peer network. Ars Technica has a good look at some of the challenges facing this group.

Are the Internet's centralized components a public good? Are they a tool for censorship? Those questions may not be answered in this Wikileaks debate, but that centralized architecture is already being challenged.