Personal Democracy Plus Our premium content network. LEARN MORE You are not logged in. LOG IN NOW >

Code Warriors Debate Whitehouse.gov Robot Commands

BY Sarah Granger | Thursday, January 22 2009

As the tech community pored over the new whitehouse.gov site, one of the first subterranean changes noted was that of a file most people would never notice called robots.txt. This file serves as a notice to search robots informing them of what files they should or shouldn't survey. Upon seeing the new version of the file, some noticed that it only had two lines of code excluding robot searches vs. the former whitehouse.gov robots.txt that had nearly 2400 lines of exclude lines by the end of the Bush administration, sparking excitement and controversy over what the change means in terms of government transparency.

The text from the new robots.txt file:

User-agent: *
Disallow: /includes/

A sampling from near the end of the previous file:

Disallow: /president/text
Disallow: /president/waronterror/iraq200404/text
Disallow: /president/waronterror/photoessay/text
Disallow: /president/winterwonderland/iraq
Disallow: /president/winterwonderland/text
Disallow: /president/world-leaders/iraq
Disallow: /president/world-leaders/text
Disallow: /president/worldunites/iraq
Disallow: /president/worldunites/text

Cory Doctorow, Editor of Boing Boing and Former Outreach Director for the Electronic Frontier Foundation was one of the first to report this finding, with just the facts followed by a bunch of commenters asking for explanations.

Proponents of the belief that the move to the vastly smaller file was a statement about transparency claimed were ecstatic. According to Patrick Thibodeau of ComputerWorld, New York blogger James Kottke "thinks that by eliminating the Bush disallow list on its first day in office, the Obama administration was sending out a symbolic message." Kottke, in his post on Tuesday, alluded to the "huge change in the executive branch of the US government." In e-mail to Thibodeau, Kottke wrote: "One of Obama's big talking points during the campaign and transition was a desire for a more transparent government, and the spare robots.txt file is a symbol of that desire."

Presenting an alternate view, Declan McCullagh of CNET News pointed out that the Bush whitehouse.gov robots.txt file followed the letter of coder law for the most part in terms of what to disallow with the exception of a few incidents that were corrected. McCullagh brought to attention the idea that perhaps the new robots.txt file is actually too short. "It doesn't currently block search pages, meaning they'll show up on search engines--something that most site operators don't want and which runs afoul of Google's Webmaster guidelines."

While most of the technical experts weighing in suggest and expect that the robots.txt file should grow, most of them explain it as just a normal process a website undergoes over time. Andy John, a search developer for DeepDyve, puts it like this: "robots.txt is just a request. Robots can do whatever they like anyway." He then went further to describe what that means. "For example, there is a program "wget" (web get). You give it a URL, it downloads it and saves the file... You can tell it to download an entire site. It honors robots.txt by default. But by just adding these parameters you can tell it to ignore robots.txt and get everything: wget -erobots=off http://www.whitehouse.gov

As to why those who developed the new whitehouse.gov site would want to code it this way, Jaelithe Judy, a Search Engine Optimization specialist and political blogger says "Google does generally encourage webmasters to use disallows to keep from having their search pages spidered; this is to help keep a Google search from returning a whole page of search results from other sites' internal search engines, instead of relevant original content. However, in some cases a search result from a site is a meaningful result. For instance, when you are searching for 'DVD recorders' and the Amazon search page for 'DVD recorders on Amazon' pops up, that might actually be useful to most users."

She added that "Google is still trying to work out how to sort annoying search-generated page results from the useful ones. The Whitehouse.gov ones may lean toward being useful. For instance, if you are a middle school student doing a report on the First Ladies, and you get a Whitehouse.gov search page for First Ladies, that has all sorts of different links to different sorts of information, that might actually be useful."

The bottom line about robots.txt? John says, "It's really more of a serving suggestion."

News Briefs

RSS Feed monday >

After Election Loss, Teachout and Wu Keep Up Net Neutrality and Anti-Comcast Merger Campaign

The Teachout/Wu campaign may have lost, but their pro net-neutrality campaign continued Monday as both former candidates participated in a rallly in New York City marking the final day to comment on the Federal Communications Commission's Internet proposals and kept up their pressure on Governor Andrew Cuomo. GO

friday >

NYC Politicians and Advocacy Groups Say Airbnb Misrepresents Sharing Economy

A coalition of New York election officials and affordable housing groups have launched an advocacy effort targeting Airbnb called "Share Better" that includes an ad campaign, a web platform, and social media outreach. GO

First POST: Data Dumps

The Internet Slowdown's impact on the FCC; Uber drivers try to go on strike; four kinds of civic tech; and much, much more. GO

thursday >

First POST: Positive Sums

How Teachout won some wealthy districts while Cuomo won some poor ones; DailyKos's explosive traffic growth; using Facebook for voter targeting; and much, much more. GO

wednesday >

First POST: Emergence

Evaluating the Teachout-Wu challenge; net neutrality defenders invoke an "internet slowdown"; NYC's first CTO; and much, much more. GO

More