Illusionary API’s


A long time ago, before Facebook was “open,” I devised the first API to extract data from the site for visualization and scientific inquery. More significant than the fact that the API was the first of its kind, it was a signal to the blogging world that Facebook and other social media websites had to become more open. At the time TechCrunch and others quickly picked up on my efforts and questioned why our data was being locked inside these sites. The effort was fairly successful for about two years an API/framework was released into the wild, officially this time.

While Facebook’s API is fairly good, giving access to all of your information, it is very restrictive on the other information that can be seen on the site about your friends. A perfect example is pictures. While you can see all of your friends pictures, through the API you can only see your own. While partially understandable, it places an incredible hindrance on what can be done with the API. What possibly useful tool can one make with just your own information? Visualize my address? Show the pictures that I took myself? Tell me how I can be reached?

Despite the hype of Facebook applications none are actually useful or pertain to the data that can be extracted with the API (except perhaps your name and the list of friends you have so they can be spammed to join the app). This is a trend that can be seen in all social media API’s. When was the last Twitter or Digg API that actually did something novel with the vast amount of data that these sites generate? Probably the worst part about it is that there is no list of restrictions that these API’s have. Yesterday I spent an entire day developing an application that became entirely useless after finding out that I could not see other people’s pictures. Was this piece of information on the API’s documentation, no.