I recently made a prediction and an evaluation of those predictions around Facebook Connect. Facebook has done a great job at updating the Facebook client and at supplying some convenient JavaScript libraries (mainly, widgets that embed core Facebook features into a web site). By their own description, this is really the problem they were trying to solve. With all the attention focused on social applications running in the Facebook canvas (inside the Facebook “frame”), it’s very easy to overlook that the client has always supported an application to be a stand-alone web application that uses the Facebook client to communicate with someone’s social graph. I even integrated Facebook login and identity into MediaWiki as a very early experiment to see how multiple platforms could play well together.

Facebook Connect, in short, allows a site owner to:

  • Use Facebook to authenticate a user, more or less in the same way OpenID enables,
  • Use some of the Facebook “core applications” (a concept we have in Ringside, where the “core applications” are built using the same architecture as a standard social application, but they are embedded in the server, much in the same way that the Friend, Share, and Comments parts of Facebook are “built-in”),
  • Use some brand new widgety bits to allow users to login and make site-specific friends without leaving the Facebook Connect-enabled web site

Another way to think about Facebook Connect is that it reverses the “normal” relationship between Facebook and a Facebook application, placing the application “in front of” Facebook (from the perspective of the user looking at the browser) instead of having the Facebook interface “in front of” the application. This isn’t a perfect analogy, since Facebook Connect still gets “in front of” the application (the web site).

Facebook Connect isn’t about connecting the Social Web, and I hope this entry clarifies that for everyone who thinks it does. It can’t, because it still assumes that there is a social network (hub) and a bunch of applications (spokes) and doesn’t (yet?) envision an environment where multiple communities come together via interesting technology. You can think about it this way: one interesting omission in the announcement is anything to do with applications. One important thing that Facebook has not enabled is running an interesting Facebook application on someone else’s web site (except for the single application that is represented by that web site). Until communities can connect to major social networks (and directly to each other, one-on-one), the vision of the Social Web will not be complete.