Beautiful Architecture

Summation

Facebook’s user-contributed social information effectively motivates the utility of most any page on http://facebook.com. However, this data is so universal that some of its best uses appear when it is integrated with the stacks of outside developers’ applications, made possible through data technologies such as Facebook Platform’s web services, data query services, and FBML. Starting from simple internal APIs that get a user’s friends or profile information, the whole range of improvements we’ve detailed in this chapter show how to reconcile continually expanding methods of data access with the expectations of the container site, especially the requirements of data privacy and site experience integrity. Each new change to the data architecture presents new problems in the web architecture, which are resolved through even more powerful improvements to the data access pattern.

Though we’ve focused entirely on the potentials and constraints of applications built using Facebook’s social data platform, new data services like these need not be limited to social information. As users contribute and consume more information that is useful across many container sites (data such as collections, reviews, location information, personal scheduling, collaboration, etc.), platform providers of all kinds can benefit from applying the ideas behind the unique data and web architecture of the Facebook Platform.