Jeremiah

As a showing sign of maturity, the latest discussion around social networks is interoperability.

When Facebook enabled Applications and turned itself into the hottest web platform, it invigorated developers. Suddenly, any web application could gain a social element. The realization of this power came with a new concern: many different entities now owned data and sharing data became important.

The problem existed before Facebook applications, but it was less apparent. I am a member of many different “social” web applications and communities (Facebook, MySpace, LinkedIn, Wallop, Virb, Flickr, Pandora, del.icio.us, to name a few). I had to create a new user account, copy and paste my standard profile, upload my standard avatar¹, and “re-friend” all of my friends for all of these sites. Projects like OpenID² are working on the single sign-on part, but there is no consensus on the rest of the problem.

The irony is that collaboration is a core value of the supposed Web 2.0 and while many innovative new services allow collaboration between users, very few services collaborate with other services.

A Billion Little Pieces

Critics of centralized social networks primarily cite data portability as the primary concern. Social networks were created as self-contained communities. Blogging, however, grew up interconnected. Journaling, forums, blogrolls, and microformats like XFN formed communities before social networks existed. Social networks centralized these activities and used various methods to notify a user’s friends when they occurred³, but they were still closed to anyone outside of the community in a very un-web way.

Facebook Applications are a great start for those interested in decentralizing data. I can now post all of my blogs, links, photos, etc. to my own content management system (unfortunately WordPress at the moment) and have it imported into my Facebook profile as part of the news feed or duplicated as part of an App. All of my data can reside on my website. It’s already published in my Atom and RSS XML feed, but the masses still don’t use news readers. My dad is on Facebook, but I doubt he’s heard of Google Reader.

Unfortunately, this is only a start. Distribution from a CMS to Facebook is a one-way street and Facebook is the only major social network with an extensive API.

Why Facebook Wins

Ultimately, the power lies with the aggregator regardless of where users store the primary data. Whether users post their own social timelines and graphs through XML formats or new APIs from a decentralized source, the centralized service that blends and connects the user with the user’s friends’ timelines wins.

Facebook brought the social graph and timeline to the masses and Facebook opened its network to allow developers to add to it. Users now have the choice to post photos through Facebook or to post photos to Flickr and have it aggregated to their profiles.

Everyone benefits from decentralization because it keeps the centralized aggregators even. Facebook already has an open API⁴. It could be more open, but it makes sense to start building a decentralized social data system with the leading social aggregator.

1 Gravatar doesn’t count as a solution for avatars because it’s centralized and prone to disruption of service.
2 Many Facebook Apps still require me to register on another site before I can use the feature of the App on my profile. Annoying!
3 LiveJournal’s friends page, Facebook’s news feed, and MySpace’s alerts did what Atom and RSS couldn’t do: get everyone to use them.
4 Mosoto uses it and has built a spiffy parallel community on top of Facebook.

Sources: Thoughts on the Social Graph, Slap in the Facebook: It’s Time for Social Networks to Open Up, Adactio: Portability, How Mark Zuckerberg Turned Facebook Into the Web’s Hottest Platform

This entry was posted on Wednesday, September 12th, 2007 at 8:46 and is tagged web development.