Ho boy, was I proud of this title ^^^ and blog post by @scobleizer. It was on May 22nd 2010, exactly three years ago. Facebook was engulfed in a major privacy scandal, the Diaspora team was wrapping up an amazing kickstarter campaing and Google was having multiple "social web" sessions at I/O and I had just been interviewed by Robert on the topic of Federated Social Networks and presenting him our work on Onesocialweb.
It felt like a perfect storm for making Federated Social Networks happen. Yet.. three years later, this dream feels further away than ever. What went wrong? Where do we go from here?
What went wrong?
Although many things eventually played against the various projects, I think we can single out three key factors:
- Loosing the leaders: A big chunk of the thought leaders got hired by major companies in a very short period of time. In fact, most of them went to Google.
- Analysis paralysis: Although we shared the same goals, the Federated Social Web community got quickly paralysed by endless debates on how to get there. XML vs JSON vs RDF, email vs uri identifier,etc...
- Building Cathedrals: We were too busy architecturing the perfect protocols and not paying enough attention to the developers (and the challenges of interoperability) and the end users.
I think this last point is crucial, and was nicely phrased by @tomcoates as the following (in CAPS indeed :-) : "THINGS THAT USERS DON'T UNDERSTAND THAT DON'T MAKE MONEY **DO NOT SUCCEED**. THEY GO BUST OR FALL AWAY AND GET REPLACED BY THINGS THAT DO MAKE MONEY AND THAT USERS GET!"
What next then?
Well, over the last months I fell in love with the #indieweb community. Which, could be summarized as: "stop talking, start coding". The idea is simple: get your own domain, host your site there, and slowly work towards federating with others.
Instead of focusing on protocols, it focuses back on YOU and your needs. You get immediate value out of it (you got a blog) and you make exciting progress with a community of likeminded folks. As an example; a couple weeks ago, we managed to federate 5 different implementations around a single comment thread. Yes, this stuff is real.
I also like the 'low-tech' approach of the indieweb. Instead of complex API, we rely on the web and some simple markup in html pages to share content in an interoperable way.