Social Graph: Een wereldwijd Social Network Storage Service

Ik heb dit zojuist op mijn engelse blog gepost, maar ben ook erg benieuwd naar reacties van Nederlandse lezers, dus vandaar dat ik deze post ook hier plaats (in het Engels). Sla ik de plank mis? Is wat ik zeg al lang bedacht? Please let me know!

I’m attending the Future of Web Apps conference at the moment and one of the main topics being talked about is the social graph and portable social networks. We have seen various developments being talked about, like OAuth, and OpenID.

But someone who really nailed it was Jyri Zengestrom. I love the thoughts of Jyri on object centered sociality and he said that we really don’t have much about portability. What’s the use of exporting your dogster’s network to LinkedIn for instance?

Jyri talked about interoperability as thé main thing. And I think he is right. But how do we accomplish interoperability? I’ve read the article of Brad Fitzpatrick on the social graph and I love it, but my thoughts on the subject are a bit different. He talks about a central organ that exchanges social networks with other sites, aggregating and making one master copy (or something). I propose a slightly different approach.

What popped in my mind (just today so forgive me of any possible ignorance) was a form of a social networking focused alternative to Amazon ’s S3 (Simple Storage Service). What S3 does is offering a central space for storage, which could be accessed through REST or SOAP protocols. What if we translate this concept to Social Networks?

We could develop a central Social Network Storage Service (SNSS) where every social networking service (or any site that uses profiles) can tap into. This goes further than central authentication and profiling that OpenId or Sxipper do. It’s about storing your network centrally, which can be used by every service available on the web through API’s. Social networking sites do not store there networks locally, but use the network stored in one central place. The social networking on their turn can focus on the object of sociality their service is focused on. Extending such a service with authentication and grouping could make it extremely powerful…

Ofcourse there are downsides to such an approach. Do we wánt a central Service that stores our networks? What about my privacy? And what if such a service cashes out (if possible). But I guess the benefits far outweigh the downsides.

But when we do build such a service, how do we come up with a protocol for social network interoperability? And how do we get the already existing networks prepared to use such a central service?

I’m interested in any thoughts on this topic or relevant material. Are the things I’m saying new or are they, in a much concrete way, already being realised?

If you enjoyed this post, make sure you subscribe to my RSS feed!