As you may know if you’ve been to Barcamp Ghent, i’ve been reading up on data portability. OpenID allows you to locate your authentication in one single place. OAuth allows you to share resources between sites without giving out your password for one site to the other site.
This, combined with the whole privacy mess that is Facebook and other such stories, has been gisting in the back of my brain. And then, bam, it hit me, after a discussion with my co-geek cousin: this is where the data portability standards are going !
There will be an open standard that allows you to have one social network (1), and other sites will communicate with it to share resources with your friends. If you want, you can even host it yourself – as long as it responds to the right queries in the right secure way (in fact, this idea is not that far from OAuth – the resource you would be sharing is the relationship).
Of course this social network should be fine-grained enough – say for instance allowing you to specify whether an acquaintance is, say, a friend, or a colleague. But specified enough, so that it would be truly interusable between websites. And you would have a say as to which of your friends get to use or see which of your resources.
The added benefit of this, besides privacy for yourself, is also to the website developers: imagine, any web2.0-ish application out there needs to implement their own version of a social network. Which means endless repetition of what should be the same social network, and a distraction from their core business (book collection/photo album/video sharing/…).
I think i might try to contact the data portability people group directly, to see where this is going (if it is going, i only found FOAF which is part of the solution). If i can collaborate, that would be great – if i can implement it as one of the standard beta developers, then i’m more than happy too.
This is probably something some of you figured out a while ago, but i was certainly happy to see where this might be leading. Adoption might take a while, as it will probably be fought/ignored by big players who’d like to keep their hands on all that valuable marketing information.
There are actually already several projects for trying to implement open distributed social networks.
One that looks nice to me is http://noserub.com
Doesn’t surprise me. The idea is so obvious once you think of it … thanks for the link !
I’m with Dieter: several initiatives going on in that space.
This is the one Chris Messina is driving:
http://diso-project.org/
It wants to bring social networking – functionality within WordPress (decentralised messaging, permissions, livestreams).
And the mentioned Noserub is in fact an open source, decentralised implementation of the much-hyped Friendfeed.
But I whole-heartedly disagree with the (perceived) goal of having one social network…
People will always have different social networks… Depending on the context (business, dating, leisure, hobbies) the definition of people you trust/want to connect with is totally different…
@Pascal. Good point, in real life we have multiple social networks, but isn’t it nice to have 1 single interface to them all? This is imo what is meant with ’1 social network’, because the different contexts and other needed features could all be modeled into 1 ‘frontend’
@Pascal: your comments are interesting, as always.
I agree we have different ‘subgroups’ of contacts, but these could be articulated within one social network. But yes, more than one is still an option. But not the situation we have now.
My point is mostly that:
1. we have endless duplications of those networks in every single application
2. we should be allowed to have control of those data – i.e. host them wherever we want.