|by Flemming Funch|
It is odd that nobody has made an open, public, platform free social network yet. Social networks *should* be independent of any particular central website, but it hasn't happened yet. They're all isolated, proprietary islands. Or, even if they're built on open source software, there's no way of exchanging much between them.
My main contribution to the field of social networks was and is the New Civilization Network. Which still exists, and I'm posting this message from my blog within it, but which is a bit dated in terms of it's software. But it has had profiles and buddy lists and blogs and forums and picture galleries for more than a decade. At the beginning, it was the intention to create a platform for social networks that would span many servers, run by different people, so that one could plug into it from many places, each having a different flavor, but accessing the aggregated resources of all of them. "The Sprawl" was the code name for this plan. Didn't actually materialize, so today NCN is just another isolated island in cyberspace.
Facebook, Ryze, Orkut, LinkedIn and all the others are also separate islands. Several of them have interfaces so that programmers can add modules to them, or access a limited amount of information from them. But nothing at all that allows you to move seamlessly between them.
People who make virtual worlds are working on standards that would allow you to go from one to the other. There's no terribly good reason you shouldn't be able to teleport from Second Life to World of Warcraft. To create something like that would require that one defines a minimum of characteristics that a virtual character would have everywhere, so that each world could implement a way of supporting them.
Social networks could do the same. If there was a shared way of representing a person/profile, what they do, and who they know, it should be something transferrable. But it is probably harder than the virtual reality scenario. Because a big part of it is the relation one has with other people. Or, we could say, with other profiles. And how do they exist, separately from a particular network site and its software?
That probably ties in with identity. How does one know that the ffunch on Facebook is the same as ffunch in Orkut? Not without some shared standard of identity, one that everybody would support.
But, assuming that the problem of a universal identity system was worked out, what would be interesting would be if somebody made the pieces of a platform-free social network.
The web is based on some standards for presenting content, and linking it together. It doesn't matter what server anything is on, it is just part of the address. So, I can link to your stuff from my webpage, hosted on any which web server, and I can even include some of your stuff on your server in my page on my server.
That is of course how a social network should work too. I don't just have friends who use the same brand of shoes as me, or who drive the same brand of car. I have friends wherever they happen to be, and whatever they do. The web way of doing it is of course that I can link to them, and make a list of them, no matter where they are, what ISP they're using, etc.
I can call your Nokia phone from my iPhone. Doesn't matter at all, as long as I have your number. That's kind of the point of network structures, that one can link freely between nodes. That's kind of the definition of a network. So, what we call Social Networks are kind of fake networks. One can list web links as before, but one can only link to people who are wearing the same shoes, who are subscribed to the same website.
So, all it takes is that somebody comes up with a way of expressing profiles and lists of contacts in a standardized format that can be put on any website. Some kind of XML thing that has most of the things you'd find in a profile on Facebook. Your interests, where have you worked, the lists of movies you like, and links to other profiles.
There is, of course, FOAF, which is indeed an open format for expressing a list of contacts. And, really, it is what the Semantic Web is supposed to do. And the World Wide Web could transform into the Giant Global Graph, connecting everybody with everybody else, no matter what server their stuff is on.
"Social Graph" is a word that expresses what Social Networks should have been, but aren't: the actual network of connections between people.
I don't know what's so hard about it, since it hasn't yet happened. Other than it maybe is more complicated than it seems. Whenever I try to look at the work of standards groups that are working on pieces of that puzzle, I walk away confused. I'm not sure if that's because it really is terribly complicated, and I'm a little slow and/or impatient, or because they're making it more complicated than it is.
Sometimes a major advance in Internet standards comes from somebody who didn't bother dealing with committees taking years to work through all the complexities. Dave Winer invented RSS and XML-RPC and other good things, basically just by deciding what would be useful to himself. RSS is very simple, and can't really be called anything other than a huge success. Somebody might come along and do the same thing with the standards needed for social graphs.
Anyway, in a week and a half I'm going to BlogTalk and Webcamp in Cork, Ireland, which will focus on these kinds of things, I'm sure, with smart people who've been much more involved in possible solutions than I have. So, hopefully I'll get a little wiser on where things are going.