As with all paradigm shifts it will be difficult for the bigger business to keep up, but the opening of the networks offers huge benefits to consumers.

Excelent I was afraid I had scared off conversation given such pointed questions. The answers to this problem I’ve seen thus far have all been of this sort. That is relatively static technology, or aproaching this problem as though we don’t already have the tools to solve it (this often arives as a part of the static technologies that have been proposed).

Any now for the enivitable pitch… my own solution involved repurposing jabber. And I would love to hear what you have to say about the idea, given that you have immenint experience in working with and designing socialy networked web spaces.

Basically I’ve written most of a web service that allows users to authenticate to their jabber server of choice and use the available tools from xmpp to manage their trust relationships (friends) on that server. What I like about this solution is that

1) it’s rather decentralized, the information you want can live on any server, and is easily retrievable from the server that you do choose,

2) it allows social networking sites to host the service on their own servers, and tie into it as they see fit, what I’ve created is simply an abstraction on top of the already available jabber libraries, given a stable set of api’s one could easily recreate this system in any language.

3) it offers bigger name social networking sites something new and useful. Primarily it gives them chat and presence. Where as before they were force to write their own messaging layer, or presence layer, this is all gotten for free from jabber.

4) because jabber is decentralized the networks can use the ease of federation to join themselves to the greater jabber community easily.

SO I think we both agree… that this is not something that by and large benefits the large or already established social networks, and that what we are proposing is a paradigm shit away from the value and comoditization of the network to the value of the tools and services offered. Which given a bit of free market theory seems to offer up much more competition by lowering the barrier to exit and entry (which would be good for the market).

I would love to hear your thoughts on this (more info at my domain).

Thanks,
~ Anders

I can’t speak for other websites, but young sites like Pownce need all the help they can get connecting new users with their friends. Identity consolidation really helps new sites. As I’ve mentioned before, “friending” takes up a lot of time because it’s valuable to users. If I can help people friend each other more quickly, all the better for me.

It seems like there will be an effort very soon to gather public friendship data in one location (a non-profit org?) and make it available to site developers. I’m not sure that this is ownership as you’d like it, but I’m guessing that people wanting to store this data on their own machines is an edge case.

As for MySpace and Facebook, I hope they realize that their value is their UI and their services, not their “graph”. We’ll see.

http://computersarehard.blogspot.com/2007/07/opensocialnetwork.html

So what I like about this solution is that it’s relatively simple, and that it doesn’t rely on screen scraping like some of the other solutions I’ve seen.

Given that you’re offering the libraries that you created for free, how do you see the ecosystem of these kinds of “open social graphing” tools interoperating?

That is, given that from what you’re offering we can collect our social data into one place (good), request that data in a web friends format easily (also good), how do we deal with the possibility of many of these types of services existing? And moving between them?

Do we care at that point that our data still lives on a single service that provides for portability across networks but that our data still lives on a server that is out of our control. How do service providers like Myspace or Facebook deal with a possibility of a wealth of new “network” providers (services like what pownce is offering that collect our networks to a singular location).

The data Pownce collects about user’s other social networks are only “claims”, meaning that they are unverified user assertions about their other profile URLs. I see 2 ways that these URL claims may be verified.

1) Via another web application’s API - the user could go from Pownce and sign in to their other web service via the API (eg Flickr Auth). Pownce doesn’t do this yet as it is a UI issue to have user’s sign in to another webservice via their API.

2) Mutual claims - a Pownce user (x) claims to have a URL on Digg (y) and Digg user (y) claims to have a Pownce URL (x). If both Pownce and Digg require the user to log in before claiming URLs, it’s safe enough to assume x and y are the same person.

About the social graph format -

Each link for a claimed URL has XFN rel=”me” and I hope it’s also parsable by Plaxo’s Online Identity Consolidator (http://www.plaxo.com/info/opensocialgraph).

In addition, Pownce (and Django-PSN) provides a JSON response with the user’s claimed URLs and friend/fan relationships (represented as user ids for graph edges in and edges out). This can be found at /interface/elsewhere_info with params id=user_id or indent=username. You can try this out with me at Pownce: http://pownce.com/interface/elsewhere_info?ident=leahculver

I copied this format from LiveJournal, guessing that it will be copied by other websites as well. I’m open to changing this format and updating Django-PSN accordingly.