Dan Brickley’s comment on an earlier post reminds me that i’ve never even introduced the FOAF (Friend of a Friend) Project. Part of this is because i’ve heard so many conflicting stories about what FOAF is and what FOAF is not that i feel too naive to properly address it.
To the best of my understanding, FOAF is a document framework that would allow you to articulate categories of and relations between people and information.
As Dan noted in his comment, FOAF poorly addresses the nuances of relationships between people. Because they couldn’t determine how to weight or contextualize the relations between people, they went with a standard “knows” relationship (similar to Friendster/Ryze, etc.). Of course, this is certainly the easiest approach, but it fails to truly capitalize on the value of nuanced relationships.
Dan continues on to remind us that FOAF is a distributed document type that has both its strengths and disadvantages. This is where i see FOAF as *potentially* having a great advantage. As a distributed tool, FOAF can easily be connected to one’s functional behavior and not simply rely on their strength in articulating their social network. What i mean is that if one is able to maintain the actual document of their relationships, it could easily be generated by their email/IM/SMS/phone behavior.
People know how to contextualize their relationships to others; they do so regularly through their behavior. The biggest weakness in Friendster is that it expects people to articulate their relationships. People are *dreadful* at this. So much of relationships is about showing face, about having relationships that are purely for mutual advantage and have nothing to do with liking one another. Professional friendships out of need. Yet to say that these relations are equivalent to that of one’s lover or one’s mother or one’s actual friends is utterly foolish. To suggest that people want to connect all of their friends with all of their other friends is naive.
My hope is that the FOAF folks will figure out how to move away from articulated social networks and emerge as a tool for people to evolve their networks through behavior. Of course, this will also require a level of privacy and non-publicness that is not currently in any of the FOAF specs i’ve seen or heard about.