I’ve been giving some thought to use cases for social-networking software. Friendster really is the most useless functionality-wise if you’re looking for people you [used to] know irl, but well, even the best (tribe.net right now, I think) are still limited.

Part of the problem is that these services are designed for ‘meeting new people.’ That’s fine and all, but I suspect that for a lot of users, the keeping track of people you know part of is a lot of the draw as well; most of these applications fail miserably at that task.

  • searches – should return all the personal information that a person wants to allow the public to see on searches (ie, there are a lot of John Doe’s; the Doe’s should be able decide if they want to be found or not, but if they want to be you should be able to try to narrow down to the one you want to find by more than a 10px thumbnail)
  • privacy control – related to the above, there needs to be a way not only to specify your searchability on the network, but also of controlling relationship/info propagation
  • groups – basically the ‘tribes’ idea, however, there needs to be a better way of filtering/organizing/subsetting these things
  • creating agents to regularly monitor based on criteria (ie, keeping a lookout for high school buddies)
  • notes – personal tagging/organization of friend’s profiles

Note, that the current centralized client/server model wouldn’t necessarily be the ideal infrastructure for this kind of application. (yeah, I still need to think of a project codename)