What problem does Diaspora solve?

This is a rant and brainstorm. Feel free to contribute in either vein.

Diaspora is getting more attention lately, but it isn’t great. I’ve come across two basic feelings, one trying to drum up excitement for it being turned over to the community (which is something I am still trying to wrap my head around), and one that is reading the obituary for the project.

I don’t really care. Diaspora is a pain in the ass to set up, and even then there are large trust loopholes that can’t be filled in with the anything I’ve seen on the roadmap. I do care about the problem(s) Diaspora tries to solve, however.

We want a secure, trusted, decentralized and sometimes private social network. Well, I do. And if that is the problem to be solved, I don’t see it working in the model of federated pods that the project proposed.

I’ve written about how the great thing about StatusNet is the lack of trust required. When I apply that goal, lack of required trust, to Diaspora, then it seems to me that we need something like off-the-record messaging. I am a network operator, so you should be worried that I am reading your messages. I know how easy it is, that is partly why I run my own services. When I need privacy, I use OTR or public-key encryption. When I use StatusNet or Diaspora, I assume all my messages will be public-viewable (with particular exceptions that defeat federation, such as in-site private messaging).

The problem the web presents to us is the faux-client. E-mail and jabber have web interfaces, but they also have obvious non-web clients. Social networking is not built that way. So maybe we should build it that way.

Who else is working on this problem? What other solutions have folks come up with?