Friday, May 14, 2010

Diaspora: Kill the Suspense!

With the Facebook debacle in full swing, people are quick on their feet to get their data out of the media giant and into a more viable solution. That's where Diaspora comes in. Diaspora, created by four students from NYU, is a completely decentralized social networking interface that uses direct peer-to-peer communication in order to establish a grid of sharing among friends. The project sports complete GPG encryption, "full-fledged" seed-to-seed communication, and will soon support VoIP, UDP integration, and Instant Messaging. For the worried Facebook user, this sounds like the panacea of privacy issues. However, there exist a number of conceptual flaws with the project, or at least with the direction I think the project is heading in, since they barely give you any information.

Diaspora is a good idea, good enough that I even backed it with $25 on Kickstarter. The primary problem is that they reveal little information about the project and what will be in it, which is very unusual considering it's an open source project. So I just wanted to assert that the opinions in this blog post are based on the best conceptualization of the project that I could derive from the current given information.

For starters, the evils of complete centralization do not imply that complete decentralization is any better. The reason everybody appears so gung-ho about the project is because Facebook essentially killed the idea of centralized social networking, and users were quick to jump to decentralized so that their data would be protected. In fact, even though I personally advocate a more decentralized social network, the path that Diaspora is taking is completely off target. They want each and every person to operate their own node in the network, and then have each node connect to other appropriate nodes to create the social network. Where is the flaw in this situation? Well, for starters, how exactly do they plan to implement a "five-minute setup" if establishing a Diaspora node involves setting up your computer as essentially a server and navigating through the forest of firewall protection into the open Internet, all without sacrificing security. Furthermore, they want to implement UDP integration, which is asking for even more trouble considering the unreliable and unorganized nature of the protocol.

The beauty of centralized social networking is that it allowed users to easily verify who they were speaking with and have one way in which to connect with that centralized server. For Facebook, you simply go to your web browser and HTTPS takes care of the rest. I am sure you could come back and say that GPG encryption will protect everybody, but exactly how many average social networking users are going to go through the trouble of properly signing all of their friends' certificates, and then ensuring that everybody else in the network gets the new signed certificate. You need some sort of central server to host everybody's certificates, and I doubt that an easy-to-use social networking service is going to use the normal channels of publishing public keys to a completely public keyserver.

Another disadvantage of decentralized social networking is the ability to effectively collaborate. Google spent months developing Google Wave because they knew that the decentralized model of email ultimately fails when more than two people are communicating. This same concept applies to social networking. Furthermore, their explanation of the project say that Diaspora will connect your seed with other existing social networking services, suck as Flickr and Facebook. By this point I am just so confused as to what the project actually does that I almost regret backing it. They employ full GPG encryption and direct, decentralized seeds in a network, but then hook that network right back up to the old centralized services that Diaspora was specifically created to improve upon. Seriously, the absolute only thing I could foresee using Diaspora for is to encourage cooperation between my various existing services. Decentralization and integration are important, but Diaspora takes it a little too far.

Unfortunately, I cannot say for sure what Diaspora will turn out to be, because I really don't know much. If anything, I would recommend that the founders release something for us to actually look at and say, "Oh, so this is what Diaspora is", because as of now my visions of Diaspora's future are foggy at best.


  1. There is mention on their Kickstarter page (and I have read it somewhere else, just can't find the URL at present) about the hosted solution that will be similar in concept to how is related to word press the software. You could run your own instance, or just host it centrally.

  2. hi i am mahmoud

    i want to speack about tourism

    in world we have many pleac to tourism

    if u want know some places and see some photot to it

    in her

  3. Has a line of code even been written yet?

  4. It's quite possible they have not even done anything. That's why I wish they would just let us see something rather than building up all this suspense.