Why I don't see Diaspora succeed
(and why I will close my diaspora account)

After a few weeks using Google+, I realized that everything currently done with G+ could be decentralized over a protocol using XMPP and HTTP: people in your circles would be people that you add to your roster. People following you are people that add you in their roster (no, it is not a symetric relation).

Want to send a message to a circle ? Each individual XMPP server will receive the message (and it's up to that server implementation to know what to do with it). Someone want to visit your private profile ? The informations are requested by XMPP and only those available to the given user are sent. Pictures could be shared the same way through private HTTP links.

It convinces me that, yes, a decentralized Social Network is possible and badly needed.

But it will not be Diaspora.

A decentralized social network is firstly a protocol. The user interface and the way people interact are a layer above that protocol. Each user interface can be different. GMail, Thunderbird, Roundcube or Evolution are all interfaces above the mail protocol. Pidgin, Empathy or Google Talk are all interfaces above the XMPP protocol.

When XMPP was developed, they ensured that you could exchange instant messages and see the online status of your contacts. Nothing fancy. Really simple. A server was implemented, immediately, a lot of clients started to bring that simple feature to the users.

Gradually, XMPP protocol evolved. Server became more complicated and a lot of innovation happened on the client side. Most of servers and clients developers were not involved in the protocol development.

Diaspora did it the opposite way. They started by developing an user interface (the client), put a server beneath it and then developing the protocol to allow two server to communicate.

They happily mixed the server and client. They reinvented their own protocol and they showed a complete lack of vision. It looks like Diaspora people are wonderful engineers. But they have no visionnary people in the team.

The result is simple: I've not seen yet any independant client for Diaspora. I did not find any competing implementation of their server. Since the first release, they didn't showed any improvement other than minor UI stuffs.

Worst, by adding support for private messages, they basically advertized that they don't even want to do anything else than a basic Facebook ripoff. They want to be to Facebook what status.net is to Twitter.

And if you call status.net a succes, you are living under a free rock, but nevertheless a rock.

Buddycloud and OneSocialWeb seems to be a bit ahead in term of vision. But they lack real implementation and community involvment.

I'm sure that we will have a real decentralized social network. It will be part the services offered by your email provider and, soon, we will take that for granted (just like we take email for granted and we will slowly consider normal for an email address to have support for XMPP).

But I'm also sure that this will not come from Diaspora. It was a great toy for freedom lovers. It teached us a lot of things. But it will fall with status.net and openid into the "never-reach-a-non-gnu-geek-audience" oblivion.

And if I'm not convinced by the long-term success of status.net, I've at least the satisfaction to make it useful now. There are lot of interesting discussions happening on it.

This is not even the case on Diaspora. Not enough people, discussions are mainly about Diaspora itself.

That's why I will close my Diaspora account soon. It is mainly a useless bookmark sitting in my browser.

Related stuffs:
Shared publiclyView activity