A Messaging Proxy and Domain as Identity

A pretty good idea

So yesterday we were talking about the problem that people in social networks have: when you’re active in social networks you are less active outside of them. You become immersed in them, so that when you’re in MySpace the people outside of MySpace get less of your Attention. If all of your friends are in MySpace, then that’s where you hang out. I told the story of a guy I met who actually signed up on MySpace so that his daughter would receive his messages.

In the comments Cori Schlegel made the seemingly innocuous suggestion that we need a messaging proxy. Send a message to the proxy, and you get it on all of your devices or services that talk to your proxy.

This is a great idea! And the more that I thought about it, the more I realized that it is a perfect extension of an idea that I wrote about last year: domain as identity. (a post which, coincidentally enough, Cori commented on).

Here’s how it would work, as far as I understand it…

Instead of web sites having domain names, and those domains having mail accounts, people have domain names and one messaging account. My domain is Bokardo, and I have services at Bokardo.com that I control. Mail would be one of those services.

When mail is sent to mail.bokardo.com, it is forwarded to any devices or services I have added to my domain. So it acts as a proxy in this way…it serves as the place that all mail is sent to, and then I control where it goes after that.

The messaging devices and I have set up on my domain could be of various types:

  • Cellphone
  • Chat programs
  • Social networks
  • PDAs
  • Traditional email accounts
  • The display on your car dashboard

The difference is subtle. Instead of having a separate messaging service for each context we’re in, we have a single messaging service provided by our own domain that routes messages for us. If we join a new social network, we still use our messaging proxy to relay the messages. We simply point the social network to our domain and it knows about us. We would have a single archive of all the messages we send, with metadata that tells us what context they were sent in. So, if I want to say “thanks for the add” to a MySpace member, I send it through my messaging proxy to the messaging proxy of the MySpace member, suggesting that it be received within a MySpace context, and then the person receives it in their MySpace interface. If they aren’t in their MySpace context, they might receive it wherever they are on their cellphone.

In this setup you would never lose mail as along as you keep your domain. If you didn’t have any services set up to receive the mail, it would sit at your domain. Right now, one of the big pains with email is that they’re often provided by your ISP, so you have something like mary@comcast.net. That’s a tie-in we don’t want because it gives the wrong domain power over the account!

The identity folks out there are probably saying “duh”. But that’s part of the problem, isn’t it? Most of us haven’t yet realized what having a solid identity would mean. Take a look at MySpace again. People are fiercely protective of their accounts there, because they’ve invested the time and energy to fill them up with information about themselves. It’s their identity. Their messaging capability is centered around the service, and they can’t interact with folks outside the service easily. That’s the pain point where identity comes in. When a new, cooler hang out spot comes along, they’ll be gone, and all of their messages and profile information will be lost. Unless that information is stored in their identity domain…

Privacy advocates will recognize that this also has benefits for privacy. When messages can be tied to an identity, and we can hold someone accountable for them, SPAM plummets. The problem with SPAM is lack of identity, and if we can create a system where every message is tied to an identity then we can start the long uphill climb of getting rid of SPAM. At least some of it.

Attention-minded folks might see this idea as personal attention streams. Route messages through a single service, and you’ve got them all right there for picking. You’ve got a single address book comprised of everyone you’ve ever sent a message to, you know where you’ve spent your attention, and that could potentially be valuable information for oneself (and perhaps for others).

All the messages come from one mouth (i.e. one mind), so why not a service to model that?

So a domain per person. A single message routing mechanism per person. It’s an interesting idea that I would love to talk more about.

If I were Steve Gillmor, I would say that it has already happened.

Published: July 20th, 2006