Marc's Voice

Outlines as a structure for distributed friending

One key aspect that was missed in Marshall Kirkpatrick’s other wise excellent coverage of my post “DiSO Dashboard Outline proposal“- was the importance of the Outline format I was proposing.

In fairness to Marshall - this is a pretty geeky, conceptual idea which I doubt a non-technical person would completely grok - at the outset.  I got lots of great feedback from folks yesterday (and we made it onto Techmeme) but I’m afraid that folks still don’t get the essence of what the proposal was all about.

1. I ASSUME that dashboards will be everywhere.  So most of the comments I got - were agreeing with and embracing that notion.  OK - coolio.

2. These dashboards take on lots of states.

Some of them look like a widget dashboard (NetVibes, iGoogle) and some of them are called dashboards (Wordpress) but they’re really control panels on some web service.  Some dashboards are a piece of a large suite of on-line apps or services (MyYahoo and whatever the hell Microsoft calls theirs.)  But technically - all these dashboards are all the same.  THAT’s what I was trying to say yesterday - that we can embrace the notion of a dashboard - as being a common user experience as a touchpoint for standardization.

3. Dashboards have been around forever - and in fact we developed a UI dashboard for our MediaBar operating system back in 1996-97.  Check out this Marc Canter Show demo which shows a bit of that dashboard.

4.  What yesterday’s post was REALLY about was something that I’ve been working on the past few months - a “pro-active contribution” I’d like to make to the DiSO effort.

DiSO stands for distributed social networking - and they’ve been gaining traction in the area of a standard set of ‘markup’ for activity streams.  But it’s more than the markup - it’s a (triple-like) combo of a verbs and object types - that can map to all the popular activity streams, coming out of MySpace, Facebook, FriendFeed, SixApart, etc.

So what I’m doing is going the next step.  I really think that if we all agree that the ‘dashboard’ metaphor can map to all systems, then we can then develop a standard kind of outline format’ - which would represent:

- the user’s State and Presence

- the user’s Profile data and Social info (social graph, bookmarks, my groups and networks, etc.)

- the user’s Content and Media - in all the forms that takes - both the payloads and pointers to where this content resides

- Feeds and Access privileges over those feeds (and over all their content and media - as well)

- and all the technical Configuration info and list of Modules, Widgets and Code that exists IN these dashboards

5.  This outline format is the lynchpin in this proposal.  By having such a beast, I believe we’ll be able to take distributed social networking to the next logical state.

Its a strange beast we’re enabling - as what we DON’T want is for Google, Facebook, Microsoft or anybody else to come in and ‘take over’ or ‘be in first leader’ position on any of this data.

So though we love Google Friend Connect - no we’re NOT happy that all the data resides on Google’s servers. And though we all LOVE Facebook - we’d like to control our own list of friends - thank you very much.

6.  So here’s one key thing that this Outline format will enable.  It’s called distributed friending.

Imagine Google’s Friend Connect - but instead of Google storing the data, it’s stored twice - once on my blog and once on my friend’s blog.

One of the challenges of doing distributed friending is that one site’s friend is another’s contact or another’s follower.  So how can we ‘normalize’ all of these different kind of relationships in one distributed system?

How can we connect any social site to another and create a master list of friends (social graph) which I control?

I really think that we can intermix reciprocated and non-reciprocated relationships, following, subscribing, contacts and any other kind of relationship type - if we do this right.  Even though the ‘buttons’ are named differently, the underlying technology is basically the same.

And as long as each dashboard stores it’s own set of relationships in it’s outline - across all these different social systems - well then the Outline is the normalizing factor.

7.  Not only can this Outline format enable distributed friending, but it can also simplify and enable inter-operability, dataportability and shared data between systems.

8. Another coolio thing I can just taste - is aggregated Groups.  I’m looking to connect all the Reggae groups, on all these different message boards, social networks, FriendFeed rooms, Flickr groups, etc. together.  To somehow glue together groups which reside in different systems and enable some sort of shared experience ACROSS these disparate systems.  I’m convinced a shared outline can do that!

9. And if I can aggregate Groups - I should be able to aggregate all my social graphs and create a master social graph.  Right?

10. The trick to all this functionality - is not what one vendor can do by themselves - but what a BUNCH of vendors can do - without really knowing who each other are.

So another purpose of the dashboard outline is discoverability.  One system has to discover what the other system is storing, what it understands and what it can do - into and out of their local data stores.

Luckily for us - that’s what XRDS is all about - so we just need to stuff some XRDS info into each dashboard’s outline.

So in closing - it’s not about finding an All-in-One Social Network.

That’s what I originally thought we’d need - five years ago.  What it’s about NOW is realizing that a dashboard metaphor is a common notion that can connect social networks, blogging environments, media repositories, ecommerce accounts, email systems, messaging and communication portals, start pages and everything else - together.

Each dashboard would have an outline - which would be the mechanism for connecting and normalizing.

Whether you be on a giant centralized social network like Facebook or MySpace, or vertical ones like a Ning Diabetes network or a business oriented networking site - like LinkedIn or Xing - this is NOT about a dashboard for managing these social networks.

This is about a dashboard which can ricohet, route and store shared distributed information - which THEN gets used for interoperability, dataportability, distributed friending, etc.

eg. people STAY on their existing social networks and blog tools and we connect these systems together.

What we can assume is that user’s will NOT WANT to pick and choose between networks, but try and maintain their presence IN MORE THAN ONE - at the same time. Some vendors will try and provide that capability and that’s how power.com got into trouble.

Each meta-aggregator platform out there (SocialThing, Profolactic, Power, etc.) will have their own approach to aggregating social networks.  And FriendFeed took a unique approach by focusing on JUST the pings - JUST the activity, while ignoring any data store or direct messaging. Its fascinating to me - to watch and see all the variations on a theme - which are constantly appearing.

But what I’m saying is (quoting Rodney King here) “can’t we all live together?”

There isn’t going to be ONE winner.

There won’t be one winner in each region (Russia, Adu Dabi, Singapore, Hong Kong) or market segment (business, health, sports) and there won’t be one winner in the mobile, gaming or home living room.

So we’ve GOT to architect our distributed standards to assume that dashboards can be a common notion and that an outline format associated with each user’s dashboard can be a mechanism to facilitate all sorts of distributed open ‘goodness’.

That’s the essence of the proposal and why I wanted to make sure everyone was focused on the outline format I’m proposing.  Yes - absolutely Dashboards are our future.  But we knew that back in 1996 - when we built the Mediabar.

What we NOW know is that the DiSO effort is achieving uptake and can be a great vehicle for developing a standard dashboard outline format, distributed friending and a whole lot more!