I love your ideas behind a user-centric web, but I have a feeling you will be waiting a very long time for it. First, I think it is too drastic a change for companies to accept quickly. We will get baby steps along the way, but it will be very incremental. The other side of it is that web sites always want to be THE place where you get your information. How else can they build a business model if they have no traffic? This is probably going to be even harder to get past due to the way most companies have always worked. It is not intuitive that you can make a successful web business by allowing users to leave your pages.

Interoperability has always been the holy grail of most software development. Typically it is very difficult and takes a long time to do correctly. With things like OpenSocial and more open APIs, we may be closer to having all of these “sub-graphs” talk to each other better.

That kind of response, and implementation, are why people love friendfeed. They do absolutely rock.

The user s the destination, exactly my thoughts (although I tend to use User-Centric Web).
I do realize a strength of the web lies in it’s nature, nodes connected by various paths.
I have several problems with current destination thinking. It obviously leads to walled gardens that force “api” development to open up. We duplicate every building block needed to provide a user service, simply because our business models force us to have our own walled garden in the web to make money. My biggest concern with destination thinking is that it distracts from the user. It focuses on the wrong parameters. The network is more important than the individual user.
In User Centric, or User Driven thinking, a lot of these issues could be resolved quite naturally. Privacy would not be implemented within a service, but is part of the user traveling the web. Instead of building walls, a service would truly become a service point. competition would be based upon user value, instead of network growth. And an interesting thought comes to mind. If important user data is part of the user, instead of part of a service, things like interoperability would be solved quite naturally.
Think of the user as a traveler, and a service provider as a restaurant or a gas station. No need for all that complexity. Provide good service and you will have customers. The interface is simple and the traveler can disclose anything he feels necessary to get an optimal service.

You now get FriendFeed direct messages (and replies to direct messages) via email. You can also comment simply by replying via email.

Interesting development. And Bret promises more to come.

Diversity is a given with the idea of an internet, every niche that can be filled will be filled and that is not bad. For example mode of communication I use at work will be different from what I use for friends. It should not be mandated that I use a single system for all my communication needs!

Interoperability is a noble goal but somewhat wishful if history is any indication. What is feasible is for these sub-graphs(Say Facebook) to put systems in place to open and reach out to other sub-graphs.

Excellent post btw. My understanding is a little clearer after reading and commenting..off to friendfeed now!