[jstimpfle]

The difficult to decentralize parts:

- hashtags (big aggregators are needed that can answer hashtag queries). Similarly, proposing new tweeters to follow, etc.

- receiving messages. This needs more than RSS, which is only a pull mechanism. It needs both push and pull.

Then there are social problems.

I don't know how far diaspora and OStatus/GNU Social have come. I think diaspora is more concentrated on building a software platform than on protocols (which if so I think is the wrong approach).

OStatus, i.e. thinking of communication protocols, is the right way to go. However it's very very difficult to agree not only on a common protocol, but also on data formats (what's a tweet, what is an album, etc.)

[tschellenbach]

At the end of the day users just don't care about decentralization. They do care about:

* The network of users present on the app * Quality of the mobile and desktop clients * Ability to have success promoting their own voice/ content

Decentralization kills innovation in the underlying protocol. (see email for a nice example of common standards holding things back). It also creates a situation were nobody can effectively monetize the platform. Which makes it hard to create a competitive product.

Sure, open standards work well for many things, don't think it will ever catch on for social networks though.

[confounded]

Another thing that they care about way more than we recognize, is

"the latest trend among tech nerds"

Focusing on the Everyman is essential if your criteria for value is a consumer mindshare natural-monopoly (and an exit for VCs) within five years.

The next five years will probably have less of those companies than the last five, and (if the tone of newspapers are anything to go by) ordinary people are becoming distrustful of large monopolistic companies.

Now seems like a good time to hack on what we want. I can't think of a better time for a new, open, social network.

[jstimpfle]

I'm not sure if I want innovation in the underlying protocol. A protocol should be exactly that, underlying.

What ideas do you have for innovation? The only problem with email that I can think of is the lack of end-to-end encryption. But that's a social, not technical problem.