FED. SOC. WEB SUM.

Session Topic: Federated Social Web Summit

Thursday 3J

Convener: Evan Prodromou

Notes-taker(s): Markus Sabadello

Goal: Provide social web functionality without dependency on a single system. In other words, have a topology like e-mail.

Functions: identity, profile information, social graph (friends, etc.), content (blog, documents, rich media), activity stream, publication and subscription model.

Technologies: FOAF, XFN, RSS/Atom/...,

Projects: StatusNet, Diaspora, Appleseed, tent.io

Current mentality: Build one piece of software that can do everything, with its own protocol, extensions, etc. This approach has resulted in some disappointment.

Better approach? Agree on an extensible protocol supported by multiple implementations. This was the idea behind OStatus (= a suite of protocols: OStatus, Webfinger, ActivityStreams,

PubSubHubbub, Salmon). This is different from earlier "monolithic" approaches.

OStatus has been successful and is widely supported, e.g. by Wordpress.

Problems of current OStatus?

• E.g. PubSubHubbub doesn't support private feeds.
• Immediate spam on publicly hosted StatusNet instances.
• Onboarding: How do I get my existing social graph from e.g. Facebook into my StatusNet instance?
• Operations requiring a "global view" on the system, e.g. monitoring a global hashtag.
• Make it look better for the user.

StatusNet and Diaspora are AGPL: Nice for free software, but less nice for people who want to make incremental improvements while protecting intellectual property.

Important feature for broader adoption: Introduce a notion of groups on the protocol level, which is independent of a specific implementation.

Goal should be not just to build a replacement for Facebook, but to build something that can do more. The Federated Social Web may look different from what we currently think of as social networking. E.g. just like blogs evolved into social networking, a federated social web may be quite different from "traditional" social networking.

Advice: Use decentralized software, e.g. publish your social content on Wordpress, support existing projects, keep experimenting.

POSSI: Publish, Own, Site, Syndicate, Everywhere

What to do to improve current shortcomings of OStatus?

Work on a new version? Consider XMPP?

Why don't Facebook, Twitter, Google+, LinkedIn talk to each other?

Incentive only for smaller communities to join a bigger internetwork, but not for big players to cooperate with each other.

So, the strategy for the Federated Social Web should be to build on small communities that want to grow into a bigger internetwork.