Friend-to-Friend Computing: Building the Social Web at the Internet Edges
The current Social Web is centralized. Large information silos store all the users’ profiles, their social links and much of the other personal data. In return for the reliable service the users allow their data and activities to be data mined by the service providers, which in this way increase their advertising revenue. As the social applications are storing increasingly more data and attracting more users, many questions about privacy, data ownership and data portability arise. In this paper we are going to critically assess the current state of the Social Web, identify several novel research problems and outline the possible solution: friend-tofriend computing (F2F). F2F is a completely decentralized architecture in which two computers can communicate only if their owners know one another. Constraining the connections to friends-only solves many of the security problems of the peer-to-peer architectures. We argue that a reliable social application platform can be built using F2F as the substrate. The platform gives the users much more control over their data than the current Social Web and ensures the level of privacy and security not possible in any centralized architecture. Groups can easily build their own ad-hoc networks and collaborate without the need for any servers or third-party services.
Record created on 2009-09-09, modified on 2016-08-08