I’ve read several books on social network analysis recently, like Networks, Crowds, and Markets: Reasoning About a Highly Connected World or Networked: The New Social Operating System. When looking at the architecture of these networks, one fact is striking: crucial information is kept locally at the networked nodes. For example, people have their personal contact lists which they store on their smartphones or other mobile devices.
However, with vendors pushing people into the cloud, this kind of data is moved to centralized servers. The vendors claim that this kind of data management is very comfortable because one can access personal data from “everywhere” and share data within the cloud. This is well exemplified by Apple’s iCloud: users have seamless access to their data from several (Apple) devices. But, and that is something that is hidden behind clever marketing, (almost) all cloud systems are closed. Once you move your data into the cloud, it is very difficult to move the data to a different platform. And, this is even more striking, it is difficult to share your data with people that are not in the cloud. Granted, Apple’s iCloud supports the sharing user data (for example pictures) with the public (photostream), but there is no mechanism to restrict access to user data for non-cloud members. A perfect example for vendor lock-in.
Ironically, Apple advertised their Macs several years ago as examples for being able to connect to different systems and to transfer data between different platforms:
I was wondering, if there was an alternative for such an centralized cloud approach, especially in the light of (1) user privacy, (2) data sharing with other users that are not in the cloud and (3) sharing data between different platforms. Basically, this it what I consider the essence of networking: empowering the user to control the “whereabouts” of personal data and being able to transfer data between different platforms.
I believe, that we can learn to build such a platform by looking at social networks and study how they work. As mentioned in the introduction, in social networks information is physically close to the owners: there is no central authority that holds and controls all the user data. This is in contrast to a social network perspective on cloud platforms. A cloud is a like a giant node in the and everybody in the network connects to the giant node and shares all data with the giant node.
In order to make the cloud more like a social network (in terms of data ownership and privacy), we need to move the data away from the giant node, back to the small nodes. However, there are two main issues to resolve. First of all, we require a global, anonymous unique identifier to handle user data. This ensures privacy: there is no giant node which knows who you are. Only those who should access your data will be able to resolve the your global identifier. Second of all, we need a place where to (temporarily) store the shared data, without keeping it. This requires a deletion mechanism (similar to a tuple space), which removes the shared data after it has been downloaded by the receivers. Thus, the cloud acts more like a web based clipboard where user can exchange data anonymously and in private.
We, at ikangai are currently working on a prototype that will allow us to set up a cloud system where users can anonymously share their data with everybody – also with users that are not in the cloud .
your ikangai development team