Interoperability for Shared Data Models
By Chelsea Barabas, Neha Narula and Ethan Zuckerman:
"Today most applications manage their own data, and have their own processes for handling authentication and access controls to their databases. As a result, it is challenging for users to switch between services that operate on similar content, or to migrate their data from one service to another. Moreover, application developers cannot easily build services that draw from a variety of data sources, because they are restricted to the information made accessible through specific platform Application Programming Interfaces (APIs). These proprietary APIs tend to operate on diverse data formats and closed protocols, which makes it challenging to repurpose data that is generated on a specific platform, creating a siloed effect. For example, Facebook stores the social profile data and interactions of its 1.7 billion users on its servers, and it isn’t available in a format users can export to another application, like Google Plus, Diaspora or Ello. Furthermore, platforms constrain access to their APIs, limiting the number of times per day they can be used, making it very difficult to maintain a social network using another company’s API.
This trend towards private data silos contributes to many of the risks to online speech we outline in this report. First, it makes it easy for governments to censor content, because they can pressure the corporations that house that content on their servers to comply with takedown requests. Even if users proactively replicate content that they fear might be taken down from a major platform like Facebook, there is no clear method for redirecting others to where that information is newly housed. For example, when the Thai government pressured Facebook to take down a satirical page from its site, the political commentators who ran the page had no clear methods for pointing their audience to an alternative location where that content could be found. More decentralized methods of data storage could make it difficult, if not practically impossible, for a single entity to censor content. For example, peer-to-peer file sharing services like Gnutella, Limewire, and Bittorrent made it almost impossible for the Recording Industry Association of America to stop the spread of illegal music downloads. In the following sections we will take a look at the InterPlanetary File System (IPFS), a project which aims to enable users to find content even if a specific website owner decides to take it down, or they opt to migrate it over to a new location (i.e. a different URL), and Solid, a framework for building applications which puts users in control of their data.
Data silos create lock-in effects, which make it challenging for users to switch between platforms or self-publish if they are excluded or unhappy with mainstream platforms, like Facebook or Twitter."
"Both IPFS and Solid have the potential to enable a more vibrant landscape of publishing options by supporting shared protocols and formats for handling data across applications. While these projects are not decentralized publishing platforms in and of themselves, they could serve as critical building blocks for more decentralized and interoperable publishing applications to be built." (http://dci.mit.edu/assets/papers/decentralized_web.pdf)
- the Inter-Planetary File System (IPFS) is an open source project that aims to
enable peer-to-peer methods for storing and disseminating information on the web.