Sigala’s architecture comprises information and communication.

Network Architecture

The basic architectural principles underlying our requirements include:

  • Scalable System – as a service it must easily cater for any number of users, from a handful to millions
  • Scalable for the user – the service must work well for any number from a handful to a few thousand connections
  • Sustainable – it needs to be easy to maintain and develop further

The kinds of requirements envisaged include:

  • Distributed operation
  • Open data – avoidance of ‘lock-in’, easy import/export of data for users, but not at the expense of privacy
  • Use of open standards to support open operation
  • Open protocols and APIs for third party developers (supporting e.g. OpenSocial) so that the SNS can be used from any kind of device, particularly mobile devices.
  • Online and offline modes of working – inspiration is drawn from Colloquia, a peer-to-peer computer-mediated learning tool developed (in Java) by Oleg Liber and Sandy Britain, http://www.reload.ac.uk/colloquia.html . The system is distinguished by the fact that activities are carried out offline, and communication uses only email protocols for the transport.
  • Multimodal Internationalisation – internationalisation (i18n) is commonly a 1-1 dictionary mapping. Many social networking sites invite you to learn how to say “hello” in different languages. However, words alone are shallow indicators of cultural practices. Greetings in some cultures involve multiple exchanges for “hello”; the sequence and rhythm of the exchanges may determine the quality of the greeting more than the actual words. Hence internationalisation should be organic, supporting different kinds of 1/many- 1/many mapping of words and gestures. 
    Internationalisation will be supported by assets, such as locally determined calendars.

It might be argued that each person has full control over their own social graph. It should be up to them how and to what extent they make this available. However, technically, there are challenges over whether a P2P distributed service can perform well enough1.

Data Ownership, Security and Privacy

A core principle of Sigala is having genuine ownership and practical control of one’s data.

  • All data should be made available to browse offline. An inspiration is the e-learning system called Colloquia, http://www.colloquia.net/ (archived), which was designed to support group working where Internet access was limited and intermittent. Content in the system was packaged up and sent by e-mail. There should ideally be facilities to manage the content, reflect on it and then go online to communicate as a deliberate cognitive step.
  • Data security: data should be secured in a way that prevents especially unauthorised access by (or leaks to) third parties. Most issues that arise do so in connection with the interactions of the host system with third party services, typically via apps, where data is not cleanly separated.

    We are particularly interested in whether we can utilise a decentralised platform, called Solid (Social Linked Data). Initially developed at MIT, it has now matured to the start-up phase in the Inrupt project. It’s led by Sir Tim Berners-Lee, who has developed a solution that gives control to data owners, ensuring clear separation of data and application, and also allowing the owner to choose where the data is stored.2

    “Solid empowers users and organizations to separate their data from the applications that use it. It allows people to look at the same data with different apps at the same time. It opens brand new avenues for creativity, problem-solving, and commerce. Learn how it came to be.” 3

Data Navigation

For Sigala we consider movement in two dimensions – vertically, in relation to how close is the relationship and horizontally, in terms of degree of connection (friend of friend etc).

In terms of SNS data, semantic zoom can be deployed in various ways:

  • At the top level it can present your closest connections – the nearest and dearest kinship and non-kinship relationships. For most people they are relatively few in number.
  • At any given level people are the central foci and data is associated accordingly, in some sense being contingent attributes, i.e. their activities, interests, memberships. So navigation is fundamentally from person to person.
  • At any given level there are horizontal connections between the relationship types, which in navigation terms means you can move horizontally to explore heterogeneous networks.

Client Support

This SNS should be easy to use anywhere.

  • Cross-platform, designed for emerging technologies, especially personally owned handheld devices such as smartphones and tablet computers as well as tablets, laptops, and desktops
  • In a similar way to the range of SNS clients currently available, it would be likely that Sigala would be variously coded according to Web standards, usually those of the World Wide Web Consortium (W3C), including HTML5, CSS, client-side JavaScript, using the most popular libraries such as React JS. Alternatively, they might be realized as native apps.
  • Where practical, they would be required to support responsive design.
  • Aware of latest interfaces, especially multi-touch and contactless gestures.

These elements of networking and data underly proposed services.
 

Notes

1 Consider e.g. YaCy, http://www.yacy.net/, as reported by the BBC News http://www.bbc.co.uk/news/technology-15935550

2 It does not use blockchain, which is another decentralized approach to data.
“For blockchain to work with Solid, it would have to be interoperable and offer multi-chain support.”
(https://bitcoinexchangeguide.com/internet-inventor-tim-berners-lee-to-launch-inrupt-decentralized-web-solution/)

3 It also has a thoughtful code of conduct.
https://github.com/solid/community/blob/master/code-of-conduct.md