Concepts

Version: v1.0.0-ea.1
Updated: 7 May 2019

Introduction

The Epiarc is a Web consensus system for discovering how we name things.

It is made of:

  • ARIs that name things.
  • Linkbases hosted on the Web.
  • User Agents that subscribe to linkbases.
  • Rules for ARI resolution that User Agents follow.
  • People - the ‘epiarchitects’ who build and use it.

The Epiarc is built on the Web, which makes it:

  • Decentralised. Anyone can make a linkbase. Anyone can share a linkbase.
  • Emergent. Social dynamics and usage patterns emerge from the ground up. There is no central authority or master architect.
  • Contextual. The Epiarc allows users to operate in their own context, in a way that is natural to them.
  • Fault tolerant. Linkbase representations can be cached in HTTP proxies at Web scale.

Private Epiarcs may emerge in private networks with no Internet access.

Archetypal Resource Identifier

The Archetypal Resource Identifier (ARI) is the principal identifier of a resource or link on the Epiarc. It is the most common natural name for something — with some extra features.

  • “mark twain”
  • “eiffel tower”

The ARI is usually expressed as a link, pointing to a URI, for example.

The ARI may be hierarchical, so that producers can organise their links underneath another link. This is supported by concatenating a parent ARI and a child ARI with the forward slash character: “Parent ARI / Child ARI”.

The ARI is:

  • Human-friendly. It is invented by humans for humans.
  • Emergent. Invented from the bottom-up, and need not conform to a top-down grand design.
  • Contextual. Its surrounding social and cultural context gives it meaning. An ARI is archetypal for its context.

A good ARI is also memorable, obvious, and concise — but not so compressed that it invents a new naming scheme.

The ARI does not need a formal structure when spoken aloud. Computers can serialize an ARI in many formats.

The ARI is a more natural and capable name format than its predecessors like URIs, domain names, or hashtags.

Linkbase

The linkbase is the principal repository of ARIs on the Epiarc. It is simply a data structure, or a representation of one, that contains one or more ARIs.

The personal linkbase is the user’s own linkbase. It is the most local and contextual linkbase, as it is formed and developed by the user.

An Epiarc User Agent can subscribe to shared linkbases. The user may assign a name to a linkbase when they subscribe to it. The linkbase name can be used as the parent ARI for child ARIs inside it.

“genbank” (NCBI Genbank linkbase):

“wiki” (Wikipedia linkbase):

When a User Agent subscribes to shared linkbases, the entries in each linkbase are archetypal for their context, and can be resolved as hierarchical ARIs.

Global linkbases may emerge over time to answer more generalized queries.

Computers can serialize a linkbase in many formats.

User Agent

The Epiarc User Agent is the client software that interacts with the Epiarc on behalf of a user.

A User Agent can subscribe to and manage linkbases.

A User Agent can perform ARI resolution (ARI → URI).

A User Agent can perform reverse resolution (URI → ARIs).

Rules

A User Agent follows rules to resolve an ARI in context.

Order of precedence

  1. Personal linkbase.
  2. Shared linkbases.

If a User Agent shows a more expansive list of results, it may have its own rules to rank and tie-break results from multiple subscribed linkbases.

Uniqueness

  • An ARI is unique in a linkbase.
  • A linkbase name is unique in a User Agent.
Cookies
Epiarc.org would like to place cookies to understand how our website is used, in accordance with our cookie policy.
×