I've thought about naming, files, hierarchies and the web a lot, and that's why names (ARIs*) on the
@epiarc:
- are archetypal resource identifiers
- are human centric
- are user chosen
- are personal, folksonomical
- are archetypal (the most suitable name)
- reflect the names people use in real life (Facebook not facebookā¢com; Metro Bank not metrobankonlineā¢coā¢uk)
- are consensus driven
- are eventually consistent (the archetypal names rise to the surface over time)
- are hierarchical
- are unique per context
- can be globally unique
On the Epiarc, here are some example global-context ARIs which map to URLs. Imagine seeing these in your browser address bar:
the BBC is "bbc"
BBC News is "bbc/news" or "bbc news"
American Broadcasting Company is "abc" or "abc (us)"
Australian Broadcasting Corporation is "australian broadcasting corporation" or "abc (australia)"
In an Australian context (linkbase) you have the above, with higher priority names like:
Australian Broadcasting Corporation is "abc" or "abc (australia)"
American Broadcasting Company is "american broadcasting company" or "abc (us)"
ARIs are a layer of abstraction for URLs, more friendly than domains and paths, and more resilient to linkrot. They're more secure and phishing resistant than URLs and domains, because they are names that people use and know in their daily life.
Users subscribe to linkbases in their browser or OS, and can make their own linkbases for naming things how they like. You can publish your linkbase under a /.well-known epiarc.json file and others can use them.
@epiapp produces linkbases for global and local contexts, and you can access them in our search engine and link checker.