webring++ is a logical extension of 1990s webring.

This spec provides a language-agnostic explanation of how to join a webring++.

In a traditional webring, websites in the collection form a linked list. Inserting into a linked list requires:

1. Creating a new node n with back pointer to n - 1 and forward pointer to n + 1
2. Adjusting n - 1's forward pointer to be n
3. Adjusting n + 1's back pointer to be n

In a webring++, there is no ring at all - we instead abstract the idea to a digraph. Each site in the webring++ points to zero or more other sites. As such, to join an existing webring++, you just need one of the other nodes to add your site to its forward set.

webring++ only specifies forward links. webring++ clients can choose how to read, render, walk, or otherwise ignore members of the digraph.

To host forward links in webring++, you must host an endpoint that serves JSON matching the schema indicated below. webring++ endpoints are scoped by hostport, which means that the following are all the same:

• a.com:8080/webring++.json
• https://a.com/webring++.json

The following are not valid, and are ignored:

• a.com/blog/webring++.json
• geocities.blog/blakeh/webring++.json

The following are all valid, but fall into different hostports and so do not collide (they exist in parallel):

• blog.coolwebshit.com/webring++.json
• forums.coolwebshit.com/webring++.json
• anything.else.really.coolwebshit.com/webring++.json

The top-level /webring++.json endpoint should serve JSON with the following schema:

Clients must permit deserialisation of unknown object keys for forward-compatibility.