webring

How the ring decides where to send you

The links above are not unique to this site. Every member ships the same four URLs. When you click /next, the ring reads the Referer header, finds the site you came from in its member list, and redirects you to the neighbour on that side. That is the entire mechanism. There is no account id in the markup, no script, and nothing to keep in sync.

The tradeoff is that the ring has to verify membership some other way, so it re-crawls every member's home page on a schedule and looks for those hrefs. A site that drops them is marked inactive and quietly falls out of the rotation. That is why the links sit on the front page of this site and not only on this one.

How the ring knows this domain is mine

Joining requires signing in as dotmavriq.life rather than as an account on somebody's server. This site does not run an IndieAuth endpoint, so it uses RelMeAuth instead: the home page links to a GitHub profile with rel="me", and that profile lists this domain in its website field. Two claims pointing at each other are the proof. The ring follows both directions and accepts the domain as the identity.

The same two links are what the ring reads to build a directory entry. It parses an h-cardoff the home page for a name, a photo and a short bio, which is why the markup there carries one.

Where else to go

The directory lists every active member with a profile, and random drops you somewhere in the ring without walking it. Both are better than the arrows if you are browsing rather than passing through.

Webrings came back because the thing they solve came back. Discovery on the open web is again a problem of finding people rather than ranking pages, and a ring solves it the cheapest way available: every member vouches for two neighbours, and the graph holds itself up without anyone running an index.