A webring with no server

A webring normally needs a tiny server to look up the current member and send a visitor to the next one. This site takes a different route: it generates every possible redirect as a static HTML file.

Each neighbour has four small pieces of data in src/_data/webring.js:

{
  slug: "your-name",
  title: "Your excellent website",
  description: "Notes on bread, buses and browsers.",
  url: "https://example.net/"
}

The build shuffles the array once, then creates pages such as:

/webring/your-name/previous/
/webring/your-name/next/

Those pages use an immediate HTML redirect and include an ordinary fallback link. JavaScript improves the hand-off but is not required.

Random enough

The order stays fixed for the lifetime of one deployment. On the next content update, the build shuffles it again. This is not request-by-request randomness, but it gives the neighbourhood an occasional reshuffle without a database, function or running process.

That constraint is useful: the ring can move from Cloudflare Pages to GitHub Pages, a basic VPS, object storage or a Raspberry Pi without changing its architecture.