Michael Taggart recently announced a new project he is working on to revive the Web 1.0 concept of webrings.
Ringspace is not an AT-based protocol, but it’s trying to solve for a lot of the same problems:
A modernized form of webrings is central to our vision for Roomy, and I wrote about it nearly two years ago (back when we were still building on Matrix):
The notion of a universal convention for personal web profiles also ties in here:
This is the essence of Ringspace:
Ringspace takes the concept of a webring—a small community of creators of like mind connecting their sites—and adds to it a trust layer providing two assurances:
Identity: “Are you who you claim to be?”
Reputation: “Are you known to act in good faith?”
Ringspace achieves this through a hybrid model of decentralized site “manifests” and a central coordination server that performs validation, updates member sites on changes in the ring, and also handles trust and safety concerns for the ring.
This sounds a whole lot like what AT already does, especially when coupled with additional utilities like endorsements /attestations by @ngerakines.me .
I will invite Michael in here to explore our common ground.
I had been talking to @wesleyfinck.org@ronentk.me about claiming websites (URLs?) as part of Semble, and in general things like Open Graph and some of the author stuff.
I need to go look at the Mastodon article author stuff and generally pull it all together in one spot.
Hi all! And thank you @erlend.sh for the invitation. I’m actually a big fan of ATProto and am excited to see what’s happening in the space.
This project was kicked off after noodling on what I wrote here. In that post, I actually call out DIDs as one of the technologies I am most excited by! That said, Ringspace was designed to function without control of a domain. Instead, the expectation is that the ring member has control of the web root of their site.
Ringspace was designed in part as a mechanism to avoid the sloppification of the internet. The idea being that a community of content creators could band together to maintain a set of community norms. Those norms could be “only use human-generated content,” but it could be anything else as well. Ring members then vote (semi-anonymously) on member sites to establish “standing”—good, bad, or neutral. A vote comes with a description.
So what does the user see of all this? When a user visits a site that participates in a ring, the Ringspace browser extension will see the presented manifest, and indicate its presence to the user. The user can choose to validate the manifest, at which point the extension reaches out to the ring server to verify the signature, and retrieve standing details. If the site is a member and the manifest validates (via asymmetric key signature), the current standing of the site is provided. Anonymized vote descriptions are provided as well. This design protects voters while informing users about the site’s behavior. And because this is all happening in a browser extension, there is no second site to visit.
I think that’s the gist. Happy to answer any questions!
“Web root” refers to the root of the directory where web content is stored from. So as an example, imagine you have a GitHub Pages site made with a static site generator. Your pages branch or your Environment needs a .well-known/webrings.json file at the top-level.