Who actually built the Solid protocol, and what its standing is
Not us. We implement none of it. solid-protocol is not-built on our side; the server running this surface is the Solid community's own, and our contribution is the integration layer around it.
Where it came from
Solid was initiated by Tim Berners-Lee and developed with researchers at MIT, then carried forward by a community that has been working on it for years.
The specification is maintained by the W3C Solid Community Group: an open group anyone may join, publishing in the open, with the work visible in public repositories.
The implementations are theirs too. The Community Solid Server (the reference implementation this surface runs) comes out of academic and community work, principally at imec and SolidLab in Ghent, with a named maintainer community. There is more than one implementation, which matters more than any single one of them.
Its standing, stated precisely
Fetched from the document itself rather than characterised by us:
Draft Community Group Report, 12 May 2024
Version 0.11.0 · published by the Solid Community Group
And the specification says of itself that it is not a W3C Standard and not on the standards track. That is their own sentence, in their own status section, and the fact that it is written there at all is the point worth taking away.
Why that distinction matters, and why it is not a criticism
A Community Group Report and a Recommendation are different things. The first is work published by a self-organising group; the second has been through a formal process with wide review and a consensus decision. Conflating them is a real error, and it is the kind of error that gets repeated by people summarising someone else's work.
But the ranking people expect from that distinction does not follow. A community report with several independent implementations and real deployments is worth more in practice than a formal Recommendation nobody ships. Process standing and practical standing are different measurements, and Solid scores differently on each.
We are not going to overstate this in either direction. Solid is not a formally ratified web standard; it is a specification with real implementations and a working community.
The mirror we are obliged to hold up
We ask exactly this precision of ourselves.
Our own identifier method is registered in the W3C DID Method Registry: an entry in a list, and not a W3C standard, which is a sentence we are required to write that way every time. The distinction we just made about Solid is the same one we make about ourselves, and a page that applied it to somebody else's work but not our own would be worthless.
On any honest reading, Solid's position is the stronger of the two: their specification has multiple independent implementations and deployments by organisations unconnected to its authors.
What we take from them, spelled out
- The storage model, resources over HTTP, in open formats.
- The access-control model, the four permission modes, enforced by their server.
- The identity binding: the identifier-and-profile pattern, which we contrast with our own and have never claimed to beat.
- The server itself.
What we added is an integration layer. That is a real contribution and a small one, and calling it anything larger would be theft.
What you can check today, and what would change our half of it
You can verify every claim on this page yourself, and none of it depends on us: fetch the specification and read its status line, read the server's source, look at the community group's public repositories. Nothing here rests on our word, which is the correct arrangement for a page about somebody else's work.
None is done, and the specification's standing does not move either way when they are.