- From: Melvin Carvalho <melvincarvalho@gmail.com>
- Date: Tue, 24 Mar 2026 17:50:45 +0100
- To: Sarven Capadisli <info@csarven.ca>
- Cc: public-solid@w3.org
- Message-ID: <CAKaEYh+Xz=bSbJ1TAsRhfh9YsM8PX1x1XeLASY6hL5ZO4TeNJA@mail.gmail.com>
Ășt 24. 3. 2026 v 13:29 odesĂlatel Sarven Capadisli <info@csarven.ca> napsal: > Ahoy hoy! > > After much deliberation and popular demand, there is now a PR: > > https://github.com/solid/web-access-control-spec/pull/134 > > proposing capability detection for extensible authorisation conditions > in WAC. Servers can signal supported condition types, and clients can > use them accordingly. > > As anticipated in > https://solidproject.org/TR/wac#authorization-extensions , this version > adds initial condition types to the core spec, issuer and client > identification. The design allows future types, e.g. time based > conditions or ODRL policies. > > Changes are backwards compatible. Servers without condition support are > unaffected. Adoption can be incremental. > > That said, we want to do this properly. This is also a call for > implementers or a call for commitment to implement so that we can align > the specification with real and open implementation experience before it > advances further. Please chime in! > > For those interested in technical writing, review of the PR would be > much appreciated. > > As usual, myself and others are available to support the community in > understanding the proposed changes. Feel free to reach out in the chats, > meetings, or catch me at in-person events where I am happy to convene an > impromptu WAC-support Group. > NACK. As a WAC implementer, I will not implement this proposal as part of WAC 1.1. The current design introduces a silent authorization bypass on WAC 1.0 servers that do not understand the new condition types. That is a security regression and violates the principle of fail-closed authorization. This is not appropriate for a minor version update. It changes the effective security model of WAC in a way that existing deployments cannot safely interpret. If pursued, this must be defined as a separate extension specification with explicit opt-in and independent implementation experience. > > -Sarven > https://csarven.ca/#i >
Received on Tuesday, 24 March 2026 16:51:04 UTC