solid / web-access-control-spec

Web Access Control (WAC)
https://solid.github.io/web-access-control-spec/
MIT License
120 stars 25 forks source link

'Request Access' html on 401 responses #40

Open michielbdejong opened 5 years ago

michielbdejong commented 5 years ago

Similar to how you can request access to a document on Google Docs, we could invent a way such that a visitor could cause an email to be sent to the owners, so they can consider whether they want to add an ACL rule for the visitor's webid.

michielbdejong commented 5 years ago

This is a suggestion that came up during the 'acl:trustedApp' session of the unconference we did in Boston on 22 March.

csarven commented 5 years ago

What aspects of the interaction or flow are expected to be prescribed (as per "spec")?

Not that I'm against email, but see the following as to make a "request" via the inherent capabilities of the platform:

csarven commented 5 years ago

As discussed elsewhere issues/chats.. one thing that could be spec'd is the "shape" (SHACL/SheX/Whatever) of the request notification that an application sends to the server's inbox eg. who wants it on what resource, what controls, validity etc.. So, the client application needs to be able to assemble the request (shape) and the server needs to be able to verify that shape, and maybe follow up (not particularly safe but maybe there are UCs or heuristics for that).. or just store it and wait for an application to consume the request by a Control user and then they act on the request.

Edit: I should add that this approach also opens up a space for new class of Solid applications ;) Which is much needed IMO.

csarven commented 5 years ago

FWIW, the way I see it for the WAC spec is to not require additional dependencies like email. Use cases around requesting should work independently - decoupled from other major protocols. Hence, I see working around shaped notifications to be more closely aligned with the current state of WAC and what the surrounding protocols and stack provides.