Hey, I have a few questions regarding pods and access, some of the concepts are a little fuzzy to me.
From what I understand, a user can have multiple pods (be it with one or different pod providers), by specifying http://www.w3.org/ns/pim/space#storage?
So how would you create a second pod without also creating a new identity/WebID? Can a Solid server solely be a Pod provider and not necessarily an idp?
Is it possible to know what resources you have been granted explicit access to, e.g. to have a container in your pod that lists all these external resources?
When accessing a piece of data from a thing, how to know what ontology was used to store that data?
E.g. I could store my street address using vcard or schema.org. Should I just look for both?
What should the architecture look like e.g. in the scenario where the government would want to keep your official documents in a Pod, should they:
Host a pod for you to which you have full control (and keep a backup elsewhere in case you were to delete or tamper with the data)
Host a pod for you to which you only have read access
Be given access to your personal pod and store a copy of the documents there
So how would you create a second pod without also creating a new identity/WebID? Can a Solid server solely be a Pod provider and not necessarily an idp?
Yes a Pod provider can provide pods only and and an IdP can provide identity only, or they can mix and match.
In my understanding, you would need to register a different WebID to get a new pod.
Is it possible to know what resources you have been granted explicit access to, e.g. to have a container in your pod that lists all these external resources?
It’s possible but basically requires crawling your entire pod. New processes are in development which would make this more doable in the future.
When accessing a piece of data from a thing, how to know what ontology was used to store that data?
E.g. I could store my street address using vcard or schema.org. Should I just look for both?
There are things called “shapes” which describe the ontology of a resource. If you don’t have a shape for a resource, yes, you basically have to guess the ontology. For profiles we are writing a spec describing the recommended ontologies for profiles. Other kinds of documents may not have such rules.
What should the architecture look like e.g. in the scenario where the government would want to keep your official documents in a Pod, should they:
All three methods you mention can be useful in different circumstances, it depends on the context.