Protocol Name - Trust Registry Query Protocol?

[darrellodonnell]

Would the name "Trust Registry Query Protocol" make more sense than "Trust Registry Protocol".

I raise this as I always getting questions from that want to understand how the TRP helps them add or update registry entries. That's a brutally hard thing to consider, while getting an answer to a simple question is, well, simple:

Does Entity X have Authorization Y, in the context of Ecosystem Governance Framework Z?

[trbouma]

Hello @darrellodonnell checking to see if this works....

The statement, IMHO above is the essence of the specification. It might be generalized further which I am now musing on.

[darrellodonnell]

we should get this statement into the Scope to hit the reader early with the basic idea.

[trbouma]

I think it can further normalized to:

Z authorizes X for action(s) Y

where:

X,Z are verifiable identifiers bound to Entity X and Entity Z

and

Y is the set of actions related to authorities/statuses/privileges granted to X in the context defined by Z

[darrellodonnell]

to your point, it is authorization or status ? that changes the question a lot

[trbouma]

I'd suggest sticking with authorization - that's what is really about in the end.

[trbouma]

Recasting Darrell's statement:

Does Entity X have Authorization Y, in the context of Ecosystem Governance Framework Z?

into something that is expressed more mathematically and can be implemented:

[X]<----(X,Y)SIGNEDBY(Z)-----[Z]

The tuple (X,Y) means that X is authorized for Y as signed by Z, which is the same as the context of Z

What is exactly Y and under what conditions is (X,Y) signed is specific to the the governance of Z.

This is just a generalized case of what happens in DNSSEC with a DS record. For DNSSEC it is just for delegation of signing keys within the context of a hierarchy. In this case it's between identifiers X and Z, and it's a graph (web) as opposed to a hierarchy'. The (X,Y)SIGNEDBY(Z) defines the authorization relationship.