We need the trust registries to focus where they are the source of truth. Professional Engineer authorities share their truth. Government issuers share theirs. Informal groups share theirs.
/6
There will be trust registries that duplicate information - and even conflict. Over time the need for these will become clear - some will vanish and some will be maintained.
/7
This web of trust registries creates a decentralized and distributed source of information systems can go and get the information needed - that is up to date, authoritative, and yes, decentralized.
/8
While a machine-readable governance approach is a file - and keeping yours up to date is a doomed-to-fail mission. You aren't the source of truth and will quickly reach the point where centralized systems fail.
/9
You will stop updating your governance file because it hits a line of diminishing returns. Adding information on that list of DIDs that your organization only needs once in a while just isn't worth it.
/10
What should you do then?
Let's recognize that for most things you aren't in control... and you don't want to be...
What you want to do is connect to the authoritative source(s). For a list of Credit Unions in the US, hit the MemberPass Digital Trust Registry (disclosure - I was CTO for that) - it is the best source for credit union DIDs.
/12
For other sources, find them and hook into them - using the Trust Registry Protocol.
Each key element (credential types, presentation requests, approved wallets, etc.) will need data structures in an API.
/16
The work that Sam has led on machine readable governance helps immensely here - we don't need to create the API spec from whole cloth.
/17
We can borrow (steal) from work already done...
/18
The data structures and concepts in Machine Readable Governance lend themselves very well to what is needed in the future work on the Trust Registry Specification.
/19
It provides some good structures for Schemas, Rules, Flows and other data structures that are needed.
/20
We can certainly create a "give me your full file" and mimic the governance file that @telegramsam proposes. That's similar to the v1.0 spec providing a full access/offline file.
/21
And if your particular solution aligns with the data replication that some propose, great. Use that. For the other systems (and there are many) a RESTful API may be your best bet.
/22
Let's get discussing...
/23
• • •
Missing some Tweet in this thread? You can try to
force a refresh
When you're talking interoperability - what do you mean?
Single or Multiple Industries/Ecosystems?
Single or Multiple Technical Stacks?
My premise is that you can only ever start with REAL interoperability in the Single Stack + Single Ecosystem realm. Then you need to move in TWO directions:
Single-to-Multiple - one at a time.
only AFTER that can you get to Multi-Stack+Multi-Ecosystem.