2024-03-18 AnonCreds Working Group Meeting

Summary

• Status Updates on AnonCreds in W3C VCDM Format projects
• Revocation Manager for ALLOSAUR and AnonCreds – Design
• Schema Object for Complex JSON Objects – Discussion
• Open Discussion

Time: 7:00 Pacific / 16:00 Central Europe
Call Link: https://zoom.us/j/97954159540?pwd=WWk3WmQ3MVh1SXBYZGVreGl0QllGdz09

Recording: 

Notices: 

This specification creating group operates under the Linux Foundation Community Specification License v1.0.

Meeting Attendees

• Stephen Curran (BC Gov / Cloud Compass Computing Inc.) <swcurran@cloudcompass.ca>

Related Specifications and Repositories:

• AnonCreds v1.0:
  • The v1.0 specification is published here: https://hyperledger.github.io/anoncreds-spec/
  • The Working Group uses this GitHub repository to manage the specification: https://github.com/hyperledger/anoncreds-spec
  • The AnonCreds Methods Registry: https://hyperledger.github.io/anoncreds-methods-registry
  • The v1.0 implementation in Rust is here: https://github.com/hyperledger/anoncreds-rs
  • The v1.0 implementation is dependent on this Rust CL Signatures implementation: https://github.com/hyperledger/anoncreds-clsignatures-rs
• AnonCreds v2.0
  • The initial framework for the v2.0 specification repository is here: https://github.com/hyperledger/anoncreds-spec-v2
  • The v2.0 implementation in Rust is here: https://github.com/hyperledger/anoncreds-v2-rs
  • Underlying AnonCreds v2.0 are cryptographic libraries in Hyperledger Labs Agora

Meeting Preliminaries:

• Welcome and Introductions
• Announcements:
  • IIW – April 16-18, 2024
  • Workshop Wed. April 24th: Zero Knowledge Proofs and ZK Programming in Blockchain Application Development
• Any updates to the Agenda?

Agenda

• AnonCreds in W3C VCDM Format - Status/Updates
  • Test Vectors Repo: https://github.com/TimoGlastra/anoncreds-w3c-test-vectors
  • Credo / Bifold Progress
  • ACA-Py Progress
• ALLOSAUR/AnonCreds Revocation Manager Component – what is it, design?  ALLOSAUR/AnonCreds Credential Status Manager (reflects 
  • Presentation
  • Hyperledger Mentorship Program
  • Proposed Hyperledger Mentorship project: AnonCreds Revocation Manager Component Implementation
  • On the Ledger:
    • RevRegDef: ID, Public Key, Accumulator, RM URLs
    • RevRegEntry: Accumulator (no list of changed status)
  • Is this NIST-compliant?
    • NIST has no approved pairing cryptography libraries, and this is based on pairing cryptography
    • Uses BLS12381 – which is used by many of the blockchains, BBS+, etc. – stronger curve – zCash and others
• Update on the Hyperledger Labs Agora Libraries
  • Contributions this week – audits are completed
    • Audit Complete: BLSFUL – https://github.com/mikelodder7/blsful - powers the ALLOSAUR library, replaces what is currently in AnonCreds v2
    • Audit Complete: Gennaro – https://github.com/mikelodder7/gennaro-dkg – distributed key generator
    • Being audited: Verifiable Secret Sharing – https://github.com/mikelodder7/vsss-rs
    • Plan is to ALLOSAUR into Agora as well
• PS vs. BBS+ Attributes
  • Aside: Rumour has it that a PS key size changes based on number of attributes to sign. If so, does that impact support for arrays in complex JSON?
    • PS must know the maximum number of claims you are going to sign at credential definition time.  Tradeoff of key size vs. number of claims.
    • PS vs. BBS+ relates to threshold signing.
    • PS has stronger security proofs.
    • PS – smaller and faster to generate
  • Nice point – could choose based on use case.
• Complex JSON in AnonCreds – ideas for the Schema Object
  • Thinking about AnonCreds V2 with complex JSON, and what the Schema would look like.
  • In V2, there are is metadata per VC attribute that provides input into the encoding for the element -- e.g., it's a string, an integer, an integer range, a date, an enumerated set, scalar (e.g. link secret).
  • Example of a simple schema to show the metadata: https://github.com/swcurran/anoncreds-v2-rs/blob/samples/examples/schema-with-linked-secret.json
  • Question – how do we manage the schema when we need to support complex JSON (structures, arrays). 
    • Idea 1: Use JSON.Path so we are back to a list of attributes
      • Schema is a flat list like we have in AnonCreds v1, but attribute names are in JSON Path format. 
      • Metadata is attached to each attribute in the (flat) list.
    • Idea 2: Use JSON-LD 
      • How much would JSON-LD help with this?
        • Probably not much
      • Is there already enough data to tell us much of those things?
        • I would guess not. It could have a lot of it, but not all that we need.
        • Completely dependent on the use case – definitely not guaranteed. Must be able to supplement it.
      • What tools are there to see for a given JSON-LD document what attributes we know about each attribute?
        • Tools are available.
      • Would it make sense to use JSON-LD and supplement attributes as needed?
        • Question for the JSON-LD crowd.
    • Idea 3: Combine the two
      • Use JSON-LD for much of the data
      • Use JSON.PATH to add additional metadata when needed to specific items.
    • Idea 4: Pure JSON-LD, adding new "features" to get support for what we need
      • Don't really know what this entails...
• Open Discussion

To Dos:

Action items

• Links to be referenced in the spec and used where needed:
  
  • From Will Abramson : https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.186.5994&rep=rep1&type=pdf
  • From Will Abramson : https://github.com/hyperledger/indy-hipe/blob/c761c583b1e01c1e9d3ceda2b03b35336fdc8cc1/text/anoncreds-protocol/README.md