WG 2022 09 05

Agenda

Time: 7:00 Pacific (fixed), 16:00 CET

Zoom Link: https://us02web.zoom.us/j/85188125213?pwd=UHRUVlgweFJ0T0g4SVdDUytYSkxJUT09

Recording: https://youtu.be/LvxhlkMTX2c

Preliminaries

• Start Recording
• Reminder: Community License v1.0 and Working Group Code of Conduct
• Welcome and Introductions
• Announcements:
  • Hyperledger Global Forum -- Sept 12-14, Dublin
  • Rebooting Web of Trust -- Sept 26-30, Den Haag
    • AnonCreds advanced paper submitted
  • IIW -- Nov 22-25, Mountain View, Calif.

Main Agenda

• Open PRs Review
  • Two ready to go -- VDR-agnostic object identifiers and the AnonCreds Objects Methods Registry
• Moving open source AnonCreds implementations out of Indy
  • Discussed and there was support expressed
    • Next steps -- continued socialization at Hyperledger Global Forum and create a project proposal after that
    • Need sponsors in submitting a project proposal to the Hyperledger Technical Steering Committee (TSC)
  • Some concern about the layers of implementation (bottom up) as AnonCreds evolves: Ursa / AnonCreds / AnonCreds Objects Methods / Agents
    • Thought experiment -- would change with the implementation of the zk-SAM revocation mechanism?
      • I think it would be the Ursa and AnonCreds implementations, but the ANOM and Agent interface would be largely unaffected?
• Latex in Spec-Up / (Re)Discovery of AnonCreds Protocols paper from Mike Lodder, Brent Zundel. I assume we can use that for the AnonCreds Protocols section of the specification.
  • Can also look at the ursa-docs repo -- pure latex
  • Will try to embed the latex in a markdown fenced code block
    • PR submitted to the Spec-Up spec enables the use of latex directly in the markdown
• Action Items
  • To Be Added
• What's left?
  • To Dos:
    • Need to investigate the ref data element and what happens if it is not an integer
    • Cred Def Generation + PRIVATE_CRED_DEF -- non revocation, and plus revocation
    • Revocation data elements -- definition
    • Normative/Non-normative references
      • Collect from documents mentioned below (under action items) and from previous meetings
      • @ale-linux to do a first cut of this.

Action Items From Previous Meetings:

• 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
• From Artur Philipp : Issue about raw/encoded values used in AnonCreds https://jira.hyperledger.org/browse/IS-786
• Challenge for AnonCreds implementations on other VDRs in supporting revocation.
  • Add the option of NOT supporting revocation? Yes
  • Tricky part is the handling of deltas of RevRegEntry sets without custom VDR support
    • Conflict between the AnonCreds proof request semantics and functionality of VDR
  • v2.0 goal - a revocation scheme that does not rely on deltas -- e.g. entire registry is published with each update
    • That removes any "ledger-side functionality"
    • We could add a comment about the use of RevocationList2020 with AnonCreds as that eliminates the deltas, but opens up correlation and ongoing monitoring.
    • Would rather see something like Revocation using zk-SAM
• Discussion of AnonCreds and the VC Data Model
  • Issuer Date is required
  • Canonicalization in AnonCreds was discussed -- the signing of the encoded integers vs. anything else
  • @TelegramSam to investigate and may propose an approach.