Minutes of February 4th Meeting

[johngray-dev]

Present: Corey Bonnell, Julien Prat, Daniel Van Geest, Felipe Ventura, Mike Ounsworth, John Gray, Nicola Tuveri, Jean-Pierre Fiset, Hendrik Brockhaus, Kiron Mirdha, Austin Lin, Abel C. H. Chen

• A little bit of discussion on the ML-DSA private keys discussions going on between Oasis and IETF. Looks like IETF may now be moving towards the more flexible format. There is more to discuss on this and it does impact the ML-DSA drafts as well as the composite ML-DSA drafts.

Round Table

• John Gray - Working on updates to the composite specifications. We are looking to submit new versions before the IETF submission cutoff date in March. Also submitted new composite -03 artifacts to the repository for two different Entrust implementations (one in Go, one in Java). They are compatible with the other implementations.

• Corey Bonnell - Will probably add linting for private keys - Probably wouldn't be hard to export to private keys. There is some complexity there for automated verification check. Won't be in person for next hackathon but will plan on participating remotely.

• Julien Prat - waiting for the stabilization of the private key format

• Daniel Van Geest - Waiting for the stabilization of the private key format before starting working on interop testing of the private key formats. Still don't know if he will be in person in Bankok but will participate.

• Felipe Ventura - Still working on version 2 of composite in LiboQS. Planning to come to hackathon remotely.

• Mike Ounsworth - PKCS11 - PKCS11v3.2 Oasis authors are happy with it, they voted to go ahead with publishing it. LAMPS should support both private key formats (expanded and seed), however discussions continue at IETF.

• JP Fiset confirmed that Crypto4A has gone through all the CMVP tests for PQ, INCLUDING EXTERNAL MU! https://csrc.nist.gov/projects/cryptographic-algorithm-validation-program/details?product=1829 also see https://www.einpresswire.com/article/782768334/crypto4a-becomes-first-company-to-achieve-full-acvp-certification-for-nist-pqc-algorithms

  • JP also mention the "Root Causes" sectigo podcast is not very positive on composites. This is kind of expected because they are specifically talking about TLS use-cases. We know others like SWIFT and Microsoft want to use composites for things like code signing or other uses.

• Nicola - Can attend in person

  • Updating OQS provider artifacts for R4 Update oqs-provider scripts for R4 #165 - will work with Felipe on getting artifacts.
  • Planning on releasing a new openSSL provider - main difference is instead of being tied to libOQS, you can plug in any provider. Using lib crust, has some formal verification. One more different algorithm implementation. Registering OIDs for standardized algorithms. This will be open source! Main target is ML-KEM 768, X25519 hybrids. Also working on the firefox hybrids. Maybe adding support for Composite or Hybrid. TLS with Authentication.

Pull Requests

• Closed Cleanup for IETF 121 #150 as team wanted to keep older artifacts r2 and r3 for historical reasons.
• Update oqs-provider scripts for R4 #165 was mentioned above - Felipe and Nicola still working on it
• Acvp auto testing #129 If no progress is made before the next hackathon, Mike and Felipe will probably work on this for the next hackathon.