Minutes of the August 6th 2024 Monthly Meeting

[johngray-dev]

Present: Abel Chen, Laurence, Mike Tsai, Austin Lin, Markku-Juhani Sarinen, Nicola Tuveri, Petr Muzikant, Daniel Van Geest, Julien Prat, Mike Ounsworth, Alexander Ralien, John Gray, Christopher Wildfeuer

Introductions of newer members

• The Telecom of Taiwan (CHT) team (Abel, Laurence, Mike T, Austin) - Happy to be working on the project, have contributed artifacts and done interop testing.
• Nicola Tuveri (first time at a Monthly meeting) - Works with Markku, had attended in Prague.
• Christohper Wildfeuer - Has come once before. From university of Zurich, working on own implementation in software and hardware in VHDL.
• Petr - Still considers himself to be ramping up, but contributed the oids.json list which is used by the automation and some providers.

Other introductory comments:

• Julien - Working with Daniel, interested in X509 for signatures and KEM and CMS
• Mike Ounsworth - See public and private key discussion below, and also talked about the automation work that was done at the past hackathon.
• Markku - Keeping track of the project, very active on pqc forums and researching implementations of the pqc algorithms.
• Alexander - Looking for CMP interoperability and starting work on CMS signatures
  • John: Was able to parse Alexanders CMP artifacts, will work on verifying an producing in next few weeks.

Discussions:

Private Key Format vs Public key format:

• ML-KEM private key contains the private key, the public key is needed for private key operations
• Always a safe assumption that it contains the public key because of the FO transform
• If you support the seed mode you can derive the public and private key.
• When Exporting a private key in a P12 format, drafts should say something about the format containing the public key in them as well. This would apply to the ML-KEM drafts and Composite KEM
• Question about why OQS keys are so big. This is because the private key does a concatenation (it is not standard). Hopefully they will update when the standards happen.

JSON Portable OID format

• the https://github.com/IETF-Hackathon/pqc-certificates/blob/master/docs/oids.json file is automatically updated from the Main OID markdown file https://github.com/IETF-Hackathon/pqc-certificates/blob/master/docs/oid_mapping.md
• The update is triggered when https://github.com/IETF-Hackathon/pqc-certificates/blob/master/src/pqc_report_writer_common.py is run, and that file is triggered when the automation of the artifacts is run
• The result file will be inside the docker imagine created during the automation run, and it is not added back into the main repository
• See Should oids.json automatically update based on Markdown table changes? #130

Hackathon Updates Discussion

• SLH-DSA in automation tables doesn't look good. In OQS they are not enabled by default.
• Does anyone know if they are just using 4 parameters sets? Is there a good reason? Only change is the pre-hash
• They claim support for SPHINCS+ in their documentation, so do they actually support SLH-DSA or is that a documentation issue?
  Could this explain the issue? Are the OIDs we use correct.
• Action John: Send an email to the OQS team to try and sort out SLH-DSA / SPHINCS+ support and clarify OIDs and why they are off by default.

CMS Artifact Updates

• Alexander is looking at signature artifacts for CMS. Will update the file format for CMS.

Pre-HASH OIDS

• Have we tested any pre-hash OIDs yet. Answer is not yet. Will probably wait for OIDs to drop from NIST in the next few weeks and then add artifacts to test these modes. See https://groups.google.com/a/list.nist.gov/g/pqc-forum/c/HmSNaPpX214/m/17oaB-a4AQAJ

Carl implemented composite KEM

• We should follow up to Pravek's OQS implementation of that as well. Interop is now possible

TLS hybrid testing

• There is interest in the group to test some of the hybrid formats like composite with TLS.