tomaswolf commented on issue #928: URL: https://github.com/apache/mina-sshd/issues/928#issuecomment-6190939136
> you'd implement ML-KEM yourself and we could try to collaborate in multiple PRs for ML-DSA? Yes. Pure ML-KEM is implemented; I'm just not pushing it right now because I have a patch release to do. It still needs an interoperability test, and so far I don't see any suitable docker container I could use. It needs to be small, multi-arch (x86_64 & arm64), and have an SSH server that does pure ML-KEM. For ML-DSA I will review PRs on the dev_3.0 branch, time permitting. It also will need an interoperability test, and again I don't see a suitable docker container yet. (Something like https://github.com/beldmit/openssh-ietf126-hackathon/blob/master/ml-dsa/Dockerfile won't do; that pulls a full Fedora 44, uses a non-standard dnf/yum repo, and installs gcc and compiles some stuff. That would make our CI build super-heavy. As above: containers used in tests should be small and multi-arch.) As for the file format for private ML-DSA keys: for PKCS#8 follow RFC 9881. For an SSH-like format following usual OpenSSH practices seems best and that would mean to just store the seed. (It's a bit moot anyway given that OpenSSH has no inclination to do pure ML-DSA at all, but if they ever do, I'd be surprised to see them storing the expanded form. On input, we might tolerate "seed followed by expanded" and just ignore the expanded part.) -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
