Post #1249780
2026-04-15 07:39 UTC
@huitema @pedromj @paulehoffman @rsalz The core issue is endorsement. It isn't about having a stable reference; having a stable reference doesn't need an RFC. It isn't about interoperability; interoperability doesn't need an RFC. The "control" argument is circular; https://archive.cr.yp.to/2026-04-10/05:38:16/1w0wAgKE9fiKZKunAg8qCyVyWYZ4j-aHgW-0aFDzgcw/https/mailarchive.ietf.org/arch/msg/tls/LqG-gHxgRvVPebE3m28D8VT7dN4/ spells this out in baby steps.
Replies (1)
-
@letoams@defcon.social 2026-04-15 15:26
@djb @huitema @pedromj @paulehoffman @rsalz funny you used the exact opposite argument for ssh sntrup761x25519-sha512, arguing a code point wasn’t enough - despite even already being almost universally available in a vendor namespace. You just wanted the “endorsement” of an RFC. Cherry pickers will cherry pick 🍒