@q@glauca.space @r@glauca.space @49016@mastodon.catgirl.cloud Oh i thought there was only one curve that mattered . okay sure yes EC does not specify a curve. hmm. FUCK . okay structure is justified. dammit
0 Boosts
0 Likes
![210561 ✨ [Q]](https://mastodon-data.content.as207960.net/accounts/avatars/108/220/736/453/606/716/original/c8b2ed3c5237c336.jpg "210561 ✨ [Q]")
@q@glauca.space @r@glauca.space @49016@catgirl.cloud @sodiboo@gaysex.cloud Is your Ed25519 private key a value that is hashed to derive the actual private key, or the actual private key?
@q@glauca.space @49016@catgirl.cloud @r@glauca.space @sodiboo@gaysex.cloud Continuing along to ML-KEM and ML-DSA...
...is your private key a seed value that is hashed to derive the actual private key...
...or the "expanded" (but not fully expanded for Reasons) private key...
...or both!
cryptographers are rightfully mad that the answer is "either or both" when the answer should have been "just the seed" or if we couldn't constrain ourselves to that because of HSMs, "just the seed or the expanded key"
But "both" is definitely the wrong answer. Because what if they're not the same key?!
@sodiboo@gaysex.cloud @r@glauca.space @q@glauca.space also having a decent format is generally nicer to work with than just "lol here's a bunch of bytes go figure it out and dont use those bytes the wrong way"