Conversation
|
Btw, I'm a little bit uncomfortable when looking at the
I suggest we should refactor the |
If we don't, we assume the key to be of a specific format (like the one created by ed25519). Not returning the public-key here could make #50 more difficult to implement.
I agree that the underlying type should be a slice instead of a fixed sized array (which again assumes we use a specific key format/size). |
If we do, |
I see your point. Then why not call it |
As a side-note: It would be fairly easy to modify the VRF lib to use sha512 instead of sha3/shake. Then we could use the exact same key type as in sign.go ( |
I think this is a fine idea. Can you open a separate ticket for it? I'd like to merge this code and I don't want the discussion to get lost.
Not sure I agree. If, for whatever reason, the public key cannot be generated from the private key at call time, ( |
Sorry, I opened #65 |
Thanks! |
This pull is a part of #29. It includes following changes:
merkletree/policy.go.utilpackage.sign.GenerateKeyshould take a reader asvrf.