README.md

Derive Secret

Aa function for deriving new secrets from an initial secret. It's defined as:

function Derive (key, feed_id, prev_msg_id, labels, length) {
  var info = ['envelope', feed_id, prev_msg_id].concat(labels)

  return HKDF.Expand(key, encode(info), length)
}

• key is a cryptographic key you're deriving from
• labels is an Array of strings
• feed_id and prev_msg_id are encoded in with the binary type-format-key (TFK) encoding
  • if there is no prev_msg_id (because you're publishing the first message for this feed), we say the "key" part of the tfk encoding for prev_msg_id is a zero buffer of the same size, e.g.

      01 00  00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
       │  │  └────────────────────┬────────────────────────────────────────────────────────────────────────┘
    type  │                hex encoded key
         format 
    

• HKDF.Expand is a hmac-like function which is specifically designed to generate random buffers of a given length.
  • we specify sha256 for hashing in HKDF-Expand
  • example of a node.js implementation : futoin-hkdf
• encode is a shallow lenth-prefixed (SLP) encoding of an ordered list

Design

We want to derive unique keys which are very unlikely to collide with other keys, where "unlikely" means:

• won't happen by chance
• won't be easy to trick a user to performing a particular derivation

By baking the context (feed_id and prev_msg_id) into the info used to derive new secrets, we make the derive function very specific.

This has the side-effect that we can use zero'd nonces for encryption, because we can be very sure that the same key will not be generated twice.