Why are the seeds different when going from seed to mnemonic and back using bx?

From an example in the documentation:

$ bx seed | bx mnemonic-new | bx mnemonic-to-seed | bx hd-new

655a1ed2d4fef69bb314198c7327f23c grab special regret prepare urge evidence slush lobster midnight odor wish ketchup 57bae342ae8e69eb63f17ef993a90a59159e0f78114b602bf0ebabfec0e5d086e883c31975bf03f8a47a32853452623094d1303fd0549745db457145e5756582 xprv9s21ZrQH143K43CUCgNp5SmXzg2axQx2P4WupA2zEnKpFM19QfqfdqfpJR3yfzAXZnsHeUaQhWQMwyuqL8DbdeLeCbdMoKNn6pCY4RUn2pK

In Mastering Bitcoin, the example given here shows it going back to the seed that was originally created; this was done using an older version of bx though. Can someone explain how the seeds from line 1 and line 3 in the example are encoded/related?

Kevin

Posted 2018-03-26T02:33:17.260

Reputation: 15

Answers

They adopted bip39 where the seed mnemonic is run through a key derivation function to derive the seed. This is an irreversible step.

Abdussamad

Posted 2018-03-26T02:33:17.260

Reputation: 1 850

I asked the same questions regarding the example in Mastering Bitcoin and here is the answer:

The old mnemonic-decode (electrum v1 mnemonics I believe) was dropped from libbitcoin before version3 and replaced with BIP-39 mnemonics. libbitcoin version4 also adds in electrum v2 mnemonics. Unlike electrum v1 mnemonics (which are symmetric), BIP-39 is asymmetric --https://github.com/libbitcoin/libbitcoin-explorer/issues/366

LIU YUE

Posted 2018-03-26T02:33:17.260

Reputation: 194