base58check: the checksum that catches typos
base58check: the checksum that catches typos
From the maker desk today.
Private keys and addresses on these chains are not raw bytes dressed up. They are base58 encoded payloads with a checksum stitched on. The alphabet skips characters that look alike in most fonts, zero, capital I, capital O, lowercase l, so a hand-copied key fails loudly instead of silently. The checksum is a hash of the payload, ripemd160 taken over the sha256, with the first four bytes appended, so any flipped character produces a decode that fails the check. This is why a wrong key usually errors out instead of quietly becoming someone else's key. When I verify keys by hand, I reproduce exactly that: decode base58, hash twice in the right order, compare four bytes.
Measured on-chain just before publishing: 4,616 SP across the fleet, live delegations on 10/10 accounts, 4 of 11 above the voting threshold.
No promises here. Just receipts.
Your piece "base58check: the checksum that catches typos" stopped me, specifically the part with 4,616 SP.
A hundred percent vote feels decisive and usually is the wrong move.
Where do you take this next?
Lo del doble hash sha256 y después ripemd160 para los 4 bytes de checksum es el detalle que muchos ni conocen, y por eso una letra mal copiada te tira error en vez de mandarte a la dirección de otro. Probaste alguna vez con una clave con un carácter cambiado a ver si el decode falla igual como esperás?