Checksum

Overview

Checksum adds redundant information derived from the original data. A verifier recalculates the value and compares it with the supplied checksum. If they differ, the data has been corrupted, mistyped, or encoded incorrectly. The exact algorithm and checksum length determine which error patterns can be detected.

User-facing blockchain formats often incorporate checksums. Base58Check appends checksum bytes before Base58 encoding, while Bech32 and Bech32m use specialized character-level checksum constructions. These mechanisms help wallets reject malformed addresses before a transaction is signed.

A checksum is not designed to stop an attacker who can modify both the data and checksum. An adversary can generate a different valid address with a correct checksum. For authentication and tamper resistance against intentional changes, systems use cryptographic signatures, message authentication codes, or trusted hash commitments.

Checksums can also fail when software applies the wrong format. A string may have a valid checksum under one network or version but still be unsuitable for the selected asset. Validation must decode the complete structure and inspect network, type, and payload rules rather than treating checksum success as a universal approval.

In operations, checksum failures should be reported as input or transport errors and should stop processing before funds move. Systems should preserve the original value for troubleshooting but avoid silently correcting it. A corrected-looking address could identify a different valid destination, creating a more dangerous error than an explicit rejection.

The probability of an accidental error passing a checksum depends on the construction and error pattern, so different formats provide different guarantees. Implementers should use the exact algorithm named by the protocol and official test vectors. Substituting a generic CRC or truncating the prescribed value can create incompatible identifiers.