Secure File Hash

Which algorithm, and why it matters

SHA-256 is the current default and what any new system should publish. SHA-512 is often faster on 64-bit hardware for the same security. MD5 and SHA-1 are cryptographically broken — practical collisions exist, meaning two different files can be constructed to share a hash — but remain perfectly adequate for detecting accidental corruption, which is a different job from detecting tampering.

What a matching hash proves

That your bytes are identical to the bytes that produced the published value. That covers a truncated download or a flipped bit completely. It does not prove authenticity: whoever can replace the file on a server can usually replace the hash listed beside it. Origin is proved by a signature verified with an independently obtained key, not by a checksum.

Hashing in the browser keeps the file local

The Web Crypto API computes the digest on your machine, so the file is never uploaded. That matters for anything confidential, where sending a file to a remote hashing service to check its integrity is self-defeating.

Any change alters the whole hash

Flipping one bit changes roughly half the output bits, so hashes are useful for detecting change and useless for locating it. A mismatch tells you the file differs and nothing about where or by how much.

Line endings break text file hashes

A text file moved between Windows and Unix systems may have its line endings converted, changing every line and therefore the hash while the content is semantically identical. FTP clients in ASCII mode do this silently, which is why a text file mismatching but opening correctly is usually this rather than corruption.