⌘ 01 2 Sept 2026 · 1 min read Finding duplicates without hashing your whole disk A three-stage filter that reads 64 KB from each file before committing to a full SHA-256 — and why 'size on disk' is the number you actually want to sort by. Performance ↗