Version: 21.5 Recognition of iso-8-1 coding category is broken. An attached file is iso-8859-5 text (mixed Russian and English). It contains only characters allowed in iso 8-bit codings. iso-8-1 detector from 21.4 detects it as such, and detecting it as iso-8-1 (at least "slightly-likely") in 21.5 is also expected. Fortunately, the beginning and end of file do not contain 8-bit characters, so debug print does not abort on it. So this is how iso-8-1 recognition fails. detect_coding_type: processing 212 bytes First 16: As I would not b 41 73 20 49 20 77 6F 75 6C 64 20 6E 6F 74 20 62 Last 16: 9 - 1865), 1854. 39 20 2D 20 31 38 36 35 29 2C 20 31 38 35 34 0A seen_non_ascii: 1 no-conversion: slightly-likely utf-8: nearly-impossible utf-8-bom: nearly-impossible ucs-4: nearly-impossible utf-16: as-likely-as-unlikely utf-16-little-endian: as-likely-as-unlikely utf-16-bom: as-likely-as-unlikely utf-16-little-endian-bom: as-likely-as-unlikely iso-7: somewhat-unlikely iso-8-designate: somewhat-unlikely iso-8-1: somewhat-unlikely iso-8-2: somewhat-unlikely iso-lock-shift: somewhat-unlikely shift-jis: nearly-impossible big5: nearly-impossible detect_coding_type: returning 0 (keep going) detected coding system: #<coding-system raw-text no-conversion>