UTF-8 is superior due to self-synchronizing features and the fact truncated or otherwise damaged sequences are obviously damaged, even if they use unknown characters.