Unicode basics: one code point per character, round-trip interconvertibility with all existing character sets. But what's a character?

• Some fonts have taken on independent meaning, e.g. ℜ is originally Fraktur, but now is math-ese for the reals
• Some otherwise equivalent characters appear in preexisting character sets, e.g. mainland Chinese sets with both traditional and simplified characters (坛 vs. 壇, 罈)
• Some characters look the same, but aren't, e.g. "o" vs. Cyrillic "о": security issue
• Composed vs. decomposed forms (accents like é, Hangul): normalize for search, comparison