You almost always need the byte count for protocols, buffers and stuff. For user-interfacing use-cases (e.g. editors) you probably want the glyph or grapheme co

You almost always need the byte count for protocols, buffers and stuff. For user-interfacing use-cases (e.g. editors) you probably want the glyph or grapheme count, not the rune count. Please remember, runes / code points, graphemes and glyphs are different things: http://www.icu-project.org/docs/papers/forms_of_unicode/ http://www.icu-project.org/docs/papers/forms_of_unicode/