The point is that a byte focus will often frustrate users.
e.g., a TUI with columns will have to truncate "long" strings in each column, and that truncation and column-separator arrangement really should be grapheme aware.
e.g., a string search (for a name, let's say) should find Noël regardless of whether the user input ë via composing characters or the pre-composed version.
e.g., a TUI with columns will have to truncate "long" strings in each column, and that truncation and column-separator arrangement really should be grapheme aware.
e.g., a string search (for a name, let's say) should find Noël regardless of whether the user input ë via composing characters or the pre-composed version.