Unicode is good. If you’re designing a data structure or protocol that has text fields, they should contain Unicode characters encoded in UTF-8. There’s another question, though: “Which Unicode characters?” The answer is “Not all of them, please exclude some.”