I don't think that's what he meant. A surrogate pair in UTF-16 is still referring to a single code point. It's a good point that it's a place where people could

I don't think that's what he meant. A surrogate pair in UTF-16 is still referring to a single code point. It's a good point that it's a place where people could screw up in dealing with Unicode, but it doesn't match his complaint. What does match is the fact that you can use a huge number of combining characters [1] to form a single glyph; each combining character and the base are a code point, so in order to figure out how many glyphs there are you have to iterate with the knowledge of what code points are combining characters. [1] https://en.wikipedia.org/wiki/Combining_character https://en.wikipedia.org/wiki/Combining_character