Your explanation is correct. UTF-16 is a hack on top of a hack, and no one uses it in new software. However, Java and Javascript are stuck with it for legacy re

Your explanation is correct. UTF-16 is a hack on top of a hack, and no one uses it in new software. However, Java and Javascript are stuck with it for legacy reasons. It wouldn't be so bad if people just understood that there are (almost) no valid uses-cases for measuring the size of a string. As someone else mentioned, it's a binary blob for all intents and purposes, and if processing of its content needs to be done (such as displaying it on the screen, or performing, say, word-wrap) then it should be handed over to libraries that have been designed for this purpose, because these things are very complicated. Simply by having a single non-ASCII character in my name, I'm seeing software fail on this on a regular basis.