It would be sad, if the slight updates of Unicode (minus the emoticons) lead to constexpr only ever supporting ASCII and no international symbols.   I get that compilation should be stable (lead to the exactly same results) over a long time and constexpr can also have an effect on ABI compatibility.   Perhaps the C++ standard has to freeze a specific Unicode version for constexpr support or define their own variant of rules, as was mentioned here in this thread every operating system and file system does.     We can use unicode letters as identifiers in C++. What if I want to do metaprogramming and change a camel case identifier into a lower case identifier? I have to create my own constexpr functions for unicode case conversion?   -----Ursprüngliche Nachricht----- Von:Thiago Macieira via Std-Proposals <std-proposals_at_[hidden]> Yes. What's the justification for their being constexpr? What version of the Unicode standard is it using? I have no problem with an ASCII-only set of functions and those being constexpr. -- Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org  Principal Engineer - Intel Platform & System Engineering