Perl supports most single-byte locales, plus UTF-8 locales. It does not support any other multi-byte locales, such as are found for some East Asian languages. On some platforms, just switching into some of these locales can cause perl to segfault (or rather it is a bug in the libc in failing to detect garbage input) Since POSIX 2001, it has been possible to detect that locales are multi-byte. And on platforms that support that, we could prohibit using them. I propose to do so, by having POSIX::setlocale() fail if perl detects that the locale is an unsupported one.Thread Next