Hello I tried to patch the built-in ruby-curses extension to use ncursesw some time ago and it would not work correctly with ruby 1.8.
The reason is that ruby relies on the locale to be "C" internally and sets locale that way. The ncurses library in turn relies on correctly set locale to determine the encoding and string length. With C locale it will report the length of the string as the number of bytes. Version 1.9 of ruby now sets locale correctly so an extension linked with ncursesw should also work correctly without any use of the wide API, just using the standard API with strings in current locale (even UTF-8) should work. It required quite a few fixes to the interpreter to allow for this, though. Until these are backported to 1.8 utf-8+ncurses+ruby1.8 is not going to work well I suspect. Thanks Michal -- To UNSUBSCRIBE, email to [email protected] with a subject of "unsubscribe". Trouble? Contact [email protected]
