Poor support for unicode file names
Recently, I created a converter that transforms ASCII texts into characters from the extended range of UNICODE that resemble letters, and I've been playing with it in various contexts. On Linux (which, like almost all modern systems and applications, operates using UTF-8 encoding) my characters are accepted and displayed correctly. I can name a file using such characters, and everything works and displays properly. Example: ᗯᓰᑎᔑᑕᑭ.Ꭲ᙭Ꭲ. Similarly, in the WinSCP interface, such filenames are displayed correctly, just as they should be. Unfortunately, if we try to rename a file from within the WinSCP interface to one containing UNICODE characters, we end up with a string of question marks ('?').
