Problem replacing characters

Question

Hi,
I'm trying to replace a # with a � in a unix environment using a String.replace, but unix won't recognise the �. It just prints out a ? if I print to screen or a \243 if I print to a file. If I move this file back to windows, it contains ��. I'm very confused. Please help.
Thanks.
Oh, here's a copy of the test code of where I'm upto at the moment after alot of messing about. Feel free to disregard this code if I'm barking completely up the wrong tree.
    public static void main(String[] args) throws Exception {
        StringBuffer stb = new StringBuffer("#1234");
        int indx = -1;
        while ((indx = stb.toString().indexOf("#")) != -1) {
            ByteArrayOutputStream baos = new ByteArrayOutputStream();
            baos = new ByteArrayOutputStream();
            baos.write((new String("�").getBytes("UTF-8")));
            String st = baos.toString("UTF-8");
            System.out.println("st = " + st);
            stb.replace(indx, indx+"#".length(), st);
        System.out.println("buffer: " + stb.toString());
        //this is just here to see what happens if I write
        //the � to a file
        FileOutputStream fos = new FileOutputStream("file.txt");
        fos.write((new String("�").getBytes("UTF-8")));
        fos.flush();
        fos.close();
    }

Answer 1

For character replacement, it's more efficient to use:
str = str.replace('#', '�'); // where str is a string
You don't need UTF-8 encoding if the Unicode characters are in the ANSI range: U+0000 - U+00FF. Just use the default encoding. When you view the file, make sure select a font that has the glyph for that charater. On Windows, Tahoma, Courier New, or Times New Roman would suffice.
For those beyond ANSI, then UTF-8 comes into play. Use OutputStreamWriter(fos, "UTF8") when write to file. Check http://java.sun.com/docs/books/tutorial/i18n/text/stream.html for usage. Win9x/Me Notepad can't read UTF-8 encoded files, but WinNT/2K/XP can.