UTF-8 support on Windows and Linux

Hi All,

I am using perl5.6.1 both on Windows 2000 and Mandrake Linux 8.1.
On Windows the UTF-8 strings are handled in a reasonable manner, that is
length() returns the number of characters and /./ matches a single
character. On Linux though, length seems to return the byte length of the
string, and /./ fails to match characters over 127. This happens even if I
use a UTF-8 locale.
How come?
What can I do about it? Is anybody experiencing similar problems on Linux?

Gregor