Cool, but it seems slightly broken (maybe in transfer?). The two last characters (before the bang) in the output gets garbled (the first of them seems to be chr(204.25), which leads me to believe the $> part is broken).