>>878 Right now I'm reading this series on Elliptic Curves. At the last lecture it gives an intelligible argument for Wiles' 5-3 trick.