when studying mathematics and computer science i would handwrite algorithms, draw all psuedocode on paper first, and reconstruct steps from published papers, all extremely useful