I don't understand the limits either. The code editor in TIC80 still counts each character as a token, and doesn't specify where the limit may be. I have no idea what "64k of compressed data" means in practice.
