Both TIS and Shenzhen are cases where I'd say make something ugly that works, then go back and fix it hours later when you have new tools and experience. IE, t

Both TIS and Shenzhen are cases where I'd say make something ugly that works, then go back and fix it hours later when you have new tools and experience. IE, the very first puzzle gets an easier solution once you learn about a hidden command. There's satisfaction to leaving a solution to come back days later and go "whoah, this better solution is so obvious" which is easy to miss when you're stuck on it for hours.