> For example, why would I want to spend time learning deeply how to write my own compiler, vs just spending that time to do tutorials for a couple different ne

> For example, why would I want to spend time learning deeply how to write my own compiler, vs just spending that time to do tutorials for a couple different new frameworks? I feel I can answer that easily: Writing a compiler is vastly more interesting and satisfying! Learning new frameworks, how boring. I'll do it for pay or prospects. Writing compilers, now that I'll do for free on the evenings and weekends. You cannot pry my fingers from the keyboard when I'm on a compiler roll! Important household tasks will go neglected! Someone should remind me to eat! I feel I speak for many here on HN with this sentiment. (See also database engines and distributed systems.) To put it back on topic: Some things are intrinsically rewarding. It feels good to go deep with them, especially to advance them in some way. That's the payoff. I think this is a significant driving factor in the expansion of human knowledge; it's not just driven by economic incentives to learn things, although economic limitations do prevent people from having capacity to go deep into things they'd like to