This section presupposes that you inevitably run out of tasks that are high-impact and easy, which hasn't been what I've observed. There are tons of little cod

This section presupposes that you inevitably run out of tasks that are high-impact and easy, which hasn't been what I've observed. There are tons of little code functions that can be tested, released via open source / internal libraries and add lots of value. I'm constantly creating new open source projects with little functions that get lots of adoption. Just added some code to validate a SQLite schema in a Pythonic manner for example. Nothing groudbreaking, but makes it easier for the folks that need to do this in the future. The Unix-philosophy is writing programs that do one thing well. Lots of little functions can be combined to build something great. Lots of "snacks" can be used to make a feast!