Often you want to fix something quickly but run into a stupid problem -- you don't have the right data locally, or the wrong version of something, or an unrelat

Often you want to fix something quickly but run into a stupid problem -- you don't have the right data locally, or the wrong version of something, or an unrelated thing is broken, or something is undocumented and the person who knows isn't immediately available. Thus you feel deflated, the energy is gone, and finally you move to something else. Instead, fix that stupid thing, once and for all. Get to the bottom of it, automate all the future times it would happen again, don't just hit it but hit it out of the park. Not always, because you won't do anything else, but often.