Ah yes. We really need some benchmarks, so we can choose CPU's according to how many colorschemes for Neovim it can generate each second. Parallelizing @ optimi

Ah yes. We really need some benchmarks, so we can choose CPU's according to how many colorschemes for Neovim it can generate each second. Parallelizing @ optimizing that task for supercomputer deployment is of the utmost importance. /kidding Read: it depends on # of users, and how often a task is done. For one-offs, or rarely done jobs, performance or energy use doesn't matter AT ALL as long as it's useable (like, user can work with it, and software does its job in acceptable timespan). Job described in the article clearly falls in this category. On the other end of the scale: if software is used daily by billions of users, then even shaving off a single CPU cycle, or a minor UI enhancement should be considered worthwhile. Most software falls in between. Optimize performance as needed, or with reasonable ratio for effort vs. userbase benefit.