I wonder how much refactoring is _not_ done because of the need for a code review. In the era before code reviews, I used to go to town on my codebase when I saw something I wanted to fix. Are folks more hesitant to make such changes now because they are afraid of the burden on the code reviewer?
Not really. But if I start doing a major refactoring while working on a new feature, I’ll generally cut a new branch, move the refactoring over to it, and do a PR of just the refactoring. That makes it easier for the reviewer to only have to look for degradation of current functionality. While the refactoring is being reviewed, I’ll then usually start working on the new functionality on top of that branch.
It can get a little unwieldy if I’ve got a lot of refactoring going on which I currently do. I inherited a codebase that never said no to any “hey. That’s a neat blog post. Let’s try that way for this feature”.
