In this light, it would be interesting to see a control study that focuses only on anonymized contributions to open source repositories, where one’s demographics are completely invisible. If we also see people from marginalized groups self-reporting greater pushback at similar rates, it would show that it is simply a difference in perception, not actual discriminatory behavior from the reviewer.
[0] https://stackoverflow.blog/2018/04/26/stack-overflow-isnt-ve...: “But how do we really know that too many developers experience Stack Overflow as an unwelcoming or hostile place? Well, the nice thing about problems that relate to how people feel is that finding the truth is easy. Feelings have no “technically correct.” They’re just what the feeler is telling you. When someone tells you how they feel, you can pack up your magnifying glass and clue kit, cuz that’s the answer. You’re done. And a lot of devs feel like Stack Overflow is an intimidating, unwelcoming place. We know because they tell us.”
I think this is quite the possibility like the child who gets picked on by parents and siblings feels like when strangers critique them it’s part of the same dynamic when it isn’t.
Everybody has one; everybody has a story. It's endemic.
I propose each time a code reviewer completes their comment they hit 'approve' and move on. Leave it to the engineer to decide what changes to embrace and which are noise. They are, after all, the expert on the code they changed.
- Don't block the commit. Review as soon as possible.
- Focus on finding bugs.
- Ask questions if you don't understand something.
- Sometimes I make suggestions.
- Before requesting a change I ask myself. "Is it wrong or is it just not the way I would do it".
But it only takes one person to ball up the whole process. Worse if they're some self-important early hire, promoted above their skills and full of idealism. Nobody will tell them to stfu. The tools enable them, since the process won't proceed until they hit a magic button.
I've suffered under this special kind of gatekeeping autocrat. My son has on several occasions. We've both left jobs in part because of it.
Nitpickers are the worst. Before someone replies with "use linters" and other such obvious things, I'd like to ensure that nitpicking often goes beyond automation. Nitpicking is imposing preferences over others. If a language or infra supports something, nitpicking one vs another is often futile in the scheme of things.
I would strongly encourage code reviewers to self reflect and see how many PRs reviewed by them actually end up in production within a day. If you don't have many, you are the problem.
My objection is to the gatekeeping self-appointed code hero that has some arbitrary plan. One they probably don't adhere to themselves, but insist everybody else toes the line. Don't free objects that way! Don't test for null like that! I know a better way I heard of in a blog or a meme or on a message board, and I will delay the project and burn runway to make myself feel important!