8 ms·

> "do you think repeating this across files will be problematic?" If you write that you clearly think it's problematic but hide it behind an unnatural question

> "do you think repeating this across files will be problematic?" If you write that you clearly think it's problematic but hide it behind an unnatural question. Then either I do what you hint at or you will push / discuss until I agree. I am not the determiner. It sounds like you are schooling me with an awkward back and forth. Honestly, to me this is beating around the bush and doesn't actually make me confident we can have a real discussion (in the context of that code review). "Do you perhaps think that doing this <tylically discouraged coding practice> is problematic?" sounds like either you want to passively insult me, or you don't consider me a peer with whom to have an actual discussion on equal footing, or you don't have the guts to share your opinion without contortions.

ysavir 4y ago
The wording above could be improved for sure. But the intention is to shift focus away from telling somehow how to solve a problem to forming consensus on whether there's a problem at all. And that starts with the reviewer expressing a concern rather than immediately providing a solution. "Is a change needed?" should come before "this is a needed change". A better example might be "I'm seeing this same code in various places--do we benefit from keeping them distinct?" or "you took an inheritance approach here, but I'm concerned about the long-term impacts as the codebase changes. How do you think this code compares with a composition approach?". If in doing so you would think that I'm being sly, that's unfortunate, but there's not much I can do in how you interpret stuff. I can't control for others assuming bad faith. And if there's bad faith, there are probably communication and/or trust problems amongst the team members that are broader than code reviews, and is something that should be tackled directly.

ysavir 4y ago

The wording above could be improved for sure. But the intention is to shift focus away from telling somehow how to solve a problem to forming consensus on whether there's a problem at all. And that starts with the reviewer expressing a concern rather than immediately providing a solution. "Is a change needed?" should come before "this is a needed change". A better example might be "I'm seeing this same code in various places--do we benefit from keeping them distinct?" or "you took an inheritance approach here, but I'm concerned about the long-term impacts as the codebase changes. How do you think this code compares with a composition approach?". If in doing so you would think that I'm being sly, that's unfortunate, but there's not much I can do in how you interpret stuff. I can't control for others assuming bad faith. And if there's bad faith, there are probably communication and/or trust problems amongst the team members that are broader than code reviews, and is something that should be tackled directly.