i keep seeing reviews where people are very active and not actually useful. lots of comments, lots of thumbs-up, not much challenge to the thing that might break later.
maybe that’s just what happens when teams start treating review as a social ritual instead of a reliability check. how do you tell the difference between a busy review and a good one?
A busy review is one where the comments mostly don’t create any “work” for the author—no new tests, no clarified assumptions, no edge cases forced into the open, just vibes and formatting. One trick I’ve used is to look at the diff after review and ask: did we add at least one thing that would’ve caught a future bug (a test, an assertion, a log/metric, a guardrail)? If the answer is usually “no, ” the team has accidentally optimized for being seen participating, because that’s lower friction than being the person who says, “This will page us at 2am. ” Do you have any norms around “must ask one risk question” or assigning a single “designated skeptic” per PR, or is it totally free-for-all?
:: Copyright KIRUPA 2024 //--