> he accidentally committed a horrible piece of code that worked great on the staging server for a few days, but when it was auto-updated to the production hard

> he accidentally committed a horrible piece of code that worked great on the staging server for a few days, but when it was auto-updated to the production hardware devices, it broke them all this anecdote says more about their pipeline than some random engineer. it shouldn't be possible for code to reach customers without multiple engineers becoming accountable for it via reviews.