[hinkley]

Some of us still recall when code reviews were more of a concept than a practice you actually did at work every day.

One of the (unspoken?) tenets of XP is that if you can't have code reviews, break your code into small enough bits that you are comfortable defending each decision individually (rather than rafting them up and trying to get through 10 lines of gold and 20 lines of crap). The blast radius is tiny, and you can't as easiely confuse yourself into thinking buggy code is fine.

So there was a time where "how to do code reviews" meant "how do you cause code reviews to occur", not "what processes should you incorporate into your reviews". Maybe OP is caught in a time warp.