Would pair programming with frequent pair rotations be better than Code Readings? It seems you get all of the value of Code Readings, but throughout the entire process.
That definitely seems possible, but I feel it might be harder to implement at an organization that wasn't already doing a lot of pair programming since in my experience pairing elicits almost philosophical support or opposition. Maybe you could start with readings and work up to pairing if it felt like the right transition?
2 comments
[ 2.7 ms ] story [ 17.6 ms ] thread