Pair Programming is Pair Development

Steve Hayes writes about the transition his team made from pair programming to pair development. I feel this to be an important distinction, so I'll use it as my basis for defending pair programming.

The main argument used against pair programming is a supposed drop in productivity. Some argue against this by trying to show that a) the productivity doesn't drop, b) the resulting quality is much better. I'll try a different line of attack.

What if the concept pair programming is misleading? Let's see the main activities while you're developing (very simplistic picture):