What are some common pair programming techniques and when to use them?

1 Driver and Navigator

The most basic pair programming technique is the driver and navigator. The driver is the one who writes the code, while the navigator is the one who reviews the code, suggests improvements, and checks for errors. The driver and navigator should switch roles frequently, to avoid fatigue and boredom, and to share knowledge and perspectives. This technique is useful for tasks that are well-defined, straightforward, or familiar, as it allows for fast and accurate coding.

2 Ping Pong

The ping pong technique is a variation of the driver and navigator, where the roles are switched after each test or feature. For example, the driver writes a failing test, then passes the keyboard to the navigator, who writes the code to make the test pass, then writes another failing test, and so on. This technique is useful for tasks that are test-driven, iterative, or creative, as it encourages feedback, communication, and innovation.

3 Strong Style

The strong style technique is another variation of the driver and navigator, where the navigator has more control over the code direction, and the driver acts as a scribe. The navigator tells the driver what to type, using high-level instructions, and the driver asks questions or clarifies details. The navigator should avoid micromanaging or dictating every keystroke, and the driver should avoid guessing or deviating from the instructions. This technique is useful for tasks that are complex, unfamiliar, or challenging, as it allows for knowledge transfer, problem-solving, and mentoring.

4 Remote Pairing

The remote pairing technique is a way of doing pair programming when the developers are not in the same physical location. It requires some tools and practices to make it work smoothly, such as a reliable internet connection, a shared screen, a video call, a chat app, and a code editor with collaboration features. The remote pairing technique can follow any of the previous techniques, depending on the task and the preferences of the pair. This technique is useful for tasks that are distributed, flexible, or diverse, as it enables access to different skills, backgrounds, and environments.

5 Solo Time

The solo time technique is not exactly a pair programming technique, but rather a complement to it. It involves taking some breaks from pairing, to work on individual tasks, such as research, experimentation, refactoring, or documentation. The solo time technique can help avoid burnout, frustration, or dependency, and can also foster autonomy, confidence, and reflection. This technique is useful for tasks that are personal, optional, or independent, as it allows for self-paced learning, exploration, and improvement.