April 30, 2006

Pair Programming

Most of today and the later half of yesterday afternoon I spent pair programming. While I've collaborated for short periods of time with other coworkers, this is by far the first concentrated time I've spent doing it. Overall I'm inclined to believe most of what the advocates claim about pair programming. We created a lot of code and most of it worked on the first run. We didn't practice a TDD approach, but did make sure that good unit tests existed. Some of the simple issues that still came up even with two sets of eyes on the code were simple compile time errors like missing imports and changed method signatures on system APIs (the downside of working with a beta of Flex 2).

Unlike a "true" paired programming session, I didn't do much typing. I think this stemmed mostly from the fact that our development systems are laptops. It isn't as easy to slide the keyboard and mouse over to your partner in this case. Even though I wasn't doing the typing, I contributed to the code by helping spot potential issues (syntactic and strategic) while duplicating knowledge on some of the more potentially error prone plumbing of the system.

Tags: pair programming tdd work