Friday, November 08 2002

Rachel Ross hasn't grasped the concept of pair programming at all. She writes about it in her article 'Pair programming' an ill-fated shotgun wedding [found via Jonas Bengtsson].

First of all, pair programming isn't a subset of XP, it's a requirement. Or rather, the effects of pair programming - its benefits - as described in the XP literature is what's required. Rachel hasn't grasped this at all.

Perhaps if you read her arguments against pair programming outside the scope of XP, you might say she makes a few good points, but I disagree. And she is writing about pair programming in the context of XP. As such, she's just plain wrong and she displays an incredible amount of ignorance on the subject. Read your homework Rachel.

... forcing two great programmers to work ...

... when you force two people to work together so closely ...

XP isn't about force. You can't possibly succeed with XP if you try to force it onto a team. XP can only be adopted if mutually agreed upon. Likewise, pair programming - as a requirement for running an XP project - can only be adopted if mutually agreed upon.

If a team trying to adopt XP chose to exclude a specific practice, that team will not succeed if it does not find a suitable substitute for the dropped practice. This of course includes the practice of pair programming.

Most of the arguments against pair programming that Rachel presents in her article rely on the assumption that the practice of pair programming is being forced onto a team by the management. This of course automatically voids them all.

Another passage in her article that I just can't leave without commenting is It's too bad the entire concept works against human nature ...

What the heck is that about? People don't work together? Humans don't live in flocks, doesn't work together in groups? She's saying that humans are all lone wolves?