Writing Software With the Grain

My JavaWorld article, Writing Software With the Grain: Exploring the Humanity of Agile Methods, is up! This is my first publication in anything bigger than a blog.* Dorky as it is, I’m pretty pumped!

The article explores several common Agile practices—to wit: user stories, lightweight documentation, and iterations—that happen to align well with normal human behavior given the kinds of creatures that we are. There is, you might be thinking, a host of philosophical assumptions that underlie this sort of reasoning, which I would be only too pleased to discuss over drinks sometime if you’re local.

I had to pick a limited number of Agile practices for analysis in the article, but I think there is more to the story than the three I’ve identified. A good example would be Extreme Programming’s Sustainable Pace. The opposite of Sustainable Pace is the discipline of death-marching developers just hard enough that they don’t quit, but hard enough that we squeeze maximum productivity out of them. If developers were impersonal units of software production akin to machines, this would be the right way to manage them.

The most obvious objection to this is the law of diminishing returns, or the idea that each marginal hour worked is an increasingly fatigued, and correspondingly less productive, hour. While there is certainly some truth to this, I think we’re forced to grant that we actually do make more progress when we work more hours. Everybody knows crunch time, and every honest developer is forced to admit that the crazy schedule of the final push to the finish actually make a difference—at least up until the time the hallucinations start, which in itself can be kind of fun, as long as it’s your coworkers doing the hallucinating.

The best objection to overwork isn’t a utilitarian one; rather, I would argue that the notion of Sustainable Pace is grounded in the right relationship between work and the other activities in human life given the kinds of creatures that human beings are. It is good and right for us to spend a certain amount of our time following some kind of economically productive vocation out in the commonwealth. It is also good and right for us to spend some amount of time with friends or family, some amount of time sleeping and eating, some amount of time standing in line at the DMV, some amount of time inside, some amount of time out-of-doors, etc. For most people, work might take up a plurality of the pie, but it should not be a significant majority. It is in our nature to spend our time in a variety of roles activities, of which work is but one part.

“The Humanity of Sustainable Pace” could definitely use a more detailed treatment than I’ve given it here, but my point is to illustrate that there is more to the topic than my article covers. I’d love to see the community do some more thinking along these lines. The trick is to realize that human beings are a particular kind of thing—we implement a common base class, if you will—so the practices and methods we develop in our vocation should respect that class’s semantics. The notion that we are uniform production units that can be squeezed into whatever organizational structure and management schema we choose is destructive not just to the ontology of the human person, but also to the well-being of the people on our teams—people who may well be our friends, and people whose healthy vocational context is certainly our responsibility.

So let’s think some more along these lines. Agile is doing great things, and hopefully more great ideas can follow from it.