Thursday, June 13, 2002 | |
Some of my most interesting referrers are the ones from Google. Here's some topics that have come up in the past couple weeks:
6:25:11 AM permalink
|
Maybe this is why you hear analyses of XP that say it works best with internal projects, because the customer and the programmer ultimately have the same boss. The article says "Communication is certainly the key component of any XP project", but I'd argue that it's the key component of any project. It's just that different projects have different modes. Alistair Cockburn describes this well in his Agile Development book. For example, the defense industry relies heavily on written communications, and RUP shares this reliance. Prose is subject to misreading and lack of specificity, so people develop things like UML to make communications specific and less prone to interpretation. XP says that this is an unnattural method of communication and goes the other direction towards communication channels that even laypeople can understand. I've worked with people who insisted on either entirely verbal or entirely written communication. Either mode is maddening; with the "talkers", everything's in the air and nothing gets nailed down. With the "writers", communication is agonizingly slow. "Writers" are often "drawers" as well, and with the "drawers", there's the additional overhead of negotiating the language. Even with UML, people understand different dialects, the language is constantly evolving, and furthermore, people have different levels of understanding of the language. I've had situations where I've wanted to work with a "drawer" and tried to express a design idea where neither of us knew the UML notation to express it. Thankfully, we both spoke English! So when you have two people who communicate expertly using, say, English, yet have less facility with UML, why struggle with UML when you could be talking? 5:09:29 AM permalink
|