My programming partner Ward Cunningham taught me to avoid complexity. In spite of my blue-chip, Silicon Valley brat credentials, I was never a very good programmer. Ward has more programming talent than I do, but he still programs simpler stuff, not because he must, but because he chooses. That is a big part of my success to date—picking development priorities and ignoring interesting side issues.

This was my first technical article. I was lucky to write it with Ward, because he had a pretty good handle on how to focus an article. That was the lesson of this paper for me—focus. I can remember discussing for days what the one single point was we wanted a reader to take away. That was a powerful lesson, and a bit painful, too. Ward and I had been working on lots of exciting stuff. I wanted to talk about all of it. Ward leaned and leaned on finding the one point that stood above all others. Finally we hit on this one.

The article introduces the Cunningham Diagram, a diagram with much the same information in it as Jacobson's Interaction Diagram, but (to my eyes, anyway) much more artistically rendered. As far as impact goes, this paper was a dud. The two good ideas here—the diagram itself and a tool for constructing it from running code—both disappeared without a trace.