April - XP introduced to the AVP. I explained XP and he seemed interested in leaning more. I gave him my copy of Kent's book (XP Explained). He read it and came back ready for more (a good sign). "How can we do this?" was his first question.
Early May - He asked me to get more books for the staff and a copy of Planning XP. Everyone has the buzz now and the responses run from skepticism to when-can-I-start.
May 24 - Final planning meeting with all managers to get started. The final to do list has been compiled and all senior management is on board. I have a lot of training to do. - marc
June 30 - Much has happened in the last month and a week. We have trained "customers", developers, project manager, the project office manager, the business analysts, and the other assorted managers in the ways of XP. We have established focus teams on areas of the application. Work rooms have been setup for the developer pairs and iteration planning protocols have been hashed rehashed and, generally, are being followed. Daily standing status meetings are held and time boxed to 15 minutes.
We have encountered some classic problems though. The customers have a history of being burned in the past. They are leery of the "make the bricks first, then build the wall" metaphor. Problems arise when they focus deeply on a single brick ignoring the fact that, in the end, the house will have thousands of bricks. We are going into some retraining with the entire team to improve stories and acceptance criteria. (This was also a problem with the first focus team. They presented stories without knowing how they, the customers, would accept them.)
Against my better judgment, 3 parallel tracks were slated. After the second track started (staggered from the first), we hit the first collision. The second track had a dependency on something that was not a priority in the first track. I have asked for a reintegration of the tracks for iteration planning. The triumvirate (me as application architect, the data architect, and the AVP) has agreed to this. This has to be sold to the rest of the team. The pros are quite clear though, so I anticipate no real resistance.
The most important item is that the entire team is still on board. There has not been any mutinous chatter. The developers are getting the hang of pairing. The customers will get better at stories. Short iterations are still too new though as is the notion of an evolutionary application (burned customer history referred above.) In short, so far so good... - marc
This page mirrored in ExtremeProgrammingRoadmap as of April 29, 2006