On April 2nd, the submission system for the Agile 2007 Conference sent out the acceptance and rejection notices to those who submitted proposals for conference sessions. Initially, some people received over 200 e-mail messages containing notifications for submitters other than themselves. Later in the day, the problem was identified and corrected. While this would seem like an embarrassing hiccup in most cases, this problem occurred in the software used by a conference promoting agile methods - methods that emphasize software quality!
Joshua Kerievsky called out the conference organizers about the level of testing of the submission software, which ruffled some feathers.
I have to agree with Joshua. We talk the talk about how software built using agile methods is of much, much higher quality, but then the Agile 2007 organizers farm out the development of the submission system to a group that, according to Ron Jeffries, didn't use agile at all. The debate then became centered around whether the debate was appropriate for the XP Yahoo group or not, even involving the moderators.
Well, I'm sorry, but I believe it was very appropriate. We can't just pick and choose when we should be agile. Like it or not, the organizers of a conference espousing agile methods have to hold themselves to a higher standard. Anything less diminishes their ability to talk about the benefits of agile development, and damages the credibility of all of us that do.
For my part, while I can't participate in the organization of next year's conference, I have volunteered to work on the submission system. I have written e-mail distribution code before... with tests. If there are problems with that system, I fully expect to be called out for them.
That's called accountability, which is what we've been saying for years is another aspect of development that's emphasized by XP.