...and so it ends. Day 4 is all wrapped up and I am very glad that I attended.
Brad Wilson, Peter Provost and Don Smith have an enthusiam that is just infectious. You can just tell they love what they do and the best thing is they have an incredible depth of knowledge to back them up.
Scott Hanselman and Ted Neward are fun to watch, their cocky arrogance makes you just love them and hate at the same time. The quote of the summit for me was Scott's line to an attendee's question about if their methodology was similar to one of Fowler's patterns "I don't read Fowler... I use success as my metric". I have great respect for Fowler, but Scott's pragmatism was a great reminder that a balance between theory and practice is essential.
Technically, I am taking away from this conference a much stronger motivation to make TDD work for me. Doing TDD, espescially on an existing codebase is hard, it forces you to do a huge amount of decoupling, which in many cases is a good thing, but it does introduce complexity where sometimes it isn't needed. The micro-pairing talk was phenomenal. I really do believe that it would have hugely beneficial effects to both the quality of the code and the productivity of the developers. A number of attendees seemed very concerned about skill level mismatches between the developers, but I personally think that the majority of experienced developers are happy to pass on their knowledge to a junior developer and nurture that developer. As long as the junior developer is intelligent, I don't see a problem. I think the problem comes with idiots and cowboy programmers, both of which are a huge problem in a regular development process anyway.
It is unfortunate that the presentations on the software factories were not much more than walkthrough's of the basic usage. That is information that you can pick up pretty easily from the available guidance on the internet. Chris Tavares' presentation, however, was exactly the kind of insight that I think these kind of conferences should provide. I can justify spending thousands of dollars when I get information and insight that just is not available anywhere else.
There were a couple of annoyances that are worth mentioning. Rick Magire did his keynote on Wednesday and didn't really leave much time for questions, but said that he would be around later to answer questions one on one. I never saw him again. Jack Greenfield did the keynote today and left no time for questions. I could have asked at least five and I am sure there were 100 other questions waiting from other people. And finally, who put Michael up there to demo Web Client Factory! He is obviously one of the less experienced presenters, demoing an alpha product that had been built 30 minutes previously. As could be seen from the later questions, he did a tremendous amount of damage to P&P's product credibility and you can be sure there are going to be a whole lot less of us looking at that factory after today.
That does lead me into one other concern that I have with P&P. There seems to have been an unusually high number of exits from developers in the P&P team. Brian Button and Scott Densmore have moved on, Brad Wilson, Jim Newkirk and Jonathan Wanagel are now on Codeplex. Jason Hogg has moved on, Peter Provost is now a program manager. Who are the developers that remain?
I would be remiss without finally thanking Keith Pleas for the excellent job he did in running the summit. From my perspective the whole event ran flawlessly. I'm sure that's not the case but Keith obviously handled any issues sufficiently so as not to impact the experience.
Here's hoping this won't be my last P&P summit and I will end with links to a couple of other blogs from guys who did a much better job of recording what actually when on at the event...