One of the things that the C# team does occaisionally is hold an appweek, where we take a week out of our schedules and devote it solely to using our product to build apps. For the last week or so, there’s been a QA appweek that the PMs have been participating in, and we’re organized into 8 or so teams.
The reason I called it “appweak” is that I’ve only been able to spend, at best, 50% of my time devoted to appweek, which means my contribution has been fairly modest. Our team has been using the profiling apis to do some program visualization (function stack, exception behavior, memory behavior, etc.). Other teams have their own projects, with at least one team doing a game (a Berzerk clone, IIRC).
Why do I mention this? Well, to explain a bit how we test our products, and to tell you that I think you’re going to be pretty happy with Whidbey – it has some nice advances that make writing code a fair bit easier.
[Update] Jack asked how we decide what kinds of apps we write, and whether we target specific feature areas.
The apps are nearly always chosen by the teams. Sometimes we give them guidance on what kind of app they write, but my experience is that it’s better to just let the teams guide themselves. Our goal is to use the product the way we expect our customer to use it, which for us meant using source code control, the new build system, and training ourselves to use refactoring