You remember your first time, right' The very first time that you tried out a new project methodology' Well, I had my first time with Scrum today. And it went pretty damn well. :-)
Scrum is very new to TG. You might say that we're TG virgins. And that's not just the developers ' that's everyone. Scrum is being implemented from the bottom-up which means that we as developers (and me as a developer manager) needs to learn Scrum while also teaching our various project managers and product managers. In order to do this, you need a large amount of buy-in from the powers that be. Luckily, I'm one of the powers and I really don't care what the other powers think. Even more lucky is that the other powers don't really care what I do as long as 'it works'.
Prior to today's meeting, there had been a quick one-hour meeting with the stakeholders (let's call them product owners), the development team and the QA team (let's call these folks the developers). During that meeting, the primary focus was teaching the product owners how to come up with their stories. They seemed a little wary that we weren't asking for strict requirements. It almost seemed like they were a bit afraid to open up and ask for what they really wanted. But after an hour, we had 8 or 10 stories kicking around.
Today's meeting was a 250 minute blockbuster that followed up on these stories. The goal was to chat about the stories to get a better idea what the product owners were looking for. The developers asked questions to clarify their understanding. After about 2 hours, we had adjusted some of the stories and consolidated the list down to 6.
Now that we had our stories, the developers got to work. They created estimates on the stories in 'story points'. Basically, they picked a story that was small and relatively well known to them. They arbitrarily gave that story 2 points and used it as a reference for estimating the rest of the stories. For those of you rolling your eyes at the 'story point' terminology, I have to say that I was pleasantly surprised at how well it worked. Using these points, the product owners were able to prioritize their stories and start to fit them into our iterations.
Now that we know the stories that we'll be working on in the first iteration, the developers will break that story down into tasks and estimate them with 'ideal hours' i.e. the time you'd have if the phone didn't ring, people didn't bug you and meetings didn't chew up your time. I'll track the tasks in our daily 'scrum' meeting and keep a constant list of how much time we have remaining in the project.
This is most definitely an experiment and something that we'll have to get better at through trial and error. Already I see a problem that we've really only got one story assigned for the first 1'week iteration period. We'll see how we can get better at that in the future.