In my What I Want to Accomplish at Agile 2005 post, I outlined my goals for the week in Denver.  Now that the last session has been held, the last beverage drank at the banquet, and I'm back in Minnesota, it's time to review those goals and see if they were met.

  • <Goal>Continue to evolve my thinking on being Agile AND Plan Driven by attending several sessions from the Agile Project Management track.

It was very enjoyable for me to hear from some of the leaders in the industry and I definitely gained some new insights.  But I also was able to reaffirm that what we are doing with wrapping agile with release planning and other structure is the right approach.  I am interested in trying out user stories, story points for estimating, value stream mapping for certain processes, and continue to work on getting the right level of ‘precision’ in our adaptive planning efforts.

  • <Goal>Determine how a Software Test organization can best add value to an Agile process.  

While there were several testing sessions at the conference, my sense is that the role of QA / Software Test is still evolving.  A useful tool from the conference is Brian Marick’s quadrant for classifying different types of tests.  In the quadrant, a type of test will be either Business Facing or Technology Facing, and it will either Support Programming or Critique the Product.  Just like the rest of the industry, we need to continue to evolve this role.  The two testers from MBF-Fargo attended many of the Test sessions and are bringing back some ideas to try.

  • <Goal>Learn about breaking large features into bite sized chunks. 

One breakthrough for my thinking on agile over the course of the conference is that I believe that reducing the ‘chunkiness’ of our work is the key for us to become more agile.  Small chunks optimize flow (from Lean theory), optimize upfront design and requirements efforts, support TDD for developers, allow QA to work closely in parallel, and enable early customer feedback (the best reason). 

I am the ScrumMaster for a team that is building a sample application on MBF.  The nature of the features for this team is that they are small increments to the product.  We work in two week sprints.  The design documentation is lightweight and reviewed with other teams efficiently.  The software test engineer is able to work in parallel and deliver within hours after dev has checked in code.  I believe an important key to the agility of this team is the size of the work increments.

Doing evolutionary design and incremental design is a challenge when building a framework like the majority of the team is doing, however.  A key goal of a framework is the stability of its interfaces.  Even before wide scale release, interface churn can be expensive.  The conference gave me some ideas to try out, but breaking our work into smaller pieces will be an interesting challenge.

  • <Goal>Hang out with other agile advocates from Microsoft. 

This was a highlight of the conference for me.  There were several other team members from my division (Developer Division) in Redmond with whom I was able to discuss common planning and development issues.  There was also a large contingent from MSN, as well as other groups.

It was also enjoyable to interact with other conference attendees in the different sessions.  I also had the opportunity to go out to dinner with over 20 Minnesotan’s on Wednesday night.  I think the premise was to discuss Agile 2006 in Minneapolis, but only socializing got done!

  • <Goal>Go see a Rockies game at Coors Field.

Mission accomplished!!  We also went to some nice restaurants and got to explore downtown Denver some.

Overall, it was a very enjoyable and educational week.  We have some ideas to try and hopefully they will prove beneficial.  The beauty of agile approaches is the ability to quickly try something and evolve as you get feedback from the practice.  I'll try to post some updates with the results as we try out some of the practices from the conference.