Patterns for Retrospectives
OOPSLA'03 workshop

This workshop was held at the OOPSLA'03 conference in October 2003.
Results of the workshop are shown below.

Organizers:
Linda Rising, Consultant, risingl@acm.org
Mary Lynn Manns, University of North Carolina at Asheville, manns@unca.edu

The theme of this workshop was a pattern language for retrospectives. Retrospectives are important, not only for object technology, but for any human activity -- to provide time for reflection, to improve on performance, and increase our understanding of the task at hand. The Principles Behind the Agile Manifesto state that, "At regular intervals, the team reflects on how to become more effective, then tunes and adjusts its behavior accordingly." There have been articles and an excellent book (Norm Kerth's Project Retrospectives, 2001) written on this topic but as is the case for many domains, experts know much that novices don't. We believe that documenting expertise in the form of patterns is an excellent way to share knowledge. The goal of this workshop was to begin documenting patterns in conducting successful retrospectives and looking for a pattern language structure.

Participation

Most participants submitted a brief description of a pattern for running a retrospective. They considered problems they have encountered in preparing and conducting retrospectives. For example, a draft follows:

= = = = = = = = = = = = = = = = = = = = = = = = = = = = = =

Every Opinion Counts

Context: The retrospective has many different people with different roles and views.

Forces:

  • The way we are colors what we see. We have our own filters and point of view. The way we interpret what we see is influenced by our experience.
  • No one knows the whole story of the project, so we have to hear all the viewpoints in order to truly understand what happened in the project.

Problem: Everyone sees things a certain way. There is a tendency for us all to believe that the way we saw the project was the truth and anyone that disagrees is wrong.

Solution: Create an atmosphere in which every opinion counts.
The following technique can be used: Create Safety (Kerth 2001: 108)

Resulting Context: The details of the project emerge. Each member of the team has the complete story, rather than only his or her own perspective.

= = = = = = = = = = = = = = = = = = = = = = = = = = = = = =

We expected participants to be familiar with Norm Kerth's book (Project Retrospectives).


Planned workshop activities

  • structured activity to brainstorm ideas for patterns
  • discuss the structure of the pattern language
  • draft more patterns from the brainstorm session
  • workshop some of the patterns brought to the workshop and drafted during the session


Pattern Drafts and Workshop Artifacts


Post-workshop activities

The patterns and other artifacts from the workshop are posted on this web site for comment. The patterns will be submitted to a PLoP conference at the next opportunity and the work will be continued in workshops in subsequent conferences. The goal is to produce a companion to Norm Kerth's book that would describe patterns in retrospectives. This book will have pattern contributions from many different retrospective facilitators. The patterns and preliminary work at this workshop was also displayed at the OOPSLA '03 poster session.


Organizers

(Manns and Rising are authors of the book, Fear Less and Other Patterns for Introducing New Ideas into Organizations, due to be published by Addison-Wesley in 2004)

Linda Rising has a Ph.D. from Arizona State University in the area of object-based design metrics. Her background includes university teaching experience as well as work in industry in the areas of telecommunications, avionics, and strategic weapons systems. She has been working with object technologies since 1983. She is the editor of A Patterns Handbook, The Pattern Almanac 2000, and Design Patterns in Communication Systems. She has experience leading retrospectives in a number of companies and academic settings around the world, as well as giving presentations on the benefits of retrospectives. She has over 20 years of academic teaching experience and over 15 years of industrial training experience. She also has experience leading workshops in pattern mining.

Mary Lynn Manns has a Ph.D. from DeMontfort University in England in the area of software patterns. She is on the faculty at the University of North Carolina at Asheville, with over 20 years of teaching experience. She has taught seminars on project retrospectives and has led numerous retrospectives in both industry and academia. She also has experience leading workshops in pattern mining.