There are many good reasons for turning co-located
projects distributed, likewise there are many good reasons for
turning traditional projects agile. In both cases there are many
obstacles to overcome and pitfalls to avoid and the
combination of agile and distributed does not make this
situation any better. In general Configuration Management
works as the infrastructure of any software project and its
concepts and principles have to be implemented in different
ways depending on the specific context. How to adapt
Configuration Management for an agile context is well
understood, how to adapt it for distribution is less understood
– and what changes and how to fix that when a co-located agile
team goes distributed is unclear. In this position paper, we
draw on our experience from academic and industrial agile
projects to give our opinions on what to look out for and
possibilities for solutions.