Reversing the reasons for changes by a-posteriori recording of decisions made

By the need for quick adaptations to meet market demands and recurring changes of the requirements many software organizations end up with maintenance processes that consume enormous resources. Frequently, such maintenance activities are performed at the code level while architectures and design documentation are not aligned to the changes made. Hence, as the system evolves, these documents are not updated accordingly. Neither the decisions that led to change at the architecture level are recorded. In order to avoid knowledge vaporization we need to apply appropriate methods and tools able to support architectural design decisions as first class entities. This work analyzes how to improve evolution activities of architectures from the results of a case study, which combines two different tools: one for recording architectural design decisions and the other a reverse engineering tool