Implementation of Software Configuration Management Process by Models: Practical Experiments and Learned Lessons

Abstract

Nowadays software configuration management process is not only dilemma which system should be used for version control or how to merge changes from one source code branch to other. There are multiple tasks such as version control, build management, deploy management, status accounting, bug tracking and many others that should be solved to support full configuration management process according to most popular quality standards. The main scope of the mentioned process is to include only valid and tested software items to final version of product and prepare a new version as soon as possible. To implement different tasks of software configuration management process, a set of different tools, scripts and utilities should be used. The current paper provides a new model-based approach to implementation of configuration management. Using different models, a new approach helps to organize existing solutions and develop new ones by a parameterized way, thus increasing reuse of solutions. The study provides a general description of new model-based conception and definitions of all models needed to implement a new approach. The second part of the paper contains an overview of criteria, practical experiments and lessons learned from using new models in software configuration management. Finally, further works are defined based on results of practical experiments and lessons learned.