The implementation of new features and improvements for each software subsystem must be in close coordination with observatory milestones, the need to rapidly respond to operational issues, regular maintenance activities and testing resources available to verify and validate new and improved software capabilities. This paper describes the main issues detected managing all these factors together and the different approaches used by the observatory in the search of an optimal solution.
In this paper, we describe the software delivery process adopted by ALMA during the construction phase and its further evolution in early operations. We also present the acceptance process implemented by the observatory for the validation of the software before it can be used for science observations. We provide details of the main roles and responsibilities during software verification and validation as well as their participation in the process for reviewing and approving changes into the accepted software versions.
Finally, we present ideas on how these processes should evolve in the near future, considering the operational reality of the ALMA observatory as it moves into full operations, and summarize the progress implementing some of these ideas and lessons learnt.