This literature wishes to shed light on the benefits software development processes can have on research software teams. Specifically, the literature wishes to gain a better understanding of software process for research software teams by analyzing whether research software community demographic factors influence their use and perceived value of software processes.
Key Points
Past literature have shown that fields in science and engineering domains increasingly use software development processes -- those of which include physics, astrophysics, and biomedical imaging.
Past literature have shown that many research software development teams use forms of agile approaches even if they themselves are unaware of those approaches.
Past literature have shown that most research practitioners have little formal SE training and tend to be self-taught, code reviews and agile methods were rarely used which can suggest a lack of collaborative development practices.
Findings
Most respondents saw value in software processes, however, they rarely used process metrics to evaluate the success of software process. This could become difficult to ensure quality control and means that teams could be using inappropriate processes and inaccurate implementations or changes.
There is a found relationship between teams and individuals for their use of software research processes. If a team uses software processes, then individuals are likely to use those processes. However, if a team does not use the processes, the individual is not likely to use those processes.
There was a significant relationship between respondents perceived value of following. defined software development process and their likelihood of individually following a defined process. This could suggest that increasing the level of software process in research can help research software developers understand the value of software processes.
Generally, larger teams (5 or more) followed a more defined software process than small teams.
Teams who working on released software followed defined software processes more than teams who are working on unreleased software. Additionally, respondents that are working on released software value processes more than those working on unreleased software.
Citation
Nasir U. Eisty, George K. Thiruvathukal, and Jeffrey C. Carver. 2019. Use of Software Process in Research Software Development: A Survey. In Proceedings of the Evaluation and Assessment on Software Engineering (EASE '19). Association for Computing Machinery, New York, NY, USA, 276–282. DOI:https://doi.org/10.1145/3319008.3319351
Use of Software Process in Research Software Development: A Survey (2019)
Sumary
This literature wishes to shed light on the benefits software development processes can have on research software teams. Specifically, the literature wishes to gain a better understanding of software process for research software teams by analyzing whether research software community demographic factors influence their use and perceived value of software processes.
Key Points
Findings
Citation
Nasir U. Eisty, George K. Thiruvathukal, and Jeffrey C. Carver. 2019. Use of Software Process in Research Software Development: A Survey. In Proceedings of the Evaluation and Assessment on Software Engineering (EASE '19). Association for Computing Machinery, New York, NY, USA, 276–282. DOI:https://doi.org/10.1145/3319008.3319351