Software development challenges and best practices in an instrument data service
The Centre for eResearch at Waipapa Taumata Rau University of Auckland is developing an Instrument Data Service, intended to ingest, store and archive data from scientific instruments. I joined this team with the project already underway and spent a significant amount of time familiarising myself wi...
Gespeichert in:
Hauptverfasser: | , |
---|---|
Format: | Video |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext bestellen |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | The Centre for eResearch at Waipapa Taumata Rau University of Auckland is developing an Instrument Data Service, intended to ingest, store and archive data from scientific instruments. I joined this team with the project already underway and spent a significant amount of time familiarising myself with the code base, technologies and development practices. This talk covers some of the issues I encountered, and how we have begun to address them. Some are specific to academia, but many are applicable to software projects in other settings as well. In the field of technology generally, staff turnover is often high. It is thus very important to onboard developers quickly and help them to become productive. Code and documentation need to be discoverable, accurate, and follow consistent standards, and systems should give developers the confidence to make changes and know they are unlikely to break something. In academic settings, developers tend to have more diverse technical and academic backgrounds, with research code being a common entry point. This can bring a valuable diversity of thought, but also a need to bridge the gap between research code and the standards expected in productiongrade development. Judicious use of tools and workflows helps to guide and upskill developers to meet these standards. There are also challenges in applying an agile framework to research-oriented development, including the sometimes disparate responsibilities of team members, and mapping long-running projects to the granularity of sprints. In this talk I will look at some ways to get the best out of development tools and workflows to address these issues, and discuss some ongoing challenges in this area. ABOUT THE AUTHOR Andrew Wilson is an eResearch Solutions Specialist at Waipapa Taumata Rau University of Auckland, contributing to the development of a new Instrument Data Service. Prior to joining the university, he worked as a research engineer and software engineer in commercial settings.For more information about eResearch NZ / eRangahau Aotearoa, visit:https://eresearchnz.co.nz/ |
---|---|
DOI: | 10.6084/m9.figshare.25333141 |