Communication Dans Un Congrès Année : 2025

Longitudinal Study of Software Environments Produced by Dockerfiles from Research Artifacts: Initial Design

Résumé

The reproducibility crisis has affected all scientific disciplines, including computer science (CS). To address this issue, the CS community has established artifact evaluation processes at conferences and in journals to evaluate the reproducibility of the results shared in publications. Authors are therefore required to share their artifacts with reviewers, including code, data, and the software environment necessary to reproduce the results. One method for sharing the software environment proposed by conferences and journals is to utilize container technologies such as Docker and Apptainer. However, these tools rely on non-reproducible tools, resulting in non-reproducible containers. In this paper, we present a tool and methodology to evaluate variations over time in software environments of container images derived from research artifacts. We also present initial results on a small set of Dockerfiles from the Euro-Par 2024 conference.

Fichier principal
Vignette du fichier
acmrep25-14.pdf (561.91 Ko) Télécharger le fichier
Origine Fichiers éditeurs autorisés sur une archive ouverte
Licence

Dates et versions

hal-05163884 , version 1 (15-07-2025)

Licence

Identifiants

Citer

Quentin Guilloteau, Antoine Waehren, Florina M. Ciorba. Longitudinal Study of Software Environments Produced by Dockerfiles from Research Artifacts: Initial Design. REP 2025 - ACM Conference on Reproducibility and Replicability, ACM, Jul 2025, Vancouver, Canada. ⟨10.1145/3736731.3746146⟩. ⟨hal-05163884⟩
91 Consultations
149 Téléchargements

Altmetric

Partager

  • More