Structure and Evolution of Package Dependency Networks

Conference Paper (2017)
Author(s)

Riivo Kikas (University of Tartu)

Georgios Gousios (TU Delft - Electrical Engineering, Mathematics and Computer Science)

Marlon Dumas (University of Tartu)

Dietmar Pfahl (University of Tartu)

Research Group
Software Engineering
DOI related publication
https://doi.org/10.1109/MSR.2017.55 Final published version
More Info
expand_more
Publication Year
2017
Language
English
Research Group
Software Engineering
Bibliographical Note
Accepted author manuscript
Pages (from-to)
102-112
ISBN (print)
978-1-5386-1545-4
ISBN (electronic)
978-1-5386-1544-7
Event
MSR 2017 (2017-05-20 - 2017-05-21), Buenos Aires, Argentina
Downloads counter
403
Collections
Institutional Repository
Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Abstract

Software developers often include available open-source software packages into their projects to minimize redundant effort. However, adding a package to a project can also introduce risks, which can propagate through multiple levels of dependencies. Currently, not much is known about the structure of open-source package ecosystems of popular programming languages and the extent to which transitive bug propagation is possible. This paper analyzes the dependency network structure and evolution of the JavaScript, Ruby, and Rust ecosystems. The reported results reveal significant differences across language ecosystems. The results indicate that the number of transitive dependencies for JavaScript has grown 60% over the last year, suggesting that developers should look more carefully into their dependencies to understand what exactly is included. The study also reveals that vulnerability to a removal of the most popular package is increasing, yet most other packages have a decreasing impact on vulnerability. The findings of this study can inform the development of dependency management tools.

Files

Ecosystems_evolution.pdf
(pdf | 0.471 Mb)
License info not available