Authorship policy

Since the performance of scientists is evaluated based on the number and quality of their publications, authorship is an important topic for software with a focus on scientific application, such as LiberTEM. Our goal is to give appropriate credit to each contribution so that the work of contributors becomes visible as a form of scientific output and is considered in the evaluation of their productivity as scientists. We hope that this can support and encourage Open Source software development as a form of scientific work.

Following established practices for scientific papers, we distinguish between creators, who are persons that contributed in a way that establishes co-authorship of LiberTEM, and contributors, who are persons that contributed in other ways, for example through discussions or help.

Creators are persons who fulfill at least one of these criteria:

• Contributed at least one commit to the repository.

• Contributed material such as code or documentation in other ways that don’t appear directly as commit, including prototype code that served as a basis for code in LiberTEM.

• Contributed to the management and direction, for example active management and support for contributors who dedicate a significant part of their working time to LiberTEM.

We maintain two files, packaging/creators.json and packaging/contributors.json, where we work to keep track of creators and contributors. For transparency, we include a short statement of the nature of the contribution. If you feel like you or someone else should be listed there or should be represented differently, please file a pull request or contact us! Accidental discrepancies happen, and many eyes and helping hands help us to keep the information up to date.

Please note that the GitHub breakdown of contributions is not always reflecting reality. As an example, commits are only linked to GitHub accounts if the Git client is configured correctly for that.

All people listed in creators.json are included as authors in publications of the software itself, for example uploads on Zenodo.

Publications about LiberTEM

For other publications about LiberTEM, such as scientific papers, all persons with at least one contribution within the past 18 months that qualifies them for being a creator of LiberTEM will be contacted during the drafting process of the publication, for example through their GitHub handle or via e-mail, and offered co-authorship. If they actively consent to being a co-author in a timely manner, they are included as co-authors under the rules of that particular publication and medium. For scientific papers, that typically includes an obligation to approve drafts before submission and being accountable for all aspects of the work. Co-authors who are not responsive may be excluded if they don’t respond to a reminder, for example to approve a draft, in a timely fashion.

For publications that only cover specific aspects of LiberTEM, for example a particular feature, only contributions to those particular aspects are considered to select creators who will be offered co-authorship.

Co-authors who are not creators of LiberTEM can be included in publications about LiberTEM following established practices for authorship, for example as contributors to the content of the publication or to other presented material, such as scientific applications.

Ordering of author lists

The creators and contributors are listed alphabetically in Acknowledgments with a short statement about their contribution.

Alphabetical ordering of author lists is uncommon for scientific papers. Instead, the position on the author list is used to indicate the relative amount of contribution of an author. The author that contributed most of the content is listed first and the author that contributed most guidance is listed last. Casual readers, in particular encountering an abbreviated author list in a reference within a citing paper, will assign most credit to the first author. An alphabetical author list would therefore be unfair towards main contributors with a name in the middle of the alphabet.

In order to resolve this issue and assign prestigious author positions to the people who deserve them in a transparent fashion, the authors agree among each other about their authorship positions for each publication individually.

Since Zenodo assigns DOIs and allows to export a citation for reference managers that are commonly used in scientific publishing, a Zenodo upload is treated as a scientific publication and the authors are included in the order of packaging/creators.json. This allows to assign prestigious first author positions to main contributors when LiberTEM is cited in scientific papers.

Authorship questions should be resolved through discussion in Issues and change proposals in the form of Pull Requests. Our goal is amicable cooperation and proper credit for all contributions.

If you have questions or would like to suggest changes to this policy, please contact us! See #460 for the initial discussion that lead to establishing this policy.