Skip to content

Make Gitlab Repo Citable via data repositories

Description

I would like a feature that makes GitLab repositories cite-able by archiving them with a trusted repository with a DOI. The goal of this feature request is to allow Zendodo to archive and apply a DOI to GitLab repositories is to make GitLab repositories cite-able, more widely disseminated, and incentivize the use of the GitLab amongst academics.

Proposal

I'm not 100% sure how to implement this, but Arfon Smith outlines the plan for GH here. Sorry! Not a developer, but would love to see this feature.

Links / references

DOI documentation

Zenodo source code

GitHub's guide for integration with Zenodo

Documentation blurb

I'm not sure what this section means?

Overview

I would like a feature that makes GitLab repositories cite-able by archiving them with a trusted repository with a DOI.

For academics, citations and metrics are key to promotion, tenure, grant applications, and more. Increasingly, researchers develop their own software as a part of their research process. Up until recently, researchers have only been able to cite software and get their software cited by writing an article about it, so they can get credit and include software in tenure packets and other areas of review.

GitHub provides functionality to make repositories (read: software) citable on its own via a webhook to Zenodo, a repository that archives the repository and gives it a DOI. Zenodo is a research data repository created by OpenAIRE and CERN to provide a place for researchers to deposit datasets. Many researchers trust this as a place to deposit data, and now for archiving code as well. Zenodo is great, but this feature could probably be implemented for many repositories, such as the Open Science Framework (source code here), Figshare, and institutional repositories maintained by Universities (predominant ones are open source -- Fedora and Hydra).

The key behind this metrics ecosystem is the dominant persistent identifier, the DOI (digital object identifier) which is indexed by many, many discovery systems including ORCID, SHARE, Web of Science, SCOPUS, and library systems. These are also used in services such as doi2bib which makes a DOI and spits back the BibTex people use to cite software in manuscripts and books.

I love GitLab, and I want it to be more widely used by academics (I'm an academic librarian). Making a repository citable would be a great addition to GitLab's already robust functionality (integrated CI is a big deal for researchers).

Use cases

Here are four predominant use cases I see across different parts of academia:

  1. A professor writes a piece of software which is hosted in a GitLab repository and they want people to cite it when they use it in papers or other research output. They archive it in Zenodo, get a DOI, and link that in the README for others to reference as a preferred citation. More citations = better chance of tenure, and visible return on investment for grant funders (which means more chance of future funding).

  2. A research engineer in a University has a three year contract and needs to prove their open source work has merit to get a contract renewed. They make a GitLab repository cite-able, track the use of the DOI associated with that repository, and presents the metrics to their supervisor, who then renews their contract.

  3. A grad student working in a lab writes an innovative piece of software to make the grunt work of data cleaning simpler. As a vulnerable person in the academic lifecycle, they want a way to publish this contribution to their field without having to write a paper, which would have to include their PI in the author list even though this grad student wrote the code on their own. They want their labour represented fairly. They discover that GitLab, which they've been using to host the code privately, lets you archive and publish code and get a DOI to track use. They make the code public and publish it via this feature, and their solo labour is acknowledged and they have a greater chance of becoming someone's PhD student.

  4. A postdoc is doing an independent research project which includes writing software for super specific scientific analyses. They need to get hired by a University at the end of their 1 or 2 year postdoc. They make their GitLab repository cite-able with Zenodo, and it' s discovered by others in their niche field. It gets cited many times over the course of the next few years, and the postdoc can track these citations via the DOI. They get hired at the end of their postdoc.

Feature checklist

Make sure these are completed before closing the issue, with a link to the relevant commit.