What the journal publishes

ExSil publishes independent examinations of scientific software and public research data: reports that test whether a widely used or consequential tool or dataset produces correct results, whether it signals when it does not, and how its maintainers responded when the issue was raised with them, or that they did not.

Articles must include a reproducible test, directly relevant to the claim being made, against a reference that is independent of what is tested.

The journal does not publish software announcements, novel methods without an independent examination, or reviews and opinion pieces that do not include a test.

Why it exists

Almost every scientific result now passes through software or a shared dataset somewhere along the way, and most of that software is only ever checked by the people who wrote it. When something does go quietly wrong, the evidence usually ends up in an issue tracker or someone’s notes, where it’s hard to find and impossible to cite. ExSil gives that evidence a permanent, peer-reviewed home, and gives the people who maintain the software a chance to answer in the same place.

How it works

• You submit by opening an issue on GitHub, and the review happens there in the open.
• Reviewers rerun your test themselves before anything is accepted.
• We invite the maintainers to respond, and publish their reply next to the article.
• Every article goes into a permanent archive on Zenodo with its own DOI.
• It costs nothing to submit, publish or read.

We're not taking submissions yet. The editorial board comes first. If you'd like to review, edit or send us something once we open, the Editors page explains how to get in touch.