Publish research like you write code.
Versioned Science is an open journal where every paper is a git repository. Every revision is public, every version has a permanent link you can cite, and anyone can question a result where the result lives.
- Open by default — every change is a public commit
- Cite any version — point readers at the revision you used
- Review in the open — questions and corrections happen in issues
- Free to read, free to publish — no fees, no paywall
An example
See what a paper looks like here
This repository holds a complete manuscript — abstract, figures, methods and supporting data — written and revised in the open.
Directed evolution of mesophilic HNA polymerases providing insight into DNA polymerase mechanisms
Paola Handal-Marquez, Leticia L. Torres, Vitor B. Pinheiro
A focused library of Phi29 DNA polymerase variants, selected for faster HNA synthesis and deep-sequenced to identify P562del — a deletion that speeds up initiation while reducing processivity and fidelity.
What you can do in it
- Read the current version, always the latest one.
- Compare it with earlier versions and see what changed.
- Open an issue to question a result or a method.
- Propose a correction as a pull request.
Every one of those actions leaves a public record. Nothing about the paper's history is hidden from the people reading it.
How it works
Four steps, no PDFs
-
Start a repository for your paper
Create one in the Versioned Science organisation, or fork an existing paper to build on work someone else already published.
-
Write the paper next to its evidence
Manuscript, figures, data and analysis code live in the same repository, so every claim sits beside what supports it.
-
Revise in the open
Edits are commits. Corrections, reanalyses and new results arrive as pull requests, so the reasoning behind each change stays part of the record.
-
Release a version when it's ready
Tag the release you want people to cite. Readers always see the current version and the full path that produced it.
You don't need to be a git expert. If you can edit a file on the GitHub website, you can publish here — git is how the history is stored, not a tool you have to learn first.
What you get
The parts of a journal that were missing
Find and solve problems with anyone
Talk to the people who did the work
Stuck reproducing a protocol? Open an issue on the paper itself. The discussion sits next to the method, and the answer is there for the next person who gets stuck.
Updates arrive as pull requests
When new results change a conclusion, the change is proposed, discussed and merged — visible to everyone, instead of buried in a corrigendum nobody reads.
Everything is on the record
Publish the results that didn't work
Negative results get a permanent, findable home. A dead end you document saves the next person a year of repeating it.
Review happens where you can check it
Objections, corrections and revisions are part of the paper's history. A closed review can't be audited afterwards; an open one can.
Credit that matches the work
See who contributed what
The history records every contribution with names and dates. Authorship itself stays a decision for the authors, made the way the community expects — the repository just makes the work behind it visible.
Always know which version you're reading
No more wondering whether the copy you downloaded is current. You see the latest version and everything it changed on the way there.
Questions
The obvious objections
Isn't this just GitHub?
Partly, and deliberately. We don't try to rebuild hosting, search, permissions or identity — GitHub already does those better than we could. What a journal adds is structure and standards: a paper format, a citable version, and a place to argue about results in public. That structure is what lives here.
Can I cite it?
Yes. Every repository has a permanent URL and every release is tagged, so you can cite the exact version you read and be sure it still resolves.
Where is the peer review?
In the open, in issues and pull requests, instead of behind closed doors. Objections are visible, the authors' responses are visible, and the resulting revision is visible. The point isn't to remove scrutiny — it's to stop the scrutiny from disappearing once the paper is accepted.
Will my committee accept it?
In most places, not yet — and that's the real obstacle, not the software. Careers are still measured by journal names and impact factors, and no amount of tooling fixes that on its own. We're building the record first, and working with the people who want how research is judged to change too.
What about human data, ethics or embargoes?
Not everything belongs in the open. Sensitive data can stay in a private repository, or be deposited in a controlled-access archive that the paper links to. The manuscript and its reasoning can be public while the data stays governed.
Do I need to know git?
No. If you can edit a file on the GitHub website, you can publish here.
Does it cost anything?
No. Reading is free, publishing is free, and there's no submission fee or paywall. Public repositories on GitHub are free to host, and that's where the journal runs.
Where we are
What exists today
This is a small project. Here is what it is right now.
-
A journal
Papers published as repositories in the Versioned Science organisation.
-
A newsletter
Announcements and discussion, on Substack.
-
A community
A Discord server where we coordinate.
Publish outside the PDF.
If you want a place to publish research that keeps its own history — or you want to help build one — come find us.