Code is Science is back: the journal code policy tracker now covers 24 journals.See the tracker
ContactCheck a journalRead the manifesto
ManifestoJournalsReview checklistBlogAboutContactCheck a journal
Tracked journals that require code with the paper 66.66667%

Code is science. Review both.

If a result depends on software, the software is part of the result. We publish the principles, policies and practical guides that help researchers open, review, archive and cite their code.

codeisscience.com / review
← Back to papersMethods 2.3FiguresDataCodeLicenseArchiveCitation
Draft · Review round 1
Soil carbon flux, analysis.py
Python 3.12lockfile ✓Tests 42 passed
41def flux(c, t):
42 rate = c / t * 86400 # per day
42 rate = c / t * 3600 # per hour, matches Fig. 3
R2Reviewer 2 · Units here disagree with the Methods text. Fixed in v1.0.1?
43 return rate.mean()
44
Code availability
Repositorypublic
LicenseMIT
ArchiveZenodo
DOI10.5281/zenodo.x
CITATION.cff✓
Approve codeRequest changes
codeisscience
Reproducible release
v1.0.1 · archived
CIRerun of Fig. 3 matches
2 min ago

The journals that already ask for code. And the 8 that still only ask nicely.

Open the tracker
JOSSReScience CGigaScienceeLifePLOS Comp BioSoftwareXJSSJORSESSDF1000GMDSci DataResearch software engineer standing in a server room
Tracker
24
journal code policies, each linked to its source

One place for research code. From first commit to final citation.

Read the manifestoOpen the tracker

The manifesto. Seven principles, plainly stated.

Paper accepted. Code: "available on request". Nobody has requested it in four years.
Code archived as v1.0.1, DOI minted, license MIT. Reviewers reran Fig. 3.
Principle 3. Incorrect code means incorrect science.

Journal policy tracker. Know the rules before you submit.

JournalPolicy
Journal of Statistical Softwaropen license
Journal of Open Research Softwopen license
The Journal of Open Source Sofopen license
ReScience Copen license
Journal of Open Source Educatiopen license
Journal of Machine Learning Reopen license
Image Processing On Line (IPOLopen license
SoftwareXopen license

Review checklist. For authors and reviewers.

  • README runs in one command
  • Versions are pinned
  • Output matches Figure 2
  • License file present
  • Release archived with DOI

Archiving guide. Links break. DOIs last longer.

Releases
soil-flux · archived on every tag
v1.0.1DOI
v1.0.0DOI
v0.9.0-preprintDOI
mainlive

License primer. Pick one, on purpose.

MITApache-2.0BSD-3 GPL-3.0MPL-2.0CC0

Get a code checklist for your paper

Describe your project in one line: language, where the data lives, your target journal. We return the steps that matter for it.

Runs in your browser. Nothing you type is sent or stored.

Most research code lives in five places at once.

A script on a laptop. An attachment in an old email. final_v3.py next to final_v3_REAL.py. A supplementary zip nobody can unpack. A repository link that now returns 404. None of it is reviewed, and none of it is cited.

  laptop/analysis.R
df <- read.csv("~/Desktop/
  data_FINAL2.csv")
# TODO: ask Sam why 0.73?
model <- lm(y ~ x + z)
LLab lead
Can you send me the script?
  results_all_runs.xlsx
  Mail
Re: updated script (attachment)
final_v3.py · 14 KB
final_v3_REAL.py · 15 KB
3
  Downloads
supplementary_S4.zip
Archive is corrupt
  github.com/oldlab/flux
404
404 Not Found
R2Reviewer 2
Is this the code behind Fig. 3?
  terminal
$ python final_v3.py
ModuleNotFoundError: No module named 'oldpkg'
$
SFormer student
It worked on my machine in 2021
One release. One record.
soil-flux v1.0.1
MIT licenselockfiletests passingarchivedDOI 10.5281/zenodo.x

Every artifact above, tagged, archived and cited from the paper.

Errors in code are errors in science. So treat them that way.

Read principle 3
“I found the sign error two years after we published. Nobody else could have found it, because nobody else ever saw the script.”Postdoc, computational ecologyIllustrative compositeRead the session notes
See the checklist
Researcher at a desk reviewing code on a laptop
Policycheck before you submit
Open the tracker

Check the policy before you submit. Not after the reviewer asks.

Open the journal tracker
“We lost three weeks at revision stage because the journal wanted an archived release with a DOI. A five-minute check would have saved us.”PhD candidate, hydrologyIllustrative compositeWrite the statement

Archived code gets cited. A broken link gets nothing.

Read the archiving guide
“Once our code had a DOI, people started citing it directly. It finally showed up as work, not as a footnote.”Research software engineer, genomicsIllustrative composite
Archiving release

soil-flux v1.0.1

Tagged 14 Oct 2026
Public, MIT license
ZDZenodo · Software Heritage

How to cite

Built for one lab.
Useful for a whole field.

Start with a single paper and a single repository. The same habits scale to a department policy or a journal's author guidelines, and to the day a lab script grows into a tool other groups depend on and you bring in an MVP development team like YuSMP Group.

When a paper is submitted
If theMethodsusecustom code
RequireOpen license+DOI
NotifyCode reviewer
Publish withavailability statement

For researchers

Checklists, license notes and an archiving walkthrough you can apply to your next submission this week.

Start with the checklist
● UK · RSE society● deRSE● US-RSE● RSE-AUNZ● The Carpentries

For editors and reviewers

A clear way to describe code policies, plus review questions that work across fields and languages.

See the policy levels

From the blog. 26 guides on open, reviewed and citable code.

All posts
Line illustration of a smartphone, a smartwatch, test tubes and a plant on a lab bench
PracticeNov 13, 202611 min

Building an iOS app for a research study: a practical guide

A study app looks like a coding problem. Most of the work is protocol, consent, data handling and maintenance. This guide walks through the decisions in the order a research team actually meets them.

PracticeNov 5, 20266 min

How to organise a research project folder so others can follow it

A research project that only its author can navigate is hard to review, hard to hand over and hard to reproduce. A simple, consistent folder layout fixes most of that, and it takes an afternoon to set up.

CommunityOct 22, 20266 min

Teaching researchers to code: what works

Most researchers learn to code by copying a colleague's script and hoping. There are better ways. This guide covers the teaching approaches that have proven themselves in research settings and how to start one where you work.

Voices from the lab

Everyone has a final_v3.py story.

Share your story
PC
PhD candidateClimate modeling
My thesis rests on 4,000 lines of Fortran. Until last year, I was the only person who had read them.
AE
Associate editorEcology
A clear code policy made reviews faster. Authors knew what to send before we had to ask.
PI
Principal investigatorEpidemiology
Our models informed real decisions. Opening the code was the least we owed the people affected by them.
P
PostdocAstronomy
Someone I had never met fixed a bug in our pipeline and sent a pull request. That is peer review.
PC
PhD candidateClimate modeling
My thesis rests on 4,000 lines of Fortran. Until last year, I was the only person who had read them.
AE
Associate editorEcology
A clear code policy made reviews faster. Authors knew what to send before we had to ask.
PI
Principal investigatorEpidemiology
Our models informed real decisions. Opening the code was the least we owed the people affected by them.
P
PostdocAstronomy
Someone I had never met fixed a bug in our pipeline and sent a pull request. That is peer review.
PC
PhD candidateClimate modeling
My thesis rests on 4,000 lines of Fortran. Until last year, I was the only person who had read them.
AE
Associate editorEcology
A clear code policy made reviews faster. Authors knew what to send before we had to ask.
PI
Principal investigatorEpidemiology
Our models informed real decisions. Opening the code was the least we owed the people affected by them.
P
PostdocAstronomy
Someone I had never met fixed a bug in our pipeline and sent a pull request. That is peer review.
PC
PhD candidateClimate modeling
My thesis rests on 4,000 lines of Fortran. Until last year, I was the only person who had read them.
AE
Associate editorEcology
A clear code policy made reviews faster. Authors knew what to send before we had to ask.
PI
Principal investigatorEpidemiology
Our models informed real decisions. Opening the code was the least we owed the people affected by them.
P
PostdocAstronomy
Someone I had never met fixed a bug in our pipeline and sent a pull request. That is peer review.
P
PostdocNeuroscience
Writing tests felt slow. Rewriting a figure after a reviewer found a bug felt slower.
RS
Research software engineerPhysics
The best review I ever got was someone running my code and telling me where they got stuck.
R
ReviewerBioinformatics
If I cannot see the code, I can only review the story the authors tell about it.
L
LibrarianResearch data services
Researchers ask us how to cite software every week. A citation file answers most of those questions.
P
PostdocNeuroscience
Writing tests felt slow. Rewriting a figure after a reviewer found a bug felt slower.
RS
Research software engineerPhysics
The best review I ever got was someone running my code and telling me where they got stuck.
R
ReviewerBioinformatics
If I cannot see the code, I can only review the story the authors tell about it.
L
LibrarianResearch data services
Researchers ask us how to cite software every week. A citation file answers most of those questions.
P
PostdocNeuroscience
Writing tests felt slow. Rewriting a figure after a reviewer found a bug felt slower.
RS
Research software engineerPhysics
The best review I ever got was someone running my code and telling me where they got stuck.
R
ReviewerBioinformatics
If I cannot see the code, I can only review the story the authors tell about it.
L
LibrarianResearch data services
Researchers ask us how to cite software every week. A citation file answers most of those questions.
P
PostdocNeuroscience
Writing tests felt slow. Rewriting a figure after a reviewer found a bug felt slower.
RS
Research software engineerPhysics
The best review I ever got was someone running my code and telling me where they got stuck.
R
ReviewerBioinformatics
If I cannot see the code, I can only review the story the authors tell about it.
L
LibrarianResearch data services
Researchers ask us how to cite software every week. A citation file answers most of those questions.
LM
Lab managerStructural biology
We now archive every release that backs a paper. New students can start from working code on day one.
MS
Master's studentEconomics
I picked a license in ten minutes once I understood the differences. I had avoided it for a year.
DS
Data scientistSocial science
Our replication package used to be a zip file. Now it is a tagged release with a DOI and a README.
PC
PhD candidateComputational chemistry
I used to think my code was too messy to share. Sharing it is what made me clean it up.
LM
Lab managerStructural biology
We now archive every release that backs a paper. New students can start from working code on day one.
MS
Master's studentEconomics
I picked a license in ten minutes once I understood the differences. I had avoided it for a year.
DS
Data scientistSocial science
Our replication package used to be a zip file. Now it is a tagged release with a DOI and a README.
PC
PhD candidateComputational chemistry
I used to think my code was too messy to share. Sharing it is what made me clean it up.
LM
Lab managerStructural biology
We now archive every release that backs a paper. New students can start from working code on day one.
MS
Master's studentEconomics
I picked a license in ten minutes once I understood the differences. I had avoided it for a year.
DS
Data scientistSocial science
Our replication package used to be a zip file. Now it is a tagged release with a DOI and a README.
PC
PhD candidateComputational chemistry
I used to think my code was too messy to share. Sharing it is what made me clean it up.
LM
Lab managerStructural biology
We now archive every release that backs a paper. New students can start from working code on day one.
MS
Master's studentEconomics
I picked a license in ten minutes once I understood the differences. I had avoided it for a year.
DS
Data scientistSocial science
Our replication package used to be a zip file. Now it is a tagged release with a DOI and a README.
PC
PhD candidateComputational chemistry
I used to think my code was too messy to share. Sharing it is what made me clean it up.

Composite voices written by the editors from common experiences. Roles only, no real people.

Open it. Review it. Cite it.