Contributing to the content of MDN Web Docs

🎉 First of all, thanks for taking the time to contribute to MDN Web Docs! 🎉

The following is a set of guidelines for contributing to the content of MDN Web Docs, which is hosted within the MDN Organization on GitHub.

Note: Just want to find a task and jump in? See Getting started on MDN for an overview of how to join, and Contributing to MDN for a filtered list of tasks.

Code of Conduct
License
Making contributions
Frequently asked questions (FAQ)
- When will my change show up on the production MDN site?
- Can I copy content from other places to put on MDN?

Code of Conduct

Everyone participating in this project is expected to follow our Code of Conduct.

License

When contributing to the content you agree to license your contributions according to our license.

Making contributions

A good place to learn about general guidelines for contributing to MDN Web Docs is the Guidelines document. For example, you can find out more about MDN's writing-style guidelines via the Writing style guide.

Setup

No matter how you wish to contribute, you'll need a GitHub account if you don't have one already. If you're not familiar with git and GitHub, you might find the MDN Git and GitHub document helpful.

There are several ways forward from this point. It's up to you. Here are some options:

Go to https://github.com/mdn/content and just use the GitHub UI. This is the easiest approach if you just want to make a simple change to a single file, like fixing a typo.
Install and use the GitHub Desktop
Install and use the GitHub CLI
Install git and use it from the command line. You might find these resources helpful:
- Using Git
- git cheatsheet

If you choose an option other than the GitHub UI, you want to install Node.js (version >=12.11) and yarn.

Fundamental concepts

These are some important things to keep in mind about the MDN content.

Documents are folders -- Documents are always represented by a folder (e.g., files/en-us/web/javascript), and that folder will contain the content of that specific document as an index.html file (e.g., files/en-us/web/javascript/index.html).
Documents are hierarchical - A document folder may contain other folders, where those folders would represent child documents (e.g., files/en-us/web/javascript/closures/index.html).
Document folders may contain image files -- A document folder may also contain image files, which are referenced within that document's index.html file.
All redirects are specified in a single file -- All of the redirects are specified within files/en-us/_redirects.txt, one redirect per line. Each line specifies a from and to URI separated by whitespace. When you move a document, you'll need to add a redirect to this file specifying that its old URI now redirects to its new URI. If both an index.html file and a redirect exist for a document, the document takes precedence and the redirect is ignored.
A document's index.html starts with "front-matter" -- Each document's index.html file must begin with some YAML called front-matter that defines some important information about the document: title, slug, and tags (if any). Here's an example that shows the front-matter from the JavaScript landing page:
```
---
title: JavaScript
slug: Web/JavaScript
tags:
  - JavaScript
  - Landing
  - Landing page
  - Learn
  - 'l10n:priority'
---
```

Simple changes

If you just want to make a simple change to a single file, like fixing a typo, the GitHub UI is the simplest way to do that. For example, if you've found a typo within the JavaScript landing page, you can sign into GitHub, go to https://github.com/mdn/content, navigate to the source file files/en-us/web/javascript/index.html, and then click on the edit (pencil) button.

Tip: Click the Source on GitHub link in the footer of any MDN page to jump to its source file on GitHub.

From there the GitHub UI will take your hand and walk you through the rest, like automatically creating a fork and branch to commit your changes to, as well as helping you reach the ultimate goal, a pull request. Your pull request represents the work you want to be reviewed, hopefully approved, and then merged into the main branch of this repository.

Note: See the pull request etiquette section for more details on creating and handling pull requests successfully.

If you're not certain of the changes that you want to make, get in touch with us first! You can chat with us or file an issue.

More substantial changes

If you need to do some work that requires changes to more than one file, like moving one or more documents, the GitHub UI is not very efficient. You'd have to make a separate pull request for every page you want to change. Instead, you're going to have to use git or one of the other git-based approaches like the GitHub Desktop.

You'll want to create a fork of this repository, so you can freely experiment with branches and changes in your own copy before submitting your changes as a pull request. Let's assume your GitHub username is octocat. Your fork would be a copy of this repository but in your own account, so https://github.com/octocat/content.
Once you've created your fork on GitHub, you'll want to clone it locally. For example, assuming again that your GitHub username is octocat, you would do something like the following:
```
cd ~/repos
git clone git@github.com:octocat/content.git mdn/content
```
You'll also want to create a remote to the main repository (https://github.com/mdn/content), which you'll use to keep your local clone as well as your fork (https://github.com/octocat/content) up-to-date. For these examples, we'll name it mdn, but you can name it anything you'd like.
```
cd ~/repos/mdn/content
git remote add mdn git@github.com:mdn/content.git
git remote -v
```
When you run the git remote -v command above, you'll see that you have two remotes: mdn and origin. The origin remote is the default name that git has assigned to your fork (https://github.com/octocat/content).
Once you've created your local clone, there's no need to do that again next time you want to make a contribution. However, each time before you start a new chunk of work make sure you update your local clone. The following checks-out your local clone's main branch, fetches the latest content from the main branch of the mdn repository and merges it into your local main branch, and finally checks out a new branch called my-work (you can call it anything you'd like) for you to work within. When you're ready, you'll push your my-work branch to your fork and use it to make a pull request.
```
cd ~/repos/mdn/content
git checkout main
git pull mdn main
git checkout -b my-work
```
Next, you'll want to start the local preview service, so you can see the changes you'll make as they would look in production. Once started, this local preview service is available at http://localhost:5000 within your browser.
```
# Switch to a separate terminal.
cd ~/repos/mdn/content
yarn
yarn start
```
When you preview a page you can press a button to open its associated document's index.html file in your preferred editor. For this to work, you need to set an environment variable called EDITOR before starting the preview server. For example, if you prefer VS Code as your editor, you'll want to do something like this:
```
export EDITOR=code
yarn start
```
Note, this is how you do it on terminals with bash (macOS and Linux). You have to do it differently on Windows.

Now, when you're previewing a page and press the Edit in your editor button, it will open the same as running:
```
code files/en-us/.../index.html
```
Instead of having to type export EDITOR=code every time prior to yarn start, you can instead store this setting in your personal .env file (this goes inside the root of your local content directory). If the .env file doesn't already exist, you can create one and include the above setting in it. Alternatively, running the following line will add the setting to the .env file automatically, creating the file if it doesn't already exist:
```
echo 'EDITOR=code' >> .env
```
Now, it should be set like that even after you've closed and started a new terminal window.
Make your desired changes to one or more index.html files using your preferred code editor. When thinking about your desired changes, it's important to keep the following in mind:
- Make sure you've read the MDN guidelines, including the Writing style guide.
- Large chunks of work can be difficult to review, so try to group your changes into the smallest logical chunks that make sense, and create a separate pull request for each logical chunk.
Once you've made and saved your changes, open a browser, and navigate to the page(s) you've changed. For example, if you changed files/en-us/web/javascript/index.html, open http://localhost:5000/en-us/docs/web/javascript in your browser.
You might have noticed that at the top of each page that you preview, for example the http://localhost:5000/en-us/docs/web/javascript page, there is a Show flaws button. Click on that button to see if your changes have introduced flaws on the page. You can also fix flaws on a particular page by running the command:
```
yarn content flaws <slug>
```
Once you're happy with your changes, add and commit them to your branch, and then push the branch to your fork. Remember, the default name that git assigned to the remote that represents your fork is origin.
```
cd ~/repos/mdn/content
git add .
git commit
git push -u origin my-work
```
You're now ready to create a pull request.
Once you've created your pull request, sit back, relax, and wait for a review. You do not need to request a review. One or more reviewers will be selected for you automatically. Your pull request will have to be reviewed and eventually approved before it's merged into the main branch, and then later (within 48 hours) published on MDN Web Docs. Along the way, you may be asked, not only to answer questions about your work, but to make changes as well. Don't worry, that's a common and natural part of the process. See the pull request etiquette section for more details on creating and handling pull requests successfully.

Pull request etiquette

Here are some important rules of etiquette to remember when working with pull requests.

When you submit a pull request, a number of tests are automatically run as GitHub Actions (see .github/workflows). If one or more of these tests fail, it is your responsibility to try and resolve the underlying issue(s). If you don't know how to resolve the underlying issue(s), you can ask for help. Your pull request will not be approved and merged if these tests are failing.
If your pull request has merge conflicts with the main branch (GitHub checks for this automatically and notifies you), you are responsible to resolve them. You can do this by merging the main branch into your branch (git pull mdn main), and then pushing the updated branch to your fork (git push).
Once you've created your pull request, never use git rebase on your branch if you need to make changes. Any changes should be made as additional commits.
Each pull request should contain a single logical change, or related set of changes that make sense to submit together. If a pull request becomes too large or contains too many unrelated changes, it becomes too difficult to review, and may begin to look suspicious (it is easier to hide malicious changes in a large pull request). In such cases, the reviewer has the right to close your pull request, and ask that you submit a separate pull request for each logical set of changes that belong together.
If your pull request contains any kind of significant complexity (it contains technical changes, and isn't just a typo fix, grammatical improvement, or formatting/structural change), please describe why you're making the change and anything else we need to know about it.
- If the pull request is simple (it is really clear what has been changed and why, and the change is obviously a good thing), you can do this in your pull request's description.
- If the pull request is complex (the changes and the reasoning behind them need a bit more explanation), then the requestor should file an issue describing the intended change first, and seek discussion/approval as needed. When the time is right to submit the PR, they should reference the issue (or an existing issue that describes the motivation for the change) in the PR. You can reference an existing issue using # followed by the issue's ID, for example #1234.
Do not re-open a pull request that a reviewer has closed.

Adding a new document

Adding a new document is relatively straightforward, especially if you can start by copying the index.html of a similar document. There are only a few things to keep in mind:

Remember that a document is represented by an index.html file within its own folder.
Determine where in the document hierarchy your document belongs. For example, if you're creating a new CSS document for a new property foo, you'll want to create a new folder files/en-us/web/css/foo/ and its files/en-us/web/css/foo/index.html file.
Remember that a document's index.html file must start with front-matter that defines the title, slug, and tags (if any) for the document. You might find it helpful to refer to the front-matter within a similar document's index.html.

As we outlined above, the step-by-step process in general would be:

Start a fresh, up-to-date branch to work within:

cd ~/repos/mdn/content
git checkout main
git pull mdn main
# Run "yarn" again just to ensure you've
# installed the latest Yari dependency.
yarn
git checkout -b my-add

Create one or more new document folders, each with their own index.html file.
Add and commit your new files, as well as push your new branch to your fork:
```
git add files/en-us/folder/you/created
git commit
git push -u origin my-add
```
And finally create your pull request.

Moving one or more documents

Moving one or more documents, or an entire tree of documents is easy, because we've created a special command that takes care of the details for you:

yarn content move <from-slug> <to-slug> [locale]

You just have to specify the slug of the existing document that you'd like to move (e.g., Learn/Accessibility), as well as the slug of its new location (e.g., Learn/A11y), optionally followed by the locale of the existing document (defaults to en-US). If the existing document that you'd like to move has child documents (i.e. it represents a document tree), the yarn content move command will move the entire tree. For example, let's say you want to move the entire /en-US/Learn/Accessibility tree to /en-US/Learn/A11y:

First, as we've outlined above, you'll start a fresh branch to work within:

cd ~/repos/mdn/content
git checkout main
git pull mdn main
# Run "yarn" again just to ensure you've
# installed the latest Yari dependency.
yarn
git checkout -b my-move

Perform the move (which will delete and modify existing files, as well as create new files):
```
yarn content move Learn/Accessibility Learn/A11y
```
Add and commit all of the deleted, created, and modified files, as well as push your branch to your fork:
```
git commit -a
git push -u origin my-move
```
Now you're ready to create your pull request.

Deleting a document

Deleting one or more documents, or an entire tree of documents is also easy, again because we've created a special command that takes care of the details for you:

yarn content delete <document-slug> [locale]

You just have to specify the slug of the existing document that you'd like to delete (e.g., Learn/Accessibility), optionally followed by the locale of the existing document (defaults to en-US). If the existing document that you'd like to delete has child documents (i.e. it represents a document tree), you must also specify the -r, --recursive option, or the command will fail. For example, let's say you want to delete the entire /en-US/Learn/Accessibility tree:

First, as we've outlined above, you'll start a fresh branch to work within:

cd ~/repos/mdn/content
git checkout main
git pull mdn main
# Run "yarn" again just to ensure you've
# installed the latest Yari dependency.
yarn
git checkout -b my-delete

Perform the delete:

yarn content delete Learn/Accessibility --recursive

Add and commit all of the deleted files, as well as push your branch to your fork:
```
git commit -a
git push -u origin my-delete
```
Now you're ready to create your pull request.

Adding images

Adding an image to a document is easy as well. All you need to do is add your image file within the document's folder, and then reference the image from within the document's index.html file using an <img> element. It's as easy as that. Let's walk through an example:

You should be getting used to this by now, as we've done it several times above. Make sure you start with a fresh working branch updated with the latest content from the main branch of the mdn remote.
```
cd ~/repos/mdn/content
git checkout main
git pull mdn main
# Run "yarn" again just to ensure you've
# installed the latest Yari dependency.
yarn
git checkout -b my-images
```
Add your image to the document folder. For this example, let's assume we're adding a new image to the files/en-us/web/css document.
```
cd ~/repos/mdn/content
cp /some/path/my-cool-image.png files/en-us/web/css/
```
Run the filecheck command on each image you add. It'll complain if something's wrong (see also the Compressing images section below). We'll automatically run this as one of the tests we run when your new pull request is created, but why wait to fix any possible issues later?
```
yarn filecheck files/en-us/web/css/my-cool-image.png
```
Use your image within the document. For example, add the following <img> element somehwere inside files/en-us/web/css/index.html:
```
<img src="my-cool-image.png" alt="My cool image">
```

Add and commit all of the deleted, created, and modified files, as well as push your branch to your fork:

git add files/en-us/web/css/my-cool-image.png files/en-us/web/css/index.html
git commit
git push -u origin my-images

Now you're ready to create your pull request.

Compressing images

When you add images to MDN, you should make sure that they are compressed as much as possible without degrading quality, to save on download size for our readers. In fact, if you don't do this our CI process will fail and the build results will warn you that some of your images are too big.

The best way to compress them is by using the built-in compression tool. You can compress an image appropriately by using the filecheck command with the --save-compression option — this compresses the specified image as much as possible, and saves the result over the top of the original.

For example:

yarn filecheck files/en-us/web/css/my-cool-image.png --save-compression

Updating a browser compatibility table

The browser compatibility table shown near the bottom of many of the MDN document pages, for example on the JavaScript Array page, is generated from the npm package created from the https://github.com/mdn/browser-compat-data repository. If you'd like to make or suggest a change to the content of the browser-compatibilitytable on any given MDN document page, you can either use the Report problems with this data on GitHub link in the upper-right corner of the table on the page itself to create an issue, or go to https://github.com/mdn/browser-compat-data and read more to learn how to contribute.

Adding code examples

There are a lot of code examples on MDN, and you'll probably want to add some as you contribute content to the project. There are four types of code example used on MDN:

Static examples — plain code blocks, possibly with a screenshot to statically show the result of such code if it were to be run.
Traditional MDN "live samples" — A macro that takes plain code blocks, dynamically puts them into a document inside an <iframe> element, and embeds it into the page to show the code running live.
GitHub "live samples" — A macro that takes a document in a GitHub repo inside the mdn organization, puts it inside an <iframe> element, and embeds it into the page to show the code running live.
Interactive examples — Our system for creating live interactive examples that show the code running live but also allow you to change code on the fly to see what the effect is.

For a complete description of this topic, see the MDN code examples page.

Also make sure that your code examples follow the MDN code example guidelines.

Archiving and unarchiving pages

Pages with the banner "This is an archived page. It's not actively maintained." at the top are archived pages — they are archived because either they are obsolete or out-of-date, or they are not in scope for MDN contributors to actively work on.

If you find a page that is marked archived and you think it shouldn't be, you can follow the below steps to unarchive it:

Submit an issue to check why the page is archived and if it should be unarchived.
Find the source code location of the archived page. Archived pages live in the Archived content repo, the structure of which is the same as this repo (except that there are also locale directories present other than en-US). As an example:
- Take the archived Mobile developer challenge page, which is at the URL https://developer.mozilla.org/en-US/docs/Mozilla/Mobile/Mobile_developer_challenge.
- This page's source code is at https://github.com/mdn/archived-content/blob/main/files/en-us/mozilla/mobile/mobile_developer_challenge/index.html.
To unarchive the page, you need to open a pull request to move the page's directory into the appropriate place inside this repo, and update the page's meta data (like slug) if needed.

Note: To archive a currently-non-archived page, you need to follow the above procedure, but move the page in question the opposite way between repos.

Note: You can have an archived page with non-archived pages below it in the hierarchy, and vice-versa. To achive this you need to have the same directory structure in both the content and archive-content repos. Whether a page is non-archived or archived depends on which repo its index.html file is in. As an example, compare the archived-content mozilla directory with the content mozilla directory. Both locations represent the same directory structure, but the index.html file appears in the latter, therefore the page is not archived.

Making a change that depends on a macro update

KumaScript macros are still used on MDN pages, even in the new platform. These are the function names surrounded by handlebars-style double curly braces that you'll see in the source code on occasion, for example {{domxref}} Eventually we have to replace them with something else, but they are here for now. They live in https://github.com/mdn/yari/tree/master/kumascript/macros.

Sometimes you'll want to make a change to the content that relies on a macro change. Take mdn#187. Florian wanted to add documentation for a new WebGL extension to MDN, but this change relied on the new feature's spec being added to https://github.com/mdn/yari/blob/master/kumascript/macros/SpecData.json. If not, the specification table on the new page would not render properly because the data it relies on would not be there.

In such situations:

Make the required PR to https://github.com/mdn/yari/blob/master/kumascript/macros/ first, and get that merged.
Add the content to this repo.

Frequently asked questions (FAQ)

When will my change show up on the production MDN site

After your pull request is merged, it may take up to 48 hours before the change goes live on the production https://developer.mozilla.org/ site, but it will usually be quicker.

The complete build runs every 24h at around 7PM US/Eastern time, and is then deployed.
Once deployment has occurred, it can take up to 24 hours for the associated CDN caches to replace their previously-cached content with the updates.

You can use https://whatsdeployed.io/s/DLi/mdn/content to see if your change has been deployed to the production site.

Can I copy content from other sources to put on MDN

In general, we do not approve of copying content from other sources and putting it on MDN. MDN should be made up of original content wherever possible. If we receive a pull request and discover that it contains plagiarised content, we will close it and request that the submitter resubmit the change with the content rewritten into their own words.

If someone wants to donate an article to MDN that they previously published on their blog or it makes sense to copy a complex reference sheet to MDN there may be justification for republishing it. In these cases you should discuss your plan with the MDN team first:

Include a comment on the relevant repo issue that explains your intention — say what you would like to copy (include the URL) and explain why you think this is appropriate.
If the content is published under a closed license:
- If you hold the rights to the content, state this, and say that you are happy to republish it on MDN.
- If you do not hold the rights to the content, include the author/publisher on the issue if possible, or include details of how they could be contacted so we can ask them for permission to republish the content.
If the content is published under an open license, say what it is, and link to the license so we can check whether it is compatible with MDN's license.
@-mention chrisdavidmills on the issue, so he can review the request.

Be warned that unless there is a good reason to republish the content, we will probably say "no". The MDN writing team's decision is final.

alvinhui/content