Skip to content

Latest commit

 

History

History
186 lines (134 loc) · 7.48 KB

CONTRIBUTING.md

File metadata and controls

186 lines (134 loc) · 7.48 KB
Raw

Contributing to Electron Forge

Electron Forge is a community-driven project. As such, we welcome and encourage all sorts of contributions. They include, but are not limited to:

We strongly suggest that before filing an issue, you search through existing issues to see if it has already been filed by someone else.

This project is a part of the Electron ecosystem. As such, all contributions to this project follow Electron's code of conduct where appropriate.

Questions about usage

If you have questions about usage, we encourage you to visit one of the several community-driven sites.

Before opening bug reports/technical issues

Debugging

Troubleshooting suggestions can be found in the support documentation.

Contribution suggestions

We use the label help wanted in the issue tracker to denote fairly-well-scoped-out bugs or feature requests that the community can pick up and work on. If any of those labeled issues do not have enough information, please feel free to ask constructive questions. (This applies to any open issue.)

Running Forge locally

The easiest way to test changes to Forge during development is by symlinking your local packages to a sample Forge project.

To create symlinks for your local Forge packages, use the yarn link:prepare command after building Forge.

yarn build
yarn link:prepare

Then, you want to initialize a new project with the electron-forge init command (which is the underlying CLI command for create-electron-app). To use the symlinks you created in the last step, pass in the LINK_FORGE_DEPENDENCIES_ON_INIT=1 environment variable.

You can choose to run this command via your local build as shown below or run the production init for versions 6.0.1 and up.

LINK_FORGE_DEPENDENCIES_ON_INIT=1 node path/to/forge/packages/api/cli/dist/electron-forge-init.js my-app

To link an existing project to your local Forge packages, use the yarn link:prepare command as listed above, and then run the following command in your project:

yarn link @electron-forge/core --link-folder=path/to/forge/.links

Forge commands executed in your my-app sample project should reflect any changes in your local Forge build. (Make sure to run yarn build:fast or yarn build between code changes.)

Documentation changes

When changing the API documentation, here are some rules to keep in mind.

  • The first line:
    • should end with a period
    • should be in imperative mood (e.g., "Create" instead of "Creates")
    • First line should not be the function's "signature"
  • The first word of the first line:
    • should be properly capitalized
    • should not be "This"

For changes to the website (electronforge.io), please file issues/pull requests at its separate repository.

Changing the Code

An example of how to make your own code edits:

git clone https://github.com/electron/forge
cd forge
# Installs all dependencies
yarn
# Builds all the TS code
yarn build

Making Commits

Please ensure that all changes are committed using semantic commit messages.

Running the Tests

The Electron Forge repository has a lot of tests, some of which take a decent amount of time to run.

yarn test

Filing Pull Requests

Here are some things to keep in mind as you file pull requests to fix bugs, add new features, etc.:

  • GitHub Actions are used to make sure that the project builds packages as expected on the supported platforms, using supported Node.js versions, and that the project conforms to the configured coding standards.
  • Unless it's impractical, please write tests for your changes. This will help us so that we can spot regressions much easier.
  • If your PR changes the behavior of an existing feature, or adds a new feature, please add/edit the package's documentation.
  • Commit messages and pull request titles should adhere to the Conventional Commits format.
  • One of the philosophies of the project is to keep the code base as small as possible. If you are adding a new feature, think about whether it is appropriate to go into a separate Node module, and then be integrated into this project.
  • Please do not bump the version number in your pull requests, the maintainers will do that. Feel free to indicate whether the changes are a breaking change in behavior.
  • If you are continuing the work of another person's PR and need to rebase/squash, please retain the attribution of the original author(s) and continue the work in subsequent commits.

Release process

This guide is for maintainers who have:

  • Push access to the electron/forge repository.
  • Collaborator access to the @electron-forge packages on npm.

1. Prepare your local code checkout

  • Switch to the tip of the main branch with git switch main && git pull.
  • Run tests locally with yarn test.
  • Check that the latest CI run passed on main on GitHub.
  • Remove all untracked files and directories from your checkout with git clean -fdx.
  • Install dependencies with yarn install.

2. Publish all npm packages

  • Log into npm with npm login.
  • Run the yarn lerna:publish command.
  • Enter your npm account's time-based one-time password (TOTP).

The lerna:publish script will automatically increment the next package version based on the Conventional Commits standard. From there, it does two things:

  1. It creates a tagged commit that bumps the version number in package.json at the root and package levels and pushes the commit and tag to GitHub.
  2. It publishes every @electron-forge/ package to npm.

3. Publish release to GitHub

Adding a new @electron-forge package

Occasionally, we add new packages to the @electron-forge monorepo. Before publishing, ensure that all version numbers for both the package itself and its dependencies match the current version of Electron Forge (e.g. if the current version is v7.0.0 and you want to add the package in v7.1.0, please publish v7.0.0 first).

Then, manually publish the package to the current Forge version using npm publish --access public. Once this version is published, you can continue with the normal release process as usual.

Note

To verify that the publish configuration is correct, first run npm publish --dry-run before publishing.

We do this manual publish step first to avoid errors with attempting to publish a non existent package with Lerna.