signal-desktop/CONTRIBUTING.md

Contributor Guidelines

Installation

• Clone the repo
• Open Chrome
• Go to chrome://extensions/
• Enable developer mode (checkbox on the top right)
• Click "Load unpacked extension..."
• Point to the repo directory

Developer Setup

For development, you should always be using the staging server Registrations on the staging server are completely partitioned from the productions server that the mobile apps use. A production app from the Play store or iTunes is hard-coded to connect to the production server. If you wish to pair your phone and computer, or test sending between the browser and mobile, you must build a mobile client that targets the staging server (see below, under Pairing).

Important! The staging server uses a self-signed ssl certificate. By default, your browser will reject this certificate as insecure. Therefore, in order to register or send and receive messages of any kind, you must first visit https://textsecure-service-staging.whispersystems.org/ in a new tab and click through the warnings to allow the certificate. If done successfully, you should get a 404 from the server. If at any time in the future you notice a console error about an "INSECURE RESPONSE" or "Handshake was canceled", repeat this step.

Pairing

Currently only the Android client supports multi-device pairing.

0. Build TextSecure for Android from source, and change its PUSH_URL to point at textsecure-service-staging.whispersystems.org. This task is 1% search and replace, 99% setting up your build environment.
1. Upon installing the extension you will be presented with a qr code.
2. Scan the qr code with an barcode/qr scanning app and open the resulting url ("tsdevice://...").
3. The phone will ask you to confirm adding the device. Click ok.
4. The browser will then ask you to confirm your phone number and enter a device name. Click ok and wait for setup to complete. Key generation can take up to a minute.

Standalone Registration

NOTE: This is only for developers and will not be presented to users.

• Open the background page and run the following command in the console: extension.install("standalone").
• Enter a real phone number (Google Voice numbers work too) and country combination and choose to send an SMS. You will receive a real SMS.
• Enter the verification code you received by SMS.
• Wait for key generation to complete.

You should now be able to use the extension.

Chrome profiles

Don't have any friends to help you test the extension? Make a couple of Chrome profiles. Each one will need its own Google account and Google Voice number. Each one will have to repeat the setup process documented above, including re-accepting the staging server cert under each profile. This is a tedious process, but once you are done you will be able to send messages back and forth between different profiles, allowing you to observe both endpoints of a conversation.

Pull requests

So you wanna make a pull request? Please observe the following guidelines.

• Rebase your changes on the latest master branch, resolving any conflicts that may arise. This ensures that your changes will merge cleanly when you open your PR.
• Run the tests locally by opening the test page in your browser. A test-breaking change will not be merged.
• Make sure the diff between our master and your branch contains only the minimal set of changes needed to implement your feature or bugfix. This will make it easier for the person reviewing your code to approve the changes. Please do not submit a PR with commented out code or unfinished features.
• Don't have too many commits. If your branch contains spurious commits along the lines of "Oops, reverted this change" or "Just experiementing, will delete this later", please squash or rebase those changes out.
• Don't have too few commits. If you have a complicated or long lived feature branch, it may make sense to break the changes up into logical atomic chunks to aid in the review process.
• Provide a well written and nicely formatted commit message. See this link for some tips on formatting. As far as content, try to include in your summary
  1. What you changed
  2. Why this change was made (including git issue # if appropriate)
  3. Any relevant technical details or motivations for your implementation choices that may be helpful to someone reviewing or auditing the commit history in the future. When in doubt, err on the side of a longer commit message.

Dependencies

Note: Unless you need to make changes to dependencies, you can skip this section and just use the checked in versions.

Dependencies are managed by bower and built with grunt. To change them, you'll need to install node and npm, then run npm install to install bower, grunt, and related plugins.

Adding a bower component

Add the package name and version to bower.json under 'dependencies' or bower install package-name --save

Next update the "preen" config in bower.json with the list of files we will actually use from the new package, e.g.:

  "preen": {
    "package-name": [
      "path/to/main.js",
      "directory/**/*.js"
    ],
    ...
  }

If you'd like to add the new dependency to js/components.js to be included on all html pages, simply append the package name to the concat.app list in bower.json. Take care to insert it in the order you would like it concatenated.

Now, run grunt to delete unused package files and build js/components.js.

Finally, stage and commit changes to bower.json, js/components.js, and components/. The latter should be limited to files we actually use.

Tests

Please write tests! Our testing framework is mocha and our assertion library is chai.

To run tests, use grunt dev or grunt connect watch to spin up a local webserver, then point your browser to localhost:9999/test/index.html and localhost:9999/libtextsecure/test/index.html