Mailspring/CONTRIBUTING.md
2016-01-21 10:42:31 -08:00

6.5 KiB

Filing an Issue

Thanks for checking out N1! If you have a feature request, be sure to check out the open source roadmap. If someone has already requested the feature you have in mind, you can upvote the card on Trello—to keep things organized, we often close feature requests on GitHub after creating Trello cards.

If you've found a bug, try searching for similars issue before filing a new one. Please include the version of N1 you're using, the platform you're using (Mac / Windows / Linux), and the type of email account. (Gmail, Outlook 365, etc.)

Contributing to N1

The hosted sync engine allows us to control adoption of N1 and maintain a great experience for our users. However, the sync engine is open source and you can set it up yourself to begin using N1 immediately. Follow instructions on the sync engine repository.

Getting Started

First, clone and build N1 from source:

git clone https://github.com/nylas/N1.git
cd N1
script/bootstrap

Read the getting started guides.

See Windows instructions here

Linux users on Debian 8 and Ubuntu 15.04 onward must also install libgcrypt11, which Electron depends on.

Running N1

./N1.sh --dev

Once the app boots, you'll be prompted to enter your email credentials.

Testing N1

./N1.sh --test

This will run the full suite of automated unit tests. We use Jasmine 1.3.

It runs all tests inside of the /spec folder and all tests inside of /internal_packages/**/spec

Creating binaries

Once you've checked out N1 and run script/bootstrap, you can create a packaged version of the application by running script/build. Note that the builds available at https://nylas.com/N1 include licensed fonts, sounds, and other improvements. If you're just looking to run N1, you should download it there!

Pull requests

We require all authors sign our Contributor License Agreement before pull requests (even minor ones) can be accepted. (It's similar to other projects, like NodeJS Meteor, or React). I'm really sorry, but Legal made us do it.

Commit Format

We decided to not impose super strict commit guidelines on the community.

We're trusting you to be thoughtful, responsible, committers.

We do have a few heuristics:

  • Keep commits fairly isolated. Don't jam lots of different functionality in 1 squashed commit. git bisect and git cherry-pick should still be reasonable things to do.
  • Keep commits fairly significant. DO squash all those little file changes and "fixmes". Don't make it difficult to browse our history. Play the balance between this idea and the last point. If a commit doesn't deserve your time to write a long thoughtful message about, then squash it.
  • Be hyper-descriptive in your commit messages. I care less about what you did (I can read the code), I want to know WHY you did it. Put that in the commit body (not the subject). Itemize the major semantic changes that happened.
  • Read "How to Write a Git Commit Message" if you haven't already (but don't be too prescriptivist about it!)

Running Against Open Source Sync Engine

N1 needs to fetch mail from a running instance of the Nylas Sync Engine. The Sync Engine is what abstracts away IMAP, POP, and SMTP to serve your email on any provider through a modern, RESTful API.

By default the N1 source points to our hosted version of the sync-engine; however, the Sync Engine is open source and you can run it yourself.

  1. Install the Nylas Sync Engine in a Vagrant virtual machine by following the installation and setup instructions.

  2. Once you've installed the sync engine, add accounts by running the inbox-auth script. For Gmail accounts, the syntax is simple: bin/inbox-auth you@gmail.com

  3. Start the sync engine by running bin/inbox-start and the API via bin/inbox-api.

  4. After you've linked accounts to the Sync Engine, open or create a file at ~/.nylas/config.cson. This is the config file that N1 reads at launch.

    Replace env: "production" with env: "local" at the top level of the config. This tells N1 to look at localhost:5555 for the sync engine. If you've deployed the sync engine elsewhere, add the following block beneath env: "local":

    syncEngine:
      APIRoot: http://mysite.com:5555
    

    Copy the JSON array of accounts returned from the Sync Engine's /accounts endpoint (ex. http://localhost:5555/accounts) into the config file at the path *.nylas.accounts.

    N1 will look for access tokens for these accounts under *.nylas.accountTokens, but the open source version of the sync engine does not provide access tokens. When you make requests to the open source API, you provide an account ID in the HTTP Basic Auth username field instead of an account token.

    For each account you've created, add an entry to *.nylas.accountTokens with the account ID as both the key and value.

    The final config.cson file should look something like this:

     "*":
       env: "local"
       nylas:
         accounts: [
           {
             server_id: "{ACCOUNT_ID_1}"
             object: "account"
             account_id: "{ACCOUNT_ID_1}"
             name: "{YOUR NAME}"
             provider: "{PROVIDER_NAME}"
             email_address: "{YOUR_EMAIL_ADDRESS}"
             organization_unit: "{folder or label}"
             id: "{ACCOUNT_ID_1}"
           }
           {
             server_id: "{ACCOUNT_ID_2}"
             object: "account"
             account_id: "{ACCOUNT_ID_2}"
             name: "{YOUR_NAME}"
             provider: "{PROVIDER_NAME}"
             email_address: "{YOUR_EMAIL_ADDRESS}"
             organization_unit: "{folder or label}"
             id: "{ACCOUNT_ID_2}"
           }
         ]
         accountTokens:
           "{ACCOUNT_ID_1}": "{ACCOUNT_ID_1}"
           "{ACCOUNT_ID_2}": "{ACCOUNT_ID_2}"
    

Note: {ACCOUNT_ID_1} refers to the database ID of the Account object you create when setting up the Sync Engine. The JSON above should match fairly closely with the Sync Engine Account object.