This is part of The Brunch.io Guide.
Let’s start exploring everything you need to know to happily use Brunch in your own projects, be they new ones or legacy codebases.
Just like other tools in this space (and like a truckload of other tools these days), Brunch is based on Node.js and gets installed through npm. You can install it globally, so as to be able to use the brunch command from anywhere:
npm install -g brunch…but I would advise you to install it locally as well, within your current project, that will need a package.json file anyway (just like with other competing tools). This way, you can use various versions of Brunch from one project to another, with no conflicts. The thing is, the globally-installed Brunch CLI will correctly defer to your project's local version for actual processing, so you get the best of both worlds!
Brunch emphasizes the concept of skeleton. Unlike generators such as those based on Yeoman, these are just Git repos suggesting a default infrastructure for a front-end app using Brunch in development. There are a number of existing skeletons, and their main usefulness lies in providing a kickstarter by creating a few folders, pre-requiring dependencies in package.json and providing a default Brunch configuration.
Brunch has a brunch new command that you pass a GitHub repo to (or any Git repo URL, for that matter), and an optional clone path. This is really just a git clone followed by an npm install…
I would advise you to start from scratch (or from your own codebase) and then define your own folders and configuration, to get a firm grip on what you’re doing.
And yes, this was a tiny chapter. Not so with the next ones!
« Previous: Brunch?! What’s Brunch? • Next: Conventions and Defaults »