by Tom Preston-Werner, Peter Pistorius, and Rob Cameron.
Redwood is an opinionated, full-stack, serverless web application framework that will allow you to build and deploy JAMstack applications with ease. Imagine a React frontend, statically delivered by CDN, that talks via GraphQL to your backend running on AWS Lambdas around the world, all deployable with just a
git push—that's Redwood. By making a lot of decisions for you, Redwood lets you get to work on what makes your application special, instead of wasting cycles choosing and re-choosing various technologies and configurations. Plus, because Redwood is a proper framework, you benefit from continued performance and feature upgrades over time and with minimum effort.
WARNING: Redwood is still in the alpha phase of development. We do not recommend that you use Redwood applications in production at this time. That said, your input can have a huge impact during this period, and we welcome your feedback and ideas! Check out the Redwood Community Forum to join in.
TUTORIAL: The best way to get to know Redwood is by going through the extensive Redwood Tutorial. Have fun!
EXAMPLES: If you'd like to see some examples of what a Redwood application looks like, take a look at the following projects:
Here's a quick taste of the technologies a standard Redwood application will use:
- Opinionated defaults for formatting, file organization, webpack, Babel, and more.
- Simple but powerful routing (all routes defined in one file) with dynamic (typed) parameters, constraints, and named route functions (to generate correct URLs).
- Automatic page-based code-splitting.
- Boilerplate-less GraphQL API construction.
- Cells: a declarative way to fetch data from the backend API.
- Generators for pages, layouts, cells, SDL, services, etc.
- Scaffold generator for CRUD operations around a specific DB table.
- Forms with easy client- and/or server-side validation and error handling.
- Hot module replacement (HMR) for faster development.
- Database migrations (via Prisma 2).
- First class JAMstack-style deployment to Netlify.
The Redwood philosophy
Redwood believes that JAMstack is a huge leap forward in how we can write web applications that are easy to write, deploy, scale, and maintain.
Redwood believes that there is power in standards, and makes decisions for you about which technologies to use, how to organize your code into files, and how to name things. With a shared understanding of the Redwood conventions, a developer should be able to jump into any Redwood application and get up to speed very quickly.
Redwood believes that traditional, relational databases like PostgreSQL and MySQL are still the workhorses of today's web applications and should be first-class citizens. However, Redwood also shines with NoSQL databases.
Redwood believes that, as much as possible, you should be able to operate in a serverless mindset and deploy to a generic computational grid. This helps unlock the next point...
Redwood believes that deployment and scaling should be super easy. To deploy your application, you should only need to commit and push to your Git repository. To scale from zero to thousands of users should not require your intervention. The principles of JAMstack and serverless make this possible.
Redwood believes that it should be equally useful for writing both simple, toy applications and complex, mission-critical applications. In addition, it should require very little operational work to grow your app from the former to the latter.
How it works
A Redwood application is split into two parts: a frontend and a backend. This is represented as two node projects within a single monorepo. We use Yarn to make it easy to operate across both projects while keeping them in a single Git repository.
The frontend project is called
web and the backend project is called
api. For clarity, we will refer to these in prose as "sides", i.e. the "web side" and the "api side". They are separate projects because code on the web side will end up running in the user's browser while code on the api side will run on a server somewhere. It is important that you keep this distinction clear in your mind as you develop your application. The two separate projects are intended to make this obvious. In addition, separate projects allow for different dependencies and build processes for each project.
The api side is an implementation of a GraphQL API. Your business logic is organized into "services" that represent their own internal API and can be called both from external GraphQL requests and other internal services. Redwood can automatically connect your internal services with Apollo, reducing the amount of boilerplate you have to write. Your services can interact with a database via Prisma's ORM, and Prisma's migration tooling provides first-class migrations that take the pain out of evolving your database schema.
The web side is built with React. Redwood's router makes it simple to map URL paths to React "Page" components (and automatically code-split your app on each route). Pages may contain a "Layout" component to wrap content. They also contain "Cells" and regular React components. Cells allow you to declaratively manage the lifecycle of a component that fetches and displays data. Other Redwood utility components make it trivial to implement smart forms and various common needs. An ideal development flow starts with Storybook entries and Jest tests, so Redwood closely integrates both, making it easy to do the right thing.
You'll notice that the web side is called "web" and not "frontend". This is because Redwood conceives of a world where you may have other sides like "mobile", "desktop", "cli", etc., all consuming the same GraphQL API and living in the same monorepo.
How can it be serverless if it involves a GraphQL API and database?
I'm glad you asked! Currently, Redwood can deploy your GraphQL API to a Lambda function. This is not appropriate for all use cases, but on hosting providers like Netlify, it makes deployment a breeze. As time goes on, "functions" will continue to enjoy performance improvements which will further increase the number of use cases that can take advantage of this technology.
Databases are a little trickier, especially the traditional relational ones like PostgreSQL and MySQL. Right now, you still need to set these up manually, but we are working hard with Netlify and other providers to fulfill the serverless dream here too.
Redwood is intentionally pushing the boundaries of what's possible with JAMstack. In fact, the whole reason I (Tom) started working on Redwood is because of a tweet I posted some time ago:
Prediction: within 5 years, you’ll build your next large scale, fully featured web app with #JAMstack and deploy on @Netlify. —@mojombo • 9 July 2018
I kept waiting for a high quality full-stack framework to arrive, but it didn't, so I decided to take matters into my own hands. And that's why Redwood exists.
If you are like minded, then I hope you'll join me in helping build Redwood and hasten the arrival of the future I predicted!
Why is it called Redwood?
(A history, by Tom Preston-Werner)
Where I live in Northern California there is a type of tree called a redwood. Redwoods are HUGE, the tallest in the world, some topping out at 115 meters (380 feet) in height. The eldest of the still-living redwoods sprouted from the ground an astonishing 3,200 years ago. To stand among them is transcendent. Sometimes, when I need to think or be creative, I will journey to my favorite grove of redwoods and walk among these giants, soaking myself in their silent grandeur.
In addition, Redwoods have a few properties that I thought would be aspirational for my nascent web app framework. Namely:
Redwoods are beautiful as saplings, and grow to be majestic. What if you could feel that way about your web app?
Redwood pinecones are dense and surprisingly small. Can we allow you to get more done with less code?
Redwood trees are resistant to fire. Surprisingly robust to disaster scenarios, just like a great web framework should be!
Redwoods appear complex from afar, but simple up close. Their branching structure provides order and allows for emergent complexity within a simple framework. Can a web framework do the same?
And there you have it.