Deployments can be found under following urls:
master branch: https://bears-18.now.sh
develop branch: https://bears-18-dev.now.sh
The application consists of several components, a rough outline:
The backend consists of the following pieces:
- MongoDB as a database, this servers as our persistant storage layer.
- NodeJS as the server side runtime(Javascript).
- Express as the web framework to handle incoming requests
- GraphQL Server as the data endpoint for the frontend application.
- The GraphQL endpoint will be responsible to present our MongoDB to our frontend in a convenient matter.
- There will only be a single endpoint and the frontend will not have to worry about where all the data is coming from or how to combine things.
- GraphQL can also aggregate the external API's and route to the client side app in a friendly manner.
- It sits inbetween our database + external information sources and the frontend app.
This will be the bulk of the backend application. There will be a small part in the serverside code that will be reponsible to bootup NextJS. More on NextJS in the Frontend section.
Frontend consists of the following pieces:
- NextJS as a framework for a server-side rendered React Application.
- NextJS solves a lot of the setup and boilerplate that you'd need to get a serverside rendered react application going.
- It provides an extendable webpack and babel configuration and takes care of the regular boilerplate.
- It's similar to create-react-app, in the sense that it tries to simplify the development experience.
- The benefit of using NextJS is a server-side rendered application, built in codesplitting which leads to a fast bootup and following navigation experience for end users.
- React as a view library to build composible UI's in a declaritive manner.
- GraphQL Client to interact with the GraphQL server and feed the data to our React Application.
This workflow is based on the Git Flow method.
We have 2 main branches, namely develop and master.
Master is meant as the release branch and our goal should be to push something meaningful to this branch that works after every sprint.
This branch is also protected from direct pushes and you always need to make changes to it via a Pull Request (PR).
The main branch you'll be using will be the develop branch.
When you create a specific feature branch, you choose develop as the starter and target it for the merge as well. You can use certain keywords to automatically close issues if your PR solves those issues.
For clarity we can add the issue number in the branch name or PR, followed by a short description. For example:
1/document-github-workflow.
So in short:
master-> For release, so one PR with all the changes from last week fromdevelopdevelop-> Accumulator branch for active development, here all the features will get merged.Feature branches-> Your own dev branch based fromdevelopto complete your tickets!