Branches¶
In Git, a branch is simply a pointer to a specific commit in your repository. Branches are used when you want to, e.g., add a new feature to your project, or implement a hotfix, while keeping your reference point (the commit or branch that your new branch has been created from) intact.
When creating a branch, you more or less tell Git: "Give me a separate starting point, including all the work from this commit and all the commits before it."
The main branch¶
After initialising a repository, you always start out with a single branch called main.
Why you should not directly commit into the main branch
For reasons, which I will get into later, it is extremely bad practice (and very frowned upon in the open source community) to actively commit into your main branch.
This branch should generally be your release version of the project only - and should therefore always be clean and bug-free. Unjustifiably committing changes into this branch could heavily disrupt your project end-user's experience.
It is best practice to (at least) have a separate development branch, in which you make your changes, which you subsequently merge into your main branch, once new features have been properly completed and tested.
Note
The initial branch of the Git repository could also be named master. This is usually the case in older Git versions. The change to the default naming convention main has occurred after the general uproar around the terms master/slave.