Branching and switching

A branch is an independent line of work. You can try something — a feature, a fix, a risky idea — without touching the stable version, then either keep it or throw it away. Branches are the foundation of working safely, alone or with others (including agents whose work you want to review before it's official).

Create and switch

The modern, clear commands are git switch and git branch:

git branch                 # list branches; * marks the current one
git switch -c add-login    # create a new branch and switch to it
# ...edit, add, commit on this branch...
git switch main            # go back to the main line

-c means "create." Without it, git switch <name> just moves to an existing branch. (You may see git checkout in older material — switch is the newer, less overloaded command for the same job.)

Why branches are cheap

A branch is essentially a lightweight pointer to a commit, so creating one is near-instant and costs almost nothing. That cheapness is the point: make a branch for every distinct piece of work. If it pans out, keep it; if not, delete it (git branch -d add-login) and your main line never knew it existed.

Commits you make on a branch stay on that branch until you merge — main is untouched while you experiment.

Bringing a branch back: merge

When a branch is done, merge it into main:

git switch main
git merge add-login

If main hasn't changed in the meantime, git does a clean "fast-forward." If both lines changed different files, git combines them automatically. If they changed the same lines, you get a merge conflict — covered in the intermediate Git module.

Where to go next

You can work in isolation and merge back. Next: working with remotes — sharing your repository with a server like GitHub so your work is backed up and collaborative.