Git basics

April 30, 2026 · View on GitHub

A git branch is a movable pointer to a commit used to work on a separate line of development.

A commit is a recorded snapshot of the repository at a point in time and is identified by a SHA hash.

A repository is the version-controlled project that contains the working tree, commits, branches, and refs.

Version control records changes over time so you can view history, revert mistakes, and collaborate safely.

Why use branches: isolate features, experiment safely, and collaborate without blocking the main history.

Why commit often: small, focused commits make review and debugging easier.

Common commands:

git status — shows staged and unstaged changes, untracked files, and branch state. Run it frequently to see what will be committed.
git add — stage changes to be included in the next commit. Use 'git add -p' to interactively stage hunks.
git commit -m "message" — record staged changes as a commit with a descriptive message. Make small, focused commits.
git log — view commit history; use --oneline or --graph for compact views.
git branch — list and manage branches; use 'git branch ' to create.
git switch — change branches safely; use 'git switch -c ' to create and switch.
git push -u origin — publish a local branch to remote and set upstream for future pushes/pulls.
git pull --ff-only — update from upstream only if fast-forward is possible; prevents unexpected merge commits.

Safety tips:

Use descriptive, short branch names (feature/desc, fix/desc).
Commit small, focused changes with clear messages.
Push branches to the remote for backup and collaboration.
Use pull --ff-only to avoid unexpected merge commits when you want a linear history.