Git works even when GitHub is down — here's why

Git is the version-control system that quietly runs modern software. Created by Linus Torvalds in 2005 in about two weeks — after the Linux kernel lost free access to its previous tool — it tracks every change to a project's files so teams can work in parallel, branch off to experiment, and merge back without overwriting each other.

The surprising part: Git is distributed. When you git clone a repository, you don't just download the latest files — you copy the entire history of every commit onto your own machine. That's why you can commit, branch, view history, and diff changes with no internet connection. Services like GitHub or GitLab are just shared meeting points; the real repository lives on every contributor's laptop.

Under the hood, Git identifies each commit not by a version number but by a SHA hash — a fingerprint computed from the file contents plus the history that came before it. Because each commit's hash depends on its parent, the history forms a tamper-evident chain: change anything in the past and every later hash changes too. This is also why Git can tell instantly whether two copies of a project are identical.

Git itself is open source (a GPLv2 project), which is a big reason it beat commercial rivals — anyone could inspect it, host it, and build tools on top, from GitHub to the git command baked into nearly every developer's terminal.