Git is a decentralized or distributed system. This means that the revisions of documents do not only exist on a server and we get to keep some local copy of the document, but that all revisions are stored on every “clone” of the “repository”.
Repository is the name of the place where the documents and their history are saved. This is usually a directory on your local file system. Inside this repository you will find a directory with the name “.git”.
This directory contains all sorts of files that make up a complicated database like structure. These files contain information about branches, revisions, tags, other clones of this repository and much more.
While it is possible to have a repository on your local system only, it is more common to have a repository on a dedicated server and have a clone of this repository on multiple local machines (used by developers for example).
One of the reasons for keeping a repository on a server that can be accessed by the people that need it is that it is easy to work together this way. It is a big improvement over sending each other files through email and trying to get things to work every time you receive an edited file.
Working with Git using a server enables us to have the same documents available at all times.
Since we are all keeping clones of the repository, an added advantage is that we also have backups of the documents (and their history) on each machine.