Getting Started

Installing git

Fedora 7, Ubuntu 8, and later: The git-core package is available through the standard package repositories. Other Linux and BSD environments should be similar, eg, emerge git on Gentoo will install the needed commands and dependencies. If your platform does not package git, you can download the latest stable release from "". Windows users are recommended to install either TortiseGit or Cygwin (the latter includes a bash shell and many other Linux programs).

Basic Tasks

First, note that you can get documentation for a command such as git log —graph with:

$ man git-log


$ git help log

With the latter command, you can use the manual viewer of your choice; see the git-help(1) man page for more information.

You should configure git with your name and preferred email address before doing any operation. The easiest way to do so is:

$ git config --global "Your Name Here"
$ git config --global

To download a new package from a git repository:

$ git clone

To avoid the username part of the above URL, use a .netrc file to store your login ID, or use an ssh public key on github and the ssh URL:

$ git clone

To update a package to the latest upstream version ("fast-forward merge"):

$ git pull origin <branch-name>

or more simply, to pull from the default branch/location from which you cloned:

$ git pull

will pull from the origin repository and default branch defined in the package-name/.git/config file.

One way to undo all local modifications:

$ git checkout -f

To check in your own local modifications (e.g. update the web site, do some refactoring, fix a bug, or apply a patch):

$ vi vct-custom.css


Don’t forget to run ‘git add’ and ‘git rm’ if adding or removing files.

To check in all local modifications to your local repository:

$ git commit -a -m "added new stylesheet and updated pelican config for enabling plugins"

Undo recent commits

If the commits are local only, then "git reset" is probably appropriate. OTOH, if the commits are already public, ie, they’ve been pushed to a remote repository and potentially cloned by someone else, then "git reset" is most likely not the right answer. That said, if you’re working completely by yourself, then any method is viable (again, depending on what your goals are).

To make one or more commits go away cleanly when working with others, the right tool is almost certainly "git revert". You can specify one commit or a range, and git will make a new commit that exactly reverts the changes made by the specified commit(s). Suppose you wanted to get rid of two commits, and you’ve already made two new commits (on top of the bad ones) that you want to keep. First, get the commit hashes for the two bad commits, then revert them:

$ git log --oneline | head -n 4

498e425 added three new drafts, still need metadata
87b2a14 latest updates to static pages
6dcdf91 added artile template with example rst metadata
6fcd2c0 Add note for ubuntu users to use apt-get version of pip instead

$ git revert 6fcd2c0..6dcdf91

Sometimes you have made a few commits, or just pulled a change, and simply want those commits to go away completely:

$ git reset --hard HEAD~2   # erase last 2 commits

This will essentially erase the top two commits, as if you had never made them. DO NOT do this, if you’ve already pushed said commits (at least not without coordination with others who may have pulled those commis). Note that this is quite different from git revert, which applies a reversed patch as an additional commit.

Listing changes in your working dir, in diff format

Display changes since last ‘git add’ or ‘git rm’:

$ git diff

Display changes since last commit:

$ git diff HEAD

Obtain summary of all changes in working dir:

$ git status

List all commits on the current branch, with descriptions:

$ git log

The ‘git log’ option "-p" shows diffs in addition to commit messages. The option "—stat" shows the diffstat.

List all commits to a specific file:

$ git log content/pages/contact.rst



List all local branches (add -a to see remote branches too):

$ git branch

Make desired branch current in working directory:

$ git checkout $branch

Create a new branch from master, and make it current:

$ git checkout -b alternate-theme master

Examine which branch is current:

$ git status

(‘git branch’ also shows you the current branch, using a "*" in front)

Obtain a diff between current branch, and master branch

In most trees with branches, .git/refs/heads/master contains the current ‘vanilla’ upstream tree, for easy diffing and merging. (in trees without branches, ‘master’ simply contains your latest changes). The following is equivalent to git diff HEAD, when used with HEAD branch:

$ git diff master..HEAD

Obtain a list of changes between current branch, and master branch:

$ git log master..HEAD

(this is equivalent to git log, when used with HEAD)

Rather than full changeset descriptions, obtain a one-line summary of each changes:

$ git shortlog master..HEAD

Merging changes from one branch to another

Suppose that you do work on two different branches, and after work on those two branches is complete, you merge the work into master:

$ git checkout master       # switch to branch master
$ git merge drafts          # merge drafts into master
$ git merge new-theme       # merge new-theme into master

Misc. Topics

Optimize your repository

git is heavily optimized for fast storage and retrieval on a per-command basis. However, over a long period of time, it can be useful to perform further optimizations, including packing all git objects into single "packfile" for fast retrieval and less wasted disk space. The following:

$ git gc

will optimize your repository. You don’t need to run this frequently — git is quite fast even without it. See the ‘git gc’ man page for more details.

Don’t forget to download tags from time to time

Doing a "git pull" only downloads new commits from the remote, and updates the requested remote head. This misses updates to the .git/refs/tags/ and .git/refs/heads/ directories. For tags, run git fetch —tags in your local repo.

Tagging a particular commit

In many cases, you will want to give interesting or significant commits a name, known as a tag. The Linux kernel uses tags for each kernel version: "v2.6.21", "v2.6.22", etc. For example, to create a new tag after a particular commit:

$ git tag my-tag

This creates a new tag named "my-tag", based on the current commit. You can also make an "annotated" tag, or a GPG-signed tag, so read the man page for more details.

Further reading

A (larger) good introduction is the Git tutorial

More complete documentation is available in the Git community book, as well as the Git Reference and git man page documentation.

And for even more information on Git, check out the Pro Git book.


This article was originally adapted and expanded from another Git Intro found on the web; I just can’t remember where :/