If you are new to git entirely, I'd recommend checking out and running through the examples they host there.

The Git-scm Book is a good read and also serves as a great online reference.

And when all else fails, ohshitgit

Create a Repository

To create a repository, just create a directory or enter the root directory of the project you want to turn into a repository, and type git init. This initializes the directory as a local repository. To add the repository to GitHub and track it remotely, you'll need to login to GitHub and click 'New Repository', name the repository the same as your root folder and continue. GitHub will provide you with the rest of the instructions, but for completeness, tweak the lines below to push your local repository to your new remote on GitHub -

git remote add origin<username>/<reponame>.git
git push -u origin master

We just created the origin remote. This is the remote that is displayed and tracked on GitHub, when you clone your repository you are on a local remote, which means until you git push <remote> <branch> your changes will only be saved and tracked on your local machine.

Using a remote via SSH such as the above<username>/<reponame>.git requires you have configured an SSH key with your GitHub account that is associated with the machine you are pushing from. If you haven't already, check out Creating SSH Login Keys and simply cat ~/.ssh/<USERKEY>.pub the public key of your user and copy it over into your GitHub settings.

If you'd rather not mess with things like this, see how to create a Person Access Token below.

Ignoring Files

Within a Git repository, .gitignore files can be seen specifying a list of files or directories that Git should ignore when tracking changes. For example, this is useful when a project is expected to contain build files generated after being cloned. We would not want the user to then make a commit publishing the files they generated when building the project for their system, we would want to provide a clean slate for the next person that clones the project.

After creating a .gitignore file, the syntax below can be followed to specify files and directories to be ignored.

# ignore all .a files

# but do track lib.a, even though you're ignoring .a files above

# only ignore the TODO file in the current directory, not subdir/TODO

# ignore all files in any directory named build

# ignore doc/notes.txt, but not doc/server/arch.txt

# ignore all .pdf files in the doc/ directory and any of its subdirectories

Git-scm - Ignoring Files

Check out GitHub - gitignore repo for some templates used in popular languages, like C, C++, or Python.

If you are ignoring a file that Git has already previously tracked, it may be necessary to remove the file (or directory) from Git's cache using the command below

# Remove cached file
git rm --cached path/to/file
# Remove cached directory
git rm -r --cached path/to/
# Remove all cached files
git rm -r --cached .

Authenticating with Git

There are many ways to authenticate with Git when pushing to remote repositories. See some of the below for examples.

Credential Caching

If you don’t want to authenticate every time you push, you can set up a “credential cache”. The simplest is just to keep it in memory for a few minutes, which you can easily set up by running git config --global credential.helper cache.

SSH Keys

You can use ssh-keygen to generate a key and the manually add the key to your GitHub profile.
To do this, run ssh-keygen -t ed25519 and follow the prompts. Be sure to include the file path when naming your key, or the key will be output into your working directory. Once generated, simply cat and copy / paste your public key into the field within your settings on GitHub.To do this, you'll need to run git remote set-url origin in order to configure your local repository to use SSH private keys when connecting to git. For more on ssh-keygen, check out Creating SSH Login Keys.

Personal Access Tokens

Alternatively, you could generate a static Personal Access Token - a token that once generated can be paired with a YubiKey or similar product. This allows you to clone / work from anywhere without having to provision or SSH keys or manage long passwords / 2FA methods. Plug your key into USB and tap your desired configuration and the static access key will be input, allowing immediate access for this one time. So, every time you push, unless you configure otherwise, you will have to enter this token by tapping the YubiKey.

Commit Guidelines

Before you commit, run git diff --check, which identifies possible whitespace errors and lists them for you.

Write your commit message in the imperative: "Fix bug" and not "Fixed bug"
or "Fixes bug." This convention matches up with commit messages generated
by commands like git merge and git revert.

Use git add --patch to partially stage files (covered in detail in )

The project snapshot at the tip of the branch is identical whether you do one commit or five, as long as all the changes are added at some point

As a general rule, your commit messages should start with a single line that’s no more than about 50 characters and that describes the changeset concisely, followed by a blank line, followed by a more detailed explanation

git log --pretty=oneline shows a terse history mapping containing the commit id and the summary

If the config option merge.summary is set, the summaries from all merged commits will make their way into the merge commit message

git pull --rebase What’s happening here? Git will rewind (undo) all of your local commits, pull down the remote commits then replay your local commits on top of the newly pulled remote commits. If any conflicts arise that git can’t handle you’ll be given the opportunity to manually merge the commits then simply run git rebase --continue to carry on replaying your local commits.

How to Avoid Merge Commits in Git - Kernowsoul

git shortlog uses summary lines in the changelog-like output it produces -
git format-patch, git send-email, and related tools use it as the subject for emails.

Reflogs, a local history accessible with git reflog, is intended to help you recover from stupid mistakes by providing the hashes along with output similar to git shortlog.


If we run ... git push <remote> serverfix
Git automatically expands the serverfix branchname out to refs/heads/serverfix:refs/heads/serverfix
Where the general sytax is local:remote..

Always check your branch with git status before pushing! Especially when pushing to a shared repository, this could have unintended consequences and result in a messy history. See Resolving Conflicts below if you are having conflict issues.

You can use this format to push a local branch into a remote branch that is named differently. If you didn’t want it to be called serverfix on the remote, you could instead run git push origin serverfix:awesomebranch to push your local serverfix branch to the awesomebranch branch on the remote project.

Should you see the errors below when attempting to push, see the Pull / Merge > Resolving Conflicts section of this page for steps on merging your branches, resolving the conflicts, and then completing your push.

[kapper@kanjaro 6-complex-numbers]$ git push origin v0.2 
 ! [rejected]        v0.2 -> v0.2 (non-fast-forward)
error: failed to push some refs to ''
hint: Updates were rejected because the tip of your current branch is behind
hint: its remote counterpart. Integrate the remote changes (e.g.
hint: 'git pull ...') before pushing again.
hint: See the 'Note about fast-forwards' in 'git push --help' for details.
Shared Remote Server Workflow

*You should also be able to share your branches by pushing them to a shared server, working with others on shared branches and rebasing your branches before they are shared. This being said, it’s possible to have a workflow where each developer has write access to their own public repository and read access to everyone else’s -

  1. The project maintainer pushes to their public repository.
  2. A contributor clones that repository and makes changes.
  3. The contributor pushes to their own public copy.
  4. The contributor sends the maintainer an email asking them to pull changes.
  5. The maintainer adds the contributor’s repository as a remote and merges locally.
  6. The maintainer pushes merged changes to the main repository.

Reversing Changes

Use git reset <remote> for local changes -

# one commit in the past
git reset HEAD^

# 2 commits in the past
git reset HEAD^^

# 2 commits in the past
git reset HEAD~2

# 3 commits in the past
git reset HEAD~3

Use git revert <remote> for changes that have already been pushed to a remote -

# 1 commit in the past
git revert HEAD

# 2 commit in the past
git revert HEAD^

# 3 commits in the past
git revert HEAD^^

# 3 commits in the past
git revert HEAD~2

# 4 commits in the past
git revert HEAD~3

# Note that HEAD could be replaced with v0.2, or any active branch that exists on the remote 


Basic git branch commands -

# checkout and create new branch if it doesnt exist
git checkout -b branchname

git log --all --graph --decorate --oneline --simplify-by-decoration will output your history in a format similar to the Network Graph on GitHub

[kapper@kanjaro cmake]$ git log --all --graph --decorate --oneline --simplify-by-decoration
* 8221652 (HEAD -> master, origin/master, origin/HEAD) merge v0.4 into master
* 27e6e1c (origin/v0.4, v0.4) Fix for tab spacing in vim
| * feb1da1 (refs/stash) WIP on master: 807e0b3 Reorganized C problem 4
* 807e0b3 (origin/v0.3) Reorganized C problem 4
* 2fc4266 (origin/v0.2) Finishing up v0.2
| * 9187276 (origin/v0.1) Cleaned up the README.
* f1b858b Initial commit of first CMake project
[kapper@kanjaro cmake]$

The output can be formatted further, and linked with aliases within git -

git config --global alias.lg "log --all --graph --pretty=format:'%Cred%h%Creset -%C(yellow)%d%Creset %s %Cgreen(%cr) %C(bold blue)<%an>%Creset' --abbrev-commit"

There are also other simpler options for similar output with less information -
git show-branch git show-branch --all

Alternatively, if you would rather a GUI - run gitk - provided it's configured correctly.

To delete a branch, local or remote, see the commands below -

# Remove a local branch
git branch -d the_local_branch

# To remove a remote branch (if you know what you are doing!)
git push origin :the_remote_branch
git push origin --delete the_remote_branch
Pull / Merge

Before merging, commit and push a 'checkpoint' to your version or feature branch. If you do not, git will squash the history from your branch into master - this commit can serve as a reference for changes merged into master later on. Should you forget to do this, the merge could still be traced with more effort.

merge master into the test first to resolve any conflicts on the test branch itself. After the test branch is clean, up-to-date, and pushed to origin, I'll git checkout master and git merge test.

git merge origin/master. If you want to fast-forward, run git merge --ff-only origin/master

The --squash option takes all the work on the merged branch and squashes it into one changeset producing the repository state as if a real merge happened, without actually making a merge commit.

Also the --no-commit option can be useful to delay the merge commit in case of the default merge process.

Resolving Conflicts

Problems pushing your local changes to a remote (origin) ?
git pull <remote> <branch> and resolve the conflicts by following the instructions below .

When attempting to pull or merge branches, there can sometimes be new changes to the same content within the same files on the two different branches. Since git wants to be sure that you retain the changes you want, it pauses our merge and prompts us to resolve these conflicts before creating a final commit to finish our merge.

[kapper@kanjaro 6-complex-numbers]$ git pull origin v0.2 
 * branch            v0.2       -> FETCH_HEAD
Auto-merging 4-Ch2-course-laucher/CMakeLists.txt
CONFLICT (content): Merge conflict in 4-Ch2-course-laucher/CMakeLists.txt
CONFLICT (add/add): Merge conflict in 4-Ch2-course-laucher/4-problems/CMakeLists.txt
Auto-merging 4-Ch2-course-laucher/4-problems/CMakeLists.txt
Automatic merge failed; fix conflicts and then commit the result.

Below, we can see that my branch has conflicts with 4-Ch2-course-launcher/CMakeLists.txt - So, to resolve these, we would run vim 4-Ch2-course-launcher/CMakeLists.txt

On branch v0.2
You have unmerged paths.
  (fix conflicts and run "git commit")
  (use "git merge --abort" to abort the merge)

Unmerged paths:
  (use "git add <file>..." to mark resolution)

	both added:      4-Ch2-course-laucher/4-problems/CMakeLists.txt
	both modified:   4-Ch2-course-laucher/CMakeLists.txt

no changes added to commit (use "git add" and/or "git commit -a")

vim <path/to/conflict/file> and you will notice syntax similar to the below has been added to your file -

Some text in a file that has no conflict.

<<<<<<< HEAD
Some text that was changed on the local HEAD.
Some text that was also changed on the remote we are attempting to merge with
>>>>>>> feature-branch

Some more text in a file that has no conflict.

All that Git is asking us to do here is delete the changes that we don't wish to keep, and then git commit -m "Commit message" to complete our merge. If you want to abort the merge, run git status to see how, or just run git merge --abort.

So, in this case if we wish to keep the changes that are on our local HEAD, and overwrite the changes on our feature-branch. Just modify the file, deleting all of the added syntax from the merge conflicts described by git, and any changes that may go with them -

Some text in a file that has no conflict.

Some text that was changed on the local HEAD.

Some more text in a file that has no conflict.

Our merge conflict is resolved. check git status, stage your changes with git add and make the commit to finish the merge.


To stash changes in git, simply run the commands below -

[kapper@kanjaro cmake]$git stash
Saved working directory and index state WIP on master: xxxxxxx Reorganized C problem 4

[kapper@kanjaro cmake]$ git stash list
stash@{0}: WIP on master: xxxxxxx Reorganized C problem 4
[kapper@kanjaro cmake]$ 

This is useful stash your local changes, pull from the remote, resolve any conflicts from your stashed changes and push back to the remote so other contributors can see your work.


Rebasing can be used to move your work to another branch without merging the branches entirely. This is useful if the branch has more development planned or if you would like to keep the branch open for other reasons. This can also be used to modify the history, if you are careful and know what you are doing.

git rebase <remote>


git rebase master
git rebase fix


git rebase --interactive or git rebase -i provides the summary for each commit in the editor it invokes

You may want to use rebase -i <commit hash> to squash your work down to a single commit, or rearrange the work in the commits to make the patch easier for the maintainer to review


To apply a patch generated by format-patch, you use git am (the command is named am as it is used to "apply a series of patches from a mailbox").

If you run a mail client that can save several emails out in mbox format, you can save entire patch series into a file and then use git am to apply them one at a time.

git diff --check, which identifies possible whitespace errors and lists them for you.