In this Git tutorial, we'll set up a GitHub account and learn how to use commands like
git add and
git commit to save our coding progress at various checkpoints. Git repositories are useful for keeping a history of your code, rolling back, collaborating with other developers, and working offline. Unlike SVN, Git is decentralized, meaning each of your developers' local machines contain the entire repository so they only need to connect to the remote repository when synching up, releasing and deploying.
Read on to learn more.
- A 32-bit or 64-bit computer with at least 8GB of RAM. 16GB or more is recommended.
- At least 50GB of free disk space.
Sign up for Github at http://github.com.
Add your SSH key to GitHub by following this tutorial.
Navigate to your home directory, by typing the following command in your shell/terminal.
If it doesn't already exist, create a
repos folder to store all of your code repositories.
Navigate into the
Create a folder for your new repository. We use all lowercase and dashes in place of spaces to match GitHub's repo naming convention.
You should now have the following folder structure.
▾ home/ ▾ repos/ ▸ my-project/
Create A New Code Repository In Github
Follow the instructions on GitHub's documentation. The following git intialization steps are also provided in the documentation, but I'll list them out here, anyway.
Initialize a new repository and link to the remote repo..
mkdir my-project cd my-project echo "# my-project" >> README.md git init git add README.md git commit -m "first commit" git remote add origin firstname.lastname@example.org:organization/my-project.git git push -u origin master
After initializing the repository, we'll be positioned at the master trunk. Think of your code history as a tree. master is the main trunk and we'll want to create branches off of that trunk. So before we move forward, just visualize a tree's main trunk lying on its side with "master" written on it, like so:
Now, let's create a branch off of master and name it "develop", by typing the following command in our shell/terminal:
git checkout -b develop
Now, picture a large branch growing out of the master with "develop" written on it.
.====develop / master
Our master branch is now safe. Anything we do to our code right now will modify develop. If, for some reason, we want to roll back or delete our branch altogether, we're free to do so.
However, we don't want to make any edits to our code while we're standing on the develop branch. The master branch and the develop branch each have their own purpose.
The Master Branch's Purpose
The master branch is reserved only for packaging a release of our software. For example, when our software passes testing and we're ready to launch, we'll create a v1.0.0 "release" version. We're not ready to do that yet, though, which is why we're not on the master branch.
The Develop Branch's Purpose
The develop branch is reserved for creating feature branches. So yes, branches can have their own branches, just like a normal tree. For now, just follow along. It will be easier to understand once we visualize the tree structure.
Type the following command in, which will push the new develop branch to the remote repository. We're essentially making a copy of our progress within GitHub.
git push origin develop
The following commands will create a branch named
rl_feature. Feel free to replace "rl" with your own initials. The naming convention for branches is
git checkout -b rl_feature git push origin rl_feature
Now, visualize this:
.==rl_feature / .====develop / master
We're free to update our code and save our progress on the
rl_feature branch. Other developers may also branch off of develop like so:
.==rl_feature / .====develop / \ master \====tr_feature-2
This allows us to merge all the branches back into develop when we're finished and stitch all of the code together like so:
.==rl_feature==. .==rl_feature-3 / \ / .====develop--------------*----*==/ / \ / master \====tr_feature-2==/
This process is a constant loop of branching off, making edits, merging back in so that develop is updated with everyone's features.
If we were to zoom in to one of these feature branches, we would see the history of the developer's progress in the form of commits. In the following example, we're going to create a new file and update it, making commits along the way.
touch test.txt git add . git commit -m "Created a text file" echo "My code" >> test.txt git commit -m "Updated the text file" git push origin rl_feature
commit commit | | .==rl_feature==*=======* / .====develop / master
Think of commits as checkpoints we can roll back to. You want to make meaningful commits when you hit certain milestones or just any time you feel like experimenting with code you'll want to undo.
These are the commands for merging our code into the develop branch so that our feature can be shared with the rest of the team.
git checkout develop git merge rl_feature git push origin develop
commit commit | | .==rl_feature==*=======*-. / \ .====develop----------------------merge / master
commit commit | | .==rl_feature==*=======*-. / \ .====develop----------------------merge / \ master-----------------------------------merge---.--- \ v1.0.0 release
Once we've merged our feature branch into develop, after a considerable amount of time has passed, we go through and clean up old branches with the following commands.
First, we make sure we're on develop or master. We can't merge or delete a branch we're currently on.```bash git checkout develop ```
Then we delete the branch from both the remote and the local repository.```bash git push -d origin rl_feature git branch -d rl_feature ```
Issue the following command. Replace
initials/feature-branch with the branch name.
git pull origin initials/feature-branch
You'll do this more often on develop because other developers will be merging their branches into develop and you may want to update it before you merge in.
git pull origin develop
# Display any edits / if the branch is up to date git status # View which branch you're on git branch # View git history git log # View git history with branches git log --graph --full-history --all --color \ --pretty=format:"%x1b[31m%h%x09%x1b[32m%d%x1b[0m%x20%s"]]]"
That concludes this Git tutorial. In part two of this series, we'll go over advanced git commands, and the real-world scenarios they address. For now, we should be fine with these commands as they satisfy the most common use cases.
- Learning Source Control with Git and SourceTree: A Hands-On Guide to Source Control for coders and non-coders
- Version Control with Git: Powerful tools and techniques for collaborative software development
- Learn Version Control with Git: A step-by-step course for the complete beginner
- Version Control with Subversion: Next Generation Open Source Version Control
- TortoiseSVN 1.7 Beginners Guide