This assignment is designed to help you create a new Git repository. The text covers both GitHub and Bitbucket. If possible, you should prefer GitHub since it is the central repository for code related to web development.
Since my students turn in assignments with Git, in some, but not all, cases it is important that their repositories be private. Free private repositories are available on both GitHub and BitBucket. GitHub has additional benefits if you sign up for the Student Pack. The student pack provides $75 - $100 in free AWS Educate credits!
NOTE: Since this document covers both GitHub and Bitbucket, you should proceed with caution. Take care to distinguish between the instructions for the two sites. Note that Git itself is an open source project entirely separate from either GitHub or BitBucket. Git is Git no matter where it is. Nevertheless, the websites for Bitbucket and GitHub have differences, so proceed with caution.
I assume the user is running on Linux, but it should be possible to follow along if you are using Windows. Especially helpful for Windows users is the Linux subsystem for Windows. Again, you should prefer Linux as it, and the macOS, are the primary platforms for Web development. Windows is a good platform for Web development, but the leaders in Web development tend to use macOS and Linux as their development platforms.
In short form, the steps to set up your GitHub account are:
All of these steps are detailed in this document.
Git is probably already installed on your system. To confirm that it is installed, type git. If you get an error, then the likely solution is to install Git. Do it like this on Lubuntu:
sudo apt-get install git
Whether Git was pre-installed on your system or not, there are a few steps you need to take to set it up. It is perhaps simplest to get these chores out of the way at the very start. Your goals will be to tell Git your:
Issue the following commands, editing as appropriate:
git config --global user.name "charlie at school"
git config --global user.email "firstname.lastname@example.org"
git config --global push.default simple
These commands are stored in your Git config file, usually found here ~/.gitconfig. You can see the contents of the file by issuing this command:
NOTE: Remember that the tilda (~) is a shorthand for your HOME directory. I should add that I'm oversimplifying a bit here. Learn more about the config files here.
co = checkout
br = branch
ci = commit
st = status
last = log -1 HEAD
it = log --graph --pretty=format:'%Cred%h%Creset -%C(yellow)%d%Creset %s %Cgreen(%cr) %C(bold blue)<%an>%Creset' --abbrev-commit --date=relative
gds = for-each-ref --sort=taggerdate --format '%(tag) %(subject)'
glog = log --date=short --pretty=format:"%h%x09%x09%ad%x09%s"
glogn = log --pretty=format:"%h%x09%an%x09%ad%x09%s"
glogc = log --pretty=format:"%h %Cblue%ad%Creset %ae %Cgreen%s%Creset"
name = Charlie on Mint Home
default = simple
tool = meld
NOTE Since I am usually not sharing my repository with a large number of users, I tend to use the name to identify the location from which I am working. For instance charlie at home, charlie at school, charlie on his laptop etc. This is more than a bit quirky, and might not be at all appropriate in a corporate setting, but I nevertheless find it occasionally useful to know where I was when I did some work.
In either case, you should sign up for the Github student pack:
GitHub is not the only place to host your repositories. Other well known sites include BitBucket and GitLab. Of these two competitors to GitHub, GitLab is currently the more popular. I use BitBucket quite frequently but have not yet tried to GitLab. (It is impossible to keep up with all the cool tech being created.) However, you should note that GitHub is one of the central sites on the web for developers, and it is almost impossible to overemphasize its importance. As a result it will be our goto tool for Git. That said, the competitors bring some nice features to the table in an attempt to differentiate themselves.
NOTE: I much prefer that you use GitHub rather than BitBucket. The reason for this is speed. GitHub has spent a lot of time and money on performance. If I have to pull 30 student repositories at one time, even small differences in performance make a difference. As a result, I ask that you use GitHub to help make your teacher's life easier.
In GitHub or BitBucket, from the home page, create a new repository:
Name your Git repository like this, where isit320 should be name of your class:
For instance, your repository might have a name like one of the following, depending on the class you are in and the current year:
NOTE: Throughout this and similar documents, the year field should be set to the current year. I may have written or last updated this document one or more years ago, but you should use your common sense when using the year field. If it is 2019, then set the year to 2019, even if my example uses some other year.
Then make the following selections:
Add these near the top of your .gitignore file:
In the above, the code that begins with a hash tag (#) are comments. The other code specifies that names of files or directories you do not want to check in to GitHub.
NOTE: On GitHub, it is possible to create a repository without any content such as a pre-initialized README.md or .gitignore file. If you do that, then the act of setting up your repository on your local machine is a bit more complex for newbies. If you select the options to put a README or other content in your repository, then you need do nothing more than clone the repository you created on GitHub site. In other words, much of what follows is unnecessary, particularly the bits about git remote add origin....
Be sure that node_modules is included in your .gitignore directory. Please understand that node_modules directories can be VERY large. Frequently they are several orders of magnitude larger than your base project. This is okay at first, but after about Week Three, repositories that don't use .gitignore to filter out node_modules become huge, taking a long time to download and taking up vast amounts of space on my hard drive. There are further complications due to subtle differences in the way these directories might look on my system and yours.
The bottom line is that I get quite frustrated by the extra work, time, and resources I expend when people forget to include node_modules in their gitignore file. Don't let this happen!
Make sure you give me at least read/write access to your repository. If you use:
Look for settings below and to the left of the word Unwatch:
Select Collaborate and type in my name, which is charliecalvert:
Let's now turn our attention from the browser to the Linux command line. Typically, this means that you will open or navigate to a terminal window in your Pristine Lubuntu instance. However, you can also perform similar work in Windows, the Linux subsystem for Windows or the Mac.
I like to keep all my repositories in a directory called Git. I first navigate to my home directory by issuing the cd command with no parameters:
I can then type ls to check if my Git repository already exists. If it does not, I create it:
If the command returns immediately with no errors then that means the directory was created successfully. If the directory already exists, then you will get an error similar to the following:
$ mkdir ~/Git
mkdir: cannot create directory ‘/home/charlie/Git’: File exists
In either case, you are good to go. The key point was to be sure you have a Git folder. If one exists, then all is well.
Once you have confirmed the presence of your ~/Git folder, navigate into it like this:
Assuming you already have a repository on GitHub or BitBucket that contains some content, you can get a local copy of it by cloning it:
git clone email@example.com:USERNAME/REPO_NAME.git git clone firstname.lastname@example.org:USERNAME/REPO_NAME.git
As explained in the previous section, I usually call git clone from my ~/Git folder. The act of cloning a repository typically looks a bit like this:
cd ~/Git git clone email@example.com:ccalvert/deleteme06.git Cloning into 'deleteme06'... remote: Counting objects: 3, done. remote: Total 3 (delta 0), reused 0 (delta 0) Receiving objects: 100% (3/3), done. Checking connectivity... done.
After cloning your repository, make sure the following are in your .gitignore file:
# IDE Files
# Package files
You can and should have more than just these items, but be sure at least this much is present.
You can read about Git URLs on Elvenware:
Make sure the Secure Shell (SSH) is installed. For instance, try issuing the ssh command and passing it -V as a parameter:
$ ssh -V
OpenSSH_7.4p1 Ubuntu-10, OpenSSL 1.0.2g 1 Mar 2016
As shown above, you should get information on the current version of ssh.
Pristine Lubuntu should have it installed, but just in case, or if you need to install it yourself on a new system, here is the command:
sudo apt-get install ssh
It won't hurt to run the command if SSH is already installed.
At the Linux command prompt Issue these commands, where the first command takes you to your home directory:
ssh-keygen -t rsa -P '' -f ~/.ssh/id_rsa
cat ~/.ssh/id_rsa.pub >> ~/.ssh/authorized_keys
Now copy your public key into the BitBucket SSH page. The first step will be to copy the key to your clipboard. Begin by navigating into the hidden .ssh directory created in the previous step.
We now need to block copy the key to the clipboard. It might be enough to simply echo the key to your bash shell, and then copy it with the mouse or keyboard:
An alternative might be to open the key in geany or some other text editor, and then select it in the editor with Ctrl-A and copy it with Ctrl-C. In any case, be sure you have the key saved to your clipboard.
Then in BitBucket:
Right now your private key is called id_rsa and your public key is called id_rsa.pub. You might consider renaming them to something more meaningful:
mv ~/.ssh/id_rsa ~/.ssh/Prog270-Fall-2016
mv ~/.ssh/id_rsa.pub ~/.ssh/Prog270-Fall-2016.pub
At this point, you should be able to load your SSH private key into memory:
If this works as expected, you should see the words Identity added in the response:
Identity added: /home/charliecalvert/.ssh/Prog270-Fall-2016 (rsa w/o comment)
At this point, you are all set to begin pushing and pulling data from your repository, as described below.
NOTE: I should have already added this code for you, but if you have trouble with ssh-add, make sure this code that loads ssh-agent is found near the bottom of your /home/username/.bashrc file:
if [ -z "$SSH_AUTH_SOCK" ] ; then
You will need to load your SSH key with ssh-add each time your start a new bash (command line) session. Or at least you will need to do so each time you reboot lubuntu.
When you want to connect to a remote machine using SSH, you have to specify which private key to use, or at least the location of your private keys. If you get a permission denied message, this is likely because:
In most cases, it is the third of these problems that causes students trouble. A simple way to make a private key accessible is to use ssh-add to register it with a program called ssh-agent. Then, when you try to access a remote site, all the keys registered with the ssh-agent are tried, and if one of them matches, you are golden.
If you don't want to perform this task manually, you could add the following code at the bottom of your ~/.bashrc file:
To add the line, open your ~/.bashrc file in either nano or geany. For instance:
Adding this line is guarrenteed to load your private key for you because your .bashrc is processed each time you open the bash shell (command line). If you ever want to manually load your .bashrc file, you can execute this code:
To learn more, search for the Linux source command with Google.
Alternatively, I have created an alias in our .bash_aliases file called sshadd. (Not ssh-add but sshadd.) It depends on a symbolic link that you must create in your ~/.ssh directory. Create the link like this:
ln -s Prog270-Fall-2016 main-key
Where Isit320-Fall-2016 is the name of the private key you want to load when you type sshadd. This second approach means you can easily load the key at any time. This is perhaps better than loading it automatically each time you start a new bash shell instance. In particular, if you load it in your bashrc, then you might end up with multiple instances of it running. If you load it by hand sshadd, you will likely end up with only one instance.
Zip up both the private and public key and upload it to Google Drive. That way you can download it at home, and use the private key at both home and school. You don't have to do it that way, but it allows you to skip the step of creating a new key at home, and uploading that public key to BitBucket. Either system works, and there are arguments in favor of each system, but I want to at least suggest that you do things this way.
Suppose you created main-key with a command like this:
ln -s MyKey main-key
When you type ls -la you see main-key in red, which means it is not value. After some thought, you realize that you made a mistake creating the key. Perhaps the name of your key was Prog270Key and not MyKey. Before trying again to create a new version of main-key, first delete it:
Now try again:
ln -s Prog270Key main-key
If you've got everything set up right, and don't have a password on your private key, you could also consider uncommenting these lines at the bottom .bashrc.
# Edit and uncomment to load a key automatically.
ssh-add -l | grep -q "main-key"
if [ $? -eq 0 ]; then
echo main-key already loaded
echo Loading main key.
This will automatically ensure that main-key is loaded everytime you open a bash shell. More specifically, it will first check if the key is loaded, and if it is not, it will load it.
If you are struggling to load your key, try this, it's easier. But please remember how you did it.
Before pushing your repository to the cloud, you should be sure you have a .gitignore file. This file should have been created when you built your Repository on GitHub, but if you missed that step, then read on.
Your .gitignore files tells Git which files to ignore, which files should not be pushed to the cloud. This may seem like an unimportant step, but for a developer to try to use Git without setting up .gitignore file would be a bit like trying to drive a car that had no break. The inevitable result: a 95% chance of rain accompanied by high winds and thunderstorms.
If it does not alread exist, in the root of your repository, create a .gitignore file with the following items in it:
You can create the file with the nano editor, or do it like this from the Bash shell:
echo node_modules >> .gitignore
echo bower_components >> .gitignore
echo coverage >> .gitignore
echo .metadata >> .gitignore
echo .idea >> .gitignore
echo .c9 >> .gitignore
echo .vscode >> .gitignore
echo .expo >> .gitignore
echo platforms >> .gitignore
echo Thumbs.db >> .gitignore
echo *.zip >> .gitignore
echo npm-debug.log >> .gitignore
echo bundle.js >> .gitignore
echo *.js.map >> .gitignore
Files that begin with a period are hidden. To see them, type ls -la or ls -a or ll
NOTE: You can learn more about nano by searching for cheat sheets.
Now add, commit and push your content:
git add .
git commit -m "Initial commit"
git push -u origin master
When we type Git add followed by a period we are asking Git to add all the changes we have made to the repository. If you want to add just one file, issue a command like this:
git add README.md
In all cases, after doing an add, you should check your work to make sure all is going well. If, for instance, you see that you have accidentally added in a node_modules directory, then you should cancel the add, adjust your .gitignore file, and try again.
You only need to use git push -u origin master the first time you commit to your repository. After that, you can simply type git push.
When upi type git push -u origin master then:
You can learn more about Git on Elvenware.
If you have problems, check these items:
Also, you can learn something about the keys that are currently loaded with this command: ssh-add -l. That's a small l for list. More details are here and here. Also, try capital L to see public keys.
This video shows how to proceed:
Other related videos are here:
Here is a deck with some information that you might find useful:
Please submit the URL of your repository in the Text Entry section of Canvas when you post this assignment. The URL you submit should, in a general way, follow this format:
git clone firstname.lastname@example.org:lastname/prog272_lastname.git
You can get the exact phrase to submit by selecting the Actions | Clone menu item at the top left of the BitBucket site.
Check to make sure you:
If you need to rename your repository, follow the instructions outlined later in this document.
Paste in the the URLs for your repository when you submit your assignment. There are two URLs that can be interesting:
Examples, with the SSH style URL listed first:
Most of the time, when I ask for your Git URL, I want the SSH URL. It should be clear, however, that either URL will allow me to access your repository.
Whenever you need to send me information when submitting an assignment, you can either enter a comment or use the Text page. When possible, use the Text page, because you can insert live hyperlinks and relatively well-formatted text using that option. However, either technique will work.
Most of us have a copy of our repository:
Nearly every day we:
We do that with these commands:
To push we typically:
git add .
git commit -m "Some message"
If you pull without pushing first, eventually you will hit merge conflicts. You will have one version of your code at home, another version in the cloud, and yet another version at school. This is bad. To avoid Merge conflicts:
You can identify a repository by looking for the hidden directory named .git. If that file exists, then you are in the root of a Git repository.
In this listing, you can tell that the file called README.md is in a Git repository because a long (l) listing (ls) of the directory with the hidden flag (a) reveals the existence of the .git folder:
$ ls -la
drwxrwxr-x 3 charlie charlie 4096 Sep 24 10:55 .
drwxrwxr-x 14 charlie charlie 4096 Sep 24 10:55 ..
drwxrwxr-x 7 charlie charlie 4096 Sep 24 10:55 .git
-rw-rw-r-- 1 charlie charlie 4 Sep 24 10:55 README.md
Don't create one repository inside another unless you are sure you know what you are doing. It is possible to nest repos, but it is special case and requires specific knowledge on how to proceed.
Simplify your life with SSH Config files.
Sample code from the BitBucket when you first create a repository. I'm including this just because it.
Set up local directory:
git remote add origin email@example.com:ccalvert/isit320-deleteme-2015.git
Create your first file, commit, and push
echo "Charlie Calvert" >> contributors.txt
git add contributors.txt
git commit -m 'Initial commit with contributors'
git push -u origin master
Here is the sample code suggested by GitHub.
Create a new repository in an existing directory from the command line:
echo "# deleteme06" >> README.md
git add README.md
git commit -m "first commit"
git remote add origin firstname.lastname@example.org:charliecalvert/deleteme06.git
git push -u origin master
Push an existing repository for the first time to a new repository:
git remote add origin email@example.com:charliecalvert/deleteme06.git
git push -u origin master
It is possible to spoof a URL or even an IP address. As a result, when you push and pull from BitBucket or GitHub, Git checks to see that the machine you are connecting to is the correct machine. In particular, it looks in this file for the fingerprint of your remote machine:
If the machine you are going to connect to is not listed in known-hosts, then the connection will be aborted.
When you first push to BitBucket or any new SSH server, you should be prompted to confirm the authenticity of the site:
git push -u origin --all
The authenticity of host 'bitbucket.org (188.8.131.52)' can't be established.
RSA key fingerprint is 97:8c:1b:f2:6f:14:6b:5c:3b:ec:aa:46:46:74:7c:40.
Are you sure you want to continue connecting (yes/no)? yes
Warning: Permanently added 'bitbucket.org' (RSA) to the list of known hosts.
When you type in yes at the prompt, the fingerprint for this server is stored in your known_hosts file. You can read more about this process on Elvenware.
If you ever need to remove a line from a known_hosts file, see this:
If you get the error "Not a git repository", then it is likely that one of two things has happened:
If you try to clone your repository, and you get a public key error then it is likely that one of the two things has happened:
Remember that when you create a repository for the first time, you need to share it with me. To share your repository, you need to know my user name. On
Though it may confuse matters somewhat, I should perhaps add that it is not wrong to pull down an existing repository onto your machine for the first time by first typing git init. It is just that I think it is simpler to pull it down by cloning it. Also, we have to be sure never to type git init outside our repository, and particular, by typing it in our home directory or in the ~/Git directory.
You can use git clone only on existing repositories. You can't use git clone if the repository does not already exist somewhere. Even if the repository is completely empty, you can still use git clone. There is, of course, a difference between an non-existent repository and an empty repository.
You can use git init to help you fetch an existing repository, or to create an entirely new repository that does not exist yet.
Once again, you can clone an empty repository, you just can't clone a non-existent repository.
To put the matter somewhat differently: git clone makes a copy of an existing repository, but gives you little control of the process. In most cases, you don't need that control, so use git clone. However, if you want more control, or more in depth understanding of how Git works, then use git init.
The following are two ways of doing the same thing. We could use either one when pulling down an existing repository with some content in it.
Use git init plus git remote plus git pull:
mkdir DeleteMe05 cd DeleteMe05 git init git remote add origin firstname.lastname@example.org:ccalvert/DeleteMe05.git git pull origin master
Or use clone:
git clone email@example.com:ccalvert/DeleteMe05.git
If we have an existing repository in the cloud, one that contains content, even if the only content is a README, then we usually don't want to use git init. Instead, we just use git clone and pull down the whole repository. We do this, in our case, from the ~/Git directory. In other words, we almost never do it from inside an existing repository, or from inside a folder that we want to make the root of a new repository. You can use either technique to pull down an existing repository, but git clone is less likely to lead to errors because it has fewer steps and less typing.
Here are the instructions that BitBucket gives us for creating a new repository:
mkdir /path/to/your/project cd /path/to/your/project git init git remote add origin firstname.lastname@example.org:ccalvert/deleteme06.git echo "Charlie Calvert" >> contributors.txt git add contributors.txt git commit -m 'Initial commit with contributors' git push -u origin master
Note that these instructions say git push while the first example I give above says git pull. Take a moment to look and see the difference. It's not obvious.
The git push shown in the BitBucket instructions only makes sense because we created an empty repository in BitBucket and we want to push new content that we created locally into it. If we have an existing repository with content, then use git pull.
One thing in favor of the code shown in my first example and in the BitBucket example is that it shows exactly what is happening, while Git clone does a lot of work behind the scenes. Still, git clone is a lot simpler to use... Perhaps BitBucket and GitHub use the somewhat more complex code simply to raise the bar a bit, and keep out complete neophytes?
The following information is for advanced users only. You do not need this if you are just getting started.
In your hidden .git folder, in a file called config, Git stores the URL for your repository. If you want to find that URL without opening up the config file, issue this command from inside your repository:
git remote -v
This should work for any repository on your system. For instance, try the above command in JsObjects.
If you want to change or set the URL you are using on your local machine, try something like this:
git remote set-url origin email@example.com:my-user-name/my-repo.git
If you are uncertain about the URL for your repository, you should able to easily find it on either GitHub or BitBucket.
Don't over complicate this subject. In the past, I have talked students through some fairly complex things that can be done when creating a new repository. But over time I decided it doesn't have to be complicated. Instead:
Don't use any commands other than clone (once) and then:
You shouldn't need any other commands. At least not at first. Don't try to set the origin or anything like that. Just clone your new repository one time. Then add, commit and push your code as needed.
After you start feeling very comfortable with the basics, then you can learn some more complex commands.
Details are here
Some students will not correctly name the repository the first time around. That is, they did not:
If you have not correctly named your repository, go to GitHub and locate your repository. Click on the Settings icon near the top right of the page.
You will be taken to the Settings page. At the top of the Settings page will be an option to rename your repository. Select this option and rename your repository. This will change the URL for your repository.
Save your work. Copy the new URL for your repository. Resubmit your assignment, being sure to include the new name, that is, the new URL for your repository.
Both at home, and at school, you want to have a ~/.ssh directory. In that directory, you should have a private key and a public key. The private key is only on your machines. Never give it to anyone else. The public key you insert into the proper dialog on GitHub (or BitBucket) or some other site as needed.
You have two choices:
Thanks to Emily Yu for inspiring this handy step by step guide to installing the key on your home machine:
The code for creating a private/public key pair can be seen here, and elsewhere in my notes:
Be sure to look at this deck:
And this document/assignment: http://www.ccalvert.net/books/CloudNotes/Assignments/GitNewRepo.html
It would definitely be a mistake to put a private key in a public Git repository. A private repository is probably okay to use, but still not recommended as eventually you may want to share your repository with others. Putting your public keys on Git is probably okay, but also probably not very useful. The best plan is to either create two or more pairs, or to create one pair and upload to Google Drive so you can download it as needed at home or at school.
If you want, it is possible to put a password on your private key.