Remotes: Sharing Code Online

A Brief Intro to Git

Page Contents

So far, everything we've done has been on your local computer--no one else can see your code or contribute to it. While cloud services like Dropbox and Google Drive are popular for sharing most files within a research team, they are not very well suited for Git repos. [1] Using a service designed specifcally for Git, like GitHub or Bitbucket, is a better solution.

Any of the projects we work on will be hosted on one of the PI's GitHub or Bitbucket accounts, so in the context of our work you won't need to worry about creating a repo online. However, you can follow these instructions if you want to practice that on your own.

1   Downloading (Cloning) a Repository

To copy a repository from the Internet to your computer, use the clone command and the URL of the online repo.

$ git clone https://github.com/dmsul/econtools.git

This will create a folder econtools/ in the current folder that contains all the info about the Git repository hosted at the given URL.

You can also tell Git to name your cloned repo something else:

$ git clone https://github.com/dmsul/econtools.git different-folder-name

This will clone the econtools Git repo into a folder called different-folder-name/ instead of econtools/.

2   Remote branches

After you've cloned a repo, you'll have a copy of it on your hard drive. Suppose that your shortlog looks like this:

$ git shortlog
* 2f40bfd -  (HEAD -> master, origin/master) Merge branch 'new-quote'
* 7d6a299 -  Added a variable to hello.py
* 823459f -  Add hello.py

By default, you have the master branch checked out after a clone. You'll notice that there's another branch label, origin/master. When you clone a repo, Git remembers the website where the repo originated and calls it origin for short. Git also creates special branches in your local copy of the repo that mirror the branches in the original online repo. These are remote branches and follow the naming convention <remote_name>/<branch_name>.

3   Update a remote branch (Fetch and Pull)

What if the online repo gets updated, how do you get the updates to your computer? You use the fetch command:

$ git fetch origin
remote: Enumerating objects: 4, done.
remote: Counting objects: 100% (4/4), done.
remote: Total 5 (delta 4), reused 4 (delta 4), pack-reused 1
Unpacking objects: 100% (5/5), done.
From github.com:dmsul/econtools
   073517e..906b941  master     -> origin/master

$ git shortlog
* 4b93ceq -  (origin/master) A new commit on the website
* 2f40bfd -  (HEAD -> master) Merge branch 'new-quote'
* 7d6a299 -  Added a variable to hello.py
* 823459f -  Add hello.py

Now origin/master is one commit ahead of our local master branch. We can merge the two using git merge origin/master just like we did before. Because this process of fetching and merging is so common, there's a shortcut command, pull, which does a fetch, then a merge.

$ git pull origin master
From github.com:dmsul/econtools
 * branch            master     -> FETCH_HEAD
Updating 073517e..906b941
Fast-forward
 econtools/util/reference.py | 5 +++++
 1 file changed, 5 insertions(+)

$ git shortlog
* 4b93ceq -  (HEAD -> master, origin/master) A new commit on the website
* 2f40bfd -  Merge branch 'new-quote'
* 7d6a299 -  Added a variable to hello.py
* 823459f -  Add hello.py

Now our local code is updated to match the code on the website.

4   Upload your changes to the remote server (Push)

If you add a commit to your local master branch, the remote branch origin/master is unaffected. After adding a commit, your shortlog looks like this:

$ git shortlog
* 4b93ceq -  (HEAD -> master) A new local commit
* 4b93ceq -  (origin/master) A new commit on the website
* 2f40bfd -  Merge branch 'new-quote'
* 7d6a299 -  Added a variable to hello.py
* 823459f -  Add hello.py

Now master is one commit ahead of origin/master. To get our new commit uploaded to origin (e.g., Github) and to update origin/master, we use the push command:

$ git push origin master
Counting objects: 12, done.
Delta compression using up to 8 threads.
Compressing objects: 100% (12/12), done.
Writing objects: 100% (12/12), 2.09 KiB | 1.05 MiB/s, done.
Total 12 (delta 10), reused 0 (delta 0)
remote: Resolving deltas: 100% (10/10), completed with 6 local objects.
To github.com:dmsul/econtools
   de8fb4e..2e49c46  master -> master

$ git shortlog
* 4b93ceq -  (HEAD -> master, origin/master) A new local commit
* 4b93ceq -  A new commit on the website
* 2f40bfd -  Merge branch 'new-quote'
* 7d6a299 -  Added a variable to hello.py
* 823459f -  Add hello.py

5   Checkout a local copy of a remote branch

[In Progress]

$ git checkout -b branch-name origin/branch-name
[1]The problem is that all information about the Git repository is stored in a hidden folder called .git/. During certain operations, Git the software makes lots of changes to files in .git/ the folder very quickly. Dropbox et al. can sometimes have trouble keeping up and will try to change these files at the same time Git is, resulting in corrupted files. Once that happens, you have to scrap your repo and re-clone it.