Git Without GitHub
Cloud Optional
Everything should be in git, but not everything needs to be in GitHub, or any other cloud repo, for that matter. This page covers the basics of using git on your personal machine in its capacity as an infinite undo. When we talk about git as a distributed version control system (DVCS), we’re going to turn the D into distributed across several personal devices so you can have some of the confidence of using a remote or offside backup location, but not.
Context & Purpose
There’s so much I’d like to cover directly in-context here, but this is just one little piece of what I’m currently layering up in my book about future-proofing yourself and your skills in the world of high technology. Things move so fast it seems like no sooner you learned something important that it’s obsolete. Effectively challenging that kind of tech churn is my mission, and this is just the page I extracted because I want it really easy to find in the short-term (not buried in the blog/book) due to its immediacy and forever-relevance.
Why Local Git Can Be Better
If you’re not using it for collaboration, then you don’t need GitHub. This is particularly true on a per-repository basis. Maybe you should be on GitHub for collaborating here and there, making edits and pull requests on public projects and engage in the free and open source software community (FOSS), but if you just need to make some file safe and available to you all the time, it might be better to employ a 1, 2, 3 backup procedure which are the same basic best practices as protecting any of your data.
Setting Up Local Git Repositories
Let’s start with making a git repository that you can pull from, push to and clone from right on your local machine, without even running any server software. If you are pulling from and pushing to your same local machine as you’re working on, you can get the files into at least two places, either on the same physical device (more risky but fewer moving parts), or onto 2 physically different devices (safer but can be flaky if one is removable media). I personally recommend two places on the same drive so that your pushes and pulls always work, and then we’ll use rsync or plain old file copying to intermittently synchronize everything across separate (often removable) devices.
The Linux Advantage
So let’s say you’re on a Linux system. Yeah, I’m going to talk Linux because that’s now the Universal Operating System! If you’re on Mac, it’s close enough. And if you’re on Windows, there’s Windows Subsystem for Linux (WSL). Sooner or later, if not this year then next or the year after that, you’ll have to face it. Linux won. It won in the professional server space. And it just doesn’t make sense for the proprietary stuff to stay-won in the personal space, except as a consumer device capable of hosting your plug-in Linux environment for development work. And I’m not talking virtual machines or even containers. I’m talking nix and guix. More on that later.
Initial Setup Steps
For now, make sure you have a directory in you ~/ home folder. I use
git-repos, but it could be anything. Just go ahead and make a directory (or
folder, if you prefer) there:
cd ~/
mkdir git-repos
And now you have where you push and pull your git repos from, exactly as if you
were using GitHub. You just need to sprinkle in the magic fairy dust that makes
it all work. Yeah, so there’s a few steps. But there is on GitHub too. So 6 of
one and a half-dozen of the other. Make your [repo].git folder…
mkdir ~/git-repos/your-repo.git
cd ~/git-repos/your-repo.git
git init --bare
Creating Your Working Repository
There! Now you have a “target” or the remote origin location that all your git pushes and pulls work from. You’re essentially replacing everything git does over the network using SSH or HTTPS protocols with your local filesystem. It’s remarkably easy and was a big awakening when I discovered it. Be your own GitHub! This at least gets your git rep, which already has its inherent infinite undo version control properties, into at least a second location, albeit on the same physical storage device. It’s not the end of the story. We’re going to work it up in layers. This is the first layer.
Initializing Your Working Directory
At this point if you haven’t already turned your working directory into a git repo folder, do that:
cd ~/repos/your-repo
git init
touch foo.txt
git add foo.txt
git commit -am "Git Kung Fu"
Connecting to Your Local Remote
And now you’re caught up with how much you can do without a remote origin. You have a local origin and you want to make that other location on the same drive your remote origin. So you simply:
git remote add origin ~/git-repos/your-repo.git
git remote set-url origin ~/git-repos/your-repo.git
You can now check that your remote location is set correctly:
git remote -v
First Push and Regular Usage
Now you can do your first push, but you have to set the location:
git push --set-upstream origin master
And now you can commit and push as usual:
git touch foo.txt
git commit -am "More git Kung Fu"
git push
Moving Repos Off Of GitHub
Okay, so that covers how to create a git repo locally from scratch and then back it up to another location on the same storage device. But let’s say you’ve got a repo already on GitHub and you want to change it’s upstream location from GitHub to local, and then delete it from GitHub. This is part of the declouding process, which I think will rise in popularity as home equipment becomes easier to maintain and people want to get the most out of the equipment they already own before subscribing to or becoming dependent upon yet another cloud service, even if it is GitHub. There’s a reason Microsoft bought them. The want… no, they need you dependent on them.
First, check current remote
cd ~/repos/exiting-repo git remote -v
Remove the GitHub remote
git remote remove origin
Add your new local remote
git remote add origin ~/git-repos/existing-repo.git
Check branch name
git branch
Push your existing repository to the new remote
git push --set-upstream origin master # or 'main' depending on your branch name