Introduction to GitInitiated by Klaatu, this open series introduces Git and the concepts behind its use in a collaborative environment.
Clacke mentioned SparkleShare in episode 2542, and it occurred to me that not everyone knows what Sparkleshare is. So here's a show about it.
To setup SparkleShare, refer to SparkleShare.org. It's available for Linux, Windows, and Mac; great for cross-platform collaboration.
The Linux installer uses FlatPak, so you do need to install that.
Once installed, launch SparkleShare in the usual way. If you have no usual way, you can use this command:$ flatpak run org.sparkleshare.SparkleShare
The first screen asks for your name and email. This doesn't have to be your real name and email, but it is what SparkleShare will use when making commits on your behalf. This name and email will be visible to anyone who can see your online Git repository.
The next screen displays the Sync Remote Project screen. You use this screen any time you want to add another share to your sparkle.
In this episode, I set up two projects: one brand new one using my home server as host, and one that mirrors an existing project on Gitlab.com.
Adding a project from Gitlab
The first thing you must do is give SparkleShare permission to access Gitlab. To do this, click on the SparkleShare icon in your system tray > SparkleShare > Client ID and copy your ID to your clipboard.
Now go to your online Git host and add this "Client ID" to your approved SSH Keys. Where this is located depends on your Git host, but in Gitlab, it's located in the left column of the Settings screen. When your SSH Key has been added, Gitlab displays a key fingerprint (actually just a string of numbers) as confirmation.
The path to your remote Git repository is the part of an URL after the host. It usually starts with your username. For example, if I have a project on Gitlab located at
gitlab.com/notklaatu/foo.gitthen the path that SparkleShare needs is
Click the Add button to add the project to your local SparkleShare folder.
Adding a project hosted on your own server
There are a lot more variables if you're hosting a Git repository on your own server. These are the things that you may need to account for:
Is your local .ssh/config file setting some "invisible" defaults for when you SSH to your server? If so, you may need to modify or add an entry for SparkleShare.
Your SparkleShare auto-generated "Client ID" is located in
Is your SparkleShare SSH key (the "Client ID" in SparkleShare lingo) in your authorized_hosts file?
Does a Git repository exist on your remote server in the location you think it exists?
Use SparkleShare exactly as you would DropBox or the NextCloud Desktop Client: drag-and-drop a file to add it, drag it to the Trash to delete it. All SparkleShare folders sync'd with any given project syncs automatically through the magickalfulness of Git hooks.
This episode is light on actual commands, and mostly a narrative analysis of what git HEAD is and why it matters to you.
Some commands you can try:
$ cat ~/path/to/git/repo/.git/HEAD
$ ls ~/path/to/git/repo/.git/refs $ cat ~/path/to/git/repo/.git/refs/master
These are all the commands covered in this episode. This is not a sequence, it's just all the commands in the episode, listed one after another.
Get changes from the remote repo:
$ git fetch
See all branches:
$ git branch --all
View a remote branch after you have fetched it:
$ git checkout origin/dev
Create a copy of a fetched remote branch in your local repo:
$ git checkout dev
Merge changes from remote origin/master into your local master branch:
$ git merge master origin/master
Fetch and merge automatically:
$ git pull
Create a new branch, and change to it:
$ git checkout -b dev
Merge dev into master:
$ git checkout master $ git merge master dev
Merge master into dev
$ git checkout dev $ git merge dev master
Delete the dev branch:
$ git branch -d dev
Instantiate a git repo:
$ mkdir alice $ cd !$ $ git init
Add a remote:
$ git remote add origin URI_OF_REMOTE
Change a remote:
$ git remote set-url origin NEW_URI
A remote can be a server, it can be a local directory, an NFS share, pretty much whatever you want.
It is a Git convention that the primary remote is called
origin. You don't have to call it that, but it's pretty common.
git add git commit -m "some useful message" git push origin HEAD