Working with remotes

There are several repository management tools that allow to take the git workflow and ease collaboration based on the cloud. These are hosting platform for git repositories and their use is supported by native git commands. The most popular are:

• Github
• Gitlab
• Bitbucket

Basic Worflow

When working with remotes there is 3 general operations performed:

1. Setup
2. Get data
3. Post data

When creating a new project or linking a local repo to the cloud for the first time, an empty remote should be created, after there is two options:

• If a local repo with data exist:

  1. Add the remote
  2. Push the data

• If its a new project

  1. Clone the empty remote
  2. Do changes
  3. Push the data

Setup

Any git repository can be configured with a remote. In order to get a list of the remotes of a repository,the following command is used

git remote -v

One (or more) can be added to the repository, renamed or deleted

Adding a remote

To add a remote the command is:

git remote add <name> <url>

Where:

• <name> is the name to give the remote. In most cases, when adding the first remote this is called origin but is just a convention
• <url> is the HTTPS or SSH url of the repository in Github, gitlab, etc.

Renaming a remote

The name of a remote can also be changed after creation with

git remote rename <old> <new>

Deleting a remote

To delete a remote, the command is:

git remote remove <name>

Getting Data

Cloning a remote

To “pull” a remote for the first time to a machine that does not have it, the clone command is used.

git clone <url>

The command :

• Copies the data from the remote
• Initializes a local git repo.
• Sets up remote tracking branch references pointing to the branches of the remote

Other branches

When cloning, only the default branch (master or main) is accessible in the local repository. This is because the cloning process, although it creates RTB references for all remote branches, just “links” the one of the default branch to a local branch.

In order to access other remote branches locally is enough to use

git switch <branch_name>

where <branch_name> is the same name as one of the remote branches. This will automatically:

• Create a new local branch called <branch_name>
• Link it with the RTB references for the remote branch

Fetching

git fetch allows to retrieve the latest changes from a remote repository without merging them into the local branch.

The command can be used as git fetch <remote> to fetch all available changes (all branches) or git fetch <remote> <branch> to fetch a specific branch.

In more detail, it only updates the RTB references but not the local branches. So if running git fetch origin master, the origin/master reference will now point to the latest version, but the local master branch won’t be affected.

Pulling

git pull is similar to git fetch, but it also merges the changes from the remote repository into the local branch. This means that your local branch is automatically updated with the latest changes from the remote repository.

The command can be used like git pull <remote> <branch> to pull changes from a specific branch. In addition, git pull will default the remote to origin and the branch to the RTB reference linked to the current branch.

In other words, it updates both the RTB reference (e.g. origin/master) and the current HEAD (e.g master)

Because of this, is critical to run the command in the correct place, as the current branch when the command is run will be the one updated.

Pushing data

In order to push or send local changes (commits) to the remote, the command push is used

git push <remote> <branch>

This command wil create a new branch on the remote (the first time) and push the changes on the local branch with the same name. For example git push origin master will push the changes from the local master branch to the remote master branch

To push to a branch with a different name, the whole syntax can be used:

git push <remote> <local_branch>:<remote_branch>