Repositories, Remotes & cloning

A source code directory by itself is nothing. It needs to be maintained by several people, iterated, enhanced, cleaned up regularily in order for the software to be presentable to the user.

In order to transform your source code directory into a git repository you may just run

$ git init

This will create the .git directory on the top level and it will contain configuration for the repository, the repo history and hooks

After this point, we're ready to add a remote repository, which acts as the storage for our source code. GitHub has made it really easy for us to do so, simply by clicking + New Repository. The repository will then be available and there will be an ssh url available to add as a remote. Remotes are aliases for urls where we can push our code to

$ git remote add origin git@github.com:falexandrou/thefotisbook.git

Of course we can name our remotes however we want, but the converntions are origin and upstream

We can add our files now, simply by running

$ git add --all

This will add all files under the current directory and this command can also be ran under a certain directory to include all files under that directory. Removing files can be accomplished with git rm and we can add or remove separate files by specifying the file's path.

If we wish to start over, we can simply

$ git reset

or if we need to temporarily remove our changes we can run

$ git stash

and then to bring them back we may

$ git stash pop

And there is also a git rm command in order to remove a file under a given path

We're now able to commit, to create a point in history where our code exists

$ git commit -m "This is my first commit"

Option -m means message and you may find many options available for commiting as well

We may now push our code to the remote repository, simply by running

$ git push origin master

This means that we pushed the master branch, default branch created to the origin remote, so our remote storage now contains our latest changes.

The commit is now represented by a unique SHA1 hash referenced as commit hash or commit id

Written on March 5, 2015