git worktree, no more stashing!

I recently discovered git-worktree and it has been a game changer. I do not use these words lightly.

One of the silent pain points of using git for me which was agonizing at times, was when working on code locally in multiple branches simultaneously. Each branch usually has a small new feature under development which wishes to add on top of the main branch, or some branches have code that is specific to certain projects or new releases, which have deviated quite a bit from the main stable branch and live on a world of its own.

When switching between branches, it was always messy. I had to stash code or commit them with half-baked commit messages which don’t fully describe the feature because hey! the feature isn’t fully done yet. Then, after I come back to that working branch, I’ve usually forgotten what I was doing. Can you relate to this? No? You guys love stashing? In that case, you can skip the rest of this post.

Enter git worktree, it allows you to maintain independent copies of the repo, where each copy is a working branch at the same time!

All you need to do to get started is, get rid of your old repo (ah ah ah not so fast!, as much as you’d like to, don’t rm -rf <repo> before you commit and push your local changes to your remote copy). So with a clean slate, let’s start with

git clone --bare <https://github.com/you/your_repo.git> your_repo

Now when you change directory to your repo, you will notice that it doesn’t look like your repo at all. There are all these files and folders like HEAD, config, refs, worktrees, etc. Don’t Panic! This is a bare worktree of your repo, and the files and other stuff you see are what’s needed in the local copy of the repo to enable you to use worktrees and take advantage of the corresponding workflow.

First, lets get our one and only main branch back. Just do

git worktree add main

This now checks out the latest commit on the main branch and creates a folder with the copy of the repo at that commit. You can do a quick ls command and you will notice the folder main with the contents of the repo after the latest commit.

Similarly, you can now add any number of copies of your repo checking out various commits from different branches, and work on them all in the same parent directory!

To list each working tree you can do

git worktree list

You will notice all the different working trees along with the (bare) one which represents the bare repo.

That’s it! We are now ready to take advantage of git worktrees! I can’t believe I didn’t know about git worktree all these years (the feature was available since 2015!).

Sigh, never too late I guess 😒.

P.S. Another cool take away is that since each branch exists in different working folders, depending on your development environment (workflow) and programming language in question, you can save the time needed to reinstall dependencies each time you want to switch between branches, assuming they have a different set of dependencies.

Furthermore, imagine you are executing some program from inside a worktree. Since you would never have to switch branches or stash code, if the program needs the state of the repo to be unchanged while it executes, then the isolation provided by worktrees becomes quite convenient.