Hacker News Reader

We have database-backed storage backends for Jujutsu at Google. There's no Git involved there. So I think it's wrong to say that it's a wrapper for Git.

Even when using the Git backend, it's only commits and lower-level objects that are stored in the Git repo, and even some parts of the commits are stored outside Git (e g. the change ID). The operation log (which powers the undo feature) is also not stored in Git.

View on HN

Ok that's pretty awesome. But by git backed I meant the data structure. Darcs & pijul shows that we can do better than a commit tree.

Do they? They're different and theoretically better in some respects, but are they better in ways that matter and truly deliver value to users?

Commutativity of patches is neat, but it's just not that big a deal. I think Pijul's best feature is first-class conflicts, an idea that Jujutsu has also incorporated.

It matters a lot when maintaining long lived forks of active upstream projects, especially with invasive changes that get tripped up on rebases.

That seems like an algorithm issue more than a fundamental data structure issue, right?

Patch commutivity would make rebasing vs. merge commits effectively the same thing. That’s where it impacts the data structure.

From the perspective of lawyers who come knocking, it seems incorrect to treat rebases and merges as the same.

Reply on news.ycombinator.com