[develop7]

> The major deciding point between the two is that Mercurial see history as an immutable truth.

That view is quite outdated[1]. It's true for _published_ history though, but that's it. Couldn't help but note Mercurial did it right, _again_ :)

1: https://www.mercurial-scm.org/doc/evolution/

[hvidgaard]

I was only talking about commits in a publishing repository. What happens on a developers machine, and between developers as draft changesets is quite a different story. We use that a lot and I really like it. Git has the concept of remote removing commits and garbage collection.

[ygra]

I guess you haven't read the link posted. Changeset evolution is not about draft commits that a developer changes locally. It's about actually changing the public, pushed history in a way that nothing breaks for other developers who already based commits on those that get altered.

[develop7]

Um, that's not quite true: https://www.mercurial-scm.org/doc/evolution/sharing.html#pub...

The limits imposed by phase are not going anywhere; you still cannot edit published history.

[hvidgaard]

Publishing repository by definition contains public changesets. You cannot alter history for them without force. Again, what happens on and between developers can and should be trimmed before it's pushed to the publishing repository.