Ilya Grigoriev

@crackcomm > The $base wasn't expanded... You are supposed to use `$output` twice, like in https://github.com/martinvonz/jj/blob/1cc271441f716549fc9e87e714aacddd2831a4b4/cli/src/config/merge_tools.toml#L12 It is a pretty easy mistake to make... **Update:** I hope #2667 will help....

Re this PR: If I understand correctly what it does, I was thinking of maybe having such an option and calling it `--parent-inherits` (perhaps `--parent-inherits-change-id`?), perhaps with the alias `-P`....

Thank you! It might make sense to add a test. Hopefully, there are already tests that are easy to adapt. I'm not sure whether this is an absolute necessity, so...

I keep postponing looking at this. I was going to double-check that the code path is the same as the one that's tested, but I'm now thinking that wouldn't be...

I'm assuming you are actively working on this. I thought I'd try encouraging people to mark the issues they're working on, but feel free to un-assign yourself at any time.

Alternatively, there could be a better way to restore from the pre-rebase version of a rebased commit. It shouldn't be hard to add a revset operator for "go back in...

I like the explanation you have in the docs. I'm not sure `--reparent` makes sense to me though. Why that word? The option changes the child, not the parent. So...

Also, I kind of like the word "verbatim". Is it confusing in some way?

I'm not sure I agree, but this makes sense. Another name this made me think of is `--verbatim-children`. I'm not sure whether that's clearer or more confusing. Idea # 3:...

My problem with `reparent` is the same as your problem with `verbatim`: every rebase changes the parent, not just the one with `--reparent`. `--just-reparent` sounds weird. OTOH, I believe the...