[tombert]

Fair enough. I feel like I do a fair amount of the more advanced features (interactive add and rebase, bisect, worktrees) without any fancy tooling and I don't have a problem much anymore, but admittedly they did confuse me at first.

[em-bee]

i don't remember confusion. i find it's mostly understanding the data model and in particular the branches and references/reflog. when i am worried i might break something then i tag the the checkout where i am at and i know i can always revert to that. i also compare the original with the new state. i usually know what that diff should look like, and even if the operations in between are confusing, if the diff looks like what i expect then i know it went all right. trust the process but verify the results.

the big thing i am missing from it is a branch history. a record for every commit to which branch it once belonged to. no improved interface can fix that. that would have to be added to the core of git.