[PATCH V2] rebase: move bookmark to destination for commits becoming empty (issue5627)

[Martin von Zweigbergk]

On Jul 24, 2017 6:03 PM, "Jun Wu" <quark at fb.com> wrote:

Per discussion with Adam, it's breaking user expectation that bookmarks are
used to identify feature branches (not to identify a single commit), and is
not that rare with rebase.rebaseskipobsolete disabled (we do currently). So
I'll send a patch.


That was my impression too, so thanks for sending a patch.


We also discovered many subtle behaviors that might be worth to reconsider.
But let's discuss them after freeze.

I saw that this patch is not public yet. So feel free to discard this.


Yeah, I actually thought I had pushed it. I discovered this morning that I
had not, but then I had also discovered the regression, so I figured I'd
wait.


Excerpts from Jun Wu's message of 2017-07-24 16:47:00 -0700:
> Good case.  I agree the previous behavior is better, but the current
> behavior also looks okay to me. I think we can reuse `adjusteddest` in D21
> to cleanly solve this (and other corner cases).
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.mercurial-scm.org/pipermail/mercurial-devel/attachments/20170724/896d39d5/attachment.html>