hg rebase killed branch name?

[Matt Mackall]

On Sat, 2015-01-17 at 12:29 +0300, anatoly techtonik wrote:
> On Sun, Jan 11, 2015 at 10:51 PM, Augie Fackler <raf at durin42.com> wrote:
> > On Jan 11, 2015, at 7:04 AM, anatoly techtonik <techtonik at gmail.com> wrote:
> >>
> >> I rebased changes in my named branch and now see that changesets don't
> >> belong to that branch anymore. I though that rebase is for moving
> >> branches relative to the tree, not to move changesets to a different
> >> branch.
> >>
> >> Is it a bug? If yes, then why --keepbranches is here?
> >
> > As you’ve surmised, what you’ve experienced is the expected behavior, and what you wanted is what you’d get with --keepbranches.
> 
> But why it is expected? Why implicitly killing branch name is expected
> behavior while rebasing?

There are a few answers to your question:

a) it just happened to work that way when rebase was created because of
the merge direction it uses internally
b) that seemed like an ok choice at the time because rebase is often
used to merge feature branches into mainline
c) ("that's the way it worked yesterday") ** 2343

Of course, the only reason that matters today is (c) because of that big
exponent. As with many of your questions on this list, you might as well
be asking why clocks turn clockwise: it's definitely not up for
discussion.

-- 
Mathematics is the supreme nostalgia of our time.