From: Junio C Hamano <gitster@pobox.com>
Date: Mon, 15 Mar 2010 05:42:03 GMT
Subject: Re: [PATCH] rebase--interactive: don't enforce valid branch
Message-ID: <7vvdcygmz8.fsf@alter.siamese.dyndns.org>
In-Reply-To: <alpine.DEB.2.00.1003142227100.796@narbuckle.genericorp.net>

Dave Olszewski <cxreg@pobox.com> writes:

>>> +test_expect_success 'rebase while detaching HEAD' '
>>> +	grandparent=$(git rev-parse HEAD~2) &&
>>> +	test_tick &&
>>> +	FAKE_LINES="2 1" git rebase -i HEAD~2 HEAD^0 &&
>>
>> What's the point of saying this?  You could instead say:
>>
>> 	git rebase -i HEAD~2
>>
>> no?
>
> There's already a test for rebasing on a previously detached HEAD.  The
> form "git rebase -i HEAD~2" specifies a non-branch upstream, but doesn't
> take the branch argument which is the point of the change.

What I meant was that if you prefer to work on a detached HEAD (and I
sometimes do), then your HEAD would likely to be detached already when you
run rebase.  IOW, I would expect that 

	git checkout HEAD^0
        ... perhaps do something, perhaps do nothing, here ...
        git rebase -i HEAD~2

would be a lot more natural thing to do, and in that case you do not need
to say HEAD^0 there.