From: Junio C Hamano <gitster@pobox.com>
Date: Wed, 24 Apr 2013 17:49:10 GMT
Subject: Re: Highly inconsistent diff UI
Message-ID: <7va9ones4p.fsf@alter.siamese.dyndns.org>
In-Reply-To: <20130424164003.GB4119@elie.Belkin>

Jonathan Nieder <jrnieder@gmail.com> writes:

> What would it mean for A..B to be treated as a revision range?

Nonsense is what it means ;-)

> Suppose I do a revision walk and come up with the commits x, y,
> and z.  What is the resulting diff?
>
> The common syntax is just a mnemonic: in the same situations as I
> might use "git log A...B", it can be handy to use "git diff A...B".

I wish it were the case, but a very common complaint that I agree
with is that

	git log origin..HEAD

is a way to show what *I* did since I forked, while saying I am *not*
interested in what they did in the meantime.  And

	git log -p origin..HEAD

is a way to view the same history as a series of individual patches,
while that output would not match what you would get from

	git diff origin..HEAD

It would be much closer to

	git diff $(git merge-base origin HEAD) HEAD

which is very often useful and got its own short-hand "origin...HEAD".

And it does not match "git log origin...HEAD" which gives both sides
of the symmetric difference of the history.  To match it, you have
to say "git log --right-only origin...HEAD" or something.