Re: [RFC/PATCH 0/3] use '--bisect-refs' as bisect rev machinery option

From: Linus Torvalds <torvalds@linux-foundation.org>
Date: Nov 4, 2009, 18:32 UTC
Message-ID: <alpine.LFD.2.01.0911041030080.31845@localhost.localdomain>
In-Reply-To: <7vljimgnaz.fsf@alter.siamese.dyndns.org>

On Wed, 4 Nov 2009, Junio C Hamano wrote:

Show 14 quoted lines

> Christian Couder <chriscool@tuxfamily.org> writes:
> 
> > So I suggest to use '--bisect-refs' instead of '--bisect' as the new
> > bisect revision machinery option, because otherwise I think we get a
> > regression when we call "git rev-list --bisect BAD --not GOOD" and we
> > are already bisecting with bisect refs different than BAD and GOOD.
> > This also simplifies the code a little bit.
> 
> Just to make sure that I read you correctly, do you mean that Linus now
> would say:
> 
>     $ git bisect ...
>     ... inside bisect session
>     $ gitk --bisect-refs arch/x86

I'd really rather prefer to keep '--bisect' for that, it doesn't make sense to export '--bisect-refs' to be the user-visible switch, and then have '--bisect' as some internal switch just for 'git rev-list'.

In fact, I'd argue that the ne wbehavior of 'git rev-list' is to some degree _better_ (ie the existing bad/good ones are implicit while doing bisection).

Yes, it is a behavioral change, but is it a bad one?

			Linus