From: Daurnimator <quae@daurnimator.com>
Date: Fri, 06 May 2016 06:06:11 GMT
Subject: Re: [PATCH v2] git-stash: Don't GPG sign when stashing changes
Message-ID: <CAEnbY+cAzLKw9GJTWy5ZP4rGsve161qSX5DgqWimcC5kYRqkyA@mail.gmail.com>
In-Reply-To: <CAPc5daXVBJ=2KJfqcdwyH1Eq54vw-QeOEBF1aGj2dn+Yihcy9A@mail.gmail.com>

On 3 May 2016 at 11:21, Junio C Hamano <gitster@pobox.com> wrote:
> On Mon, May 2, 2016 at 4:56 PM, Daurnimator <quae@daurnimator.com> wrote:
>> On 3 May 2016 at 07:57, Junio C Hamano <gitster@pobox.com> wrote:
>>
>> I agree quiltimport should not, but I think filter-branch possibly
>> should... at least for your *own* commits.
>> I often think of filter-branch as an "advanced" `git commit --amend`
>
> But it does not do the gpgsign thing only for your own commits right now, and
> even if it did, you would not necessarily want it to always use the configured
> setting. That means that the command, if it wants to work well with the signed
> commits, must learn to honor command line option to enable or disable passing
> --gpgsign option to underlying commit-tree *anyway*.
>
> So I think it is a right thing to fix commit-tree to ignore
> commit.gpgsign at least
> by default.

Okay, that makes sense.
Will you work on a patch?