Date: Wed, 9 Mar 2016 07:13:09 -0700
From: Jonathan Corbet <corbet@lwn.net>
To: Laszlo Ersek <lersek@redhat.com>
Cc: David Woodhouse <dwmw2@infradead.org>, linux-kernel@vger.kernel.org,
        Randy Dunlap <rdunlap@infradead.org>
Subject: Re: [PATCH 2/7] Docs: Bring SubmittingPatches more into the git era
Message-ID: <20160309071309.36141068@lwn.net>
In-Reply-To: <56E00C9A.4060703@redhat.com>
References: <1419352327-4517-1-git-send-email-corbet@lwn.net>
	<1419352327-4517-3-git-send-email-corbet@lwn.net>
	<1457516710.118898.188.camel@infradead.org>
	<56E00C9A.4060703@redhat.com>
Organization: LWN.net
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 8bit
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 832
Lines: 22

On Wed, 9 Mar 2016 12:44:26 +0100
Laszlo Ersek <lersek@redhat.com> wrote:

> Namely, do signed tags serve the purpose that a higher level maintainer
> can pull from a trusted, lower level maintainer without looking?
> 
> At these higher levels of the patch flow, does "trusted identity"
> replace "review"?

No, I really don't think so.  Signed tags just verify the origin of the
pull request.

Think of it as a form of defense in depth.  Anybody who merges code into
the kernel merges bugs on a regular basis, even if they carefully review
every line.  Review is a defense against threats like the deliberate
insertion of malevolent code, but it is not an absolute defense.  Signed
tags, one might hope, will at least keep code from deliberately forged
pull requests out of the stream of code needing review.

Or so I see it.

jon