Date: Mon, 9 Jan 2012 15:10:17 +0100
From: Gregor Zattler <telegraph@gmx.net>
To: notmuch <notmuch@notmuchmail.org>
Subject: is there a default reply function? (was: Re: [PATCH v2 5/6] emacs:
	bind 'r' to reply-to-sender and 'R' to reply-to-all)
Message-ID: <20120109141017.GA5306@shi.workgroup>
Mail-Followup-To: notmuch <notmuch@notmuchmail.org>
References: <cover.1325794371.git.jani@nikula.org>
	<cover.1326058946.git.jani@nikula.org>
	<f2c641e47bed1bb76def7ffaed25143071ee7713.1326058946.git.jani@nikula.org>
	<CA+eQo_3kr3L8Nuz3pOTeUfT-Gi2i46yV-Odgo3gmTyOmEi+mTg@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To:
 <CA+eQo_3kr3L8Nuz3pOTeUfT-Gi2i46yV-Odgo3gmTyOmEi+mTg@mail.gmail.com>
User-Agent: Mutt/1.5.21 (2010-09-15)
Precedence: list

Hi Jeremy, notmuch -developers,
* Jeremy Nickurak <not-much@trk.nickurak.ca> [08. Jan. 2012]:
> On Sun, Jan 8, 2012 at 14:48, Jani Nikula <jani@nikula.org> wrote:
>> It seemed to me that most people wanted this, and nobody spoke for keeping
>> the old binding now that we have reply-to-sender. This as a separate patch
>> so it's easy to drop if needed.
> 
> FWIW, I generally prefer reply-all as the default. 

Actually I think there is no default at all.  The perception of a
default here is because two different functions are on "r" and
"R" key respectively while the "r" key feels as a default.

Other MUAs have different key bindings:

message mode has "r" as reply and "w" as "wide-reply"
mutt         has "r" as reply and "g" as "group-reply" and "L" as list-reply

I prefer the mutt way of doing this with three different functions.

> In my experience,
> when a message is sent to a bunch of people, it's usually treated as a
> "forum" discussion where everybody wants to be in on
> everything. 

I think you are right for the majority of cases.  But last week I
got in trouble because I did exactly this.  Perhaps here it is
more important which mistake has more potential for disaster.
Both mistakes could be important but I think there is more
disaster potential in giving information to people who shouldn't
get it than the other way around.  In my perception an UI which clearly
distinguishes two or three reply-functions encourages to think
in advance about which to use.  But last week...

> Also,
> once you've started composing, it's much easier to delete people you
> don't want included than to add people who are no longer referenced in
> the new buffer at all.

True.

Ciao, Gregor
-- 
 -... --- .-. . -.. ..--.. ...-.-