From: Aaron Ecay <aaronecay@gmail.com>
To: David Edmondson <dme@dme.org>, Adam Wolfe Gordon <awg+notmuch@xvx.ca>,
	notmuch@notmuchmail.org, awg@xvx.ca
Subject: Re: [PATCH 0/4] Quoting HTML-only emails in replies redux
In-Reply-To: <cunsjjg59dp.fsf@hotblack-desiato.hh.sledj.net>
References: <1326009162-19524-1-git-send-email-awg+notmuch@xvx.ca>
	<cun4nvx71uk.fsf@hotblack-desiato.hh.sledj.net>
	<m239bgcd1d.fsf@gmail.com>
	<cunsjjg59dp.fsf@hotblack-desiato.hh.sledj.net>
User-Agent: Notmuch/0.11+62~ge41b0d3 (http://notmuchmail.org) Emacs/24.0.92.1
	(i386-apple-darwin10.8.0)
Date: Mon, 16 Jan 2012 17:31:11 -0500
Message-ID: <m2mx9nb7ps.fsf@gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
Precedence: list

On Mon, 16 Jan 2012 08:39:30 +0000, David Edmondson <dme@dme.org> wrote:
> On Mon, 16 Jan 2012 02:38:38 -0500, Aaron Ecay <aaronecay@gmail.com> wrot=
e:
> > - Greater flexibility in the construction of address lists.  For exampl=
e,
> >   there are some email lists where I want replies to list mail to go on=
ly
> >   to the list, not also to the original sender.
>=20
> Is there a mechanistic way to determine the correct behaviour in this
> respect? I suspect that it's exactly the kind of thing that Carl wanted
> to be included in 'notmuch' itself, so that other UIs can benefit.

I think it requires some amount of configuration, but it can be done
sensibly.  I am much more proficient with Elisp than with C, and
Emacs has prejudiced me towards solutions that allow me to have a
Turing-complete configuration language.  :)

I think a good starting point for thinking about mailing lists is what
Mutt does:
http://www.mutt.org/doc/manual/manual-4.html#using_lists

Notmuch at the CLI/C code level could aim for a comparable level of
expressiveness, and I think it would suffice for most people (including
me).

[...]
>=20
> You can do both of these things today using `message-send-hook' (I
> do).

I avoided that, as it seemed to me that just before the message is sent
is too late to be doing these things (I=E2=80=99d like to see confirmation =
when
writing the message that the address/signature changes were applied
correctly).  But =E2=80=9CM-x apropos RET message hook RET=E2=80=9D shows t=
hat there are
some earlier points to hook into as well.  Thanks.

> I dislike supercite, so no support from me in that direction :-)

One advantage of supercite is that it allows non-English speakers to set
up the =E2=80=9COn X, X wrote=E2=80=9D line as they prefer.  Notmuch=E2=80=
=99s current approach
(a hard-coded C string) is the opposite of internationalized.  So it
would be nice to support some customization of that as well, even if we
don=E2=80=99t go the supercite route.

--=20
Aaron Ecay