Sender: Aaron Ecay <aaronecay@gmail.com>
From: Aaron Ecay <ecay@sas.upenn.edu>
To: Adam Wolfe Gordon <awg+notmuch@xvx.ca>
Subject: Re: [PATCH v3 5/5] emacs: Use message-citation-line-format in reply
In-Reply-To:
 <CAMoJFUtn2Ls4maJGecY5Y8HzdOJSR6ZfVdUxnCZ6s4Jhq8r9yw@mail.gmail.com>
References: <1326995217-27423-1-git-send-email-awg+notmuch@xvx.ca>
	<1326995217-27423-6-git-send-email-awg+notmuch@xvx.ca>
	<m2k44n4jlz.fsf@wal122.wireless-pennnet.upenn.edu>
	<CAMoJFUtn2Ls4maJGecY5Y8HzdOJSR6ZfVdUxnCZ6s4Jhq8r9yw@mail.gmail.com>
User-Agent: Notmuch/0.11+73~gd51b784 (http://notmuchmail.org) Emacs/24.0.92.2
	(i386-apple-darwin10.8.0)
Date: Fri, 20 Jan 2012 00:53:39 -0500
Message-ID: <m2mx9i3onw.fsf@wal122.wireless-pennnet.upenn.edu>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
Cc: notmuch@notmuchmail.org
Precedence: list

On Thu, 19 Jan 2012 21:46:46 -0700, Adam Wolfe Gordon <awg+notmuch@xvx.ca> =
wrote:
> On Thu, Jan 19, 2012 at 11:45, Aaron Ecay <ecay@sas.upenn.edu> wrote:
> > Shouldn=E2=80=99t this just use message-insert-formatted-citation-line?
>=20
> Yes, good idea.  I just tried this and it almost works.  The only
> issue is that the default message-mode-citation-line-format has a
> newline at the end, and this function inserts an *additional* newline,
> so we end up with a blank line before the beginning of the quoted
> text.  This is fixable by the user, of course, but it means the
> default out-of-the-box setup will create funny-looking replies, which
> is probably bad.  Thoughts?

(let ((message-citation-line-format
       (remove ?\n message-citation-line-format)))
  ...)

(Or, if you think someone might have a newline other than at the end of
the string, you could do something a little more complicated to remove
only a newline at the end.)

Or you could introduce a new defcustom
=E2=80=98notmuch-message-citation-line-format=E2=80=99 (with newline-less d=
efault), and
let-bind m-l-c-f to that value.  But that is pretty ugly =E2=80=93 we don=
=E2=80=99t want
to have to =E2=80=9Cwrap=E2=80=9D every variable whose default we don=E2=80=
=99t like.

> I've also tried using message-cite-original to create the reply body,
> and it also almost works.  The issue, again, is one of defaults.  In
> addition to the blank line I mentioned above, the default
> message-citation-line-function inserts the "plain" citation line "Foo
> <foo@bar.baz> writes:" instead of the formatted one.  This is a big
> change from the current notmuch default.
>=20
> If everyone's OK with this and willing to customize it themselves,
> then I'm happy to go with this solution, but I'm pretty reluctant to
> change the default in such a significant way.

I=E2=80=99m personally of the opinion that notmuch should just say =E2=80=
=9Cthe mail
composition facility is provided by message mode (here is the
documentation on customizing it)=E2=80=9D.  Message mode is 8,000 lines of
elisp.  We have the choice to:
- write our own message composition mode
- wrap (as explained above) the default message-mode variables we don=E2=80=
=99t
  like in notmuch-prefixed variants, with suitable let-bindings.
- use only the parts of message-mode that we like
- pass composition off to message mode

The first option just doesn=E2=80=99t make sense.  The second is also a lit=
tle
nuts.  The third is what we are trying now, but it=E2=80=99s painful =E2=80=
=93 the
message mode code has its own structure and its own defaults, which
don=E2=80=99t always agree with notmuch=E2=80=99s.  I am in favor of the fo=
urth =E2=80=93
notmuch=E2=80=99s elisp code should pass data to message mode in the most l=
ow
level form it can, and let message mode do any extra work in the way
it already does.  And users should customize message composition via
message mode, not notmuch.  There=E2=80=99s a tradeoff to this approach =E2=
=80=93 by
controlling more within notmuch, we can have nicer defaults at the
expense of more brittle code and/or fewer user-visible customizations.

This is not in any way a criticism of your work (which is great) =E2=80=93
you=E2=80=99re right that you need =E2=80=9Cpermission=E2=80=9D to uproot t=
he defaults, and I=E2=80=99m
advocating for it to be given.

One possible step that might ease the transition pain could be for
notmuch=E2=80=99s emacs interface to have a configuration file (similar to
Wanderlust=E2=80=99s ~/.wl; I believe Gnus also uses a ~/.gnus).  The idea =
is
that this file contains elisp code, and is loaded by notmuch the first
time any notmuch-related commands are invoked by the user.  If the file
does not exist, notmuch could create it with default content that sets
message-citation-function, message-citation-line-format,
message-yank-prefix (to get rid of the ugly default whereby message-mode
indents the original message by four spaces instead of inserting =E2=80=9C>=
 =E2=80=9D),
etc.  If there is interest in this approach, I=E2=80=99d be happy to work o=
n a
patch for it.

I=E2=80=99ve sort of stumped for this idea before
(id:"m239bgcd1d.fsf@gmail.com") and it didn=E2=80=99t exactly get rave revi=
ews.
So I=E2=80=99ll shut up if it=E2=80=99s really not something people want to=
 see.

I=E2=80=99ll close with an example of a nice feature that message mode has
(which I=E2=80=99ve been really wanting since the reply keybindings changed)
that notmuch would get for free if it hooked into message mode better:
the function message-widen-reply takes a reply-to-sender message and
makes it reply-to-all.

--=20
Aaron Ecay