To: Jameson Graef Rollins <jrollins@finestructure.net>,
 notmuch@notmuchmail.org
Subject: Re: [PATCH 1/3] emacs: Don't return the button from
	`notmuch-show-insert-part-header'.
In-Reply-To: <87bopszxna.fsf@servo.finestructure.net>
References: <1327052612-1040-1-git-send-email-dme@dme.org>
	<1327409620-10133-1-git-send-email-dme@dme.org>
	<1327409620-10133-2-git-send-email-dme@dme.org>
	<87k44hym3i.fsf@servo.finestructure.net>
	<cun8vkwq4ww.fsf@hotblack-desiato.hh.sledj.net>
	<87bopszxna.fsf@servo.finestructure.net>
User-Agent: Notmuch/0.11+105~g08b3b0f (http://notmuchmail.org) Emacs/24.0.92.1
	(x86_64-pc-linux-gnu)
From: David Edmondson <dme@dme.org>
Date: Wed, 25 Jan 2012 06:15:06 +0000
Message-ID: <cun4nvkpatx.fsf@hotblack-desiato.hh.sledj.net>
MIME-Version: 1.0
Content-Type: multipart/signed; boundary="=-=-=";
	micalg=pgp-sha1; protocol="application/pgp-signature"
Precedence: list

--=-=-=
Content-Type: text/plain
Content-Transfer-Encoding: quoted-printable

On Tue, 24 Jan 2012 11:52:09 -0800, Jameson Graef Rollins <jrollins@finestr=
ucture.net> wrote:
> On Tue, 24 Jan 2012 19:25:19 +0000, David Edmondson <dme@dme.org> wrote:
> > On Tue, 24 Jan 2012 10:46:57 -0800, Jameson Graef Rollins <jrollins@fin=
estructure.net> wrote:
> > > Is there a reason it's really necessary to make this change?  Can't
> > > callers just ignore the returned button if they don't care about it
> > > further?  I can see that maybe it's nice to be able to specify
> > > parameters at creation time, but I'm not sure why that requires throw=
ing
> > > out the returned object as well.
> >=20
> > Patches 2 and 3 in that series can result in the button not being
> > inserted.
>=20
> Can patches 2 and 3 be rewritten so they are compatible with the button
> being returned by the button creation function?

No. The whole purpose of 2 and 3 is that they don't insert a button in
some circumstances. If no button is inserted then there is no button to
return.

The changes to `notmuch-show-insert-part-multipart/signed' and
`notmuch-show-insert-part-multipart/encrypted' in patch 1 could be
re-written to anticipate that a button might not be returned, but the
resulting code was more complex that that I sent.

> > > I can see that maybe it's nice to be able to specify parameters at
> > > creation time, but I'm not sure why that requires throwing out the
> > > returned object as well.

Do you have a specific use case for this?

--=-=-=
Content-Type: application/pgp-signature

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)

iEYEARECAAYFAk8fneoACgkQaezQq/BJZRYxLwCfVU7BLe2QG91SPjoehqJ1YLZM
FBgAnAzPRbV4a8T5YG+fxP5/+lT4yWa7
=fZ6e
-----END PGP SIGNATURE-----
--=-=-=--