From: Carl Worth <cworth@cworth.org>
To: Sebastian Spaeth <Sebastian@SSpaeth.de>, Jameson Rollins
	<jrollins@finestructure.net>, Notmuch Mail list <notmuch@notmuchmail.org>
In-Reply-To: <87ljes0yy0.fsf@SSpaeth.de>
References: <87ska1vh7r.fsf@servo.finestructure.net>
	<87ljes0yy0.fsf@SSpaeth.de>
Date: Wed, 24 Feb 2010 10:26:47 -0800
Message-ID: <87vddmwkuw.fsf@yoom.home.cworth.org>
MIME-Version: 1.0
Content-Type: multipart/signed; boundary="=-=-=";
	micalg=pgp-sha1; protocol="application/pgp-signature"
Subject: Re: [notmuch] [PATCH] Simplify "unread" tag handling in emacs UI.
Precedence: list

--=-=-=

On Wed, 17 Feb 2010 14:33:11 +0100, "Sebastian Spaeth" <Sebastian@SSpaeth.de> wrote:
> On Tue, 19 Jan 2010 17:54:16 -0500, Jameson Rollins <jrollins@finestructure.net> wrote:
> > This patch is intended to greatly simplify the handling of the
> > "unread" tag in the emacs UI.  This patch adds a new function
> > 'notmuch-show-mark-read', that removes the "unread" tag in
> > notmuch-show-mode.  This function is then executed as a
> > notmuch-show-hook, and by notmuch-show-next-message.  All of the
> > functions that explicitly marked messages as unread are removed or
> > renamed.

Hi Jameson,

Thanks for contributing the patch. This exact feature, (removing all
commands with "and mark read" in their names), has been on my todo list
for too long, and I'm anxious to remove it from that. But...

> It then checks the unread status in order to decide whether to proceed
> to the next again. So with your patch notmuch-show-next-unread-message
> will skip through all messages in a thread thinking they are all read
> (and actually marking all as read).

...that seems like a fatal bug in this script. Thanks for noting that
Sebastian.

I'm very interested in augmenting the test suite such that we could
write emacs-based tests that would capture this bug. It seems that with
the --batch, --load, and --funcall options it shouldn't be too hard to
write a modular function exercising emacs code and then execute it
no-interactively. Then we could either inspect the state of the database
to ensure the operation worked as desired, or else have the test
function write a buffer out to a file and test that the file has the
desired contents.

So that's something I'll be working on soon. And of course, I'll always
be glad to accept any help.

-Carl

--=-=-=
Content-Type: application/pgp-signature

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)

iD8DBQFLhW9n6JDdNq8qSWgRAg1SAJ42ckxe1mjQ8EGtTYY4F82oppri8gCeL8vx
TAWYawtFdqm0VCPVvUzww/k=
=3Ud/
-----END PGP SIGNATURE-----
--=-=-=--