From: Rob Browning <rlb@defaultvalue.org>
To: notmuch@notmuchmail.org
Subject: Mail summary comparison with Gnus
Date: Sat, 11 Sep 2010 16:49:18 -0500
Message-ID: <87eid0q7c1.fsf@raven.defaultvalue.org>
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/23.2 (gnu/linux)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Precedence: list


Now that I'm able to start using notmuch with my actual mail, I thought
I might provide some impressions from the perspective of a Gnus user.

Here's what a recent bit of the notmuch list looks like in my notmuch
buffer:

      May 03 [7/7]   (null)               [PATCH] Customize saved search ord
      May 04 [2/2]   (null)               [PATCH] emacs: Pretty print the nu
      May 04 [4/4]   (null)               [PATCH] emacs: More functionality 
      May 07 [4/4]   (null)               Use of "notmuch count" in notmuch-
      May 08 [1/1]   (null)               Fix missing References from broken
      May 10 [1/1]   (null)               rfc: improved MIME handling (multi

and here's what it looks like in my Gnus summary buffer with threads
expanded, which is my default:

   [  39: Keith Packard          ] [PATCH] Customize saved search order separ
     [  19: David Edmondson        ] 
       [  65: Carl Worth             ] 
         [  13: Michal Sojka           ] 
           [  92: Michal Sojka           ] [PATCH RFC] notmuch-hello: What's 
         [  45: David Edmondson        ] [PATCH] Customize saved search order
           [  25: David Edmondson        ] 
   [ 100: David Edmondson        ] [PATCH] emacs: Pretty print the numbers of
     [  17: Carl Worth             ] 
   [  34: David Edmondson        ] [PATCH] emacs: More functionality for `not
     [  18: David Edmondson        ] 
       [  37: David Edmondson        ] 
         [  22: Carl Worth             ] 
   [  35: Jason White            ] Use of "notmuch count" in notmuch-hello.el
     [  24: Jameson Rollins        ] 
       [   7: Jason White            ] 
     [  34: Carl Worth             ] 
   [  23: Arvid Picciani         ] Fix missing References from broken muas.
   [  33: David Edmondson        ] rfc: improved MIME handling (multipart)
   [  50: Michal Sojka           ] [PATCH v2 0/5] Git-based modularization of
     [ 874: Michal Sojka           ] [PATCH v2 2/5] Update test framework for
     [ 444: Michal Sojka           ] [PATCH v2 5/5] test: Set all times to UT
     [1655: Michal Sojka           ] [PATCH v2 1/5] Copy test framework from 
     [ 109: Michal Sojka           ] [PATCH v2 3/5] test: Update helper funct
     [2235: Michal Sojka           ] [PATCH v2 4/5] Convert the actual tests 

(Any articles I've replied to would have an 'A' in the far left column,
 flagged messages would have an exclamation point, and would have red
 text by default, etc.)

The notmuch output probably doesn't have authors because I pulled this
mail in via the mail archives (using Gnus 'G f').  I assume that Gnus
must be able to parse the obfuscated "keithp at keithp.com (Keith
Packard)" style addresses in the archive.  I might see if I can add that
to notmuch.

The notmuch output is definitely more compact, but for me, it favors
information that I don't normally care as much about.  I'd rather see
more subject and author information, and less date.  Though perhaps
that's just me, and I can always customize it.

Also, while I'm not saying that the Gnus approach is ideal, it does make
it easy to get a rough impression at a glance about who's talking, to
whom, and about what, and you can instantaneously collapse/expand
threads (or toggle threading altogether).  That might eventually be a
nice feature for notmuch.

Note too, that Gnus tracks subject changes (at least in this case)
within the thread.  Whether that's what you really want or not is
another question.  (Gnus has a lot of information about how it threads
and how threading can be adjusted in its info pages (c.f. info gnus "3.9
Threading...")).

Hope this is useful
-- 
Rob Browning
rlb @defaultvalue.org and @debian.org
GPG as of 2002-11-03 14DD 432F AE39 534D B592 F9A0 25C8 D377 8C7E 73A4