[1930] in athena10
Re: mutt config
daemon@ATHENA.MIT.EDU (Greg Price)
Sat Mar 28 21:22:55 2009
Date: Sat, 28 Mar 2009 21:22:25 -0400
From: Greg Price <price@MIT.EDU>
To: Shreevatsa R <vatsa@mit.edu>
Cc: debathena@mit.edu
Message-ID: <20090329012225.GC21912@vinegar-pot.mit.edu>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20090325172034.GA49454@WHITAKER-THREE-TEN.MIT.EDU>
On Wed, Mar 25, 2009 at 01:20:34PM -0400, Shreevatsa R wrote:
> > set folder="imap://$USER@`hesinfo $USER pobox | awk '{print $2}'`"
> > set spoolfile="imap://$USER@`hesinfo $USER pobox | awk '{print $2}'`"
> > set record=""
> >
> > $folder is the root of the user's mail tree. $spoolfile is the inbox;
> > I'd set it to just "+INBOX", where "+" is a Mutt abbreviation for the
> > value of $folder.
> > Recommendation: definite keep.
>
> [One thing I don't understand is why I need that on Athena and
> set folder = "imaps://$USER.mail.mit.edu/"
> set imap_user = "$USER"
> set spoolfile = "+INBOX"
> on my laptop -- is there some configuration that would work on both?]
Does the other configuration not work on your laptop? Do you have the
hesinfo program? That's pulled in and configured by
debathena-hesiod-config.
Other than that, the configurations differ by vatsa.mail.mit.edu <->
po14.mit.edu and otherwise look equivalent.
> I just use
> my_hdr bcc vatsa@mit.edu
> to bcc all emails to myself so that they get threaded with replies etc.,
> but maybe that's an unpopular personal preference. :)
>
> It's probably best to not set it, so that it defaults to ~/sent, which
> is better than nothing.
Yep, ended up setting it to +INBOX.sent-mail to match Alpine and to
keep it in IMAP.
> > set sort=threads
> > Recommendation: drop
>
> Agreed. Most users are confused by threads, anyway, and without
> set sort_aux='last-date-received'
> one ends up having new email at random locations in the inbox.
Well, if we wanted to set this we could set that too. (As I do.)
But it's a matter of preference that possibly upstream should do
differently but certainly isn't MIT-specific.
> 1. Each time one quits mutt, one is prompted:
> Move read messages to /afs/athena.mit.edu/user/v/a/vatsa/mbox? ([no]/yes):
> This is quite annoying. I just "set move=no", and I wonder if any users
> would actually want this. The move is also hard to reverse; it's easy to
> accidentally hit 'y' and not know how to move mail back into MIT mail.
Thanks, included this.
> 2. Caching: Large mailboxes take quite a while to load. It would be nice
> to enable caching by default. Something like
> set header_cache=~/.muttcache
> It can make quite a bit of difference.
I also do this. Maybe we should. It's true that without it, Mutt on
a big mailbox is nearly unusable unless you run it in a screen and
~never restart it.
I have occasionally had instability problems where Mutt crashes upon
trying to open my (huge) archive folder, and rm'ing the header cache
so it gets regenerated fixed things, so I'd worry a bit about users
hitting that and not knowing about the cache. But that was on the
mutt-1.5.12 in the sipb locker, and maybe that bug is gone now.
> 3. "set imap_check_subscribed"
> So that the user doesn't have to manually poll all folders to check for
> new mail.
This looks like an IMAP feature I didn't know about. It seems like a
mistake for Mutt upstream not to set this by default -- if the user
configured their IMAP account to know what folders they want to see
new mail in, it's silly and contrary to the one-account, many-clients
principle of IMAP for Mutt to demand that they enter the same
information in its own configuration.
I'd try to pursue this in the Debian BTS, or in Mutt upstream.
Greg
