Mutt

The best MUA I know.
Mutt homepage
'official' Mutt wiki

My Patches

patch-1.5.6.nr.crypt-autohook
patch-1.5.8.nr.crypt-autohook
It should be a (better) replacement for constructs like this:

source 'gpg --list-keys | perl -e "while(<>){/(\\b|[<(])([^ \\t]+@[^ \\t]+\.[^ \t]+)(\\b|[>)])/ and \$a{\$2}=1;}print map qq/send-hook \"~C \$_\" set pgp_autoencrypt\n/, keys %a;"|'

The patch introduces 3 new options:

       6.3.128  crypt_autohook_sign

       Type: boolean

       Default: no

       Setting this variable will cause Mutt to sign messages if there is a key for
       every recipient. This is done immediately after executing the send-hooks
       (and also only once after getting the initial list of recipients).  Only work-
       ing with pgp at the moment.  (Crypto only)

       6.3.129  crypt_autohook_encrypt

       Type: boolean

       Default: no

       Setting this variable will cause Mutt to encrypt messages if there is a key for
       every recipient. This is done immediately after executing the send-hooks
       (and also only once after getting the initial list of recipients).  Only work-
       ing with pgp at the moment.  (Crypto only)

       6.3.130  crypt_autohook_strongonly

       Type: boolean

       Default: yes

       If this variable is set only 'strong' keys are considered for ``$crypt_auto-
       hook_encrypt (section 6.3.129  , page 89)'' and ``$crypt_autohook_sign (section
       6.3.128  , page 89)'' Only working with pgp at the moment.  (Crypto only)

I like this appoach better than the above snippet for some reasons.

• No need to parse gpg output, the existing code from mutt is used. It should work with any pgp implementation mutt can work with.
• It notices keyring changes which happen while mutt is running.
• Only strong keys are considered (optional).
• It's easier to change with folder- and send-hooks.
• If you send a mail to multiple recipients, for some of whom you have keys and some not, you are not asked for nonexistent keys, since encryption is not selected.

Old Stuff

I don't maintain these things any longer, either because nobody seems to use them (including me), because they are included in mutt or because they are superseded by something better.

patch-1.3.28.nr.threadcomplete
It adds an new pattern ~a. I see only two possible uses of it, "T~a\n" and "^T! ~a\n". The first one tags all messages which are in threads containing any tagged message, the second one untags all messages which are in threads containing any untagged messages. I should add that ~a and !~a are not inverse. The not applies to the tag status of the other messages in the thread.

It seems this patch can also be useful together with <limit>.

This patch should work also work with newer versions of mutt
Recent versions of mutt include a more general feature ~(PATTERN).

patch-1.3.23.2.nr.color-questions
patch-1.3.28.nr.color-questions-plain-mutt
patch-1.3.28.nr.color-questions
You can color the questions mutt asks you with color question after applying this patch. I just noticed that this patch does not apply to a plain mutt. I use it with the FreeBSD port of mutt, where it applies cleanly. I produced a version for use with plain mutt, but I haven't tested it.
It seems this patch does not apply to recent version of mutt (I tried 1.5.3) Since I no longer use it, and it seems that nobody else uses it, I don't try to update it.

patch-1.3.23.2.nr.tag_prefix_cond
patch-1.3.28.nr.tag_prefix_cond
patch-1.5.4.nr.tag_prefix_cond
This one adds 2 new functions, they are only usefull inside macros or push-commands, the first one tag-prefix-cond does exactly the same as tag-prefix, with one difference, if there are no tagged messages, it "eats" the rest of the macro, aborting its execution.
The second one end_cond, does nothing (its a noop), but if tag-prefix-cond "eats" an "end_cond", it stops "eating" the rest of the macro, which will then be executed as normal.
end_cond was named special in the patches before 1.3.28
I haven't tried it, but at least 2 mutt-users told me, that this patch works with mutt-1.4.
Since version 1.5.5 this patch is included in mutt, but end_cond has been changed to end-cond.

patch-1.3.27.nr.unmailboxes.1
patch-1.3.28.nr.unmailboxes.1
This adds an unmailboxes command.
This patch is included in recent versions of mutt (I think since 1.5.2).

This changes "old" messages to "new" messages (works only for Maildirs)
old2new.pl
I mark messages I want to read later as old, and later I run this script, and I can cycle through all folders with (now) "new" mails again.

This moves old mails from one mbox to another. It handles gzipped and bzip2ed mbox files. This one works only for mboxes and needs mutt_dotlock and grepmail.
move_mail.sh

Last change (on this page): $Date: 2007/06/14 22:09:42 $