headers
Kyle Wheeler | 3 Nov 20:31

Re: canonicalizing and de-canonicalizing

On Friday, October 31 at 11:21 PM, quoth TAKAHASHI Tamotsu:
>So here is another patch.
>%e is a pretty version of %f in $folder_format.

Hmm... there's a minor problem... in browsing around an IMAP server, 
going to the parent folder is normally labelled "..", but after going 
through mutt_pretty_mailbox, it gets transformed into the parent 
directory of the local disk. It's easy to special-case that, of 
course... but is that the right thing to do?

~Kyle
--

-- 
It is said that power corrupts, but actually it's more true that power 
attracts the corruptible. The sane are usually attracted by other 
things than power.
                                                          -- David Brin
Permalink | Reply | 1 comment |
headers
TAKAHASHI Tamotsu | 4 Nov 00:47
Picon

Re: canonicalizing and de-canonicalizing

* Mon Nov  3 2008 Kyle Wheeler <kyle-mutt-dev <at> memoryhole.net>
> On Friday, October 31 at 11:21 PM, quoth TAKAHASHI Tamotsu:
>> So here is another patch.
>> %e is a pretty version of %f in $folder_format.
>
> Hmm... there's a minor problem... in browsing around an IMAP server,  
> going to the parent folder is normally labelled "..", but after going  
> through mutt_pretty_mailbox, it gets transformed into the parent  
> directory of the local disk.

Oops, I'm sorry.

> It's easy to special-case that, of  
> course... but is that the right thing to do?

grepping '\"\.\.' *.c tells me that it's normal.

imap/browser.c gives you "../" (not ".."),
so strcmp(s,"../") will fix this problem.

      if (op == 'e' && mutt_strcmp (s, "../"))

--

-- 
tamo
Permalink | Reply | 0 comments |
headers
Thomas Parmelan | 4 Nov 15:58
Picon
Favicon

False envelope From saved via fcc when using envelope_from_adress

Hi,

I am using the following configuration (real domains masked) :

set from="tom <at> friendlydomain"
set use_from
set use_envelope_from

send-hook . 'unmy_hdr from'
send-hook . 'set from="tom <at> friendlydomain"; set realname="Thomas Parmelan"
send-hook . 'unset envelope_from_address'

[here, other send-hooks possibly modifying from or
envelope_from_address, then : ]

send-hook       '~C "@workdomain"'       'set envelope_from_address=tom <at> friendlydomain'
send-hook       '~C "@workdomain"'       'my_hdr From: "PARMELAN, Thomas" <thomas.parmelan <at> workdomain>'

It works as I want : when sending mail to a coworker, the From: is my
work adress, but the enveloppe from is my "friendlydomain" address.
(Yes, this is somewhat a bizarre configuration, but it works and is my
only want to send work mail from mutt instead of the outlook/exchange
nightmare).

The problem I have is that in this case the saved copy of my sent mails
have not the real enveloppe from, but the one derived from the From:.

I am using Mutt 1.5.18 (more precisely mutt 1.5.18-4 on a lenny
(testing) Debian/GNU Linux amd64 host.
(Continue reading)

Permalink | Reply | 5 comments |
headers
Kyle Wheeler | 4 Nov 16:49

Re: False envelope From saved via fcc when using

On Tuesday, November  4 at 03:58 PM, quoth Thomas Parmelan:
> The problem I have is that in this case the saved copy of my sent 
> mails have not the real enveloppe from, but the one derived from the 
> From:.

Mutt doesn't store the envelope from in the saved copy of the message. 
The envelope is not part of the message any more than the envelope 
with the stamp on it is part of a letter that your grandma might send 
you.

Why do you want the saved "sent" message to be something other than 
was actually sent?

> Am I doing something wrong, or is this a real bug ?

You aren't doing anything wrong, nor is it a bug. It seems to be that 
you're expecting something to happen that is not supposed to happen. 
It's like saying "I made coffee to drink on my way to work, and filled 
my car's gas tank with gasoline, but when I got to work, for some 
reason the gas tank still only had gasoline rather than coffee in it." 
That's what I would *expect* to be the case; why did you expect the 
gas tank to magically be filled with coffee? Why did you expect the 
envelope to be stored in the message rather than the original headers 
of the message?

~Kyle
--

-- 
No nation is permitted to live in ignorance with impunity.
                                                    -- Thomas Jefferson
(Continue reading)

Permalink | Reply | 3 comments |
headers
Elimar Riesebieter | 4 Nov 17:06
Picon
Favicon

Re: False envelope From saved via fcc when using

* Thomas Parmelan [081104 16:00 +0100]
> Hi,
> 
> I am using the following configuration (real domains masked) :
> 
> set from="tom <at> friendlydomain"
> set use_from
> set use_envelope_from
> 
> send-hook . 'unmy_hdr from'
> send-hook . 'set from="tom <at> friendlydomain"; set realname="Thomas Parmelan"
> send-hook . 'unset envelope_from_address'
> 
> [here, other send-hooks possibly modifying from or
> envelope_from_address, then : ]
> 
> send-hook       '~C "@workdomain"'       'set envelope_from_address=tom <at> friendlydomain'
> send-hook       '~C "@workdomain"'       'my_hdr From: "PARMELAN, Thomas" <thomas.parmelan <at> workdomain>'
> 

Try instead of your settings:

set use_from
set use_envelope_from
set realname="Thomas Parmelan"

unhook *

folder-hook . unmy_hdr *
folder-hook . my_hdr From: <tom <at> friendlydomain>
(Continue reading)

Permalink | Reply | 0 comments |
headers
Thomas Parmelan | 5 Nov 15:11
Picon
Favicon

Re: False envelope From saved via fcc when using

Le mardi 04 novembre 2008 à 16:49, d'après
Kyle Wheeler <kyle-mutt-dev <at> memoryhole.net> :

> Mutt doesn't store the envelope from in the saved copy of the message.  

Yes it does. I use mbox folders. The From separator usually uses the
envelope sender. In fact, it really does for incoming messages...

> Why do you want the saved "sent" message to be something other than was 
> actually sent?

On the contrary, I want it to be exactly what was sent!

> Why did you expect the envelope to be stored in the message rather
> than the original headers of the message?

I expected the envelope sender to be stored in the From separator line
of my mbox folder, and the header stored in the From: header, just like
it is for incoming messages :

    From tom <at> friendlydomain.com  Wed Nov  5 12:00:08 2008 
    From: "PARMELAN, Thomas" <thomas.parmelan <at> workdomain.com>
    Date: Wed, 5 Nov 2008 11:58:44 +0100
    Subject: ....

I suspect it can be useful also for postponed messages or messages
created from a saved message (via <resend-message>)...

--

-- 
Thomas Parmelan
(Continue reading)

Permalink | Reply | 2 comments |
headers
Kyle Wheeler | 5 Nov 15:38

Re: False envelope From saved via fcc when using

On Wednesday, November  5 at 03:11 PM, quoth Thomas Parmelan:
>> Mutt doesn't store the envelope from in the saved copy of the 
>> message.
>
> Yes it does. I use mbox folders. The From separator usually uses the 
> envelope sender. In fact, it really does for incoming messages...

Ahh, sorry, I was thinking of Maildirs.

In any case, the problem you face here is that mboxes are not well 
defined (there's no official standard). Imagine moving a message from 
a Maildir into an mbox---what should be used to generate the From_ 
line?

Granted, mutt could probably work harder to preserve envelope 
information when working exclusively with mboxes, and you can submit 
an enhancement request to the mutt-dev folks.

That said, I don't think that relying on the From_ separator for 
useful information is a good idea. It's just too unreliable. The From_ 
line has no reliable format specification, so extracting information 
from it is, at best, a guessing game. With certain software 
combinations, you can get lucky and have a predictable format there, 
but... it's perfectly legitimate to use a separator line in your mbox 
that begins with "From an mbox generator!" or contains random ascii 
characters after the "From " prefix.

> I suspect it can be useful also for postponed messages or messages 
> created from a saved message (via <resend-message>)...
(Continue reading)

Permalink | Reply | 1 comment |
headers
TAKAHASHI Tamotsu | 6 Nov 03:34
Picon

inconsistent expansion of mailbox shortcuts

If your $spoolfile points to a Maildir, you cannot
tab-complete "!" correctly.

 :set spoolfile=/somewhere/Maildir
 c!<tab><tab><tab>

will give you "!///".

This is caused by a conflict between mutt_complete()
and mutt_expand_path().

"+" and "=" always point to a directory, not a file.
So "+foo" may/should expand to "/somewhere/Mail/foo"
even when your $folder is set to "/somewhere/Mail".
Note it doesn't end with "/".

But ~ < > ! - ^ and @ not always mean a directory.
Actually, mutt_expand_path() expands "!foo" to
"/somewhere/Maildirfoo" instead of "/somewhere/Maildir/foo"
if your $spoolfile is "/somewhere/Maildir" (without
a trailing slash, again).

Reading through complete.c, I thought that it was good
to expand these special shortcut chars and tell users
what they are doing while completing.

So here is a PoC patch.

--

-- 
tamo
(Continue reading)

Permalink | Reply | 1 comment |
headers
Vincent Lefevre | 6 Nov 17:13
Gravatar

Re: False envelope From saved via fcc when using

On 2008-11-05 08:38:13 -0600, Kyle Wheeler wrote:
> In any case, the problem you face here is that mboxes are not well  
> defined (there's no official standard). Imagine moving a message from a 
> Maildir into an mbox---what should be used to generate the From_ line?

Mutt uses the Return-Path address, which should be a copy of the
envelope From.

> That said, I don't think that relying on the From_ separator for useful 
> information is a good idea. It's just too unreliable.

I agree. Return-Path is better.

--

-- 
Vincent Lefèvre <vincent <at> vinc17.org> - Web: <http://www.vinc17.org/>
100% accessible validated (X)HTML - Blog: <http://www.vinc17.org/blog/>
Work: CR INRIA - computer arithmetic / Arenaire project (LIP, ENS-Lyon)
Permalink | Reply | 0 comments |
headers
Mutt | 7 Nov 20:35

[Mutt] #3132: askfrom: prompt for from-address before editing the body of an outgoing message

#3132: askfrom: prompt for from-address before editing the body of an outgoing
message
------------------------------+---------------------------------------------
 Reporter:  pebo              |       Owner:  mutt-dev
     Type:  enhancement       |      Status:  new     
 Priority:  major             |   Milestone:          
Component:  mutt              |     Version:  1.5.18  
 Keywords:  askfrom ask-from  |  
------------------------------+---------------------------------------------
 Hi,

 when one has to deal with multiple from-addresses, `alternates` and
 `reverse_name` are a great time-saver. To make the situation perfect,
 there is just one component missing -- `askfrom`. When replying to a
 message `reverse_name` does its job very well, but when composing a new-
 one, you always have to remember changing the From: header through using
 `<ESC>-f` after editing the body. Obviously, I often forget to do this.

 `askcc` & `askbcc` exist and there was even an `askfrom` patch back in
 2002:

 http://www.mail-archive.com/mutt-users <at> mutt.org/msg28714.html

 but I couldn't figure out why it wasn't accepted back then.

 Thanks,

 Petar

--

--
(Continue reading)

Permalink | Reply | 1 comment |

Gmane