headers
Mutt | 1 Jan 21:50

Re: [Mutt] #3002: mutt-1.5.17: 'N'ew flag keeps being set on mails

#3002: mutt-1.5.17: 'N'ew flag keeps being set on mails in IMAP folders

Changes (by pw):

  * owner:  mutt-dev => brendan
  * component:  mutt => IMAP

Comment:

 I've been frustrated by this too.  It is repeatable.  Get a few new
 messages at the bottom of the folder (at least 3).  Delete the last
 one, reading it or not.  Mark the previous ones as read, either by
 reading or explicitly turning off the New flag.  Quit.  Upon rereading
 the folder, you'll see that the first non-deleted mail has been marked
 seen, but the others are still marked New.

 The problem is in imap_make_msg_set, so I'm changing the component
 of this ticket to IMAP.  It is called by sync_helper in imap_sync_mailbox
 to update the five flags of each message that may have changed.  The
 code at the end of imap_make_msg_set sets "setstart" to the first message
 that has been read, but then defers extending the buffer with each
 subsequent message, hoping to make a range.  Eventually it finds the
 last message, which has been deleted, remember, but fails to terminate
 the range with ":%u" of the previous messages, because this last deleted
 message is not ->active.

 I'm a bit wary about making a patch as I'm not sure why this test on
 active is in there.  Changing line 946 in imap/imap.c 1.5.17 to read:

 {{{
(Continue reading)

Permalink | Reply | 0 comments |
headers
Brendan Cully | 3 Jan 09:00
Gravatar

mutt: new changeset

New changeset in mutt:

http://dev.mutt.org/hg/mutt/rev/3cf8bc21bc3d
changeset:   5336:3cf8bc21bc3d
branch:      HEAD
tag:         tip
user:        Johan Svedberg <johan <at> svedberg.com>
date:        Wed Jan 02 11:47:33 2008 -0800
summary:     Updated Swedish translation

--

-- 
Repository URL: http://dev.mutt.org/hg/mutt
Permalink | Reply | 0 comments |
headers
Mutt | 3 Jan 16:03

[Mutt] #3007: cache doesn't set imap flags correctly

#3007: cache doesn't set imap flags correctly

 mutt-1.5.17
 When using the header cache, most of the time the messages that were read
 in some mailbox get the wrong flags. For example, one reads about a dozen
 new messages. They are all marked as read. Mailbox is synced ($). Exit the
 mailbox and reenter: the same messages are marked as new again, or
 sometimes old.

 In fact, I'm not 100% sure it's due to the header cache code: I basically
 can't use mutt + imap without this cache, so I didn't give it a try in
 that way. But it happens for me and my colleagues.

 I do have some external patches applied:
 - nntp
 - xterm-title
 - compressed folders

 If no-one else has seen this behaviour, I'm willing to give it a try
 without these patches.

--

-- 
Ticket URL: <http://dev.mutt.org/trac/ticket/3007>
Permalink | Reply | 3 comments |
headers
Mutt | 3 Jan 16:29

Re: [Mutt] #3007: cache doesn't set imap flags correctly

#3007: cache doesn't set imap flags correctly

Comment (by Fabian Groffen):

 {{{
 On 03-01-2008 15:03:28 -0000, Mutt wrote:
 > #3007: cache doesn't set imap flags correctly
 >
 >  mutt-1.5.17
 >  When using the header cache, most of the time the messages that were
 read
 >  in some mailbox get the wrong flags. For example, one reads about a
 dozen
 >  new messages. They are all marked as read. Mailbox is synced ($). Exit
 the
 >  mailbox and reenter: the same messages are marked as new again, or
 >  sometimes old.

 Sounds a lot like http://dev.mutt.org/trac/ticket/3002
 I'm testing the patch suggested there
 (http://overlays.gentoo.org/proj/alt/browser/trunk/prefix-overlay/mail-
 client/mutt/files/mutt-1.5.17-new-flag-hang.patch)
 and sofar it has worked wonderfully.  I haven't noticed any side effects
 yet.  I just don't have to remark mail as read all the time, which is
 nice.

 }}}

--

-- 
Ticket URL: <http://dev.mutt.org/trac/ticket/3007#comment:>
(Continue reading)

Permalink | Reply | 0 comments |
headers
Mutt | 3 Jan 16:32

Re: [Mutt] #3007: cache doesn't set imap flags correctly

#3007: cache doesn't set imap flags correctly

Comment (by Bruno Postle):

 {{{
 On Thu 03-Jan-2008 at 15:03 -0000, Mutt wrote:
 >#3007: cache doesn't set imap flags correctly
 >
 > When using the header cache, most of the time the messages that were
 read
 > in some mailbox get the wrong flags. For example, one reads about a
 dozen
 > new messages. They are all marked as read. Mailbox is synced ($). Exit
 the
 > mailbox and reenter: the same messages are marked as new again, or
 > sometimes old.

 I see this occasionally on an unpatched mutt with courier-imap and
 header cacheing.  Specifically it seems to go something like this:
 read a message (clearing any 'new' flag), mark it as 'new' with 'N',
 then mark it as 'read' with 'N'.  Close the mailbox, reopen and the
 message is 'new' again.

 I tend to keep the same mailbox open for weeks on end, so possibly
 this only happens when the mailbox is closed abruptly.
 }}}

--

-- 
Ticket URL: <http://dev.mutt.org/trac/ticket/3007#comment:>
(Continue reading)

Permalink | Reply | 0 comments |
headers
Mutt | 3 Jan 16:43

Re: [Mutt] #3002: mutt-1.5.17: 'N'ew flag keeps being set on mails

#3002: mutt-1.5.17: 'N'ew flag keeps being set on mails in IMAP folders

Comment (by ahasenack):

 I see the same thing here with 1.5.17, and have been seeing it for quite a
 while in older versions (just don't recall exactly since when).

--

-- 
Ticket URL: <http://dev.mutt.org/trac/ticket/3002#comment:2>
Permalink | Reply | 0 comments |
headers
Mutt | 3 Jan 16:44

Re: [Mutt] #3007: cache doesn't set imap flags correctly

#3007: cache doesn't set imap flags correctly

Changes (by ahasenack):

  * status:  new => closed
  * resolution:  => duplicate

Comment:

 Yes, it's the same as ticket #3002, I'll close this one.

--

-- 
Ticket URL: <http://dev.mutt.org/trac/ticket/3007#comment:3>
Permalink | Reply | 0 comments |
headers
Mutt | 3 Jan 16:56

Re: [Mutt] #3002: mutt-1.5.17: 'N'ew flag keeps being set on mails

#3002: mutt-1.5.17: 'N'ew flag keeps being set on mails in IMAP folders

Changes (by ahasenack):

  * version:  => 1.5.17
  * milestone:  => 1.6

Comment:

 Setting this for 1.6 (I hope I can do that :)

--

-- 
Ticket URL: <http://dev.mutt.org/trac/ticket/3002#comment:3>
Permalink | Reply | 0 comments |
headers
Mun Johl | 4 Jan 20:23
Favicon

Crashes with the "indent" patch on Mutt v1.5.17

Hi all,

Quite a while ago, Jimmy Mäkelä wrote an "indent" patch for Mutt that I
have been using for a long time.  However, when I applied it to Mutt
v1.5.17, it caused a crash.  I've attached the patch for reference.
Note that I compiled with gcc v3.3.2 on a Sun Solaris 8 system.

I'm hoping there may be someone out there that is familiar with the
patch, or just willing to help with the debug.

I last applied the indent-patch to mutt 1.5.15 .  I didn't try 1.5.16;
but when applied to 1.5.17 mutt crashes when I try to reply to a
message.  Note that there are no other patches applied.  Here is the
backtrace from the core file:

------------------------------ Delimiter BEGIN --------------------------------
core 'core' of 6419:    ./mutt
 0007f02c mutt_FormatString (ffbe8ba8, 7f, 0, 69, 2e004, 13f078) + c0
 0002e4f4 _mutt_copy_message (e30b8, e30a8, 13ec00, 13ec60, 167, 86) + 68
 0002ea98 mutt_copy_message (ffffffff, 127c58, 13ec00, 167, 86, b0) + 38
 0006e860 include_reply (0, 13ec00, e30b8, e2af0, e3000, e3400) + a0
 0006f328 generate_body (ffffffff, 1eaf88, 1, 127c58, 13ec00, ffffff8a) + 25c
 00070ccc ci_send_message (0, 0, 0, 127c58, 1, 0) + 1344
 0003547c mutt_index_menu (1, 0, 0, 0, 4, 0) + 3ffc
 0004ba18 main     (1, ffbea6ec, 0, b9000, 0, e3400) + b5c
 0001de8c _start   (0, 0, 0, 0, 0, 0) + 5c
------------------------------- Delimiter END ---------------------------------

And below is a gdb session where I set a breakpoint close to the point of
failure and then single stepped the code until the crash occurred.
(Continue reading)

Permalink | Reply | 2 comments |
headers
Jose Celestino | 4 Jan 22:06
Picon
Gravatar

Re: Crashes with the "indent" patch on Mutt v1.5.17

Words by Mun Johl [Fri, Jan 04, 2008 at 11:23:38AM -0800]:
> Hi all,
> 
> Quite a while ago, Jimmy Mäkelä wrote an "indent" patch for Mutt that I
> have been using for a long time.  However, when I applied it to Mutt
> v1.5.17, it caused a crash.  I've attached the patch for reference.
> Note that I compiled with gcc v3.3.2 on a Sun Solaris 8 system.
> 
> I'm hoping there may be someone out there that is familiar with the
> patch, or just willing to help with the debug.
> 
> 
> I last applied the indent-patch to mutt 1.5.15 .  I didn't try 1.5.16;
> but when applied to 1.5.17 mutt crashes when I try to reply to a
> message.  Note that there are no other patches applied.  Here is the
> backtrace from the core file:
> 
> ------------------------------ Delimiter BEGIN --------------------------------
> core 'core' of 6419:    ./mutt
>  0007f02c mutt_FormatString (ffbe8ba8, 7f, 0, 69, 2e004, 13f078) + c0
>  0002e4f4 _mutt_copy_message (e30b8, e30a8, 13ec00, 13ec60, 167, 86) + 68
>  0002ea98 mutt_copy_message (ffffffff, 127c58, 13ec00, 167, 86, b0) + 38
>  0006e860 include_reply (0, 13ec00, e30b8, e2af0, e3000, e3400) + a0
>  0006f328 generate_body (ffffffff, 1eaf88, 1, 127c58, 13ec00, ffffff8a) + 25c
>  00070ccc ci_send_message (0, 0, 0, 127c58, 1, 0) + 1344
>  0003547c mutt_index_menu (1, 0, 0, 0, 4, 0) + 3ffc
>  0004ba18 main     (1, ffbea6ec, 0, b9000, 0, e3400) + b5c
>  0001de8c _start   (0, 0, 0, 0, 0, 0) + 5c
> ------------------------------- Delimiter END ---------------------------------
>
(Continue reading)

Permalink | Reply | 1 comment |

Gmane