headers
Thomas Roessler | 1 Jul 2004 05:00

[2004-07-01] CVS repository changes

This message was generated and sent automatically.  It contains a
summary of the CVS commits over the last 48 hours.  These changes
should be propagated to the public repository within at most a day
or two.  Most probably, they have already been propagated.
Permalink | Reply | 0 comments |
headers
Mads Martin Joergensen | 1 Jul 2004 14:43
Picon

source /path/to/rc| (with pipe)

Hey Thomas,

It seems that the sanity checks added sometime in the 1.5 series to
check if it's possible to stat an rc file before proceeding did not take
care of the fact that it could be there's a pipe at the end of it.

Here's a patch against 1.5.6:

--- init.c
+++ init.c
 <at>  <at>  -1366,15 +1366,21  <at>  <at>  static int source_rc (const char *rcfile
   int line = 0, rc = 0;
   BUFFER token;
   char *linebuf = NULL;
+  char *rrcfile = strdup(rcfile);
   size_t buflen;
   pid_t pid;
   struct stat s;

-  if (stat (rcfile, &s) < 0)
+  buflen = strlen(rcfile);
+  if (rcfile[buflen-1] == '|')
+    rrcfile[buflen-1] = '\0';
+  if (stat (rrcfile, &s) < 0)
   {
-    snprintf (err->data, err->dsize, _("%s: stat: %s"), rcfile, strerror (errno));
+    snprintf (err->data, err->dsize, _("%s: stat: %s"), rrcfile, strerror (errno));
+    free(rrcfile);
     return (-1);
   }
(Continue reading)

Permalink | Reply | 2 comments |
headers
TAKAHASHI Tamotsu | 1 Jul 2004 15:27
Picon

Re: source /path/to/rc| (with pipe)

On Thu, Jul 01, 2004 at 02:43:41PM +0200, Mads Martin Joergensen wrote:

> Hey Thomas,

Although I am not Thomas,

> It seems that the sanity checks added sometime in the 1.5 series to
> check if it's possible to stat an rc file before proceeding did not take
> care of the fact that it could be there's a pipe at the end of it.

That has been fixed in CVS.
See "2004-02-07 21:36:41" entry in ChangeLog.

Thanks,
--

-- 
tamo
Permalink | Reply | 1 comment |
headers
Danielle KANDEL | 1 Jul 2004 15:32
Picon

problem


 Hello,
I wonder if you could help me:
I amm reading my mail on a server with mutt 1.2.5 and I just put my client
computer with fedora mutt 1.4
I am in France and I can't send accents and for example éèàçù
Of course I could use another mail software like evolution which comes 
with fedora but I like mutt
Thank you if you can help me . Sincerely yours
                                               Danielle Kandel
Permalink | Reply | 2 comments |
headers
Bob Bell | 1 Jul 2004 17:56
Picon

Re: limit by personal reply feature

On Wed, Jun 30, 2004 at 03:44:38PM -0700, Edward Peschko <esp5 <at> pge.com> wrote:
> I was wondering if mutt had a limit to 'personal reply' feature...  By
> that I mean I want to see all messages and threads that I either
> replied to, or initiated, bound to a key like ~N is bound to new
> messages and ~O is bound to old.  Does this feature exist? If not, how
> easy would this be to implement? ( I - and certainly others - would
> use this feature all the time )

~P || ~Q for messages (not threads) that are either addresses to you or
to which you've replied.  That's as close as it comes, I think.

> ( ps - another thing - are there programmable ~ keys? in other words,
> I'd like to map ~P to messages matching pattern 1 or 2 or 3 or 4...
> that way, I can keep track of certain users without having to remember
> all their names..)

No, but the limit command does have a history, so you can hit the "up"
key and retrieve your pattern.  You could also define a macro to enter
the pattern for you.

--

-- 
Bob Bell <bbell <at> users.sourceforge.net>
Permalink | Reply | 0 comments |
headers
Mads Martin Joergensen | 1 Jul 2004 18:35
Picon

Re: source /path/to/rc| (with pipe)

* TAKAHASHI Tamotsu <ttakah <at> lapis.plala.or.jp> [Jul 01. 2004 15:32]:
> Although I am not Thomas,
> 
> > It seems that the sanity checks added sometime in the 1.5 series to
> > check if it's possible to stat an rc file before proceeding did not take
> > care of the fact that it could be there's a pipe at the end of it.
> 
> That has been fixed in CVS.
> See "2004-02-07 21:36:41" entry in ChangeLog.

I see, thanks.

I wonder if we're ever going to see 1.6, or maybe just 1.5.7?

--

-- 
Mads Martin Joergensen, http://mmj.dk
"Why make things difficult, when it is possible to make them cryptic
 and totally illogical, with just a little bit more effort?"
                                -- A. P. J.
Permalink | Reply | 0 comments |
headers
Thomas Roessler | 2 Jul 2004 05:00

[2004-07-02] CVS repository changes

This message was generated and sent automatically.  It contains a
summary of the CVS commits over the last 48 hours.  These changes
should be propagated to the public repository within at most a day
or two.  Most probably, they have already been propagated.
Permalink | Reply | 0 comments |
headers
TAKAHASHI Tamotsu | 2 Jul 2004 09:28
Picon

Re: [PATCHES] awaiting...

On Sun, May 23, 2004 at 08:27:22PM +0900, TAKAHASHI Tamotsu wrote:
> I recently posted multiple patches.
> Please commit these changes to cvs if you don't find a bug in them.
> These really fix problems, and I have not experienced any side effect.
> I have been testing them for at least three days.

My unbind patch has been discussed and refused.
(IIRC, Dave will write better one.)
But nobody has objected to the rest two patches.
I try to explain these patches again.

> Date: Fri, 7 May 2004 16:23:57 +0900
> Subject: Re: Using the builtin editor
> 
> --- compose.c~	Tue Apr 13 09:01:33 2004
> +++ compose.c	Fri May  7 16:10:37 2004
>  <at>  <at>  -733,7 +733,7  <at>  <at> 
>  	/* fall through */
>        case OP_COMPOSE_EDIT_HEADERS:
>  	if (op == OP_COMPOSE_EDIT_HEADERS ||
> -	    (op == OP_COMPOSE_EDIT_MESSAGE && option (OPTEDITHDRS)))
> +	    (op == OP_COMPOSE_EDIT_MESSAGE && option (OPTEDITHDRS) && (mutt_strcmp ("builtin", Editor) != 0)))
>  	{
>  	  char *tag = NULL, *err = NULL;
>  	  mutt_env_to_local (msg->env);

You will be here if you do "edit-message" on compose menu.

:      case OP_COMPOSE_EDIT_MESSAGE:
:	if (Editor && (mutt_strcmp ("builtin", Editor) != 0) && !option (OPTEDITHDRS))
(Continue reading)

Permalink | Reply | 0 comments |
headers
Thomas Roessler | 2 Jul 2004 08:47

bug#1876: mutt-1.5.6i: Mutt doesn't handle invalid characters when replying to a mail

On 2004-05-31 11:56:06 +0200, Vincent Lefevre wrote:

>>  On Friday, May 21, 2004 at 3:29:32 PM +0200, Vincent Lefèvre
>>  wrote:

>>> I have a mail with an application/x-tex attachment, where the
>>> tex file has ISO-8859-1 characters but no charset
>>> information. Mutt doesn't convert the non-ASCII characters
>>> into UTF-8 when viewing the attachment in the internal pager.

How should mutt know what character set to use?  Do you suggest we
start to look for \usepackage{inputenc} lines in application/x-tex
attachment?

> This is even worse: Mutt stops the parsing at the first invalid
> character (at least when this is an ISO-8859-1 character and Mutt
> was started with UTF-8 locales), and in the case of an invalid
> "From:" header, an incorrect address is generated when replying,
> because of that.

This is a typical symptom of a badly installed iconv library.  If
you're using the iconv library that comes with libc, please make
sure all the i18n support data are installed.

--

-- 
Thomas Roessler			      <roessler <at> does-not-exist.org>
Permalink | Reply | 3 comments |
headers
Thomas Roessler | 2 Jul 2004 10:03

bug#1876: mutt-1.5.6i: Mutt doesn't handle invalid characters when replying to a mail

On 2004-07-02 09:58:29 +0200, Vincent Lefevre wrote:

> It comes from Debian's libc6 package, and I assume that it is
> complete. Are you sure that this isn't a bug in Mutt (or in the
> compat patch)?

Is the "locales" package installed?

--

-- 
Thomas Roessler			      <roessler <at> does-not-exist.org>
Permalink | Reply | 1 comment |

Gmane