Re: long 'Sending message...'

2000-12-01 Thread Suresh Ramasubramanian

Ken Weingold proclaimed on mutt-users that: 

 When mutt sits for a while on 'Sending message...', is it a mutt or
 sendmail issue?  Or something else?

Sendmail mostly - check your DNS settings (resolv.conf, nsswitch.conf and such)
- and install a caching nameserver (there's a caching-nameserver-rpm in
  redhat at least

-- 
Suresh Ramasubramanian + Wallopus Malletus Indigenensis
mallet @ cluestick.org + Lumber Cartel of India, tinlcI
The American Dental Association announced today that most plaque tends
to form on teeth around 4:00 PM in the afternoon.

Film at 11:00.



Re: long 'Sending message...'

2000-12-01 Thread Pete Wyckoff

[EMAIL PROTECTED] said:
 On Fri, Dec 01, 2000 at 12:22:57AM -0500 or thereabouts, Ken Weingold wrote:
 
  When mutt sits for a while on 'Sending message...', is it a mutt or
  sendmail issue?  Or something else?
 
 This is the same problem that I have mentioned earlier.  Many thought
 it was a sendmail problem.  However, I have made some trials with
 other MUAs, even Xemacs, and they all send normally without problems
 or waiting.  I have also gone over my sendmail rc files, and
 everything over again, and all see to be as it was previously.
 
 I can't figure it out.   

Being a control freak, I patched mutt to have it show me the sendmail
dialog.  Somehow I thought there was a configuration option which would
do this, but I could not find it, hence the ugly patch.  Maybe it will
help you uncover your mutt/sendmail interaction issue.

To use:  set sendmail_wait=-1.  It's not pretty:  the curses handling is
a mess and overloading a well defined variable is a bad idea.  Caveat
patcher.
-- Pete


diff -ruN mutt-1.3.12-stock/sendlib.c mutt-1.3.12/sendlib.c
--- mutt-1.3.12-stock/sendlib.c Tue Nov 28 11:35:05 2000
+++ mutt-1.3.12/sendlib.c   Tue Nov 28 11:36:27 2000
@@ -1740,6 +1740,7 @@
   if ((pid = fork ()) == 0)
   {
 struct sigaction act, oldalrm;
+int fdmax;
 
 /* save parent's ID before setsid() */
 ppid = getppid ();
@@ -1751,16 +1752,20 @@
   
 /* next we close all open files */
 #if defined(OPEN_MAX)
-for (fd = 0; fd  OPEN_MAX; fd++)
-  close (fd);
+fdmax = OPEN_MAX;
 #elif defined(_POSIX_OPEN_MAX)
-for (fd = 0; fd  _POSIX_OPEN_MAX; fd++)
-  close (fd);
+fdmax = _POSIX_OPEN_MAX;
 #else
-close (0);
-close (1);
-close (2);
+fdmax = 3;
 #endif
+for (fd = 0; fd  fdmax; fd++) {
+  if (fd == 1  SendmailWait == -1) continue;
+  close (fd);
+}
+
+/* for showing sendmail output, end window mode */
+if (SendmailWait == -1)
+   mutt_endwin(0);
 
 /* now the second fork() */
 if ((pid = fork ()) == 0)
@@ -1782,6 +1787,11 @@
if (dup (1)  0)
  _exit (S_ERR);
   }
+  else if (SendmailWait == -1)
+  {
+if (dup(1)  0)
+ _exit (S_ERR);
+  }
 
   execv (path, args);
   _exit (S_ERR);
@@ -1795,7 +1805,8 @@
 
 /* SendmailWait  0: interrupt waitpid() after SendmailWait seconds
  * SendmailWait = 0: wait forever
- * SendmailWait  0: don't wait
+ * SendmailWait = -1: wait forever, and send stdout/err to console
+ * SendmailWait  -1: don't wait
  */
 if (SendmailWait  0)
 {
@@ -1811,7 +1822,7 @@
   sigaction (SIGALRM, act, oldalrm);
   alarm (SendmailWait);
 }
-else if (SendmailWait  0)
+else if (SendmailWait  -1)
   _exit (0xff  EX_OK);
 
 if (waitpid (pid, st, 0)  0)
@@ -1856,6 +1867,15 @@
 st = S_ERR;/* error */
 
   mutt_unblock_signals_system (1);
+
+  /* for showing sendmail output, end window mode */
+  if (SendmailWait == -1) {
+mutt_any_key_to_continue(0);
+clear();
+cbreak();
+noecho();
+refresh();
+  }
 
   return (st);



Re: long 'Sending message...'

2000-11-30 Thread Gary

On Fri, Dec 01, 2000 at 12:22:57AM -0500 or thereabouts, Ken Weingold wrote:

 When mutt sits for a while on 'Sending message...', is it a mutt or
 sendmail issue?  Or something else?

This is the same problem that I have mentioned earlier.  Many thought
it was a sendmail problem.  However, I have made some trials with
other MUAs, even Xemacs, and they all send normally without problems
or waiting.  I have also gone over my sendmail rc files, and
everything over again, and all see to be as it was previously.

I can't figure it out.   
-- 
Best regards,
Gary