Your message dated Sat, 11 Mar 2006 11:54:09 +0100
with message-id <[EMAIL PROTECTED]>
and subject line Bug#355749: exim4-base: conffile prompt for etch upgrade
has caused the attached Bug report to be marked as done.

This means that you claim that the problem has been dealt with.
If this is not the case it is now your responsibility to reopen the
Bug report if necessary, and/or fix the problem forthwith.

(NB: If you are a system administrator and have no idea what I am
talking about this indicates a serious mail system misconfiguration
somewhere.  Please contact me immediately.)

Debian bug tracking system administrator
(administrator, Debian Bugs database)

--- Begin Message ---
Package: exim4-base
Version: 4.60-4
Severity: important

The attached diff was presented to me while updating exim4 from
version 4.60-3; I didn't make the indicated changes, so I shouldn't
need to be prompted about them.

If you're not familiar with conffile prompting issues, see my message
to bug #345112 and friends.
--- /etc/init.d/exim4   2005-08-16 12:07:46.000000000 -0400
+++ /etc/init.d/exim4.dpkg-new  2006-01-28 09:56:53.000000000 -0500
@@ -13,6 +13,8 @@
   set -x
 fi
 
+ENV="env -i LANG=C PATH=$PATH"
+
 #read default file
 QUEUERUNNER='combined'
 QUEUEINTERVAL='30m'
@@ -42,8 +44,9 @@
   exit 0
 fi
 
-DAEMON=/usr/sbin/exim4
-NAME=exim4
+
+DAEMON="/usr/sbin/exim4"
+NAME="exim4"
 
 ##test -x $DAEMON || exit 0
 test -x /usr/lib/exim4/exim4 || exit 0
@@ -52,24 +55,24 @@
 {
   case ${QUEUERUNNER} in
     combined)
-      start-stop-daemon --start --pidfile "$PIDFILE" \
+      $ENV start-stop-daemon --start --pidfile "$PIDFILE" \
         --exec $DAEMON -- -bd -q${QFLAGS}${QUEUEINTERVAL} \
         ${COMMONOPTIONS} ${QUEUERUNNEROPTIONS} ${SMTPLISTENEROPTIONS}
       ;;
     separate)
-      start-stop-daemon --start --pidfile "$PIDFILE" \
+      $ENV start-stop-daemon --start --pidfile "$PIDFILE" \
         --exec $DAEMON -- -bd ${SMTPLISTENEROPTIONS} ${COMMONOPTIONS}
-      start-stop-daemon --start --pidfile "$QRPIDFILE" \
+      $ENV start-stop-daemon --start --pidfile "$QRPIDFILE" \
         --exec $DAEMON -- -oP "$QRPIDFILE" \
         -q${QFLAGS}${QUEUEINTERVAL} ${QUEUERUNNEROPTIONS} ${COMMONOPTIONS}
       ;;
     queueonly)
-      start-stop-daemon --start --pidfile "$QRPIDFILE" \
+      $ENV start-stop-daemon --start --pidfile "$QRPIDFILE" \
         --exec $DAEMON -- -oP "$QRPIDFILE" \
         -q${QFLAGS}${QUEUEINTERVAL} ${QUEUERUNNEROPTIONS} ${COMMONOPTIONS}
       ;;
     no|ppp)
-      start-stop-daemon --start --pidfile "$PIDFILE" \
+      $ENV start-stop-daemon --start --pidfile "$PIDFILE" \
         --exec $DAEMON -- -bd ${SMTPLISTENEROPTIONS}
       ;;
     nodaemon)
@@ -110,7 +113,7 @@
 {
   SIG="${1:-TERM}"
   for pid in $(pidof $NAME); do
-    if [ "$(readlink /proc/$pid/root)" == "/" ]; then
+    if [ "$(readlink /proc/$pid/root)" = "/" ]; then
       kill -$SIG $pid
     fi
   done

--- End Message ---
--- Begin Message ---
On 2006-03-07 Justin Pryzby <[EMAIL PROTECTED]> wrote:
> On Tue, Mar 07, 2006 at 06:59:39PM +0100, Andreas Metzler wrote:
[...]
>> I cannot reproduce this. Neither of
>> fresh install of 4.50.8 --> 4.60-4
>> fresh install of 4.50.8 --> 4.60-3 --> 4.60-4
>> fresh install of 4.60-3 --> 4.60-4
>> shows any unnecessary conffile prompts. Do you know which version you
>> upgraded from? Can you reproduce this bug?
> See above: 4.60-3; I just attempted to reproduce it, and failed, while
> upgrading from 4.60-3 on a different machine..  Yesterdays upgrade on
> a 3rd machine also had no such prompts.
[...]
> but the nearly-general agreement that conffile
> prompts shouldn't happen unnecessarily, and different mechanisms by
> which they can be avoided.

I perfectly agree with that.

> I'm including an interdiff of the dpkg prompt diff vs the diff between
> the conffile included in the 2 relevant versions of the package; do
> you recognize it?  This isn't change I would have made.

> diff -u /etc/init.d/exim4 /etc/init.d/exim4
> --- /etc/init.d/exim4 2006-01-28 09:56:53.000000000 -0500
> +++ /etc/init.d/exim4 2006-03-07 14:17:30.000000000 -0500
> @@ -45,7 +45,6 @@
>  fi


> -
>  DAEMON="/usr/sbin/exim4"
>  NAME="exim4"

> @@ -113,7 +112,7 @@
>  {
>    SIG="${1:-TERM}"
>    for pid in $(pidof $NAME); do
> -    if [ "$(readlink /proc/$pid/root)" = "/" ]; then
> +    if [ "$(readlink /proc/$pid/root)" == "/" ]; then
>        kill -$SIG $pid
>      fi
>    done

The /=/==/ change is between 4.53-1 and 4.54-1, but the whitespace
change is no part of that diff.

The init-script has not changed a lot:
4.50-8 up to and including 4.53-1 is identical
4.54-1 fixes the bashism (one line change s/==/=/
4.54-1 up to and including 4.60-3 is identical
4.60-4 quotes some variables and starts exim4 with LANG=C.

Indeed if I apply exim4-dpkg-conffile-diff (as supplied in the initial
bugreport) to 4.50-8's initscript I end up with a file that is almost
identical to 4.60-4's initscript
-----------------------
[EMAIL PROTECTED]:/tmp/$ diff -u exim4-4.60-4 afterapplying-supplieddiff
--- exim4-4.60-4        2006-03-11 11:12:04.727339904 +0100
+++ afterapplying-supplieddiff  2006-03-11 11:37:45.368126904 +0100
@@ -45,6 +45,7 @@
 fi


+
 DAEMON="/usr/sbin/exim4"
 NAME="exim4"

-----------------------

OTOH if I edit 4.50-8's initscript, deleting a newline ...
-----------------------
--- exim4-4.50-8        2005-05-27 09:59:31.000000000 +0200
+++ afterapplying-supplieddiff  2006-03-11 11:44:14.032041008 +0100
@@ -42,7 +42,6 @@
   exit 0
 fi

-
 DAEMON=/usr/sbin/exim4
 NAME=exim4

-----------------------
... then exim4-dpkg-conffile-diff (as supplied in the
initial bugreport) applies cleanly and I get exactly 4.60-4's
initscript.

exim4's maintainerscripts do not modify the init-script (as this would
be a serious bug.)

For these reasons I very strongly suspect that you (accidentally)
modified exim4's init-script in a editor (deleting a blank line)
somewhere from 4.50-8 to 4.54-1 and therefore simply got a correct
conffile-changed-prompt when upgrading from this version to 4.60-4.

I am therefore closing this bug-report, feel free to re-open it if yu
think this to be incorrect.
  cu andreas

-- 
The 'Galactic Cleaning' policy undertaken by Emperor Zhark is a personal
vision of the emperor's, and its inclusion in this work does not constitute
tacit approval by the author or the publisher for any such projects,
howsoever undertaken.                                (c) Jasper Ffforde

--- End Message ---

Reply via email to