exim4-config: Debconf infinite loop

[Debian Bug Importer]

Automatically imported from Debian bug report #246742 http://bugs.debian.org/246742

[Andreas Metzler (ametzler-downhill)]

On 2004-04-30 Adam C Powell IV <email address hidden> wrote:
> Package: exim4-config
> Version: 4.31-2
> Severity: important

> While upgrading a system from woody to sarge, I simultaneously installed
> exim4, all using dselect. In the preconfigure step, exim4 asked:
> * Split configuration into multiple files? <no>
> * What type of host? <local>
> * Split configuration into multiple files? <no>
> * What type of host? <internet>
> So it switched back and forth between these two questions, until I
> answered "yes" to "Split configuration into multiple files?" then it
> proceeded normally.

Hello,
I cannot reproduce this. - Can you? If you are able to can you show me
the output when looping with 'DEBCONF_DEBUG=developer' or
DEBIAN_FRONTEND=readline /usr/share/debconf/frontend sh -x \
 /var/lib/dpkg/info/exim4-config.config configure

> Does exim4 need sarge's debconf in order to work properly?
[...]

Definititely not. I am providing (and using) woody backports which use
the identical debconf-code.
                thanks, cu andreas
--
"See, I told you they'd listen to Reason," [SPOILER] Svfurlr fnlf,
fuhggvat qbja gur juveyvat tha.
Neal Stephenson in "Snow Crash"

[hazelsct (hazelsct)]

On Sat, 2004-05-01 at 04:30, Andreas Metzler wrote:
> On 2004-04-30 Adam C Powell IV <email address hidden> wrote:
> > Package: exim4-config
> > Version: 4.31-2
> > Severity: important
>
> > While upgrading a system from woody to sarge, I simultaneously installed
> > exim4, all using dselect. In the preconfigure step, exim4 asked:
> > * Split configuration into multiple files? <no>
> > * What type of host? <local>
> > * Split configuration into multiple files? <no>
> > * What type of host? <internet>
> > So it switched back and forth between these two questions, until I
> > answered "yes" to "Split configuration into multiple files?" then it
> > proceeded normally.
>
> Hello,
> I cannot reproduce this. - Can you? If you are able to can you show me
> the output when looping with 'DEBCONF_DEBUG=developer' or
> DEBIAN_FRONTEND=readline /usr/share/debconf/frontend sh -x \
> /var/lib/dpkg/info/exim4-config.config configure

Hmm, can't reproduce it now that it's sarge...

Another data point: this was in a chroot. Could that affect things?
/proc may not have been mounted when I did the upgrade, that affected
some packages.

I'll see if this shows up on other machines, and if so, will stop and
try again with the options above.

Feel free to tag "unreproducible" and downgrade the severity :-(

Sorry,

-Adam P.

GPG fingerprint: D54D 1AEE B11C CE9B A02B C5DD 526F 01E8 564E E4B6

Welcome to the best software in the world today cafe!
http://lyre.mit.edu/~powell/The_Best_Stuff_In_The_World_Today_Cafe.ogg

[Andreas Metzler (ametzler-downhill)]

On 2004-05-02 Adam C Powell IV <email address hidden> wrote:
> On Sat, 2004-05-01 at 04:30, Andreas Metzler wrote:
> > On 2004-04-30 Adam C Powell IV <email address hidden> wrote:
> > > Package: exim4-config
> > > Version: 4.31-2
> > > Severity: important
> >
> > > While upgrading a system from woody to sarge, I simultaneously installed
> > > exim4, all using dselect. In the preconfigure step, exim4 asked:
> > > * Split configuration into multiple files? <no>
> > > * What type of host? <local>
> > > * Split configuration into multiple files? <no>
> > > * What type of host? <internet>
> > > So it switched back and forth between these two questions, until I
> > > answered "yes" to "Split configuration into multiple files?" then it
> > > proceeded normally.

> > I cannot reproduce this. - Can you? If you are able to can you show me
> > the output when looping with 'DEBCONF_DEBUG=developer' or
> > DEBIAN_FRONTEND=readline /usr/share/debconf/frontend sh -x \
> > /var/lib/dpkg/info/exim4-config.config configure

> Hmm, can't reproduce it now that it's sarge...

:-(

> Another data point: this was in a chroot. Could that affect things?
> /proc may not have been mounted when I did the upgrade, that affected
> some packages.

It is not unpossible but unlikely, imho. I am using chroots, too and
ocasionally have forgotten to mount /proc in there but have never
perceived loops. (The initscript will fail to stop the daemon, but
that's it.)

> I'll see if this shows up on other machines, and if so, will stop and
> try again with the options above.

Thanks, I'd appreciate that. It is possible that there is a hidden
error somewhere, it has been reported once before, but I was not able
to find it yet.

> Feel free to tag "unreproducible" and downgrade the severity :-(

> Sorry,

Thanks for the effort.
             cu andreas
--
"See, I told you they'd listen to Reason," [SPOILER] Svfurlr fnlf,
fuhggvat qbja gur juveyvat tha.
Neal Stephenson in "Snow Crash"

[Alexander Koch (alexander-koch-ilk)]

Hello,

I also ran in the infinitely loop with a backported exim4-daemon-light
and debconf frontend dialog. After setting 'DEBCONF_DEBUG=developer' and
DEBIAN_FRONTEND=readline all works fine :-/

<snip src=typescript>
$ sudo apt-get install exim4-daemon-light
Reading Package Lists... Done
Building Dependency Tree... Done
The following extra packages will be installed:
  exim4-base exim4-config
The following packages will be REMOVED:
  ssmtp
The following NEW packages will be installed:
  exim4-base exim4-config exim4-daemon-light
0 packages upgraded, 3 newly installed, 1 to remove and 0 not upgraded.
Need to get 0B/1264kB of archives. After unpacking 2699kB will be used.
Do you want to continue? [Y/n]
Preconfiguring packages ...
debconf (developer): starting /tmp/config.284821 configure
debconf (developer): <-- GET exim4/dc_eximconfig_configtype
debconf (developer): --> 0
debconf (developer): <-- GET exim4/use_split_config
debconf (developer): --> 0
debconf (developer): <-- SET exim4/dc_eximconfig_configtype local
delivery only; not on a network
debconf (developer): --> 0 value set
debconf (developer): <-- SET exim4/mailname daiquiri.ilk.net
debconf (developer): --> 0 value set
debconf (developer): <-- GET exim4/dc_readhost
debconf (developer): --> 0
debconf (developer): <-- GET exim4/dc_smarthost
debconf (developer): --> 0
debconf (developer): <-- GET exim4/dc_local_interfaces
debconf (developer): --> 0 notset
debconf (developer): <-- GET exim4/dc_minimaldns
debconf (developer): --> 0 false
debconf (developer): <-- SET exim4/use_split_config false
debconf (developer): --> 0 value set
debconf (developer): <-- VERSION 2.0
debconf (developer): --> 0 2.0
debconf (developer): <-- CAPB backup
debconf (developer): --> 0 multiselect
debconf (developer): <-- INPUT medium exim4/use_split_config
debconf (developer): --> 0 question will be asked
debconf (developer): <-- GO
Configuring Exim4-config
------------------------

The Debian exim4 packages can either use a single monolithic file
(/etc/exim4/exim4.conf.template) or about 40 small files in
/etc/exim4/conf.d/ to generate the final configuration.

The former is better suited for large modifications and is generally
more stable, whereas the latter offers a comfortable way to make smaller
modifications but is more fragile and might break if modified
extensively.

If you are unsure then you should not use split configuration.

Split configuration into small files? [no]

debconf (developer): --> 0 ok
debconf (developer): <-- INPUT critical exim4/dc_eximconfig_configtype
debconf (developer): --> 0 question will be asked
debconf (developer): <-- GO
Select the configuration type that best meets your needs.

Systems with dynamic IP addresses, including dialup systems, should
generally be configured to send outgoing mail to another machine, called
a "smart host" for delivery. You can choose to receive mail on such a
system; or to have no local mail delivery, except mail for root and
postmaster.

  1. internet site; mail is sent and received directly using SMTP 3.
mail sent by smarthost; no local mail 5. manually convert from
handcrafted Exim v3 configuration
  2. mail sent by smarthost; rec...

[Adrian Bunk (bunk-fs)]

severity 246742 grave
thanks

I can reproduce it starting from a woody chroot:

<-- snip -->

# apt-get --purge install exim4-daemon-light
Reading Package Lists...
Building Dependency Tree...
The following extra packages will be installed:
  exim4-base exim4-config libc6 libc6-dev libdb1-compat libdb3 libgcrypt7
  libgnutls10 libgpg-error0 liblzo1 libopencdk8 libpcre3 libtasn1-2
  linux-kernel-headers zlib1g zlib1g-dev
The following packages will be REMOVED:
  exim*
The following NEW packages will be installed:
  exim4-base exim4-config exim4-daemon-light libdb1-compat libgcrypt7
  libgnutls10 libgpg-error0 liblzo1 libopencdk8 libtasn1-2
  linux-kernel-headers
6 packages upgraded, 11 newly installed, 1 to remove and 140 not
upgraded.
Need to get 0B/11.8MB of archives. After unpacking 11.1MB will be used.
Do you want to continue? [Y/n]
Preconfiguring packages ...
head: /etc/mailname: No such file or directory

<-- snip -->

It seems to depend on the "Split configuration into small files?"
question:

<-- snip -->

debconf:
Split configuration into small files? [no]
General type of mail configuration: [local delivery only]
Split configuration into small files? [no]
General type of mail configuration: [local delivery only]
Split configuration into small files? [no]
General type of mail configuration: [local delivery only]
Split configuration into small files? [yes]
General type of mail configuration: [local delivery only]
System mail name:

<-- snip -->

Raising the severity since such a problem in the default MTA
should definitely be fixed before the release of Debian 3.1 .

cu
Adrian

--

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed

[Debian Bug Importer (debzilla)]

Automatically imported from Debian bug report #246742 http://bugs.debian.org/246742

[Debian Bug Importer (debzilla)]

Message-Id: <email address hidden>
Date: Fri, 30 Apr 2004 15:43:11 -0400
From: Adam C Powell IV <email address hidden>
To: Debian Bugs <email address hidden>
Subject: exim4-config: Debconf infinite loop

Package: exim4-config
Version: 4.31-2
Severity: important

Greetings,

While upgrading a system from woody to sarge, I simultaneously installed
exim4, all using dselect. In the preconfigure step, exim4 asked:
      * Split configuration into multiple files? <no>
      * What type of host? <local>
      * Split configuration into multiple files? <no>
      * What type of host? <internet>
So it switched back and forth between these two questions, until I
answered "yes" to "Split configuration into multiple files?" then it
proceeded normally.

Does exim4 need sarge's debconf in order to work properly? If so, it
should probably depend on a new debconf, and prevent the
preconfiguration step from happening unless debconf is new enough (e.g.
have the config script check the debconf version and exit if it's too
old, so configuration runs only in postinst after debconf is
configured).

Thanks,

-Adam P.

GPG fingerprint: D54D 1AEE B11C CE9B A02B C5DD 526F 01E8 564E E4B6

Welcome to the best software in the world today cafe!
http://lyre.mit.edu/~powell/The_Best_Stuff_In_The_World_Today_Cafe.ogg

[Debian Bug Importer (debzilla)]

Message-ID: <email address hidden>
Date: Sat, 1 May 2004 10:30:25 +0200
From: Andreas Metzler <email address hidden>
To: Adam C Powell IV <email address hidden>, <email address hidden>
Subject: Re: Bug#246742: exim4-config: Debconf infinite loop

On 2004-04-30 Adam C Powell IV <email address hidden> wrote:
> Package: exim4-config
> Version: 4.31-2
> Severity: important

> While upgrading a system from woody to sarge, I simultaneously installed
> exim4, all using dselect. In the preconfigure step, exim4 asked:
> * Split configuration into multiple files? <no>
> * What type of host? <local>
> * Split configuration into multiple files? <no>
> * What type of host? <internet>
> So it switched back and forth between these two questions, until I
> answered "yes" to "Split configuration into multiple files?" then it
> proceeded normally.

Hello,
I cannot reproduce this. - Can you? If you are able to can you show me
the output when looping with 'DEBCONF_DEBUG=developer' or
DEBIAN_FRONTEND=readline /usr/share/debconf/frontend sh -x \
 /var/lib/dpkg/info/exim4-config.config configure

> Does exim4 need sarge's debconf in order to work properly?
[...]

Definititely not. I am providing (and using) woody backports which use
the identical debconf-code.
                thanks, cu andreas
--
"See, I told you they'd listen to Reason," [SPOILER] Svfurlr fnlf,
fuhggvat qbja gur juveyvat tha.
Neal Stephenson in "Snow Crash"

[Debian Bug Importer (debzilla)]

Message-Id: <email address hidden>
Date: Sun, 02 May 2004 16:47:44 -0400
From: Adam C Powell IV <email address hidden>
To: Andreas Metzler <email address hidden>
Cc: <email address hidden>
Subject: Re: Bug#246742: exim4-config: Debconf infinite loop

On Sat, 2004-05-01 at 04:30, Andreas Metzler wrote:
> On 2004-04-30 Adam C Powell IV <email address hidden> wrote:
> > Package: exim4-config
> > Version: 4.31-2
> > Severity: important
>
> > While upgrading a system from woody to sarge, I simultaneously installed
> > exim4, all using dselect. In the preconfigure step, exim4 asked:
> > * Split configuration into multiple files? <no>
> > * What type of host? <local>
> > * Split configuration into multiple files? <no>
> > * What type of host? <internet>
> > So it switched back and forth between these two questions, until I
> > answered "yes" to "Split configuration into multiple files?" then it
> > proceeded normally.
>
> Hello,
> I cannot reproduce this. - Can you? If you are able to can you show me
> the output when looping with 'DEBCONF_DEBUG=developer' or
> DEBIAN_FRONTEND=readline /usr/share/debconf/frontend sh -x \
> /var/lib/dpkg/info/exim4-config.config configure

Hmm, can't reproduce it now that it's sarge...

Another data point: this was in a chroot. Could that affect things?
/proc may not have been mounted when I did the upgrade, that affected
some packages.

I'll see if this shows up on other machines, and if so, will stop and
try again with the options above.

Feel free to tag "unreproducible" and downgrade the severity :-(

Sorry,

-Adam P.

GPG fingerprint: D54D 1AEE B11C CE9B A02B C5DD 526F 01E8 564E E4B6

Welcome to the best software in the world today cafe!
http://lyre.mit.edu/~powell/The_Best_Stuff_In_The_World_Today_Cafe.ogg

[Debian Bug Importer (debzilla)]

Message-ID: <email address hidden>
Date: Mon, 3 May 2004 08:54:02 +0200
From: Andreas Metzler <email address hidden>
To: Adam C Powell IV <email address hidden>, <email address hidden>
Subject: Re: Bug#246742: exim4-config: Debconf infinite loop

On 2004-05-02 Adam C Powell IV <email address hidden> wrote:
> On Sat, 2004-05-01 at 04:30, Andreas Metzler wrote:
> > On 2004-04-30 Adam C Powell IV <email address hidden> wrote:
> > > Package: exim4-config
> > > Version: 4.31-2
> > > Severity: important
> >
> > > While upgrading a system from woody to sarge, I simultaneously installed
> > > exim4, all using dselect. In the preconfigure step, exim4 asked:
> > > * Split configuration into multiple files? <no>
> > > * What type of host? <local>
> > > * Split configuration into multiple files? <no>
> > > * What type of host? <internet>
> > > So it switched back and forth between these two questions, until I
> > > answered "yes" to "Split configuration into multiple files?" then it
> > > proceeded normally.

> > I cannot reproduce this. - Can you? If you are able to can you show me
> > the output when looping with 'DEBCONF_DEBUG=developer' or
> > DEBIAN_FRONTEND=readline /usr/share/debconf/frontend sh -x \
> > /var/lib/dpkg/info/exim4-config.config configure

> Hmm, can't reproduce it now that it's sarge...

:-(

> Another data point: this was in a chroot. Could that affect things?
> /proc may not have been mounted when I did the upgrade, that affected
> some packages.

It is not unpossible but unlikely, imho. I am using chroots, too and
ocasionally have forgotten to mount /proc in there but have never
perceived loops. (The initscript will fail to stop the daemon, but
that's it.)

> I'll see if this shows up on other machines, and if so, will stop and
> try again with the options above.

Thanks, I'd appreciate that. It is possible that there is a hidden
error somewhere, it has been reported once before, but I was not able
to find it yet.

> Feel free to tag "unreproducible" and downgrade the severity :-(

> Sorry,

Thanks for the effort.
             cu andreas
--
"See, I told you they'd listen to Reason," [SPOILER] Svfurlr fnlf,
fuhggvat qbja gur juveyvat tha.
Neal Stephenson in "Snow Crash"

[Debian Bug Importer (debzilla)]

Message-ID: <email address hidden>
Date: Fri, 23 Jul 2004 22:09:53 +0200
From: Alexander Koch <email address hidden>
To: <email address hidden>
Subject: Bug#246742: exim4-config: Debconf infinite loop (exim4-daemon-light 4.34-2 backported)

Hello,

I also ran in the infinitely loop with a backported exim4-daemon-light
and debconf frontend dialog. After setting 'DEBCONF_DEBUG=developer' and
DEBIAN_FRONTEND=readline all works fine :-/

<snip src=typescript>
$ sudo apt-get install exim4-daemon-light
Reading Package Lists... Done
Building Dependency Tree... Done
The following extra packages will be installed:
  exim4-base exim4-config
The following packages will be REMOVED:
  ssmtp
The following NEW packages will be installed:
  exim4-base exim4-config exim4-daemon-light
0 packages upgraded, 3 newly installed, 1 to remove and 0 not upgraded.
Need to get 0B/1264kB of archives. After unpacking 2699kB will be used.
Do you want to continue? [Y/n]
Preconfiguring packages ...
debconf (developer): starting /tmp/config.284821 configure
debconf (developer): <-- GET exim4/dc_eximconfig_configtype
debconf (developer): --> 0
debconf (developer): <-- GET exim4/use_split_config
debconf (developer): --> 0
debconf (developer): <-- SET exim4/dc_eximconfig_configtype local
delivery only; not on a network
debconf (developer): --> 0 value set
debconf (developer): <-- SET exim4/mailname daiquiri.ilk.net
debconf (developer): --> 0 value set
debconf (developer): <-- GET exim4/dc_readhost
debconf (developer): --> 0
debconf (developer): <-- GET exim4/dc_smarthost
debconf (developer): --> 0
debconf (developer): <-- GET exim4/dc_local_interfaces
debconf (developer): --> 0 notset
debconf (developer): <-- GET exim4/dc_minimaldns
debconf (developer): --> 0 false
debconf (developer): <-- SET exim4/use_split_config false
debconf (developer): --> 0 value set
debconf (developer): <-- VERSION 2.0
debconf (developer): --> 0 2.0
debconf (developer): <-- CAPB backup
debconf (developer): --> 0 multiselect
debconf (developer): <-- INPUT medium exim4/use_split_config
debconf (developer): --> 0 question will be asked
debconf (developer): <-- GO
Configuring Exim4-config
------------------------

The Debian exim4 packages can either use a single monolithic file
(/etc/exim4/exim4.conf.template) or about 40 small files in
/etc/exim4/conf.d/ to generate the final configuration.

The former is better suited for large modifications and is generally
more stable, whereas the latter offers a comfortable way to make smaller
modifications but is more fragile and might break if modified
extensively.

If you are unsure then you should not use split configuration.

Split configuration into small files? [no]

debconf (developer): --> 0 ok
debconf (developer): <-- INPUT critical exim4/dc_eximconfig_configtype
debconf (developer): --> 0 question will be asked
debconf (developer): <-- GO
Select the configuration type that best meets your needs.

Systems with dynamic IP addresses, including dialup systems, should
generally be configured to send outgoing mail to another machine, called
a "smart host" for delivery. You can choose to receive mail on such a
system; or to ha...

[Debian Bug Importer (debzilla)]

Message-ID: <email address hidden>
Date: Mon, 26 Jul 2004 00:03:56 +0200
From: Adrian Bunk <email address hidden>
To: <email address hidden>
Cc: <email address hidden>
Subject: This bug is 100% reproducible for me

severity 246742 grave
thanks

I can reproduce it starting from a woody chroot:

<-- snip -->

# apt-get --purge install exim4-daemon-light
Reading Package Lists...
Building Dependency Tree...
The following extra packages will be installed:
  exim4-base exim4-config libc6 libc6-dev libdb1-compat libdb3 libgcrypt7
  libgnutls10 libgpg-error0 liblzo1 libopencdk8 libpcre3 libtasn1-2
  linux-kernel-headers zlib1g zlib1g-dev
The following packages will be REMOVED:
  exim*
The following NEW packages will be installed:
  exim4-base exim4-config exim4-daemon-light libdb1-compat libgcrypt7
  libgnutls10 libgpg-error0 liblzo1 libopencdk8 libtasn1-2
  linux-kernel-headers
6 packages upgraded, 11 newly installed, 1 to remove and 140 not
upgraded.
Need to get 0B/11.8MB of archives. After unpacking 11.1MB will be used.
Do you want to continue? [Y/n]
Preconfiguring packages ...
head: /etc/mailname: No such file or directory

<-- snip -->

It seems to depend on the "Split configuration into small files?"
question:

<-- snip -->

debconf:
Split configuration into small files? [no]
General type of mail configuration: [local delivery only]
Split configuration into small files? [no]
General type of mail configuration: [local delivery only]
Split configuration into small files? [no]
General type of mail configuration: [local delivery only]
Split configuration into small files? [yes]
General type of mail configuration: [local delivery only]
System mail name:

<-- snip -->

Raising the severity since such a problem in the default MTA
should definitely be fixed before the release of Debian 3.1 .

cu
Adrian

--

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed

[Andreas Metzler (ametzler-downhill)]

## On 2004-07-26 Adrian Bunk <email address hidden> wrote:
## > severity 246742 grave
##
## grave: makes the package in question unusable or mostly so, or causes
## data loss, or introduces a security hole allowing access to the
## accounts of users who use the package.
##
## Does not match. There are literally thousands of succesful
## installations of exim4, so ths bug clearly does not maske the package
## unusable.
##
## I'll be nice for the moment and use "in the package maintainer's
## opinion"
##
## serious: is a severe violation of Debian policy (roughly, it violates
## a "must" or "required" directive), or, in the package maintainer's
## opinion, makes the package unsuitable for release.

severity 246742 serious

## debconf in sid and sarge isabasically identical, no need to let this
## bug influence sarge migration.

tags 246742 sarge sid

[Andreas Metzler (ametzler-downhill)]

On 2004-07-26 Adrian Bunk <email address hidden> wrote:
> severity 246742 grave
> thanks
>
> I can reproduce it starting from a woody chroot:
[...]

Can you reproduce it with DEBCONF_DEBUG=developer? Do you install
sarge or sid? Is it only reproduce if you start with exim installed?
If yes, how is exim configured?
              cu andreas
--
"See, I told you they'd listen to Reason," [SPOILER] Svfurlr fnlf,
fuhggvat qbja gur juveyvat tha.
Neal Stephenson in "Snow Crash"

[Debian Bug Importer (debzilla)]

Message-ID: <email address hidden>
Date: Mon, 26 Jul 2004 08:29:19 +0200
From: Andreas Metzler <email address hidden>
To: Adrian Bunk <email address hidden>, <email address hidden>
Subject: Re: Bug#246742: This bug is 100% reproducible for me

On 2004-07-26 Adrian Bunk <email address hidden> wrote:
> severity 246742 grave
> thanks
>
> I can reproduce it starting from a woody chroot:
[...]

Can you reproduce it with DEBCONF_DEBUG=developer? Do you install
sarge or sid? Is it only reproduce if you start with exim installed?
If yes, how is exim configured?
              cu andreas
--
"See, I told you they'd listen to Reason," [SPOILER] Svfurlr fnlf,
fuhggvat qbja gur juveyvat tha.
Neal Stephenson in "Snow Crash"

[Debian Bug Importer (debzilla)]

Message-ID: <email address hidden>
Date: Mon, 26 Jul 2004 08:18:36 +0200
From: Andreas Metzler <email address hidden>
To: <email address hidden>
Subject: Re: Bug#246742: This bug is 100% reproducible for me

## On 2004-07-26 Adrian Bunk <email address hidden> wrote:
## > severity 246742 grave
##
## grave: makes the package in question unusable or mostly so, or causes
## data loss, or introduces a security hole allowing access to the
## accounts of users who use the package.
##
## Does not match. There are literally thousands of succesful
## installations of exim4, so ths bug clearly does not maske the package
## unusable.
##
## I'll be nice for the moment and use "in the package maintainer's
## opinion"
##
## serious: is a severe violation of Debian policy (roughly, it violates
## a "must" or "required" directive), or, in the package maintainer's
## opinion, makes the package unsuitable for release.

severity 246742 serious

## debconf in sid and sarge isabasically identical, no need to let this
## bug influence sarge migration.

tags 246742 sarge sid

[Adrian Bunk (bunk-fs)]

On Mon, Jul 26, 2004 at 08:29:19AM +0200, Andreas Metzler wrote:
> On 2004-07-26 Adrian Bunk <email address hidden> wrote:
> > severity 246742 grave
> > thanks
> >
> > I can reproduce it starting from a woody chroot:
> [...]
>
> Can you reproduce it with DEBCONF_DEBUG=developer? Do you install

Yes.

> sarge or sid?

I tried sid.

> Is it only reproduce if you start with exim installed?
> If yes, how is exim configured?

I purged exim before, and the problem is still present.

> cu andreas

cu
Adrian

Full log (it asked exim4/dc_other_hostnames after I answered "yes" to
the small files question):

<-- snip -->

# apt-get --purge install exim4-daemon-light
Reading Package Lists...
Building Dependency Tree...
The following extra packages will be installed:
  exim4-base exim4-config libc6 libc6-dev libdb1-compat libdb3
libgcrypt7
  libgnutls10 libgpg-error0 liblzo1 libopencdk8 libpcre3 libtasn1-2
  linux-kernel-headers zlib1g zlib1g-dev
The following NEW packages will be installed:
  exim4-base exim4-config exim4-daemon-light libdb1-compat libgcrypt7
  libgnutls10 libgpg-error0 liblzo1 libopencdk8 libtasn1-2
  linux-kernel-headers
6 packages upgraded, 11 newly installed, 0 to remove and 138 not
upgraded.
Need to get 0B/11.8MB of archives. After unpacking 12.5MB will be used.
Do you want to continue? [Y/n]
Preconfiguring packages ...
debconf (developer): starting /tmp/config.121121 configure
debconf (developer): <-- GET exim4/dc_eximconfig_configtype
debconf (developer): --> 0
debconf (developer): <-- GET exim4/use_split_config
debconf (developer): --> 0
debconf (developer): <-- SET exim4/dc_eximconfig_configtype local
delivery only; not on a network
debconf (developer): --> 0 value set
debconf (developer): <-- GET exim4/mailname
debconf (developer): --> 0
debconf (developer): <-- SET exim4/mailname r063144.stusta.swh.mhn.de
debconf (developer): --> 0 value set
debconf (developer): <-- GET exim4/dc_readhost
debconf (developer): --> 0
debconf (developer): <-- GET exim4/dc_smarthost
debconf (developer): --> 0
debconf (developer): <-- GET exim4/dc_local_interfaces
debconf (developer): --> 0 notset
debconf (developer): <-- GET exim4/dc_minimaldns
debconf (developer): --> 0 false
debconf (developer): <-- SET exim4/use_split_config false
debconf (developer): --> 0 value set
debconf (developer): <-- VERSION 2.0
debconf (developer): --> 0 2.0
debconf (developer): <-- CAPB backup
debconf (developer): --> 0 multiselect backup
debconf (developer): <-- INPUT medium exim4/use_split_config
debconf (developer): --> 0 question will be asked
debconf (developer): <-- GO
debconf (developer): --> 0 ok
debconf (developer): <-- INPUT critical exim4/dc_eximconfig_configtype
debconf (developer): --> 0 question will be asked
debconf (developer): <-- GO
debconf (developer): --> 30 backup
debconf (developer): <-- INPUT medium exim4/use_split_config
debconf (developer): --> 0 question will be asked
debconf (developer): <-- GO
debconf (developer): --> 0 ok
debconf (developer): <-- INPUT critical exim4/dc_eximconfig_configtype
debconf (developer): --> 0 question will be asked
debconf (developer): <-- GO
debconf (developer): --> 30 backup
debc...

[Adrian Bunk (bunk-fs)]

I'm not a debconf expert, but updating debconf to the version in sid
before installing exim4 fixed this issue for me.

Could it be that you use some debconf features that are not supported be
the debconf package in woody (not to mention debconf 0.5 exim4-config
depends on)?

cu
Adrian

--

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed

[Andreas Metzler (ametzler-downhill)]

On 2004-07-26 Adrian Bunk <email address hidden> wrote:
> I'm not a debconf expert, but updating debconf to the version in sid
> before installing exim4 fixed this issue for me.

> Could it be that you use some debconf features that are not supported be
> the debconf package in woody (not to mention debconf 0.5 exim4-config
> depends on)?

Not knowingly.[1] - I have run woody on my primary computer until about
two months ago and therefore did almost all the exim development under
woody. (I still provide backports.)

Thanks for your help, I now have hope to find this bug. (Just not
today, I am too tired.)
             cu andreas
[1] I use one new feature, db_settitle, but I check whether it is
available before trying to use it.
--
"See, I told you they'd listen to Reason," [SPOILER] Svfurlr fnlf,
fuhggvat qbja gur juveyvat tha.
Neal Stephenson in "Snow Crash"

[Debian Bug Importer (debzilla)]

Message-ID: <email address hidden>
Date: Mon, 26 Jul 2004 21:16:05 +0200
From: Adrian Bunk <email address hidden>
To: Andreas Metzler <email address hidden>
Cc: <email address hidden>
Subject: Re: Bug#246742: This bug is 100% reproducible for me

On Mon, Jul 26, 2004 at 08:29:19AM +0200, Andreas Metzler wrote:
> On 2004-07-26 Adrian Bunk <email address hidden> wrote:
> > severity 246742 grave
> > thanks
> >
> > I can reproduce it starting from a woody chroot:
> [...]
>
> Can you reproduce it with DEBCONF_DEBUG=developer? Do you install

Yes.

> sarge or sid?

I tried sid.

> Is it only reproduce if you start with exim installed?
> If yes, how is exim configured?

I purged exim before, and the problem is still present.

> cu andreas

cu
Adrian

Full log (it asked exim4/dc_other_hostnames after I answered "yes" to
the small files question):

<-- snip -->

# apt-get --purge install exim4-daemon-light
Reading Package Lists...
Building Dependency Tree...
The following extra packages will be installed:
  exim4-base exim4-config libc6 libc6-dev libdb1-compat libdb3
libgcrypt7
  libgnutls10 libgpg-error0 liblzo1 libopencdk8 libpcre3 libtasn1-2
  linux-kernel-headers zlib1g zlib1g-dev
The following NEW packages will be installed:
  exim4-base exim4-config exim4-daemon-light libdb1-compat libgcrypt7
  libgnutls10 libgpg-error0 liblzo1 libopencdk8 libtasn1-2
  linux-kernel-headers
6 packages upgraded, 11 newly installed, 0 to remove and 138 not
upgraded.
Need to get 0B/11.8MB of archives. After unpacking 12.5MB will be used.
Do you want to continue? [Y/n]
Preconfiguring packages ...
debconf (developer): starting /tmp/config.121121 configure
debconf (developer): <-- GET exim4/dc_eximconfig_configtype
debconf (developer): --> 0
debconf (developer): <-- GET exim4/use_split_config
debconf (developer): --> 0
debconf (developer): <-- SET exim4/dc_eximconfig_configtype local
delivery only; not on a network
debconf (developer): --> 0 value set
debconf (developer): <-- GET exim4/mailname
debconf (developer): --> 0
debconf (developer): <-- SET exim4/mailname r063144.stusta.swh.mhn.de
debconf (developer): --> 0 value set
debconf (developer): <-- GET exim4/dc_readhost
debconf (developer): --> 0
debconf (developer): <-- GET exim4/dc_smarthost
debconf (developer): --> 0
debconf (developer): <-- GET exim4/dc_local_interfaces
debconf (developer): --> 0 notset
debconf (developer): <-- GET exim4/dc_minimaldns
debconf (developer): --> 0 false
debconf (developer): <-- SET exim4/use_split_config false
debconf (developer): --> 0 value set
debconf (developer): <-- VERSION 2.0
debconf (developer): --> 0 2.0
debconf (developer): <-- CAPB backup
debconf (developer): --> 0 multiselect backup
debconf (developer): <-- INPUT medium exim4/use_split_config
debconf (developer): --> 0 question will be asked
debconf (developer): <-- GO
debconf (developer): --> 0 ok
debconf (developer): <-- INPUT critical exim4/dc_eximconfig_configtype
debconf (developer): --> 0 question will be asked
debconf (developer): <-- GO
debconf (developer): --> 30 backup
debconf (developer): <-- INPUT medium exim4/use_split_config
debconf (developer): --> 0 question wi...

[Debian Bug Importer (debzilla)]

Message-ID: <email address hidden>
Date: Mon, 26 Jul 2004 21:25:30 +0200
From: Adrian Bunk <email address hidden>
To: Andreas Metzler <email address hidden>
Cc: <email address hidden>
Subject: Re: Bug#246742: This bug is 100% reproducible for me

I'm not a debconf expert, but updating debconf to the version in sid
before installing exim4 fixed this issue for me.

Could it be that you use some debconf features that are not supported be
the debconf package in woody (not to mention debconf 0.5 exim4-config
depends on)?

cu
Adrian

--

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed

[Debian Bug Importer (debzilla)]

Message-ID: <email address hidden>
Date: Mon, 26 Jul 2004 22:20:12 +0200
From: Andreas Metzler <email address hidden>
To: Adrian Bunk <email address hidden>, <email address hidden>
Subject: Re: Bug#246742: This bug is 100% reproducible for me

On 2004-07-26 Adrian Bunk <email address hidden> wrote:
> I'm not a debconf expert, but updating debconf to the version in sid
> before installing exim4 fixed this issue for me.

> Could it be that you use some debconf features that are not supported be
> the debconf package in woody (not to mention debconf 0.5 exim4-config
> depends on)?

Not knowingly.[1] - I have run woody on my primary computer until about
two months ago and therefore did almost all the exim development under
woody. (I still provide backports.)

Thanks for your help, I now have hope to find this bug. (Just not
today, I am too tired.)
             cu andreas
[1] I use one new feature, db_settitle, but I check whether it is
available before trying to use it.
--
"See, I told you they'd listen to Reason," [SPOILER] Svfurlr fnlf,
fuhggvat qbja gur juveyvat tha.
Neal Stephenson in "Snow Crash"

[Dan Weber (dan-mirrorlynx)]

I managed to reproduce the bug easily. Pretty much a woody chroot,
then apt-get --purge install exim4-daemon-light triggered it. From
what I have found, this only happens in dialog mode and not any of the
other modes. This is a debconf bug with woody, this bug should be
downgraded and tagged wontfix, esp. since sarge will be stable soon.
Anyway, I made a script log as Andreas had requested. It is attached,
but note, you will need a fullscreen/maximized xterm or using a tty.
To watch the motion picture use the following command:

perl -ne '$|=1;for(split(//)){select(undef,undef,undef,0.01);print}' <
typescript

(Thanks Stric)

Otherwise you can set your baud real low with getty and just cat it.

Dan

[Debian Bug Importer (debzilla)]

Message-ID: <email address hidden>
Date: Mon, 26 Jul 2004 18:58:45 -0400
From: Dan Weber <email address hidden>
To: <email address hidden>
Subject: Re: Bug#246742: This bug is 100% reproducible for me

--/WwmFnJnmDyWGHa4
Content-Type: multipart/mixed; boundary="J2SCkAp4GZ/dPZZf"
Content-Disposition: inline

--J2SCkAp4GZ/dPZZf
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline

I managed to reproduce the bug easily. Pretty much a woody chroot,
then apt-get --purge install exim4-daemon-light triggered it. From
what I have found, this only happens in dialog mode and not any of the
other modes. This is a debconf bug with woody, this bug should be
downgraded and tagged wontfix, esp. since sarge will be stable soon.
Anyway, I made a script log as Andreas had requested. It is attached,
but note, you will need a fullscreen/maximized xterm or using a tty.
To watch the motion picture use the following command:

perl -ne '$|=1;for(split(//)){select(undef,undef,undef,0.01);print}' <
typescript

(Thanks Stric)

Otherwise you can set your baud real low with getty and just cat it.

Dan

--J2SCkAp4GZ/dPZZf
Content-Type: application/octet-stream
Content-Disposition: attachment; filename="exim4-heisenbug-reproduce.log.gz"
Content-Transfer-Encoding: base64
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...

[Andreas Metzler (ametzler-downhill)]

[Debconf-infinite loop]
Thanks to Adrian Bunk and Dan Weber we now know the source of this
bug.

It is only reproducible under these circumstances:
* The installed version of debconf is old. (e.g. woody, sarge is ok.)
* The debconf frontend is set to dialog (both readline and gnome
  work).
* The config script is run by dpkg-preconfigure. (The invocation by
  postinst works.)

I glared at the respective code for a long time, inserted, echo, set
-x and other stuff, and still cannot see a bug in the script, it has
to be one the older version of debconf itself. I've tried some trivial
changes to not trigger the bug (no backup, move the first question out
of the backup-loop) but failed to get rid of the bug.

I therefore prpose this patch.
-----------------------
diff -Nur old/DEBIAN/config foo/DEBIAN/config
--- old/DEBIAN/config 2004-04-23 19:17:30.000000000 +0200
+++ foo/DEBIAN/config 2004-07-27 11:44:40.000000000 +0200
@@ -3,6 +3,13 @@

 alias stripwhitespace="sed -e 's/^[[:blank:]]*//' -e 's/[[:blank:]]*$//'"

+installeddebconfversion=`dpkg -s debconf | sed -ne '/^Version/s/^Version: //p'`
+if dpkg --compare-versions "${installeddebconfversion}" "le" "1.4" && \
+ [ "$exim4postinstisrunning" != "true" ] ; then
+ echo "Installed debconf version is broken. Aborting preconfigure."
+ exit 1
+fi
+
 . /usr/share/debconf/confmodule

 [ -n "$EX4DEBUG" ] && set -x
diff -Nur old/DEBIAN/postinst foo/DEBIAN/postinst
--- old/DEBIAN/postinst 2004-04-26 09:57:51.000000000 +0200
+++ foo/DEBIAN/postinst 2004-07-27 11:39:57.000000000 +0200
@@ -1,6 +1,7 @@
 #!/bin/sh

 set -e
+export exim4postinstisrunning=true
 . /usr/share/debconf/confmodule

 [ -n "$EX4DEBUG" ] && set -x
-----------------------

This will force the script to abort if it is running in preconfigure
mode with an old debconf, delaying configuration to postinst.
                       cu andreas
--
"See, I told you they'd listen to Reason," [SPOILER] Svfurlr fnlf,
fuhggvat qbja gur juveyvat tha.
Neal Stephenson in "Snow Crash"

[Debian Bug Importer (debzilla)]

Message-ID: <email address hidden>
Date: Tue, 27 Jul 2004 12:16:20 +0200
From: Andreas Metzler <email address hidden>
To: Adrian Bunk <email address hidden>, <email address hidden>, Alexander Koch <email address hidden>,
 Adam C Powell IV <email address hidden>, Dan Weber <email address hidden>
Subject: Re: Bug#246742: This bug is 100% reproducible for me

[Debconf-infinite loop]
Thanks to Adrian Bunk and Dan Weber we now know the source of this
bug.

It is only reproducible under these circumstances:
* The installed version of debconf is old. (e.g. woody, sarge is ok.)
* The debconf frontend is set to dialog (both readline and gnome
  work).
* The config script is run by dpkg-preconfigure. (The invocation by
  postinst works.)

I glared at the respective code for a long time, inserted, echo, set
-x and other stuff, and still cannot see a bug in the script, it has
to be one the older version of debconf itself. I've tried some trivial
changes to not trigger the bug (no backup, move the first question out
of the backup-loop) but failed to get rid of the bug.

I therefore prpose this patch.
-----------------------
diff -Nur old/DEBIAN/config foo/DEBIAN/config
--- old/DEBIAN/config 2004-04-23 19:17:30.000000000 +0200
+++ foo/DEBIAN/config 2004-07-27 11:44:40.000000000 +0200
@@ -3,6 +3,13 @@

 alias stripwhitespace="sed -e 's/^[[:blank:]]*//' -e 's/[[:blank:]]*$//'"

+installeddebconfversion=`dpkg -s debconf | sed -ne '/^Version/s/^Version: //p'`
+if dpkg --compare-versions "${installeddebconfversion}" "le" "1.4" && \
+ [ "$exim4postinstisrunning" != "true" ] ; then
+ echo "Installed debconf version is broken. Aborting preconfigure."
+ exit 1
+fi
+
 . /usr/share/debconf/confmodule

 [ -n "$EX4DEBUG" ] && set -x
diff -Nur old/DEBIAN/postinst foo/DEBIAN/postinst
--- old/DEBIAN/postinst 2004-04-26 09:57:51.000000000 +0200
+++ foo/DEBIAN/postinst 2004-07-27 11:39:57.000000000 +0200
@@ -1,6 +1,7 @@
 #!/bin/sh

 set -e
+export exim4postinstisrunning=true
 . /usr/share/debconf/confmodule

 [ -n "$EX4DEBUG" ] && set -x
-----------------------

This will force the script to abort if it is running in preconfigure
mode with an old debconf, delaying configuration to postinst.
                       cu andreas
--
"See, I told you they'd listen to Reason," [SPOILER] Svfurlr fnlf,
fuhggvat qbja gur juveyvat tha.
Neal Stephenson in "Snow Crash"

[Andreas Metzler (ametzler)]

Source: exim4
Source-Version: 4.34-4

We believe that the bug you reported is fixed in the latest version of
exim4, which is due to be installed in the Debian FTP archive:

exim4-base_4.34-4_i386.deb
  to pool/main/e/exim4/exim4-base_4.34-4_i386.deb
exim4-config_4.34-4_all.deb
  to pool/main/e/exim4/exim4-config_4.34-4_all.deb
exim4-daemon-heavy_4.34-4_i386.deb
  to pool/main/e/exim4/exim4-daemon-heavy_4.34-4_i386.deb
exim4-daemon-light_4.34-4_i386.deb
  to pool/main/e/exim4/exim4-daemon-light_4.34-4_i386.deb
exim4_4.34-4.diff.gz
  to pool/main/e/exim4/exim4_4.34-4.diff.gz
exim4_4.34-4.dsc
  to pool/main/e/exim4/exim4_4.34-4.dsc
exim4_4.34-4_all.deb
  to pool/main/e/exim4/exim4_4.34-4_all.deb
eximon4_4.34-4_i386.deb
  to pool/main/e/exim4/eximon4_4.34-4_i386.deb

A summary of the changes between this version and the previous one is
attached.

Thank you for reporting the bug, which will now be closed. If you
have further comments please address them to <email address hidden>,
and the maintainer will reopen the bug report if appropriate.

Debian distribution maintenance software
pp.
Andreas Metzler <email address hidden> (supplier of updated exim4 package)

(This message was generated automatically at their request; if you
believe that there is a problem with it please contact the archive
administrators by mailing <email address hidden>)

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Format: 1.7
Date: Tue, 27 Jul 2004 16:38:54 +0200
Source: exim4
Binary: eximon4 exim4-daemon-custom exim4-daemon-heavy exim4-base exim4 exim4-daemon-light exim4-config
Architecture: source i386 all
Version: 4.34-4
Distribution: unstable
Urgency: high
Maintainer: Exim4 Maintainers <email address hidden>
Changed-By: Andreas Metzler <email address hidden>
Description:
 exim4 - An MTA (Mail Transport Agent)
 exim4-base - EXperimental Internal Mailer -- a Mail Transport Agent
 exim4-config - Debian configuration for exim4
 exim4-daemon-heavy - Exim (v4) with extended features, including exiscan-acl
 exim4-daemon-light - Lightweight version of the Exim (v4) MTA
 eximon4 - X monitor for the Exim (v4) mail transport agent
Closes: 246742 261014
Changes:
 exim4 (4.34-4) unstable; urgency=high
 .
   * Urgency high because upgrades from woody were broken.
   * Exim4 triggers a bug in woody's debconf. - With dialog frontend, invoked
     by dpkg-preconfigure you are stuck in a loop, always being asked the same
     two questions (split config, and basic configtype) again and again until
     you give up and choose split_config=yes although being discouraged from
     doing so. I am working around this by making the config-script abort if
     debconf is old and we are running in preconfigure mode. (Thanks to Dan
     Weber, Adrian Bunk and whoever else wasted brainpower on this.)
     (Closes: #246742) (am).
   * Arabic (ar.po) translation of debconf templates by the translation team
     of Arabeyes.org (Abdulaziz Al-Arfaj). (Closes: #261014)
   * Change maintainer address to a mailinglist, add myself to uploaders (am)
   * Quote ${dc_mailname} in exim4-config.config. (am)
   * Fix grammar error in the original English templates (found by Ad...

[Debian Bug Importer (debzilla)]

Message-Id: <email address hidden>
Date: Tue, 27 Jul 2004 12:32:07 -0400
From: Andreas Metzler <email address hidden>
To: <email address hidden>
Subject: Bug#246742: fixed in exim4 4.34-4

Source: exim4
Source-Version: 4.34-4

We believe that the bug you reported is fixed in the latest version of
exim4, which is due to be installed in the Debian FTP archive:

exim4-base_4.34-4_i386.deb
  to pool/main/e/exim4/exim4-base_4.34-4_i386.deb
exim4-config_4.34-4_all.deb
  to pool/main/e/exim4/exim4-config_4.34-4_all.deb
exim4-daemon-heavy_4.34-4_i386.deb
  to pool/main/e/exim4/exim4-daemon-heavy_4.34-4_i386.deb
exim4-daemon-light_4.34-4_i386.deb
  to pool/main/e/exim4/exim4-daemon-light_4.34-4_i386.deb
exim4_4.34-4.diff.gz
  to pool/main/e/exim4/exim4_4.34-4.diff.gz
exim4_4.34-4.dsc
  to pool/main/e/exim4/exim4_4.34-4.dsc
exim4_4.34-4_all.deb
  to pool/main/e/exim4/exim4_4.34-4_all.deb
eximon4_4.34-4_i386.deb
  to pool/main/e/exim4/eximon4_4.34-4_i386.deb

A summary of the changes between this version and the previous one is
attached.

Thank you for reporting the bug, which will now be closed. If you
have further comments please address them to <email address hidden>,
and the maintainer will reopen the bug report if appropriate.

Debian distribution maintenance software
pp.
Andreas Metzler <email address hidden> (supplier of updated exim4 package)

(This message was generated automatically at their request; if you
believe that there is a problem with it please contact the archive
administrators by mailing <email address hidden>)

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Format: 1.7
Date: Tue, 27 Jul 2004 16:38:54 +0200
Source: exim4
Binary: eximon4 exim4-daemon-custom exim4-daemon-heavy exim4-base exim4 exim4-daemon-light exim4-config
Architecture: source i386 all
Version: 4.34-4
Distribution: unstable
Urgency: high
Maintainer: Exim4 Maintainers <email address hidden>
Changed-By: Andreas Metzler <email address hidden>
Description:
 exim4 - An MTA (Mail Transport Agent)
 exim4-base - EXperimental Internal Mailer -- a Mail Transport Agent
 exim4-config - Debian configuration for exim4
 exim4-daemon-heavy - Exim (v4) with extended features, including exiscan-acl
 exim4-daemon-light - Lightweight version of the Exim (v4) MTA
 eximon4 - X monitor for the Exim (v4) mail transport agent
Closes: 246742 261014
Changes:
 exim4 (4.34-4) unstable; urgency=high
 .
   * Urgency high because upgrades from woody were broken.
   * Exim4 triggers a bug in woody's debconf. - With dialog frontend, invoked
     by dpkg-preconfigure you are stuck in a loop, always being asked the same
     two questions (split config, and basic configtype) again and again until
     you give up and choose split_config=yes although being discouraged from
     doing so. I am working around this by making the config-script abort if
     debconf is old and we are running in preconfigure mode. (Thanks to Dan
     Weber, Adrian Bunk and whoever else wasted brainpower on this.)
     (Closes: #246742) (am).
   * Arabic (ar.po) translation of debconf templates by the translation team
     of Arabeyes.org (Abdulaziz Al-Arfaj). (Closes: #...

[Matt Zimmerman (mdz)]

This patch is awful. Surely there is a more robust way to fix this. Did anyone
ask joeyh?

[Fabio Massimo Di Nitto (fabbione)]

I talked with nutmeg yesterday on irc and he wanted to upload a fast fix and get
time to work on a proper patch. I will take care of this one coordinating with
him. Apparently we already spotted the problem but he had the release ready for
upload and didn't want to delay it anylonger.

[Fabio Massimo Di Nitto (fabbione)]

The problem has been investigated and all info reported to the maintainer. The
bug is located in whiptail (woody) that uses a wrong default in multiselect.
This default combined with CAPB backup causes the loop. Since there is no
win-to-win situation here I didn't provide a patch, but i rather prefer the
maintainer to choose the best solution.

[Fabio Massimo Di Nitto (fabbione)]

Fixed with exim4_4.34-5ubuntu1.

The bug has been tracked down by Joeyh in debconf <-> perl interaction that
happens only on 2.6 kernels. Since there is no way to push the change back to
woody and be sure that all the users will have the lastest one we agreed with
the maintainer that the fix he applied is the less worst, since it still ensure
the CAPB during woody->* upgrade.