Bug#1057061: The service roundcube-cleandb should depend on mariadb.service

[Dmitry Katsubo]

On 2023-11-30 00:22, Guilhem Moulin wrote:
> If the impact is only that the first run after a reboot or during a
> RDBMS restart fails, then I don't think it's worth trying to fix the
> race.  The service might fail for various other reasons, but as long as
> garbage doesn't pile up forever (i.e., as long as it doesn't *always*
> fail) this is not an issue.

I generally agree. Positive is that we have shared possible solutions if 
somebody sees the same issue.

-- 
With best regards,
Dmitry

Bug#1057061: The service roundcube-cleandb should depend on mariadb.service

[Dmitry Katsubo]

On 2023-11-29 20:17, Guilhem Moulin wrote:
> On Wed, 29 Nov 2023 at 19:48:09 +0100, Dmitry Katsubo wrote:
>> After= is not the same as Requires=
>> If the service is not present, it is just noop.
>> You might wish to add all supported RDBMS into After=.
> 
> One could also imagine systems where one (or more) of these .service
> files exists but isn't used by Roundcube.  In that case it's not a noop.
> I'll need to check what other maintainer do.

I agree, that adds more work to the scheduler, but roundcube-cleandb still 
works correctly.

>> Otherwise I have no good idea how to cure the error I got... making an 
>> explicit delay?
> 
> You can also add the After= relevant for your setup as a separate file
> override.

Right, that what I've done: added After=mariadb.service to 
roundcube-cleandb.service.

> What's the impact of this anyway? The race condition might
> cause the first run to fail but subsequent ones should succeed no?

Well, the impact is that during startup the unit is marked as failed:

# systemctl --all --failed
  UNIT  LOAD   ACTIVE SUBDESCRIPTION
● roundcube-cleandb.service loaded failed failed Purge Roundcube database: 
remove old records that were marked as deleted

For the subsequent calls I ma not sure – I've got an impression that this 
service is run only once at system startup.

-- 
With best regards,
Dmitry

Bug#1057061: The service roundcube-cleandb should depend on mariadb.service

[Dmitry Katsubo]

On 2023-11-29 02:23, Guilhem Moulin wrote:
> I don't think it's that simple: MariaDB is only one of several supported
> RDBMS, and the server might be on a different system.

After= is not the same as Requires=
If the service is not present, it is just noop.
You might wish to add all supported RDBMS into After=.

Otherwise I have no good idea how to cure the error I got... making an explicit 
delay?

TimeoutStartSec=70
ExecStartPre=/bin/sleep 60

-- 
With best regards,
Dmitry

Bug#1057061: The service roundcube-cleandb should depend on mariadb.service

[Dmitry Katsubo]

Package: roundcube-core
Version: 1.6.3+dfsg-1~deb12u1

The service roundcube-cleandb should be run after MySQL/MariaDB is started:

=== file /lib/systemd/system/roundcube-cleandb.service ===

[Unit]
After=mariadb.service

=== end ===

otherwise the following error may occur during the startup:

2023-11-29 00:12:13.677906 systemd[1]: roundcube-cleandb.service: Main process 
exited, code=exited, status=1/FAILURE
2023-11-29 00:12:13.678149 cleandb.sh[1034]: ERROR: SQLSTATE[HY000] [2002] No 
such file or directory
2023-11-29 00:12:13.680342 cleandb.sh[1034]: ERROR: Failed to connect to 
database

-- 
With best regards,
Dmitry

Bug#1053452: The packaged plugin "keyboard_shortcuts" does not work with roundcube v1.6.1

[Dmitry Katsubo]

Package: roundcube-plugins-extra
Version: 1.4.10+1-4

The plugin "keyboard_shortcuts" should be either fixed or replaced with another 
plugin.

Error log:

=== cut ===
[03-Oct-2023 17:23:13 UTC] PHP Warning:  Undefined property: rcmail::$imap in 
/usr/share/roundcube/plugins/keyboard_shortcuts/keyboard_shortcuts.php on line 
104
[03-Oct-2023 17:23:13 UTC] PHP Fatal error:  Uncaught Error: Call to undefined 
method rcmail::imap_connect() in 
/usr/share/roundcube/plugins/keyboard_shortcuts/keyboard_shortcuts.php:105
Stack trace:
#0 /usr/share/roundcube/program/lib/Roundcube/rcube_plugin_api.php(518): 
keyboard_shortcuts->html_output()
#1 /usr/share/roundcube/program/include/rcmail_output_html.php(1456): 
rcube_plugin_api->exec_hook()
#2 [internal function]: rcmail_output_html->xml_command()
#3 /usr/share/roundcube/program/include/rcmail_output_html.php(1322): 
preg_replace_callback()
#4 /usr/share/roundcube/program/include/rcmail_output_html.php(825): 
rcmail_output_html->parse_xml()
#5 /usr/share/roundcube/program/include/rcmail_output_html.php(654): 
rcmail_output_html->parse()
#6 /usr/share/roundcube/program/include/rcmail.php(296): 
rcmail_output_html->send()
#7 /usr/share/roundcube/index.php(278): rcmail->action_handler()
#8 {main}
  thrown in 
/usr/share/roundcube/plugins/keyboard_shortcuts/keyboard_shortcuts.php on line 
105

[04-Oct-2023 10:19:41 UTC] PHP Warning:  Undefined property: rcmail::$imap in 
/usr/share/roundcube/plugins/keyboard_shortcuts/keyboard_shortcuts.php on line 
104
[04-Oct-2023 10:19:41 UTC] PHP Warning:  Undefined property: rcmail::$imap in 
/usr/share/roundcube/plugins/keyboard_shortcuts/keyboard_shortcuts.php on line 
107

[04-Oct-2023 10:19:41 UTC] PHP Fatal error:  Uncaught Error: Call to a member 
function get_capability() on null in 
/usr/share/roundcube/plugins/keyboard_shortcuts/keyboard_shortcuts.php:107
Stack trace:
#0 /usr/share/roundcube/program/lib/Roundcube/rcube_plugin_api.php(518): 
keyboard_shortcuts->html_output()
#1 /usr/share/roundcube/program/include/rcmail_output_html.php(1456): 
rcube_plugin_api->exec_hook()
#2 [internal function]: rcmail_output_html->xml_command()
#3 /usr/share/roundcube/program/include/rcmail_output_html.php(1322): 
preg_replace_callback()
#4 /usr/share/roundcube/program/include/rcmail_output_html.php(825): 
rcmail_output_html->parse_xml()
#5 /usr/share/roundcube/program/include/rcmail_output_html.php(654): 
rcmail_output_html->parse()
#6 /usr/share/roundcube/program/include/rcmail.php(296): 
rcmail_output_html->send()
#7 /usr/share/roundcube/index.php(278): rcmail->action_handler()
#8 {main}
  thrown in 
/usr/share/roundcube/plugins/keyboard_shortcuts/keyboard_shortcuts.php on line 
107
=== cut ===

The following patch solves the critical issues:

=== cut ===
--- /usr/share/roundcube/plugins/keyboard_shortcuts/keyboard_shortcuts.php.orig 
   2023-10-04 12:19:19.182581433 +0200
+++ /usr/share/roundcube/plugins/keyboard_shortcuts/keyboard_shortcuts.php 
2023-10-04 12:31:59.561591110 +0200
@@ -101,12 +101,12 @@
 $c .= "  ";
 $c .= "";

-if(!is_object($rcmail->imap)){
-  $rcmail->imap_connect();
+if(!is_object($rcmail->storage)){
+  $rcmail->storage_connect();
 }
-$threading_supported = 
$rcmail->imap->get_capability('thread=references')
-  || $rcmail->imap->get_capability('thread=orderedsubject')
-  || $rcmail->imap->get_capability('thread=refs');
+$threading_supported = 
$rcmail->storage->get_capability('thread=references')
+  || $rcmail->storage->get_capability('thread=orderedsubject')
+  || $rcmail->storage->get_capability('thread=refs');

 if ($threading_supported) {
   $c .= "".$this->gettext("threads")."";
=== cut ===


-- 
With best regards,
Dmitry

Bug#1052529: proftpd: provide debug symbols for amd64 platform in sid

[Dmitry Katsubo]

On 2023-09-26 10:42, Preuße, Hilmar wrote:
> On 24.09.2023 00:38, Dmitry Katsubo wrote:
> 
> Hello Dmitry,
> 
>> Would be great to provide proftpd-*-dbgsym packages for amd64 platform. 
>> Current [1] platforms are:
>>
>> alpha hppa ia64 m68k ppc64 riscv64 sh4 sparc64 x32
>>
>> [1] https://packages.debian.org/search?keywords=proftpd
>>
> 
> I just rebuilt the proftpd packages and the debug packages were generated. 
> Further I see them on [2]. Maybe refer to [3] to learn how to get these 
> symbols.
> 
> Hilmar
> 
> [2] https://deb.debian.org/debian-debug/pool/main/p/proftpd-dfsg/
> [3] https://wiki.debian.org/HowToGetABacktrace

Thanks, I was able to install the packages. However I am stuck with how to make 
advantage of that. I would expect that the symbols are resolved directly in 
ProFTPD log:

ProFTPD killed (signal 15)
ProFTPD 1.3.8 standalone mode SHUTDOWN
ProFTPD 1.3.8 (stable) (built Sun Jul 2 2023 10:56:40 UTC) standalone mode 
STARTUP
(localhost[127.0.0.1]): -BEGIN STACK TRACE-
(localhost[127.0.0.1]): [0] /usr/lib/proftpd/mod_tls.so(+0x19ef5) 
[0x7f3b60371ef5]
(localhost[127.0.0.1]): [1] /usr/lib/proftpd/mod_tls.so(+0x19ef5) 
[0x7f3b60371ef5]
(localhost[127.0.0.1]): [2] /usr/lib/proftpd/mod_tls.so(+0x26a7e) 
[0x7f3b6037ea7e]
(localhost[127.0.0.1]): [3] proftpd: (accepting 
connections)(modules_session_init+0x57) [0x55950185bbc7]
(localhost[127.0.0.1]): [4] proftpd: (accepting connections)(+0x204ad) 
[0x55950182e4ad] <-- trying to resolve this address
(localhost[127.0.0.1]): [5] proftpd: (accepting connections)(+0x20ea1) 
[0x55950182eea1]
(localhost[127.0.0.1]): [6] proftpd: (accepting connections)(main+0x5f8) 
[0x55950182cb98]
(localhost[127.0.0.1]): [7] /lib/x86_64-linux-gnu/libc.so.6(+0x271ca) 
[0x7f3b606461ca]
(localhost[127.0.0.1]): [8] 
/lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0x85) [0x7f3b60646285]
(localhost[127.0.0.1]): [9] proftpd: (accepting connections)(_start+0x21) 
[0x55950182d1c1]
(localhost[127.0.0.1]): -END STACK TRACE-

Then I tried the advise from [4]:

# addr2line -e /usr/sbin/proftpd 0x55950182e4ad
??:0

and with gdb:

# gdb /usr/sbin/proftpd
Reading symbols from /usr/sbin/proftpd...
Reading symbols from 
/usr/lib/debug/.build-id/f2/6641fdbd8d794169752f091b130234807cdcfb.debug...
(gdb) info symbol 0x55950182e4ad
No symbol matches 0x55950182e4ad.

Any ideas? Thanks in advance!

[4] http://www.proftpd.org/docs/howto/Compiling.html#DeveloperOptions

-- 
With best regards,
Dmitry

Bug#1053107: Provide squid-dbgsym packages for amd64 platform

[Dmitry Katsubo]

Package: squid-dbgsym
Version: 6.1-2

Would be great to provide squid-dbgsym packages for amd64 platform. Current [1] 
platforms are:

alpha hppa ia64 m68k ppc64 riscv64 sh4 sparc64 x32

[1] https://packages.debian.org/sid/squid-dbgsym

-- 
With best regards,
Dmitry

Bug#1052529: proftpd: provide debug symbols for amd64 platform in sid

[Dmitry Katsubo]

Package: proftpd-core-dbgsym
Version: 1.3.8+dfsg-8
Severity: wishlist

Dear ProFTPd maintainers,

Would be great to provide proftpd-*-dbgsym packages for amd64 platform. Current 
[1] platforms are:

alpha hppa ia64 m68k ppc64 riscv64 sh4 sparc64 x32

[1] https://packages.debian.org/search?keywords=proftpd

-- 
With best regards,
Dmitry

Bug#749646: x11-xserver-utils depends on cpp

[Dmitry Katsubo]

Dear community,

Would be great to improve the dependency. What about marking cpp as optional 
package for x11-xserver-utils? Or maybe there could be another additional 
package x11-xserver-utils-nocpp which does not
depend on cpp ?

-- 
With best regards,
Dmitry

Bug#703190: closed by Patrick Matthäi (Closing already long fixed bugs)

[Dmitry Katsubo]

Hi,

Just detected that fbxine does not stop/exit by Ctrl-C. Not crucial, but I had 
to kill it via "killall fbxine" from another console.

(sorry to report that in the scope of this issue)

Bug#703190: closed by Patrick Matthäi (Closing already long fixed bugs)

[Dmitry Katsubo]

On 2022-09-21 10:14, Patrick Matthäi wrote:
> Hi,
> 
> I have uploaded here an untested backport for you for the next days: 
> https://cloud.linux-dev.org/index.php/s/jBZmioEZWRP9ebb
> 
> You have to update libxine and xine-ui

Many thanks. I confirm that xine plays video on framebuffer console just fine. 
Fantastic!

-- 
With best regards,
Dmitry

Bug#978701: wireshark: Please package version 2.6.20 with GTK support

[Dmitry Katsubo]

On 01/01/2021 15:33, Bálint Réczey wrote:
> I've pushed the new packaged upstream to the debian/buster branch on Salsa.
> 
> If you are (or anyone else is) interested please test the package on
> Buster get an approval for the upload, to follow:
> https://www.debian.org/doc/manuals/developers-reference/pkgs.html#upload-stable
> 
> (I don't have Buster setups now.)
> 
> Cheers,
> Balint

Many thanks, I was able to compile and install Wireshark v2.6.20 from that Git 
repository.

Great!

-- 
With best regards,
Dmitry

Bug#978701: wireshark: Please package version 2.6.20 with GTK support

[Dmitry Katsubo]

On 12/30/2020 17:14, Bálint Réczey wrote:
> Control: tags -1 wontfix
> 
> Hi Dmitry,
> 
> Buster-backports receives backports from testing where the version is
> already at 3.4.0-1, thus 2.6.x can't be uploaded there.
> 
> Also the 2.6.x branch reached EOL on 2020.10.18:
> https://www.wireshark.org/docs/relnotes/wireshark-2.6.20.html
> 
> It is unlikely to have a newer 2.6.x version uploaded to Buster
> because all the security issues are marked as not being important
> enough to warrant an upload:
> https://security-tracker.debian.org/tracker/source-package/wireshark

Thanks, Bálint, for looking into the issue.

There was a small hope that there will be a Christmas version of
version 2.6.20 for those who prefer GTK to QT.

I am not sure how difficult is to prepare Debian source package
for self-compilation for version 2.6.20 based on 2.6.8, probably
I would be able to do it myself.

Thanks!

Bug#978701: wireshark: Please package version 2.6.20 with GTK support

[Dmitry Katsubo]

Package: wireshark
Version: 2.6.8-1.1
Severity: wishlist

It would be nice if the latest Wireshark 2.6.x with GTK support appears in 
buster/buster-backports.

-- 
With best regards,
Dmitry

Bug#783305: closed by Xavier (Bugs not related to current version of cyrus-imapd)

[Dmitry Katsubo]

Package: cyrus-imapd
Version: 3.0.8-6+deb10u4

The same potential problem is present in one of latest Cyrus releases. It could 
be by co-incidence that saslauthd is started before Cyrus, because otherwise it 
just fails to start (requires saslauthd
socket to be present/opened).

-- 
With best regards,
Dmitry

Bug#945539: udisks2: Make it possible to use zram device for /tmp

[Dmitry Katsubo]

Package: udisks2
Version: 2.8.4-1
Severity: wishlist

It would be nice if the service script zram-setup@.service could support 
creating partitions for /tmp. I think at the end of the day that is a 
nice usecase!


I have implemented the service /etc/systemd/system/zram-setup@.service 
as follows:


[Unit]
Description=Setup zram based device %i
After=dev-%i.device
Requires=dev-%i.device
Before=local-fs.target umount.target

[Service]
Type=oneshot
RemainAfterExit=yes
EnvironmentFile=-/etc/zram.conf.d/%i
ExecStart=/bin/sh -c 'if [ -n "$ZRAM_COMPRESSION" ]; then echo 
$ZRAM_COMPRESSION > /sys/class/block/%i/comp_algorithm; fi'
ExecStart=/bin/sh -c 'if [ -n "$ZRAM_STREAMS_NUM" ]; then echo 
$ZRAM_STREAMS_NUM > /sys/class/block/%i/max_comp_streams; fi'
ExecStart=/bin/sh -c 'if [ -n "$ZRAM_DEV_SIZE" ];then echo 
$ZRAM_DEV_SIZE> /sys/class/block/%i/disksize; fi'
ExecStart=/bin/sh -c 'if [ "$SWAP" = "y" ];  then mkswap /dev/%i 
&& swapon /dev/%i; fi'
ExecStart=/bin/sh -c 'if [ "$TMP" = "y" ];   then mke2fs -q -m 0 
-b 4096 -O sparse_super -L %i /dev/%i && mount -t ext2 /dev/%i /tmp && 
chmod 1777 /tmp; fi'

ExecStop=-/bin/sh -c 'echo 1 > /sys/class/block/%i/reset'

[Install]
WantedBy=local-fs-pre.target

/etc/zram.conf.d/zram0:

ZRAM_COMPRESSION=lz4
ZRAM_DEV_SIZE=100M
SWAP=y

/etc/zram.conf.d/zram1:

ZRAM_COMPRESSION=lz4
ZRAM_DEV_SIZE=100M
TMP=y

After that comment out mounting of /tmp in fstab and:

systemctl enable zram-setup@zram0.service
systemctl enable zram-setup@zram1.service

I have encountered a problem that there is no "mode=1777" mount option 
for ext2, that is why I call "chmod 1777" explicitly after mounting it.


Any further improvements are welcomed.

--
With best regards,
Dmitry

Bug#924758: php7.0: Port to Debian Buster

[Dmitry Katsubo]

Thanks for the hint, Ondřej!

I have followed the instructions in README.txt [1] and it worked like a
charm. You've saved me a day, thanks!

On 2019-09-29 17:13, Ondřej Surý wrote: 

> Hi,
> 
> Debian always provides one PHP version in a release. If you want to
> run PHP 7.0 from official repository, you'll need to use Debian
> Stretch.
> 
> If you are willing to use external repository you can always use my
> https://deb.sury.org/ project that provides PHP 5.6, 7.0, 7.1-7.4 with
> *security* updates for Jessie,Stretch and Buster (also for Ubuntu).
> 
> I am closing this bug...
> 
> Ondřej
> --
> Ondřej Surý 
> 
>> On 29 Sep 2019, at 13:57, Sylvain Rochet  wrote:
>> 
>> Hi Dmitry,
>> 
>> I doubt your wish will be granted, PHP 7.0 upstream became end of life
>> around December 2018, thus packaging it is not relevant anymore.
>> 
>> Anyway, reality catches up with us… I need to keep PHP 5.6 for a while
>> at TuxFamily, in order to prevent breaking "a couple of" (thousands…)
>> legacy websites. Looking for similar issues drove me toward this bug
>> report, the build issues you had with PHP 7.0 looks like to be exactly
>> the same I had with PHP 5.6.
>> 
>> So you might be interested by our work to port PHP 5.6 Jessie Packages
>> to Buster:
>> 
>> https://git.tuxfamily.org/vhffs4/vhffs.git/tree/vhffs-patches/php5?id=28c40430856181b6d56694c0091995e44af8acd4
>> 
>> Sylvain

-- 
With best regards,
Dmitry 

Links:
--
[1] https://packages.sury.org/php/README.txt

Bug#637744: proftpd-mod-ldap: LDAPServer scope doesn't work

[Dmitry Katsubo]

On 2019-07-18 20:38, Hilmar Preuße wrote:
> On 10.04.17 22:27, Hilmar Preuße wrote:
>> Am 16.02.2017 um 00:53 tastete Dmitry Katsubo:
> 
> Hi Dimitry,
> 
> The proftp 1.3.6 is meanwhile available in Debian stable. Could you
> confirm that is works as expected?
> 
> Hilmar

Hello,

I confirm that v1.3.6-4 works fine with setting:

LDAPServer ldap://localhost??sub

Hence this issue is resolved.

-- 
With best regards,
Dmitry

Bug#886855: xfce4-session: verbose logging is always enabled

[Dmitry Katsubo]

I've encountered the same problem being discussed: xfce4-session v4.12.1-6 
(Debian buster) creates ~/.xfce4-session.verbose-log even though XFSM_VERBOSE 
variable is not set.

Bug#647053: Reverse dependencies of virtual package not handled correctly

[Dmitry Katsubo]

I've hit the same problem when listing dependencies for package tomcat8, which 
depends on tomcat8-common, which depends on "default-jre-headless | 
java8-runtime-headless | java8-runtime,
libtomcat8-java" but java8-runtime-headless is not shown:

# apt-rdepends --state-follow=Installed --state-show=Installed tomcat8

tomcat8
  Depends: tomcat8-common (>= 8.5.38-2)
...
tomcat8-common
  Depends: libtomcat8-java (>= 8.5.38-2)


# apt-cache showpkg java8-runtime-headless
...
Reverse Provides:
openjdk-8-jre-headless 8u212-b01-1~deb9u1 (= )
openjdk-11-jre-headless 11.0.3+1-1~bpo9+1 (= )
openjdk-8-jre-headless 8u181-b13-2~deb9u1 (= )
default-jre-headless 2:1.8-58 (= )
openjdk-11-jre-headless 11.0.3+1-1 (= )
default-jre-headless 2:1.11-71 (= )

-- 
With best regards,
Dmitry

Bug#924758: php7.0: Port to Debian Buster

[Dmitry Katsubo]

Package: php7.0
Version: 7.0.33-0+deb9u3
Severity: wishlist

Dear PHP maintainers,

There is a demand to run both PHP 7.0 and 7.3 on the same Debian server to 
allow transition of applications. Most of the applications are PHP 7.1 ready 
and sort of PHP 7.2 friendly, but 7.3 is really
a big jump.

Co-existence of these versions is almost possible, however I faced a problem 
with php7.0-curl which depends on libcurl3 which on some reason cannot be 
installed next to libcurl4 which is needed by
php7.3-curl (if you know the reason, please share).

So I have decided to recompile PHP for Debian Buster. In it almost worked 
except the following:

* libicu57 instead of libicu63 should be installed. This is because ./configure 
relies on /bin/icu-config which is not present in libicu-dev_63. Would be nice 
if PHP 7.0 can be linked against libicu63.
* On some reason autodetection of completion_matches() presence in readline.h 
does not work and HAVE_RL_COMPLETION_MATCHES stays undefined. This causes the 
following error:

/home/packages/php/php7.0-7.0.33/ext/readline/readline.c:34:31: error: 
conflicting types for ‘completion_matches’
 #define rl_completion_matches completion_matches
   ^~
In file included from 
/home/packages/php/php7.0-7.0.33/ext/readline/readline.c:38:
/usr/include/editline/readline.h:183:15: note: previous declaration of 
‘completion_matches’ was here
 char**completion_matches(/* const */ char *, rl_compentry_func_t *);
   ^~
make[2]: *** [Makefile:1663: ext/readline/readline.lo] Error 1
make[2]: Leaving directory '/home/packages/php/php7.0-7.0.33/ext-build'

I don't know how to fix that correctly.

I wonder if somebody could port PHP 7.0 to Debian Buster. Many thanks!

-- 
With best regards,
Dmitry

Bug#924679: guacamole: Does not start on Tomcat8 with message "Error deploying configuration descriptor"

[Dmitry Katsubo]

Package: guacamole
Version: 0.9.9+dfsg-1

After update from Stretch (Tomcat 8.5.14-1+deb9u3) to Buster (Tomcat 8.5.38-2) 
Guacamole does not deploy anymore:

[2019-03-14 23:42:08] [info] 14-Mar-2019 23:42:05.234 SEVERE 
[localhost-startStop-1] org.apache.catalina.startup.HostConfig.deployDescriptor 
Error deploying configuration descriptor 
[/etc/tomcat8/Catalina/localhost/guacamole.xml]
[2019-03-14 23:42:08] [info]  java.lang.IllegalStateException: 
ContainerBase.addChild: start: org.apache.catalina.LifecycleException: Failed 
to start component 
[StandardEngine[Catalina].StandardHost[localhost].StandardContext[/guacamole]][2019-03-14
 23:42:08] [info] #011at 
org.apache.catalina.core.ContainerBase.addChildInternal(ContainerBase.java:758)
[2019-03-14 23:42:08] [info] #011at 
org.apache.catalina.core.ContainerBase.addChild(ContainerBase.java:730)
[2019-03-14 23:42:08] [info] #011at 
org.apache.catalina.core.StandardHost.addChild(StandardHost.java:734)
[2019-03-14 23:42:08] [info] #011at 
org.apache.catalina.startup.HostConfig.deployDescriptor(HostConfig.java:629)
[2019-03-14 23:42:08] [info] #011at 
org.apache.catalina.startup.HostConfig$DeployDescriptor.run(HostConfig.java:1839)
[2019-03-14 23:42:08] [info] #011at 
java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
[2019-03-14 23:42:08] [info] #011at 
java.util.concurrent.FutureTask.run(FutureTask.java:266)
[2019-03-14 23:42:08] [info] #011at 
java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
...
[2019-03-14 23:42:08] [info] 14-Mar-2019 23:42:05.241 INFO 
[localhost-startStop-1] org.apache.catalina.startup.HostConfig.deployDescriptor 
Deployment of configuration descriptor 
[/etc/tomcat8/Catalina/localhost/guacamole.xml] has finished in [690] ms

The problem reports I found ([1], [2]) suggest updating to Guacamole v0.9.14 or 
later (see bug #887465).

[1] 
http://mail-archives.apache.org/mod_mbox/guacamole-user/201808.mbox/%3cCALKeL-Pmqh7VvF=jdw02xcqd3axuta-qachistmosum4mcw...@mail.gmail.com%3e
[2] https://www.mail-archive.com/user@guacamole.apache.org/msg01168.html

-- 
With best regards,
Dmitry

Bug#913342: libapache2-mod-svn: Lower the severity of the problem from ERROR to WARN in case if config file does not point to one provided by the package

[Dmitry Katsubo]

Package: libapache2-mod-svn
Version: 1.9.5-1+deb9u2
Severity: minor

Dear package maintainers,

In my case I am using the "custom" configuration file of dav_svn module 
which is

located here:

/etc/apache2/mods-available/dav_svn.conf.my

I have reconfigured the symlink so that it points to that file:

/etc/apache2/mods-enabled/dav_svn.conf --> 
../mods-available/dav_svn.conf.my


With above configuration when libapache2-mod-svn package is updated, the
setting-up step fails:

Setting up apache2 (2.4.25-3+deb9u5) ...
Installing new version of config file /etc/init.d/apache-htcacheclean 
...

info: Executing deferred 'a2enconf apache2-doc' for package apache2-doc
info: Executing deferred 'a2enmod authz_svn' for package 
libapache2-mod-svn
ERROR: Config file dav_svn.conf not properly enabled: 
/etc/apache2/mods-enabled/dav_svn.conf exists but does not point to 
/etc/apache2/mods-available/dav_svn.conf, not touching it

ERROR: Could not enable dependency dav_svn for authz_svn, aborting
dpkg: error processing package apache2 (--configure):
 subprocess installed post-installation script returned error exit 
status 1


I suggest the check that symlink points to "right" location becomes more 
relaxed
and does not fail the package upgrade. For example, it could be just a 
warning.


--
With best regards,
Dmitry

Bug#886344: guacamole: Package alternative authentication modules (guacamole-auth-cas, guacamole-auth-duo, guacamole-auth-jdbc, guacamole-auth-header, guacamole-auth-ldap)

[Dmitry Katsubo]

Package: guacamole
Version: 0.9.9+dfsg-1
Severity: minor

Please package alternative authentication modules:

* guacamole-auth-cas
* guacamole-auth-duo
* guacamole-auth-jdbc
* guacamole-auth-header
* guacamole-auth-ldap
* ...

-- 
With best regards,
Dmitry

Bug#637744: proftpd-mod-ldap: LDAPServer scope doesn't work

[Dmitry Katsubo]

On 2017-02-15 22:37, Hilmar Preuße wrote:
> On 14.08.2011 03:02, Dmitry Katsubo wrote:
> 
> Hi Dimitry,
> 
>> The search scope is always base (scope=0):
>>
>> slapd: SRCH base="cn=persons,cn=server" scope=0 deref=0
>> filter="(uid=dmitry)"
>>
>> regardless the LDAPServer setting:
>>
>> LDAPServer ldap://localhost??sub
>>
>> The workaround is to use LDAPServer+LDAPSearchScope separately:
>>
>> LDAPServer localhost
>> LDAPSearchScope subtree
>>
>> This bug is in some sense the opposite bug for:
>>
>> http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=500731
>>
> I've noticed that http://bugs.proftpd.org/show_bug.cgi?id=4289 has been
> reported @upstream. AFAICT this is a different one, right?
> 
> Is your issue still reproducible?

Hi,

I think this is about the same matter. If I remember correctly, I have
originally using "LDAPSearchScope subtree", which at some moment was broken
(bug#500731), and then I tried "LDAPServer ldap://localhost??sub; which also
didn't work, but after updating to v1.3.2 it worked fine. Then I believe it was
broken again (bug#637744).

Maybe I miss something (quite some time had passed), however I confirm that
"LDAPServer ldap://localhost??sub; works for me since v1.3.2 up to v1.3.5a-1
which I am using now.

Do you want to focus on "LDAPSearchScope subtree" issue at the moment?
Well, I quickly tried

LDAPServer localhost

LDAPSearchScope subtree 


and it seems to work as expected. I cannot be 100% sure, but I'll let one
know in bugzilla if I find any problem.

-- 
With best regards,
Dmitry

Bug#774164: libocrad-dev: libocrad.a contains non-reallocatable code

[Dmitry Katsubo]

On 2016-12-16 10:16, Alexander Alemayhu wrote:
> On Thu, Dec 15, 2016 at 09:42:08PM +0100, Petter Reinholdtsen wrote:
>>
>> Personally I use tesseract these days for my OCR work, and do not need ocrad.
> 
> I have unfortunately stopped using ocrad in favour of proprietary
> solutions.
> 
> Thanks for adding me to the recipient list, but I don't have the
> motivation to help here.
> 
> Added Jakub Wilk to copy, IIRC he made the package.

I see from the build log [1] that the build policy for .a had changed... I think
that is for good. Relocatable code makes object a bit bigger, but on current
systems that does not matter at all. Briefly:

1. One solution would be to make .a code reallocatable, so that .so library
that refers libocrad.a can compile.

2. Another solution is to provide libocrad.so within libocrad package.

[1] 
https://ci.debian.net/data/packages/unstable/amd64/o/ocrad/20161215_073147.autopkgtest.log.gz

-- 
With best regards,
Dmitry

Bug#741487: mozjpeg 2.1 package

[Dmitry Katsubo]

On 2016-11-21 23:45, Pierre Rudloff wrote:
By default mozjpeg is installed in /opt/mozjpeg/. You don't even need 
to use LD_PRELOAD as the binary it installs is automatically linked to 
/opt/mozjpeg/lib64/libjpeg.so.


(The conflict issue I was raising is only if want to include mozjpeg in 
Debian and install it in /usr/.)


Actually I just thought that utility tools in bin/ could be statically 
linked like in the example below:


root@debian:~# ./configure --disable-shared; make clean install
root@debian:~# ldd /opt/mozjpeg/bin/cjpeg
linux-gate.so.1 (0xb7768000)
libpng12.so.0 => /lib/i386-linux-gnu/libpng12.so.0 (0xb7728000)
libm.so.6 => /lib/i386-linux-gnu/i686/cmov/libm.so.6 
(0xb76e2000)
libc.so.6 => /lib/i386-linux-gnu/i686/cmov/libc.so.6 
(0xb7537000)

libz.so.1 => /lib/i386-linux-gnu/libz.so.1 (0xb751a000)
/lib/ld-linux.so.2 (0x8002f000)

Now one can package _only_ binaries and omit the (conflicting) library. 
Wouldn't this be a solution to finally allow mozjpeg into Debian repo?


--
With best regards,
Dmitry

Bug#741487: mozjpeg 2.1 package

[Dmitry Katsubo]

On 2016-11-21 20:48, Pierre Rudloff wrote:
>> Hello Pierre,
>>
>> Could you please share the mozjpeg Debian package sources (with 
>> debian/control,
>> ... files)? For example as a branch in github? That will simplify creating a
>> package to play with for me.
>>
>> Also I wonder what exactly the conflict with the standard libjpeg leads to. 
>> Does
>> it mean that if mozjpeg is installed to /usr/local subtree it will interfere
>> with software that uses libjpeg?
>>
>> Many thanks in advance!
>>
> Hello,
> 
> Sorry, I don't have the source anymore (and it has been removed from mentors) 
> but it is pretty easy to build.
> 
> IIRC mozjpeg installs its own libjpeg.so library so it can't be installed 
> alongside upstream libjpeg.

Many thanks for the information!

The question is: if I need to install mozjpeg together with libjpeg, what are 
the
options? Do I need to change the prefix to /usr/local for mozjpeg, and then use

$ LD_PRELOAD=/usr/local/lib/libjpeg.so jpegtran -copy all  

? How have you played with it?

Also if I may ask you, what options are meant to be for maximum compression? I 
see
that "-optimize" and "-arithmetic" are mutually exclusive... I would assume that
compression should have a mode "try all possible".

P.S. Just discovered that non of modern browsers support arithmetic-encoded 
JPEGs

https://bugzilla.mozilla.org/show_bug.cgi?id=680385

which makes "-arithmetic" option non-attractive.

-- 
With best regards,
Dmitry

Bug#741487: mozjpeg 2.1 package

[Dmitry Katsubo]

> From: Pierre Rudloff 
> To: 741...@bugs.debian.org
> Subject: mozjpeg 2.1 package
> Date: Sun, 14 Sep 2014 02:49:22 +0200
> 
> Hello,
> 
> I have built a package for mozjpeg 2.1:
> http://mentors.debian.net/package/mozjpeg
> 
> But, as I said, it conflicts with the standard libjpeg.
> 
> Regards,

Hello Pierre,

Could you please share the mozjpeg Debian package sources (with debian/control,
... files)? For example as a branch in github? That will simplify creating a
package to play with for me.

Also I wonder what exactly the conflict with the standard libjpeg leads to. Does
it mean that if mozjpeg is installed to /usr/local subtree it will interfere
with software that uses libjpeg?

Many thanks in advance!

-- 
With best regards,
Dmitry

Bug#840154: genisoimage: dirsplit displays the reference to array and not array contents (as supposed)

[Dmitry Katsubo]

Package: genisoimage
Version: 9:1.1.11-3

The script /usr/bin/dirsplit attempts to output the contents of the array in
this case:

$ dirsplit -s47724M -e2 -l /cd
Building file list, please wait...
Too large object(s) (118714487893) for the given max size: ARRAY(0xa28d0fc) 
(maybe coalesced in arrays, check manually)

To print values I think the easiest way is to use Data::Dumper.

-- 
With best regards,
Dmitry
--- /usr/bin/dirsplit   2006-11-26 00:13:29.0 +0100
+++ dirsplit2016-10-08 23:46:24.905411610 +0200
@@ -15,6 +15,9 @@
 use File::Basename;
 use File::Path;
 use Cwd 'abs_path';
+use Data::Dumper;
+
+$Data::Dumper::Indent = 0;
 
 my $ret=0;
 my $max="4488M";
@@ -118,7 +121,7 @@
 # check for pointless requests
 my $testsize=0;
 for(@sizes) {
-   die "Too large object(s) ($_) for the given max size: @{$names{$_}} (maybe 
coalesced in arrays, check manually)\n" if($_>$max);
+   die "Too large object(s) ($_) for the given max size: " . 
Data::Dumper->Dump($names{$_}, ["names"]) . " (maybe coalesced in arrays, check 
manually)\n" if($_>$max);
 
$testsize+=$_;
 }

Bug#834858: php-common: Process with given PID might have be terminated, hence causing an error

[Dmitry Katsubo]

Package: php-common
Version: 1:41

The script /usr/lib/php/sessionclean in line 48 iterates through PIDs which
might terminate in meanwhile. As a result the system administrator receives the
following error mail from cron:

find: ‘/proc/23493/fd’: No such file or directory

Perhaps this situation can be ignored:

find "/proc/$pid/fd" -ignore_readdir_race ... 2>/dev/null

-- 
With best regards,
Dmitry

Bug#827423: procps: Failure in systemd-sysctl.service causes procps update to fail

[Dmitry Katsubo]

Package: procps
Version: 2:3.3.11-3

Scenario:

# apt-get upgrade
Setting up procps (2:3.3.11-3) ...
Job for systemd-sysctl.service failed because the control process exited with 
error code.
See "systemctl status systemd-sysctl.service" and "journalctl -xe" for details.
invoke-rc.d: initscript procps, action "start" failed.
dpkg: error processing package procps (--configure):
 subprocess installed post-installation script returned error exit status 1

Problem:

# journalctl -xe
systemd[1]: Starting Apply Kernel Variables...
systemd-sysctl[11882]: Couldn't write ':windows:M::MZ::/usr/bin/wine:' to 
'fs/binfmt_misc/register', ignoring: Invalid argument
systemd[1]: systemd-sysctl.service: Main process exited, code=exited, 
status=1/FAILURE
systemd[1]: Failed to start Apply Kernel Variables.

In this particular case (as far as I can see) binfmt_misc module rejects to
register a handler if there is already one (which was registered at system
startup). As a result a lot of the packages could not be configured, because
they depend on procps.

I would expect that procps performs systemd-sysctl.service restart, but
does not stop package installation if it fails.

-- 
With best regards,
Dmitry

Bug#827402: udev-230-2: /var/lib/dpkg/info/udev.prerm called with unknown argument 'deconfigure'

[Dmitry Katsubo]

Package: udev
Version: 230-2

I have the problem when doing "apt-get dist-upgrade":

dpkg: considering deconfiguration of ifupdown, which would be broken by 
installation of systemd ...
dpkg: yes, will deconfigure ifupdown (broken by systemd)
dpkg: considering deconfiguration of udev, which would be broken by 
installation of systemd ...
dpkg: yes, will deconfigure udev (broken by systemd)
Preparing to unpack .../systemd_230-2_i386.deb ...
De-configuring udev (228-2) ...
/var/lib/dpkg/info/udev.prerm called with unknown argument 'deconfigure'
dpkg: error processing archive /var/cache/apt/archives/systemd_230-2_i386.deb 
(--unpack):
 subprocess installed pre-removal script returned error exit status 1
addgroup: The group `systemd-journal' already exists as a system group. Exiting.
dpkg: considering deconfiguration of systemd, which would be broken by 
installation of ifupdown ...
dpkg: yes, will deconfigure systemd (broken by ifupdown)
Preparing to unpack .../ifupdown_0.8.13_i386.deb ...
De-configuring systemd (228-2) ...
Unpacking ifupdown (0.8.13) over (0.7.54) ...
Replacing files in old package systemd (228-2) ...
Preparing to unpack .../archives/udev_230-2_i386.deb ...
Unpacking udev (230-2) over (228-2) ...
...
Errors were encountered while processing:
 /var/cache/apt/archives/systemd_230-2_i386.deb
E: Sub-process /usr/bin/dpkg returned an error code (1)

Bug #809744 and #809743 read that the problem is fixed in v228-3, but I am
installing 230-2. I was able to recover/continue using "apt-get -f install".

-- 
With best regards,
Dmitry

Bug#825783: cyrus-imapd-2.4: Cyrus does not create directory in /var/run if it differs from default one (/var/run/cyrus)

[Dmitry Katsubo]

Thanks for reply Ondřej,

After analysing /usr/lib/cyrus/bin/init-helper I see what is the problem.
>From one side everything is implemented correctly: once I use the default
imap.conf, the script should also work OK (and it indeed creates
/var/run/cyrus/socket and other stuff). However I was not 100% exact with
problem description -- sorry for that. The actual value is:

lmtpsocket: /var/run/postfix/cyrus/lmtp

I used different from default once location because later /var/run/postfix
is bound to /var/spool/postfix/var/run (to make it possible for chrooted
postfix to communicate with cyrus).

That is why I re-formulate the problem:

Cyrus does not create directory in /var/run if it differs from default
one (/var/run/cyrus).

Expected that the actual values from configuration file /etc/imapd.conf are
considered in function fixdirs().

On 2016-05-30 09:46, Ondřej Surý wrote:
> Hi Dmitry,
> 
> each init script runs:
> 
> /usr/sbin/cyrus init-helper start
> 
> that creates /var/run/cyrus and /var/run/cyrus/socket directories.
> 
> Could you run:
> 
> bash -x /usr/lib/cyrus/bin/init-helper start
> 
> on your system to see what went wrong?
> 
> Cheers,
>
> -- Ondřej Surý <ond...@sury.org> Knot DNS (https://www.knot-dns.cz/)
>>> On Sun, May 29, 2016, at 21:43, Dmitry Katsubo wrote:
>>> Package: cyrus-imapd-2.4
>>> Version: 2.4.17+nocaldav-2+b1
>>> 
>>> Hello,
>>> 
>>> I have noticed that Cyrus does not create necessary directory in /var/run
>>> when needed. For example, when the following is specified in
>>> /etc/imapd.conf
>>> 
>>> lmtpsocket: /var/run/cyrus/lmtp
>>> 
>>> the directory /var/run/cyrus is not created, even though Cyrus starts up
>>> and functions ok.
>>> 
>>> Expected: the directory /var/run/cyrus (or any other necessary top-level
>>> directory) is created or Cyrus complains that directory is not writeable.
>>> 
>>> Solution: Do it in init.d script (very draft suggestion is attached) or
>>> via /etc/tmpfiles.d/cyrus.conf:
>>> 
>>> d /var/run/cyrus 0755 cyrus mail

-- 
With best regards,
Dmitry

Bug#825783: cyrus-imapd-2.4: Cyrus does not create directory in /var/run

[Dmitry Katsubo]

Package: cyrus-imapd-2.4
Version: 2.4.17+nocaldav-2+b1

Hello,

I have noticed that Cyrus does not create necessary directory in /var/run
when needed. For example, when the following is specified in /etc/imapd.conf

lmtpsocket: /var/run/cyrus/lmtp

the directory /var/run/cyrus is not created, even though Cyrus starts up and
functions ok.

Expected: the directory /var/run/cyrus (or any other necessary top-level
directory) is created or Cyrus complains that directory is not writeable.

Solution: Do it in init.d script (very draft suggestion is attached) or via
/etc/tmpfiles.d/cyrus.conf:

d /var/run/cyrus 0755 cyrus mail

-- 
With best regards,
Dmitry
--- /etc/init.d/cyrus-imapd.orig2015-03-24 12:10:20.0 +0100
+++ /etc/init.d/cyrus-imapd 2016-05-29 21:31:56.947282286 +0200
@@ -61,6 +61,14 @@
 
 SYNC_CLIENT=/usr/lib/cyrus/bin/sync_client
 SYNCSHUTDOWN="$(gawk '/^sync_shutdown_file:[[:blank:]]/ { print $2 }' $CONF)"
+SOCKET="$(gawk '/^lmtpsocket:[[:blank:]]/ { print $2 }' $CONF)"
+
+if [ ! -z "$SOCKET" ]
+then
+   SOCKET_DIR=`dirname $SOCKET`
+   [ -d "$SOCKET_DIR" ] || mkdir -p "$SOCKET_DIR"
+   chown cyrus.mail "$SOCKET_DIR"
+fi
 
 # Load the VERBOSE setting and other rcS variables
 . /lib/init/vars.sh

Bug#812128: motion: Incorrect ownership for /var/lib/motion directory

[Dmitry Katsubo]

On 2016-01-22 16:49, Ximin Luo wrote:
> Try this: uninstall motion, then `sudo deluser motion`. This should
> also automatically delete the motion group as well. If it doesn't
> also run `sudo deluser --group motion`.
> 
> Then when you reinstall motion, that directory should be created by
> `adduser` when it creates the motion user again. (It is created
> motion:motion; if you want motion:adm instead you should probably
> edit /etc/adduser.conf, see `man adduser.conf` for more details...)
> 
> I think what happened to you was that you purged motion (or otherwise
> removed /var/lib/motion) before. Am I right? The previous purge
> behaviour removed /var/lib/motion, which I fixed here: [1] but I
> didn't think about what would happen to users that had already done
> this.
> 
> To fix this, I will add some logic to the postinst to re-create the
> motion user's $HOME if it doesn't already exist *and* it's mentioned
> in motion.conf. This condition should allow us to avoid interfering
> where the sysadmin removed /var/lib/motion on purpose. Let me know if
> you think that will cause problems.
> 
> X
> 
> [1] 
> https://anonscm.debian.org/cgit/users/infinity0/motion.git/commit/?id=7a2c309b65400a24aa68bc502ff94f783e9536cb

I confirm that after removing the package and user as you have described,
and installation of the package afterwards, /var/lib/motion has appeared.

P.S. Is there any /etc/logrotate.d/motion script that should take care of
/var/log/motion/motion.log ?

-- 
With best regards,
Dmitry

Bug#812128: motion: Incorrect ownership for /var/lib/motion directory

[Dmitry Katsubo]

On 2016-01-21 00:42, Ximin Luo wrote:
> Hi Dmitry, I believe I fixed this just yesterday with 3.2.12+git20140228-8. 
> Can you try it and confirm?

I have downloaded the package from sid. Then I did:

# rm -R /var/lib/motion
# dpkg -i motion_3.2.12+git20140228-8_i386.deb 
Unpacking motion (3.2.12+git20140228-8) over (3.2.12+git20140228-7) ...
Setting up motion (3.2.12+git20140228-8) ...
Processing triggers for systemd (228-2) ...
Processing triggers for man-db (2.7.5-1) ...
# ls -ld /var/lib/motion
ls: cannot access /var/lib/motion: No such file or directory
# /etc/init.d/motion start
Starting motion (via systemctl): motion.service.
# ls -ld /var/lib/motion
ls: cannot access /var/lib/motion: No such file or directory

I am not sure at what moment it should appear...

-- 
With best regards,
Dmitry

Bug#812128: motion: Incorrect ownership for /var/lib/motion directory

[Dmitry Katsubo]

Package: motion
Version: 3.2.12+git20140228-7
Severity: minor

When motion daemon is first started, it cannot write to /var/lib/motion
directory. Expected: this directory is created with motion:adm ownership
and 750 permissions during the package installation.

[1] [ERR] [ALL] [Jan 20 21:30:53] put_picture: Can't write picture to file 
/var/lib/motion/01-20160120213053-01.jpg - check access rights to target 
directory
Thread is going to finish due to this fatal error: Permission denied
[1] [NTC] [ALL] [Jan 20 21:30:54] motion_loop: Thread exiting

-- 
With best regards,
Dmitry

Bug#810287: fail2ban: Error in jail.conf configuration file in [roundcube-auth] section

[Dmitry Katsubo]

On 2016-01-08 01:08, Yaroslav Halchenko wrote:

Thanks!

it was fixed upstream awhile back in
d278fbca306d8bdcc5b3ffe34b1cfc3cd8963f0b
I guess we should cook up a new upstream release (0.9.4) and update 
package in

Debian

Cheers


Hi Yaroslav,

If you have a chance, could you look into how systemd logs the error 
message in case when fail2ban server fails to start? It was really a 
challenge to find out what went wrong.


Thanks.

--
With best regards,
Dmitry

Bug#810287: fail2ban: Error in jail.conf configuration file in [roundcube-auth] section

[Dmitry Katsubo]

Package: fail2ban
Version: 0.9.3-1
Severity: minor

The section [roundcube-auth] contains the error for log definition. The line

logpath  = logpath = %(roundcube_errors_log)s

should read as

logpath = %(roundcube_errors_log)s

otherwise it triggers an error:

# /usr/bin/fail2ban-client -x start
ERROR  No file(s) found for glob logpath =
ERROR  Failed during configuration: Have not found any log file for
roundcube-auth jail

Unfortunately, this error trace is not shown in systemd output, which
makes it difficult to resolve the problem using systemctl:

# systemctl status fail2ban.service
..
systemd[1]: fail2ban.service: Service hold-off time over, scheduling
restart.
systemd[1]: Stopped Fail2Ban Service.
systemd[1]: fail2ban.service: Start request repeated too quickly.
systemd[1]: Failed to start Fail2Ban Service.
systemd[1]: fail2ban.service: Unit entered failed state.
systemd[1]: fail2ban.service: Failed with result 'start-limit'.


-- 
With best regards,
Dmitry

Bug#809769: roundcube: Error log is filled with messages that config file cannot be loaded

[Dmitry Katsubo]

On 03/01/2016 22:31, Vincent Bernat wrote:
>> The mentioned config files exist, but are empty (perhaps that is the reason 
>> of the error).
> 
> Yes, since 1.0.0, roundcube now checks if $config is an array. Plugin
> configurations are expected to have something like $config['something']
> = ''. Maybe putting $config=array() in each file would work.

Placing $config=array() into each of these config files actually helped,
thank you! Can it be also fixed in Debian package? Thanks!

-- 
With best regards,
Dmitry

Bug#809769: roundcube: Error log is filled with messages that config file cannot be loaded

[Dmitry Katsubo]

Package: roundcube
Version: 1.1.4+dfsg.1-1
Severity: minor

The error log of roundcube (/var/log/roundcube/error) has a lot of message like 
that:

> [03-Jan-2016 20:43:38 +0100]:  PHP Error: Failed to load config 
> from /var/lib/roundcube/plugins/zipdownload/config.inc.php in 
> /usr/share/roundcube/program/lib/Roundcube/rcube_plugin.php on line 157 (GET 
> /roundcube/)
> [03-Jan-2016 20:43:38 +0100]:  PHP Error: Failed to load config 
> from /var/lib/roundcube/plugins/jqueryui/config.inc.php in 
> /usr/share/roundcube/program/lib/Roundcube/rcube_plugin.php on line 157 (GET 
> /roundcube/)

The mentioned config files exist, but are empty (perhaps that is the reason of 
the error).

-- 
With best regards,
Dmitry

Bug#783306: cyrus-imapd-2.4: Cyrus IMAPd fails to install/start

[Dmitry Katsubo]

On Mon, 11 May 2015 16:09:12 +0200 rollop...@gmail.com wrote:
 Same problem here, but I have solved with the command:

 update-rc.d cyrus-imapd enable

 Gabriel

Thanks, that helped me. I thought that Debian 8 is systemd-managed
system. The command above updates links in /etc/rcX.d -- how does it
help systemd?

BTW, I have discovered similar bug#739318 but that one seems to be fixed
year ago.


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#788997: cyrus-imapd-2.4: user_deny.db gets wrong ownership when it is created

[Dmitry Katsubo]

Package: cyrus-imapd-2.4
Version: 2.4.17+caldav~beta10-18
Severity: minor

It seems like Cyrus does not create user_deny.db with correct
ownership which causes the following to appear in syslog each time on
client login:

 cyrus/imap[14182]: IOERROR: opening /var/lib/cyrus/user_deny.db: Permission 
 denied
 cyrus/imap[14182]: DENYDB_ERROR: opening /var/lib/cyrus/user_deny.db: cyrusdb 
 error

The given file is owned by root:root but should be cyrus:mail. Removing
this file does not help, as it is re-created again.

Additional information:

# file /var/lib/cyrus/user_deny.db
/var/lib/cyrus/user_deny.db: Cyrus skiplist DB

# ll /var/lib/cyrus/user_deny.db
-rw--- 1 root root 144 Jun 15 06:25 /var/lib/cyrus/user_deny.db

-- 
With best regards,
Dmitry


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#785056: roundcube-core: Missing Net_LDAP3 from the vendor/kolab dir which is required for ldap support

[Dmitry Katsubo]

Vincent Bernat ber...@debian.org:
 Unfortunately, this needs to be packaged separately. I don't find it on
 PEAR. It seems to be available through composer but I don't know if
 that's easy to package. For some reason, just finding a tarball seems to
 be difficult.

Would be nice if php-kolab-net-ldap3 is installed as part of Roundcube
until a proper package for it is created.

Related: http://trac.roundcube.net/ticket/1490066

-- 
With best regards,
Dmitry


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#785666: mldonkey-server: Starts OK but reports an error to systemd

[Dmitry Katsubo]

Package: mldonkey-server
Version: 3.1.5-2+b2

Output of `systemctl status mldonkey-server.service` reads:

 systemd[1]: mldonkey-server.service: control process exited, code=exited 
 status=1
 systemd[1]: Failed to start LSB: Server for the mldonkey peer-to-peer 
 downloader..
 systemd[1]: Unit mldonkey-server.service entered failed state.

However mldonkey server starts and I can access it's Web UI. It seems to
be fully functional.

-- 
With best regards,
Dmitry
2015/05/18 20:01:13 [cO] Logging in .//var/log/mldonkey/mlnet.log
2015/05/18 20:01:13 [cO] Started logging...
2015/05/18 20:01:13 [dcCO] LETS reverse clients list NOW
2015/05/18 20:01:13 [cCO] Options correctly saved
2015/05/18 20:01:13 [dMain] loading geoip.dat from web_infos/GeoIP.dat.gz
2015/05/18 20:01:13 [Geo] country edition database loaded
2015/05/18 20:01:13 [dMain] loading server.met from web_infos/server.met.gz
2015/05/18 20:01:13 [EDK] server.met loaded from 
http://www.gruk.org/server.met.gz
2015/05/18 20:01:13 [EDK] 10 servers found, 2 new ones inserted
2015/05/18 20:01:13 [dMain] Check http://www.mldonkey.org for updates
2015/05/18 20:01:13 [dMain] enabling networks: 
2015/05/18 20:01:13 [dMain]  enabling Donkey 
2015/05/18 20:01:13 [EDK] loading sources completed
2015/05/18 20:01:13 [dMain] using port 17727 (client_port TCP)
2015/05/18 20:01:13 [dMain] using port 17731 (client_port UDP)
2015/05/18 20:01:13 [dMain]  enabling FileTP 
2015/05/18 20:01:13 [dMain]  enabling interfaces 
2015/05/18 20:01:13 [dMain] using port 4080 (http_port)
2015/05/18 20:01:13 [dMain] using port 4000 (telnet_port)
2015/05/18 20:01:13 [dMain] using port 4001 (gui_port)
2015/05/18 20:01:13 [dMain] disabled networks: BitTorrent Direct Connect 
2015/05/18 20:01:13 [dMain] To command: telnet 127.0.0.1 4000
2015/05/18 20:01:13 [dMain] Or with browser: http://127.0.0.1:4080
2015/05/18 20:01:13 [dMain] For a GUI check out 
http://sancho-gui.sourceforge.net
2015/05/18 20:01:13 [dMain] Connect to IP 127.0.0.1, port 4001
2015/05/18 20:01:13 [dMain] If you connect from a remote machine adjust 
allowed_ips
2015/05/18 20:01:13 [dMain] mldonkey is now running as user mldonkey
2015/05/18 20:01:14 [cCO] Options correctly saved
2015/05/18 20:01:14 [dMain] Core started
2015/05/18 20:01:14 [dMain] Core started
2015/05/18 20:01:14 [cWeb] request geoip.dat 
(http://www.maxmind.com/download/geoip/database/GeoLiteCountry/GeoIP.dat.gz)
2015/05/18 20:01:14 [cWeb] request guarding.p2p 
(http://www.bluetack.co.uk/config/level1.gz)
2015/05/18 20:01:14 [cWeb] request server.met 
(http://www.gruk.org/server.met.gz)
2015/05/18 20:01:14 [cWeb] request hublist 
(http://dchublist.com/hublist.config.bz2)
2015/05/18 20:01:14 [cWeb] Failure(Unknown kind [nodes.gzip]) while loading 
http://update.kceasy.com/update/fasttrack/nodes.gzip
2015/05/18 20:01:29 [cWeb] downloading newer web_infos/GeoIP.dat.gz, HTML 
header (Wed, 06 May 2015 21:00:20 GMT)
2015/05/18 20:01:29 [Geo] country edition database loaded
2015/05/18 20:01:30 [EDK] server.met loaded from 
http://www.gruk.org/server.met.gz
2015/05/18 20:01:30 [EDK] 10 servers found, 0 new ones inserted
2015/05/18 20:01:30 [HTTPcl] 410: bad reply for: 
http://www.bluetack.co.uk/config/level1.gz
2015/05/18 20:01:30 [cWeb] local file web_infos/level1.gz not found, HTTP 
request failed (error 410)
2015/05/18 20:01:33 [dcInt] DirectConnect module is disabled, ignoring...
2015/05/18 20:01:33 [HTTPcl] Exception Not_found in loading downloaded file 
web_infos/hublist.config.bz2
2015/05/18 20:01:33 [bS] exec closed : unexpected exn Not_found

Bug#783321: systemd opens file in /var/run and not in /run

[Dmitry Katsubo]

On 2015-04-27 17:13, Simon McVittie wrote:
 On 26/04/15 13:12, Dmitry Katsubo wrote:
 Indeed other files could be opened from /var, but in single mode that
 is very limited. The only service that lock it is NFS mount (rpcbind).
 And I can always stop these services, thus allowing me to unmount
 /var. But that is not the case with process with PID=1.
 
 If you're booting into single-user mode to do sufficiently low-level
 filesystem surgery that you want /var not mounted, I would really
 recommend doing it from the initramfs or a live-CD/live-USB/etc.
 environment, not the running system. jessie's initramfs-tools puts fsck
 in the initramfs.
 
 In particular, if you suspect that there might be disk corruption, using
 the maybe-corrupted system to repair itself seems much less than ideal:
 the critical path here has quite a lot of files in it (fsck, libc,
 ld.so, bash, e2fslibs...)

Thanks, Simon, for the suggestion. I haven't explored initramfs-tools so
far, but indeed live-CD/USB is much more flexible and reliable. In my
case I am confident about root partition (/) as it is mounted on SSD,
but /var could problematic.

 For the initramfs, only two files need to be intact (the kernel and the
 initramfs), and AIUI both of those are compressed data with a built-in
 checksum, so if it boots at all, you can be reasonably confident that
 it's good.
 
 If you would like a more elaborate recovery environment, I usually go
 for a small secondary installation of Debian stable in its own partition
 at the end of the disk, but grml and Debian Live are also good choices.

P.S. Answering Michael Biebl question concerning dbus version which he
asked some time ago: I have v1.8.16-1


-- 
With best regards,
Dmitry


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#783321: systemd opens file in /var/run and not in /run

[Dmitry Katsubo]

On 26/04/2015 01:05, Michael Biebl wrote:
 Am 26.04.2015 um 00:36 schrieb Michael Biebl:
 Am 26.04.2015 um 00:05 schrieb Dmitry Katsubo:
 
 Afterwards the process systemd opens a file in /var/run, thus
 not allowing me to unmount /var:
 
 systemd opens a file in /run, thus allowing the administrator
 to umount and e.g. repair /var volume.
 
 According to codesearch [1], we have quite a few locations where 
 /var/run/dbus/system_bus_socket is hard-coded.
 
 
 After thinking about this some more, I'm actually not convinced
 this is a bug at all. /var/run/dbus/system_bus_socket is by far not
 the only resource, which could be openend from /var.
 
 In the end, I think it's rather questionable if it's a good idea to
 run fsck on a system partition in single user mode. And if you want
 to do so, you'll just need to shut down all services, sockets and
 processes manually, e.g. by running systemctl stop foo.socket.
 
 If you want to run a forced fsck, there are much better facilities,
 like passing fsck.mode=force on the kernel-command-line [1].
 
 Given this, I'm not sure if it's worth the effort to move the dbus 
 socket around and I'm inclined to just close this bug report.
 
 I'll leave the decision to Simon though.
 
 Michael
 
 [1] man 8 systemd-fsck

Michael, I agree that is not a bug but rather an improvement. /run
was introduced long time ago, and it is known to be tmpfs-mounted
system. Thus I think that using this path should be preferred over
/var/run.

Indeed other files could be opened from /var, but in single mode that
is very limited. The only service that lock it is NFS mount (rpcbind).
And I can always stop these services, thus allowing me to unmount
/var. But that is not the case with process with PID=1. Should I just
kill it? :)

I haven't tried fsck.mode=force. I have entered single mode after
hard drive crash when fsck was forced at boot but complained that
filesystem cannot be repaired automatically, thus interactive mode is
necessary. So I don't think that passing fsck.mode=force will help:
I will be kicked to single mode again.

Repairing in single mode is easier then starting in another level (3)
and then stopping all services. Using mouse in single mode is not a
must, but that speeds up my work (e.g. copy-paste text). So ability to
use it is a plus. I can think about another services that don't use
/var and could be helpful, but once started they lock /var because of
systemd... For example: networking + ssh server.


-- 
With best regards,
Dmitry


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#783322: [Pkg-samba-maint] Bug#783322: winbindd crashes if LDAP password fails to be retrieved from secrets.tdb

[Dmitry Katsubo]

On 26/04/2015 01:27, Jelmer Vernooij wrote:
 Please set the LDAP database password using pdbedit. 
 
 This indeed shouldn't cause a segfault, merely a big error message in the 
 logs.
 
 Cheers,
 
 Jelmer

Thanks for suggestion, Jelmer.

Indeed after I have set the password smbpasswd -w password the
problem has gone. And you're right: the message of the issue is that
winbindd should not segfault, but should gracefully exit, or keep
retrying (but not aggressively, e.g. every 5 mins).

-- 
With best regards,
Dmitry


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#783322: winbindd crashes if LDAP password fails to be retrieved from secrets.tdb

[Dmitry Katsubo]

Package: winbind
Version: 2:4.1.17+dfsg-2

Hello,

winbind reports Failed to retrieve LDAP password from secrets.tdb and
crashes creating a core dump.

Expected:

winbind correctly handles this situation, e.g. exits with error code.


-- 
With best regards,
Dmitry
  STATUS=daemon 'winbindd' finished starting up and ready to serve 
connectionsGot sig[15] terminate (is_parent=0)
  STATUS=daemon 'winbindd' finished starting up and ready to serve 
connectionsfetch_ldap_pw: neither ldap secret retrieved!
[2014/06/22 23:25:10.038062,  0] 
../source3/passdb/pdb_ldap.c:6427(pdb_init_ldapsam_common)
  pdb_init_ldapsam_common: Failed to retrieve LDAP password from secrets.tdb
[2014/06/22 23:25:10.038313,  0] 
../source3/passdb/pdb_interface.c:178(make_pdb_method_name)
  pdb backend ldapsam:ldap://localhost did not correctly init (error was 
NT_STATUS_NO_MEMORY)
[2014/06/22 23:25:10.038500,  0] ../source3/lib/util.c:785(smb_panic_s3)
  PANIC (pid 3254): pdb_get_methods: failed to get pdb methods for backend 
ldapsam:ldap://localhost
  
[2014/06/22 23:25:10.054345,  0] ../source3/lib/util.c:896(log_stack_trace)
  BACKTRACE: 31 stack frames:
   #0 /usr/lib/i386-linux-gnu/libsmbconf.so.0(log_stack_trace+0x29) [0xb6cd4429]
   #1 /usr/lib/i386-linux-gnu/libsmbconf.so.0(smb_panic_s3+0x28) [0xb6cd4528]
   #2 /usr/lib/i386-linux-gnu/libsamba-util.so.0(smb_panic+0x3a) [0xb76116ba]
   #3 /usr/lib/i386-linux-gnu/libpdb.so.0(+0x2221f) [0xb732f21f]
   #4 /usr/lib/i386-linux-gnu/libpdb.so.0(pdb_capabilities+0x8) [0xb73311a8]
   #5 /usr/sbin/winbindd(_lsa_EnumTrustedDomainsEx+0x1c) [0xb772c61c]
   #6 /usr/sbin/winbindd(+0xb0b6f) [0xb7732b6f]
   #7 /usr/sbin/winbindd(+0x7d7dd) [0xb76ff7dd]
   #8 
/usr/lib/i386-linux-gnu/libdcerpc-binding.so.0(dcerpc_binding_handle_raw_call_send+0xa8)
 [0xb7390c28]
   #9 
/usr/lib/i386-linux-gnu/libdcerpc-binding.so.0(dcerpc_binding_handle_call_send+0x1d7)
 [0xb7391457]
   #10 
/usr/lib/i386-linux-gnu/libdcerpc-binding.so.0(dcerpc_binding_handle_call+0x6f) 
[0xb73916df]
   #11 
/usr/lib/i386-linux-gnu/samba/libdcerpc-samba.so.0(dcerpc_lsa_EnumTrustedDomainsEx_r+0x4d)
 [0xb741399d]
   #12 
/usr/lib/i386-linux-gnu/samba/libdcerpc-samba.so.0(dcerpc_lsa_EnumTrustedDomainsEx+0x5c)
 [0xb7413c2c]
   #13 /usr/sbin/winbindd(rpc_trusted_domains+0x93) [0xb76cb393]
   #14 /usr/sbin/winbindd(+0x4f759) [0xb76d1759]
   #15 /usr/sbin/winbindd(+0x34bf2) [0xb76b6bf2]
   #16 /usr/sbin/winbindd(winbindd_dual_list_trusted_domains+0x4a) [0xb76bea7a]
   #17 /usr/sbin/winbindd(+0x5220d) [0xb76d420d]
   #18 /usr/lib/i386-linux-gnu/libtevent.so.0(+0x8f0f) [0xb6b34f0f]
   #19 /usr/lib/i386-linux-gnu/libtevent.so.0(+0x713e) [0xb6b3313e]
   #20 /usr/lib/i386-linux-gnu/libtevent.so.0(_tevent_loop_once+0xa0) 
[0xb6b2f350]
   #21 /usr/sbin/winbindd(+0x54ca3) [0xb76d6ca3]
   #22 /usr/sbin/winbindd(+0x554ed) [0xb76d74ed]
   #23 /usr/lib/i386-linux-gnu/libtevent.so.0(+0x3fb0) [0xb6b2ffb0]
   #24 
/usr/lib/i386-linux-gnu/libtevent.so.0(tevent_common_loop_immediate+0xe8) 
[0xb6b2fc38]
   #25 /usr/lib/i386-linux-gnu/libtevent.so.0(+0x8caf) [0xb6b34caf]
   #26 /usr/lib/i386-linux-gnu/libtevent.so.0(+0x713e) [0xb6b3313e]
   #27 /usr/lib/i386-linux-gnu/libtevent.so.0(_tevent_loop_once+0xa0) 
[0xb6b2f350]
   #28 /usr/sbin/winbindd(main+0xd51) [0xb769c861]
   #29 /lib/i386-linux-gnu/i686/cmov/libc.so.6(__libc_start_main+0xf3) 
[0xb6999a63]
   #30 /usr/sbin/winbindd(+0x1b01e) [0xb769d01e]
[2014/06/22 23:25:10.057303,  0] ../source3/lib/util.c:797(smb_panic_s3)
  smb_panic(): calling panic action [/usr/share/samba/panic-action 3254]
[2014/06/22 23:25:10.400890,  0] ../source3/lib/util.c:805(smb_panic_s3)
  smb_panic(): action returned status 0
[2014/06/22 23:25:10.401174,  0] ../source3/lib/dumpcore.c:317(dump_core)
  dumping core in /var/log/samba/cores/winbindd

Bug#783321: systemd opens file in /var/run and not in /run

[Dmitry Katsubo]

Package: systemd
Version: 215-16

Hello,

I run Debian jessie in single mode (recovery mode). In this mode I would
like to start gpm service:

# /etc/init.d/gpm start

Afterwards the process systemd opens a file in /var/run, thus not
allowing me to unmount /var:

# lsof | grep /var
systemd  1 root   25u unix 0xf4760300  0t0  24111
/var/run/dbus/system_bus_socket

Expected:

systemd opens a file in /run, thus allowing the administrator to umount
and e.g. repair /var volume.

-- 
With best regards,
Dmitry


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#783305: cyrus-imapd-2.4: saslauthd is not listed as After service

[Dmitry Katsubo]

Package: cyrus-imapd-2.4
Version: 2.4.17+caldav~beta10-18

Hello,

I have noticed that saslauthd.service is not listed as service that has
to be started before Cyrus. If that service is active, it should be
started before as in my particular situation it is used for users'
authentication.

Expected that /lib/systemd/system/cyrus-imapd.service reads as follows:

After=local-fs.target network.target saslauthd.service

-- 
With best regards,
Dmitry


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#783306: cyrus-imapd-2.4: Cyrus IMAPd service fails to start

[Dmitry Katsubo]

Package: cyrus-imapd-2.4
Version: 2.4.17+caldav~beta10-18

Dear all,

I have the same problem as mentioned in bug#739318. In particular, Cyrus
is listed as service to be started, but it does not start:

# chkconfig --list cyrus-imapd
cyrus-imapd   0:off  1:off  2:on   3:on   4:on   5:on   6:off

After boot, no cyrus-imapd is listed in systemctl output. I have to
start it manually:

# /etc/init.d/cyrus-imapd start

After that it is listed as started:

# systemctl status cyrus-imapd.service
● cyrus-imapd.service - Cyrus IMAP/POP3 daemons
   Loaded: loaded (/lib/systemd/system/cyrus-imapd.service; disabled)
   Active: active (running) since Sat 2015-04-25 14:44:47 CEST; 3h 11min ago
  Process: 2406 ExecStartPre=/usr/sbin/cyrus init-helper start
(code=exited, status=0/SUCCESS)
 Main PID: 2432 (cyrmaster)
   CGroup: /system.slice/cyrus-imapd.service
   ├─2432 /usr/sbin/cyrmaster -l 32 -C /etc/imapd.conf -M
/etc/cyrus.conf
   ├─2763 notifyd
   ├─8591 imapd -U 30
   ├─8593 imapd -U 30
   └─8594 imapd -U 30

-- 
With best regards,
Dmitry


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#774164: libocrad-dev: libocrad.a contains non-reallocatable code

[Dmitry Katsubo]

Package: libocrad-dev
Version: 0.24-1

Hi,

It turned out that the -dev package contains the code in libocrad.a
which is not compiled with -fPIC. These object files can be used when
linking the executable, but not when linking the dynamic library (my case).

As far as I can see libocrad.a is used to create ocrad CLI. Can Makefile
be re-worked so that sources are compiled twice: once without PIC for
CLI and another time with PIC for .a?

Thanks!

Relative bug#583595.

-- 
With best regards,
Dmitry


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#582575: Introduce gocr-dev package

[Dmitry Katsubo]

Here is yet another update. What was fixed:

- Only minimal set of headers (*.h) is installed
- Sources for libPgm2asc.a are compiled with -fPIC

-- 
With best regards,
Dmitry


gocr_0.50-1.debian.tar.xz
Description: Binary data


signature.asc
Description: OpenPGP digital signature

Bug#582575: Introduce gocr-dev package

[Dmitry Katsubo]

Here is the update. The attached rules can build gocr-dev. Can anyone
push it to repo?

-- 
With best regards,
Dmitry


gocr_0.50-1.debian.tar.xz
Description: Binary data


signature.asc
Description: OpenPGP digital signature

Bug#770895: zoneminder: Push version 1.28.0 to experimental

[Dmitry Katsubo]

Package: zoneminder
Version: 1.26.5-3

Hello,

Could you please push v1.28.0 to experimental? I have made some adaptive
changes, but I cannot figure out what to do with systemd
mysqld.service and httpd.service. Are those put on systemd rails?

 https://github.com/dmak/ZoneMinder/commit/5f26c945a8eed58d15cb7e296e98980707797efb

Also may turned out that httpd.service should be optional dependency
because of bug#759504.

-- 
With best regards,
Dmitry


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#760303: squid 3.3.8-1.2 segfaults during initscript start/restart

[Dmitry Katsubo]

 On Fri, 12 Sep 2014 08:01:25 -0500 Dale Schroeder wrote:
 I don't understand how a log rotation can fail, yet succeed.
 
 The way the log rotation integrates with external log management tools
 is that Squid leaves the external tool to do file renaming or moving,
 and just re-opens the filedescriptors it has for the log files.
 
 So the logs have been changed successfully by logrotated. But the squid
 binary is asserting at some point in the followup process.
 
 The assertion seems to be caused by memory management for some globals
 allocated for the ICAP services.

Hello,

I have noticed the same error during log rotation. I have icap_enable
off (default) and no squidclamav (almost default config) but I have

  ecap_service gzip_service respmod_precache ...
  loadable_modules /usr/lib/squid3/ecap_adapter_gzip.so
  adaptation_access gzip_service allow GZIP_HTTP_STATUS
  ^ correlates with bug#4057 mentioned above

Now when I run squid from command line (with another instance running in
background), it segfauls:

 # gdb /usr/sbin/squid3
 Reading symbols from /usr/sbin/squid3...Reading symbols from 
 /usr/lib/debug//usr/sbin/squid3...done.
 (gdb) run
 Starting program: /usr/sbin/squid3
 [Thread debugging using libthread_db enabled]
 Using host libthread_db library 
 /lib/i386-linux-gnu/i686/cmov/libthread_db.so.1.
 
 Program received signal SIGABRT, Aborted.
 0xb7fde424 in __kernel_vsyscall ()
 (gdb) bt
 #0  0xb7fde424 in __kernel_vsyscall ()
 #1  0xb78f9307 in __GI_raise (sig=sig@entry=6) at 
 ../nptl/sysdeps/unix/sysv/linux/raise.c:56
 #2  0xb78fa9c3 in __GI_abort () at abort.c:89
 #3  0x801470a4 in xassert (msg=0x80438250 size == 
 StrPoolsAttrs[i].obj_size, file=0x8043f7c4 mem.cc, line=282) at 
 debug.cc:566
 #4  0x801f0fc7 in memFreeString (size=721, buf=0x806c87e0) at mem.cc:282
 #5  0x80233e1a in String::clean (this=0x806ccd0c) at String.cc:159
 #6  0x80399033 in Adaptation::AccessRule::~AccessRule (this=0x806ccd08, 
 __in_chrg=optimized out) at AccessRule.cc:16
 #7  0x80399f4a in Adaptation::Config::FreeAccess () at Config.cc:281
 #8  0x8039ad64 in Adaptation::Config::freeService (this=0x806b34c0 
 Adaptation::Icap::TheConfig) at Config.cc:148
 #9  0x8039adb8 in Adaptation::Config::~Config (this=0x806b34c0 
 Adaptation::Icap::TheConfig, __in_chrg=optimized out) at Config.cc:307
 #10 0x803b5d04 in Adaptation::Icap::Config::~Config (this=0x806b34c0 
 Adaptation::Icap::TheConfig, __in_chrg=optimized out) at Config.cc:54
 #11 0xb78fc121 in __run_exit_handlers (status=status@entry=0, 
 listp=0xb7a723c4 __exit_funcs, run_list_atexit=run_list_atexit@entry=true)
 at exit.c:82
 #12 0xb78fc17d in __GI_exit (status=0) at exit.c:104
 #13 0x801ee9cf in watch_child (argv=optimized out) at main.cc:1687
 #14 SquidMain (argc=1, argv=0xba94) at main.cc:1452
 #15 0x800d5c79 in SquidMainSafe (argv=0xba94, argc=1) at main.cc:1260
 #16 main (argc=1, argv=0xba94) at main.cc:1252

In /var/log/squid3/cache.log:

 2014/11/11 01:30:36| Squid is already running!  Process ID 22357
 2014/11/11 01:30:36| assertion failed: mem.cc:282: size == 
 StrPoolsAttrs[i].obj_size

I am running squid 3.4.8-2 / jessie

-- 
With best regards,
Dmitry


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#580336: gnome-icon-theme: Too many dependencies

[Dmitry Katsubo]

 Le mercredi 05 mai 2010 à 13:25 +0400, Roman V. Nikolaev a écrit : 
 I wan to use gnome icons in web project (rtpg-www package). Then I
 attempt install gnome-icon-theme on web server (no x-org server, no
 gnome on it, just console applications and daemons). And I get too more
 dependences like: fontconfig, libgtk2.0-0, x11-common, ...
 Can you decrease count of dependences?
 
 Theoretically it’s possible to remove the dependencies on
 librsvg2-common and libgtk2.0-bin, but it would possibly break a lot of
 applications which use icons from this package, expecting it to just
 work. And this doesn’t “just work” without those installed; the former
 for SVG support, the latter for icon cache support.
 
 Josselin Mouette

I agree with original bug reporter. The dependency on librsvg2-common
could be explained, but what dependency on libgtk-3-bin is needed for?
It takes 30+ dependencies, which include colord, sane-utils,
glib-networking and many others which have no direct relation to icons.


-- 
With best regards,
Dmitry


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#751003: squidguard: Unnecessary dependencies on perl modules

[Dmitry Katsubo]

Package: squidguard
Version: 1.5-1

squidguard package lists the following ones as dependencies:

* liburi-perl
* libwww-perl

These modules are not needed for squidguard to run in minimal
configuration. These modules are needed only if one use some examples
from /usr/share/doc/squidguard/examples/

So I think they can be marked as optional.

-- 
With best regards,
Dmitry


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#721395: libxml-twig-perl: warnings with new perl

[Dmitry Katsubo]

I suffer from the same problem. Would be fantastic if one can pull the
patched package to test repo.

 $[ used in numeric lt () (did you mean $] ?) at /usr/share/perl5/XML/Twig.pm 
 line 7292.
 $[ used in numeric lt () (did you mean $] ?) at /usr/share/perl5/XML/Twig.pm 
 line 7304.


-- 
With best regards,
Dmitry


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#694523: libxml-twig-perl: Few mistypings

[Dmitry Katsubo]

It seems like this bug is the same as bug#721395.

-- 
With best regards,
Dmitry


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#732210: Some serious incompatibilities with wheezy php 5.4

[Dmitry Katsubo]

On 18/12/2013 20:01, Thijs Kinkhorst wrote:
 On Tue, December 17, 2013 02:15, Dmitry Katsubo wrote:
 In case somebody will try to install SquirrelMail 1.5.1, there are two
 issues with it:

 1) PHP Fatal error:  Call to undefined function session_unregister() in
 /usr/share/squirrelmail/functions/global.php on line 111

 2) PHP Fatal error:  Cannot redeclare hex2bin() in
 /usr/share/squirrelmail/plugins/mail_fetch/functions.php on line 309

 Attached patches solve these problems.

 Also after installation one need to pickup sqspell_config.php from
 latest stable (e.g. 1.4.23) to make aspell dictionaries working.

 I personally need 1.5.1 to be able to STARTTLS for IMAP/SMTP server.
 Otherwise PLAIN authentication is forbidden.
 
 If you want to use 1.5 branch, you should take a snapshot from
 squirrelmail.org, as the 1.5.1 release is very dated and indeed suffers
 from the problems above.
 
 Thijs

Would it be difficult for package maintainers to take
squirrelmail-20131220_0200-SVN.devel.tar.bz2 and pull it to experimental
as is (provided the rest infrastructure is taken from 1.4.23)? If it
does not need patches above I will be happy to test it. For me the
difficulty is that I need to combine Debian-specific features from
1.4.23 with latest sources and I will certainly do something wrong.

-- 
With best regards,
Dmitry


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#732210: Some serious incompatibilities with wheezy php 5.4

[Dmitry Katsubo]

In case somebody will try to install SquirrelMail 1.5.1, there are two
issues with it:

1) PHP Fatal error:  Call to undefined function session_unregister() in
/usr/share/squirrelmail/functions/global.php on line 111

2) PHP Fatal error:  Cannot redeclare hex2bin() in
/usr/share/squirrelmail/plugins/mail_fetch/functions.php on line 309

Attached patches solve these problems.

Also after installation one need to pickup sqspell_config.php from
latest stable (e.g. 1.4.23) to make aspell dictionaries working.

I personally need 1.5.1 to be able to STARTTLS for IMAP/SMTP server.
Otherwise PLAIN authentication is forbidden.
--- /usr/share/squirrelmail/functions/global.php.orig   2006-07-08 
21:01:27.0 +0200
+++ /usr/share/squirrelmail/functions/global.php2012-12-19 
22:27:47.433789672 +0100
@@ -92,9 +92,9 @@
 
 sqsession_is_active();
 
-$_SESSION[$name] = $var;
+session_start();
 
-session_register($name);
+$_SESSION[$name] = $var;
 }
 
 /**
@@ -107,8 +107,6 @@
 sqsession_is_active();
 
 unset($_SESSION[$name]);
-
-session_unregister($name);
 }
 
 /**
--- /usr/share/squirrelmail/plugins/mail_fetch/functions.php.orig   
2006-07-08 21:09:31.0 +0200
+++ /usr/share/squirrelmail/plugins/mail_fetch/functions.php2012-12-19 
22:32:26.273782351 +0100
@@ -293,21 +293,6 @@
 }
 // end of hooked functions
 
-/**
- * hex2bin - document me
- */
-function hex2bin( $data ) {
-
-/* Original code by j...@superfork.com */
-
-$len = strlen($data);
-$newdata = '';
-for( $i=0; $i  $len; $i += 2 ) {
-$newdata .= pack( C, hexdec( substr( $data, $i, 2) ) );
-}
-return $newdata;
-}
-
 function mf_keyED( $txt ) {
 
 global $MF_TIT;

Bug#706388: freetds: Include the fix for long passwords

[Dmitry Katsubo]

Hi Steve,

Yes, all my references are from long-long ago. However I have reported
the problem (correctly) against latest at the time of bug report repo
version 0.82.

As patch is trivial and tds.h has not changed in 0.91, I am 99.9% sure
the problem is actual for that version as well.

On 11.09.2013 1:14, Steve Langasek wrote:
 This thread is from 2002, long before freetds 0.82 was released,
 and shows an upstream developer stating that the bug had been fixed
 back then.

The sources (0.82, 0.91) still state:

#define TDS_MAX_LOGIN_STR_SZ 30

 So why is this patch needed?  What protocol version are you using?

My /etc/freetds.conf says:

=== cut ===
[test]
host = 10.1.2.14
port = 1433
tds version = 7.1
=== cut ===

Also I have compiled FreeTDS with following command line:

CPPFLAGS=-I$HOME/include ./configure --prefix=$HOME --with-iodbc=$HOME
--enable-krb5=gssapi_krb5 --with-tdsver=7.1

 Is this problem reproducible with freetds 0.91, which released with
 wheezy? It looks like the code has changed significantly, and there
 are no string limits at all in the password field now.

Please, follow TDS_MAX_LOGIN_STR_SZ constant in all sources and you
will see that the limit is still there.


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#702580: libxine2: Looks for channels.conf in unexpected location hangs

[Dmitry Katsubo]

Thanks,

I really meant non-responsive as no UI buttons work. The only way to
continue is to kill the app. What if user wants to continue, open some
video file from hard drive? Have you succeeded with that?

On 29/05/2013 14:43, Darren Salt wrote:
 thanks
 
 linuxtv.org is not xine-project.org ☺
 
 The page in question does mention the correct location for channels.conf for
 xine-lib 1.2 (though it could probably do so more clearly since 1.1 has been
 deprecated for a while and I now regard it as obsolete).
 
 It definitely doesn't hang (or I don't see it hanging – you clearly mean
 non-responsive), and it does tell you which file was not found; any problems
 with display of this information is down to the front end (but both xine and
 gxine are fine here).
 
 ~/.config/xine-lib is correct for xine-lib 1.2.


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#582575: Introduce gocr-dev package

[Dmitry Katsubo]

Version 0.50 is available for download. Unfortunately, the package does
not install gocr.h (the corresponding line is commented out) however
build for .so and .a is there.

gocr.h in its turn includes pnm.h, unicode.h, list.h which in
their turn include more headers.

Could we still overcome this limitation and assemble the list of headers
(perhaps *.h would be OK) and finally package -dev?

Thanks!


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#706388: freetds: Include the fix for long passwords

[Dmitry Katsubo]

Package: freetds-dev
Version: 0.82-7

UUID-based passwords are 36 characters long. Unfortunately FreeTDS
limits the length to 30 characters. Attached patch fixes the problem.

Relative information:

http://lists.ibiblio.org/pipermail/freetds/2002q3/008105.html
http://stackoverflow.com/questions/15369304/

-- 
With best regards,
Dmitry
--- include/tds.h   2013-04-26 11:17:07.0 +0200
+++ include/tds.h.orig  2012-07-17 10:16:16.0 +0200
@@ -499,7 +499,7 @@
 
 #define TDS_ALIGN_SIZE sizeof(tds_align_struct)
 
-#define TDS_MAX_LOGIN_STR_SZ 50
+#define TDS_MAX_LOGIN_STR_SZ 30
 typedef struct tds_login
 {
DSTR server_name;   /** server name (in freetds.conf) */

Bug#706392: DBD::Sybase: Does not support passwords longer then 32 chars

[Dmitry Katsubo]

Package: libdbd-sybase-perl
Version: 1.00-3
Block: 706388

Unfortunately DBD::Sybase limits the length to 32 characters. Attached
patch fixes the problem, pushing the limit to 50 characters (49 to be
more exact with respect to terminating NULL).

-- 
With best regards,
Dmitry

--- dbdimp.h.orig   2013-04-03 22:16:20.0 +0200
+++ dbdimp.h2013-04-29 17:24:21.0 +0200
@@ -51,6 +51,9 @@
 
 #define MAX_SQL_SIZE 255
 
+/* The same as TDS_MAX_LOGIN_STR_SZ */
+#define MAX_LOGIN_STR_SZ 50
+ 
 /* Define dbh implementor data structure */
 struct imp_dbh_st {
dbih_dbc_t com; /* MUST be first element in structure   */
@@ -71,8 +74,8 @@
int lasterr;
int lastsev;
 
-   char uid[32];
-   char pwd[32];
+   char uid[MAX_LOGIN_STR_SZ];
+   char pwd[MAX_LOGIN_STR_SZ];
 
char server[64];
char charset[64];
--- dbdimp.c.orig   2013-04-03 22:16:20.0 +0200
+++ dbdimp.c2013-04-29 17:24:56.0 +0200
@@ -1147,10 +1147,10 @@
imp_dbh-server[63] = 0;
}
 
-   strncpy(imp_dbh-uid, uid, 32);
-   imp_dbh-uid[31] = 0;
-   strncpy(imp_dbh-pwd, pwd, 32);
-   imp_dbh-pwd[31] = 0;
+   strncpy(imp_dbh-uid, uid, MAX_LOGIN_STR_SZ);
+   imp_dbh-uid[MAX_LOGIN_STR_SZ - 1] = 0;
+   strncpy(imp_dbh-pwd, pwd, MAX_LOGIN_STR_SZ);
+   imp_dbh-pwd[MAX_LOGIN_STR_SZ - 1] = 0;
 
sv_setpv(DBIc_ERRSTR(imp_dbh), );
 

Bug#703190: xine-console: fbxine is causing Video port failed

[Dmitry Katsubo]

Package: xine-console
Version: 0.99.7-1

I am trying to launch Xine on FrameBuffer, and I am suffering from the
following problem:

When I am trying to start fbxine it literally shows the following:

centurion:~$ fbxine /mnt/video/film.avi
centurion:~$ iled.

After strace'ing it, I got the following result:

ioctl(4, FBIOPUT_VSCREENINFO, 0x98e04b4) = -1 EINVAL (Invalid argument)
ioctl(4, FBIOPUT_VSCREENINFO, 0x98e04b4) = -1 EINVAL (Invalid argument)
ioctl(4, FBIOPUT_VSCREENINFO, 0x98e04b4) = 0
ioctl(4, FBIOGET_VSCREENINFO, 0x98e04b4) = 0
ioctl(4, FBIOGET_FSCREENINFO, 0x98e0554) = 0
ioctl(4, FBIOPUTCMAP, 0xbfa392c8)   = -1 EINVAL (Invalid argument)
write(2, Video port failed.\n, 19Video port failed.
)= 19

and when I have compared this trace with video_out_fb.c sources, I come
to conclusion that set_fb_palette() is failing
(see
https://github.com/Raider05/enigma2pc/blob/master/xine-lib/src/video_out/video_out_fb.c#L776).

Additional information:

=== uname -a ===
Linux centurion 3.2.0-4-686-pae #1 SMP Debian 3.2.35-2 i686 GNU/Linux

=== fbset -i ===
mode 1024x768
geometry 1024 768 1024 768 32
timings 0 0 0 0 0 0 0
accel true
rgba 8/16,8/8,8/0,8/24
endmode

Frame buffer device information:
Name: inteldrmfb
Address : 0xd002
Size: 3145728
Type: PACKED PIXELS
Visual  : TRUECOLOR
XPanStep: 1
YPanStep: 1
YWrapStep   : 0
LineLength  : 4096
Accelerator : No

-- 
With best regards,
Dmitry


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#702868: dvb-apps: Wrong URL to source repository

[Dmitry Katsubo]

Package: dvb-apps
Version: 1.1.1+rev1483-2

At the moment the description file
http://ftp.de.debian.org/debian/pool/main/l/linuxtv-dvb-apps/linuxtv-dvb-apps_1.1.1+rev1483-2.dsc
contains the references to source repository which does not exist:

 Vcs-Browser: 
 http://svn.debian.org/wsvn/pkg-vdr-dvb/dvb/linuxtv-dvb-apps/trunk/
 Vcs-Svn: svn://svn.debian.org/pkg-vdr-dvb/dvb/linuxtv-dvb-apps/trunk/

The repository itself is 8 years old. That creates a confusion
concerning what exactly was used as a source for .orig.tar.bz2

-- 
With best regards,
Dmitry


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#702580: libxine2: Looks for channels.conf in unexpected location hangs

[Dmitry Katsubo]

Package: libxine2
Version: 1.2.2-4

When xine UI is switched to DVB-T viewing by clicking the DVB button
on control panel it looks for channels.conf not in location which is
mentioned in wiki (http://linuxtv.org/wiki/index.php/Xine).
Actually searched in:  ~/.config/xine-lib/channels.conf
But expected location: ~/.xine/channels.conf

The side effect is that xine shows only Sorry, No valid channels.conf
found and it hangs.

Several issues here:
* Xine becomes non-responsible, only kill helps.
* Message is not verbose to understand the problem.
* ~/.xine/ is more expected location for channels.conf

-- 
With best regards,
Dmitry


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#604468: imapd crashes with SIGSEGV in mboxlist.c:221

[Dmitry Katsubo]

The bug can be closed. The reason turned out to rooted in mailboxes.db,
where tabs need to be replaced by spaces. Probably that was caused by
export+import of mailboxes.db with buggy ctl_mboxlist.

See also:
 http://www.mail-archive.com/cyrus-devel@lists.andrew.cmu.edu/msg00305.html
 http://asg.andrew.cmu.edu/archive/message.php?mailbox=archive.info-cyrusmsg=54338

-- 
With best regards,
Dmitry


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#691026: backuppc: deprecated perl syntax in Text.pm and Lib.pm

[Dmitry Katsubo]

Package: backuppc
Version: 3.2.1-3
Severity: minor

When BackupPC server is started, the following is printed to console:

=== cut ===
Starting backuppc...
Use of qw(...) as parentheses is deprecated at
/usr/share/backuppc/lib/BackupPC/Storage/Text.pm line 302.
Use of qw(...) as parentheses is deprecated at
/usr/share/backuppc/lib/BackupPC/Lib.pm line 1413.
=== cut ===

Perl code should be updated to correspond to Perl interpreter in sid.

-- 
With best regards,
Dmitry


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#689705: grc: SIGPIPE should be suppressed

[Dmitry Katsubo]

Package: grc
Version: 1.1
Severity: minor

In scenario when grcat output is fed via pipe to some receiver and that
receiver exits, grcat displays the Broken pipe message:

=== cut ===
$ cat /log/tomcat.log | grcat ~/.grc.conf | head -10

Traceback (most recent call last):
  File /home/dmitry/bin/grcat, line 212, in module
print nline
IOError: [Errno 32] Broken pipe
=== cut ===

Probably SIGPIPE should be ignored.

-- 
With best regards,
Dmitry


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#663969: ppmd: Utility exits if large file is given as input

[Dmitry Katsubo]

Package: ppmd
Version: 10.1-4

When trying to process the large file (3Gb in my case), ppmd utility
silently exits without any processing:

# ll in.txt
-rwxr-xr-x 1 root root 3264023644 Mar  9 11:51 in.txt

# time ppmd e -s -ftest.ppmd in.txt
real0m0.038s
user0m0.000s
sys 0m0.036s


-- 
With best regards,
Dmitry
execve(/usr/bin/ppmd, [ppmd, e, -s, -ftest.ppmd, in.txt], [/* 37 
vars */]) = 0
brk(0)  = 0x88ff000
access(/etc/ld.so.nohwcap, F_OK)  = -1 ENOENT (No such file or directory)
mmap2(NULL, 8192, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 
0xb789
access(/etc/ld.so.preload, R_OK)  = -1 ENOENT (No such file or directory)
open(/etc/ld.so.cache, O_RDONLY)  = 3
fstat64(3, {st_mode=S_IFREG|0644, st_size=35843, ...}) = 0
mmap2(NULL, 35843, PROT_READ, MAP_PRIVATE, 3, 0) = 0xb7887000
close(3)= 0
access(/etc/ld.so.nohwcap, F_OK)  = -1 ENOENT (No such file or directory)
open(/usr/lib/i386-linux-gnu/libstdc++.so.6, O_RDONLY) = 3
read(3, \177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0P\246\4\0004\0\0\0..., 
512) = 512
fstat64(3, {st_mode=S_IFREG|0644, st_size=922060, ...}) = 0
mmap2(NULL, 951720, PROT_READ|PROT_EXEC, MAP_PRIVATE|MAP_DENYWRITE, 3, 0) = 
0xb779e000
mprotect(0xb787a000, 4096, PROT_NONE)   = 0
mmap2(0xb787b000, 20480, PROT_READ|PROT_WRITE, 
MAP_PRIVATE|MAP_FIXED|MAP_DENYWRITE, 3, 0xdc) = 0xb787b000
mmap2(0xb788, 26024, PROT_READ|PROT_WRITE, 
MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0xb788
close(3)= 0
access(/etc/ld.so.nohwcap, F_OK)  = -1 ENOENT (No such file or directory)
open(/lib/i386-linux-gnu/i686/cmov/libm.so.6, O_RDONLY) = 3
read(3, \177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0\2604\0\0004\0\0\0..., 
512) = 512
fstat64(3, {st_mode=S_IFREG|0644, st_size=148996, ...}) = 0
mmap2(NULL, 151680, PROT_READ|PROT_EXEC, MAP_PRIVATE|MAP_DENYWRITE, 3, 0) = 
0xb7778000
mmap2(0xb779c000, 8192, PROT_READ|PROT_WRITE, 
MAP_PRIVATE|MAP_FIXED|MAP_DENYWRITE, 3, 0x23) = 0xb779c000
close(3)= 0
access(/etc/ld.so.nohwcap, F_OK)  = -1 ENOENT (No such file or directory)
open(/lib/i386-linux-gnu/libgcc_s.so.1, O_RDONLY) = 3
read(3, \177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0\260#\0\0004\0\0\0..., 
512) = 512
fstat64(3, {st_mode=S_IFREG|0644, st_size=115472, ...}) = 0
mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 
0xb000
mmap2(NULL, 118544, PROT_READ|PROT_EXEC, MAP_PRIVATE|MAP_DENYWRITE, 3, 0) = 
0xb775a000
mmap2(0xb7776000, 4096, PROT_READ|PROT_WRITE, 
MAP_PRIVATE|MAP_FIXED|MAP_DENYWRITE, 3, 0x1b) = 0xb7776000
close(3)= 0
access(/etc/ld.so.nohwcap, F_OK)  = -1 ENOENT (No such file or directory)
open(/lib/i386-linux-gnu/i686/cmov/libc.so.6, O_RDONLY) = 3
read(3, \177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0\240o\1\0004\0\0\0..., 
512) = 512
fstat64(3, {st_mode=S_IFREG|0755, st_size=1401000, ...}) = 0
mmap2(NULL, 1415544, PROT_READ|PROT_EXEC, MAP_PRIVATE|MAP_DENYWRITE, 3, 0) = 
0xb760
mprotect(0xb7753000, 4096, PROT_NONE)   = 0
mmap2(0xb7754000, 12288, PROT_READ|PROT_WRITE, 
MAP_PRIVATE|MAP_FIXED|MAP_DENYWRITE, 3, 0x153) = 0xb7754000
mmap2(0xb7757000, 10616, PROT_READ|PROT_WRITE, 
MAP_PRIVATE|MAP_FIXED|MAP_ANONYMOUS, -1, 0) = 0xb7757000
close(3)= 0
mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 
0xb75ff000
set_thread_area({entry_number:-1 - 6, base_addr:0xb75ff6d0, limit:1048575, 
seg_32bit:1, contents:0, read_exec_only:0, limit_in_pages:1, seg_not_present:0, 
useable:1}) = 0
mprotect(0xb7754000, 8192, PROT_READ)   = 0
mprotect(0xb779c000, 4096, PROT_READ)   = 0
mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 
0xb75fe000
mprotect(0xb787b000, 16384, PROT_READ)  = 0
mprotect(0xb78ae000, 4096, PROT_READ)   = 0
munmap(0xb7887000, 35843)   = 0
brk(0)  = 0x88ff000
brk(0x892)  = 0x892
open(test.ppmd, O_RDONLY)  = -1 ENOENT (No such file or directory)
open(., O_RDONLY|O_NONBLOCK|O_LARGEFILE|O_DIRECTORY|O_CLOEXEC) = 3
getdents(3, /* 24 entries */, 32768)= 656
stat64(./in.txt, {st_mode=S_IFREG|0755, st_size=3264023644, ...}) = 0
getdents(3, /* 0 entries */, 32768) = 0
close(3)= 0
exit_group(0)   = ?

Bug#655528: awstats: Decoding URL in Pages-URL table

[Dmitry Katsubo]

On 12.01.2012 19:49, Sergey B Kirpichev wrote:
 I'm not sure if this is a correct an approach at all.  Do you
 assume that URL-encoded string is in UTF-8, right? Why?

I myself is not gig in specifications, one should carefully read RFC
1738  RFC 3986. I found that HTML 4.0 recommends the use of UTF-8 to
encode the query string and the same is mentioned here:
http://en.wikipedia.org/wiki/Query_string#URL_encoding

 Anyway, I'll lower the severity of this bug to wishlist.

Yes, wishlist is most appropriate. Hopefully somebody looking for
solution will find this patch. Decoding of URLs was crucial for me.

-- 
With best regards,
Dmitry



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#655528: awstats: Decoding URL in Pages-URL table

[Dmitry Katsubo]

Package: awstats
Version: 7.0~dfsg-2

Dear awstats maintainer,

I have noticed that awstats does not decode e.g. URLs containing
Cyrillic characters, leaving them URL-encoded:

/w/%D1%81%D0%B5%D0%BC%D1%8C%D1%8F

For me it is nice to see this link decoded:

/w/семья

I am attaching simple patch that solves this problem.

-- 
With best regards,
Dmitry
--- awstats.pl.orig 2011-10-16 16:29:59.728857013 +0200
+++ awstats.pl  2011-11-03 12:49:18.945971378 +0100
@@ -18,6 +18,7 @@
   ; # use Time::Local 'timelocal_nocheck' is faster but not supported by all 
Time::Local modules
 use Socket;
 use Encode;
+use URI::Escape; # Debian package: liburi-perl
 
 #--
 # Defines
@@ -8790,7 +8791,7 @@
print a href=\
  . XMLEncode($newkey)
  . \ target=\url\
- . XMLEncode($nompage) . /a;
+ . XMLEncode(uri_unescape($nompage)) . /a;
}
elsif ( $newkey =~ /^\// )
{ # URL seems to be an url extracted from a web or wap 
server log file
@@ -8805,7 +8806,7 @@
print a href=\
  . XMLEncode($urlprot://$SiteDomain$newkey)
  . \ target=\url\
- . XMLEncode($nompage) . /a;
+ . XMLEncode(uri_unescape($nompage)) . /a;
}
else {
print XMLEncode($nompage);

Bug#646015: libpam-modules: pam_env reports missing /etc/environment, which it is deprecated

[Dmitry Katsubo]

Package: libpam-modules
Version: 1.1.3-2

pam_env reports that /etc/environment is missing, but that file is
deprecated in favour of /etc/default/locale. To my opinion PAM
configuration files should be updated to ignore the missing /etc/environment

 pam_env(su:session): Unable to open env file: /etc/environment: No such file 
 or directory
 pam_env(sshd:setcred): Unable to open env file: /etc/environment: No such 
 file or directory

-- 
With best regards,
Dmitry



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#581039: tightvncserver: New upstream release 2.0.x

[Dmitry Katsubo]

I vote for this bug.
v2.0.4 is already available.

-- 
With best regards,
Dmitry



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#638228: cyrus-imapd: Package rebuild is not possible as the source tree is not returned to initial stage

[Dmitry Katsubo]

Package: cyrus-imapd-2.4
Version: 2.4.9
Severity: minor

It is clear that this error is not reproducible on build server, as it
always starts a build in a clean environment. However for development
reasons it is handy when the clean rule of the build cycle reverts all
changes back to original stage, so that e.g. dpkg-source does not
complain. However that was not the case for cyrus-imapd-2.2, and the
same for cyrus-imapd-2.4.

Expected: It is possible to rebuild a project again e.g. via
dpkg-buildpackage.
Actual: When dpkg-buildpackage is start again, it fails.

See the attached console log.

-- 
With best regards,
Dmitry
dpkg-source: error: cannot represent change to 
cyrus-imapd-2.4-2.4.9/config.guess:
dpkg-source: error:   new version is symlink to 
/usr/share/automake-1.11/config.guess
dpkg-source: error:   old version is plain file
dpkg-source: error: cannot represent change to cyrus-imapd-2.4-2.4.9/config.sub:
dpkg-source: error:   new version is symlink to 
/usr/share/automake-1.11/config.sub
dpkg-source: error:   old version is plain file
dpkg-source: error: cannot represent change to cyrus-imapd-2.4-2.4.9/install-sh:
dpkg-source: error:   new version is symlink to 
/usr/share/automake-1.11/install-sh
dpkg-source: error:   old version is plain file
dpkg-source: warning: ignoring deletion of file imap/pushstats.c
dpkg-source: warning: ignoring deletion of file imap/lmtpstats.h
dpkg-source: warning: ignoring deletion of file imap/lmtpstats.c
dpkg-source: warning: ignoring deletion of file imap/pushstats.h
dpkg-source: warning: ignoring deletion of file lib/imapopts.h
dpkg-source: warning: ignoring deletion of file lib/imapopts.c
dpkg-source: warning: ignoring deletion of directory autom4te.cache
dpkg-source: warning: ignoring deletion of file autom4te.cache/requests
dpkg-source: warning: ignoring deletion of file autom4te.cache/output.1
dpkg-source: warning: ignoring deletion of file autom4te.cache/traces.0
dpkg-source: warning: ignoring deletion of file autom4te.cache/output.0
dpkg-source: warning: ignoring deletion of file autom4te.cache/traces.1
dpkg-source: warning: ignoring deletion of file man/sieveshell.1
dpkg-source: warning: ignoring deletion of file man/imapd.conf.5
dpkg-source: error: cannot represent change to 
cyrus-imapd-2.4-2.4.9/doc/murder.png: binary file contents changed
dpkg-source: error: add doc/murder.png in debian/source/include-binaries if you 
want to store the modified binary in the debian tarball
dpkg-source: error: unrepresentable changes to source
dpkg-buildpackage: error: dpkg-source -b cyrus-imapd-2.4-2.4.9 gave error exit 
status 2

Bug#637744: proftpd-mod-ldap: LDAPServer scope doesn't work

[Dmitry Katsubo]

Package: proftpd-mod-ldap
Version: 1.3.4~rc2-3

The search scope is always base (scope=0):

slapd: SRCH base=cn=persons,cn=server scope=0 deref=0
filter=(uid=dmitry)

regardless the LDAPServer setting:

LDAPServer ldap://localhost??sub

The workaround is to use LDAPServer+LDAPSearchScope separately:

LDAPServer localhost
LDAPSearchScope subtree

This bug is in some sense the opposite bug for:

http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=500731

-- 
With best regards,
Dmitry



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#637050: Package perl-suid v5.12.x is not available in repositories [sid] and [experimental]

[Dmitry Katsubo]

On 08.08.2011 11:53, Dominic Hargreaves wrote:
 Yes, this is correct; suidperl has been removed upstream, so it has
 gone from the Debian pacakge too. Any package using suidperl in Debian
 should now have a bug open against it, if it hasn't already been fixed
 to not use suidperl.
 
 It's regrettable that this functionality was removed, but there was no
 interest in maintaining it upstream, so its removal was inevitable.

Dominic, thanks for you explanations.

What are the proposed workarounds? E.g. in my case I need to run few
Perl scripts under root (at least to elevate to disk group), which are
called by Cacti grapher to collect S.M.A.R.T HDD information.

Is /etc/sudoers the right way for that?

-- 
With best regards,
Dmitry



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#637044: When Cyrus package is rebuild again, the build fails

[Dmitry Katsubo]

Package: cyrus-common-2.2
Version: 2.2.13p1-15

Dear maintainers,

Currently when I rebuild the Cyrus package after the 1st successfull
build using
# dpkg-buildpackage -rfakeroot -b

the following error occurs:

=== cut ===
Applying patch 99-update-autoconf.dpatch
patching file config.h.in
patching file configure
Hunk #1 FAILED at 1.
Hunk #2 FAILED at 83.
Hunk #3 FAILED at 231.
...
143 out of 146 hunks FAILED -- rejects in file configure
patching file install-sh
Patch 99-update-autoconf.dpatch does not apply (enforce with -f)
make: *** [debian/stamp-patched] Error 1
dpkg-buildpackage: error: debian/rules build gave error exit status 2
=== cut ===

Patching ./configure file seems to be bad idea, as quilt reverts
patches before dh_auto_configure and applies after. Perhaps this patch
should be applied directly from debian/rules (what is the best practise?)

P.S. You will not see this error on build server logs, as build server
always start a build from pure installation.

-- 
With best regards,
Dmitry



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#637050: Package perl-suid v5.12.x is not available in repositories [sid] and [experimental]

[Dmitry Katsubo]

Package: perl-suid
Version: 5.10.1-17squeeze2

Dear package maintainers,

Currently when migrating from [squeeze] to [sid] I have faced the
following problem:

- The package perl was upgraded from 5.10 to 5.12 [OK]
- The package perl-suid was simply removed (as no corresponding
version was found in [sid] or [experimental])

This causes troubles in installations that rely on perl-suid.

Package perl:
 * lenny (oldstable) (perl): Larry Wall's Practical Extraction and Report 
 Language
   5.10.0-19lenny5 [security]: alpha amd64 arm armel i386 ia64 mips mipsel 
 powerpc s390 sparc
   5.10.0-19lenny3: hppa
 * squeeze (stable) (perl): Larry Wall's Practical Extraction and Report 
 Language
   5.10.1-17squeeze2 [security]: amd64 armel i386 ia64 kfreebsd-amd64 
 kfreebsd-i386 mips mipsel powerpc s390 sparc
 * wheezy (testing) (perl): Larry Wall's Practical Extraction and Report 
 Language
   5.12.4-2: amd64 armel i386 ia64 kfreebsd-amd64 kfreebsd-i386 mips mipsel 
 powerpc s390 sparc
 * sid (unstable) (perl): Larry Wall's Practical Extraction and Report Language
   5.12.4-2: amd64 armel armhf i386 ia64 kfreebsd-amd64 kfreebsd-i386 m68k 
 mips mipsel powerpc s390 sh4 sparc sparc64
   5.12.4-1 [debports]: alpha powerpcspe
   5.12.3-6: hurd-i386
   5.10.1-18: hppa
 * experimental (perl): Larry Wall's Practical Extraction and Report Language
   5.14.1-1: amd64 armel i386 ia64 kfreebsd-amd64 mips mipsel powerpc s390 
 sparc
   5.14.0-1: hurd-i386 kfreebsd-i386

Package perl-suid:
 * lenny (oldstable) (perl): Runs setuid Perl scripts
   5.10.0-19lenny5 [security]: alpha amd64 arm armel i386 ia64 mips mipsel 
 powerpc s390 sparc
   5.10.0-19lenny3: hppa
 * squeeze (stable) (perl): runs setuid Perl scripts
   5.10.1-17squeeze2 [security]: amd64 armel i386 ia64 kfreebsd-amd64 
 kfreebsd-i386 mips mipsel powerpc s390 sparc
 * sid (unstable) (perl): runs setuid Perl scripts
   5.10.1-19: alpha
   5.10.1-18: hppa

-- 
With best regards,
Dmitry



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#307298: Include patch that fixes the problem of CONNECT via SSL

[Dmitry Katsubo]

 But squeeze+1 will (hopefully) have v2.4, anyway.

Stefan, unfortunately that did not happen: [sid] does not contain 2.4.
Nor 2.3.

 And I don't think the code is tested enough to include in the Debian
 package.

True, but if nobody includes the feature, then nobody can test it. The
patch is already more then 3 years available...

Is the following issue fixed or not (after 1 year)?

 2) If the backend closes the connection, but the client doesn't, the apache
 thread hangs and will not close the connection to the client.

-- 
With best regards,
Dmitry



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#560207: closed by Ondřej Surý ond...@sury.org (Bug#560207: Cyrus package from sid is not compilable)

[Dmitry Katsubo]

I confirm that the package 2.2.13p1-5 compiles just fine.

On 15.04.2011 19:00, Debian Bug Tracking System wrote:
 This is an automatic notification regarding your Bug report
 which was filed against the cyrus-imapd-2.2 package:
 
 #560207: Cyrus package from sid is not compilable
 
 It has been closed by Ondřej Surý ond...@sury.org.
 
 Their explanation is attached below along with your original report.
 If this explanation is unsatisfactory and you have not received a
 better one in a separate message then please contact Ondřej Surý 
 ond...@sury.org by
 replying to this email.

-- 
With best regards,
Dmitry



--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#621006: lintian: Variable substitution in Maintainer field in causes warnings and no-maintainer-field error

[Dmitry Katsubo]

Dear Neils,
Dear Russ,

Thank you for your comments. I have understood the situation: the
substitution does not take place, and lintian is reporting the problem
correctly.

This bug#621006 can be closed with won't fix.

On 08.04.2011 19:53, Russ Allbery wrote:
 dpkg-genchanges only supports a substvars file for one specific purpose,
 namely setting the Format field
 
 Does it look like a bug for dpkg-source / dpkg-genchanges? I have found
 this bug report:
 http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=589609
 saying that it's only dpkg-gencontrol (and not dpkg-source /
 dpkg-genchanges) that performs the substitution.
 
 Yup, that's correct.
 
 I can agree with you. Does it mean, that it is only Maintainer field
 which is problematic? What if I want to substitute Homepage or
 Vcs-Browser or Version field?
 
 I'm pretty sure you can't.

-- 
With best regards,
Dmitry



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#621006: lintian: Variable substitution in Maintainer field in causes warnings and no-maintainer-field error

[Dmitry Katsubo]

Hi Niels,

On 07.04.2011 18:52, Niels Thykier wrote:
 Hi
 
 I have the feeling you have not read #615132, which is basically asking
 for a check against this particular usage of substvars.

To my opinion, emitting a warning is fine.

Jakub says that use substitution variables in the source is always a
packaging bug, however the documentation concerning debian/control
file keeps silence concerning the fields which are not allowed to be
substituted:

http://www.debian.org/doc/debian-policy/ch-source.html#s-substvars

 This most likely happens because dpkg command (exact name eludes me at
 the moment, possibly dpkg-source or dpkg-genchanges) building the dsc
 does not include the Maintainer field.

You are right here: the generated .dsc file does not contain
Maintainer field at all, and dpkg-genchanges generated the .changes
file with Maintainer not substituted:
=== cut ===
Maintainer: ${common:Maintainer}
=== cut ===

I have explicitly passed the substvars file to use for dpkg-source /
dpkg-genchanges like that:
debuild -sa -us -uc --changes-option=-Tdebian/substvars
but it does not help:

=== console ===
   debian/rules override_dh_gencontrol
make[1]: Entering directory `/root/tmp/build/osra-1.3.8'
dh_gencontrol -- -Tdebian/substvars
...
 dpkg-genchanges -sa -Tdebian/substvars ../osra_1.3.8-1_i386.changes
dpkg-genchanges: including full source code in upload
 dpkg-source -Tdebian/substvars --after-build osra-1.3.8
dpkg-buildpackage: full upload (original source is included)
Now running lintian...
E: osra source: no-maintainer-field
...
=== console ===

As from the logs I also override dh_gencontrol, which by default is
passing package-specific substvars file:
dh_gencontrol -Tdebian/osra.substvars
while I want to use one substvars for all packages.

As I mentioned, all DEBIAN/control files in the .deb packages look OK.

Does it look like a bug for dpkg-source / dpkg-genchanges? I have found
this bug report:
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=589609
saying that it's only dpkg-gencontrol (and not dpkg-source /
dpkg-genchanges) that performs the substitution. But still -T option
is there... I've got a bad feeling :)

 I assume it does this if the Maintainer field is empty, which is very
 likely as the debian/substvars file is usually removed by the clean
 target (e.g. by dh_clean) and is therefore not present when the source
 package is created.

No, debian/substvars is not removed (at least explicitly by me) at
dh_clean step and it stays in debian/ after the build.

 Note we also have a tag for keeping the debian/substvars file in the
 source package because doing so may lead to complications/weird results.

Could you provide more information on that (which lintian option / tag?)

 Obviously we cannot satisfy this and #615132 at the same time.
 Personally I am more inclined to satisfy #615132, since these
 substitutions can be rather non-trivial and (as you experienced) lead to
 some interesting issues when done in source fields.

I can agree with you. Does it mean, that it is only Maintainer field
which is problematic? What if I want to substitute Homepage or
Vcs-Browser or Version field?

Thanks.

-- 
With best regards,
Dmitry



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#621006: lintian: Variable substitution in Maintainer field in causes warnings and no-maintainer-field error

[Dmitry Katsubo]

Package: lintian
Version: 2.4.3

When debian/control file reads:

=== cut ===
Maintainer: ${common:Maintainer}
=== cut ===

and in my debian/substvars I have:

=== cut ===
common:Maintainer=Dmitry Katsubo dm...@mail.ru
=== cut ===

the lintian produces the following warnings for osra source package
and no-maintainer-field error (but reports nothing for other real
packages):

 Now running lintian...
 Use of uninitialized value $maintainer in substitution (s///) at 
 /usr/share/lintian/checks/nmu line 139.
 Use of uninitialized value $maintainer in string ne at 
 /usr/share/lintian/checks/nmu line 97.
 Use of uninitialized value $maintainer in pattern match (m//) at 
 /usr/share/lintian/checks/nmu line 109.
 E: osra source: no-maintainer-field

Expected: the maintainer field is analysed with respect to variable
substitution.



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#588248: Provide library for static, linkage (libopenbabel.a)

[Dmitry Katsubo]

I have discovered that Openbabel supports configure --disable-shared.
Yes, it will double the assembly time and perhaps that is not trivial to
have both shared and non-shared deliverables on one debuild run.

-- 
With best regards,
Dmitry



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#616688: mdadm: crontab's array check should not report the error

[Dmitry Katsubo]

Package: mdadm
Version: 3.1.4-1+8efb9d1
Severity: whishist

/etc/cron.d/mdadm should call checkarray with -Q to ignore the error
message, which is reported by crontab with email:

 Subject:  Cron root@centurion if [ -x /usr/share/mdadm/checkarray ]  [ 
 $(date +%d) -le 7 ]; then /usr/share/mdadm/checkarray --cron --all --idle 
 --quiet; fi (failed)

 checkarray: E: MD subsystem not loaded, or /proc unavailable.



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#616688: mdadm: crontab's array check should not report the error

[Dmitry Katsubo]

On 06.03.2011 18:49, martin f krafft wrote:
 Why do you have mdadm installed but no MD subsystem loaded?

I think, this package is installed by default, even though I am not
using it. I suppose, crontab job should not fail if no RAID is setup in
the system.



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#613457: openbabel: Request to package v2.3.0

[Dmitry Katsubo]

Package: openbabel
Version: 2.2.3-1

On 2010-10-26 Open Babel 2.3.0 Released  was released. Please, package
it. Release notes: http://openbabel.org/wiki/Open_Babel_2.3.0

-- 
With best regards,
Dmitry



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#611740: Please package cyrus-imapd 2.4.6

[Dmitry Katsubo]

Package: cyrus-imapd-2.4
Version: 2.4.5-1

On 12/20/2010 cyrus-imapd-2.4.6 was released. Please, package it.

Release notes:

We are pleased to announce the immediate availability of Cyrus IMAPd
version 2.4.6. This is primarily a bugfix release. We recommend all
users of the Cyrus 2.4.x series update to this release, particularly if
they use OpenBSD or murder.

-- 
With best regards,
Dmitry



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#611642: ld.so.conf configuration file for cuneiform

[Dmitry Katsubo]

Package: cuneiform
Version: 1.0.0+dfsg-1

It would be nice if Debian package includes the following file:

=== cut /etc/ld.so.conf.d/cuneiform.conf ===
/usr/lib/cuneiform
=== end of cut ===

plus runs ldconfig after installation. Otherwise when Cuneiform library
is used (linked against), dh_shlibdeps reports that it cannot find
libcuneiform.so.1.0.0

-- 
With best regards,
Dmitry



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#611645: Provide information for static library linking

[Dmitry Katsubo]

Package: tesseract-ocr-dev
Version: 2.04-2
Severity: wishlist

It would be nice if together with .la files Debian development package
also provides the configuration files for pkgconfig (location:
/usr/lib/pkgconfig), which is easier to consume via CLI utility (e.g.
pkg-config --static tesseract_api).

These files could be generated by configure script using the following
template:

=== template ===
prefix=@prefix@
exec_prefix=@exec_prefix@
libdir=@libdir@
includedir=@includedir@

Name: @PACKAGE_NAME@
Description: ... put something here...
Version: @PACKAGE_VERSION@

Libs: @LDFLAGS@ @LIBS@
Cflags: @CFLAGS@
=== end of template ===

-- 
With best regards,
Dmitry



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#598616: cuneiform: development package is needed

[Dmitry Katsubo]

On 28.01.2011 0:55, Jakub Wilk wrote:
 Hi Dmitry,
 
 * Dmitry Katsubo dm...@mail.ru, 2011-01-21, 13:36:
 I attach the trivial solution for the bug: the library header
 cuneiform.h is packaged as libcuneiform-dev.deb
 
 Apart from being trivial it is also wrong. :)

It is just a proposal, which of course might not be perfect. But I am
happy we both understand the direction of improvements.

 With this patch applied it is (still) not possible to link to
 libcuneiform without jumping through extra hops.

Sorry, I can't guess what you really mean here. Perhaps you need to give
a concrete example about what does not work well.

 Worse still, it is not possible to build a policy-compliant package
 linking to libcuneiform at all; please consult Debian Policy 8.

I've opened Debian Policy Manual Chapter 8 - Shared libraries:
http://www.debian.org/doc/debian-policy/ch-sharedlibs.html
It has many subchapters. Perhaps you refer section 8.4 and the rule that
there should be a symlink for shared library without a version number. I
fully agree with this. I hope this is easy to fix (please do so).

 All in all, package layouts should look like this:
 
 * cuneiform:
   /usr/share/man/man1/cuneiform.1.gz
   /usr/bin/cuneiform
 
 * libcuneiformsoversion:
   /usr/lib/libcuneiform.so.soversion
   /usr/lib/cuneiform/libexc.so.soversion
   /usr/lib/cuneiform/librsadd.so.soversion
   /usr/lib/cuneiform/librlings.so.soversion
   [... - and so on; it's better to keep these auxiliary libraries in a
 private
 directory, as other software is not supposed to link to them]
 
 * libcuneiform-dev:
   /usr/include/cuneiform.h
   /usr/lib/libcuneiform.so - /usr/lib/libcuneiform.so.soversion

Absolutely agreed. Looks very nice.

 Apart from that, two things are worrying me:
 
 1. A few types and constants defined in cuneiform.h have very generic
 names (enum Languages, Bool, LANG_ENGLISH etc.). This is not
 something acceptable in a public header file, and these names could
 easily conflict with other names in user's code.

Yes, this might be a problem, and I am already facing this problem with
other libraries. If you can, please, raise this question in mainstream,
in particular, adding a CPP namespace for all declared types.

 2. SONAME is currently libcuneiform.so.1.0.0. Does it mean that it's
 going to be libcuneiform.so.upstream-version, i.e. changing with
 every upstream release? That would mean that with every upstream release:
 - Debian package would have to pass through the NEW queue.
 - Packages linking to the shared library (if any) would have to be rebuilt.
 This is not something I'd be happy about. SONAME should change if and
 only if the new version actually breaks API.

You're absolutely right here: the SO version number should be increased
only when API changes. So do not change it with every upstream release
(perhaps only the minor number).

-- 
With best regards,
Dmitry



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#611086: tesseract-ocr-dev should not depend on tesseract-ocr

[Dmitry Katsubo]

Package: tesseract-ocr-dev
Version: 2.04-2

I suppose the libraries in tesseract-ocr-dev are not calling any
binaries in tesseract-ocr. It should be visa versa: binaries in
tesseract-ocr might use the libraries in tesseract-ocr-dev. In this case
tesseract-ocr-dev should be split into:
* libtesseract-ocr (*.so only)
* libtesseract-ocr-dev (*.a, *.la, *.h, ...)
otherwise the dependency should be simply dropped.

Perhaps tesseract-ocr-dev depends on shared packages, like
tesseract-ocr-language.

-- 
With best regards,
Dmitry



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#598616: cuneiform: development package is needed

[Dmitry Katsubo]

Version: 1.0.0+dfsg-1

I attach the trivial solution for the bug: the library header
cuneiform.h is packaged as libcuneiform-dev.deb

Please consider for inclusion.

Perhaps it also makes sense to split cuneiform package into
cuneiform (binary only) and libcuneiform (.so only) and update the
dependencies.
diff -du -ruN debian.orig/control debian/control
--- debian.orig/control 2010-09-10 16:34:51.0 +0200
+++ debian/control  2011-01-21 11:13:21.0 +0100
@@ -33,3 +33,19 @@
  Swedish, Turkish and Ukrainian.
  .
  This package contains the common files.
+
+Package: libcuneiform-dev
+Architecture: any
+Section: libdevel
+Depends: ${misc:Depends}, cuneiform
+Description: multi-language OCR system (headers)
+ Cuneiform is an OCR system. In addition to text recognition it also does 
layout
+ analysis and text format recognition.
+ . 
+ The following languages are supported: Bulgarian, Croatian, Czech, Danish,
+ Dutch, English, Estonian, French, German, Hungarian, Italian, Latvian,
+ Lithuanian, Polish, Portuguese, Romanian, Russian, Serbian, Slovak, Spanish,
+ Swedish, Turkish and Ukrainian.
+ .
+ This package contains the development header files.
+
diff -du -ruN debian.orig/libcuneiform-dev.install 
debian/libcuneiform-dev.install
--- debian.orig/libcuneiform-dev.install1970-01-01 01:00:00.0 
+0100
+++ debian/libcuneiform-dev.install 2011-01-21 10:56:38.0 +0100
@@ -0,0 +1 @@
+/usr/include/cuneiform.h