Debian Bug report logs - #733974
Crash when browsing "Public Serverlist"

Toggle useless messages

Message #5 received at submit@bugs.debian.org (full text, mbox, reply):

From: Mantas Kriaučiūnas <mantas@akl.lt>

To: submit@bugs.debian.org

Subject: New version - minetest 0.4.9 released, please update Debian packages and test if new packages doesn't crash when browsing "Public Serverlist"

Date: Thu, 2 Jan 2014 20:46:26 +0200

Package: minetest
Version: 0.4.8+repack2-3

Minetest package is buggy (crashes when selecting Client->"Public Serverlist"; doesn't work Mods->"Online mod repository") and outdated in Debian - new minetest 0.4.9 was released few days ago.
0.4.9 release contains some improvements and bugfixes.
It would be nice if you test if new packages doesn't crash when selecting Client->"Public Serverlist" and doesn't show an error when pressing Mods->"Online mod repository" in startup screen.

0.4.8 → 0.4.9 New Features
Logistic changes
    SQLite rollback (Mario Barrera & ShadowNinja)
    Implement HTTPFetch (kahrl)
    Replace SimpleThread with JThread (sapier) 
Visual changes
    Shaders rework (RealBadAngel)
    Add configurable font shadow (xyz)
    Directional fog + horizon colors (Taoki) 
Other things
    Implement modstore search tab and version picker (sapier) 

Bug Fixes
 Fix line_of_sight() (sapier)
 Fix modstore/favourites hang by adding asynchronous Lua (sapier)
 Fix LevelDB maps (sfan5)
 Fix Lua mapgen override param handling (kwolekr)
 Fix leak and possible segfault in minetest.set_mapgen_params (kwolekr)
 Fix segfault in indev cave generation due to uninitialized variable (kwolekr)
 Check if width, height or start index of a list[] is negative (PilzAdam)
 Fix single character formspec field labels (BlockMen)
 Handle Lua errors in on_generate callbacks instead of throwing SIGABRT (kwolekr)
 Update mapgen params in ServerMap after Mapgen init (kwolekr) 

Modding-related Changes
 Add area parameters back to calc_lighting() and set_lighting() (kwolekr)
 Add get_light_data() and set_light_data() to LuaVoxelManip (kwolekr)
 Add minetest.swap_node (Novatux)
 Assume a selection box for fences (0gb.us)
 Decoration: Add schematic Y-slice probability support (kwolekr)
 Add sneak and sneak_glitch in set_physics_override() (PilzAdam)
 Use a table in set_physics_override() (PilzAdam)
 Add 'on_prejoinplayer' callback (kaeza)
 Make line_of_sight return blocking node position (stujones11)
 Remove support for optdepends.txt (ShadowNinja)
 Add map feature generation notify Lua API (kwolekr) 

See http://dev.minetest.net/Changelog


P.S. I'm pasting terminal output after 0.4.8+repack2-3 crash when selecting Client->"Public Serverlist":
(crashes only official Debian packages, but deb packages from https://launchpad.net/~minetestdevs/+archive/stable works fine)
baltix@linux:~$ minetest
Irrlicht Engine version 1.8.0
Linux 3.11.0-12-generic #19-Ubuntu SMP Wed Oct 9 16:20:46 UTC 2013 x86_64
Using renderer: OpenGL 3.0
Mesa DRI Intel(R) Haswell Server : Intel Open Source Technology Center
OpenGL driver version is 1.2 or better.
GLSL version: 1.3
Loaded texture: /usr/share/games/minetest/games/minetest_game/menu/header.png
Loaded texture: /usr/share/games/minetest/games/minetest_game/menu/icon.png

19:42:47: ERROR[main]: MAINMENU ERROR: LuaError: C++ exception
19:42:48: ERROR[main]: ========== ERROR FROM LUA ===========
19:42:48: ERROR[main]: Failed to load and run script from 
19:42:48: ERROR[main]: /usr/share/games/minetest/builtin/mainmenu.lua:
19:42:48: ERROR[main]: C++ exception
19:42:48: ERROR[main]: ======= END OF ERROR FROM LUA ========
19:42:48: ERROR[main]: GUIEngine: unable to load builtin menu

In thread 7f9f3eadc7c0:
/build/buildd/minetest-0.4.8+repack2/src/guiEngine.cpp:199: GUIEngine::GUIEngine(irr::IrrlichtDevice*, irr::gui::IGUIElement*, IMenuManager*, irr::scene::ISceneManager*, MainMenuData*, bool&): Assertion '"no future without mainmenu" == 0' failed.
Debug stacks:
DEBUG STACK FOR THREAD 7f9f3eadc7c0:
#0  int main(int, char**)
Aborted (core dumped)
baltix@linux:~$ 

-- 
Labanaktis/Good luck,
Mantas Kriaučiūnas	 Jabber ID: mantas@akl.lt 	 GPG ID: 43535BD5
Public organization "Open Source for Lithuania" - www.akl.lt
Geriausios biuro programos verslui ir namams - http://openoffice.lt
Prekyba naujais ir atnaujintais kompiuteriais su Linux OS - http://tinklas.eu
Naudok Baltix GNU/Linux sistemą savo kompiuteryje - http://baltix.lt

Message #10 received at 733974@bugs.debian.org (full text, mbox, reply):

From: Martin Quinson <martin.quinson@loria.fr>

To: Mantas Kriaučiūnas <mantas@akl.lt>, 733974@bugs.debian.org

Subject: Re: Bug#733974: New version - minetest 0.4.9 released, please update Debian packages and test if new packages doesn't crash when browsing "Public Serverlist"

Date: Sat, 4 Jan 2014 21:36:55 +0100

package minetest
retitle 733974 Crash when browsing "Public Serverlist"
thanks (control@b.d.o is bcc'ed)

Hello,

Please open one bug per issue. It's easier to handle on my side. 0.4.9
is packaged, but Mods->"Online mod repository" still results in the
same error. I will try to find the time to investigate, but any help
is welcomed.

Bye, Mt.

On Thu, Jan 02, 2014 at 08:46:26PM +0200, Mantas Kriaučiūnas wrote:
> 
> P.S. I'm pasting terminal output after 0.4.8+repack2-3 crash when selecting Client->"Public Serverlist":
> (crashes only official Debian packages, but deb packages from https://launchpad.net/~minetestdevs/+archive/stable works fine)
> baltix@linux:~$ minetest
> Irrlicht Engine version 1.8.0
> Linux 3.11.0-12-generic #19-Ubuntu SMP Wed Oct 9 16:20:46 UTC 2013 x86_64
> Using renderer: OpenGL 3.0
> Mesa DRI Intel(R) Haswell Server : Intel Open Source Technology Center
> OpenGL driver version is 1.2 or better.
> GLSL version: 1.3
> Loaded texture: /usr/share/games/minetest/games/minetest_game/menu/header.png
> Loaded texture: /usr/share/games/minetest/games/minetest_game/menu/icon.png
> 
> 19:42:47: ERROR[main]: MAINMENU ERROR: LuaError: C++ exception
> 19:42:48: ERROR[main]: ========== ERROR FROM LUA ===========
> 19:42:48: ERROR[main]: Failed to load and run script from 
> 19:42:48: ERROR[main]: /usr/share/games/minetest/builtin/mainmenu.lua:
> 19:42:48: ERROR[main]: C++ exception
> 19:42:48: ERROR[main]: ======= END OF ERROR FROM LUA ========
> 19:42:48: ERROR[main]: GUIEngine: unable to load builtin menu
> 
> In thread 7f9f3eadc7c0:
> /build/buildd/minetest-0.4.8+repack2/src/guiEngine.cpp:199: GUIEngine::GUIEngine(irr::IrrlichtDevice*, irr::gui::IGUIElement*, IMenuManager*, irr::scene::ISceneManager*, MainMenuData*, bool&): Assertion '"no future without mainmenu" == 0' failed.
> Debug stacks:
> DEBUG STACK FOR THREAD 7f9f3eadc7c0:
> #0  int main(int, char**)
> Aborted (core dumped)
> baltix@linux:~$ 
> 

Message #17 received at 733974@bugs.debian.org (full text, mbox, reply):

From: "Andreas B. Mundt" <andi.mundt@web.de>

To: Debian Bug Tracking System <733974@bugs.debian.org>

Subject: minetest: gdb backtrace and strace

Date: Sun, 02 Feb 2014 20:14:13 +0100

[Message part 1 (text/plain, inline)]

Package: minetest
Version: 0.4.9+repack-4
Followup-For: Bug #733974


Dear Martin,

please find attached a gdb trace from the crash.

In addition, I made a strace which is quite large:  When starting
minetest first no window opens (strace 'runs' as crazy), but if I
switch the screen and back again the window is there ("awesome" window
manager).   I do not know if the issues are related.

Best regards,

     Andi



-- System Information:
Debian Release: jessie/sid
  APT prefers testing
  APT policy: (500, 'testing')
Architecture: amd64 (x86_64)

Kernel: Linux 3.12-1-amd64 (SMP w/4 CPU cores)
Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash

Versions of packages minetest depends on:
ii  libc6            2.17-97
ii  libcurl3-gnutls  7.35.0-1
ii  libfreetype6     2.5.2-1
ii  libgcc1          1:4.8.2-14
ii  libirrlicht1.8   1.8.1+dfsg1-1
ii  libjsoncpp0      0.6.0~rc2-3
ii  libluajit-5.1-2  2.0.2+dfsg-1
ii  libogg0          1.3.1-1
ii  libopenal1       1:1.14-4
ii  libsqlite3-0     3.8.2-1
ii  libstdc++6       4.8.2-14
ii  libvorbis0a      1.3.2-1.3
ii  libvorbisfile3   1.3.2-1.3
ii  minetest-data    0.4.9+repack-4
ii  zlib1g           1:1.2.8.dfsg-1

minetest recommends no packages.

Versions of packages minetest suggests:
pn  minetest-mod-moreblocks  <none>
pn  minetest-mod-moreores    <none>
pn  minetest-mod-pipeworks   <none>
ii  minetest-server          0.4.9+repack-4

-- no debconf information

[minetest.gdb (text/plain, attachment)]

[minetest.strace (text/plain, attachment)]

Message #22 received at 733974@bugs.debian.org (full text, mbox, reply):

From: Martin Quinson <martin.quinson@loria.fr>

To: "Andreas B. Mundt" <andi.mundt@web.de>, 733974@bugs.debian.org

Subject: Re: Bug#733974: minetest: gdb backtrace and strace

Date: Sun, 2 Feb 2014 21:35:26 +0100

Dear Andreas,

many thanks for that detailed information, that's really helping. The
only bit of information missing here is a console output (as it seems
that the segfault occured while reporting an error message). I assume
that the output is similar to the first reporter, but it would be
great if you could doublecheck, please.

As for the other potential issue, I've no idea. Actually, I reported
the bug to upstream, seeking for help. I'm myself useless and clueless
here.

Thanks, Mt.

On Sun, Feb 02, 2014 at 08:14:13PM +0100, Andreas B. Mundt wrote:
> Package: minetest
> Version: 0.4.9+repack-4
> Followup-For: Bug #733974
> 
> 
> Dear Martin,
> 
> please find attached a gdb trace from the crash.
> 
> In addition, I made a strace which is quite large:  When starting
> minetest first no window opens (strace 'runs' as crazy), but if I
> switch the screen and back again the window is there ("awesome" window
> manager).   I do not know if the issues are related.
> 
> Best regards,
> 
>      Andi


-- 
I think that God in creating Man somewhat overestimated his ability.
          --- Oscar Wilde

Message #27 received at 733974@bugs.debian.org (full text, mbox, reply):

From: "Andreas B. Mundt" <andi@debian.org>

To: 733974@bugs.debian.org

Subject: cmdline trace

Date: Sun, 2 Feb 2014 23:05:26 +0100

Dear Martin,

find the console trace below.  I tried to cherry pick some suspicious
commits from upstream to find out more, but without success.

Something I observed:  Termination happens when I choose (in the main
menu) either:

        Mods -> Online mods repository
or
        CLient -> Public Serverslist

Comparision with the first reporter shows that the error messages are
for some reason not identical.

Best regards,

     Andi


---------------------
Console output below, only the last lines seem to be relevant:

    terminate called after throwing an instance of 'LuaError'
      what():  C++ exception
    Aborted


andi@flashgordon:~$ minetest --trace
Enabling trace level debug output
22:44:40: INFO[main]: Using system-wide paths (NOT RUN_IN_PLACE)
22:44:40: INFO[main]: path_share = /usr/share/games/minetest
22:44:40: INFO[main]: path_user  = /home/andi/.minetest
22:44:40: INFO[main]: minetest with SER_FMT_VER_HIGHEST_READ=26,
VER=0.4.9 BUILD_TYPE=Release RUN_IN_PLACE=0 USE_GETTEXT=1 USE_SOUND=1
USE_CURL=1 USE_FREETYPE=1 USE_LUAJIT=1
STATIC_SHAREDIR=/usr/share/games/minetest
22:44:40: INFO[main]: logfile    = /home/andi/.minetest/debug.txt
22:44:40: VERBOSE[main]: httpfetch_init: parallel_limit=8
22:44:40: INFO[main]: Message locale is now set to:
LC_CTYPE=en_US.UTF-8;LC_NUMERIC=C;LC_TIME=en_US.UTF-8;LC_COLLATE=en_US.UTF-8;LC_MONETARY=en_US.UTF-8;LC_MESSAGES=en_US.UTF-8;LC_PAPER=en_US.UTF-8;LC_NAME=en_US.UTF-8;LC_ADDRESS=en_US.UTF-8;LC_TELEPHONE=en_US.UTF-8;LC_MEASUREMENT=en_US.UTF-8;LC_IDENTIFICATION=en_US.UTF-8
Irrlicht Engine version 1.8.1
Linux 3.12-1-amd64 #1 SMP Debian 3.12.6-2 (2013-12-29) x86_64
Using renderer: OpenGL 3.0
Mesa DRI Intel(R) Haswell Desktop : Intel Open Source Technology Center
OpenGL driver version is 1.2 or better.
GLSL version: 1.3
22:44:40: INFO[main]: text_height=16
22:44:40: INFO[main]: Searching worlds...
22:44:40: INFO[main]:   In /home/andi/.minetest/worlds:
22:44:40: INFO[main]: 0 found.
22:44:40: INFO[main]: Waiting for other menus
22:44:40: INFO[main]: Waited for other menus
22:44:40: INFO[main]: Audio: Initializing...
22:44:40: INFO[main]: Audio: Vorbis extension NOT present
22:44:40: INFO[main]: Audio: Initialized: OpenAL 1.1 ALSOFT 1.14, using OpenAL Soft
22:44:40: INFO[main]: GUIEngine: Initializing Lua
22:44:40: INFO[AsyncWorkerThread_0]: Initializing Asynchronous environment
22:44:40: INFO[AsyncWorkerThread_1]: Initializing Asynchronous environment
22:44:40: INFO[AsyncWorkerThread_2]: Initializing Asynchronous environment
22:44:40: INFO[main]: SCRIPTAPI: initialized mainmenu modules
22:44:40: INFO[AsyncWorkerThread_3]: Initializing Asynchronous environment
22:44:40: VERBOSE[main]: Loading and running script from /usr/share/games/minetest/builtin/mainmenu.lua
22:44:40: INFO[main]: Searching worlds...
22:44:40: INFO[main]:   In /home/andi/.minetest/worlds:
22:44:40: INFO[main]: 0 found.
Loaded texture: /usr/share/games/minetest/games/minetest_game/menu/header.png
22:44:40: VERBOSE[main]: OpenALSoundManager::maintain(): 0 playing sounds, 0 sound names loaded
22:44:40: INFO[main]: Audio: Error opening /home/andi/.minetest/sounds/main_menu.0.ogg for decoding
22:44:40: INFO[main]: Audio: Error opening /home/andi/.minetest/sounds/main_menu.1.ogg for decoding
22:44:40: INFO[main]: Audio: Error opening /home/andi/.minetest/sounds/main_menu.2.ogg for decoding
22:44:40: INFO[main]: Audio: Error opening /home/andi/.minetest/sounds/main_menu.3.ogg for decoding
22:44:40: INFO[main]: Audio: Error opening /home/andi/.minetest/sounds/main_menu.4.ogg for decoding
22:44:40: INFO[main]: Audio: Error opening /home/andi/.minetest/sounds/main_menu.5.ogg for decoding
22:44:40: INFO[main]: Audio: Error opening /home/andi/.minetest/sounds/main_menu.6.ogg for decoding
22:44:40: INFO[main]: Audio: Error opening /home/andi/.minetest/sounds/main_menu.7.ogg for decoding
22:44:40: INFO[main]: Audio: Error opening /home/andi/.minetest/sounds/main_menu.8.ogg for decoding
22:44:40: INFO[main]: Audio: Error opening /home/andi/.minetest/sounds/main_menu.9.ogg for decoding
22:44:40: INFO[main]: Audio: Error opening /home/andi/.minetest/sounds/main_menu.ogg for decoding
22:44:40: INFO[main]: Audio: Error opening /usr/share/games/minetest/sounds/main_menu.0.ogg for decoding
22:44:40: INFO[main]: Audio: Error opening /usr/share/games/minetest/sounds/main_menu.1.ogg for decoding
22:44:40: INFO[main]: Audio: Error opening /usr/share/games/minetest/sounds/main_menu.2.ogg for decoding
22:44:40: INFO[main]: Audio: Error opening /usr/share/games/minetest/sounds/main_menu.3.ogg for decoding
22:44:40: INFO[main]: Audio: Error opening /usr/share/games/minetest/sounds/main_menu.4.ogg for decoding
22:44:40: INFO[main]: Audio: Error opening /usr/share/games/minetest/sounds/main_menu.5.ogg for decoding
22:44:40: INFO[main]: Audio: Error opening /usr/share/games/minetest/sounds/main_menu.6.ogg for decoding
22:44:40: INFO[main]: Audio: Error opening /usr/share/games/minetest/sounds/main_menu.7.ogg for decoding
22:44:40: INFO[main]: Audio: Error opening /usr/share/games/minetest/sounds/main_menu.8.ogg for decoding
22:44:40: INFO[main]: Audio: Error opening /usr/share/games/minetest/sounds/main_menu.9.ogg for decoding
22:44:40: INFO[main]: Audio: Error opening /usr/share/games/minetest/sounds/main_menu.ogg for decoding
22:44:40: INFO[main]: OpenALSoundManager: "main_menu" not found.
Loaded texture: /usr/share/games/minetest/games/minetest_game/menu/icon.png
Loaded texture: /usr/share/games/minetest/games/minimal/menu/icon.png
22:44:43: INFO[main]: GUIFormSpecMenu: Not allowing focus change.
terminate called after throwing an instance of 'LuaError'
  what():  C++ exception
Aborted
andi@flashgordon:~$

Message #32 received at 733974@bugs.debian.org (full text, mbox, reply):

From: Baltix Linux <baltix@gmail.com>

To: Martin Quinson <martin.quinson@loria.fr>, 733974@bugs.debian.org

Subject: Please tell upstream bugreport for Minetest crashes when selecting Client->"Public Serverlist" and Mods->"Online mod repository"

Date: Fri, 14 Feb 2014 14:58:42 +0200

Hi Martin,

Sun, 2 Feb 2014 21:35 Martin Quinson wrote:
> Actually, I reported the bug to upstream, seeking for help.
> I'm myself useless and clueless here.

I can't find your bugreport in upstream BTS:
https://github.com/minetest/minetest/issues
Please tell upstream bugreport URL for Minetest crashes when selecting
Client->"Public  Serverlist" and Mods->"Online mod repository"

Thanks,
-- 
Prekyba kompiuteriais su Linux OS - http://tinklas.eu/prekyba
Naudokite laisvą Linux operacinę sistemą savo kompiuteryje -
http://baltix.akl.lt
Use Baltix GNU/Linux OS !

Message #37 received at 733974@bugs.debian.org (full text, mbox, reply):

From: Martin Quinson <martin.quinson@loria.fr>

To: Baltix Linux <baltix@gmail.com>, 733974@bugs.debian.org

Subject: Re: Bug#733974: Please tell upstream bugreport for Minetest crashes when selecting Client->"Public Serverlist" and Mods->"Online mod repository"

Date: Fri, 14 Feb 2014 15:22:34 +0100

[Message part 1 (text/plain, inline)]

On Fri, Feb 14, 2014 at 02:58:42PM +0200, Baltix Linux wrote:
> Hi Martin,
> 
> Sun, 2 Feb 2014 21:35 Martin Quinson wrote:
> > Actually, I reported the bug to upstream, seeking for help.
> > I'm myself useless and clueless here.
> 
> I can't find your bugreport in upstream BTS:
> https://github.com/minetest/minetest/issues
> Please tell upstream bugreport URL for Minetest crashes when selecting
> Client->"Public  Serverlist" and Mods->"Online mod repository"

I used the IRC for that so far, but I just opened
https://github.com/minetest/minetest/issues/1147

Thanks for the reminder,
Mt

http://irc.minetest.ru/minetest-dev/2014-02-02



-- 
The nice thing about standards is that there are so many to choose from.  And if you
really don't like all the standards you just have to wait another year until the one
arises you are looking for.     -- Tannenbaum, 'Introduction to Computer Networks'

[signature.asc (application/pgp-signature, inline)]

Message #50 received at 733974@bugs.debian.org (full text, mbox, reply):

From: ERIIX Blaike <eriixblaike@gmail.com>

To: 733974@bugs.debian.org

Subject: Partial diagnosis.

Date: Wed, 11 Jun 2014 04:29:46 -0700

I can't say I'm entirely happy with my result so far, but I have
figured out the general thrust of this bug.  Minetest seems
incompatible with the libjsoncpp in Debian.  Solving this issue
properly may be a mess, but the proof for the location of the problem
also suggests a temporary workaround.

If you have the Minetest source available (Upstream source --
"git://github.com/minetest/minetest"), it contains "src/json/".  From
within, you can simply "g++ -shared jsoncpp.cpp -o libjsoncpp.so.0"
and use LD_LIBRARY_PATH to force the current Debian package to load
that preferentially to the system-wide version.  Suddenly, the server
list and online mod repo work as they do upstream.

An obvious temporary solution is to keep the bundled libjsoncpp, but a
long-term solution will require figuring out why the system libjsoncpp
fails and either making it work correctly or making Minetest use it
correctly, dependent on which is behaving improperly.

I haven't looked too deeply into which is responsible, though I'd
suspect Minetest.  I'll look into it later, but it is quite late right
now, so I figured I'd send in what I have and get to bed.

Message #55 received at 733974@bugs.debian.org (full text, mbox, reply):

From: Martin Quinson <martin.quinson@loria.fr>

To: ERIIX Blaike <eriixblaike@gmail.com>, 733974@bugs.debian.org

Subject: Re: Bug#733974: Partial diagnosis.

Date: Wed, 11 Jun 2014 14:54:58 +0200

[Message part 1 (text/plain, inline)]

WOOT! thanks a lot! your analysis is incredibly useful. It makes
perfectly sense. I feel bad about using the embeeded version of the
lib, and I'd love if you could find some time to investigate further.
I'm rather underwater right now, unfortunately.

Thanks, Mt.

On Wed, Jun 11, 2014 at 04:29:46AM -0700, ERIIX Blaike wrote:
> I can't say I'm entirely happy with my result so far, but I have
> figured out the general thrust of this bug.  Minetest seems
> incompatible with the libjsoncpp in Debian.  Solving this issue
> properly may be a mess, but the proof for the location of the problem
> also suggests a temporary workaround.
> 
> If you have the Minetest source available (Upstream source --
> "git://github.com/minetest/minetest"), it contains "src/json/".  From
> within, you can simply "g++ -shared jsoncpp.cpp -o libjsoncpp.so.0"
> and use LD_LIBRARY_PATH to force the current Debian package to load
> that preferentially to the system-wide version.  Suddenly, the server
> list and online mod repo work as they do upstream.
> 
> An obvious temporary solution is to keep the bundled libjsoncpp, but a
> long-term solution will require figuring out why the system libjsoncpp
> fails and either making it work correctly or making Minetest use it
> correctly, dependent on which is behaving improperly.
> 
> I haven't looked too deeply into which is responsible, though I'd
> suspect Minetest.  I'll look into it later, but it is quite late right
> now, so I figured I'd send in what I have and get to bed.
> 

-- 
Je préfère glisser ma peau sous les draps pour le plaisir des sens
que la risquer sous les drapeaux pour le prix de l'essence.

[signature.asc (application/pgp-signature, inline)]

Message #60 received at 733974@bugs.debian.org (full text, mbox, reply):

From: "mbekkema97 ." <mbekkema97@gmail.com>

To: Martin Quinson <martin.quinson@loria.fr>, 733974@bugs.debian.org

Subject: Re: Bug#733974: Partial diagnosis.

Date: Sun, 22 Jun 2014 02:24:18 +1000

I discovered that adding the line:
      return valueToString( value_.int_ );

underneath line 668 of
jsoncpp-src-0.6.0-rc2/src/lib_json/json_value.cpp of the debian
libjsoncpp source package fixes server list, but modstore only loads
titles and gives these errors on the console:
02:20:49: ERROR[AsyncWorkerThread_3]: readModStoreModDetails: missing author id
02:20:49: ERROR[AsyncWorkerThread_3]: readModStoreModDetails: missing
author username
02:20:49: ERROR[AsyncWorkerThread_3]: readModStoreModDetails: missing license id
02:20:49: ERROR[AsyncWorkerThread_3]: readModStoreModDetails: missing
license short
02:20:49: ERROR[AsyncWorkerThread_3]: readModStoreModDetails: missing id
02:20:49: ERROR[AsyncWorkerThread_3]: readModStoreModDetails: missing title
02:20:49: ERROR[AsyncWorkerThread_3]: readModStoreModDetails: missing basename

Message #65 received at 733974@bugs.debian.org (full text, mbox, reply):

From: Matthew Bekkema <mbekkema97@gmail.com>

To: Martin Quinson <martin.quinson@loria.fr>, 733974@bugs.debian.org

Subject: Re: Bug#733974: Partial diagnosis.

Date: Sun, 22 Jun 2014 17:01:15 +1000

[Message part 1 (text/plain, inline)]

I found out what changes to libjsoncpp allow everything to work. I
don't have access to Debian Sid at the moment so I've only tested this
on Wheezy.

[patch (application/octet-stream, attachment)]

Message #70 received at 733974@bugs.debian.org (full text, mbox, reply):

From: Martin Quinson <martin.quinson@loria.fr>

To: Matthew Bekkema <mbekkema97@gmail.com>, 733974@bugs.debian.org

Subject: Re: Bug#733974: Partial diagnosis.

Date: Thu, 3 Jul 2014 09:02:18 +0200

[Message part 1 (text/plain, inline)]

On Sun, Jun 22, 2014 at 05:01:15PM +1000, Matthew Bekkema wrote:
> I found out what changes to libjsoncpp allow everything to work. I
> don't have access to Debian Sid at the moment so I've only tested this
> on Wheezy.

I just opened a bug against libjsoncpp (#753560) to ask for the
inclusion of that patch. Thanks very much for sorting that out, your
precious work is really appreciated.

Bye, Mt.

-- 
I had a headache just by looking at the data structures of your linear-time 
optimal algorithm. No doubt an exhaustive algorithm would be more efficient in
practice. -- Bastard Reviewer From Hell

[signature.asc (application/pgp-signature, inline)]

Message #79 received at 733974@bugs.debian.org (full text, mbox, reply):

From: Markus Koschany <apo@gambaru.de>

To: 733974@bugs.debian.org, 753560@bugs.debian.org

Subject: Crash when browsing the public server list in Minetest

Date: Mon, 17 Nov 2014 19:20:34 +0100

[Message part 1 (text/plain, inline)]

On Thu, 3 Jul 2014 09:02:18 +0200 Martin Quinson
<martin.quinson@loria.fr> wrote:
> On Sun, Jun 22, 2014 at 05:01:15PM +1000, Matthew Bekkema wrote:
> > I found out what changes to libjsoncpp allow everything to work. I
> > don't have access to Debian Sid at the moment so I've only tested this
> > on Wheezy.
> 
> I just opened a bug against libjsoncpp (#753560) to ask for the
> inclusion of that patch. Thanks very much for sorting that out, your
> precious work is really appreciated.
> 
> Bye, Mt.


Hi,

I have recently spent some time to learn more about Minetest. This bug
is a serious issue because it prevents players from browsing the public
server list. Since Minetest is a multiplayer online game, this is really
a show-stopper for a lot of people including myself. Obviously the bug
itself lies in libjsoncpp.

I am going to merge #733974 with #753560 and raise the severity to
serious. The patch from Matthew Bekkema works. I am attaching the
debdiff to this bug report.

Regards,

Markus

[libjsoncpp.debdiff (text/plain, attachment)]

[signature.asc (application/pgp-signature, attachment)]

Message #102 received at 733974@bugs.debian.org (full text, mbox, reply):

From: Martin Quinson <martin.quinson@loria.fr>

To: Markus Koschany <apo@gambaru.de>, 769965@bugs.debian.org

Cc: 733974@bugs.debian.org

Subject: Re: Bug#769965: minetest: please support init.d scripts and a global configuration file

Date: Wed, 19 Nov 2014 09:27:21 +0100

[Message part 1 (text/plain, inline)]

On Tue, Nov 18, 2014 at 11:16:51PM +0100, Markus Koschany wrote:
> 
> I agree that https://bugs.debian.org/733974 is a problem that should be
> fixed for Jessie. I suggest we wait a few more days until Friday and
> give the maintainer a chance to respond to the bug report. However the
> package looks rather neglected to me hence I would like to suggest that
> you just grab my attached debdiff from #733974 and NMU the package. I'm
> confident that the changes qualify for an unblock by the release team. I
> hope your last upload to unstable didn't introduce to many changes then
> I think we can get both, the Hurd FTBFS fix and the libjsoncpp fix, past
> the release team.

I just uploaded the fixed package for libjsoncpp. Thanks for preparing
the debdiff, I always feel unconfident to work with packages that are
neither in a git not a svn.

I confirmed that this is enough to fix the problem in minetest. We
just need to get the new libjsoncpp in Jessie, the current minetest
can stay.

If you quickly know how to request that unblock, I'd be really
grateful if you could do so, please.
 
> I think we should get in contact with Dominik George who already
> packaged some Minetest mods in the past. Perhaps he might be interested
> to package one of the other popular mods from this list
> 
> http://minetest.net/mods
> 
> He probably just needs a sponsor. Michael Dorrington also advertised his
> interest in packaging some Minetest mods on the mailing list not very
> long ago.

I'm ok with helping if I'm contacted, but I wont seek for more work
myself ;)

[...]

> I intend to finish the systemd integration until Friday. I think we
> could try to achieve something similar like Fedora and use instantiated
> services, so that you could run multiple Minetest servers at the same time.
> 
> http://pkgs.fedoraproject.org/cgit/minetest.git/tree/minetest@.service
> 
> I will also add logrotate support to the server package, so that all log
> files are rotated on a regular basis.

That'd be perfect, indeed.

The last issue that naggles me about minetest is whether we should
switch to freeminer at some point. The upstream community have forked,
and I cannot decide which version is the best one. We should maybe
package both, but that also seem to unnecessary burden the Debian
archive. There too, your advice would be very welcome.

Thanks for your time,
Mt.

-- 
Technique sans politique n'est que renforcement des nuisants.
     -- Rabelais (traduction en français moderne).

[signature.asc (application/pgp-signature, inline)]

Message #103 received at 753560-close@bugs.debian.org (full text, mbox, reply):

From: Martin Quinson <mquinson@debian.org>

To: 753560-close@bugs.debian.org

Subject: Bug#753560: fixed in libjsoncpp 0.6.0~rc2-3.1

Date: Wed, 19 Nov 2014 09:56:40 +0000

Source: libjsoncpp
Source-Version: 0.6.0~rc2-3.1

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

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

Thank you for reporting the bug, which will now be closed.  If you
have further comments please address them to 753560@bugs.debian.org,
and the maintainer will reopen the bug report if appropriate.

Debian distribution maintenance software
pp.
Martin Quinson <mquinson@debian.org> (supplier of updated libjsoncpp package)

(This message was generated automatically at their request; if you
believe that there is a problem with it please contact the archive
administrators by mailing ftpmaster@ftp-master.debian.org)


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

Format: 1.8
Date: Wed, 19 Nov 2014 09:14:06 +0100
Source: libjsoncpp
Binary: libjsoncpp0 libjsoncpp-dev
Architecture: source amd64
Version: 0.6.0~rc2-3.1
Distribution: unstable
Urgency: medium
Maintainer: José Luis Segura Lucas <josel.segura@gmx.es>
Changed-By: Martin Quinson <mquinson@debian.org>
Description:
 libjsoncpp-dev - Library for reading and writing JSON for C++ (devel files)
 libjsoncpp0 - Library for reading and writing JSON for C++
Closes: 753560
Changes:
 libjsoncpp (0.6.0~rc2-3.1) unstable; urgency=medium
 .
   [ Markus Koschany ]
   * Non-maintainer upload.
   * Add value-as-string.patch. (Closes: #753560)
     Fix conversion of numerical values to string that caused crashes in
     reverse-dependencies such as Minetest.
Checksums-Sha1:
 e452a1a046a59532a0573f94ad9440519eaf418d 1891 libjsoncpp_0.6.0~rc2-3.1.dsc
 5a6b180ed0ad20a29b0b0a2a88167a4733c8ae2a 4852 libjsoncpp_0.6.0~rc2-3.1.debian.tar.xz
 41e60e93412823800ae7ce336d52c689f8f25dbf 71826 libjsoncpp0_0.6.0~rc2-3.1_amd64.deb
 7724f96a37ed008f4427ffbbc0d83930dbd75ee4 216536 libjsoncpp-dev_0.6.0~rc2-3.1_amd64.deb
Checksums-Sha256:
 09ac4c902e575276abcfc876191f5b6d96f2cd61f8d13a9653f8ec7c7faaf9b0 1891 libjsoncpp_0.6.0~rc2-3.1.dsc
 d2abe156f2c2e57661da480b3c8fd4a4618f78597742e7195907c28c4ba68de1 4852 libjsoncpp_0.6.0~rc2-3.1.debian.tar.xz
 5daae105b5d8300661ef5c18ead332b513913bf332296b38f3008f247cce9fb0 71826 libjsoncpp0_0.6.0~rc2-3.1_amd64.deb
 b050d84e611c6283beb26a59f229063bf8ca76848a7ab5ecf70acef965f0904c 216536 libjsoncpp-dev_0.6.0~rc2-3.1_amd64.deb
Files:
 6be9166cc752466c531a199ffb23b285 71826 libs extra libjsoncpp0_0.6.0~rc2-3.1_amd64.deb
 778463c64affdd4c7a738eee7b7387a4 216536 libdevel extra libjsoncpp-dev_0.6.0~rc2-3.1_amd64.deb
 e99c8903001ca8f8adcead25c9116121 1891 libs extra libjsoncpp_0.6.0~rc2-3.1.dsc
 e9341b399247f0c27833fb55681e847c 4852 libs extra libjsoncpp_0.6.0~rc2-3.1.debian.tar.xz

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2
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=MVWX
-----END PGP SIGNATURE-----

Message #108 received at 733974@bugs.debian.org (full text, mbox, reply):

From: Markus Koschany <apo@gambaru.de>

To: Martin Quinson <martin.quinson@loria.fr>

Cc: 769965@bugs.debian.org, 733974@bugs.debian.org

Subject: Re: Bug#769965: minetest: please support init.d scripts and a global configuration file

Date: Wed, 19 Nov 2014 15:07:08 +0100

[Message part 1 (text/plain, inline)]

On 19.11.2014 09:27, Martin Quinson wrote:
> On Tue, Nov 18, 2014 at 11:16:51PM +0100, Markus Koschany wrote:
>>
>> I agree that https://bugs.debian.org/733974 is a problem that should be
>> fixed for Jessie. I suggest we wait a few more days until Friday and
>> give the maintainer a chance to respond to the bug report. However the
>> package looks rather neglected to me hence I would like to suggest that
>> you just grab my attached debdiff from #733974 and NMU the package. I'm
>> confident that the changes qualify for an unblock by the release team. I
>> hope your last upload to unstable didn't introduce to many changes then
>> I think we can get both, the Hurd FTBFS fix and the libjsoncpp fix, past
>> the release team.
> 
> I just uploaded the fixed package for libjsoncpp. Thanks for preparing
> the debdiff, I always feel unconfident to work with packages that are
> neither in a git not a svn.

Thank you.


> I confirmed that this is enough to fix the problem in minetest. We
> just need to get the new libjsoncpp in Jessie, the current minetest
> can stay.
> 
> If you quickly know how to request that unblock, I'd be really
> grateful if you could do so, please.

It's quite easy

reportbug release.debian.org

and then a short explanation why you want that unblock + the debdiff
against the version in testing. I'll take care of it.



>> I think we should get in contact with Dominik George who already
>> packaged some Minetest mods in the past. Perhaps he might be interested
>> to package one of the other popular mods from this list
>>
>> http://minetest.net/mods
>>
>> He probably just needs a sponsor. Michael Dorrington also advertised his
>> interest in packaging some Minetest mods on the mailing list not very
>> long ago.
> 
> I'm ok with helping if I'm contacted, but I wont seek for more work
> myself ;)


Ok, fair enough.

[...]

> The last issue that naggles me about minetest is whether we should
> switch to freeminer at some point. The upstream community have forked,
> and I cannot decide which version is the best one. We should maybe
> package both, but that also seem to unnecessary burden the Debian
> archive. There too, your advice would be very welcome.
> 

I think we should stick with Minetest for the time being because I
perceive the goals of freeminer as rather vague and I don't see any kind
of big improvements that would justify a switch. I'm also quite
reluctant when it comes to forks. They should really prove that they are
the better alternative and I can't see that currently with freeminer.
There is also only one server at the moment

http://servers.freeminer.org/

and the Minetest community is much more vital and active.

> Thanks for your time,
> Mt.

Regards,

Markus

[signature.asc (application/pgp-signature, attachment)]

Send a report that this bug log contains spam.