Minetest 0.4.13

[TenPlus1]

2015-08-31 10:22:55: ERROR[Main]: ServerError: Lua: Runtime error from mod 'test' in callback on_prejoinplayer(): /home/xanadu/.minetest/mods/test/init.lua:52: attempt to get length of local 'info' (a nil value)

the 'local info = minetest.get_player_information("TenPlus1")' table contains nothing at all ?!?!

[est31]

Ah, I see, the problem was the "pre" in "prejoinplayer". The following code would work (I've tested it), but there is an engine bug right now, leading in a crash, I will fix it, then you can try it once updated:

Code: Select all

minetest.register_on_joinplayer(function(player)
	local pname = "TenPlus1"
	if player:get_player_name() == pname then
	local info = minetest.get_player_information(pname)
		minetest.chat_send_player(pname, "Hi, the output is: " .. dump(info.connection_uptime) .. " and " .. dump(info.version_string))
	end
end)

[Linuxdirk]

@Linuxdirk can you try to do an in-place build, without installing? So doing:

Code: Select all

git clone https://github.com/minetest/minetest.git && cd minetest && cmake -DRUN_IN_PLACE=1 && make -j 8 && LANGUAGE=de_DE bin/minetest

It’s all in English and it now says USE_GETTEXT=0.

What we found out in the other thread (besides marking bugfixes as low-priority enhancements) is that minetest expects the locales to be in an unusual place instead of the default /usr/share/locale tree.

When linking the whole locale tree in /usr/share/minetest/locale the installed version of the games properly uses the translation. So I guess iths either an ignored configuration option (a bug in Minetest) a not set compiling option (a bug in my distributions build file for the package) or something else that behaves weird.

Anyway, here’s the log file of the verbose startup I did after compiling it with your one-liner.

Code: Select all

$ LANGUAGE=de_DE bin/minetest --verbose
2015-08-31 19:12:58: INFO[Main]: Using relative paths (RUN_IN_PLACE)
2015-08-31 19:12:58: INFO[Main]: Detected share path: /home/dirk/mt/minetest/bin/..
2015-08-31 19:12:58: INFO[Main]: Detected user path: /home/dirk/mt/minetest/bin/..
2015-08-31 19:12:58: INFO[Main]: path_share = /home/dirk/mt/minetest/bin/..
2015-08-31 19:12:58: INFO[Main]: path_user  = /home/dirk/mt/minetest/bin/..
2015-08-31 19:12:58: INFO[Main]: minetest with SER_FMT_VER_HIGHEST_READ=26, VER=0.4.13-dev-714b68a BUILD_TYPE=Release RUN_IN_PLACE=1 USE_GETTEXT=0 USE_SOUND=1 USE_CURL=1 USE_FREETYPE=1 USE_LUAJIT=0 STATIC_SHAREDIR="."
2015-08-31 19:12:58: INFO[Main]: logfile = debug.txt
2015-08-31 19:12:58: VERBOSE[Main]: httpfetch_init: parallel_limit=8
2015-08-31 19:12:58: INFO[Main]: Message locale is now set to: LC_CTYPE=de_DE.utf8;LC_NUMERIC=C;LC_TIME=en_DK.utf8;LC_COLLATE=de_DE.utf8;LC_MONETARY=de_DE.utf8;LC_MESSAGES=de_DE.utf8;LC_PAPER=de_DE.utf8;LC_NAME=de_DE.utf8;LC_ADDRESS=de_DE.utf8;LC_TELEPHONE=de_DE.utf8;LC_MEASUREMENT=de_DE.utf8;LC_IDENTIFICATION=de_DE.utf8
2015-08-31 19:12:58: VERBOSE[Main]: Determining world path
2015-08-31 19:12:58: INFO[Main]: Searching worlds...
2015-08-31 19:12:58: INFO[Main]:   In /home/dirk/mt/minetest/bin/../worlds: 
2015-08-31 19:12:58: INFO[Main]: 0 found.
2015-08-31 19:12:58: INFO[Main]: Creating default world at [/home/dirk/mt/minetest/bin/../worlds/world]
2015-08-31 19:12:58: VERBOSE[Main]: Determining gameid/gamespec
2015-08-31 19:12:58: INFO[Main]: Using default gameid []
2015-08-31 19:12:58: ERROR[Main]: Subgame specified in default_game [minetest] is invalid.
2015-08-31 19:12:58: INFO[Main]: Using commanded world path [/home/dirk/mt/minetest/bin/../worlds/world]
2015-08-31 19:12:58: INFO[Main]: Irrlicht: Irrlicht Engine version 1.8.1
2015-08-31 19:12:58: INFO[Main]: Irrlicht: Linux 4.1.6-1-ARCH #1 SMP PREEMPT Mon Aug 17 08:52:28 CEST 2015 x86_64
2015-08-31 19:12:58: INFO[Main]: Irrlicht: Using renderer: OpenGL 3.3.0
2015-08-31 19:12:58: INFO[Main]: Irrlicht: GeForce 9600 GT/PCIe/SSE2: NVIDIA Corporation
2015-08-31 19:12:58: INFO[Main]: Irrlicht: OpenGL driver version is 1.2 or better.
2015-08-31 19:12:58: INFO[Main]: Irrlicht: GLSL version: 3.3
2015-08-31 19:12:58: INFO[Main]: text_height=18
2015-08-31 19:12:58: INFO[Main]: Waiting for other menus
2015-08-31 19:12:58: INFO[Main]: Waited for other menus
2015-08-31 19:12:58: INFO[Main]: Audio: Initializing...
2015-08-31 19:12:58: INFO[Main]: Audio: Vorbis extension NOT present
2015-08-31 19:12:58: INFO[Main]: Audio: Initialized: OpenAL 1.1 ALSOFT 1.16.0, using OpenAL Soft
2015-08-31 19:12:58: INFO[Main]: GUIEngine: Initializing Lua
2015-08-31 19:12:58: VERBOSE[AsyncWorker-0]: Loading and running script from /home/dirk/mt/minetest/bin/../builtin/init.lua
2015-08-31 19:12:58: INFO[AsyncWorker-0]: Initializing Asynchronous environment
2015-08-31 19:12:58: VERBOSE[AsyncWorker-1]: Loading and running script from /home/dirk/mt/minetest/bin/../builtin/init.lua
2015-08-31 19:12:58: INFO[AsyncWorker-1]: Initializing Asynchronous environment
2015-08-31 19:12:58: VERBOSE[AsyncWorker-2]: Loading and running script from /home/dirk/mt/minetest/bin/../builtin/init.lua
2015-08-31 19:12:58: VERBOSE[AsyncWorker-3]: Loading and running script from /home/dirk/mt/minetest/bin/../builtin/init.lua
2015-08-31 19:12:58: INFO[AsyncWorker-2]: Initializing Asynchronous environment
2015-08-31 19:12:58: INFO[Main]: SCRIPTAPI: Initialized main menu modules
2015-08-31 19:12:58: VERBOSE[Main]: Loading and running script from /home/dirk/mt/minetest/bin/../builtin/init.lua
2015-08-31 19:12:58: INFO[AsyncWorker-3]: Initializing Asynchronous environment
2015-08-31 19:12:58: INFO[Main]: Searching worlds...
2015-08-31 19:12:58: INFO[Main]:   In /home/dirk/mt/minetest/bin/../worlds: 
2015-08-31 19:12:58: INFO[Main]: 0 found.
2015-08-31 19:12:58: VERBOSE[Main]: OpenALSoundManager::maintain(): 0 playing sounds, 0 sound names loaded
2015-08-31 19:12:58: INFO[Main]: Audio: Error opening /home/dirk/mt/minetest/bin/../sounds/main_menu.0.ogg for decoding
2015-08-31 19:12:58: INFO[Main]: Audio: Error opening /home/dirk/mt/minetest/bin/../sounds/main_menu.1.ogg for decoding
2015-08-31 19:12:58: INFO[Main]: Audio: Error opening /home/dirk/mt/minetest/bin/../sounds/main_menu.2.ogg for decoding
2015-08-31 19:12:58: INFO[Main]: Audio: Error opening /home/dirk/mt/minetest/bin/../sounds/main_menu.3.ogg for decoding
2015-08-31 19:12:58: INFO[Main]: Audio: Error opening /home/dirk/mt/minetest/bin/../sounds/main_menu.4.ogg for decoding
2015-08-31 19:12:58: INFO[Main]: Audio: Error opening /home/dirk/mt/minetest/bin/../sounds/main_menu.5.ogg for decoding
2015-08-31 19:12:58: INFO[Main]: Audio: Error opening /home/dirk/mt/minetest/bin/../sounds/main_menu.6.ogg for decoding
2015-08-31 19:12:58: INFO[Main]: Audio: Error opening /home/dirk/mt/minetest/bin/../sounds/main_menu.7.ogg for decoding
2015-08-31 19:12:58: INFO[Main]: Audio: Error opening /home/dirk/mt/minetest/bin/../sounds/main_menu.8.ogg for decoding
2015-08-31 19:12:58: INFO[Main]: Audio: Error opening /home/dirk/mt/minetest/bin/../sounds/main_menu.9.ogg for decoding
2015-08-31 19:12:58: INFO[Main]: Audio: Error opening /home/dirk/mt/minetest/bin/../sounds/main_menu.ogg for decoding
2015-08-31 19:12:58: INFO[Main]: OpenALSoundManager: "main_menu" not found.
2015-08-31 19:13:04: INFO[Main]: Audio: Deinitializing...
2015-08-31 19:13:04: INFO[Main]: Audio: Deinitialized.
2015-08-31 19:13:04: INFO[Main]: GUIEngine: Deinitializing scripting
2015-08-31 19:13:04: INFO[Main]: Searching worlds...
2015-08-31 19:13:04: INFO[Main]:   In /home/dirk/mt/minetest/bin/../worlds: 
2015-08-31 19:13:04: INFO[Main]: 0 found.
2015-08-31 19:13:04: INFO[Main]: Selected world: [--world parameter] [new] [/home/dirk/mt/minetest/bin/../worlds/world]
2015-08-31 19:13:04: ERROR[Main]: Provided world path doesn't exist: /home/dirk/mt/minetest/bin/../worlds/world
2015-08-31 19:13:04: VERBOSE[Main]: httpfetch_cleanup: cleaning up

[est31]

Lets continue discussion there.

[Linuxdirk]

As requested here I applied pull/3108 to the latest available Git version and compiled with the following options:

Code: Select all

cmake . \
    -DCMAKE_INSTALL_PREFIX=/usr \
    -DBUILD_SERVER=0 \
    -DENABLE_GETTEXT=1 \
    -DENABLE_FREETYPE=1 \
    -DENABLE_LEVELDB=0 \
    -DENABLE_REDIS=0 \
    -DCUSTOM_LOCALEDIR=/usr/share/locale/

Then I started it in verbose mode (truncated):

Code: Select all

$ ./minetest --verbose
2015-09-03 20:18:55: INFO[Main]: minetest with SER_FMT_VER_HIGHEST_READ=26, VER=0.4.13-dev-6c5bc14-dirty BUILD_TYPE=Release RUN_IN_PLACE=0 USE_GETTEXT=1 USE_SOUND=1 USE_CURL=1 USE_FREETYPE=1 USE_LUAJIT=0 STATIC_SHAREDIR="/usr/share/minetest"
2015-09-03 20:18:55: INFO[Main]: Message locale is now set to: LC_CTYPE=de_DE.utf8;LC_NUMERIC=C;LC_TIME=en_DK.utf8;LC_COLLATE=de_DE.utf8;LC_MONETARY=de_DE.utf8;LC_MESSAGES=de_DE.utf8;LC_PAPER=de_DE.utf8;LC_NAME=de_DE.utf8;LC_ADDRESS=de_DE.utf8;LC_TELEPHONE=de_DE.utf8;LC_MEASUREMENT=de_DE.utf8;LC_IDENTIFICATION=de_DE.utf8

All strings are properly translated to German.

Just for comparison I compiled the latest available Git version with the same options but without applying pull/3108 and started it the same way. … Even there the strings are all properly translated(?!)

[est31]

Just for comparison I compiled the latest available Git version with the same options but without applying pull/3108 and started it the same way. … Even there the strings are all properly translated(?!)

It might be that it has found the locale files another way, maybe from the locale/ subdir of your working directory, or from /usr/share. In order to test you'll have to repeat the steps of your custom PKGBUILD, manually or automatically.

[kaeza]

2015-08-31 10:22:55: ERROR[Main]: ServerError: Lua: Runtime error from mod 'test' in callback on_prejoinplayer(): /home/xanadu/.minetest/mods/test/init.lua:52: attempt to get length of local 'info' (a nil value)

the 'local info = minetest.get_player_information("TenPlus1")' table contains nothing at all ?!?!

I've seen this asked a few times already, and is something that maybe needs clarifying somewhere.

The fact is, `on_prejoinplayer` callbacks are run before the game actually creates a player object. That is, at that point, there's not even a player by that name as far as the engine is concerned, so any other function that depends on a player object being available (whether internal to the engine or exposed to Lua) returns no meaningful information.

In fact, you should not even count on the player name being "valid" since the engine has not even performed authentication yet.

Its main purpose is quick filtering player names/IPs for whatever reason from Lua (e.g. the xban, no_guests, and similar mods).

If you need any other info about a player other than name or source IP, you need to use the regular `on_joinplayer` callbacks as pointed by est.

[BrunoMine]

This version is cool (VERY COOL)

Spoiler

[benrob0329]

This version is cool (VERY COOL)

Spoiler

Lol

[paramat]

Ugh, please remove those images, especially since this is a news thread.

[BrunoMine]

Ugh, please remove those images, especially since this is a news thread.

OK. Sorry.

[est31]

TenPlus1, note that you can test it now with an updated engine: https://github.com/minetest/minetest/co ... 40a8def401

[Linuxdirk]

Too bad it can’t be compiled on an up-to-date system right now :(

[benrob0329]

Just downgrade your Irrlicht package.

[Linuxdirk]

Just downgrade your Irrlicht package.

That won’t happen. I’m not downgrading anything on my machine. Never. I do’tn even know where to find an outdated package for Irrlicht and won’t compile one by myself …

I rather stay with 0.4.12 as long as either the Minetest package maintainer, the Irrlicht package maintainer, the Irrlicht developers or the Minetest developers provide a patch for either Minetest, or the Minetest package, or Irrlicht, or the Irrlicht package :)

If there aren’t irrlicht-1.8.2-3, irrlicht-1.8.3-1 or minetest.0.4.12-3 or minetest.0.4.13-1 fixing or circumventing the bug all my testing is on hold and I keep using the already compiled 0.4.12 version I have installed.

(My system isn’t a testing system so I only do very limited testing from time to time.)

[est31]

I rather stay with 0.4.12 as long as either the Minetest package maintainer, the Irrlicht package maintainer, the Irrlicht developers or the Minetest developers provide a patch for either Minetest, or the Minetest package, or Irrlicht, or the Irrlicht package :)

The irrlicht package maintainer is aware of the crash bug, and has provided an initial
patch, irrlicht version 1.8.2-2, more see below.

If there aren’t irrlicht-1.8.2-3, irrlicht-1.8.3-1 or minetest.0.4.12-3 or minetest.0.4.13-1 fixing or circumventing the bug all my testing is on hold and I keep using the already compiled 0.4.12 version I have installed.

From what it seems, the crash on startup issue occurs if you compile irrlicht 1.8.2 headers with gcc 5.2. So if you have irrlicht 1.8.2-2, and use clang/llvm to compile minetest, you should be fine (passing -DCMAKE_CXX_COMPILER to a fresh clone).

[benrob0329]

I prefer clang anyways! What do I do to tell it to use clang when using the AUR PACKAGEBUILD?

[est31]

Just add a -DCMAKE_CXX_COMPILER=clang++ option to the cmake step.

[Silwncer]

When we're going to get weather without mods ?

[Linuxdirk]

Since clang can properly compile Minetest Here’s my testing results for the translation functionality.



All variants were compiled with the following base set of options plus -DCUSTOM_LOCALEDIR=/usr/share/locale/ where necessary due to testing.

Code: Select all

-DCMAKE_INSTALL_PREFIX=/usr
-DBUILD_SERVER=0
-DENABLE_GETTEXT=1
-DENABLE_FREETYPE=1
-DENABLE_LEVELDB=0
-DENABLE_REDIS=0
-DCMAKE_CXX_COMPILER=clang++

Before compiling the previously installed version was removed.

All compiled versions were properly installed by the package manager. Before starting each version the ~/.minetest directory was deleted.

[TenPlus1]

est13: it displays "Hi, the output is: 2 and \13"

[rubenwardy]

Linuxdirk, the crash is Irrlicht's fault not Minetest's. Does 0.4.12 work when 0.4.13 doesn't?

[Linuxdirk]

Linuxdirk, the crash is Irrlicht's fault not Minetest's. Does 0.4.12 work when 0.4.13 doesn't?

All already compiled versions work. Only the self-compiled variants don’t.

[rubenwardy]

It must be because the precompiled versions already use Irrlicht 0.8.1.
Irrlicht 0.8.2 is the one with the bug.

[est31]

@Linuxdirk: I have some trouble interpreting your table, but if I read it right, the translations aren't read, so the patch doesn't work?