19:41:15: ERROR[main]: ERROR: An unhandled exception occurred: ...worldmods/Minetest-WorldEdit/worldedit_commands/init.lua:621: attempt to perform arithmetic on a nil value

For more info see minetest/minetest#2481. I'm not sure if this is a mod bug or an engine bug so... I reported it here and in the engine repository.

I know this one sounds totally obscure, but here goes:

The server in question runs the Realtest game, though this may not actually matter. A player signed on as "You" (capital "Y") just as a joke, though this also may not matter. No problems when he signed on. He signed off before the media had finished downloading, and thus without actually ever touching worldedit or the worldedit GUI. The server crashed, citing the following errors:

terminate called after throwing an instance of 'ServerError'
what(): ...ealtest_World/worldmods/worldedit/worldedit_gui/init.lua:137: attempt to index local 'player' (a nil value)
stack traceback:
...ealtest_World/worldmods/worldedit/worldedit_gui/init.lua:137: in function 'update_main_formspec'
...ealtest_World/worldmods/worldedit/worldedit_gui/init.lua:145: in function 'func'
/usr/local/share/minetest/builtin/misc.lua:17: in function </usr/local/share/minetest/builtin/misc.lua:9>
/usr/local/share/minetest/builtin/misc_register.lua:348: in function </usr/local/share/minetest/builtin/misc_register.lua:336>
/home/minetest/Scripts/minetestserver-realtest.sh: line 27: 18551 Aborted

This may be related to some discussion Sapier and I had had in #minetest-dev where the server and client could get out of sync with one another during the initial connection if the server is sufficiently fast, causing certain parts of the initial connection data to be exchanged at the wrong time. Thus, this may actually be a server bug, rather than a Worldedit bug.

//transpose, and as a result //rotate, do not work for non-square regions. So it turns out that in-place matrix transposition is something that is actually much more difficult than it seems.

Square matrices are easy, though, so for now make sure a square region is selected before transposing.

WARNING: Undeclared global variable "unified_inventory" accessed at .../minetest_game/mods/worldedit/worldedit_gui/init.lua:76

if mod not installed

Hi, Now Minetest server is actually exposing some bugs in mods and here is one worldedit_infinity

WARNING: Undeclared global variable "worldedit" accessed at ...Server/bin/../mods/WolrdEdit/worldedit_infinity/init.lua:1

http://paste.debian.net/plain/132623

Always moving in the look direction despite the axis selection.

Currently worldedit.player_notify is used, but all that does now is prepend "WorldEdit -!- ", which is obvious given the context of the command.

terminate called after throwing an instance of 'ServerError'
what(): ...vanessae_game/mods/worldedit/worldedit/visualization.lua:138: bad
argument #4 to 'add_node' (string expected, got nil)
stack traceback:
[C]: in function 'add_node'
...vanessae_game/mods/worldedit/worldedit/visualization.lua:138: in func
tion 'restore'
...vanessae_game/mods/worldedit/worldedit_commands/init.lua:911: in func
tion 'func'
...ssae_game/mods/worldedit/worldedit_gui/functionality.lua:575: in func
tion 'on_select'
...ames/vanessae_game/mods/worldedit/worldedit_gui/init.lua:220: in func
tion <...ames/vanessae_game/mods/worldedit/worldedit_gui/init.lua:210>
/usr/local/share/minetest/builtin/misc_register.lua:330: in function </u
sr/local/share/minetest/builtin/misc_register.lua:318>
/home/minetest/Scripts/minetestserver-survival.sh: line 25: 25671 Aborted

I had a few thoughts about making region selection a bit more user friendly.

I would like to add the following types of commands (thought not necessarily with these names)

//setactivenode -- select the active region node (1 or 2)
//toggleactivenode -- switch to the other region node
//setactivefore -- move the active node n units (1 by default) in the nearest horizontal direction from the camera (NHDC)
//setactiveback -- move the active node n units (1 by default) backward NHDC
//setactiveleft -- move the active node n units (1 by default) left NHDC
//setactiveright -- move the active node n units (1 by default) right NHDC
//setactiveup -- move the active node up n units (1 by default) (y axis)
//setactivedown -- move the active node down n units (1 by default) (y axis)

Then I am going to try to patch minetest to allow binding commands to shortcut keys as I suggested here: https://forum.minetest.net/viewtopic.php?id=8611

It could be nice to have a set of shortcuts bound to the keypad, like
Up //setactivefore
Down //setactiveback
Left //setactiveleft
Right //setactiveright
PgUp //setactiveup
PgDn //setactivedown
Ins //toggleactivenode

(Could also imagine using shift+arrow keys for moving the inactive anchor, ctrl+arrow for moving both anchors at once, ctrl+click to set anchors without needing to set them etc)

Is this worth pursuing in your opinion?

GUI closes without any changes every time, when i try to choose smth from drop-down list. tested on 0.4.10

Our server has crashed a few times over the past couple months due to something with WE's chat commands. The crashes have occurred when no admins were ingame (only admins have the WE privs).

Not enough info to be able to reproduce this crash.

Log snippet from most recent crash:

19:05:36: ACTION[ServerThread]: BluAngel places node default:cobble at (1389,-186,714)
19:05:36: ACTION[ServerThread]: facedir: 0
19:05:36: ACTION[ServerThread]: BluAngel places node default:cobble at (1389,-185,714)
19:05:36: ACTION[ServerThread]: facedir: 0
19:05:43: ACTION[ServerThread]: hoog digs default:grass_4 at (-315,20,805)
19:05:59: ERROR[ConnectionSend]: con(5/1)RunTimeouts(): Peer 24 has timed out. (source=peer->timeout_counter)
19:05:59: ERROR[ConnectionSend]: con(5/1)RunTimeouts(): Peer 40 has timed out. (source=peer->timeout_counter)
19:05:59: ERROR[ConnectionSend]: con(5/1)RunTimeouts(): Peer 73 has timed out. (source=peer->timeout_counter)
19:05:59: ERROR[ConnectionSend]: con(5/1)RunTimeouts(): Peer 54 has timed out. (source=peer->timeout_counter)
19:05:59: ERROR[ConnectionSend]: con(5/1)RunTimeouts(): Peer 90 has timed out. (source=peer->timeout_counter)
19:05:59: ERROR[ConnectionSend]: con(5/1)RunTimeouts(): Peer 92 has timed out. (source=peer->timeout_counter)
19:05:59: ERROR[ConnectionSend]: con(5/1)RunTimeouts(): Peer 96 has timed out. (source=peer->timeout_counter)
19:05:59: ERROR[main]: ERROR: An unhandled exception occurred: ...../mods/worldedit_2014_03_15/worldedit_commands/init.lua:36: bad argument #1 to 'chat_send_player' (string expected, got nil)
19:05:59: ERROR[main]: stack traceback:
19:05:59: ERROR[main]:  [C]: in function 'chat_send_player'
19:05:59: ERROR[main]:  ...../mods/worldedit_2014_03_15/worldedit_commands/init.lua:36: in function 'player_notify'
19:05:59: ERROR[main]:  ...../mods/worldedit_2014_03_15/worldedit_commands/safe.lua:42: in function 'func'
19:05:59: ERROR[main]:  ...test/minetest-0.4.9/bin/../builtin/game/chatcommands.lua:27: in function <...test/minetest-0.4.9/bin/../builtin/game/chatcommands.lua:16>
19:05:59: ERROR[main]:  ...minetest/minetest-0.4.9/bin/../builtin/game/register.lua:348: in function <...minetest/minetest-0.4.9/bin/../builtin/game/register.lua:336>

In thread 7fdeccc1f780:
/home/minetest/new_minetest-0.4.9/src/main.cpp:1870: int main(int, char**): Assertion '0' failed.
Debug stacks:
DEBUG STACK FOR THREAD 7fdec4ec4700:
#0  virtual void* CurlFetchThread::Thread()
DEBUG STACK FOR THREAD 7fdeccc1f780:
#0  int main(int, char**)
(Leftover data: #1  Dedicated server branch)
(Leftover data: #2  virtual void ServerMap::save(ModifiedState))
(Leftover data: #3  virtual void Database_SQLite3::saveBlock(MapBlock*))
(Leftover data: #4  void ItemStack::serialize(std::ostream&) const)

//save will save nodes.
//load will load those nodes.
But air isn't saved. So if you //load into a place that isn't completely filled with air already, the area will be a combination (or merge) of the first and second places.

If you do this:

//p set (first place)
//save file
//p set (second place)
//load file
//save file
//p set (first place)
//load file

you won't end up with the first place being the same because it will load nodes saved from the second place that weren't cleared when loading from the first place.

Now to explain it in pictures:

//p set //save test

//pos1 //pos2 //load test

//save test //p set //load test

//copy //copy behaves differently than //save //load //save //load:

//p set

//copy x 6

//p set //copy x -6

//save //load ~= //copy

Somehow figure out a way to save all node metadata in the WorldEdit save.

Also make sure everything is backwards-compatible with the current version of the .we format, as well as the first version.

Probably should also add some sort of version marker to WE files too.

mauvebic has expressed interest in combining the WE API with the Multinode interface to make some sort of ultimate node-manipulation tool.

Should probably have both interfaces available for use if possible. Not sure how that would work.

There was an... incident on VanessaE's Creative Server involving a moderator accidentally making a 1000m long 10000m deep trench. I was thinking that this could have been prevented if WorldEdit had some sort of confirmation system that would prompt the user before executing any command that would affect more than 5000 nodes (configurable). An example of how I think this should work:

//pos1
WorldEdit -!- Position 1 set to (position)
//pos2
WorldEdit -!- Position 2 set to (position)
//set default:stone
WorldEdit -!- WARNING: This would affect 13348 nodes. Type the command again if you are sure you want to do this.
//set default:stone
WorldEdit -!- 13348 nodes set

WorldEdit GitHub commit, 2014_10_24, fb49cc4
MIneteset engine, GitHub commit, 2014_12_09, dcc48976ce

I tried three times to transfer a small build (286 nodes) from one singleplayer world to another using //save and //load. I saved the build (//save shack), then copied the "shack.we" file to the schems folder of another singleplayer world, and then tried to place the build in that world using "//load shack". At this point, Minetest crashed each time.

Here are error messages from the terminal output:

05:17:17: ERROR[main]: ServerError: ...ty_mods/WorldEdit_2014_10_24/worldedit/serialization.lua:172: bad argument #2 to 'insert' (number expected, got nil)
05:17:17: ERROR[main]: stack traceback:
05:17:17: ERROR[main]:  [C]: in function 'insert'
05:17:17: ERROR[main]:  ...ty_mods/WorldEdit_2014_10_24/worldedit/serialization.lua:172: in function 'load_schematic'
05:17:17: ERROR[main]:  ...ty_mods/WorldEdit_2014_10_24/worldedit/serialization.lua:219: in function 'deserialize'
05:17:17: ERROR[main]:  ...ty_mods/WorldEdit_2014_10_24/worldedit_commands/init.lua:978: in function 'func'
05:17:17: ERROR[main]:  ...f/01_minetest/lazyj/bin/../builtin/game/chatcommands.lua:40: in function <...f/01_minetest/lazyj/bin/../builtin/game/chatcommands.lua:29>
05:17:17: ERROR[main]:  ...stuff/01_minetest/lazyj/bin/../builtin/game/register.lua:348: in function <...stuff/01_minetest/lazyj/bin/../builtin/game/register.lua:336>

The save command doesn't always work. In the following map:
?????????????
??A??????????
?????????????
?????????B???
?????????????
Doing a //set pos from A to B and saving could work, but if you do it from B to A it won't. (so one of them will work and the other won't, depending on the way you are facing). When it doesn't work, it says 0 blocks were saved. The problem is how those while's in the serialize function work, I think.

see https://forum.minetest.net/viewtopic.php?p=148803#p148803

Hi, I'm working on adding cuboid manipulation commands such as outset, inset, etc. To match those in the minecraft version of worldedit. I already have some working. I was wondering if you would be interested in accepting the changes.

I already have:

• Outset
• Inset
• Shift
• Expand (WIP)

The code still needs some cleaning, but it shouldnt take too much time.

This fits the style of the icons used in Unified Inventory. Originally made for that mod, but it wont actually be used there (not directly usable in that mod I mean). Original source files for this image can be obtained from http://commons.wikimedia.org/wiki/File:Erioll_world_2.svg and http://www.clker.com/clipart-2256.html .

(EDIT: changed the icon a bit from the initial "commit" to add some shadowing)

If I add

spawned = false
minetest.register_globalstep(function(dtime)
        if spawned then
                print(worldedit.player_axis('singleplayer'))
        end
end)

and then add

spawned = true

to any command, then when I run it, it gives me a live ?-to-axis mapping.

Here's how it maps:
Anything above 45° up maps to y.
Looking down doesn't map to y.
From halfway between +z and +x, to +x, to -z, to about 30° clockwise past -z, maps to x.
From about 30° clockwise past -z, to about 30° before -x, maps to y.
From about 30° before -x, to -x, to +z, to halfway between +z and +x, maps to z.

So, with axes like this:
....+z....
-x....+x
....-z....

it maps like this:
zzzzz.
z...|...x
z...|...x
z--+--x
z...|...x
y...|...x
yyxxxx

with y on top. (looking more that 45° vertical maps to y)

Also, looking down (about 45°) on the y side (-x and -z) maps to either x or z (instead of y) depending on your horizontal angle.

Code in question:

worldedit.player_axis = function(name)
        local dir = minetest.env:get_player_by_name(name):get_look_dir()
        if dir.x > dir.y then
                if dir.x > dir.z then
                        return "x"
                end
        elseif dir.y > dir.z then
                return "y"
        end
        return "z"
end

When I first noticed it, I was looking in the -z direction and my microcontrollers stacked in the +x direction.
Speaking of which, the ? axis should return both positive and negative axes, with the functions that use it stripping off the negative sign. When //stacking, the sign of the axis matters.

This issue has been resolved.

The .we files were not being downloaded in a raw format, but rather included html code.

The wget used in linux was not working the same as right-click save as in windows.

My apologies if I put anyone to unnecessary work.

I'm attempting to make a mod (well, subgame) use ".we" files for its world generation. For this I'm using worldedit.deserialize directly from within my mod.

However, when I attempt to do this I obtain the following error:

Lua: Runtime error from mod 'default' in callback ScriptApiServer::on_chat_message(): 
   ...worldedit/serialization.lua:27: attempt to call method 'find' (a nil value)
   stack traceback:
   ...worldedit/serialization.lua:27: In function 'read_header'
   ...worldedit/serialization.lua:115: In function 'load_schematic'
   ...worldedit/serialization.lua:220: In function 'deserialize'

I'm actually not sure if this is really an issue that I should be posting here or is a mistake I made (I'm not very familiar with minetest development).

The schematic loads fine when I use the //load command from the console, and I'm basically using a very similar code to the one implemented in that command. It's just that I call it from a command implemented in my mod (the 'default' mod of my subgame).

What could be causing 'find' to be nil? Do I need to import some library or something?

The //scale parameter should be used like this:

//scale 2 to double the size of the structure, so a 1 cube becomes 2^3, and so on.

So, using this, you can make a small tree say 5 blocks high, then scale it 2x, it should then be 10 blocks tall, plus the other directions times two.

ERROR[ServerThread]: Map::setNode(): Not allowing to place CONTENT_IGNORE while trying to replace "air" at (21,79,29) (block (1,4,1))

The schematic is placed just fine (I think). I didn't use my own file so I will test again and reply.

Edit: Nvm the file must have been messed up or something.

(might not actually be worldedit's fault)

I teleported out to 10000,10000 because I needed a large open space, then tried to set worldedit's pos1 there before it had generated that chunk. The server then crashed with this error:

terminate called after throwing an instance of 'std::bad_alloc'
what(): std::bad_alloc

Ok, not the most descriptive subject line, but that's because I have no clue how this happened or why, as I wasn't the one who actually triggered the crash. This might be related to the //inspect command, as I'd instructed one of my moderators to use that feature while investigating some vandalism on a server.

Here's the backtrace:

terminate called after throwing an instance of 'ServerError'
what(): ...st_World/worldmods/worldedit/worldedit_commands/init.lua:36: bad argument #1 to 'chat_send_player' (string expected, got nil)
stack traceback:
[C]: in function 'chat_send_player'
...st_World/worldmods/worldedit/worldedit_commands/init.lua:36: in function 'player_notify'
...st_World/worldmods/worldedit/worldedit_commands/safe.lua:42: in function 'func'
/usr/local/share/minetest/builtin/chatcommands.lua:25: in function </usr/local/share/minetest/builtin/chatcommands.lua:16>
/usr/local/share/minetest/builtin/misc_register.lua:348: in function </usr/local/share/minetest/builtin/misc_register.lua:336>
/home/minetest/Scripts/minetestserver-minetest.sh: line 25: 28052 Aborted /usr/local/bin/minetestserver --worldname Minetest_World --config /home/minetest/.minetest/minetest-server-minetest.conf --gameid minetest_game --port 30005 --disable-unittests --logfile $DEBUG

I can't change pages in the creative inventory if the WorldEdit GUI button is installed. It flashes to the page I selected, then back to page 1.

would be awesome if the //save and //load supported the metadata features. It would make it a lot easier to move parts around where copy doesn't work (eg, you want to copy to x+10,y+20, //copy only takes 1 directional argument).

When a filled chest is //moved or //copyed it keeps its inventory.

When the filled chest is //saved and //loaded it loses its inventory.

When the filled chest is //metasaved and //metaloaded it loses the capability of having inventory, instead of being either filled or empty.

Make a region of 50 x 5 x 50
Select it
//stack y 10

...the whole thing runs for a while, says it's stacked N nodes, then continues to spin the CPU at 100% and never finishes.

Meanwhile, //copy seems to work fine for the same amount of data - just a matter of several seconds. The slowness is to the point that, even for a relatively small area (say 10x5x5 nodes) I can //copy the structure repeatedly (with increasing coordinates) and get the job done faster than using //stack to do the same thing.

Currently the spiral code is taken from RosettaCode. I really doubt this is compatible with the license of the WorldEdit mod. According to RosettaCode, all content on the site is under the GFDL, so we probably need to rewrite the spiral function.

Also it's sort of inscrutable anyways, check out this monster of a line:

return (2*l-1)^2+4*l+2*l*sn(x+y)+sn(y^2-x^2)*(l-(av(y)==l and sn(y)*x or sn(x)*y)) -- OH GOD WHAT

Does anyone actually know what that does?

Changing the axis from "Look direction" to any other axis causes player's inventory to appear instantly.

https://github.com/Uberi/Minetest-WorldEdit/blob/master/worldedit_commands/init.lua#L940
I saw that this command looks for the name the admin of the server uses.
If this admin didn't set a password or set a bad one (eg. "1") and he/she leaves the server, a player could join as admin and has access to the his/her computer ("//lua os.execute('notify-send hi')"). Without worldedit this person could only change some stuff of the minetest server (without having the mesecons commando block or terminal mod installed).
I think the ip address of the player should be "127.0.0.1" if he/she is allowed to use //lua.
https://github.com/minetest/minetest/blob/master/doc/lua_api.txt#L1355

local admin = minetest.setting_get("name")
if not (admin and name == admin and minetest.get_player_ip(name) == "127.0.0.1") then
worldedit.player_notify(name, "this command can only be run by the server administrator")
return
end

Could you create a function in WE that when you place a //pos1 (-> the summit), it create a mountain under the pos to the ground ?

It must be place //pos1 for the summit, and //pos2 (pos1 <-> pos2 = radius) on the ground for the mountain width.

Thanks !

Add //sphere, //cone, and //pyramid commands for simpler building.

Hi, I couldnt track down when this happens, I wasnt online when it crashed.

21:16:59: ACTION[ServerThread]: Jessch8281 uses fake_fire:flint_and_steel, pointing at [node under=-502,-1565,991 above=-502,-1565,990]
21:16:59: ACTION[ServerThread]: Gronx digs default:stone at (-488,-1565,1020)
21:16:59: ACTION[ServerThread]: Gronx digs default:stone at (-488,-1565,1021)
21:17:00: ACTION[ServerThread]: Gronx digs default:stone at (-489,-1565,1020)
21:17:00: ACTION[ServerThread]: Gronx digs default:stone at (-489,-1565,1021)
21:17:01: ACTION[ServerThread]: Gronx digs default:stone at (-490,-1565,1019)
21:17:02: ACTION[ServerThread]: Gronx digs default:stone at (-490,-1565,1020)
21:17:02: ACTION[ServerThread]: Gronx digs default:stone at (-490,-1565,1021)
21:17:02: ACTION[ServerThread]: Jessch8281 uses fake_fire:flint_and_steel, pointing at [node under=-511,-1566,990 above=-511,-1565,990]
21:17:03: ACTION[ServerThread]: Gronx digs default:stone at (-490,-1565,1022)
21:17:03: ACTION[ServerThread]: Gronx digs default:stone at (-491,-1565,1022)
21:17:04: ACTION[ServerThread]: Gronx digs default:stone at (-492,-1565,1022)
21:17:05: ACTION[ServerThread]: Gronx digs default:stone at (-494,-1566,1022)
21:17:06: ACTION[ServerThread]: Gronx digs default:stone at (-494,-1565,1022)
21:17:06: ACTION[ServerThread]: CHAT: <Jessch8281> Yay! Furnaces
21:17:08: ACTION[ServerThread]: Jessch8281 uses fake_fire:flint_and_steel, pointing at [node under=-515,-1563,993 above=-514,-1563,993]
21:17:10: ACTION[ServerThread]: Jessch8281 uses fake_fire:flint_and_steel, pointing at [node under=-511,-1565,993 above=-511,-1564,993]
21:17:13: ERROR[main]: ERROR: An unhandled exception occurred: std::bad_alloc
21:17:13: ERROR[main]: stack traceback:
21:17:13: ERROR[main]:  [C]: in function 'read_from_map'
21:17:13: ERROR[main]:  ...Server/bin/../mods/WolrdEdit/worldedit_commands/mark.lua:68: in function 'mark_region'
21:17:13: ERROR[main]:  ...Server/bin/../mods/WolrdEdit/worldedit_commands/mark.lua:25: in function 'mark_pos1'
21:17:13: ERROR[main]:  ...Server/bin/../mods/WolrdEdit/worldedit_commands/init.lua:242: in function 'callback'
21:17:13: ERROR[main]:  /home/minetest/Server/bin/../builtin/game/item.lua:383: in function </home/minetest/Server/bin/../builtin/game/item.lua:376>

In thread b689b220:
/home/minetest/Server/src/main.cpp:1955: int main(int, char**): Assertion '0' failed.
Debug stacks:
DEBUG STACK FOR THREAD b3817440:
#0  virtual void* CurlFetchThread::Thread()
DEBUG STACK FOR THREAD b689b220:
#0  int main(int, char**)
(Leftover data: #1  Dedicated server branch)
(Leftover data: #2  virtual void ServerMap::save(ModifiedState))
(Leftover data: #3  void ItemStack::serialize(std::ostream&) const)
Aborted

11:54:37: ERROR[main]: ERROR: An unhandled exception occurred: ...test/mods/Minetest-WorldEdit/worldedit_commands/init.lua:36: bad argument #1 to 'chat_send_player' (string expected, got nil)
11:54:37: ERROR[main]: stack traceback:
11:54:37: ERROR[main]: [C]: in function 'chat_send_player'
11:54:37: ERROR[main]: ...test/mods/Minetest-WorldEdit/worldedit_commands/init.lua:36: in function 'player_notify'
11:54:37: ERROR[main]: ...test/mods/Minetest-WorldEdit/worldedit_commands/safe.lua:42: in function 'func'
11:54:37: ERROR[main]: /usr/share/minetest/builtin/chatcommands.lua:25: in function </usr/share/minetest/builtin/chatcommands.lua:16>
11:54:37: ERROR[main]: /usr/share/minetest/builtin/misc_register.lua:348: in function </usr/share/minetest/builtin/misc_register.lua:336>

In thread b6c96700:
/build/buildd/minetest-201404190451/src/main.cpp:1887: int main(int, char**): Assertion '0' failed.
Debug stacks:
DEBUG STACK FOR THREAD b4bffb40:
#0 virtual void* CurlFetchThread::Thread()

DEBUG STACK FOR THREAD b6c96700:
#0 int main(int, char**)

(Leftover data: #1 Dedicated server branch)
(Leftover data: #2 virtual void ServerMap::save(ModifiedState))
(Leftover data: #3 virtual void Database_SQLite3::saveBlock(MapBlock*))
(Leftover data: #4 void ItemStack::serialize(std::ostream&) const)

When loading a WE file in the original WorldEdit format, nodes are loaded upside-down/flipped. This probably has something to do with worldedit.deserialize_old, but can't discount the possibility of the file itself being saved upside down.

Solution, obviously, is to either fix the function in the former case, or flip it in the latter case.

Only happens with //set

Like 3d max and other 3d modeling tools

It seems to place a couple of levels, then it places the rest of them up higher...

Hi there,

I am writing an ebuild for Gentoo and have a slight problems with installation.
Do you know where exactly I should place mod files if I am installing system-wide ?

I've tried to put them into /usr/share/games/minetest/mods/minetest/ dir, but they became invisible for game.

Do you know if there any setting I should tweak or set to make it work ?

Cheers,
Iegor

Hi,
I think a undo function to rollback the last change made by WorldEdit would be a good addition to the tool.
Thank you for considering it :-)

Hi,

I am currently packaging the "popular" minetest mods for Debian. For Debian (and most likely others), AGPL-3 is an issue, because it has incompatibilities with older GPL licenses.

Most importantly, the plugin cannot be distributed with other plugins that are GPL-2 only.

Also, the AGPL-3 is not fulfillable as it requires the UI to expose a download button to the source code. Being a Minetest plugin, doing so is almost impossible and makes no sense at all (yes, being able to run Minetest through a remote dispaly connection does mean it communicates with the user over a network).

It is highly adviseable to re-license WorldEdit under GPL-3 (or GPL-2+) to make it fully compatible with Minetest itself and others.

-nik

Hi, Now Minetest server is actually exposing some bugs in mods and here is one on worldedit_commands

07:51:41: WARNING: Undeclared global variable "check_set" accessed at ...Server/bin/../mods/WolrdEdit/worldedit_commands/init.lua:831
07:51:41: WARNING: Undeclared global variable "check_set" accessed at ...Server/bin/../mods/WolrdEdit/worldedit_commands/init.lua:842

http://paste.debian.net/plain/132623

//save <filename> works great.
It seems that //save isn't called at all, since even the folder schems is not being created.