Time  Nick         Message
08:34 Megaf        Folks, we need ChromeOS support
08:35 Megaf        I just learned that some players on my server would like to player form their chromebooks
09:34 crazy_baboon Hi
09:36 crazy_baboon Anyone knows what is the function of wieldmesh.h and wieldmesh.cpp ?
09:48 Megaf        eh, build without them and see
09:48 Megaf        I'd guess is related to something to do with something relating the regardness of items and hands and stuff and tools and visual
09:50 crazy_baboon thanks!
09:51 crazy_baboon the codebase is a right mess without many useful comments
09:52 Megaf        so, to make myself more clear
09:52 Megaf        I have no idea what it is for
09:52 Megaf        though I believe it could be related to items held in hand
09:52 Megaf        and I agree on the comments
10:12 shivajiva    it's undoubtedly related to wielding meshes in the hand, lack of comments in the codebase means you need to follow and understand the code long before you make any changes, I think it's a test :')
10:24 crazy_baboon thanks guys!
10:27 crazy_baboon also the more recent code 0.5.0 is huge and barerly good to start learning
10:27 crazy_baboon I have gone back to previous much older versions
10:28 Megaf        0.3 was a masterpiece
10:28 Megaf        very simple code
10:29 crazy_baboon cool!
10:29 crazy_baboon gonna check that out
10:29 crazy_baboon Ive been working with 0.4.2
10:31 crazy_baboon yes, I am looking at it now
10:32 crazy_baboon with around 150 source files
10:32 crazy_baboon looks neat
10:35 shivajiva    I'd love to see skybox transitions and audio fade in/out instead of all these abrupt changes we have currently
10:39 shivajiva    I mean I can play a fade in and time it to introduce the continuation sound and trigger the fade out at the end but holy moly that's a lot of effort to get a sound to fade in and out :/
10:43 crazy_baboon hum when I try to compile 0.3-stable, I get several errors of the type: error: call of overloaded ‘stoi(std::__cxx11::string&)’ is ambiguous
10:44 crazy_baboon any idea how to fix this?
10:44 crazy_baboon What do you mean by audio fade in/out? we only have digging samples don't we?
10:45 shivajiva    degrade your gcc version, it compiled at the time
10:45 crazy_baboon poop
10:45 crazy_baboon I cannot do that
10:45 crazy_baboon it's difficult as hell
10:45 shivajiva    we have play sound, we can set it to loop but that's it
10:46 ANAND        crazy_baboon: Here's a page from the Dev Wiki which gives a short non-technical explanation about the engine architecture
10:46 ANAND        https://dev.minetest.net/Core_Architecture
10:47 crazy_baboon we need music composer like those of neverball
10:47 crazy_baboon thanks ANAND
10:49 crazy_baboon it's a shame 0.3-stable won't compile
10:49 shivajiva    I'd settle for a flag that fades the looped sound in, easy to handle the fadeout, the fadein is where it gets complicated
10:49 crazy_baboon 0.3-stable: 150 source files, 0.4.1: 500 files
10:49 ANAND        wow
10:49 crazy_baboon what the hell!
10:52 crazy_baboon i would much rather have a code that is less efficient but it is easy to follow and well commented,
10:53 shivajiva    crazy_baboon: use a vm set up with a suitable version of gcc for 0.3?
10:53 crazy_baboon shivajiva I am already using a vm
10:54 crazy_baboon but it might be the only solution, thanks!
10:57 crazy_baboon gonna try with Ubuntu 12.04 Precise Pangolin!
11:06 crazy_baboon it looks like the code base from 0.3-stable to 0.4.1 was inflated a lot
11:07 crazy_baboon in terms of single player gaming, I cannot really tell why
11:18 celeron55    well, that's the entire modding system and all that
11:19 celeron55    it takes a lot of code to make things customizable
11:20 crazy_baboon thanks celeron55!
11:21 crazy_baboon I would like to ask you a question, I am trying to add irrlicht lighting on minetest
11:21 crazy_baboon with shadows
11:21 celeron55    it's nice to hear 0.3 being called a masterpiece given that most of it was made by me, lol
11:21 celeron55    you need to start that by removing the current lighting, it's incompatible with irrlicht's hardware lighting
11:22 crazy_baboon so far i have enabled the physically based lighting system
11:22 celeron55    or use some shader techniques
11:22 celeron55    (i'm not a shader expert in the slightest so can't really help)
11:22 crazy_baboon and managed to add a cube that cast shadows on minetest voxels
11:23 celeron55    sounds like a reasonable start
11:23 crazy_baboon but the only thing missing is the shadow casting from the voxel cubes - i cannot get it for the life of me
11:24 crazy_baboon thanks!
11:24 celeron55    the voxels are handled quite specially and the entire voxel world is a single irrlicht scene node
11:25 celeron55    that just extends as things are loaded and unloaded
11:25 ANAND        celeron55, any specific reason for choosing Irrlicht over others?
11:25 crazy_baboon do you know the name of that irrlicht scene node?
11:25 celeron55    i guess the class is ClientMap or something like that
11:26 celeron55    ANAND: it was the only reasonable choice at the time, i compared it to ogre3d but ogre3d used lgpl at the time
11:26 celeron55    we're talking of 2010 here
11:26 celeron55    a long time ago
11:27 ANAND        Ah, I see..
11:27 crazy_baboon thanks!
11:28 celeron55    (it might be a bit difficult today to get a good picture of the libraries available in 2010...)
11:28 celeron55    (but it wasn't the same)
11:30 ANAND        Irrlicht doesn't seem to provide the best performance. But I guess it's near-impossible to jump to another one now... :)
11:31 Megaf        Irrlich is ok
11:31 Megaf        Minetest is getting to out of hand
11:31 Megaf        too complex
11:31 ANAND        I once tried MC (demo). My FPS was pinned to 60 even when I had turned up the graphics settings
11:32 Megaf        I believe we got to a point where we could have our own engine
11:32 Megaf        we kinda do already
11:32 Megaf        because we dont use a lot of stuff from irrlich, we are just using our own stuff
11:32 ANAND        Irrlicht might be OK, but it probably isn't the best for MT
11:32 celeron55    it's entirely possible to strip out irrlicht, but it's a lot of work
11:33 crazy_baboon irrlicht is nice so far i can tell, the problem is the added complexity that is being added to the code
11:34 crazy_baboon for example, check out the leap from 0.4.17 to 0.5.0
11:34 celeron55    and doesn't really make sense that way; what would make sense is to fork irrlicht, remove the parts MT doesn't use and include the rest within the project
11:34 celeron55    and then start developing it alongside
11:34 Megaf        ok, I have something good to say about MT to contrast my usual complaining behaviour
11:34 shivajiva    heh sounds like a plan celeron
11:35 Megaf        I am currently redoing my compiling/updating protocol for my server, and just made a new clean build of minetest server
11:35 Megaf        minetest@Minetest:~$ du -ms MinetestBin/
11:35 Megaf        16      MinetestBin/
11:35 Megaf        16 Megabytes!
11:35 Megaf        (with minimal subgame added)
11:36 celeron55    the MT server already uses only some headers from irrlicht, so really not irrlicht at all
11:36 celeron55    just some basic types for coordinates and colors
11:37 Megaf        I have no games folder and of course no minimal subgame there, so where the heck minetestserver is running minimal from?
11:37 celeron55    that has meant blocking some suggested features that aren't possible there as it is
11:38 celeron55    probably you didn't build a run-in-place version and it's using your global MT files
11:38 Megaf        OH! /home/minetest/MinetestBin/share/minetest/games/minimal/
11:38 Megaf        I knew it! I knew something had changed!
11:38 celeron55    or... something weird like that
11:39 Megaf        thy moved games from mt_root/games to that thing
11:39 Megaf        WHY!>
11:39 Megaf        s/>/?
11:39 Megaf        that's why my server had gone haiwire and my script had stopped working
11:39 Megaf        (╯°□°）╯︵ ┻━┻
11:40 celeron55    well, they didn't
11:40 Megaf        how come? Whats the share folder doing there?
11:40 celeron55    use RUN_IN_PLACE=1, like i said, just like always
11:40 celeron55    and it'll work just like always
11:41 Megaf        oh my
11:41 * Megaf      feels ashamed
11:41 Megaf        I had completely forgotten about run in place
11:41 Megaf        hold on
11:41 shivajiva    picnic error
11:42 shivajiva    as TommyTreasure would say
11:42 Megaf        In Portuguese we have a saying, error description for that stuff, we call BIOS
11:42 Megaf        it's a BIOS problem
11:43 Megaf        BIOS = Bixo Ignorante Operando o Sistema. (Stupid[ignorante] animal[bixo] using[operando] the system)
11:43 shivajiva    picnic = problem in chair not in computer
11:44 Megaf        :D
11:44 Megaf        cmake ../MinetestSource -DCMAKE_BUILD_TYPE=Release -DBUILD_SERVER=1 -DBUILD_CLIENT=0 -DCMAKE_INSTALL_PREFIX=../MinetestBin -DENABLE_SYSTEM_JSONCPP=1 -DRUN_IN_PLACE=1
11:44 Megaf        That should do ^^^^
11:45 shivajiva    wondered why you mentioned that prefix :)
11:46 Megaf        what do you mean?
11:47 shivajiva    simply it didn't reflect the path I use and if it's run in place why is it necessary?
11:48 Megaf        well, make install will copy the compiled stuff and install in the prefix
11:48 Megaf        and run from there then
11:48 Megaf        so I have a clean folder without source or unecessary files
11:49 Megaf        until know I actually use a script to copy only the bare minimum required to run the server, but i want to be less dependent on my own scripts and use the work from other people
11:50 Megaf        minetest@Minetest:~/MinetestBin$ du -ms ../MinetestBin/
11:50 Megaf        16      ../MinetestBin/
11:50 Megaf        we still at 16 MB
11:50 Megaf        minetest@Minetest:~/MinetestBin$ ls
11:50 Megaf        bin  builtin  client  clientmods  doc  fonts  games  minetest.conf.example  mods  textures  unix
11:50 Megaf        and there's my beloved tree
11:52 shivajiva    ah I see, just seemed complicated given you retain the src anyway for updating and worlds can use worldmods if it's present giving you the separation you want
11:53 shivajiva    horses for courses
11:54 Megaf        so basically I had games, mods and worlds folder in a safeplace
11:55 Megaf        once the server was built and bareminimum copied to the binary folder, a symbolic link would be made from games,mods and worlds to the new binary folder
11:55 Megaf        and the old source and build files removed
11:56 Megaf        but now, my plane is keeping mods, games and worlds in a safe place, thou the install is done by make install/local, server is built with make minetestserver and I will actually copy, duplicating everything to the new binary folder
11:56 Megaf        will also keep the source and build files, as well a compilation cache
11:57 Megaf        the reason I was so bare minimum before is because at first I had only 4 GB of storage, so source code was actually downloaded to a ram disk, and build was done in ram disk as well
11:57 Megaf        and backups were transfered to other computers
13:45 crazy_baboon voxelands compiles fine and has got around 350 source files
13:46 crazy_baboon not quite as good as minetest 0.3-dev
13:46 crazy_baboon but much better than minetest 0.5.0
13:46 rubenwardy   Would you really prefer less files but longer?
13:47 rubenwardy   Minetest is a pretty small project really
13:47 rubenwardy   Compared to things like the Linux kernel
13:47 crazy_baboon now it's getting over to 1000 files
13:47 crazy_baboon everything is relativ ei know
13:48 rubenwardy   For me, it would make it easier if everything was in directories like  server, client, then common components
13:48 rubenwardy   It's heading that way
13:48 rubenwardy   Not quite though
13:49 crazy_baboon everything would be easier if the code was at least commented decently and some of the key variables have nonsense names
13:50 crazy_baboon but i agree with you
13:50 crazy_baboon for example, rendering code is scattered all over the place!
13:50 crazy_baboon camera.cpp, mapblock_mesh.cpp, tile.h, ... you name it
13:51 rubenwardy   That's because rendering is a pipeline, not a single operation
13:52 rubenwardy   That's because rendering is a pipeline, not a single operation
13:52 crazy_baboon it looks way too convoluted, especially in 0.5.0
13:53 rubenwardy   Rendering is complicated
13:53 crazy_baboon but that's my opinion, hardly worth anything as I have never contributed a single line to minetest!
13:53 rubenwardy   And 5.0 has made some improvements
13:54 crazy_baboon like what for example?
13:54 rubenwardy   But really it would be good to have more up to date overarching documentation for structure
13:54 rubenwardy   Splitting up the server mega class for example
13:56 crazy_baboon splitting up classes, is that a good thing?
13:56 crazy_baboon hum...
13:56 crazy_baboon ok
13:56 rubenwardy   Yes
13:56 rubenwardy   *hmmm
13:57 rubenwardy   Read up on coupling and cohesion
13:58 rubenwardy   Your approach to lighting is unlikely to work
13:58 rubenwardy   Lighting is a massive part of rendering
13:58 rubenwardy   It's probably worth making a separate program and testing out different rendering algorithms
13:58 rubenwardy   Maybe even in raw opengl
14:00 crazy_baboon IMO, the easier solution to get shadows is probably to go back to 0.3-stable and work from there
14:01 crazy_baboon nowadays, the codebase is just reaching a unpleasant degree of complexity
14:01 rubenwardy   More features means more complexity, unfortunately
14:03 crazy_baboon yep, I understand, but, personally, I never got into the modding thing
14:04 crazy_baboon but I understand those that like that, of course, the code has to be larger to accomodate that
14:06 rubenwardy   As a game engine, that's what we aim for
14:17 crazy_baboon does anyone know which is the most recent version of gcc that can compile minetest 0.3-stable?
14:17 Megaf        hold on
14:19 crazy_baboon ok
14:21 rubenwardy   alternatively, you could fix the issue
14:22 Megaf        I will try to build it here now
14:23 crazy_baboon I tried to fix it, unsuccessfully ;)
14:23 crazy_baboon thanks!
14:25 Megaf        rubenwardy: http://paste.debian.net/plainh/b679b70c
14:26 Megaf        any quick fix for that?
14:32 Megaf        crazy_baboon: libc6 and/or gcc 4.7 should build it
14:32 Megaf        Minetest 0.3 on Debian Wheezy was built with gcc4.7 https://packages.debian.org/wheezy/minetest
14:33 Krock        so what's the challenge with chromebooks? can't the compile it?
14:33 crazy_baboon thanks! I'll try to find the Ubuntu version that corresponds to that!
14:33 Krock        *they
14:33 rubenwardy   Megaf: delete the function in utility
14:33 rubenwardy   basically, Minetest declared their own stoi function before C++ added it
14:34 rubenwardy   the order compiler has an order stdliv
14:34 Krock        shivajiva: sound fading has been implemented already..
14:35 crazy_baboon ubuntu 16.06 should do it
14:36 Krock        don't you mean 16.04?
14:37 crazy_baboon yes, rubenwardy, fixing the issue  with minetest03-stable would be so infinitely better! But the commit is like 7 years old, so will it get updated?
14:37 crazy_baboon yes
14:37 crazy_baboon sorry 16.04
14:37 crazy_baboon i have typing issues ;)
14:39 Krock        ANAND: switching from irrlicht to something else was already discussed many times and it's just too much effort for the current code base
14:39 Megaf        now I broke the source o/
14:39 Krock        not saying that it's impossible
14:39 Krock        Megaf: rm -rf, git clone :P
14:41 Megaf        eh, stoi is declared in lots of files
14:41 Krock        declared or used?
14:41 Megaf        why not just add a single stoi header and use it
14:41 Megaf        Krock: FYI I have no idea what I am talking about
14:41 Megaf        /dev/shm/minetest/src/inventory.h: In constructor ‘IMoveAction::IMoveAction(std::istream&)’:
14:41 Megaf        /dev/shm/minetest/src/inventory.h:586:18: error: call of overloaded ‘stoi(std::__cxx11::string&)’ is ambiguous
14:41 Megaf        count = stoi(ts);
14:42 Krock        -> PR
14:42 rubenwardy   no need
14:42 rubenwardy   0.3.0 isn't supported
14:42 Krock        oh, we're still talking about 0.3.0 here?
14:42 rubenwardy   releasing 0.3.2 six years later would be a bit weird
14:42 Krock        because I see __cxx11 there, which wasn't enabled in 0.3.x
14:43 Megaf        fixed the code and broke it again
14:43 Megaf        /dev/shm/minetest/src/inventory.h:581:30: error: expected ‘;’ at end of member declaration
14:43 Megaf        IMoveAction(std::istream &is)
14:43 Megaf        o/
14:43 Megaf        rubenwardy: it would show commitment
14:43 rubenwardy   I guess it could be fixed better in CMake
14:43 rubenwardy   wait
14:43 rubenwardy   did you clear the cache?
14:43 rubenwardy   you need to clear the cmake cache when switching between   0.5 and 0.3
14:43 Megaf        99,8536% sure I did
14:43 rubenwardy   rm -rf CMakeFiles CMakeCache.txt
14:44 Megaf        im building in a different folder
14:44 Megaf        actually, building mt for the first time here
14:44 Megaf        in this machine
14:44 Krock        crazy_baboon: FYI the next release will be 5.0.0 - kinda moving the minor version number to major now
14:44 Megaf        s/machine/os install
14:45 Megaf        I can't code
14:45 Megaf        (╯°□°）╯︵ ┻━┻
14:45 Megaf        ¯\_(ツ)_/¯
14:45 Krock        then just download an ancient windows binary and run it in wine...
14:45 crazy_baboon dio porco
14:45 Megaf        porca putana
14:45 * Megaf      runs and hides
14:46 crazy_baboon dio canne!
14:46 Megaf        porca troia!
14:46 Krock        guys, please.
14:46 Krock        there's #minetest-chat for this
14:46 crazy_baboon sorry
14:46 crazy_baboon :)
15:06 crazy_baboon ok, I am also having issues with virtualbox
15:06 crazy_baboon and besides that, virtualized environments are never good for working
15:07 crazy_baboon it would be great if minetest-stable-0.3 was updated to work with modern c++ compilers, IMO
15:42 bodqhrohro   Are glowing entities already supported?
15:46 Krock        kinda, yes. they glow but don't emit light
15:47 Krock        the camera just lightens it
15:49 bodqhrohro   Okay, I will try it
15:50 p_gimeno     crazy_baboon: have you tried the Debian version? they often patch issues that arise in the compilation
15:51 crazy_baboon p_gimeno: I have not tried it, can you compile it from source?
15:51 * p_gimeno   tries
15:56 p_gimeno     missin jthread, installing...
15:57 p_gimeno     hm, it aborts there even if I install jthread, maybe it requires a too old version
16:00 p_gimeno     ok, it searched in the wrong directory, fixed by editing FindJthread.cmake, now going to compile
16:13 p_gimeno     meh, it errs on JMutex and JMutexAutoLock
16:21 p_gimeno     maybe this commit fixes the stof issue? https://github.com/minetest/minetest/commit/c8211ceb3
16:21 p_gimeno     you may need to apply it by hand
16:28 crazy_baboon p_gimeno: yep it looks like it might fix it
16:29 crazy_baboon thanks!
18:23 Krock        merging somewhat trivial #7668 in 5 mins
18:23 ShadowBot    https://github.com/minetest/minetest/issues/7668 -- Fixed comment by crazyBaboon
18:25 Krock        #7592 too
18:25 ShadowBot    https://github.com/minetest/minetest/issues/7592 -- get_node_drops: Make empty drop return empty table by tenplus1
18:31 Krock        merging
18:32 Krock        done
20:28 shivajiva    Krock: good to know, any more info on what the float value represents?
20:28 Krock        sorry.. context please?
20:29 shivajiva    fade
20:29 Krock        oh right. it should be documented in the Lua API. It's the volume decrease per second
20:30 Krock        or increase, if it's positive
20:30 Krock        https://github.com/minetest/minetest/blob/master/doc/lua_api.txt#L813
20:31 shivajiva    thanks not exactly clear in the doc that it relates to per second
20:35 Krock        !next
20:35 ShadowBot    Another satisfied customer. Next!
20:38 shivajiva    not satisfied yet...so if I set it to 0.1 how long to get to full volume, 10s?
20:39 shivajiva    because that doesn't seem to be the case
20:40 Krock        checking the source...
20:42 Krock        linear decrease, a "step" per second https://github.com/minetest/minetest/blob/master/src/client/sound_openal.cpp#L623
20:42 shivajiva    and if the sample is shorter than 10s what happens when it loops
20:43 Krock        oh I see...
20:43 Krock        fade is actually the amount per second, whereas "gain" notes the final gain
20:44 shivajiva    so the loop length shouldn't matter I guess
20:46 Krock        same OpenAL handler/ID, so loops will behave the same as regular sounds
20:50 Krock        say you play a sound with gain=0.8, then will fade=-0.1, gain=0.2 take 6s to fade out according to what I see in the code
20:57 shivajiva    thanks for checking it, seems okay, pink noise is a bit hard to discern the change but it has definitely softened the transition
21:00 Krock        I originally tested the PR with music with large gain changes over a long timespan
21:00 Krock        and the difference was there. Docs need tweaking however
21:01 shivajiva    yea, it's far more apparent on a 1Khz tone than a rain sample