Time Nick Message 06:20 hmmmm hey guys i'm here 06:20 hmmmm hrmm somebody wanted me to review somethings 07:34 est31 you want to review something? 07:34 est31 hmmmm, you can have a look at #4249 07:34 ShadowBot https://github.com/minetest/minetest/issues/4249 -- Add multithreading ability to mapblock mesh creation by ElementW 07:34 est31 if you are still around :) 08:46 VanessaE 2016-07-13 04:46:22: ERROR[Main]: ...st/.minetest/games/minetest_game/mods/default/mapgen.lua:1526: attempt to call field 'get_mapgen_setting' (a nil value) 08:46 VanessaE someone broke minetest_game ... 08:48 VanessaE oh, new engine function 08:48 VanessaE *sigh* 08:51 VanessaE why must you guys always fuck around with the mapgen? 08:51 VanessaE my maps are all dis-contiguous enough as it is 09:37 everamzah Still billed as a multiplayer game, FOUND IT heheh https://github.com/minetest/minetest/blob/master/misc/minetest.appdata.xml#L7 12:10 paramat nore i'll merge game#1189 game#1102 soon. 1102 is fairly trivial so will merge if no objection but you might want to look at it 12:10 ShadowBot https://github.com/minetest/minetest_game/issues/1189 -- Doors: Fix missing node parameter passing. by t4im 12:10 ShadowBot https://github.com/minetest/minetest_game/issues/1102 -- Remove constant respawn reads, save respawn pos entering bed by tenplus1 12:13 nore paramat: lgtm 12:13 paramat ok 12:13 nore don't forget to squash the commits 12:13 paramat i won't 12:17 paramat nore any comments on game#1190 ? 12:17 ShadowBot https://github.com/minetest/minetest_game/issues/1190 -- Move Nyancats into a separate mod by rubenwardy 12:18 nore paramat: sorry, don't have the time to look now 12:18 paramat ok 13:04 Krock hehe paramat "If it works" - yes it does :P 13:14 Megaf Folks, cmake is failing the c++ test here when I use the variables CC and CXX, am I doing anything wrong here? http://paste.debian.net/plain/780869 13:18 Megaf It failes with clang 3.8, clang 3.6, gcc 6.1.1, gcc 5.4.0, gcc 4.9.3, gcc 4.8.4 13:18 Megaf fails* 13:19 Megaf hm, if I let it detect on it's own it will find cc (gcc 5.4) and c++ as compilers 13:21 Megaf ok, I think I might have figured out the issue. /usr/bin/c++: symbolic link to /etc/alternatives/c++, /etc/alternatives/c++: symbolic link to /usr/bin/g++, /usr/bin/g++: symbolic link to g++-5 13:24 Megaf I don't know why it says clang is broken. 13:24 Megaf -- Check for working CXX compiler: /usr/bin/clang++ -- broken 13:27 Megaf my libstdc libs were not installed with the compilers... 13:30 Megaf all good now, http://paste.debian.net/plain/780875 13:30 Megaf So basically apt doesn't install the libstdc++ files when you ask it to install new compilers. 13:35 Megaf #1934 fixed 13:35 ShadowBot https://github.com/minetest/minetest/issues/1934 -- TAB auto complete will no longer work on console. 13:36 paramat ok 13:36 paramat now merging game#1102 game#1189 13:36 ShadowBot https://github.com/minetest/minetest_game/issues/1102 -- Remove constant respawn reads, save respawn pos entering bed by tenplus1 13:36 ShadowBot https://github.com/minetest/minetest_game/issues/1189 -- Doors: Fix missing node parameter passing. by t4im 13:39 Megaf anyone has any update on #1556 ? 13:39 ShadowBot https://github.com/minetest/minetest/issues/1556 -- InvalidPositionExceptions thrown by some env functions 13:53 paramat merged 14:02 rubenwardy game#1190 14:03 ShadowBot https://github.com/minetest/minetest_game/issues/1190 -- Move Nyancats into a separate mod by rubenwardy 14:19 paramat error in docs 15:43 paramat hi hmmmmm this came up #4314 perhaps 'noise_params' in 'register ore' has to be the values, not a variable? not a huge problem though 15:43 ShadowBot https://github.com/minetest/minetest/issues/4314 -- Local variable ignored in ores definition 15:49 rubenwardy hmmmmm, you expressed approval for concept, may you review and approve if the implementation is ok? game#1190 15:49 ShadowBot https://github.com/minetest/minetest_game/issues/1190 -- Move Nyancats into a separate mod by rubenwardy 15:50 rubenwardy *please may 15:53 Megaf rubenwardy: I don't see any big advantage in moving nyam cat to a seperate mod 15:53 paramat i think i'll merge that this evening if there are no objections 15:54 paramat nore ^ 15:54 Megaf I have to say that I don't see any reason not do it either 15:54 paramat before conflicts arise 15:54 paramat it does help, makes stuff more flexible 15:55 Megaf there are other ways of making it more flexible, a simple configuration option in minetest.conf would do 15:55 Megaf but again, I see no reason to not do it. So yeah, lets move it to it's own mod, why not 15:56 Megaf changing topics now, is there a way to benchmark Minetest? I'd like to benchmark it, more preciselly the compiler used, GCC 4 vs GCC5 vs GCC6 vs CLANG 3 15:57 rubenwardy callgrind 15:57 rubenwardy or the inbuilt profiler 15:57 paramat i mean flexible as in making it easier to work on subgames 15:58 Megaf paramat: yep, I see nothing wrong in moving it, I think maybe we could make as much as possible in mods, so everything can be easily removed 15:58 Megaf and the engine streamlined 15:59 rubenwardy Ideally everything would be in mods, except for things used in biomes (dirt, stone, trees, water) 15:59 rubenwardy But: compatibility 16:02 Megaf well, I'm compiling 4 minetests with the mentioned compilers, http://uploadpie.com/APl58 then I will check the binary sizes, loading time and maybe use the built in profiler 19:36 Megaf Hm, so, I have these 4 minetest builds, each compiled with a different compiler. When I create a new minimal game, using the same world name, same seed and same mapgen I get totally different worlds. Is this considered a bug? 19:37 Megaf nrzkt: nore: ^ 19:37 Krock I think so. Did you verify if the seeds are identical? 19:37 T4im do you get the same with the same build? 19:37 Megaf good question Krock 19:38 Megaf ops, good question T4im 19:38 Megaf both made good questions, let me check 19:38 nore Megaf: yes, it everything is the same 19:38 Krock we're only asking good questions so far :D 19:39 Megaf Krock: Yes, they are 19:39 Megaf T4im: No, I don't 19:39 Megaf same build, same seed, different world 19:40 Megaf I think at this point anyone can confirm the bug 19:40 Megaf create two minimal v6 worlds, use the same seed, you will get two different worlds 19:40 T4im does it happen to a vanilla minetest_game? maybe it's just some minimal legacy issue 19:40 Megaf once you confirm it here I will open a bug report 19:41 Megaf T4im: let me get the vanilla 19:42 Megaf I think is a bug, If I'm not mistaken back then I could create and identical world in minimal to a vanilla one if I used the same seed 19:43 Megaf T4im: same happens with minetest_game 19:44 Megaf same build, same subgame, same mapgen, same seed, different worlds 19:44 Megaf nore: ^ 19:44 T4im I'm not having that issue here though 19:45 nore Hmmm 19:45 T4im player seems to start in slightly different positions, but always in the same beach ehre 19:45 nore This is quite strange 19:45 T4im here* 19:45 Megaf oh 19:45 T4im is your position the same? 19:45 Megaf wait 19:45 T4im heh 19:45 Megaf no, it mightn ot be 19:45 Megaf I just checked if I was looking at the same direction 19:45 Megaf hold on 19:47 Megaf T4im: nore, Krock, sorry, just the player position was changing 19:47 nore Oh, good 19:47 Megaf Still interesting thing though 19:48 T4im well, the bugs that don't need fixing are the best ones :D 19:48 Krock I assumed you teleported to the same spot every time 19:48 Megaf forgot about that 19:48 Krock Yes, there is a random spawning when no spawn is defined! 19:51 Megaf heh, my second bug that I speak of here that don't need fixing. I'm so good at spoting bugs aint I? 19:51 Megaf anyway, but here are some real bugs reported by me that still not quite fixed. https://github.com/minetest/minetest/issues/created_by/Megaf 19:51 Megaf I'm currentely further investigating #2408 19:52 ShadowBot https://github.com/minetest/minetest/issues/2408 -- Server's clock seems to change acording to dedicated_server_step 19:52 Megaf And the title is terrible 19:52 T4im what kind of clock are you using for #2408? 19:52 ShadowBot https://github.com/minetest/minetest/issues/2408 -- Server's clock seems to change acording to dedicated_server_step 19:52 T4im oops, that was unnecessary 19:53 Megaf T4im: by that I mean the /time command, the time that is reported 19:53 Megaf so if I join at 12:10, leave at 12:11 and return at 12:20 I expect the server to be at 12:20 too, but that don't happen 19:53 Hijiri is there already an issue for recyclable textures? 19:53 T4im well /time is get_timeofday, i.e. gametime 19:53 T4im not real time 19:54 T4im that's imo correct to pause when noone is there 19:54 Krock ^ 19:54 Megaf I will check if it actually pauses or not, I'm waiting 10 minutes 19:54 Megaf I left the server at 50 minuts, will return there at the next full hour 19:54 T4im there is simply no point in wasting cpu cycles when noone is around :D 19:54 Megaf then lets see what time the server will report 19:55 Megaf T4im: still, server still uses up to 40% of CPU with nobody online 19:55 Megaf s/uses/wastes 19:55 T4im so the inverse might be the bug 19:55 T4im ;) 19:55 T4im "stuff runs, when it wouldn't have to" 19:56 Megaf my theory is that there are machines that are keeping blocks active, and I don't really mind about that, I mind about a correct time keeping if the server is active 19:56 T4im that seems a likely reason, yea 19:56 T4im time keeping to what end though? 19:57 Megaf my server runs in real time, so its supposed to be in real time 19:57 Megaf by real time I mean, UTC -3 19:57 T4im then /time is the wrong place though 19:57 T4im minetest.get_time_us is what you are looking for 19:57 Megaf so the sun sets and rises at the "same moment" as it does in real life 19:57 T4im (us is actually Ms/µs) 19:57 Megaf How is it calculated? 19:58 T4im from the system clock 19:58 Megaf so it should indeed be tracking the time correctly 19:58 Megaf IMHO 19:59 T4im get_us_time* sorry, switched name around 19:59 T4im well yes and no, it's kinda overflowing all the time :p 19:59 Megaf get us time, dat name dough 19:59 T4im you probably don't even need it that precise 19:59 Megaf I'd be happy with a second of precision 20:00 Megaf it would maybe use less cycles then? 20:00 T4im os.time might be enough 20:00 Megaf full hour now, opening hte server console again 20:00 Megaf hm 20:00 T4im hm nah, if you run os.time all the time, it might be even more expensive (not sure though) 20:00 T4im it's more of a calendar 20:01 T4im there is os.clock, but it's cputime 20:01 Megaf spot on... 2016-07-13 16:50:09: [Server] Megaf sets time to 16:50 when I left the server 20:01 Megaf urrent time is 17:00 when reopened the console 20:01 Megaf maybe it takes more than 10 minutes for it to go crazy 20:01 T4im hm? 20:02 Megaf I don't get it either 20:03 T4im I don't get what you meant :D