Time Nick Message 00:28 Calinou https://0x0.st/zH8w.txt 00:28 Calinou does anyone know if those global moreblocks functions are being relied upon in other mods? 00:28 Calinou I'd like to make them local to make luacheck pass 00:28 Calinou (so I can add Travis CI to moreblocks without adding more ignored functions :P) 00:28 Calinou VanessaE: ^ 00:30 VanessaE I don't think any of my mods use those 00:30 VanessaE (too lazy to check ;) ) 00:30 sofar +1 for luachekc 00:39 Calinou it's complaining about the function being unused if I make it local :| 00:42 rubenwardy !book luacheck 00:42 MinetestBot rubenwardy: Automatic Error Checking - https://rubenwardy.com/minetest_modding_book//en/quality/luacheck.html 00:42 rubenwardy (doesn't answer your question 00:47 Calinou I could add "-- luacheck: ignore" above every function considered unused but that's meh 00:48 Calinou how does one typically declare local functions that will be used in other files/by other mods? 00:48 sofar through a table 00:48 sofar but I usually don't 00:50 Calinou I forgot how modname: prefixes even work for functions too 00:55 Calinou anyway, I added `-- luacheck: no unused` to various functions fix the issue for now 01:03 Calinou there, Travis CI is set up on More Blocks now :) 01:14 VanessaE CI... on a Lua mod. 01:14 VanessaE something just sounds wrong about that. 01:14 VanessaE :) 01:58 kaeza luacheck++ 02:00 kaeza ~karma c 02:00 ShadowBot kaeza: Karma for "c" has been increased 4 times and decreased 0 times for a total karma of 4. 02:45 kaeza Reposting from #lua : [20:24] For those using luarocks.org, there was a security vulnerability reported that could potentially compromise the integrity of all packages on the site. We don't see any evidence of someon exploiting it, but I've just added bunch of security and audit tools so people check over their packages: https://luarocks.org/security-incident-march-2019 03:21 GreenDimond a lot 03:22 GreenDimond they pull from the same repos debian does 03:22 GreenDimond context: I keep forgetting to [off] 04:01 sofar Try ClearLinux! :D 04:01 sofar 04:02 VanessaE debian or gtfo. :) 04:02 sofar they don't pay me to use debian 04:02 VanessaE heh 04:02 sofar money or gtho 04:03 VanessaE eh, money isn't everything :) 04:03 * sofar sits in dry, warm house 04:04 sofar more snow outside today 04:04 sofar *g* 04:04 VanessaE yeach... that ugly, cold, white shit? no thanks. you can have your snow and you can KEEP it. :) 04:07 Emerald2 lmao sofar 04:08 sofar we had 8 inches in the last 3 days, enough to make me have to use chains again once 04:09 rubenwardy ( ͡° ͜ʖ ͡°) 04:31 * kaeza hasn't seen snow, ever. 04:33 kaeza Well, not in person. 04:34 sofar I can see snow every day of the year (clouds permitting) 05:42 benrob0329 sofar: I'm about ready to complain to management about the unruly amount of advertisements in the channel :P 07:33 paramat IhrFussel i replied here http://irc.minetest.net/minetest/2019-03-09#i_5508357 08:33 Krock o/ 09:11 IhrFussel I will reply to paramat later...too early in the morning for that now... is damage completely server side? 09:11 IhrFussel Or can modified clients ignore all damage? 09:15 IhrFussel One of my staff members told me they saw a player who apparently ignored damage by mobs and fall damage 09:18 ANAND I've seen hacked clients that ignore fall damage 09:18 ANAND I've no idea why fall damage is client-side (or is it?) 09:20 IhrFussel Fall damage is client side cause you can easily 'cheat' it with sneak at the right moment 09:20 IhrFussel (depends on physics mode on server) 09:20 IhrFussel AKA new move code vs old move code 09:28 IcyDiamond minetest's autojump is way better than minecraft's 09:28 IcyDiamond XD 09:41 ANAND Ah, good point... 09:42 ANAND I guess this is better than having it server-side - at least players will have to use a hacked client now. :) 09:42 Calinou I remember when you could sometimes avoid fall damage just by sneaking and jumping anywhere 09:42 Calinou (when landing) 09:42 Calinou that was in 0.3.x era 09:42 ANAND :O 09:43 ANAND Tried that once and my brain splattered all over the floor :| 10:22 IhrFussel Does set_hp() on players call 'on_punchplayer()' ? 10:22 IcyDiamond is there a way to make entities not go black when inside a node? 10:23 IcyDiamond i want it to behave with a nodebox 10:23 IcyDiamond but the entity is just rendered as a black model "inside" it 10:25 IcyDiamond also no way to use animated textures for entity? 10:26 IcyDiamond yeah ok that was the problem 10:26 Krock IhrFussel: no, because there's no punch action 10:27 Krock call it manually 10:27 Krock IcyDiamond: set their glow to 15 10:27 Krock or set a Y light position offset, if that's merged already 10:29 IhrFussel Krock, so set_hp() only calls on_hpchange() right? 10:30 IhrFussel I'm trying to find a bug in my mod code that's why I wonder 10:31 Krock yes 10:33 IcyDiamond setting texture on entity does nothing 10:34 IhrFussel Then it must be somewhere in that callback... thanks 10:36 IcyDiamond sigh i hate working with entities, i can never get them to work the way i want 10:41 Emerald2 Is there more info about how the noise params in mapgen effect the generated terrain? (already looked at the lua_api) 11:21 IcyDiamond https://i.lunasqu.ee/2019-03-09_13-20-05.mp4 11:26 Emerald2 What's that Icy? 11:39 IcyDiamond https://i.lunasqu.ee/2019-03-09_13-38-29.mp4 11:42 Emerald2 Oh hey that's a cool thing you're working on. 11:49 Krock nice fluids 12:43 IcyDiamond The goddamn torch keeps crashing my game 12:44 IcyDiamond I don't know what to do about it 12:44 IcyDiamond s/game/server/ 12:48 IcyDiamond Ok protectors is crashing it 12:49 IcyDiamond When you place a torch in a protected area the torch causes a crash 12:59 IcyDiamond https://github.com/minetest/minetest_game/issues/2330 13:28 Krock IcyDiamond: gdb 13:59 * Krock opens email 13:59 * Krock sees first player offset complain emails 14:14 Fixer Krock: closes email 14:16 Krock Fixer: got a directory for that. 95 emails so far 14:16 Krock sorry, 106 14:26 Fixer Sokomine: usually it is someone who used that google engine to display some maps, so I can't switch anything 14:48 IhrFussel Either I'm dumb or player:set_hp() DOES call on_punchplayer() 14:49 IhrFussel In my case I want a node to hurt a player when they dig it... I do that via set_hp() but punchplayer() seems to log this event 14:51 IcyDiamond does anyone else have this issue with hbsprint where it doesnt regenerate? 15:01 IhrFussel Either (my version of) lucky_block mod has a critical bug or set_hp() calls punchplayer() but ONLY if it gets called from a on_dignode() 15:11 IhrFussel This happens when I dig a lucky block and the random function that runs is explosion (note there is not a single :punch() call in the entire lucky block file) 15:11 IhrFussel 2019-03-09 16:09:18: ACTION[Server]: punchplayer(): IhrFussel tries to punch IhrFussel 15:11 IhrFussel 2019-03-09 16:09:18: ACTION[Server]: hp_change: IhrFussel -> -23 15:28 Fixer Krock: just two more years and then you will be haaaaappppyyy 15:31 IhrFussel Okay I'm completely confused...if I call set_hp() directly inside the on_dig() callback it works as expected without punchplayer() 15:33 Krock Fixer, IhrFussel, TommyTreasure: DO you have any input on #8282 ? 15:33 ShadowBot https://github.com/minetest/minetest/issues/8282 -- Key settings: Clear with backspace or delete by SmallJoker 15:34 Krock delete and backspace will now erase the assigned key 15:37 Fixer Krock: i guess it makes sense... 15:37 Fixer Krock: it will be nice to have it 15:47 TommyTreasure personally, i've only used key assignments twice. to move 'drop items' and to turn on 'cinematic view'. sounds good though 15:50 IhrFussel I only use default controls but the idea sounds good 15:53 benrob0329 Inventory -> Tab 15:58 Fixer Krock: "I extended that request to remove all default values from the config file" now that's nice, config will look much slimmer, if it affects all values... 16:01 Krock affects all key settings 16:02 Krock not keen on doing the same for all settings 16:04 Krock thanks for the feedbacks 16:05 Krock IhrFussel: I cannot reproduce the u_i performance issue with a debug build 16:06 Krock peak execution time 9ms 16:11 IhrFussel Krock, are you sure that you did it exactly like in my videos? 16:11 IhrFussel You need to remove a few slots of items first before pressing the button again 16:13 xerox123 o/ 16:44 Krock IhrFussel: yes, I even wrote statistics for it. start/stop measuring 16:48 Krock IhrFussel: https://krock-works.uk.to/u/u_i_speed-2019-03-09_17.45.56.mp4 16:50 Krock the first action (fills all fields) took nearly 12ms 16:51 Krock and each stack movement took in average 65ms 16:51 Krock * 65us 16:55 Krock IhrFussel: https://krock-works.uk.to/u/patches/unified_inventory_speed_measure.diff 16:55 Krock rudimentary measuring but it works 17:00 IhrFussel Krock, you don't use the button at the bottom to switch to the craft grid...maybe that's it about? 17:01 Krock it's the same functionality 17:02 IhrFussel Is it about the network then? 17:27 Krock I don't think so. The network should not affect Lua execution 17:56 IhrFussel Did you use 0.4.17.1? 17:58 Krock 5.0.0-dev 17:58 Krock * 5.1.0-dev 17:59 Krock testing in 0.4.17.1 now 18:00 Krock initial copy: 8ms 18:01 Krock two books removed: 5ms 18:04 IhrFussel Try to remove all but 1 18:04 IhrFussel For example only leave the top center item in 18:08 Krock IhrFussel: https://krock-works.uk.to/u/u_i_test04171-2019-03-09_19.07.12.mp4 18:09 Krock same result when keeping the middle item 18:13 IhrFussel 2019-03-09 19:12:45: ACTION[Server]: CRAFT TIMER: IhrFussel moved 1 default:bookshelf to craft grid (239781 µs) 18:13 IhrFussel When I only leave the center item in (very center) 18:14 IhrFussel And note I only press the '1' button 18:14 IhrFussel And that button causes 0.2 secs lag ... how is that possible then? 18:16 IhrFussel Another note: The lag does not happen if I either leave the very first slot or very last slot in 18:17 IhrFussel Wait...testing some more 18:18 Krock did you also test the PR locally, without any glibberish modifications around it? 18:18 Krock I cannot debug your code 18:20 IhrFussel I didn't change the file in any other way after applying it to my u_i version 18:21 IhrFussel Only added minetest.log() 18:21 IhrFussel I can download it to my local PC and test 18:30 IhrFussel Welp... I can't reproduce it locally with my server version of u_i either 18:30 IhrFussel Then the question is WHAT can make itemstack moving lag so much 18:30 Krock *shrug* 18:30 IhrFussel lua is singlethreaded so that should mean nothing can run while the button loops through the inventory of a player 18:31 Krock hello kaeza :D 18:31 IhrFussel So...nothing should be able to bother the loop 18:31 IhrFussel Right? 18:31 Krock IhrFussel: right 18:31 Krock the network however is threaded afaik. quite complicated structure there 18:37 IhrFussel Krock, could you change your server step? 18:37 IhrFussel To 0.15 (that's the value on my server) 18:37 IhrFussel I cannot do that somehow...it won't accept my SP server step value 18:38 Krock singleplayer does not use the server step 18:39 Krock needs [x] Host Server to be checked 18:40 Krock full craft grid copy: 7552us @ step 0.3 18:41 Krock although that does not matter at all 18:41 Krock those Lua callbacks are run async to the server step 18:42 IhrFussel Do the callbacks wait for the network? 18:42 IhrFussel That would mean something kinda slows down the execution 18:43 Krock I hope not.. those packets are usually sent through a queue which is then processed independently 19:19 Sokomine any speakers of other languages than english or german present (and willing to add translations for my travelnet mod?) 19:21 sofar yes (no) 19:21 Sokomine maybe french, (brazilian?) portugese and russian might be most urgent 19:23 kaeza Sokomine: I can do Spanish. 19:23 Sokomine that would be great, keaza 20:00 Krock Sokomine: https://github.com/Sokomine/travelnet/pull/35 20:26 IhrFussel Krock, any idea what could cause it on my server then? 20:27 Krock *shrug* I guess you'll have to debug it yourself 20:27 Krock I don't know 20:28 Krock although.. you could try to compare the bookshelf recipe with one that does not have any groups inside 20:28 Krock if the no-groups-recipe does not suffer from the performance issues, then it would be a clear case 20:28 Krock otherwise it might be network/C++ related 20:32 IhrFussel It happens with carpets too which only require 6x green wool 20:32 IhrFussel But it only happens when the craft grid was not empty before 20:34 scr267 There is a bug in skinsdb which cause textures to appear 50% below the player. I've openned an issue at https://github.com/minetest-mods/skinsdb/issues/23 20:34 IhrFussel Maybe I could check if the entire craft grid is empty and otherwise abort the moving and display a message to the user? 20:34 scr267 But if someone could point me in the correct direction, perhaps I might be able to push a patch to it 20:39 Krock scr267: already reported a year ago https://github.com/minetest-mods/skinsdb/issues/12 20:39 Krock update minetest_game and use the 5.0.0 branch 20:39 scr267 Oops, ill close the issue i openned then. 20:39 scr267 I did update minetest_game and am using the 5.0.0 branch 20:40 Krock also update 3d_armor if you've got that 20:40 IhrFussel Krock, could this 'fix' the lag for now? If I add before the moving loop 'if not player_inv:is_empty("craft") then return end' 20:40 Krock IhrFussel: well yes, but then you'd skip the entire filling function 20:41 Krock and if you still experience lags, then it's not a problem of u_i 20:41 IhrFussel Well I don't want the server to lag 20:41 Krock or at least not of my code 20:41 IhrFussel Like I said if the craft grid is EMPTY before I press the moving buttons the lag is less than 10ms 20:41 IhrFussel But if it#s non-empty the lag can go up to 300ms 20:46 IhrFussel It happens with recipes that contain groups and ones that only require single items 20:46 IhrFussel I'm not sure but I feel like the lag gets more the more slots are missing 20:46 IhrFussel So if I take only 1 item slot from the craft grid it still executes faster than when I take 8 out 20:47 IhrFussel But if I take all 9 out the execution time is almost instant 20:48 scr267 Krock: Thanks for the info. As per the issue I'm trying out the dev branch by bell07... seems to work :) 20:51 Krock i.e. the opposite of the execution time which should actually happen 20:53 * Sokomine stares angrily at always changing doors 20:58 IhrFussel My temp hack works 20:59 IhrFussel You can only move things if your grid is empty else error 20:59 IhrFussel And with that method even moving 10 takes only ~ 5ms 22:34 sofar wow, no mt 5.0.0 builds for centos yet 22:38 benrob0329 People game on centos? 22:38 sofar oh great gcc-4.8.5 22:38 sofar no, run the server 22:38 benrob0329 Then they can probably compile it :P 22:38 sofar CMake Error at CMakeLists.txt:217 (message): 22:38 sofar Insufficient gcc version, found 4.8.5. Version 4.9 or higher is required. 22:38 sofar nope 22:39 benrob0329 Or..they can't because Centos is ancient 22:40 sofar I'm very tempted to swupd take over this system 22:40 sofar since, screw RH 22:45 sofar # ./ld-linux-x86-64.so.2 ./swupd 22:45 sofar FATAL: kernel too old 22:45 sofar argh, fail 22:46 Emerald2 rip 22:48 sofar so no gcc-4.9, and no clearlinux 22:48 sofar 5.0.0 is going to be annoying on this system 23:05 sofar compiling a compiler 23:05 sofar what a flashback 23:10 sofar -g -O2 -O2 -g -O2 23:11 sofar maybe I need to add -g -O2 some more to make sure it's compiled correctly 23:16 Sokomine finally anaged to update my travelnet mod 23:16 * Sokomine hides because it took so long and many good PRs accumulated 23:17 Sokomine sofar: good luck! hope compiling won't take too long 23:26 Calinou compiling GCC on an high-end system takes about 40-50 minutes IIRC 23:36 Fixer Calinou: by all the cores 23:36 Fixer o/ o/ o/