Time Nick Message 00:00 paramat there is no other solution 00:00 octacian It's a fix (not a hack) because it is the only solution - aside from making the server deal with looking for nodes. 00:00 Hijiri I think I would favor a stricter version of rubenwardy's suggestion: Just require the search only require nodes that are visible to the player already 00:00 paramat ok now i know what the fuss was about, and opening an issue .. 00:01 octacian Hijiri: That'd be great 00:01 Hijiri I don't think it's any more of a solution than just disabling it though 00:01 Hijiri they're both solutions, just one allows more mod use 00:01 Hijiri And why can't "the only solution" be a hack? 00:02 octacian Disabling removes the feature and makes it useless. Actually fixing the underlying problem (to as much extent possible), makes it still useful to players while relatively safe. 00:02 Hijiri octacian: My yardstick has been, "does it make the cheat useless, while not making other mod features useless" 00:02 Hijiri Originally I thought that reducing the radius would make it as useful to non-cheat mods as it is to cheat mods 00:03 Hijiri But if it can only return things in the line of sight then that would be fine I think 00:03 octacian Same. Only allowing visible nodes to be searched for would make the cheat useless, while not making other mods features useless as you said 00:03 octacian So, I'm all for that concept 00:04 Hijiri The real solution for ore-hiding would be to not send ores until they are visible, though 00:04 Hijiri that would protect against modified clients as well 00:05 Hijiri and visible can mean "exposed to air", or else it might be too expensive to do 00:05 octacian The only solution is to not send nodes to the client whatsoever if they are not visible 00:05 red-001 \o/ pushed non run_in_place win64 build to itch.io 00:05 Hijiri (for ore-hiding, not CSM limits) 00:05 octacian However, that would mean that every time you broke a node the node beneath it would have to be loaded - pretty inefficient 00:05 Hijiri octacian: Yes, it would be expensive to hide every node 00:05 Hijiri But you would get most of the value of node hiding by just hiding the ores 00:06 octacian I think that only allowing searching for nodes that are visible is the best solution 00:06 Hijiri You could even reveal a whole vein if one node gets visible, since the player would have found those anyway 00:07 Hijiri octacian: I think it's the best solution involving only CSM, but not sending ores will prevent cheating by modified clients (I already said this) 00:07 Hijiri And I don't think it would be expensive if you only apply it to certain nodes, and reveal an entire connected group when found 00:08 Hijiri (or maybe a block-aligned piece of the group, whatever is more efficient) 00:08 octacian Also, damn my horrible internet... 00:08 Hijiri octacian: It doesn't have to be per-player, either 00:08 octacian I gotta right a script to auto-ghost 00:08 octacian Hijiri: True 00:09 Hijiri If a player finds valuable ore they are likely to mine it straight away anyway 00:09 Hijiri so it won't be too valuable to handle the case where another player finds ore that another player dug earlier 00:09 Hijiri It might be possible to hack in a solution for this mod side in current API, even 00:10 Hijiri replace ores with default:stone with private metadata 00:10 Hijiri then transform them into ores if they are exposed to air 00:10 octacian Still a bit complex. 00:10 Hijiri But possible 00:10 Hijiri not sure about performance 00:11 Hijiri It would only need to be done once 00:11 Hijiri by done once I mean coded once 00:12 Hijiri The part I'm not sure of is how to efficiently check if it's exposed to air 00:12 Hijiri It is easy to see if it's generated exposed to air, by LBM 00:13 Hijiri And if someone digs something you could use a dig callback to check nearby nodes 00:13 octacian Would be terrible performance wise though 00:13 Hijiri but it could be a problem if things other than players modify nodes, like by moving sand 00:13 Fixer yeah, leveldb is not influencing 00:13 Hijiri octacian: It wouldn't be that bad, it would only happen once every dig 00:13 octacian uhhh.... That'd be bad :P 00:14 Hijiri that's like max once a tick per player, though 00:14 Hijiri and usually you don't have every player on a server digging constantly 00:14 Hijiri If you check on dig you only need to check the nodes around the dug node 00:15 Hijiri a few searches in a tiny radius and metadata gets per tick is not that expensive 00:15 Hijiri Minetest API overhead isn't THAT bad 00:16 Hijiri I think less than once a tick per player, dig time was minimum like 0.15 seconds right? 00:17 octacian Dig time can be immediate 00:17 Hijiri yeah, but not for most nodes 00:17 Hijiri you won't have most players constantly digging immediate dig time nodes 00:17 octacian True. But there could be tools with immediate digtime 00:17 Hijiri by immediate do you mean bypassing normal digging and digging through on_use? 00:18 octacian No, you can make the tool capabilities so that it will dig all or some nodes instantly 00:18 Hijiri oh 00:18 Hijiri Is it literally immediate though? 00:18 Hijiri does the line of nodes immediately disappear as soon as you click? 00:18 octacian Yeah, literally immediate 00:19 Hijiri Ok, but that's still limited by the tool range 00:19 octacian Yes 00:21 Hijiri Server operators could choose not to have mods that add things that do that for common nodes, though 00:22 Hijiri And then add a further limitation to the mod that "exposed ore" means bordered by a non-ore non-stone node 00:22 Hijiri then you just need to check when stone and ore is dug 00:25 Hijiri do mapgen callbacks get called before or after ore generation? 00:29 Fixer paramat: by using mesh_gen_interval at 15ms I've missed pretty big stutter improvement it seems, numbers confirm, not sure if celerons cache thing or fix of mem leak of it or whatever, but it does look much better and obvious, stutter still happens but less and much less annoying 00:30 Fixer i can tune down my mesh gen lag to perhaps 5 ms 00:31 paramat i use 0ms and c55s work smoothed it for me 00:33 octacian Is this the issue where sometimes when you click many times in a small time period with high latency you get seemingly teleported around? 00:34 Fixer numbers confirm that stutter was reduced before 0.4.16 release 00:35 * octacian will test later 00:35 Fixer i'm very excited it went into release 00:35 Fixer gn 00:35 octacian It's always been incredibly annoying for me when harvesting crops 00:42 Calinou 4 more karma, and I have 6,000 reddit link karma! :D 00:43 Calinou but I'm going to sleep, night 00:57 Shara Really can't play the game with 'new' sneak. 00:58 Shara And already had one player ask if I am okay with him using old client just because he hates it too. 00:59 Shara I haven't updated the server yet, so didn't add in code to enable old move... I plan to when updating though. 00:59 rdococ hi 00:59 Shara Hi rdococ 01:15 bigfoot547 .back 01:15 * bigfoot547 had one job... 01:15 * octacian welcomes bigfoot547 to the backness 01:16 * rdococ says hi 01:16 * octacian has no clue what he's saying 01:16 * bigfoot547 talks in 3rd person more. 01:17 * rdococ eats cheese chickens 01:17 * octacian eats a donut 01:17 * rdococ eats a duocylinder 01:17 * bigfoot547 writes a game in c with ncurses. 01:17 * octacian eats noodles 01:18 * rdococ drinks a 3-sphere 01:18 * octacian eats rdococ 01:18 octacian Or wait, is that going too far? 01:18 * bigfoot547 thinks that maybe it is? 01:18 * rdococ proceeds to eat octacian, resulting in a feedback loop where rdococ is in octacian is in rdococ is in octacian is in rdococ, and so on. 01:18 * octacian is going to assume so seeing as rdococ didn't answer 01:19 octacian Oh boy... 01:19 * octacian didn't see that one coming 01:19 * bigfoot547 eats ncurses 01:19 * rdococ makes the universe implode on itself, taking the opportunity of everything else travelling towards it to eat bigfoot547. 01:19 * octacian eats minetest 01:20 * bigfoot547 enters an inter-dimensional portal to hide from rdococ trying to eat me. 01:20 * octacian goes *boom* 01:20 * rdococ eats the interdimensional portal. 01:20 rdococ tastes like magic. 01:21 * bigfoot547 is stranded, and the portal's other end was in rdococ's stomach, solving his end despite my best efforts. 01:21 * octacian was able to get in a spacesuit before he went *boom* 01:21 * octacian is therefore still alive 01:21 * rdococ eats the spacesuit 01:22 * bigfoot547 gets digested. 01:22 * rdococ suddenly turns into the universe 01:22 * octacian escapes into bigfoot547's portal 01:22 * rdococ is actually a gigantic creature whose interior is the alternate dimension that all portals link with 01:23 bigfoot547 g_map[i+1][curs_pos.y] = selected; 01:23 bigfoot547 Wait, I wasn't in my IDE? 01:23 bigfoot547 Oh, no... 01:24 * octacian finds a dangerous space crystal in the alternate dimension, accidentally eats it, and is transformed into a *I give up* 01:24 octacian I got more important things to do with my weekend now, so.... Yeah. 01:25 rdococ aw. now I'm sad. 01:25 bigfoot547 My weekend is over in 5 hours. 01:25 bigfoot547 :( 01:26 octacian Practically same for me... 01:26 rdococ I'm sad now 01:27 bigfoot547 It's ok, rdococ, it's just monday... 01:27 rdococ oh god, not monday 01:28 bigfoot547 Ok, time to stop lying to myself... monday sucks! 01:31 bigfoot547 Does anyone have experience with pair_content() in ncurses? 01:38 rdococ nope. I don't even have experience with ncurses. 01:39 bigfoot547 Ok, thanks anyway. 01:44 rdococ hi. 04:29 octacian Night all! :D 05:42 Raven262 Greetings everyone 09:45 IhrFussel I'm freaking out!! It worked for a few hours and now it starts to disconnect again...and my hoster expects me to pay $ 10 / month JUST to get a 24/7 support 11:20 red-001 wb Zeno` 11:22 Shara Hi Zeno` 11:23 Zeno` thanks 11:23 Raven262 Hello Shara Zeno` red-001 11:23 Shara Hi Raven262 11:30 CWz Zeno`, 11:31 Zeno` hi 11:32 CWz Zeno, minetest has compile issues 11:32 CWz https://www.hastebin.com/kukiroruva.vbs 11:33 CWz what the issue? 11:34 red-NaN Calinou, your appimage exits just after displaying the window 11:34 Calinou red-NaN: what's the console output? 11:34 Calinou it might be because you have a bogus font_path 11:34 Calinou but pastebin the console output anyway 11:35 red-NaN the old appimage works btw 11:35 Shara Hi tenplus1 11:35 nerzhul CWz, use a C++11 compiler. 11:35 tenplus1 hi folks 11:36 tenplus1 hey Shara 11:36 Shara Yay, beat you for once :P 11:36 nerzhul 0.4.16 was released, now we switch to C++11 11:36 tenplus1 ehehe 11:36 red-NaN https://pastebin.com/fcSwRBEq 11:36 tenplus1 hi nrz, wait... wasnt 0.4.16 compiled using c++11 ? 11:36 tenplus1 hi red 11:36 nerzhul tenplus1, no 11:36 nerzhul we said we will do a break release after 0.4.16 and we will use C++11 11:36 tenplus1 then what did I install last night ? could have sworn that was c++11 11:37 nerzhul i don't know what you do, but if you follow master there can be breakage starting now, we are in the 0.5 cycle. The c++11 breakage is only for compiler, there will not be any functional change atm 11:38 red-NaN btw is there something like appimage that doesn't pack all the data into a single file? 11:38 red-NaN that would work better with itch.io 11:38 Fixer nerzhul: btw i had success with building on MSYS2/mingw64 (win7), preC+11 minetest, GCC v6.3.0, had linking (?) problem for libleveldb 11:38 tenplus1 hi fixer 11:39 Fixer gonna try newest git with updated libraries 11:39 nerzhul Fixer, update your leveldb lib 11:39 Fixer hi tenplus1 11:39 nerzhul sfan5 updated it in the build process for our CI 11:39 CWz how do i switch an 11 prosessor 11:39 red-NaN but not the build bot? 11:39 CWz *compiler 11:39 Fixer nerzhul: ok, i'm writing manual for people who might want build on windows without installing that MSVC begemoth 11:39 Fixer msys2/mingw64 is much smaller 11:40 tenplus1 https://code.launchpad.net/~minetestdevs/+archive/ubuntu/daily-builds/+packages built with c++11 11:40 nerzhul tenplus1, but trusty and vivid failed because no C++11 compiler 11:40 nerzhul (by default) 11:41 tenplus1 yeah was wondering about those :P 11:41 nerzhul just update to 16.04 11:41 * tenplus1 is happy to be running latest under new compiler :pPpP 11:41 nerzhul no reason to keep 14.04 LTS 11:41 red-NaN Calinou, could my existing install if 0.4.15 be whats breaking the appimage? 11:41 red-NaN of* 11:43 Calinou red-NaN: maybe, not sure 11:43 tenplus1 hi Cal 11:43 Calinou nerzhul: AppImages should be made to run everywhere, not just on recent distros 11:43 Calinou I'm aware many desktop users use 2016+ distros now, but not all, sadly 11:44 Calinou anyway, 0.4.16 stable can be built on 14.04 fine 11:44 Calinou but Git can't now, I guess 11:44 tenplus1 hi krock with an underscore :P 11:44 nerzhul appimage doesn't no run on old distro because they dont have appimage packages 11:44 Calinou hi tenplus1 btw 11:44 Calinou nerzhul: AppImage does not require any sort of runtime... 11:44 Calinou the distro does not need to explicitly support it, this is exactly the point 11:45 Calinou (since most distro developers have been uncooperative) 11:45 tenplus1 am surprised that anyone is using anything prior to ubuntu 16.04 lts 11:45 nerzhul then just build it with libstdc++ with c++11 support :p 11:45 nerzhul unoffial images are not our problem 11:45 nerzhul but if you have a simple and working process to generate appimage i can add it to gitlab CI 11:46 nerzhul i'm starting to implement automated windows daily builds 11:46 Calinou it's also a problem for server users, ie. RHEL/CentOS 7 users 11:46 Calinou which was IIRC released in 2013 or so 11:46 Calinou and will be supported for 10 years or so 11:46 Calinou also Debian oldstable, which is still Debian 7 right now 11:46 Calinou (soon 8) 11:46 red-NaN nerzhul, could you build both run_in_place and non-run_in_place builds? 11:47 tenplus1 how much bigger is the run in place builds ? 11:47 Calinou most users don't use run-in-place actually 11:47 Calinou they don't need portable builds 11:47 Calinou (as in, the kind of builds you carry on an USB stick and are entirely self-contained) 11:47 Calinou that's a minority use case for gaming, really 11:48 tenplus1 wb Krock_ 11:48 Krock_ well great 11:48 Krock_ hello tenplus1 11:48 CWz so g++ 4.9 is too old 11:48 Krock_ "Krock" is locked because I missed to type in the password now that I have to set up my configuration <.< 11:48 tenplus1 ouch 11:49 Krock_ hmm.. now I only must manage to get my PCI wireless card working and I'm happy with this system 11:49 Calinou CWz: yeah, ideally you should have GCC 6.x for full C++14 support 11:49 Calinou (it's also enabled by default as of 6.x) 11:49 Calinou 5.x is enough for C++11, I believe 11:52 CWz man VanessaE is going to be pissed 11:53 nerzhul g++4.9 is too old for move constructor. Use Clang 3.6 or GCC 5.1 at least 12:01 tenplus1 is ClearLinux compiled using c++14 to gain max speed/optimization from the system ? 12:04 Jordach http://imgur.com/gallery/aSoQE :thinking; 12:05 nerzhul c++14 without using all c++14 features doesn't mean free performance 12:05 tenplus1 lol 12:05 tenplus1 am curious to how clearlinux has better performance than most in certain aspects ;P 12:07 nerzhul clearlinux has marketing performanec 12:07 nerzhul and also needs specific CPU to work as intended, in fact the performance gain is also due to the fact you must use a specific CPU generation to make all compiled code use all new x86 features, it's same if your compile your own programs using -march=native -mtune=native 12:08 tenplus1 ahh kewl :P 12:08 nerzhul program will not be portable across CPU 12:08 tenplus1 hi talus 12:09 tenplus1 hi DS-minetest 12:09 Jordach nerzhul, smells of intel 12:09 Jordach £99 adapter to USE RAID 12:09 DS-minetest ahh, tenplus1 is back. hi! 12:09 Jordach and must be INTEL drivers 12:10 Jordach AMD Threadripper coming through 12:10 tenplus1 threadripper ??? 12:11 nerzhul i'm for AMD because it cost less, and i can use my money i other useful things. Intel has sufficient money no need for me 12:11 Jordach Ryzen on mega steroids 12:11 Jordach it's also noted the mid range Ryzen destroys intel's i5 and i7 ranges 12:11 tenplus1 ahh nice... amd really is getting major performance gains lately, especially open-source gfx :PPP 12:11 Jordach tl;dr their IPC is at Haswell level 12:11 nerzhul yes, i run bioshock infinite in ultra mode on Linux with my 7970 (4 years old cards) 12:12 Jordach but since Threadripper and Ryzen 17xx supports HEDT 12:12 Jordach (intel's i9 doesn't support it properly) 12:12 tenplus1 phoronix did a comparison and os driver is better than binary blob in most cases 12:14 Jordach tenplus1, planning on going DDR4, M.2, Ryzen / Threadripper, AMD RX 580 12:14 tenplus1 o.O sounds like you'll rip a hole in space and time with that lot... 12:14 Jordach tenplus1, you can take some of my spare parts, or Shara can 12:14 tenplus1 hopefully soon mt-engine will make use of more than 1 core 12:15 Zeno` it already does 12:15 tenplus1 it does ? thought it was cave generation only ? 12:15 tenplus1 also hi Zeno 12:15 Zeno` http://dev.minetest.net/Engine_structure#Threads 12:15 Zeno` hi :) 12:16 Fixer hi Zeno` 12:16 Jordach http://i.imgur.com/VYVbrGw.mp4 12:17 * tenplus1 did it 1st time as well 12:17 CWz gonna need to figure out how to build using version 6 compiler without sudo 12:18 Jordach GIB 12:18 Jordach http://i.imgur.com/rZoxs9i.mp4 12:19 Fixer kitsunemimi is so slow at times 12:19 Fixer wonder what sfan is seeding 12:19 Jordach 4 letters, p 12:19 tenplus1 ocd senses tingling 12:19 Fixer speed is back 12:20 Fixer doing clean buildbot64.sh via msys2/mingw for testing 12:20 Fixer if all goes well I will move to compiling via msys2/mingw, so much faster than in VM 12:22 Krock_ copying files part II... bbl 12:22 Fixer Krock_: installing Windows 98 SE? 12:22 tenplus1 lol, I still got mah 98 cd 12:23 Fixer also MSYS2 does not support Windows XP anymore, rip 12:23 Fixer tenplus1: multiwindows dvd* 12:23 Fixer and cd* 12:23 tenplus1 slow but sure ppl are moving on from xp... 12:25 Fixer tenplus1: xp history will repeat with 7 12:25 tenplus1 true, but they say 70% + still use win7. 12:26 Fixer tenplus1: win7 will live as long as xp 12:26 red-NaN btw I pushed the windows-32bit 0.4.16 and appimage to itch.io 12:27 Fixer minetest_game cloning is so slow, wth... it is tiny 12:27 tenplus1 someone mentioned about separating mt-enging and mt-game in linux 12:27 tenplus1 so users can install their own mt-game spinoff 12:27 Fixer 20 kb/sec 12:28 Fixer done, compilation 12:28 Fixer 4 cores 12:28 Fixer *threads 12:29 Fixer 4 core i5 costs a fortune 12:29 tenplus1 agreed, 12:29 Fixer anyway I plan to upgrade in 202X if all goes well 12:29 Fixer no need to rush 12:30 tenplus1 my last amd was athlong xp 2200+ 12:30 tenplus1 *athlon :P 12:30 Fixer tenplus1: Duron 900MHz 12:30 Fixer my last one 12:30 Fixer AMD Athlons 1/64 was good 12:30 Fixer K6 too 12:31 Fixer friend had that PC 12:31 tenplus1 supposidly the intel has better head dispertion 12:31 Fixer had memory errors ,_, 12:31 Fixer damn it is fast, already 71% 12:32 Fixer it is amazing Win98SE + Office 97 run on 16MB memory 12:32 tenplus1 wow, nicely done :PPP 12:33 Fixer had kernel.dll errors though, probably fan(fun) was not spinning on K6-2 CPU 12:33 Fixer it had HL installed, but I have not looked into it 12:33 Fixer 100% linking 12:34 Fixer create package 12:34 Fixer done 12:34 Fixer about 5 minutes to compile 12:34 Fixer great success 12:34 Fixer nerzhul: yep, updated libleveldb and it worked 12:35 Fixer dfd79c08 12:36 tenplus1 whats the best map format to use ? 12:38 sfan5 Fixer: the server has a (shared) gbit connection so it's unlikely that the problem is on my side 12:38 Fixer sfan5: probably on my side 12:39 TommyTreasure tenplus1, suggestion, request, inquiry. you decide. lol. possibility of adding a 'hide' function to exposed protection blocks. similar to the way 'sound blocks' can be hidden. 12:40 nerzhul sfan5, i have working gitlab-ci windows daily build for win32, i'm doing it for win64 now (using the buildbot script) 12:40 Fixer sfan5: probably, since github speed also fluctuated a lot like kitsunemimi 12:40 tenplus1 hi tommy, and yeah that sounds a nice idea 12:40 nerzhul sfan5, => https://gitlab.com/nerzhul/minetest/pipelines/8775767 12:40 nerzhul i will provide a PR within an hour to add this capability, this will permit to have windows automated daily build on the gitlab infrastructure :) 12:40 Fixer looking 12:41 nerzhul in ~5 mins win builds should be finished and artifacts downloadable 12:41 sfan5 nerzhul: make sure to add the required DLLs as the build will otherwise not run 12:41 nerzhul maybe you can help me to finish i don't know the ending. CPack does a zip i see, i added it as artifact 12:42 tenplus1 am still building mine: https://www.ft.com/__origami/service/image/v2/images/raw/http%3A%2F%2Fftalphaville.ft.com%2Ffiles%2F2015%2F04%2FHamstergif.gif?source=Alphaville 12:42 tenplus1 hi nathan 12:42 NathanS21 hi tenplus1 12:42 nerzhul seems i set the wrong path for artifact :p 12:43 nerzhul sfan5, https://gitlab.com/nerzhul/minetest/builds/17900972 12:43 nerzhul can you download the artifact and verify if the zipfile inside it is good ? 12:44 sfan5 you can try yourself in wine 12:44 nerzhul sounds reasonable heh 12:44 sfan5 also you need libgcc*.dll libstdc++-6.dll and libwinpthreads-1.dll 12:45 sfan5 (if wine complains about missing crypt32.dll then that's not a problem) 12:45 tenplus1 am still having problems getting wine to install .net runtimes 4.0 12:45 sfan5 yup the zip is missing it 12:48 nerzhul sfan5, are they in the mingw distro we install for building ? 12:49 sfan5 /usr/i686-w64-mingw32/bin or /usr/x86_64-w64-mingw32/bin 12:51 nerzhul nice 12:51 nerzhul i don't know if i can split the build and the package methods 12:51 nerzhul could be nice to have same pipeline for linux and windows (except install 12:51 nerzhul it can permit to have intermediate unit tests before packaging 12:53 tenplus1 hi octacian 12:54 octacian Morning tenplus1 :D 12:54 nerzhul sfan5, can i remove make package -j2 from buildbot and call it directly from script.sh ? 12:55 nerzhul (i talk about a mergeable fix into minetest on github) 12:56 sfan5 why? 12:56 nerzhul to permit consistency in my gitlab pipeline where linux build have build step and after build they spawn a new docker container which only package 12:57 nerzhul or maybe add an option to disable packaging, could be better 12:57 sfan5 yea 12:57 sfan5 because otherwise the buildbot would no longer do what its name implies 12:57 nerzhul :) 12:57 sfan5 add a variable like NO_PACKAGE which causes it to call make instead of make package 12:58 nerzhul yeah i called it like this heh 12:58 nerzhul if [ "x$NO_PACKAGE" = "x" ]; then 12:58 nerzhul make package -j2 12:58 nerzhul fi 13:07 nerzhul interesting, make package also builds :p 13:16 tenplus1 hi RobbieF 13:16 sfan5 always run make -j2 and optionally make package 13:20 nerzhul y 13:23 tenplus1 bbl 13:25 Fixer btw 13:25 Fixer sfan5: why not "make package -j$(nproc)" ? 13:25 nerzhul for the CI it's not optimal 13:25 nerzhul because you see more CPU than available 13:28 RobbieF hi tenplus1! 13:54 nerzhul nice the pipeline is working properly hehe sfan5 13:55 nerzhul https://gitlab.com/nerzhul/minetest/pipelines/8776980 13:55 nerzhul i just cleanup build paths 13:55 nerzhul and i should find how to add the missing DLL 13:56 Fixer My msys2/mingw64 minetest compilation instruction, rate and comment: https://forum.minetest.net/viewtopic.php?f=42&t=17797 14:27 Fixer sfan5: would you like to compile nonluajit 0.4.16 stable? :) 14:47 sfan5 later today 15:22 Fixer preoctacian hardware 15:34 * benrob0329 steps in 15:35 * DS-minetest tells benrob0329 to take his shoes off 15:36 benrob0329 Quite literary too :P 16:30 Jordach AAAAAAAAAAAAAAARGGHHH 16:30 Jordach AMD Ryzen 7 CPUs Get up to 23% Price Cut 16:30 Jordach *autistic screeching* 16:30 Jordach and i've got no money 16:42 garywhite I had an idea: What about an I'm Feeling Lucky button that joins a random server? (Joke, not intended to be taken seriously) 16:44 Krock looks like the PCI wireless card is detected now \o/ 16:46 tenplus1 allo :P 16:48 benrob0329 tenplus1: https://youtu.be/jfLvXhN4uCo 16:49 red-NaN nerzhul, it finnaly happened https://github.com/DS-Minetest/csm_com 16:49 tenplus1 hi benrob 16:49 red-NaN well we better add an offical version of this before that mod becomes a must have to play on servers 16:50 garywhite what mod? 16:50 tenplus1 ah crap... gotta have a setting to disable CSM for servers or have it enable CSM with only mods specified by server 17:05 CWz i just witnessed a player break though steel door 17:05 tenplus1 lag ? or completely removed ? 17:05 CWz lag i guess 17:06 Shara Steel door is really easy to get through 17:06 tenplus1 yeah... protected doors are much better... :DDD 17:07 garywhite I've broken through steel doors before...now if they used no tools or a wood tool, then I'd be concerned 17:07 Shara Stick two protected doors back to back, so they are touching.. almost impossible to lag through that 17:07 tenplus1 and enable protector.flip and it's impossible 17:08 * DS-minetest never liked to break through steel, best entrance is roof 17:08 tenplus1 plus steel doors are next to useless without protection mod anyhoo 17:08 Shara Protector flip is horrible 17:08 tenplus1 works tho :DDD 17:08 Shara :P 17:08 tenplus1 maybe steel doors could be made so that you require a tool to rmove... breaking does nothing 17:09 Shara If tool checks owner name, then yes 17:09 tenplus1 wb Raven262 17:09 Raven262 Hello tenplus1 17:09 tenplus1 exactly shara... so nothing will break door apart from that tool and owner 17:10 Shara Hmm, might look at doing a mod along those lines eventually. 17:10 Shara Special armoured doors and armour plating for buildings :P 17:10 tenplus1 hi lisac_ 17:11 tenplus1 can have technic wrench (or similar) and on_punch to drop it 17:11 Raven262 Thats lisac's computer that just regained internet connection 17:11 Shara I don't really know technic well enough to know what that does :P 17:12 tenplus1 a wrench of sorts (craftitem), doesnt have to do anything, but doors on_punch will check and drop if owner :D 17:12 DS-minetest items have an on_use and an on_secondary_use 17:13 Shara I'd like to look at more than just doors 17:13 Shara Walls need protection as well 17:13 tenplus1 steel plates like you say :D 17:13 tenplus1 craft bolts + plates = protection plates 17:13 Shara So players can make proper armoured bunkers :P 17:13 DS-minetest hmm, what about a protector mod that just protects nodes in a plate form 17:13 tenplus1 and without protection mod as well 17:14 DS-minetest so, protect walls but all inside is not protected 17:14 tenplus1 players could still teleport inside and raid your stuff :) so protection still needed 17:15 Shara DS-minetest: this would be about preventing entry, rather than about preventing theft/destruction of content 17:15 DS-minetest the best thing is to teleport players away when they digged 17:15 Shara so the usual protection is a different thing 17:15 Shara Na, then someone tries to pick a flower and finds themself thrown across the map? 17:16 DS-minetest lol, just some nodeslenghts away 17:16 Shara I don't like stuff like that. 17:16 DS-minetest but it works good 17:16 Shara Yea, it's annoying. 17:16 DS-minetest or freeze the player a moment 17:17 Shara And my rule on servers is that I won't expect players to put up with anything I wouldn't put up 17:17 tenplus1 that's why I added protector.flip so if they dig all that happens is they flip around and it stops them breaking blocks 17:17 Shara flip is better than teleport at least 17:17 tenplus1 I disabled player.drop .. players using it to double up on items 17:17 tenplus1 lag unfortunately 17:17 Shara player.drop? 17:18 DS-minetest you mean they turn around their own axis like kebab? 17:18 tenplus1 yeah, when you dig a protected block it drops the tool you are using :P 17:18 DS-minetest ah, ok 17:18 DS-minetest but thats mean 17:18 Shara That's horrible anyway. Too easy not to realise something is protected 17:18 tenplus1 yeah DS... it flips them around... and if you break through ceiling it stops you falling through :D 17:19 DS-minetest ive got another idea 17:19 tenplus1 ?? 17:19 Shara The line between stopping troublemakers and punishing good players along with them always seem very narrow to me 17:19 Shara seems* 17:20 tenplus1 typically a players protected area is marked with a boundary, so others will know 17:20 DS-minetest when someone digs in protected are, make him smaller and freeze him, he will hit the ground when the node is already back. when the freeze is over, make him big again and teleport him up 17:21 tenplus1 too much involved in that one DS-minetest, especially when other mods controlling player physica 17:21 tenplus1 *s 17:21 DS-minetest hm, or why not use tnt.boom? 17:21 tenplus1 I have a protector.damage in place tohurt player also :) same as boom would 17:22 DS-minetest meh 17:22 Shara tenplus1: from my experience, they are often not marked 17:22 DS-minetest boom is cooler 17:22 Shara I had huge trouble with this on Xanadu, because literally everywhere is protected 17:23 Shara ad you can't always easily notice 17:23 tenplus1 yeah, I left players to do their thing after spawn was finished... instead of simple protecting the roads leading to places they did the surrounding land as well... not too great 17:23 Shara On DL, players bury the protectors because it looks nicer 17:23 tenplus1 +1... it's ncier to hide them 17:24 DS-minetest yeah, but it's good to see them 17:24 cheapie The protectors in use on VE's servers tie in with the areas mod, and can (optionally) be dug while keeping the area protected. 17:24 cheapie (you can of course use the commands instead and not use the protector blocks, if desired) 17:25 Shara tenplus1: It's just that sort of thing makes flip and drop behaviour more of a punishment for players who didn't realise the area is protected 17:25 cheapie It's this protector block mod: https://github.com/cheapie/areasprotector 17:25 cheapie (requires https://github.com/ShadowNinja/areas) 17:25 tenplus1 github.com/tenplus1/protector 17:26 DS-minetest do you remember when the loud voice saying that you are banned if you don't stop? 17:26 DS-minetest the voice was good, the banning not 17:27 * Shara never plays with sound 17:27 Jordach when did windows media player support FLAC ripping 17:27 tenplus1 w10 ?? 17:27 Jordach yes 17:27 tenplus1 they included a lot of codecs 17:27 tenplus1 cause the 1st thing ppl tend to do after installing windows was install ffdshow :D 17:28 Jordach klite master race 17:28 Jordach or VLC 17:28 red-NaN and then firefox/chrome 17:28 tenplus1 heh 17:29 red-NaN http://i.imgur.com/9TxWoa9.png 17:29 tenplus1 hah 17:35 lisac_ Hi tenplus1 17:35 tenplus1 :P hi hi 17:35 lisac_ Half the Serbian telecom staff are working overtime. That last power loss damaged most of the local internet modems. 17:35 lisac_ so now the internet is going down randomly 17:36 red-NaN wait so the internet is broken in literal all of serbia? 17:36 lisac_ red-NaN, Just the parts bordering Bosnia 17:36 lisac_ mostly Zlatibor region 17:37 red-NaN ah lol, I was hoping this would be more epic in scope 17:37 DS-minetest ah, another thing came to my mind. players that want to break trough protected nodes dig that node very often, that could be used to make better punishment 17:38 red-NaN like that time someone accidently broke a major fiber optic link and disconnected large parts of north america from the rest of the world 17:38 tenplus1 eek 17:39 red-NaN Armenia** 17:39 red-NaN lol 17:39 red-NaN large diffrence 17:39 lisac_ DS-minetest, why not enforce clientside mod which prevents digging protected nodes at all? 17:39 Fixer quote from linux forum: "minetest devs spend time on technology but not on actual content: sounds, textures, UI, mobs - looks bad" 17:39 lisac_ and give is_protected access to CSM? 17:40 DS-minetest lisac_: idk if that would work, i will see 17:40 Fixer i do like technology part, but that guy is right, MTG should do some kind of revolution motion 17:40 DS-minetest If any function returns true, the node isn't dug 17:40 red-NaN Fixer, well it's pretty serious if GNU/Linux people are complaining 17:40 Fixer red-NaN: not really 17:41 DS-minetest can we have client side builtin? 17:41 Fixer red-NaN: i have same vibe, engine gets better and better with each day, BUT WHAT ABOUT MINETEST_GAME? 17:41 Fixer MTG stands still 17:41 Fixer adding few blocks here and there is not enough 17:41 Fixer what is goals of MTG? 17:41 Fixer are* 17:41 red-NaN yeah mt-game is kinda dead 17:41 red-NaN if kinda means mostly 17:41 Shara No one can agree on goals for it, it seems... 17:42 Fixer red-NaN: it is not dead, but developed in certain "framework" of simple sandbox thing? 17:42 lisac_ DS-minetest, the problem is probably enforcing it 17:44 Sokomine as for the future of mtg...there are a lot of nice subgames. in particular adventuretest (perhaps in combination with ethereal?). it is far closer to what players in general like and ask for. yet most people do not seem to be aware of its existence 17:44 tenplus1 hi Sokomine 17:45 Sokomine hi ten 17:45 tenplus1 I think that's why a few ppl wanted minetest and minetest-game as separate parts, so you could add other games to repositories etc 17:45 Shara Problem is MTG is the only 'official' game, and so it's what everyone starts with 17:46 Sokomine imho adventuretest is - while no far from beeing finished or bugfree - an excellent game. yet most people even here do not seem to be aware of its existence 17:46 tenplus1 brb 17:48 red-NaN we really need to include more games with minetest 17:49 red-NaN whatever happened to https://forum.minetest.net/viewtopic.php?f=15&t=9116 17:49 red-NaN that was 3 years ago 17:49 DS-minetest btw other games have in synaptic two packages, one is -data. why doesn't minetest have this? 17:51 Sokomine a -data package with lots of mods and subgames would be nice. after all download size isn't really a consideration today anmoyre 17:51 red-NaN ^ 17:53 * DS-minetest has got an idea how to implement the client side protection already now: on join the server makes the player try to dig a protected node. if it works the player is kicked 17:53 DS-minetest but that wont work 17:54 DS-minetest >_< players could make a clientmod that only returns true on first seconds / first dig 17:59 DS-minetest ahh, now i know, always kick player when they violate protection. but it's really needed to have acces to protection in csm 17:59 rubenwardy or validate player moves using raytracing server side 18:00 Sokomine aehem. does not sound very reasonable, ds-minetest. please keep in mind that placing/digging in protected areas can also happen frequently due to not beeing aware of the area beeing protected - or due to the wish to find out who built it 18:01 DS-minetest Sokomine: yeah, but if there was a clientmod that prevents happening of violence 18:01 Sokomine happening of violence? 18:04 DS-minetest Sokomine: If any function off the ones from register_on_dignode returns true, the node isn't dug 18:05 DS-minetest most likely the on_violence of server won't be triggered 18:05 Sokomine that function always ought to get triggered. if it isn't that's reason for a bug report 18:08 DS-minetest Sokomine: but if csm stops the dig before it happens 18:09 DS-minetest isnt there a new roadmap of csm? 18:29 sfan5 Fixer: https://kitsunemimi.pw/tmp/minetest-0.4.16-noluajit-win64.7z here you go 18:30 Fixer sfan5: ty 18:30 Fixer goes to my signature 18:39 Fixer "Microsoft Leak Reveals New Windows 10 Workstation Edition For Power Users" has no telemetry nor update control /throws in a bin 18:40 * Sokomine eyes the bin distastefully and throws it into lava 18:44 cheapie Cool, now we just need "Windows 10 Server" and "Windows 10 Advanced Server" :P 18:44 red-NaN pro-tip: windows 10 updates are downgrades 18:50 paramat "minetest devs spend time on technology but not on actual content: sounds, textures, UI, mobs .." true unfortunately, not through lack of effort, just we have a low number of quality subgame and content creators. even i'm more interested in engine side of mapgen and am not a natural subgame creator 18:51 paramat i'm always hoping that others make the amazing subgame content my mapgen work allows 18:59 garywhite "Windows 10 Server Family" 18:59 rubenwardy I think a DDoS of Github is starting 19:00 rubenwardy slow to respond, https://status.github.com/ shows a recent spike 19:00 rubenwardy or maybe this time just gets more visitors 19:01 paramat "is it possible to set the meat of ignore?" =D 19:05 Hijiri red-NaN: There could be cases where the client does not know about protection on the server 19:05 Hijiri The server would have to send all relevant information for protection and keep it up to date all the time 19:05 Hijiri That will involve race conditions, for example if someone starts digging something and then the area is protected before they finish 19:15 red-NaN Hijiri? 19:25 garywhite GitHub.com is recovering 19:26 red-NaN \o/ 19:29 Calinou GitHub right now: http://users.kymp.net/feuer/etcomic/064.jpg 19:30 CWz why the ddos 19:31 cheapie 'load "tank",8,1' :P 19:31 cheapie I'll check back in a few hours. 19:38 Hijiri red-NaN: ? 19:45 Jordach http://i.imgur.com/NVsaUBK.jpg 19:46 Fixer Jordach: i9 suspiciously looks like 1999 19:46 Calinou i999 19:47 Fixer exactly 19:48 DS-minetest will there be a [CSM] Third Roadmap issue? 19:48 Jordach time for a threadripper build 19:48 Hijiri They should just have a Rolling Roadmap Issue 19:48 Jordach we're not gentoo 19:48 Fixer Jordach: threadripper furnace build* 19:49 Jordach Fixer, NOCTUA FANS TO THE RESCUE 20:13 Calinou paramat: maybe MT needs to use punkbuster ;) 20:13 Calinou hahaha 20:14 nerzhul paramat, please read PR carefully and stop telling unrelated things. 20:14 Fixer stop talking about overclocking 20:15 Fixer and please no punkbuster 20:16 lisac_ hey any idea how to make player completely invulnerable? 20:17 lisac_ player:set_armor_groups({immortal=1}) doesn't work 20:17 Jordach Fixer, use BattlEye Kappa 20:17 lisac_ it doesn't prevent lava and falldamage, for example. 20:19 nerzhul immortal armor group is only existant if all mod agreed it shoud ignore all damages 20:26 calcul0n lisac, you could register a playerevent handler to catch hp changes, and always reset it to 20 20:27 calcul0n i don't know if that would prevent death though 20:31 lisac http://dev.minetest.net/minetest.register_playerevent 20:31 lisac function 20:31 lisac This article is a stub. 20:31 lisac great :P 20:32 calcul0n yes, there are no docs iirc, if found that in some other mod :) 20:32 calcul0n so the idea is core.register_playerevent(player_event) 20:32 calcul0n then local function player_event ( player, event ) 20:33 lisac alright. I gotta go sleep now 20:33 lisac Thanks for your help 20:33 lisac night. 20:33 calcul0n where event is a string which can be "health_changed" or "health_changed" 20:33 calcul0n :\ 20:34 Fixer mmmmmmm.... ZIP 20:34 Fixer that 90s feel 20:34 Fixer still can't into some cyrillic chars 20:38 paramat server owners and moderators please support https://github.com/minetest/minetest/issues/5915 "CSM: Allowing a server to protect itself against clients running client-provided clientmods" 20:45 VaultyTowers im playing Modded Minecraft if anybody is willing to join on PC 20:45 VaultyTowers @ArcaneClassic 20:47 DS-minetest yey 20:47 DS-minetest easiest way to tell client that something is protected: tell him via chat on punch 20:51 garywhite Guys, would this be possible: A server owner is allowed to set whether or not they will allow client-side mod loading. If they choose not to allow, & a player with such mods joins, then they get auto-kicked? 20:52 DS-minetest garywhite: i wouldn't like that 20:52 garywhite Just putting it out there as a possibility, not saying we should do it 20:52 * DS-minetest has now made something with csm that the server would like 20:53 red-NaN I would suggest a three level system like this: All CSM functionsc an be used by local mods, less exploitable functions (so no get_nodes_in_area, etc) and only server provided mods 20:53 DS-minetest but i'm stupid, i should depend on the other mod i just made instead of doing the same 20:54 DS-minetest red-NaN: how would you be able to get a node in an unloaded area? 20:55 red-NaN if we want to make this system more advance I would suggest adding a way for servers to add specific functions to thier white/black lists 20:55 red-NaN DS-minetest, try random locations till you find ignore? 20:55 red-NaN dunno 20:57 DS-minetest red-NaN: i mean, the unload of nodes already makes get_node only possible in an area, right? 21:07 Jordach what if minetest was framerate dependant 21:08 sfan5 red-NaN: what you are arguing for does not change that your argument is bad 21:09 red-NaN are you saying it's easy for using to make one trivial c++ change but hard to make another similarly trivial c++ change? 21:09 red-NaN s/using/users 21:10 red-NaN since it's either that or your are misunderstanding my point 21:14 nerzhul i prefer feature based blacklist 21:14 nerzhul it's less restrictive and permits client to customize their client without remove the whole API 21:17 Calinou Jordach: fps_max = 333 in minetest.conf ;) 21:20 red-NaN eh do I need to do something extra to build minetest using c++11? 21:20 sfan5 nope 21:20 red-NaN or would it just fail to build if my compiler wasn't using it? 21:20 nerzhul it just failed 21:20 nerzhul if you use embedded jsoncpp it will fail 21:21 nerzhul with patchset 3 & 4 merged it will failed quickly, without only at linking 21:21 nerzhul i hope a nice coredev can add the approval on those patchsets :p 21:25 nerzhul https://files.mastodon.social/media_attachments/files/000/597/321/original/media.mp4 21:30 Fixer i've officially moved to building on MSYS2/mingw64 on windows 7 since today 21:30 Fixer gcc.exe (Rev3, Built by MSYS2 project) 6.3.0 21:30 Fixer GNU Make 4.2.1 21:36 * DS-minetest finished his mod: https://github.com/DS-Minetest/csm_protection 21:36 Fixer stackexchange.com is so massively helpful 21:37 Fixer no need to read 100 page manual of ZIP 21:37 Fixer google, first link, DONE 21:38 nerzhul man zip 21:38 nerzhul not need for google 21:38 Fixer nerzhul: no man pages for some reason on msys2 21:38 Fixer only --help 21:39 nerzhul change the OS => fixed 21:39 Fixer in fairness, ZIP cmdline help looks as good as in 199X 21:39 Fixer pretty short, as I remember it 20 years ago 21:39 Fixer if only it worked with cyrillics right 21:40 Fixer it works, except two chars, or one %) 21:40 Fixer still used 21:55 Jordach Calinou, e x p l a i n 21:59 nerzhul Jordach, it's a very hacky way 21:59 sfan5 but it works :P 22:00 Jordach nerzhul, hacky, but nearly un-counterable since the handshake could be anything and compiled as LuaC 22:00 nerzhul oh yes it works but it's ugly, CSM is disabled by default on every client he will just kick all clients lol 22:01 Jordach nerzhul, but anyone on a PC will enable it ;D 22:01 nerzhul why i should enable it i dont want it, it doesn't do anything interesting for my user experience 22:02 nerzhul but you can use that hack if you want while we have proper client/server negociation, it's just another bad mod because core doesn't have the feature 22:02 nerzhul :) 22:02 Jordach isn't the point of Minetest the fact that you use mods ;) 22:02 Shara I didn't see a CSM mod yet that made me care enough to try it 22:02 nerzhul mods != doing ugly things 22:03 Shara I'm sure some will show up that I will like eventually. But not yet. 22:05 nerzhul Shara, yes too early stage, i like the who formspec i saw recently 22:05 nerzhul it looks like World of Warcraft who feature :) 22:05 Shara I'm fine with /status for checking players 22:06 Shara I'm hoping to eventually see good ways to handle weather or.. something from this. 22:06 Shara But I don't really know the potential yet 22:06 nerzhul sfan5, will you get time for PR this evening or not ? (just a question because i have more patchset to finish and open and i'm waiting those for being merged) 22:06 red-NaN who isn't that usefull until the player list is fixed 22:07 nerzhul red-NaN, yeah the new packet is interesting 22:07 red-NaN by who I mean the csm "who" and "who_plus" mod 22:07 nerzhul send the list at connection, send modifications at player connection on disconnection 22:08 nerzhul i don't remember if we have on_player_connection/disconnection in CSM ? it can be an extended case with that packet 22:08 red-NaN I made a PR for it a while back but closed it due to the poor way that minetest handled the player list 22:08 nerzhul :) 22:08 sfan5 nerzhul: probably not 22:09 red-NaN nerzhul, yes this PR is designed to allow for such a PR 22:09 red-NaN thats why there is a seperate PLAYER_ADD_INIT and PLAYER_ADD 22:09 nerzhul red-NaN, it's why i prefer update_player_list package name :) 22:09 red-NaN nerzhul, I updated it 22:09 nerzhul and just name: INIT, ADD, REMOVE 22:10 nerzhul and use a list of player names not just a player name 22:10 nerzhul , 22:10 nerzhul it will permit to send bulk updates if needed 22:11 red-NaN does minetest not handle a lot of packets at once well? 22:11 nerzhul every packet trigger many calls 22:11 nerzhul here it's just in case we need to extend this with bulk 22:12 red-NaN alright I will make init use a list 22:12 nerzhul all cases 22:12 nerzhul use single for your current use case 22:12 nerzhul but at a point we can optimize some purposes, for example mass kick can use this purpose of mass removal 22:12 nerzhul overhead is just 1 uint32_t 22:12 nerzhul and don't forget to type your enum 22:13 red-NaN sure I see your point 22:13 red-NaN it will make it more future proof 22:13 nerzhul (then you can read u8 instead of u32 because we can type our enum to be the right int in C++11 :p) 22:14 red-NaN I'm trying that and it doesn't seem to be working correctly 22:14 nerzhul u8 or uint8_t because i wrongly said uint8_t (i'm using cstdint in many projects :p) 22:15 nerzhul oh you set playerListModifier in networkprotocol that's not good 22:15 nerzhul can you move it to a better header ? 22:15 nerzhul i should go to bed, see you, and thanks for the pr :) 22:15 red-NaN is there another header that both packet handlers share? 22:16 nerzhul i don't think there isn't yes... generally we don't have a such type of field, it's linked to player or end 22:16 nerzhul env* 22:16 nerzhul maybe you can let it there then 22:16 nerzhul but i don't think u8 is known in this header 22:38 Jordach http://i.imgur.com/8mYhuDQ.png 22:42 VanessaE extreme long view on VE-S: http://i.imgur.com/AZzGxR7.png (using HDX-256 yet) 22:43 VanessaE (that's near the west end of West St, looking back east after a long run, 1000m range and still getting 16 fps) 23:28 Fixer https://i.imgur.com/dG3rpyc.png 23:28 Fixer https://i.imgur.com/XkhoM2t.png 23:28 Fixer good seedc 23:47 Fixer good night, have fun