Time Nick Message 00:00 nepugia a fosdem stand would be nice already ;) 00:00 erlehmann maybe one could connect it to another event 00:01 rubenwardy I tried to find people to run a fosdem stand with, but no one was willing 00:02 rubenwardy I've ran stalls at another conference: https://pbs.twimg.com/media/DrE5n3MW4AEt5fZ?format=jpg&name=4096x4096 00:05 paramat i highly recommend you compile MT 5.1.0, or at least 5.0.1, it is so far ahead of 0.4.x, which now feels like ancient history =) 00:05 nepugia I can second that 00:06 paramat rails, papyrus, bookshelf are still in MTGame, and we finally have minecarts to run on the rails 00:06 erlehmann wow, minetest has autogenerated sky dungeons now 00:06 erlehmann amazing 00:07 paramat ? i thought i fixed floating dungeons 00:07 erlehmann i am on 4.17, seed 2019-11-27 00:07 erlehmann i will install the latest one 00:09 paramat in the past though, dungeons would occasionally project out of the ground, or float near the ground, but people complained and i fixed them. i liked them though 00:10 erlehmann the dungeon comes out of the side of a mountain 00:10 erlehmann there is a huge cliff 00:10 erlehmann is „build-essential“ on debian recommended because you are building a debian package? 00:11 erlehmann If you do not plan to build Debian packages, you don't need this 00:11 erlehmann package. Starting with dpkg (>= 1.14.18) this package is required 00:11 erlehmann for building Debian packages. 00:11 nepugia build-essential pulls in gcc aswell as some development packages 00:11 nepugia for debian it is a usefull shorthand 00:12 paramat ha, i like your commit '+ unknown block is sick' https://github.com/minetest/minetest/blob/17a47b3335cc8797b3ce15b6023e3a5eef89d89d/data/unknown_block.png 00:16 paramat wow, oerkki used to be an 'omsk bird' https://github.com/minetest/minetest/blob/50021073bf23ce5f796624cea34a66c5d1357ce4/data/oerkki1.png 00:16 paramat commit '* enemies are now omsk birds' 00:17 paramat looks like oerkki sort of evolved from omsk bird 00:18 erlehmann hahaha 00:18 erlehmann welcome to omsk 00:18 paramat 'the dungeon comes out of the side of a mountain' ah yes, protruding. i made dungeons generate at all y, not just below water level 00:19 erlehmann nepugia, i think one should actually list the packages that are the dependencies then 00:19 erlehmann because if i autoremoved build-essential (since i do not build debian packages), it would remove the gcc again? 00:19 erlehmann is the unknown block still in there? :D 00:20 erlehmann i have now checked out git tag 5.1.0 00:20 paramat it is different now, has text on it 'unknown block' 00:20 erlehmann oh sad 00:20 paramat make sure to use MTG 5.1.0 too, they are synced 00:20 erlehmann the only MTG i know is magic the gathering 00:22 paramat sorry, 'Minetest Game' the default lua game for MT 00:22 paramat *Lua 00:22 nepugia erlehmann, apt does not autoremove packages per default :) 00:22 nepugia but surely, the dependency list could be more concise 00:22 paramat omsk bird https://knowyourmeme.com/memes/winged-doom-welcome-to-omsk 00:23 erlehmann nepugia, i propose replacing “build-essential” in the README with “g++ make libc6-dev” 00:23 erlehmann does that sound about right? 00:23 erlehmann or would it have to be libc-dev? 00:23 paramat by the way, if c55 sees this, how is 'oerkki' pronounced? 00:27 erlehmann in my head it was always /ˈœrk.ki/ 00:27 erlehmann (IPA) 00:27 erlehmann https://en.wiktionary.org/wiki/Appendix:English_pronunciation if you are confused 00:29 erlehmann building minetest 5.1.0 now 00:29 erlehmann apparently replacing „build-essential“ with „g++ make libc6-dev“ was the right thing to do 00:30 erlehmann nepugia paramat could you commit that change or would i have to submit a patch somewhere? 00:30 paramat best either open an issue with all the info or submit a PR 00:31 erlehmann rationale: build-essential also depends on dpkg-dev, which depends on fakeroot, gnupg | gnupg2, gpgv | gpgv2 and perl 00:31 paramat hm, if you are on github 00:31 erlehmann i stopped using it after they hellbanned me once 00:31 erlehmann can i just send you a patch or pull request using git-request-pull? 00:31 erlehmann or via email or pastebin? 00:31 erlehmann i mean it is 3 words 00:31 erlehmann oh, of course, the explanation is longer 00:32 erlehmann it would be “minetest accidentally depends on perl” 00:32 paramat no prob, i will make an issue from logs of this channel. pastebin the patch wuld be good 00:32 erlehmann ;) 00:32 paramat *would 00:33 nepugia erlehmann, i don't really have a github account for that, sorry :/ 00:33 paramat and i will make sure this gets attended to and a PR made 00:34 erlehmann thx 00:42 erlehmann > /home/nils/share/src/games/minetest/src/util/srp.cpp:432:8: warning: ‘void* memcpy(void*, const void*, size_t)’ specified size 4294967295 exceeds maximum object size 2147483647 [-Wstringop-overflow=] 00:42 erlehmann do i need to worry? 00:43 erlehmann > /home/nils/share/src/games/minetest/src/util/srp.cpp:432:8: warning: ‘void* memcpy(void*, const void*, size_t)’ pointer overflow between offset [0, 2147483647] and size [4294967295, 2147483647] [-Warray-bounds] 00:45 paramat warning during compilation? i cannot judge if that is something to worry about 00:46 erlehmann yes 00:46 paramat 'warning' suggests nothing critical 00:46 erlehmann [100%] Linking CXX executable ../bin/minetest 00:46 erlehmann [100%] Built target minetest 00:46 erlehmann so 00:47 p_gimeno sounds critical-ish to me 00:49 erlehmann i'll write my patch, then worry about it 00:55 p_gimeno so it's apparently believing that the default case here can happen: https://github.com/minetest/minetest/blob/master/src/util/srp.cpp#L341 00:56 erlehmann yeah so 00:56 p_gimeno I don't think it can 00:56 erlehmann it can what? 00:56 erlehmann this code looks weird to me 00:56 erlehmann either -1 is a perfectly fine value or is it not 00:56 p_gimeno happen that the algorithm passed to hash_length() is other than CSRP_USE_SHA1 or CSRP_USE_SHA256 00:56 erlehmann if it is not, why does it compile? 00:57 p_gimeno -1 is being converted to uint32_t and that's what causes the alert 00:58 erlehmann this sounds really fishy 00:58 p_gimeno it's the default value of the switch, I don't think it can happen 00:58 erlehmann i guess it should not compile at all instead of returning -1? 01:00 p_gimeno I don't know, maybe 01:03 p_gimeno to be clear, the compiler is complaining because -1 is one of the possible return values of the function, not because it's always returned. In fact I think it would never be returned. 01:04 paramat this is MT 5.1.0 stable release or latest dev version? 01:04 p_gimeno that was from master 01:04 p_gimeno I doubt it has changed in a while though 01:06 p_gimeno yeah, the file was last changed on January 01:07 erlehmann paramat, 5.1.0 01:07 erlehmann i compiled 5.1.0 01:07 erlehmann can it be that you have … undefined behaviour? 01:07 paramat yes, stable release then 01:08 paramat i cannot judge this issue sorry, i do not have much general programming talent, i am just the mapgen guy =) 01:08 p_gimeno it's converting -1 to uint32_t and then back to size_t which is signed, and the last one causes an overflow 01:09 erlehmann paramat, here is my patch, you can apply it with „git am“ https://mister-muffin.de/p/FYAf.diff 01:09 erlehmann or http://mister-muffin.de/p/FYAf.diff if you want 01:10 erlehmann i made a commit and then did „git format-patch 'HEAD^1'“ 01:10 erlehmann okay so 5.1.0 still has the „Hauptmen�“ bug 01:11 erlehmann also apparently it cannot into internet 01:12 erlehmann 2019-11-27 02:11:47: ERROR[AsyncWorker-0]: servers.minetest.net/list?proto_version_min=37&proto_version_max=38 not found (Timeout was reached) (response code 0) 01:12 erlehmann what am i doing wrong? 01:12 erlehmann if i use curl that thing does give something back 01:13 p_gimeno erlehmann: maybe that's on your side? I downloaded your patch and I see the characters encoded in something like an ISO-8859 encoding but not UTF-8 01:13 erlehmann p_gimeno, the pastebin is bad, sorry. 01:13 p_gimeno makes me wonder if the title bar requires Unicode 01:14 erlehmann srsly 01:14 erlehmann it does 01:14 erlehmann pastebin ruined my utf-8? 01:14 erlehmann i'll upload it again 01:15 p_gimeno erlehmann: what OS and WM? 01:15 erlehmann p_gimeno, Debian GNU/Linux 10 and i3. and i can definitely have a terminal that has „Hauptmenü“ in the title bar 01:15 erlehmann want a screenshot? 01:15 erlehmann font is pango:Unifont 01:16 p_gimeno hm ok... maybe the title is sent in the wrong encoding somewhere 01:16 paramat diff looks ok here, there are just a few characters that are boxes containing '0093', '0094', '0096' 01:16 erlehmann i assume that 01:16 paramat only the commit message contains those characters 01:16 erlehmann i will send a new one 01:16 erlehmann wait 01:17 erlehmann paramat p_gimeno here is the patch with the correct encoding: https://web.archive.org/web/20191127011650/https://mister-muffin.de/p/fXeo.diff 01:17 erlehmann i use this pastebin https://mister-muffin.de/paste 01:18 p_gimeno that link works for me 01:18 erlehmann and apparently when i upload text it changes the encoding 01:18 erlehmann paramat, that patch works? 01:19 paramat new paste is better thanks, downloaded 01:20 erlehmann i'll try to find the source of the Hauptmen� bug 01:20 erlehmann but maybe later 01:20 erlehmann it is 2:20:20 here 01:20 paramat i will make sure this becomes a PR and is reviewed/merged 01:20 nepugia erlehmann, hauptmenü rendering wrong might be a window manager bug 01:21 erlehmann nepugia, why should it be one? 01:21 nepugia the window bar title is rendered by the wm no? 01:21 erlehmann hmmmm 01:21 erlehmann i checked with xprop 01:21 paramat anyway, good to see you here, hope you can drop in occasionally 01:22 nepugia what is your LC_ stuff set to? i could test against mine here 01:23 nepugia hmm, don't seem to get the window title to appear in german.. maybe i didn't compile something that is required for localization 01:24 erlehmann i found the reason 01:24 erlehmann wait for the explanation 01:24 erlehmann ; xprop |grep WM_NAME 01:24 erlehmann WM_NAME(COMPOUND_TEXT) = "Minetest 5.1.0 [Hauptmenü]" 01:25 erlehmann so i set up a window that has an rc shell with a similarly named folder in it 01:25 erlehmann ; xprop |grep 'WM_NAME' 01:25 erlehmann WM_NAME(STRING) = "rc: /tmp/Minetest 5.1.0 [Hauptmenü]" 01:25 erlehmann _NET_WM_NAME(UTF8_STRING) = "rc: /tmp/Minetest 5.1.0 [Hauptmenü]" 01:25 erlehmann COMPOUND_TEXT is a legacy X11 text encoding 01:25 erlehmann and not UTF-8 01:25 erlehmann https://www.x.org/releases/X11R7.6/doc/xorg-docs/specs/CTEXT/ctext.html 01:25 p_gimeno heh 01:25 nepugia an X11 specific text encoding? 01:26 erlehmann so the issue is two-fold 01:26 erlehmann first, minetest only states its own name in the legacy WM_NAME property 01:26 erlehmann second, it uses an encoding my window manager does not understand 01:27 erlehmann i guess using _NET_WM_NAME and UTF-8 would be the way to go? 01:27 nepugia Argueably, with the expectations of backwards compat in X11 the bug is on both sides :) 01:27 erlehmann in addition to the old one, of course 01:27 erlehmann you see the terminal has both properties 01:27 erlehmann nepugia, indeed 01:28 erlehmann i found a good description https://metacpan.org/pod/Encode::X11 01:28 erlehmann > Compound text is found in window properties of type COMPOUND_TEXT. It's not usual to use it outside that context. Compound text consists of ISO-2022 style escape sequences switching among various basic charsets, including the ISO-8859 series, JIS, KSC, and GB. 01:28 erlehmann so i must admit i have no idea how to fix this 01:29 nepugia _NET_WM_NAME as utf-8 looks good to me? but i don't have that much experience with X11 either :) 01:30 erlehmann yeah someone open an issue or something i guess i don't know 01:30 erlehmann i'll tell you when i have a patch! 01:30 paramat i will 01:30 erlehmann if i have a patch at one point 01:30 erlehmann thx paramat 01:31 erlehmann nepugia, paramat you can probably reproduce it using LANG=de_DE.UTF-8 01:31 paramat ok 01:31 erlehmann if you have that locale installed 01:32 erlehmann do you need a screenshot? 01:32 paramat probably not 01:32 erlehmann ok 01:32 nepugia just recompiling with gettext to reproduce it, i would think that the xprop output is more than enough though 01:33 erlehmann yeah, but reproducability is a thing 01:34 p_gimeno fuck, my disk has mounted read-only, there's a raid problem 01:34 erlehmann next bug: 2019-11-27 02:34:29: ERROR[Main]: https://content.minetest.net/api/packages/?type=mod&type=game&type=txp&protocol_version=38&hide=nonfree&hide=desktop_default not found (Timeout was reached) (response code 0) 01:34 erlehmann i mentioned that already 01:34 erlehmann p_gimeno, i hope you do not have one of these HPE SSDs 01:35 p_gimeno erlehmann: no, it's a toshiba disk drive 01:36 erlehmann https://support.hpe.com/hpsc/doc/public/display?docId=emr_na-a00092491en_us 01:36 erlehmann here, for a RAID horror story: https://old.reddit.com/r/sysadmin/comments/e1x729/uh_oh_hpe_issues_firmware_fix_to_stop_certain_sas/f8s9yw9/ 01:36 nepugia hw raid sucks anyway :) 01:36 erlehmann > We had this issue and 6 SSDs died in the span of 15 minutes. Some data was lost and we had an outage that lasted for a few days. Not to forget we had 6h CTR warranty on that machine and they didn't have 6 SSDs on stock so it took them a few days to ship the drives. Before that we replaced the RAID controller, Expander card and cables just to be sure as nobody belived that 6 drives died in such short time. 01:38 erlehmann so any idea why minetest can not into internet? 01:40 rubenwardy Try changing the server list setting to be non HTTPS? 01:40 rubenwardy Can't remember if it is by default 01:40 erlehmann LOL 01:40 erlehmann if i use ltrace it works 01:40 erlehmann wtf 01:40 rubenwardy It's not possible to access contentdb over HTTP 01:40 erlehmann a true heisenbug 01:40 rubenwardy Oh, weird 01:41 erlehmann oh now it works without ltrace 01:41 erlehmann i guess some part is racy and now it is cached? 01:41 erlehmann FYI: i had done this 01:41 erlehmann ltrace -e '*curl*' ./bin/minetest 01:42 erlehmann and now it stopped working again 01:42 erlehmann wtf 01:42 nepugia ugh, minetest fails to link for me against gettext 01:42 erlehmann it works with ltrace 01:42 nepugia guess i won't be reproducing the issue 01:43 erlehmann this does not make sense at all 01:43 erlehmann nepugia, which OS? 01:44 erlehmann ok, then i will have to ltrace minetest all the time 01:44 erlehmann wtf 01:44 nepugia loonix 01:44 erlehmann which minetest game is a good one? 01:44 p_gimeno changing to a spare drive, bbl, good luck 01:44 nepugia but alpine linux, so not glibc :) 01:44 nepugia depends what you like... there are some popular modpacks, dreambuilder one comes to mind 01:45 erlehmann i mean, the descriptions are quite weird 01:45 rubenwardy erlehmann: perhaps it's just hitting the 5 second timeout 01:45 erlehmann > Game without Thrones 01:45 erlehmann something something agile organization 01:45 erlehmann rubenwardy, in the span of less than 5 seconds? i doubt it 01:46 rubenwardy erlehmann: nodecore is one with super strange crafting, mineclone is Minecraft like 01:46 erlehmann super strange crafting? 01:46 rubenwardy I've heard good things about legend of minetest 01:46 nepugia Maybe i'll work on my subgame a bit more heh 01:46 rubenwardy You craft things by placing them in the world 01:46 rubenwardy There are no GUIs in nodecore 01:46 erlehmann oh 01:46 erlehmann nice 01:47 nepugia i sure would like some non-blocking GUI in minetest 01:47 erlehmann rubenwardy, what would you suggest when i want a good minecraft-experience and what would you suggest, when i want a surreal experience? 01:47 nepugia had something like that implemented in gmod, but then they changed their EULA to an unacceptable one :/ 01:47 erlehmann nepugia, as someone who has drawn around 500 blocky emoji for GNU unifont, i vehemently disagree 01:47 rubenwardy Former would be mineclone2, latter idk 01:47 erlehmann i can see the pixels and it is a good thing 01:47 erlehmann also my interface font is unifont everywhere 01:48 nepugia erlehmann, with non-blocking i mean you can move around while a UI element is open 01:48 nepugia as in maybe seeing content or beeing able to interact with objects without loosing camera and movement control 01:48 erlehmann ah 01:48 nepugia basically requires that i be able to capture different inputs though, like arrow keys for instance 01:48 erlehmann sorry 01:49 nepugia I am not quite sure what you interpreted my sentence as, but i certainly don't take any offense in someone disagreeing ;) 01:49 paramat make sure to have 'Minetest Game' 5.1.0 installed, not exciting but a high quality general sandbox 01:49 nepugia I really like the new formspec real coordinates system 01:52 erlehmann nepugia, i thought of „non-blocky“ 01:52 erlehmann instead of „non-blocking“ 01:52 erlehmann paramat, i have it 01:53 paramat no mobs in it though, but intended 01:53 erlehmann my friend liked that minetest had no mobs by default 01:54 erlehmann she said that her problem with minecraft was always the mobs appearing at night 01:54 paramat they will be optional if added 01:54 paramat heh 01:54 erlehmann which kinda ruined it for her 01:54 erlehmann i mean, it is like playing with lego 01:54 erlehmann and then a crazy child comes and kicks your lego model and bites your arm 01:54 erlehmann and explodes 01:55 erlehmann also it is suddenly night 01:55 paramat i would certainly like to add the oerkki 01:55 paramat ... a 3D version 01:56 paramat worlds now have a start time of early morning, it is also settable 01:56 erlehmann so it says it is downloading and installing mineclone 2 01:56 erlehmann but there is no progress bar 01:56 erlehmann how big is the mineclone? 01:57 erlehmann i mean, according to nethogs, the minetest binary is downloading something 01:57 nepugia Ah, blocky or not is just a visual preference imo :) 01:57 nepugia for a non-blocking UI for my stuff i'd basically need the serverprovided clientside mods first though, and then a method to capture some input 01:57 nepugia heh, mobs are optional in minecraft though, atleast hostile ones 01:57 erlehmann oh i have it now 01:57 erlehmann a progress bar would be nice though 01:59 paramat i do find default day length far too short though, only 10 mins of daylight, default is still set to the same as MC 02:02 nepugia Hmm, the pr i had made for default stack size settings is still open 02:03 erlehmann rubenwardy, oh you build the contentdb? 02:03 erlehmann nice 02:03 erlehmann built 02:03 rubenwardy Yes 02:03 rubenwardy It was a university project 02:03 nepugia paramat, personally i just really hate nights in minetest, far more than i do in minecraft, its just way too dark too do anything usefull and torches help very little 02:04 rubenwardy Minetest is way too dark even with torches 02:04 paramat lol 02:06 nepugia The pr i had made (https://github.com/minetest/minetest/pull/8873) is pretty much targeted at server owners... not sure why it was bashed for not beeing an UI feature :) 02:07 nepugia surely it would be better to have that per world/gamemode, but that isnt a problem with this setting, it just is a problem with settings in generall :) 02:10 rubenwardy nepugia: wrong link? 02:10 rubenwardy That pr is by Krock 02:10 nepugia rubenwardy, no, correct link 02:10 nepugia i don't have a github account :) 02:11 rubenwardy Ohh 02:15 erlehmann does minetest have a mailing list or something? 02:16 paramat no, forum has a news section though 02:19 nepugia Hmm, i once had a pr for crafting groups that went overboard, 02:19 nepugia what about adding something similar to group: that refferences another modification? as in like:default:stone 02:20 nepugia so one can include compatible recipes, and declare ones own crafting components to be compatible with an item from another modification 02:20 paramat bye all, hope to see you here again erlehmann 02:20 erlehmann :) 17:48 Krock will just re-use "Documentation needed" for #9147 17:48 ShadowBot https://github.com/minetest/minetest/issues/9147 -- pointed_thing_to_face_pos behaves incorrectly for nodes with nodebox type other then regular 17:48 Krock makes less sense to add more labels 18:09 Krock https://i.imgur.com/XlZNT7E.png 18:50 erlehmann what do i need to compile minetest with sound? 18:50 pyrollo speakers 18:50 pyrollo and a compilator 18:52 pyrollo Actually, it should compile with sound out of the box. There is a few packages to install before compilation depending on your operating system 18:53 pmp-p like asound/ogg/vorbis/sndio/openal 19:18 sfan5 openal, libogg, libvorbis 19:30 Krock well so, I fixed the attachment stuff, but this breaks mods 19:40 rubenwardy pushing trivial fix at some point: https://github.com/rubenwardy/minetest.github.io/commit/ca98634876274ab3c7398fdc02a800bdd582329f 20:18 Krock rubenwardy: it's time 20:18 Krock also #9150 20:18 ShadowBot https://github.com/minetest/minetest/issues/9150 -- Attachments: Fix interpolation from (0,0,0) after detach by SmallJoker 21:38 Andrey01 hello, I`ve installed Irrlicht and I`ve a problem: I compiled the lib and trying to link all obj files with my obj file using it running `make -I ~/irrlicht-1.8.4/source/Irrlicht all` and it gives a following error: 21:41 Andrey01 it throws out error in CAnimatedMeshMD2.h line 8 #include can not find IAnimatedMeshMD2.h because of incoorect path as if it is located in the current dir, but it is in /include 21:42 Andrey01 and it`s strange why is the path specified incorrectly in that file? 21:45 Andrey01 my files to review (Makefile and the file): https://ufile.io/y135jdza 23:39 erlehmann oh i accidentally muted minetest 23:44 rubenwardy Oh no