Time Nick Message 01:33 erlehmann hey kilbith what's your favourite minetest drama 01:36 kilbith do you feel always forced to open your mouth 01:36 kilbith get to a bar or something 01:37 erlehmann sorry to everyone, i wanted to write this in #minetest and misclicked :( 01:38 erlehmann i agree that #11988 is a good idea. any way to add the version? 01:38 ShadowBot https://github.com/minetest/minetest/issues/11988 -- Add game name to server status string by sfan5 01:39 erlehmann it's a bit difficult to assist with debugging when sometimes the server owners themselves have no idea what version a mod is 01:39 MTDiscord not within 5.5 timeframe 01:48 erlehmann regarding #11866 – the mcl* map generators (which, surprisingly, diverged a bit from each other) all contain code that may have unwelcome results at the map border, if this is not fixed. i have not investigated this too much lately because a fix is in the pipeline, but if anyone feels they need more horror stories to be convinced that it's good, tell me please. 01:49 ShadowBot https://github.com/minetest/minetest/issues/11866 -- Raise max mapgen limit constant to align with MapBlock by sfan5 01:50 erlehmann … that it's good to approve the PR i mean 01:54 erlehmann rubenwardy thx for publishing the security advisories. any idea when/if one or more CVEs will be assigned? 07:20 sfan5 erlehmann: the only concept of version numbers MT has is when you downloaded something from ContentDB 07:21 sfan5 so not really useful in most cases 07:24 erlehmann oops you are right 07:24 erlehmann so games need to get version numbers first 07:24 erlehmann in the game.conf i guess 07:25 erlehmann the thing why i would find it important is because players sometimes hold off playing until bugs are fixed 07:25 erlehmann and after the bugfix there is the “admin does not respond, do we have to trigger the bug to find out?” period 07:51 Baytuch sfan5, i connect my game server to servers.minetest.net over https and work fine now. Thank you so much. 09:14 Baytuch Guys, how limiting allocate mem in the client? so far, is const. Can i set less ram for client, like ~100-150 mb. I have 1 Gb ram on my machine (PI 3) and have some problem with this. 09:16 Baytuch Client allocate mem, when i move on the map 09:16 Baytuch can set limite for this? 09:34 erlehmann Baytuch, loaded mapblocks kept in memory can be set IIRC 09:34 erlehmann i have no idea what has changed about dynamic media, but it seems that creating maps is now racy, whereas it was not in 5.4.1 and following 09:35 erlehmann newest minetest + newest mineclone5 = some textures do not get sent to the client 09:35 MTDiscord racy? like, race conditions? 09:35 erlehmann yes i guess so 09:35 erlehmann making the first map usually works. subsequent ones are hit or miss. 09:36 erlehmann i.e. the texture is generated on the server, but the client does not receive it somehow. 09:36 erlehmann at least it does not display it and shows an error message 09:36 erlehmann once you reconnect, you see all the old maps, but its still racy 09:42 erlehmann also minetest now seems to use a bit more memory and processing power for doing nothing in 5.5-dev, but i have not investigated that in-depth 09:42 erlehmann the maps issue is extremely impractical though 09:46 erlehmann Sublayer plank, do reproduce: when you download mineclone5 from cdb with newest minetest, get a map from creative inv and right click while holding it. 09:46 erlehmann do that repeatedly 09:46 erlehmann if you have the bug, sometimes, the texture will not load 09:47 erlehmann and by sometimes, i mean: most of the time, for me. whereas it loaded fine in 5.4.1, unless the texture was actually not present on the server. 09:47 erlehmann it is generated in 5.5-dev though! 09:55 MTDiscord huh I tried generating a map in mcl5 with minetest 5.5.0-rc1 and the TGA files it generates aren't even valid 09:56 erlehmann are you sure? 09:57 erlehmann can you post such a file? 09:58 erlehmann they should be in the world directory in mcl_maps folder or something 09:58 erlehmann https://mister-muffin.de/paste is a pastebin you can use for arbitrary stuff 09:58 MTDiscord do discord attachment links get sent over the irc bridge? 09:58 MTDiscord https://cdn.discordapp.com/attachments/747163566800633906/936560967825903616/mcl_maps_map_texture_0.tga 09:59 erlehmann i see it 09:59 MTDiscord it does not load neither in minetest or in gimp 09:59 erlehmann let me check 09:59 erlehmann thank you 09:59 MTDiscord in-game it gives a 'Client: Cannot create image from data of file "mcl_maps_map_texture_0.tga"' error 09:59 erlehmann yes, i get the same 10:00 erlehmann but when i shutdown and start again, i do not get it 10:00 MTDiscord huh 10:00 erlehmann i would be happy if this was somehow mcl5's fault, encoder changes this that 10:00 erlehmann uh no idea what that file is o.0 10:01 erlehmann Subplayer plank, does the same thing with minetest 5.4.1 generate valid files or are they similarly broken? 10:02 MTDiscord I don't know, I can check 10:02 MTDiscord I generated like 22 maps and all of them are identical 10:09 erlehmann please check with minetest 5.4.1 then. i just rechecked, using mcl5 from contentdb and i was not able to get any corrupted files 10:09 erlehmann which system are you on? 10:09 erlehmann i am on debian gnu/linux 10:09 MTDiscord ok wtf, mcl5 generates these broken TGA files even on 5.4.1 10:10 MTDiscord I'm on windows right now 10:10 erlehmann uh oh 10:10 erlehmann which mineclone5, the cdb version? 10:10 MTDiscord TGA generated on 5.4.1: 10:10 MTDiscord https://cdn.discordapp.com/attachments/747163566800633906/936563839745527818/mcl_maps_map_texture_0.tga 10:11 MTDiscord equally broken, plus now it gives another error in addition to the cannot create image error: 'Irrlicht: Unsupported TGA file type: _tempreadfile' 10:11 MTDiscord off to eat lunch now but I'll be back in a while 10:11 erlehmann this is super weird 10:12 erlehmann i look at the file 10:17 erlehmann hahaha 10:17 erlehmann this is exactly the thing where i need my devtest tga textures for 10:19 erlehmann Subplayer plank i would be interested if you can load all of these https://github.com/minetest/minetest/pull/11978 10:19 erlehmann (there are 10) 10:20 erlehmann to quote myself 10:20 erlehmann > it is probably going to be useful if a developer can ask ordinary users to start devtest and make a screenshot of some texture test nodes. 10:28 erlehmann OH LOL 10:30 erlehmann ; file mcl_maps_map_texture_0.tga; LANG=C dos2unix -f mcl_maps_map_texture_0.tga; file mcl_maps_map_texture_0.tga 10:30 erlehmann mcl_maps_map_texture_0.tga: data 10:30 erlehmann dos2unix: converting file mcl_maps_map_texture_0.tga to Unix format... 10:30 erlehmann mcl_maps_map_texture_0.tga: Targa image data - RGB - RLE 128 x 128 x 24 10:32 MTDiscord @77 73 6F 72 is wrong 10:32 erlehmann wdym 10:32 MTDiscord #11988 can very well get a version string, the question is just whether it'll be what we want 10:32 ShadowBot https://github.com/minetest/minetest/issues/11988 -- Add game name to server status string by sfan5 10:32 erlehmann @77 73 6F 72 is wrong 10:32 MTDiscord CDB stores a "version" field in the game.conf I believe 10:32 erlehmann what is @77 73 6F 72 ? 10:33 MTDiscord it's wsor's/Jonathon's new edgy nick 10:33 erlehmann ah ok, well 10:33 MTDiscord perhaps the hex codes for wsor? 10:33 erlehmann luatic can you see if the \n to \r\n mangling also happens for modlib output? 10:34 MTDiscord hmm wdym? 10:34 erlehmann Sublayer plank generated some TGAs with tga_encoder from mineclone5 on windows and they came out garbage, but i applied dos2unix to it 10:35 erlehmann i know why this has not been a problem before, because the signature did not contain \n before i improved tga_encoder to use RLE 10:35 sfan5 why the fuck are you talking about mineclone in here 10:35 sfan5 go away 10:35 MTDiscord please don't tell me TGA encoder doesn't use wb 10:35 erlehmann sfan5, because minetest on windows mangles written files 10:35 MTDiscord you probably forget to set the binary flag when writing 10:35 sfan5 come here when you have debugged the bug 10:35 erlehmann luatic may just have debugged it 10:35 sfan5 but you don't have to do your debugging session here 10:36 MTDiscord yep, https://git.minetest.land/MineClone2/MineClone2/src/branch/master/mods/CORE/tga_encoder/init.lua#L87 is the cause of the bug 10:36 erlehmann ok, then i do them in #minetest ig 10:36 MTDiscord should be "wb" 10:36 erlehmann luatic thank you! 10:36 MTDiscord np, now back to the version 10:36 sfan5 so much for "Minetest mangles files" 10:36 MTDiscord sfan5: couldn't you just read the version field from game.conf? 10:36 sfan5 the version field is just a number isn't it? 10:37 MTDiscord huh 10:37 sfan5 also that field is not present when not installed from CDB so it's of no use on the vast majority of server installs 10:37 erlehmann sfan5, i am very sorry, it is news to me that writing a file on windows silently changes the bytestream. i will ask for help about debugging engine issues in #minetest from now on. 10:37 MTDiscord now I'm confused, doesn't CDB store the version in game.conf anymore 10:37 MTDiscord sfan5: modders could manually set version = something though 10:37 MTDiscord I used to do this 10:37 MTDiscord because I expected it could be of use in the future ;) 10:38 MTDiscord oh my god, has mineclone's tga encoder been completely busted on windows for all this time? 10:38 MTDiscord well, that explains it then, CRLF moment 10:38 sfan5 erlehmann: it becomes an engine issue once debugging has concluded that the source lies in the engine, before that point it's not a topic for here 10:38 erlehmann sfan5, ok 10:38 erlehmann Subplayer plank lets pls move over to #minetest 10:38 MTDiscord yeah 10:42 Baytuch I think, i find it: while (!mapblock_queue.empty() && (mapblock_queue.size() > max_loaded_blocks 10:42 Baytuch where var max_loaded_blocks 10:43 erlehmann Baytuch i believe for you the same is true as for me, move to #minetest if it is tech support and not development 10:44 celeron55 Baytuch: setting client_mapblock_limit 10:44 Baytuch g_settings->getS32("client_mapblock_limit") 10:44 Baytuch client_mapblock_limit = 100 10:44 Baytuch client alloc 170 Mb 10:46 Baytuch I create request for minetest_game this lang an wait... 10:48 Baytuch on this machine MT 5.4.1 work vary slow - i used 5.3.0 10:55 erlehmann Baytuch do /join #minetest and i try to help you there 11:59 MTDiscord @Luatic not everyone ise cdb. And the version field is a reserved key for cdb. 11:59 MTDiscord *uses 11:59 erlehmann celeron55 i hereby want to propose a “minetest-stable” branch that is exactly like minetest is now, only that it never breaks the API in minor releases. i would volunteer for maintaining it. 12:00 erlehmann basically it would be like minetest without all the reasons why i open issues 12:00 erlehmann celeron55 what do you think of that proposal? 12:00 MTDiscord Sounds stupid with stable-5 12:00 erlehmann what is stable-5? 12:00 erlehmann does such a branch already exist? 12:00 MTDiscord Lol 12:01 MTDiscord Maybe go do your research first 12:01 rubenwardy you ask us to deprecate and remove in a previous issue, and then complain when that was done 12:07 MTDiscord what about #10841 is unconfirmed? 12:07 ShadowBot https://github.com/minetest/minetest/issues/10841 -- 3rd Person Bugs 12:09 rubenwardy updated 14:32 MTDiscord ? 15:57 sfan5 ok so what's still open 15:58 sfan5 #11831 which we could leave out but I might have time for an alternate solution 15:58 ShadowBot https://github.com/minetest/minetest/issues/11831 -- Don't go out of the map during raycast by savilli 15:58 sfan5 #11699 is done, just needs approval 15:58 ShadowBot https://github.com/minetest/minetest/issues/11699 -- Disable dynamic shadows for 5.5.0 release by SmallJoker 15:59 MTDiscord has 3d mode pageflip been fixed or its references removed yet? 15:59 sfan5 #11937 has a fix ready I can push if desired 15:59 ShadowBot https://github.com/minetest/minetest/issues/11937 -- Weird Android workaround in MapDatabaseSQLite3 15:59 sfan5 appears to be still broken, so no 16:00 sfan5 #11992 depends on whether anyone can figure out a solution soon, not critical 16:00 ShadowBot https://github.com/minetest/minetest/issues/11992 -- Heap UAF in craft recipes 16:03 sfan5 the credits tab needs to be updated, I'll prepare a PR for that now 16:03 sfan5 also merging #11988 16:03 ShadowBot https://github.com/minetest/minetest/issues/11988 -- Add game name to server status string by sfan5 16:04 sfan5 also https://github.com/minetest/minetest/pull/11976/commits/46366e6ea8c332b5f3b823e83a23876446be5ecf for good measure 16:04 sfan5 in 10m 16:12 Pexin_ #11969 is trivial, wont break anything, and I really want it for partial future-proofing.. 16:12 ShadowBot https://github.com/minetest/minetest/issues/11969 -- Use absolute value for bouncy in collision by pecksin 16:13 erlehmann https://github.com/minetest/minetest/pull/11934 is documentation purely 16:14 sfan5 Pexin_: ping some coredevs and ask them for their opinion 16:15 erlehmann if this short doc fix is not approved i will look forward to when facedir is changed to right-handed rotation and then someone says, but it was undefined, we still rotate around the same axis! https://github.com/minetest/minetest/pull/11934/files 16:18 MTDiscord I'd like to see #11955 be merged but it requires changes before it can be merged and I do not know how to solve them myself 16:18 ShadowBot https://github.com/minetest/minetest/issues/11955 -- Update and improve warnings when no games are found or when only devtest is installed by rollerozxa 16:20 Pexin_ Krock rubenwardy v-rob nrz celeron55: #11969 opinions? 16:20 ShadowBot https://github.com/minetest/minetest/issues/11969 -- Use absolute value for bouncy in collision by pecksin 16:20 sfan5 as for mtg we have game#2920 and game#2921 which could use a cursory look @luatic 16:20 ShadowBot https://github.com/minetest/minetest_game/issues/2920 -- Fix and update Swedish translation. by rollerozxa 16:20 ShadowBot https://github.com/minetest/minetest_game/issues/2921 -- Add ukrainian (uk) translation by baytuch 16:22 erlehmann i want to once again highlight that aligning the world end with map blocks is highly beneficial https://github.com/minetest/minetest/pull/11866 16:22 sfan5 I wonder what happened to the idea of not shipping devtest in release builds 16:23 sfan5 ¯\_(ツ)_/¯ 16:37 sfan5 [0]: should I move you back to the core developer list in credits? 17:08 erlehmann minetest is on the radar of the debian security-team now, thx rubenwardy! https://salsa.debian.org/security-tracker-team/security-tracker/-/commit/3db1fe1ab2c140906022a463cf18046ebbdd8aca#922f24d99bcbb71f467600d4a765aa2ac5ee31f5_31_31 17:25 sfan5 someone stole the round number 17:26 sfan5 HuguesRoss, @Warr1024: please check https://github.com/minetest/minetest/pull/12001/files to see if you are credited the way you like 17:27 sfan5 ~tell x2048 please check https://github.com/minetest/minetest/pull/12001/files to see if you are credited the way you like 17:27 ShadowBot sfan5: OK. 17:27 HuguesRoss seems fine to me 17:27 sfan5 thanks 17:28 MTDiscord lgtm 17:29 erlehmann what is this list for? 17:30 erlehmann oh, ingam 17:43 MTDiscord is 5.5.0 going to be able to be released in f-droid? I remember there has been some changes to the android build process but I don't remember if it now makes it possible to build minetest entirely from source including libraries 17:43 rubenwardy that's fixed now, requires someone to update F-Droid scripts 18:05 MTDiscord @x2048 18:05 Krock I think he was coredev for a short period? 20:19 MTDiscord I think he is a coredev? 23:58 MTDiscord He is 23:58 MTDiscord Warr1024 is the was 23:58 MTDiscord who is no longer