Time Nick Message 03:59 frostsnow Anyone know how to debug: "AsyncErr: Lua: Invalid vector (expected table got nil).". It's almost certainly from an old mod. Worked on 5.3.0. No stack trace or line number is given, though. 04:10 Blockhead256 frostsnow: how big is the mod set you are currently working with/do you know what mod causes the error? 04:11 frostsnow Blockhead256: Not sure how to quantify, but I know which mod causes the error and how to trigger it consistently. 04:11 Blockhead256 Cool, so are you having a hard time tracking down where exactly in the mod it comes from due to no stack trace? 04:12 frostsnow Yup! 04:12 Blockhead256 would you mind linking me to the mod if it's publically available? 04:12 Blockhead256 In this case I would also be wary since sometimes it's the combination of mods that can cause problems. Too early for me to tell 04:12 frostsnow https://github.com/clinew/pyramids 04:13 frostsnow Doubtful in this case. Find a pyramid, let the mummy fall into lava. Shortly after the game will crash. Worse, reloading the world will cause the game to crash again. 04:14 Blockhead256 well hopefully the repeated crashing lets one rapidly iterate lol 04:14 frostsnow I'm hoping there's a trick to getting a stack trace. 04:15 Blockhead256 Maybe the debug mod https://github.com/appgurueu/dbg will help since I can't quite recall but I think it catches some stack traces 04:15 Blockhead256 I mean crash conditions 04:21 Blockhead256 does anyone know why my "R" key isn't toggling unlimited render distance any more on 5.7.0-dev-eee347a37?? 04:25 Blockhead256 frostsnow: I'm having a hard time finding a pyramid, do you have a seed and mapgen I can use? 04:26 Blockhead256 spoke too soon.. 04:26 frostsnow Blockhead256: Not sure if pyramid-generation is deterministic. Will -- nevermind 04:26 frostsnow Trying to remember where to set secure.trusted_mods >.> 04:26 Blockhead256 seed "pyramid3", flew a bit west 04:26 Blockhead256 I would just search "secure" in the settings menu 04:28 Blockhead256 I got a stack trace, mummy.lua:188 mummy.lua:294 04:29 frostsnow Huh. Not sure why I didn't. 04:29 Blockhead256 looking at the code, my first guess is the object is used after it's already supposed to have been deleted. I've seen that type before 04:30 Blockhead256 lines 205-208: the function doesn't return after the mummy's object is removed, so self.object becomes nil and causes the crash 04:31 Blockhead256 "it's a wonder it ever worked in the first place" as they say 04:32 frostsnow "Undefined behavior" is always fun 04:32 frostsnow Thanks for looking at it. Now I'm curious why I'm not getting a stack trace... 04:33 Blockhead256 hmm I added a return there but still got a crash - this may not fix existing crashing worlds... 04:33 frostsnow Good thing I'm running it in a test world :) 04:38 Blockhead256 Frostsnow: wait a minute, let me back up for a bit. This is a really old version of the Pyramids mod. Have you tried Wuzzy's version? https://forum.minetest.net/viewtopic.php?f=11&t=10336 04:40 frostsnow Blockhead256: Nope, though I saw it while searching for the issue. I'll have to see if it looks good and is backwards-compatible. 04:41 Blockhead256 oh dear, that causes a bunch of unknown nodes when transitioning forward 04:41 frostsnow Yikes 04:45 Blockhead256 ah that was caused by the mod being renamed pyramids to tsm_pyramids 04:45 Blockhead256 aliases seems to have been left for the old nodes, however the mummy and mummy spawner entities that existed in the old version have not transitioned 04:46 Blockhead256 on the other hand, a new mummy (tsm_pyramids:mummy) has spawned within an old pyramid for me 04:48 Blockhead256 frostsnow: so I guess for a perfect transition some custom code still needs to be written to turn old entities into new ones, but that would be some work. New pyramids will work fine. 04:50 Blockhead256 the other change is the addition of sandstone pyramids.. anyway I'm sure you're competent to investigate these things 04:52 frostsnow Blockhead256: More or less. I'll probably call it a night before too long, but will note to look into tsm_pyramids as a path forward. Thanks again for looking into it! 04:52 Blockhead256 not a problem. Old mods are a bit of a worry sometimes, but glad I could find you something that looks promising compared to fixing every old bug :) 04:53 Blockhead256 You may want to post onto the forum thread of tsm_pyramids as I linked and ask Wuzzy if he will do something about this as well 04:53 frostsnow It was only difficult because of the lack of a stack trace! Interestingly enough, re-emerging (Gentoo-ism) with USE="-lua_single_target_lua5_1 lua_single_target_luajit" gives me a stack trace. 04:54 Blockhead256 ah. I always recommend LuaJIT for Minetes yes, though it works with PUC Lua 04:55 frostsnow Lua5.1 is the default, which is why I was using it. Apparently LuaJIT is the way to go, though! 04:56 Blockhead256 LuaJIT will improve performance. I'm fairly sure Debian's Minetest packages use LuaJIT. Perhaps less care was taken for Gentoo 04:57 muurkha I didn't realize it was possible to use LuaJIT for Minetest! 04:57 muurkha I like LuaJIT plenty 04:57 Blockhead256 I've only used Gentoo in passing before, I'm not sure how they choose their default emerge flags 04:58 Blockhead256 Minetest goes to a lot of effort to make sure they support both PUC Lua and LuaJIT 04:58 muurkha what are the hard parts of that? 05:00 Blockhead256 LuaJIT doesn't support ARM I think. I think I can pull up some issues on GitHub that show PUC Lua vs LuaJIT discussion hang on.. 05:01 Blockhead256 this is in the midst of a larger discussion but here it is: https://github.com/minetest/minetest/issues/12778#issuecomment-1250255332 05:02 Blockhead256 plus you can just search the issue tracker for "puc lua" or "luajit" for other things 05:41 muurkha thanks! 11:10 sfan5 Blockhead256: the infinite distance key is no longer bound by default in dev versions 14:16 FavoritoHJS Is there a way to limit how far I can see player names on the client? 15:45 sfan5 client? no server? yes 17:44 FavoritoHJS that's... bad 17:45 FavoritoHJS for some godforsaken reason most servers seem to have a player view distance of yes, so some directions are basically completely occluded with player names from players too far away to ever reach 17:45 MinetestBot 02[git] 04x2048 -> 03minetest/minetest: Use multiple threads for mesh generation (#13062) 1389e7f72 https://github.com/minetest/minetest/commit/89e7f72c929fbeef8ad755bc85db22ae6102787d (152022-12-27T17:44:18Z) 17:48 FavoritoHJS :eyes: does that mean less lag due to block updates? 20:07 nyje can anyone point me to working instructions to build minetest git sourcecode on debian? using the instructions in readme.txt it builds IrrlichtMT then at 71% it fails with make[1]: *** [src/CMakeFiles/minetest.dir/all] Error 2 20:07 nyje Makefile:149: recipe for target 'all' failed 20:07 nyje make: *** [all] Error 2 20:08 Krock please provide the full log on pastebin.com or alikes 20:12 nyje https://pastebin.com/k3xg24We 20:13 nyje is that ok? or where would I find the full log? 20:22 nyje that's the full console output from following the compile instructions in readme.txt linked from the minetest.net download page... 20:30 Krock install libzstd-dev 20:31 Krock I'm surprised that CMake did not complain about it 20:32 Krock hold on. it might be a version mismatch 20:34 Krock the version after zstd v1.5.2 should have the correct header 20:34 Krock also https://github.com/facebook/zstd/pull/1174 20:36 nyje ok, thanks for your time, I'll try 20:36 Krock sorry, please ignore my comment about "v1.5.2". it seems that the commits always show the newest tag instead of the previous/next tag 20:37 Krock the change was merged in June 2018, so any version staring from v1.3.6 should include that change. you might want to enable backports for that package or compile zstd yourself 20:58 nyje thank you Krock I built it and make installed it, and the build completed just fine... 21:00 nyje thank you for your time and expertize, I could not have solved it by myself... 21:01 nyje I now just have to deal with the "FrankenDebian" fallout 21:02 Krock np. you're welcome 21:02 nyje but at least I have MT ;-) thanks again... 21:11 nyje happy holidays Krock and I wish you and your family a happy new year with health, wealth and anything else you desire.... 21:11 Krock yes sure thanks. happy new year's to you as well (when time comes)