Time Nick Message 01:38 garywhite Be back on Monday 07:36 ANAND Hello, I wanted to know if there's a way to modify the server status that's displayed to a player on join. 07:36 ANAND Would over-riding the /status command suffice? 08:09 Calinou I think it's defined in builtin/, no? 08:09 Calinou the message displayed on join is the "motd" property of minetest.conf 08:09 Calinou I think core handles it, not sure… 08:10 ANAND No not the Motd, I want to add some extra details to the status message 08:11 ANAND What I'm asking is whether over-riding /status command (builtin/game/chatcommands.lua) will help me achieve that 08:12 Calinou I guess… look at the implementation in chatcommands.lua 08:13 ANAND I'll make a few changes and I'll see if it reflects... I'll be back in a minute or two 08:16 ANAND oh.. /status just invokes core.get_server_status() 08:29 ANAND I can't go modifying the engine just to add a little piece of info specific to a server. Thanks for the help though, Calinou! 10:52 * rdococ yays for quantum computing 11:12 IhrFussel !tell paramat Couldn't we just check in MTG fire mod whether or not the "to replace" node is air or not and then decide between set_node() swap_node()? I just demonstrated last night that swap_node() is THREE times faster vs set_node() when replacing 50 nodes *without* any callbacks/meta 11:12 MinetestBot IhrFussel: I'll pass that on when paramat is around 11:24 Shara Fussel: why not just open an issue so everyone gets to see and comment? 11:26 IhrFussel Shara, I only suggest something on github if I'm 100% certain that it makes sense... I have not much clue about the fire mod so I wanted to ask him on here first before suggesting something stupid 11:28 Shara If it's stupid, it's equally stupid here 11:29 IhrFussel On github it's visible to a broader audience 11:30 IhrFussel I just looked through tenplus1's PR changes and I can only facepalm when I read "but this air node might contain metadata" comment 11:30 IhrFussel Not once did I ever witness an AIR node with metadata 11:31 Shara He actually uses airlike nodes that have metadata for something 11:31 Shara Just because you don't do something, doesn't mean no one will 11:32 IhrFussel If there is a big fire going on, the server might well lag for 10+ ms (if it has to replace 50+ nodes), maybe it should be an option 11:33 IhrFussel "fire_enable_fastreplacing" or something like that 11:35 IhrFussel I honestly didn't know that set_node() is actually THAT slow even when the node has no custom callbacks/meta ... it's terrible 11:36 Shara If it's as bad as you suggest, maybe it would make more sense to try and check why it's that bad 11:37 Shara (assuming you didn't yet) 11:38 IhrFussel Shara, only did a benchmark and 50 air -> dirt replacements took 9ms with set_node() and 3ms with swap_node() 11:39 IhrFussel Maybe an alternative could be to add an optional "keep meta" to swap_node() ... like minetest.swap_node(pos, node, true) 11:39 IhrFussel Oops I mean delete meta* 11:41 IhrFussel Not sure how much faster it then still would be over set_node() 11:57 xerox123_ hello everyone o/ 12:19 IcyDiamond o/ 12:34 Krock o/ 12:59 * rdococ yays, and turns into corium 13:06 rdococ Unicornizing 13:51 rdococ hm, Snap! is kind of weird, I can't tell if the continuations it has are delimited or not 15:04 MinetestBot paramat: May-12 11:12 UTC Couldn't we just check in MTG fire mod whether or not the "to replace" node is air or not and then decide between set_node() swap_node()? I just demonstrated last night that swap_node() is THREE times faster vs set_node() when replacing 50 nodes *without* any callbacks/meta 15:04 paramat hmm 15:07 IhrFussel Earlier today I also suggested as an alternative to add a "delete meta" flag to swap_node() if that would make it better 15:08 paramat no that's messy, checking a node to see what it is to decide between set and swap, if done everywhere that would be code bloat. also because you now have an additional node check it will probably remove any performance gain 15:09 IhrFussel The node is already known it's a simple node.name check AFAIK...it will take max a few µs 15:09 paramat fire nodes aren't replaced 50 at a time and you didn't add in 50 extra node checks 15:10 IhrFussel Are you sure that a big fire on a server cannot run set_node() lots of times in a row? 15:12 paramat it's a mistake to try too hard to use swap, and some modders are using it where it 'likely' to be ok, which is bad 15:12 IhrFussel It doesn't affect my server personally since I disabled fire spreading and igniting altogether...but a x3 performance boost sounds like a lot for servers with fire enabled and probably aggressive players who ignite everything they see 15:13 IhrFussel Then what about my alternative solution? Add optional minetest.swap_node(pos, node, true) [true = delete meta] 15:14 paramat we could possibly have another node setter because there are 2 options: meta removal, and running callbacks 15:14 paramat ah if it's already known, maybe .. 15:15 IhrFussel Or add 2 flags to swap_node() ... one for meta another for callbacks 15:16 paramat yeah 15:16 paramat well 15:17 IhrFussel But settings both flags at once will make set_node() obsolete 15:19 IhrFussel Maybe minetest.swap_node(pos, node, {meta=true}) and minetest.swap_node(pos, node, {callbacks=true}) 15:20 paramat for performance it's best to keep 'set node' without flags, to be as simple as possible (less arguments to process), duplication is ok here 15:20 paramat swap could become a version with options 15:21 IhrFussel I wonder how much faster swap_node() is when meta gets deleted 15:21 IhrFussel Or callbacks are run 15:21 ANAND sorry to interrupt, quick question: Is it possible to set ratios for the //mix command in WE? 15:22 IhrFussel ANAND, hack: type the ones atht should be more common multiple times 15:22 paramat it depends on whether separate control of meta and callbacks is useful or not, maybe 15:23 paramat no problem, there is no 'interrupt' on IRC :) 15:23 ANAND :D 15:23 IhrFussel Clearing meta without running callbacks is exactly what we are looking for in the fire mod...or not? 15:24 ANAND IhrFussel, I'd ended up passing default:stone 137 times along with one of each ore 15:24 ANAND ... in order to get the natural ratio 15:25 IhrFussel ANAND, the mix command accepts that many parameters? 15:25 ANAND yes 15:26 ANAND I did not use exactly 137, but I did copy-paste default:stone way too many times 15:26 sfan5 ANAND: there isn't 15:26 ANAND :( 15:27 paramat Ihr yeah perhaps. anyway it's worth an issue to suggest 2 arguments for 'swap node': meta removal, run callbacks. we need more input on how useful it is 15:27 sfan5 you're welcome to make a pull request ;) 15:27 ANAND sfan5 I was actually thinking of that... 15:27 ANAND I'll see if it can be done 15:28 ANAND Thank you though! 15:28 sfan5 it can, you just need to touch a little more code 15:28 sfan5 since worldedit.set is not laid out for this at all 15:29 ANAND Cool! I'll check out the code right away... 15:29 Krock minetest even ships a function to generator ores 15:29 Krock *generate 15:29 ANAND Krock, c++ or lua? 15:30 sfan5 though, if you just want stone turned into ores according to some fixed distribution, you might want to send a PR to https://github.com/sfan5/we_env/ instead 15:30 Krock ANAND, both. in Lua it's like minetest.generate_ores(vm) and done 15:30 ANAND I see 15:30 sfan5 Krock: is that guaranteed to only touch stone nodes? 15:31 Krock sfan5, it touches whatever the ores are defined to touch 15:31 sfan5 hm 15:31 sfan5 I should have expected this 15:31 ANAND Is it documented in lua_api.txt? 15:31 Krock ANAND, Ctrl+F is your friend. Yes. 15:31 ANAND haha ok 15:32 Krock however, this ore generation will rely on the Y position in the world to generate the stuff, so you'll end up manually creating terrain like the map generator 15:32 Krock might not be the desired output if you want to have diamond clusters at Y=10 15:37 sfan5 hm that function does gravel, dirt and stuff too 15:39 Krock perfect underground life :) 16:52 xerox123 how does this look for a CTF floating island? http://0x0.st/sjDQ.png 16:54 Krock visually looks very good but it won't help much to cover from snipers 17:00 Jordach snipin's a good job mate 17:01 xerox123 can't just spoonfeed the players their defences 17:01 xerox123 :P 17:03 Jordach it's challengin' work, outta doors, i'll guarantee you 17:03 Jordach 'll not go hungry 17:04 sfan5 Krock: generate_ores is broken 17:05 rdococ concept: technology themed CTF map. players, rather than having resources like cobble and wood to build their base and protect their flag, would construct highly advanced security mechanisms to ensure that the other team doesn't get into their base and/or out again 17:06 Krock sfan5, did you call vm:write_to_map afterwards() 17:06 sfan5 yes and that's not the issue 17:06 sfan5 get_emerged_area says min {x = -256, y = -112, z = -128} max {x = -225, y = -81, z = -81} 17:06 sfan5 but in the end placeAllOres() is called with nmin = (-241,-97,-112) nmax = (-240,-96,-97) 17:07 sfan5 where does it even get these nonsense values? 17:07 Krock 5 nodes in each direction.. hmm 17:07 Krock I think it does that because it was designed to work for on_generate only 17:12 sfan5 also it's not a reliable way to generating ores either 17:12 sfan5 if the manip you're using happens to belong to a desert biome, it will do nothing at all 17:31 sfan5 Krock: https://github.com/sfan5/we_env/commit/f73cbc1de72b1f76cde3b06fc8e1ed8e4d255580#diff-d7d74284e2e39f6c06d199c8e2bedbd0R121 17:31 sfan5 most of this method are just terrible workarounds for generate_ores being unflexible and sorta broken 17:31 sfan5 still better than nothing I guess 17:32 sfan5 ^ ANAND I've implemented your request in we_env 17:33 Krock interesting solution. running the generating function and then undo all unwanted changes 17:34 rubenwardy Krock: why don't some entries in modList.php have `author`? 17:34 rubenwardy eg: my crafting mod 17:34 Krock looking 17:35 Krock "Unlockable Crafting [crafting] - No craft grid"? it has an author 17:37 rdococ Neologism used to be a neologism 17:41 Krock rubenwardy, if there's an author missing, then it'll be missing for all of their mods. Authors that couldn't be found are called "ERROR_500" just to see that something failed 17:42 rubenwardy nvm, firefox is just stupid 17:43 Krock yes, it won't be shown if you search for a term. all the other keys will be sorted out :3 17:44 rubenwardy yay 17:45 rubenwardy the auto meta importer for the contentdb now finds the forum topic ID for you 17:45 Krock o/ 17:50 xerox123 rubenwardy: working on a floating islands map 17:51 xerox123 http://0x0.st/sjkB.png 17:51 xerox123 the islands so far 17:52 ANAND Nice! 17:52 ANAND Are those cobbles placed on purpose? 17:52 ANAND on the sides 17:53 xerox123 yes 17:54 ANAND ok 17:54 ANAND and the buildings house the team bases? 17:55 xerox123 I don't want to give the bases too much initial protection 17:56 xerox123 I think a house is a bit generous 17:56 ANAND IMHO, the players should decide how they want their base to be built 17:58 xerox123 you can connect on the same address and see them 17:59 Shara xerox123... what happens if players fall? 17:59 xerox123 if I'm feeling nice I may put water at the bottom 17:59 Shara then how do they get up? 18:00 Shara It's quite poor gameplayer is you either fall endlessly, or land somewhere that prevents a way of quickly rejoining the action 18:00 rdococ you could have an invisible death barrier at the bottom of maps with problems like that 18:01 rdococ e.g. not for the current maps where you can reach the bottom of the map without getting trapped, but only for maps like xerox123's floating islands map 18:01 Shara rdococ: which makes it sound less suitable for CTF 18:01 Shara But just my opinion anyway 18:01 rdococ Shara: why, and what else would you do? 18:01 Shara I'd not hav e amap like this 18:01 Shara Doesn't feel well suited 18:02 ANAND Lava could be added 18:02 ANAND Won't look good 18:03 Shara Yea, already thought of and dimisseed that 18:03 ANAND But will kill the player, forcing them to re-spawn at base 18:03 Shara And if players can survive the drop, you'll just end up with them making lovely stacks of dirt to climb back up wherever they like 18:04 xerox123 water streams from the islands? 18:04 Shara Players could remove them 18:08 ANAND If the lava river's deep enough, players can't point at anything to place the dirt on 18:10 rdococ Again, invisible death barrier. unless you can give a legitimate reason as to why it doesn't feel well suited. 18:11 Shara Because it's not a thing that exists in CTF 18:11 Shara You'd need to get that added before adding maps that depend on it 18:11 ANAND rdococ try convincing rubenwardy first :) 18:24 Fixer make a mooove 18:24 Fixer rdococ: fancy UI 19:30 Tenplus1 hi folks 19:40 ANAND Hey Tenplus1! 19:40 Tenplus1 o/ anand 19:41 Krock o/ Tenplus1 19:41 Tenplus1 I see that apples regrow in mtg now :) nice shara 19:41 Tenplus1 hey krock 19:41 Krock IhrFussel, dump their inventory into the bin 19:42 Shara :) 19:42 Shara Hi Tenplus1 19:42 Shara Fussel: I just tell them that's their own fault then, and don't even bother checking 19:42 Tenplus1 hi shara :) 19:43 Tenplus1 I found a way so that apple_mark node can replace oranges and banana's too :D 19:43 Shara Tenplus1: yea, but not doing it for MTG 19:43 Shara And I don't want to use metadata 19:43 Tenplus1 it's not metadata 19:43 Shara What then? 19:44 Tenplus1 I can dig a banana and replace it with apple_mark with param2 = 2 and override the node to check for param2 and replace with banana :P 19:44 Krock a node for each fruit 19:44 Tenplus1 I have all those rotatiuonal positions to work with :D 19:44 Shara Ahh, sure 19:44 Tenplus1 1 node, many rotations 19:44 Tenplus1 ehehe 19:44 Shara Just seems bettr to register nodes though :P 19:44 Shara But that's a nice way of doing it as well 19:45 Krock so could I create a fruit spawner when using the screwdriver correctly? 19:45 Tenplus1 I dont like wasting nodes if possible, would also like beds mod to use 1 single node per bed instead of 2 19:45 Shara you can target the mark node 19:45 Shara can't* 19:45 Tenplus1 lol krock, you cant point to airlike with screwdriver 19:45 Tenplus1 but yeah, if it were possible it would be fun to do 19:45 Krock no, I meant targetting the fruit before diggin ti 19:46 Krock *digging it 19:46 Tenplus1 param2 only works with apple_mark (airlike node) 19:46 Shara I assuming the mark is being set based on type of fruit, not its rotation 19:46 Tenplus1 yeah, if I dig apple then place apple_mark with 0 param2, if banana 1 param2, if orange 2 param2 etc 19:47 Shara Problem of course is that if you do it this way for ethereal fruit trees, how is some other mod to klnow what you are using? 19:47 Tenplus1 they can add their own nodes :) eheheheh 19:47 Shara why should they if you don't? 19:47 Shara This isn't reserved for you :P 19:47 Tenplus1 am kidding :) they could... 19:48 Tenplus1 ethereal only has 3 items to regrow (banana, orange, coconut), you got a lot of param2 space left 19:48 Shara I really want regrowth added to ethereal, but please do it in a way that isn't so likely to make conflicts with other mods 19:48 Shara I guess I'm forking ethereal when I do my next big update then :P 19:48 Tenplus1 was gonna add 'regrow' to ethereal 19:50 Tenplus1 1 node to rule them all 19:51 Shara ethereal is going to be annoying to update for me anyway 19:51 Tenplus1 how so ? 19:51 Shara because I need to manually switch settings every time, plus bonemeal change, and I think some other things you did since I last updated 19:52 Tenplus1 I've only removed bonemeal so that servers use the actual bonemeal mod which has a lot more features and api 19:52 Shara features I didn't want 19:52 Tenplus1 what bonemeal ? 19:52 Shara Having the lighter version in ethereal itself was useful 19:52 Shara yes 19:52 Shara I iddn't want the other types 19:53 Tenplus1 it's the same thing, only different versions (mulch, bonemeal, fertilizer) 19:53 Tenplus1 just working a lot better with better mod support 19:53 Shara I still don't want the other versions 19:53 Tenplus1 o.O 19:54 Shara Just because you want something added to a mod, doesn't mean everyone else who uses it does 19:54 Shara and it's your mod, so it's fine, but it means I will have to make my own version on next update, since you changed ethereal as well 19:55 Tenplus1 I dont get your dislike of mulch and fertilizer items 19:55 Shara It wasn't a problem while that still had bonemeal 19:55 Tenplus1 it still HAS bonemeal 19:55 Shara Because I consider it too basic a feature to justify having three different items for it 19:55 Tenplus1 mulch from leaves but only works in 2x2 area, bonemeal from dirt/bones but works on 3x3 area, mulch + bonemeal = fertilizer that works in 5x5 area 19:56 Tenplus1 it's a progression 19:56 Shara You removed bonemeal from ethereal, yes? 19:56 Tenplus1 the aliases are there if you are running bonemeal 19:56 Tenplus1 it still works... 19:56 Shara which is not what I want 19:56 Tenplus1 *sigh* 19:56 Shara I don't want mulch or fertiliser 19:57 Shara Pretty sure I told him that way back when he changed it... but oh well 20:07 rdococ I once modified one of the farming mods to add random mutations that would affect their growth speed 20:08 Shara Hmm, just growth speed, or anything else? 20:08 rdococ I don't think there was anything else 20:08 Shara I guess I'm surprised there's nothign random about it by default 20:08 rdococ Although I think it would be better if it affected other things too 20:09 Shara Growth speed should be variable 20:10 rdococ This would probably be infeasible for a Lua mod, but imagine crop evolution that, after many generations of slight changes to the crop's appearance and functionality, would result in entirely different plant species 20:10 rdococ Or at least visible changes that would be enough to distinguish different varieties of the crop, similarly to how you have different varieties of plants in real life 20:11 Shara Would be pretty intersting 20:11 rdococ I guess you could do it to a limited extent with the node coloring feature 20:12 Shara Or just lots of alternative crops registered, which are different stage versions, and there's a small chance of growing a next stage version 20:12 rdococ That would also work, but depending on variety could result in a lot of nodes being registered 20:12 Shara Of course 20:13 Shara Just that it's not impossible 20:26 rdococ You would use selective breeding to breed only the crops that have the properties you want (higher yield, speed, etc.), and maybe the properties could be somewhat mutually exclusive (for example, a crop can have a high yield and low speed, or a low yield and high speed, but it can't have high both, only medium both). 20:33 Fixer 270 players on 2b2t... 20:33 Fixer wonder if minetest can handle that much players on one server 20:34 Calinou I thought there was a MT server with 270 players online right now :P 20:35 Krock it would be punished by the server list 20:36 Calinou yeah, max_players > 100 is punished IIRC 20:36 Krock 127 is the max 20:46 Fixer lol 20:50 IhrFussel IMO MT isn't that bad of an engine...it's a OSS game that supports ~ 50 players connected at once with its network code without huge lag caused by the engine itself 20:53 IhrFussel I bet most servers wouldn't lag at all anymore if they removed all lua code (except for builtin) 20:53 lisac IhrFussel: they would lag even less if you removed the map loading and handling of player movement 20:54 IhrFussel But who wants to play on a barebones server 20:54 lisac maybe just make a nicely colored chat 20:54 IhrFussel I mean it's not MT's fault that lua is slower than c++ 20:54 paramat "I bet most servers wouldn't lag at all anymore if they removed all lua code" so true :D 20:55 lisac I also bet if minetest only allowed C++ mods, it would be a lot less alive 20:56 paramat problem is many mods are badly coded and overintensive, and users don't know which are intensive. unfortunately it sems to be the popular ones 20:56 paramat *seems 20:56 IhrFussel Back when there were barely any mods for servers and MTG just began development, I could imagine that servers had 20+ players without lag 20:57 paramat also usually too many mods are added to servers, like hundreds 20:58 IhrFussel paramat, I have 180 but most of them aren't worthy of the name "mod" really...they just add 1 or 2 nodes to the game 20:59 IhrFussel The worst are mods that run intensive tasks on each server step when it's not necessary 20:59 Calinou IhrFussel: it does tend to lag with 50 online players in my experience 21:00 Calinou LuaJIT is not that slow, too 21:00 Calinou I mean, some people accuse the LuaJIT developer of being an alien due to how fast it can be :P 21:01 IhrFussel Calinou, I know but luaJIT -> c++ -> luaJIT is still slower than c++ -> c++ -> c++ ... at least I think so 21:02 Calinou it is, yes 21:02 Dumbeldor IhrFussel: it depends, few people can write good C++. So better than beginners doing bad lua than bad C++ :P 21:02 Calinou earlier 0.4 versions felt much snappier though 21:02 Calinou (less mods, less intensive mods, but also less things to do) 21:03 IhrFussel For example I modified mesecons globalstep because AFAIK there is no node/mechanic that requires the 0.1 sec precision...so I changed it to 0.2 sec 21:06 IhrFussel Did ABMs always exist? 21:06 Krock Did Minetest always exist? 21:06 Krock No. 21:06 Calinou did life always exist? 21:06 IhrFussel I mean since MT dev 21:06 Calinou ABMs were already a thing in December 2011 IIRC 21:07 IhrFussel Was 2011 already the version with lua API? 21:08 IhrFussel When I first heard of "ABMs are slow to execute" I assumed they would be in lua xP but they are in the c++ code so I wonder how they can be so slow/intensive 21:10 Fixer since 0.4 in early 2012 iirc 21:11 Calinou the Lua API was added on 2011-11-30 21:11 Calinou 4 days after, I launched the first version of More Blocks <3 21:20 Fixer https://www.reddit.com/r/linux/comments/8iupdz/caution_the_are_malware_snaps_in_ubuntu_snaps/ 21:21 Fixer 2011 was year of vanilla for me 21:21 Fixer also, officially 7 years with minetest 21:51 paramat no need for [off] for on-topic 22:42 ircSparky minetest.override_item is wrecking connected nodes (fences and glass panes are no longer connecting) any way to fix?