Time Nick Message 00:03 GreenDimond but I dont think I am going to get far in the way of fixing digtron 00:18 GreenDimond IhrFussel: Tested on a compiled version, still no change. 00:19 IhrFussel GreenDimond, if possible make a video showcasing it ... and then I will try to to everything exactly as you in the video 00:19 IhrFussel to do* 00:25 GreenDimond IhrFussel: https://nofile.io/f/7mTJhrBRSF5/sharkfail.mp4 00:31 IhrFussel Thanks will do the exact same in a moment 00:40 benrob0329 Is there a way to change the max length of a username? 00:43 GreenDimond IhrFussel: Are you on Linux or Windows? 00:45 IhrFussel Linux (Xubuntu 18.04) 00:46 GreenDimond rofl 00:46 GreenDimond same here 00:50 GreenDimond Identical OS, identical Minetest version, identical mods, different result. 01:02 GreenDimond IhrFussel: Anything? 01:10 IhrFussel Sorry I'm on CTF right now... and the server seems to ignore hits a lot cause of the lag likely... I wonder why nobody tried to optimize the code there...on my server lag with 200 mods is mostly 0.3-0.4 sec max 01:27 ANAND Hehe 01:27 ANAND rubenwardy: ^ :) 01:29 IhrFussel You should enable the mod profiler to see what's wrong...CTF is partly unplayable...like seriously 01:30 IhrFussel It might help to increase the server step a little... in case it#s the default (90-100ms) right now 01:33 IhrFussel Or decrease active_block_range (but that might be a very drastic change) 01:36 rubenwardy it's not due to CPU, IhrFussel 01:36 rubenwardy it's a network issue 01:37 IhrFussel /status shows the internal lag caused by mods..and it is constantly at 0.5-0.6 seconds 01:37 rubenwardy no process or core goes above 40% 01:38 IhrFussel Just enable the mod profiler once and you see what causes so many ms lag 01:38 rubenwardy I have 01:38 rubenwardy and it's sauth 01:38 IhrFussel That makes sense 01:39 rubenwardy it's a lot worse without sauth 01:39 rubenwardy almost a second of no packets every 18 seconds 01:39 IhrFussel shivajiva told me he's working on a solution...possibly changing the DB structure... one table for each letter so the mod needs to look through less records in on_prejoin() 01:39 rubenwardy also, a high thing in status doesn't mean it's mods 01:39 rubenwardy it could be network I/O 01:39 rubenwardy *blocking network I/o 01:39 rubenwardy or just blocking things generally 01:40 IhrFussel sauth does cause the most lag on my server too... 200ms when a player connects is unacceptable 01:42 rubenwardy I wonder what would happen if I switched to built-in and disabled writing 01:42 rubenwardy like, it would all be in memory 01:43 IhrFussel I bet you have much more records in your auth DB...I have 150,000 01:44 rubenwardy 295638 01:45 rubenwardy should probably make a clean up script 01:45 rubenwardy delete all players with < 100 score 01:45 rubenwardy well, CTF doesn't actually save player data at all 01:45 rubenwardy only auth data 01:50 rubenwardy https://gist.github.com/rubenwardy/649c921947c64a60bf123ba4e3eb2bb4 01:50 rubenwardy IhrFussel ^ 01:50 rubenwardy most of it is in s* 01:51 rubenwardy but there's probably room for improvement in a lot of the mods, such as gauges 01:51 rubenwardy don't understand why no_guests takes so long 01:53 IhrFussel sauth and sban are the most intensive ones, cause they cause unexpected lag in prejoin() which happens a lot more than many other callbacks I'd say..basically whenever someone tries to connect 01:55 IhrFussel I guess no_guests also needs auth so it causes the same lag as sauth itself 01:55 rubenwardy oh wait, I do a player_exists check in no_guests 01:55 rubenwardy maybe without that it'd be faster 01:55 rubenwardy heh 01:57 rubenwardy it would be nice if sauth had an in-memory cache of a certain size for the last N online players 01:58 rubenwardy or maybe if it could use LuaJIT's extension for C-sided memory to get around the limit 01:58 rubenwardy then you could have all players in memory with sqlite for serialisaiton 01:58 rubenwardy hey 01:58 rubenwardy *heh 02:00 IhrFussel That's exactly what shivajiva talked to me about in msg yesterday ... he plans to 1. Add a cache that remembers the last X auth entries and 2. wants to possibly change the data structure so that the records are not in 1 huge table but each letter gets its own like table A table B etc 02:03 IhrFussel But then he thought 37 tables might be too much and he thought about merging certain letters 02:03 rubenwardy that's bad practice 02:04 rubenwardy surely sqlite would have a proper index for that? 02:06 GreenDimond IhrFussel: Plz test 02:08 IhrFussel I will now 02:10 IhrFussel GreenDimond, did you try it with another world yet? Or new world? 02:15 rubenwardy !title https://github.com/MT-CTF/capturetheflag/issues/320 02:15 MinetestBot rubenwardy: Performance Issues · Issue #320 · MT-CTF/capturetheflag · GitHub 02:18 GreenDimond IhrFussel: Occurs in fresh world as well 02:18 IhrFussel GreenDimond, https://nofile.io/f/9jqjWbmXC7p/mobs_debug.mp4 02:20 IhrFussel I probably should've increased the recording resolution...I hope it's clear enough 02:20 GreenDimond you are substituting the var inside that output right? 02:20 GreenDimond not just printing "it is false"? 02:21 IhrFussel minetest.chat_send_all("after_activate() runs ... self._megalodon is '"..dump(self._megalodon).."'") 02:21 GreenDimond well then I dont know what to do 02:22 GreenDimond IhrFussel: You are printing that before you set it right? 02:23 IhrFussel The after_activate() runs directly at the start of the callback but self._megalodon tells that it's nil first and the 'set to false' 02:24 GreenDimond Is your print statement at the beginning ofr the after_activate() function 02:24 GreenDimond If not its an invalid print 02:24 IhrFussel Now I changed the false/true random number to 1,2 ... in theory it should be false or true now randomly 02:25 GreenDimond All im saying is print the value right at the start of after_activate 02:26 GreenDimond It should say its nil the first time 02:26 GreenDimond and false/true the rest 02:26 GreenDimond print the value and then check for it/set it 02:27 GreenDimond so the first time the mob loads it should print nil 02:27 GreenDimond the rest it should print a consistent value 02:29 IhrFussel https://nofile.io/f/6JbmHsw3OaH/shark_debug2.mp4 02:30 GreenDimond so yeah 02:30 GreenDimond yours works 02:30 GreenDimond wait 02:31 GreenDimond IhrFussel: You arent using 0.4.17.1 02:31 IhrFussel Let me check multiple activates ina row 02:31 GreenDimond What commit of Minetest are you using 02:33 IhrFussel AFAIK the devs didn't touch entity code between my commit and 0.4.17.1 ... let me check from when mine is 02:39 GreenDimond The commit hash should be on the window title 02:40 GreenDimond IhrFussel: I just need to commit hash 02:41 IhrFussel I will compile 0.4.17.1 02:41 GreenDimond Guess ill test with 0.4.16 then 02:43 GreenDimond No difference 02:45 IhrFussel git checkout 0.4.17.1 is enough right? After that I just execute cmake & make 02:46 GreenDimond you dont need to cmake 02:46 GreenDimond er 02:46 GreenDimond if you already have the build 02:46 GreenDimond If its a fresh build then yeah 02:46 GreenDimond I was downgrading from an already existing build so I didnt need to cmake 02:47 IhrFussel But the -b flag is not needed in checkout? 02:48 GreenDimond correct 02:48 GreenDimond unless you want it as a new branch 02:51 IhrFussel Compiling takes a bit on my local PC... 65% now 02:58 rubenwardy https://i.rubenwardy.com/TT4lg.png 02:58 rubenwardy note that 100 score can be gained in a single kill 03:00 GreenDimond how much of that is bots before you added bot handling 03:04 IhrFussel Why do you still call them 'bots'? They are regular people like me and you...they just happen to use an app to connect to servers and that app gives them (by default) a randomly generated username and randomly generated password... 99% of them don't bother/don't know to change that name and just choose a server 03:05 IhrFussel Their app tends to give them a NEW generated name once in a while so that explains why one IP sometimes has 30+ different names 03:06 GreenDimond IhrFussel: What? I mean bots. 03:06 IhrFussel They are likely confused as heck cause they don't know that they actually play on a 'Minetest' server...since their app is called 'Exploration 3D' or something like that 03:06 GreenDimond Im gonna stop you right there 03:06 GreenDimond I mean literal bots 03:07 IhrFussel There are barely any real bots on servers 03:07 GreenDimond Oh really 03:07 GreenDimond I beg to differ :) 03:07 GreenDimond A player that sits at the exact location they spawned in and punches air for 10 minutes is a bot 03:07 IhrFussel I have my server since almost 3 years and am online up to 10 hours per day... I should know when I have bots 03:08 GreenDimond a player that wanders at random going no further than 10 meters in a bot 03:08 GreenDimond HOMETOWN had bots 03:08 GreenDimond They exist 03:08 GreenDimond And just because your server was lucky to not have them doesnt mean they dont exist 03:08 GreenDimond I know for a fact they exist 03:09 GreenDimond My server has a rules interface that sends a chat message when someone hits disagree or agree 03:09 IhrFussel No those are not bots...bots are the ones that advertize something in the chat...or there were bots that 'float' in one direction ... randomly walking players are no bots 03:09 GreenDimond Roughly 3 times per day a user will log on and press both buttons at the same time 03:09 GreenDimond Thus sending both chat messages at once 03:10 GreenDimond > bots are the ones that advertize in chat 03:10 GreenDimond bs 03:10 GreenDimond Most of the time they dont 03:10 GreenDimond > randomly walking players are no bots 03:10 GreenDimond When they go in literal circles and punch air, they are 03:10 GreenDimond They go no further than 10 blocks from spawn 03:10 GreenDimond There is no exploratio 03:10 GreenDimond +n 03:10 GreenDimond they dont even look at other players, they dont seem to have any goal 03:10 GreenDimond they move around at random and punch nothing 03:11 IhrFussel No they could just be confused ... I've never seen players walking in circles then... mine walk in one direction for a bit at least before they stop and turn around or walk backwards again... I see no bot behavior in that 03:11 GreenDimond Just because you've never seen it doesnt mean it doesnt exist 03:11 GreenDimond So stop using it as a reason 03:11 GreenDimond :) 03:12 IhrFussel Most of the ese randomly generated names are NOT bots... I am right in that 03:12 GreenDimond I dont care about that 03:12 IhrFussel So blocking them means you block most mobile users 03:12 GreenDimond That's fine 03:13 GreenDimond None of that was my point 03:13 GreenDimond My point is bots exist 03:13 GreenDimond I almost found it offensive when you said "Why do you still call them 'bots'?" 03:13 GreenDimond I never did 03:14 GreenDimond You immediately jumped to the conclusion that I call every randomly generated name a bot 03:14 GreenDimond I call bots 'bots' 03:14 GreenDimond When 99% of players that join my server with a randomly generated name press 2 buttons at once, they are bots. 03:21 IhrFussel It's not 99% ... maybe you have a bug in your formspec then 03:21 IhrFussel It still works in 0.4.17.1 ... data is set to true/false 03:21 GreenDimond IhrFussel, do not tell me what is and what isnt. 03:22 GreenDimond 99% of the players that join with a randomly generated name press both buttons at once 03:22 GreenDimond There is no bug 03:22 GreenDimond Because it works fine for other players 03:22 IhrFussel Maybe their app code is broken...did you know that clients are able to HALT map sending for everyone? 03:23 GreenDimond Yes I do know that 03:23 IhrFussel Those are also not bots 03:23 GreenDimond And I refuse to chalk it up to app code 03:23 GreenDimond If their app does that then they wont be able to play the server anyway 03:23 GreenDimond Stop trying to defend them 03:24 GreenDimond You do things your way, ill do things my way. 03:24 GreenDimond Thank you for trying to help me with the mob problem, I have no idea why it works for you and not me. 03:25 IhrFussel Maybe you just spend more time with them to see who they really are...I find it disrespectful to call them 'bots' just because they might use a broken app on iOS which has no official MT version... it's not their fault in this case ... most app users I see seem to know what they do... they can look at you and follow you...they can read signs or just look around ... don't tell me 'bots' could do that 03:25 IhrFussel Maybe you should just* 03:26 IhrFussel You only see if they look at you when you use 'playeranim' 03:28 GreenDimond IhrFussel, stop. You clearly dont get what I am saying and I refuse to entertain this argument. 03:28 GreenDimond iOS users are iOS users and bots are bots. 03:28 GreenDimond And that's all I have to say. 03:30 IhrFussel 99% are not bots dude...this cannot be true in any shape or form...more like 99% are mobile users and 1% bots maybe ... but believe what you want to believe ... I doubt that you spend enough time on your server to know that so many bots connect to it 03:30 GreenDimond lol 03:30 GreenDimond Did I say 99% are bots? 03:30 GreenDimond Were you reading? 03:31 GreenDimond I spend almost every waking hour watching my server 03:31 GreenDimond Maybe I shold repeat myself 03:32 GreenDimond 99% of the players that join with a randomly generated name press both b 03:32 GreenDimond buttons at once* 03:32 IhrFussel You said they are bots just because they somehow manage to press both buttons at once ... that can have many reasons... for example I think some clients don't receive a certain formspec on my server at all after they join 03:32 GreenDimond Did I say 99% of players are bots? 03:33 IhrFussel After you join you are asked to either enable or disable PvP ... and many of them press neither 03:33 GreenDimond If you didnt tell it to not exist when they press escape, then they can press neither. 03:34 IhrFussel You said that if 99% of random generated names press both buttons they are bots 03:34 GreenDimond Yay you read it 03:35 IhrFussel Yes so you meant 99% of those are bots 03:35 GreenDimond Yes good you got it :) 03:35 GreenDimond Every player gets the formspec, that I am certain of. 03:35 GreenDimond Thats easily seen 03:35 GreenDimond The chat message is only sent when they press a button 03:36 GreenDimond So no formspec, no button, no chat message. A chat message is always sent, so they always get the formspec. 03:37 IhrFussel No no you don't know whether or not their APP displays the formspec ... some of those fake apps are badly programmed I tell you 03:37 GreenDimond Ok? 03:37 GreenDimond So they shouldnt be playing 03:37 GreenDimond Works for me 03:38 GreenDimond But yes, it does display. 03:39 GreenDimond If you really think that a player can randomly tap a screen and manage to press 2 very specifically placed buttons at once, I would rethink a few things. 03:39 IhrFussel That's what no_guests is for ... but server that use this mod and are not CTF are largely dead when using it ... cause FACE IT... ~ 90% of MT userbase is mobile players 03:39 GreenDimond When did we start talking about that 03:39 GreenDimond All my users receive the formspec 03:40 GreenDimond And every time a user with a randomly generated name joins, both buttons are pressed at once. 03:40 IhrFussel Again you don't know that...the client is instructed to do what you want but if it's badly programmed it might not do what you want...it might actually even crash 03:40 GreenDimond *sigh* 03:40 GreenDimond You dont get it do you 03:40 GreenDimond If they cant see the formspec, THEY CANT PRESS BUTTONS 03:40 GreenDimond Especially 2 at the same time 03:41 GreenDimond But they do 03:41 GreenDimond which means it CAN 03:41 GreenDimond and they DO 03:42 IhrFussel That could be part of the "convenience" gameplay ... why do you think apps started to give their users random names and passwords? Cause more and more server disallowed empty names and many didn't know what to do when that happens 03:42 GreenDimond "convenience gameplay".... 03:43 GreenDimond being..? 03:43 IhrFussel You should try out such a fake app... it gives you a nice random name, random password like X5GhjIop pre-determined so you only have to click connect 03:43 GreenDimond "press all buttons in your formspec at once!" 03:43 GreenDimond I have, Ihr. 03:43 GreenDimond I have. 03:43 GreenDimond I get how it all works. 03:43 GreenDimond I KNOW this stuff. 03:43 GreenDimond You dont need to tell me. 03:43 GreenDimond and you know what else I know? 03:43 GreenDimond I KNOW that the players with randomly generated names ALWAYS press both at once 03:44 GreenDimond So you mean to tell me they ALL use the SAME app that does the SAME "convenience function" (lmao) to press BOTH buttons at once that they might not even SEE? 03:44 GreenDimond Let me know when pigs fly 03:44 IhrFussel That the bad clients press all buttons at once could be bad code... I don't see that happen on my server btw...players either enable or disable PvP and that displays in the chat...not once did I see both at the same time... so I cannot confirm what you say 03:45 GreenDimond > I cannot confirm what you say 03:45 GreenDimond So? 03:45 GreenDimond Just because you cant confirm it doesnt mean it doesnt happen 03:46 GreenDimond Look at it this way 03:46 GreenDimond I've never seen a mobile client that presses all buttons at once 03:46 GreenDimond Ever 03:46 GreenDimond Which is why its a bot 03:46 GreenDimond because its NOT a mobile client 03:46 IhrFussel It means for now it only happens on your server until someone else confirms the same behavior on theirs 03:46 GreenDimond That what happens? 03:46 GreenDimond Bots? 03:46 GreenDimond lol 03:46 GreenDimond they happened on HOMETOWN too 03:46 GreenDimond and CTF 03:46 GreenDimond and literally every other server 03:47 IhrFussel Then ask yourself what a 'bot' would gain from connecting to random MT servers and pressing all buttons, randomly moving in circles and punching... what is the point?! 03:48 GreenDimond Who cares? 03:48 GreenDimond Take up server resources for one 03:48 GreenDimond Scrape data for two 03:51 IhrFussel I recommend to install 'playeranim' and THEN watch them again while they walk around...watch their heads... if they always look in the same direction or move in certain patterns then it might be a non-human ... but anyone who's able to look at me is no bot 03:51 GreenDimond No :) 03:51 GreenDimond I dont need to do extra work to over confirm what I already know and what you dont know 03:51 GreenDimond And thats fine with me 03:52 GreenDimond You dont need to play my server 03:52 GreenDimond And I dont expect you to 03:53 IhrFussel I just get mad at server owners calling most mobile players 'bots' cause some of those 'bots' built stuff no regular name managed to do ... and Shara agrees there 03:53 GreenDimond Thats a good thing 03:53 GreenDimond mobile players arent bots 03:53 GreenDimond bots are though :D 03:53 GreenDimond Good day :) 04:00 IhrFussel I never claimed that MT has no bots...but calling 99% of the players with random names "bots" is bit too far-fetched 04:48 sofar why would someone spider the PRMS and download a file every x seconds, taking 4 days to download a pitiful amount of data (maybe a few mb, lol) 04:55 sofar I don't even understand how they managed to get a good sha1 list in the first place without a MT client 04:55 sofar bizarre 11:11 Shara Guh at Fussel dropping my name to try and win.. whatever that was. 11:13 Shara I can only go by what I've seen personally, and I've never yet seen player behaviour that matches what GreenDimond describes. I have seen players who run about banging into things, then pause, and who seem hopelessly lost, but the behaviour isn't structured or patterned in any way. Just seems like young kids bashing buttons. 11:14 Shara Left long enough, a lot of them just start standing in spawn changing their skins over and over again, because they can't read the signs but do eventually figure out how to do that. 11:39 Krock o/ 11:50 nerzhul i just released a new android APK based on the stable-0.4 11:50 nerzhul https://github.com/minetest/minetest/releases/download/0.4.17.1/Minetest-0.4.17.3.apk 11:50 nerzhul can users with problems test it ? 11:50 nerzhul i just tested it on my phone at works it works as intendee 15:31 IhrFussel Shara: I didn't use your name to win an argument ... I said that you and me both saw players with random names that build better than most players with "normal" names 15:32 IhrFussel Oh and "GUI scaling is fine" -> ihrfussels-server.tk/apkss.png 15:33 IhrFussel nerzhul ^ 15:33 nerzhul IhrFussel: fine without backporting all fixes ? nice :D 15:34 IhrFussel You ignore the fact that this bugfix release is partly unusable on many devices and you don't care either it seems 15:37 nerzhul i didn't looked at the screenshot 15:38 nerzhul this release fix crashes on many devices 15:38 nerzhul now we can go ahead and backport the other fix 15:38 nerzhul do you remember who don't have scaling problems but app crashing on startup ? 15:38 nerzhul with that at least we ensure this release is valid 15:38 nerzhul (for this problem) 15:39 IhrFussel You should look at the screenshot 15:39 IhrFussel It is HORRIBLE 15:40 IhrFussel Of course fixing a crash is more important... But why not fix as many things as possible in a "bugfix release" ? 15:40 nerzhul time 15:41 nerzhul and push the first fix asap they are available and tested 15:41 IhrFussel Now I have to tell my players "sorry you need to wait for another apk update to adjust your settings, this one only fixes crashes on sone phones" ... This will sound really silly 15:41 nerzhul do you remember when the scaling problem occurs ? 15:41 nerzhul or is this an historical problem ? 15:42 IhrFussel It was always there and was fixed in 5.0.0-dev so I really hoped that it will make it into the very next bugfix release 15:43 nerzhul okay, then users never played on android because it never works 15:43 nerzhul then they don't wait for a regression fix 15:43 nerzhul they wait for gameplay on android 15:43 nerzhul the main problem on play store is crash reports 15:43 nerzhul (on the reviews) 15:44 nerzhul if someone without scaling problem but crash can confirm i can push 15:45 nerzhul we should really have more tiny dev cycles 15:45 nerzhul i don't know if we can apply it to core engine or not 15:46 nerzhul and if other coredev can agree to have a faster dev cycle, faster than 2 years dev cycle 15:49 nerzhul these days i'm very busy at work and today is a special day for me then i cannot do more this week than reviewing because of this special day, sorry IhrFussel. I know you want all, but we should be more agile and push fixes on the rythm we can :( 15:50 IhrFussel Many many just don't use Minetest Android cause the controls SUCK...like literally, cause 0.4.X still uses those stupid arrows for movement 15:50 rubenwardy we don't normally have a 2 years dev cycle 15:50 rubenwardy it's normally 6 months 15:50 nerzhul controls are in 5.0.0 15:50 nerzhul and they are interesting 15:50 nerzhul rubenwardy i know 15:51 nerzhul rubenwardy: can you take time to polish your remaining features blocking release this week or next week ? 15:52 IhrFussel I know that pretty much everything got improved in the 5.0.0 apk ... I'm saying many will still not install/use the newest bugfix apk cause the controls still suck there 15:52 nerzhul controls will not be part of 0.4.17 15:55 IhrFussel I didn't expect it to... I listen to what my players say...and I hear a lot about "i use [enter 3rd-party client here] cause controls are not good in Minetest for android" 15:56 IhrFussel So controls are a very big reason why so few use the official app *right now* 15:56 nerzhul i understand the point, but we are not talking about 5.0.0 now :( 15:57 IhrFussel We talked about why people don't use the official app kinda so I wanted to add that reason 15:58 nerzhul yeah but both releases has different goals :) 15:59 nerzhul i hope we can release the 0.4.17.X on android before 5.0.0, else it will be waste of time 15:59 nerzhul i wait 1 or 2 days and promote the previous 0.4.17.3 release to production if i don't see a major issue on my phone 16:00 nerzhul and next i will try to take time to look to backport the scaling issue, but maybe it will be better for focus on the 5.0.0 release 16:01 IhrFussel The crashes only affected "older" Android versions right? So we need people with old devices... Average lifespan of a smartphone nowadays is 9 months last I heard 16:04 nerzhul android 5 and older if i remember 16:05 nerzhul i hope it's not 9 months, or it's ridiculous, i keep mine 2 years at least :p 16:06 IhrFussel A week ago I had someone on my server with that INVALID_ENUM bug ... He said he cannotread the chat cause it is full of errors ... I told him to install 0.4.16 in the meantime...when I see him I can ask him to install -3 16:07 IhrFussel You also fixed the ENUM chat flood right? 16:08 nerzhul the release was done with a very clear machine in relese mode 16:09 jas_ wow impressive post by xiong on stu's unofficial andrioid thread. 16:10 jas_ other interesting bug reports/feedback 16:23 jas_ hi 18:06 sofar Calinou: do you have appimages of 0.4.13 or 0.4.11? 18:46 Calinou sofar: nope 18:46 Calinou if you need to run them, I guess you should build them from a container if current compilers are too new 18:46 Calinou (or run the Windows binary in WINE :P) 18:47 Calinou I should build a 0.4.17.1 AppImage by the way 19:00 sofar trying to repro something with an old client, lol 19:00 sofar I found a 32bit appimage, but, my distro doesn't like it much XD 19:03 sofar heh, somehow I had the foresight to clone a 0.4.13 tree at some point 19:03 sofar let's see if it still compiles 19:04 sofar amazig 19:04 sofar whodathunk that it still just compiles! 19:06 GreenDimond Shara: I dont know of many servers with a rule formspec on_join that kicks/grants when the deny/accept buttons are pressed respectively and send a chat message to the server _and_ Discord where it is there to read later. So I am sad but not surprised that I am the only one experiencing this :/ 19:07 sofar btw about that 19:07 sofar I just realized that even fsc won't properly protect against that 19:08 sofar it'll be a 50-50 chance that they get through 19:08 GreenDimond Against what? 19:08 sofar essentialy it's a formspec input abuse tactic 19:08 sofar you would have to ban on anyone who clicks the wrong button 19:09 GreenDimond I just kick them 19:09 sofar if they click accept and reject within a second I would ban them 19:09 sofar if they click reject first, just kick 19:09 sofar if they accept, and reject, ban 19:09 GreenDimond Oh thats an interesting idea 19:09 sofar here's what I would do 19:10 sofar when they click, wait 1 second to process the input 19:10 GreenDimond The problem is that both messages are sent and are pretty spammy 19:10 sofar capture all inputs 19:10 sofar after 1 second, if both responses are received, ban 19:10 GreenDimond Good idea 19:10 sofar if only one choice, kick or grant 19:11 GreenDimond I had a funny thought 19:11 sofar just have the formspec handler store the input in a table 19:11 GreenDimond I could add hidden buttons in a few places that only bots would press 19:11 sofar and run a minetest.after after each input to handle all input 19:12 GreenDimond Right 19:13 sofar I would also seriously think about logging client version and IP address 19:13 GreenDimond Already logging IP addresses 19:13 Shara GreenDimond: it's more about the moving in circles and punching air thing. 19:13 GreenDimond Im going to add a logger with a ratio of kicks to bans with randomly generated names 19:13 GreenDimond Shara: Hm? 19:14 Shara Was replying to: 19:14 Shara GreenDimond: Shara: I dont know of many servers with a rule formspec on_join that kicks/grants when the deny/accept buttons are pressed respectively and send a chat message to the server _and_ Discord where it is there to read later. So I am sad but not surprised that I am the only one experiencing this :/ 19:14 GreenDimond Yes I figured that part 19:14 GreenDimond What's more about the moving in circles and punching air thing? 19:14 GreenDimond Most people experience that? Or dont? 19:14 Shara I've seen plenty of random/weird movement and punching from players, but have even spoken with a few of them after watching them, and it's usually that the mpa hasn't loaded well and they have no clue what they're doing 19:15 GreenDimond Yes, but when the player doesnt move far enough for the chunk to not load? 19:15 Shara I've never seen anyone walk in what seems like bot-like movement (apart from you know who that one time when he was doing you-know-what) 19:15 GreenDimond heh 19:16 Shara The chunk can always not load 19:16 GreenDimond When I say "circle" I mean sparatic movement with no clear goal 19:16 Shara Useless connection/device isn't so rare sadly 19:16 GreenDimond usually ending up where theyve already been 19:16 Shara It's not methodical enough to make me think it's automated when I see things like that 19:17 GreenDimond Things dont have to be methodical 19:17 GreenDimond Random values determining direction make things seem natural 19:17 Shara Yet I spoke to enough of them after they finally got their act together to know there are genuine players who baheave just like this 19:17 GreenDimond especially random values within a given range 19:17 Shara behave* 19:17 Shara I do know how you can program somethign to have random values/movement intervals 19:18 Shara But it all seems very unlikely given there's absolutely no reason for it 19:18 Shara And since, as I said, a lot of these players ended up actually playing once they sorted themselves out 19:18 GreenDimond Shara, I dont deny that some could be normal players. Most of the "random-walkers" on HOMETOWN would mosey around for about 10 minutes around spawn doing nothing and then log off. 19:19 GreenDimond And we'd have 10 of those at once 19:19 GreenDimond > absolutely no reason for it 19:19 GreenDimond players take resources 19:19 Shara Deciding there is a bot epedemic of bots that do nothing and serve no purpose seems bordering on paranoia to me to be honest 19:19 Shara Sure,t hey take resources... what does that gain "them" 19:19 GreenDimond Who decided that? 19:19 Shara ? 19:19 GreenDimond Did you decide there is a bot epidemic? Because I didnt. 19:19 GreenDimond I said bots exist 19:20 GreenDimond Does it really matter what they gain? 19:20 Shara Just seems to be where we've ended up before when people start casually calling those kinds of players bots 19:20 GreenDimond Like I said to Ihr, I am not one of those people. 19:21 Shara That's great then. 19:21 Shara But still, people don't generally make bots for no reason. 19:21 Shara It eats someone else's resources too after all. Not just the server they join. 19:21 GreenDimond Not my problem. What is my problem is they do make bots and I dont like it. 19:22 Shara Better to assume players, not bots, until proven otherwise. 19:22 GreenDimond As far as I am concerned, pressing 2 buttons with microsecond precision is proving otherwise. 19:22 Shara You can still kick them if they're a bother or you need to make space for players who actually can play, of course. 19:23 Shara Seems more likely to be some stupidity in the app to me. 19:23 GreenDimond *sigh* 19:23 GreenDimond And thats where I draw the line 19:23 Shara Yes, sigh. 19:23 GreenDimond I refuse to chalk it up to "stupidity in the app" 19:23 GreenDimond If thats the case, they cant play anyway. 19:23 GreenDimond Cant press one button at once? Then you cant play. 19:23 Shara Well, I saw enough of that already that I gave up thinking "stpid apps won't be so stupid" 19:24 Shara I just said kicking them is fine either way. 19:24 GreenDimond When you have one core and one thread running a heavy server you need to keep useless users at just about zero 19:24 Shara I haven't argued with that, have I? 19:25 GreenDimond Indirectly 19:25 Shara No 19:25 GreenDimond > you can still kick them if theyre a bother or you need to make space for players who can actually play 19:25 GreenDimond implying I should wait until those conditions arise 19:25 Shara Not at all. 19:25 GreenDimond Sure looks like it 19:25 * Krock grabs some popcorn 19:25 Shara If your server is busy, you need the space right away, yes? 19:26 GreenDimond Generally 19:26 Shara So the moment you have any marker they can't play, whatever it is, kicking is fine. 19:26 GreenDimond Fine 19:26 Shara Player or bot, doesn't matter. 19:27 * Shara steals Krock's popcorn. 19:27 * Krock runs after Shara 19:27 Shara Nooo, mine now! 19:27 Krock naah but it's mine 19:27 GreenDimond tl;dr: Not all "wanderers" are bots but bots exist, and GreenDimond wants to find away to keep the bots out :) 19:28 Krock you mean thief you 19:28 Shara I know right :( 19:28 * GreenDimond dusts of popcorn machine 19:28 GreenDimond off* 19:28 GreenDimond I cen spel 19:28 Shara I could if I had space to not have my keyboard sideways :D 19:29 Krock time for a larger desk 19:29 Shara Yea, would love one. 19:29 Shara But then I'd need space for that too 19:29 GreenDimond time for a larger workspace 19:30 Shara Or I could just rearrange things, but... time/effort 19:31 Krock you'd be thankful later 19:31 Shara This room is stupidly shaped. I have tried a couple of times and it never ends up well 19:33 Shara What I actually need to do is set up stuff for work in another room, so I don't have my stuff + work stuff all in the same space 19:34 Krock or have it in easily accessible (front) stackable boxes 19:34 Shara I don't think that works for computer equipment I need access to every day... but getting kidn of offtopic anyway. 19:34 Shara kind* 19:34 GreenDimond Can I ban a player and remove them 19:35 GreenDimond or do they need to be in the database to be banned 19:35 Krock built-in banning? 19:35 Shara Depends if whatever you use records the ban info somewhere 19:36 GreenDimond well I use xban 19:36 Krock banning will most likely check whether the auth exists, so delete after banning 19:36 GreenDimond right 19:36 GreenDimond I was just thinking if it checks if they are banned and they dont exist it might not work 19:37 Shara I think sban lets you ban an IP address even if it never connected, but would have to look to be sure 19:37 GreenDimond but xban should be fine with it 19:37 GreenDimond I dont currently use sban because I think it was too much work to set up 19:37 Krock banning is done with the on_(pre)joinplayer so auth doesn't matter if the name matches 19:37 Shara I just like the information I can pull from it more 19:37 GreenDimond Yes sban info is great 19:37 Shara Plus xban was constantly corrupting no matter what I did toward the end 19:37 GreenDimond HT started using sban 19:37 GreenDimond Yes 19:38 GreenDimond xban was corrupting a lot on HT as well xD 19:38 Krock inb4 there's now safe_write_file 19:38 GreenDimond I should probably start using it 19:38 Krock but sqlite3 for multi-megabyte databases is good too 19:39 GreenDimond Doesnt look too hard to set up 19:39 GreenDimond oh gosh 1700 lines all in one file 19:40 GreenDimond um 19:40 GreenDimond sban has no way to access the ban method outside of it 19:41 GreenDimond xban has a ban function 19:41 Krock GreenDimond: I can tell you that biome_lib was 750-ish lines but very hard to understand 19:41 GreenDimond Can confirm 19:41 Krock whereas sban should be well structured 19:41 GreenDimond it is 19:41 GreenDimond and ctrl+f is e z 19:41 Krock xban has a ban functio 19:41 Krock o rly? 19:41 GreenDimond yes 19:42 GreenDimond xban.ban_player 19:42 Krock that was sarcasm 19:42 GreenDimond sban does not 19:42 GreenDimond fine 19:42 Krock ah, I see that you're talking about API functions 19:42 GreenDimond xban has a global* ban function 19:44 GreenDimond sheesh 19:44 GreenDimond ima have go globalize it myself I guess 19:44 GreenDimond not hard 19:44 GreenDimond hecc maybe I should PR it too 19:47 tumeninodes sounds like Shara needs a rolodex-like workspace design 19:47 Shara Hi tumeninodes :) 19:48 tumeninodes Hi 19:49 tumeninodes o...m...g I swear, every single time I just hop onto irc... my phone rings 19:50 Krock maybe mentions in IRC use the ring sound? 19:50 Krock tumeninodes ^ 19:50 * tumeninodes is suspicious 19:50 tumeninodes hahaha, that would be extremely funny 19:50 Krock Kowalski: Analysis. 19:51 GreenDimond oof 19:51 tumeninodes ring ring. "Hello?" (dial tone) "WHO IS THIIIIS!!!??" 19:51 Krock This is Patrick 19:52 tumeninodes I thought you were stupid? 19:52 Krock That's where you're wrong (kiddo) 19:52 tumeninodes "Nooo, I'm Texas!" 19:53 Krock I don't know a meme sentence for that one 19:53 tumeninodes you are not the SpongeBob fan you presented yourself to be 19:53 tumeninodes fraud 19:54 tumeninodes dammit, this one's relentless, I have to answer it brb 20:07 GreenDimond sofar: one second seems like too long 20:07 GreenDimond half a second might be better 20:10 sofar tune it to your liking 20:40 rdococ aw, Gun Party doesn't seem to be up 20:41 tumeninodes the what? ^ 20:41 rdococ the "Gun Party" server 20:41 tumeninodes ohhhh, phew 20:41 rdococ lol 20:41 GreenDimond Krock: There, I made the PR https://github.com/shivajiva101/sban/pull/23 20:42 tumeninodes why has no one just made a random_ban mod? 20:43 Krock GreenDimond: I think the local scope is intended for a few of those functions 20:44 Krock such as update_login 20:44 tumeninodes btw something on the idea of this might be more fun for servers, and weed out the riff-raff https://www.youtube.com/watch?v=q5PiPIffTxg&list=PLbKW2l97ZjJKdiaX-piRebshJYkEhuHyo&index=16&t=0s 20:44 tumeninodes !title 20:44 MinetestBot tumeninodes: Last To Leave Minecraft Server Wins $1000 - Challenge - YouTube 20:45 tumeninodes minus a cash reward but, when spawned into a server, before one can even read the rules, they have to pass a challenge 20:45 GreenDimond local function randomban() if math.random(1,1337) == 1337 then for _,player in pairs(minetest.get_connected_players()) do if math.random(1,42) == 42 then minetest.ban_player(player:get_player_name()) return end end end 20:45 GreenDimond oh whoops 20:45 tumeninodes or a puzzle of some sort 20:45 GreenDimond add minetest.after(60, function() randomban()) before the last end there 20:46 tumeninodes woops sorry to interrupt dudes 20:46 GreenDimond tumeninodes: Theres your random ban mod^ 20:46 tumeninodes haha sweet 20:46 GreenDimond Krock: Which ones should I un-globalize 20:48 rdococ make people solve calculus problems before being allowed into the server 20:48 GreenDimond get_id, next_id, update_login, reset_orphan_record Krock? 20:49 tumeninodes Calculus was a tyrant 20:49 tumeninodes 3rd ruler of Algebra 20:53 GreenDimond Krock: Removed those; also fixed unsban.ban_player ;P 21:06 Krock GreenDimond: I'd rather ask shiva instead of me. I didn't write the mod 21:06 Krock rdococ: regular text input sucks if you're asked for a formula 21:19 * Sokomine bans the random ban mod 21:22 * benrob0329 has an idea for world portals in Lua 21:27 tumeninodes aww :( 21:30 benrob0329 is there a way to mark nodes/entities as visible or non visible per-player? 21:31 rubenwardy no 21:31 benrob0329 or to make them fully culled when viewed from behind? 21:32 benrob0329 So that if you have say, a sphere with inverted normals you can't see the inside from the outside? 21:40 rubenwardy that's possible if you enable(?) backface culling 21:43 benrob0329 This is a different problem, because while backface culling will not render the backs of polygons, it will render any forward facing polygons that would otherwise be behind the back of the mesh 21:43 benrob0329 rubenwardy: this could be considered a bug, but probably not as I believe this is pretty standard behavior 21:45 benrob0329 If we had an option to enable hidden-face culling or something, that would be useful 21:46 benrob0329 ah, hidden surface detection seems to be the general purpose name 21:46 benrob0329 *determination 21:48 benrob0329 I do wonder though, with server-side mapblock culling, would a hidden node simply not be sent to the client? 21:50 rubenwardy culling is on a per-mapblock basis, not pernode 21:50 rubenwardy all nodes in a mapblock are sent 21:51 benrob0329 I see 22:02 rdococ russian roulette 22:03 rdococ instead of dying, you get banned 22:24 rdococ I wish there was a PvP duel server 22:30 benrob0329 rubenwardy: which do you think would be more likely to be included, per player entity visibility or hidden face culling? 22:30 benrob0329 if I made a feature request 22:31 benrob0329 or should I make both as they are both useful and see if one gets a PR? 22:48 garywhite Happy (late) New Year 23:29 benrob0329 #8045 23:29 ShadowBot https://github.com/minetest/minetest/issues/8045 -- Allow Changing Entity Visibility On A Per-Player Basis