Time Nick Message 00:14 Out`Of`Control what could cause : 00:14 Out`Of`Control WARNING: saveBlock: Block failed to save (12,2,-6): SQL logic error or missing database 00:15 Out`Of`Control all loads fine, when i try edite something, this starts 00:21 est31 Out`Of`Control, perhaps problem with rights on the sql file? 00:22 Out`Of`Control how can i check ? 00:22 Out`Of`Control i chown all files 00:23 est31 i cant help if i dont have the full error 00:26 Out`Of`Control when i try edite or place block, i get this error 00:26 Out`Of`Control lots of them 00:56 paramat will push soon game#594 00:56 ShadowBot https://github.com/minetest/minetest_game/issues/594 -- Default: Add papyrus schematic by paramat 00:59 paramat http://i.imgur.com/KQ84FpT.png mgv7 mountain with triple stone biome blend 01:02 paramat http://i.imgur.com/5W7jz0k.png ideal for conversion to a base with a nice view 01:25 paramat now pushing to game 01:29 paramat complete 04:11 bobomb how do you test if a node is of a certian group, like walkable? 04:54 hmmmm walkable is a node property, not a group 04:54 hmmmm nodedef->get(node_here).walkable 05:11 bobomb in lua? 05:12 hmmmm no, in the core 05:12 hmmmm this channel is for core development 05:13 bobomb oh 06:58 cornernote any ideas on how this can happen? https://github.com/cornernote/minetest-inventory_plus/issues/4#issuecomment-125466494 06:58 cornernote init.lua:60: attempt to index a nil value 06:59 cornernote line 60 is player:get_inventory():set_width("craft", 3) 07:01 kahrl perhaps player is nil? 07:04 hmmmm cornernote, this is clearly not minetest core development 07:41 cornernote hmmmm, sorry i'll take it to #minetest 07:58 rubenwardy Krock, http://irc.minetest.ru/minetest-dev/2015-07-27#i_4341933 07:58 rubenwardy Client side modding is a good idea 07:58 rubenwardy increased prediction, guis, replaced formspecs 07:58 Krock rubenwardy, how about the hacking problem? 07:59 rubenwardy hacking is not a valid counter argument: 1. mods will be supplied from server. 2. server side authority still applies 3. security from obsurity, kind of 08:01 rubenwardy http://dev.minetest.net/Client_scripting_plans 08:04 rubenwardy anyway, gtg 09:06 VanessaE nrzkt: I was looking at the Minetest play store app's comments, and there's a couple here you need to address/respond to. 09:08 nrzkt VanessaE: i don't look at play store very often because comments are not very constructive many time, but ok, i will look at it later 09:09 bluegreen can someone tell me why this code crashes minetest http://paste.ubuntu.com/11952479/ ? 09:09 VanessaE one of them is accusing this of being a direct 100% fork of minecraft, with "minor" changes to "avoid copyright". 09:09 bluegreen debug.txt is empty 09:10 VanessaE bluegreen: that's a #minetest question 09:10 VanessaE however I don't see anything wrong with the code. 09:10 nrzkt VanessaE: i already answer to same question later in comments if i remember 09:11 VanessaE nrzkt: hm, ok. interesting that it doesn't show up as a reply to that post. 09:11 * VanessaE continues reading 09:12 nrzkt maybe not same person 09:25 nore this needs to be fixed: "AWFUL!!! I started underground and I didn't have flying privilige" 09:25 VanessaE agreed. 09:25 VanessaE though I thought there was already some work done on that recently? 09:26 nore hm... I don't know :/ 09:27 nore some comments about the menu being mobile unfriendly too 09:28 nrzkt yes, maybe having a android native menu could be better no ? 09:28 nrzkt if we can have some java parts in our minetest build maybe we can add more from android-sdk ? 09:28 nrzkt like a public server list using customized listviews 09:30 nore nrzkt: if someone is able to do it, then it could be useful 09:35 nrzkt i know how to do it natively using Android-SDK but maybe it will replace all our current interface with Android native interface 09:36 nrzkt and only launch minetest when connecting to server or singleplayer 09:36 bluegreen that sounds like a good idea 09:36 Calinou use Material Design :P 09:37 nrzkt material design isn't supported everywhere 09:37 Calinou please don't make 2.3-era ugly GUIs 09:37 Calinou any app that is being redone today is redone in Material Design, period 09:40 nrzkt yes but many of our users don't support it :p 09:40 nrzkt and i don't say i will do it now 09:41 nrzkt i have many things to do and having an interaction between minetest and android is very complex because this isn't native java :( 09:43 VanessaE nrzkt: too late, you've already been volunteered ;) 09:47 nrzkt xD 16:09 kilbith https://github.com/minetest/minetest/tree/nrzforked_experiment 16:09 kilbith huh ?? 16:11 nrzkt fuck of 16:11 nrzkt thanks :) 16:11 nrzkt fixed :p 17:38 bluegreen Hi everyone 17:40 VanessaE hi. 17:52 est31 it seems google has posted a library for material design backwards compatibility. 17:52 est31 giving compat back to api level 7 17:52 est31 we support up to 9 17:52 est31 so it shouldnt be an issue 17:52 est31 from 9* 17:56 hmmmm so what blocking bugs do we have 17:56 hmmmm as far as i can remember there's... 17:56 hmmmm - textures don't work on android 17:57 hmmmm - graphical glitches on android 17:57 est31 - textures don't work on android might be a setup problem of me 17:57 hmmmm - security flaws with network data handling 17:57 VanessaE #2913 17:57 ShadowBot https://github.com/minetest/minetest/issues/2913 -- Unexplained, random crashes (segfaults, aborts, OOM) 17:57 hmmmm uhh 17:57 hmmmm yeah ^ 17:57 est31 are there flaws? 17:57 hmmmm - the memory leaks 17:57 hmmmm yeah I am working on them 17:57 est31 those unexplained crashes arent new, are they? 17:58 est31 its just now that somebody started putting them into a bug, no? 17:58 VanessaE est31: no, but someone marked it as a blocker. 17:58 hmmmm - the inefficient resource usage of minimap 17:58 est31 hmmmm, can you explain the last one? 17:58 hmmmm well 17:58 hmmmm the entire minimap generation algorithm is incredibly inefficient 17:58 est31 we do a prescan 17:59 est31 then we take those prescannned values to use them in the main generation 17:59 hmmmm but what's worst about the current setup is that the prescan takes a ~30% performance hit on mesh making 17:59 est31 the alternative would be to copy the mapblocks -- again 17:59 hmmmm and then there's the FPS regression as well, with and without shaders 17:59 hmmmm and then there's the ATI shaders bug 18:00 hmmmm holy crap we have so many problems 18:00 VanessaE oh also: this is hearsay, but someone reported that Android performance is...well.. pathetic in the official app, compared to some of the clones (to the tune of 40fps versus 10 on the same hardware) 18:00 hmmmm interesting 18:00 est31 the ATI shaders bug is the thing pointed out by kaeza? 18:00 hmmmm yeah 18:00 hmmmm we can't release with that... 18:00 est31 and the fps regression is the thing with the looking at a wall? 18:00 hmmmm no 18:00 hmmmm the FPS regression is just a 25% regression across the board 18:00 est31 agreed we cant release with the ati shaders bug 18:00 hmmmm in any scene 18:01 hmmmm when your FPS is already like 20 18:01 est31 has that been bisected? 18:01 hmmmm a 25% performance hit hurts a lot 18:01 hmmmm yes I bisected it 18:01 hmmmm it's RBA's recent shaders rework 18:01 est31 can you make a bug report? 18:01 hmmmm meh okay 18:01 est31 its good if we can track these things in github 18:02 est31 then we have a list and dont forget stuff 18:02 VanessaE inb4 RBA blames the regression on drivers :) 18:02 VanessaE oh wait, he did that already :P 18:02 est31 also explain what minimap generation could do better in your eyes. 18:02 hmmmm well I already explained it 18:03 hmmmm last I saw of RBA on the channel he admitted that the prescan was a shitty idea and he's working on it 18:03 hmmmm i don't think that's necessarily a blocker though 18:03 est31 we do have a prescan already now 18:03 VanessaE no offense to RBA, but frankly, he's always "working on" whatever, for days to weeks at a time :( 18:03 hmmmm because... 18:03 hmmmm i added the ability to actually disable the minimap 18:04 hmmmm so if it's absolutely horrifying we can just set enable_minimap = false 18:04 est31 great 18:04 hmmmm but yeah 18:04 hmmmm minimap on android? that'd be a killer 18:04 hmmmm as if it wasn't bad enough on PC 18:05 est31 it did work, back when I tested it. 18:05 est31 just no button for showing it 18:05 hmmmm the no. 1 complaint about android is lag 18:06 hmmmm it doesn't help if we 1). add yet another thread and 2). decrease meshmaking throughput by a solid 30% 18:07 est31 even my ultra old i9100 has dualcore 18:08 hmmmm yes but we have the render thread, meshmaking thread, connection thread, minimap thread, and a gazillion other threads from the rest of the OS all competing for that time slot 18:08 hmmmm apart from the connection thread, all those are slurping up CPU 18:09 hmmmm also: battery life 18:09 est31 yea agreed 18:12 hmmmm ahhhh :D 18:12 hmmmm VanessaE 18:12 VanessaE hm? 18:12 * VanessaE hides 18:12 hmmmm you got a backtrace of the infamous refcount < 0 bug 18:12 VanessaE I did? 18:12 hmmmm yes 18:12 VanessaE \o/ 18:13 hmmmm that was the one sapier added when he redid the network 18:13 VanessaE speaking of sapier, wtf ever happened to him? 18:14 hmmmm I think he got pissed off with minetest 18:14 VanessaE ohh 18:15 hmmmm the OOM crash was Lua related 18:15 hmmmm lol wtf sfan 18:15 hmmmm "Backtrace of crashed thread contains Connection and Profiler, crash was probably caused by one of them." 18:15 hmmmm no.... no 18:15 est31 ^ 18:15 hmmmm just because you see a thing you recognize in a blob of computer output doesn't mean that must be it 18:15 VanessaE sorry to spam the bug with all those backtraces, but the thought got into my head that enough of them would reveal a pattern 18:16 hmmmm all of them are different lmao 18:16 VanessaE yeah 18:16 hmmmm the first one I can't really do anything about though 18:16 hmmmm you can clearly see it's crashing somewhere in the constructor of an object we create with operator new 18:16 VanessaE which made me think of that old memory corruption bug that's sorta "floated around" for a while now 18:17 hmmmm just because a segfault happens in new doesn't mean it's memory corruption 18:17 VanessaE oh I know 18:17 hmmmm new doesn't just malloc() memory; it also calls the object constructor 18:17 hmmmm in this instance it crashed in the ctor 18:17 nrzkt which ATI bug ? I have a AMD card too and i didn't notice any bug 18:17 hmmmm but there's no backtrace because no doubt the stack got smashed 18:17 est31 nrzkt, dunno why he calls it ati bug 18:18 est31 for me its related to old opengl version 18:18 hmmmm it's when you use shaders on certain ATI cards it'll make the bottom portion of the screen red 18:19 hmmmm ahh 18:19 hmmmm actually it seems i was wrong 18:19 nrzkt okay, i have a AMD 4890 it's not very old but it's two or tree generations before actual 18:19 hmmmm there is a backtrace for the new() segfault 18:20 kilbith fyi, kaeza uses an intel gpu (not amd/ati) : https://github.com/minetest/minetest/issues/2866 18:20 hmmmm well it seems like sapier's refcounting has some sort of major problem with it 18:20 hmmmm i'll have to work on all of these 18:21 hmmmm but the "out of memory" bug seriously looks like a mod problem to me 18:23 VanessaE pretty sure a genuine OOM is impossible on the Nostalgia server :) 18:23 hmmmm like i said, it's a bug 18:24 hmmmm maybe some integer that's supposed to be a count gets set to a bogus value somehow and is passed as a parameter to a lua api 18:24 hmmmm and the lua api dutifully attempts to allocate that amount 18:24 VanessaE could be *shrug* 18:25 hmmmm you see the "out of memory" string whenever too many tables are being allocated typically 18:25 hmmmm or too many entries 18:25 VanessaE I don't know the details of that subgame too well, so no idea whether it could actually be a game/mod bug or not 18:25 hmmmm like try creating a 500x500x500 3d noise object in lua 18:26 hmmmm when you see std::bad_alloc, on the other hand, that means there's an allocation in some kind of loop that went awry 18:26 hmmmm deeper in the core 18:27 hmmmm like for example if somebody forgot to initialize a variable that happens to be the count for a loop that allocates EmergeThreads :) 18:27 hmmmm it'll try to allocate 545930495495 or however many the uninitialized variable is, and then after the 200,000,000th emerge thread it'll run out of virtual memory that it's able to reserve 18:28 hmmmm so no, you're not out of memory 18:28 hmmmm it's just that the program did something wrong and stupid that there isn't enough virtual memory to satisfy 18:28 VanessaE ah right 18:30 VanessaE oh btw, est31 18:30 jin_xi meanwhile expiring particle spawner ids still not handled, so vector of ids is ever growing 18:31 hmmmm yes, we need to merge the particle spanwer fixes 18:31 VanessaE I'm getting occasional client exits when the server quits/shuts down. not if there's a pattern yet 18:31 hmmmm jin_xi: is your PR still a work in progress? 18:33 jin_xi nope, i have given up. see branch linked in PR for latest stage 18:33 hmmmm oh :( 18:33 hmmmm you should mention that in your PR then 18:33 hmmmm we'll take over 18:34 est31 VanessaE, thats autokick, no? 18:34 hmmmm or is there soemthing any of us can do to help? 18:34 jin_xi it is mentionned 18:34 VanessaE est31: yeah, instead of a dialog warning that the server went down, sometimes the client just exits. I have not grabbed a backtrace of it yet, though 18:35 est31 VanessaE, didnt you say one time that you never can exit minetetst to the main menu, because it always crashes? 18:36 est31 perhaps the same bug 18:36 VanessaE est31: sometimes it crashes when exiting to the menu, yes 18:36 jin_xi yes, look at this branch and let me know if its completely misguided: https://github.com/obneq/minetest/tree/scenenodeparticles 18:36 VanessaE not always though 18:52 est31 ah it seems it can compile libiconv 18:52 est31 but somehow fails after that 18:52 est31 interesting 18:52 est31 very interesting 19:23 H-H-H also est31 could you check sound on latest head in android ? 19:51 est31 H-H-H, one moment 20:11 paramat nore sfan5 any comments/approval for game#588 ? 20:11 ShadowBot https://github.com/minetest/minetest_game/issues/588 -- Default: Thicker snow nodebox by paramat 20:11 luizrpgluiz It would be possible in the future to launch a new api to make mods for the minetest? Ie launch 100% to api minetest? 20:12 est31 luizrpgluiz, do you use google translate? 20:12 est31 i barely understand you 20:12 est31 ermm I dont understand you at all 20:13 paramat .. btw next i will prepare the mgv5/v7 biome commit 20:13 luizrpgluiz yes i using google translate 20:15 luizrpgluiz a friend told me it would be great launch api minetest 100% to facilitate the creation of mods in the game 20:15 bluegreen minetest has a lua api for mod to create mods 20:15 est31 yes 20:18 TBC_x I believe that networking code is slow beacause it throws whenever it gets an opportunity, and that throw generates std::String. 20:21 paramat okay thanks guess i'll push thicker snow, if sfan doesn't object 20:28 Megaf paramat: I do 20:28 Megaf paramat: I do like the way it is 20:29 Megaf but it would be cool if we had snow fall and it acumulated 20:32 paramat snowfall needs better particle performance which may come in future 20:33 Red-001 https://github.com/minetest/minetest_game/pull/595 20:34 TBC_x game#595 20:34 ShadowBot https://github.com/minetest/minetest_game/issues/595 -- Change a book_written into a book if empty. by red-001 22:08 est31 seems very good, I seem to have fixed libiconv... 22:11 H-H-H sorry est32 if you answered my question regarding sound i disconected 22:11 Red-001 game#595 22:11 ShadowBot https://github.com/minetest/minetest_game/issues/595 -- Change a book_written into a book if empty. by red-001 22:12 est31 H-H-H, still on some other issue 22:12 est31 fixing iconv right now 22:12 H-H-H :) i didnt miss your reply then good :) 22:12 est31 I dont know why i wrote such shit 22:12 est31 its really that 22:12 est31 complete shit 22:13 H-H-H coz it does/did the job :) 22:31 TBC_x this one is funny https://forum.minetest.net/viewtopic.php?f=6&t=12924 22:33 Red-001 is quantum mechanics included? 22:34 Red-001 or this that an extra mod? 22:35 TBC_x nah, just simple quantum computer emulator 22:36 Red-001 :( 22:37 TBC_x The guy uses words exactly like me when I have no idea what I'm talking about. 22:37 Red-001 lol 22:38 Red-001 you should probily fix the glich that a allowed you to fly when you say /grant singleplayer fly 22:38 TBC_x But I can make octopus square! 22:39 Red-001 lol 22:39 TBC_x Everyone can 22:40 TBC_x It just depends on individuals that realize that they can 22:40 Red-001 btw does minetest have a thermal velocity? 22:40 TBC_x thermal? 22:40 Red-001 nvm 22:41 TBC_x there's so much velocities transported over network It may be even there 22:41 Red-001 i meant terminal velocity 22:42 Red-001 sorry 22:42 TBC_x terminal velocity? deacceleration? 22:42 TBC_x I must research what perdictions the client even does 22:43 Red-001 as in what is the max speed when your falling 22:43 paramat no terminal velocity as far as i know 22:43 paramat to keep stuff simple 22:43 Red-001 does it just keep accelerating? 22:43 TBC_x no 22:43 TBC_x you hit the bottom of unloaded mapblock 22:43 TBC_x :) 22:44 Red-001 we don't need a terminal velocity 22:44 TBC_x i mean top 22:44 Red-001 we have ignore 22:44 paramat yeah. you have to load in a huge vertical space then fall through it 22:45 * Red-001 thinks about skyblock 22:45 TBC_x I think that client shouldn't be able to request mapblocks from server 22:45 TBC_x instead, just send its rendering distance 22:46 Red-001 why? 22:46 Red-001 hacked clients? 22:46 TBC_x that's probably another reason 22:46 Red-001 downloading the whole map? 22:46 Red-001 and casing a crash 22:47 Red-001 wow 22:47 Red-001 mg_vilages looks a lot cooler then last time 22:47 TBC_x the server should, aside from simulating the world, also predict and load mapblocks 22:48 Red-001 yeah 22:49 Red-001 won't that break connecting with old clients? 22:49 TBC_x will have to research that one 22:50 TBC_x if client accepts any mapblock server serves, client's request can be completely ignored 22:51 Red-001 is there and editor for .mts files minetest uses? 22:51 TBC_x well... or at least calculate its viewing distance from the request 22:51 TBC_x .mts? 22:52 Red-001 schematics 22:52 est31 paramat, ^ 22:52 est31 your question :) 22:52 paramat erm 22:53 Red-001 like how do it set chance and all? 22:53 paramat no, but the way to create a schematic is shown here https://github.com/paramat/saveschems 22:54 Red-001 what about changing existing schematics ? 22:54 paramat and see https://forum.minetest.net/viewtopic.php?f=9&t=12011 22:55 Red-001 ok thanks 22:55 paramat to change existing you can use minetest.serialize to convert from .mts to a lua table, then edit and convert back 22:55 Red-001 ok thanks 22:56 paramat more info in lua_api.txt 22:58 Red-001 bye 23:16 est31 good thing that I've made the commit msg of 33a633a6317d272ffd06320c92416644a367b186 so verbose