Time Nick Message 02:54 paramat est are you able to +1 #2827 ? i tested it and it works well 02:54 ShadowBot https://github.com/minetest/minetest/issues/2827 -- Allow overriding the hand by TeTpaAka 03:13 paramat will merge game#1409 in a moment 03:13 ShadowBot https://github.com/minetest/minetest_game/issues/1409 -- Farming/fire: Add tool break sounds by paramat 03:15 paramat merging 03:18 paramat complete. wooo 11 PRs 03:30 est zeno gave a +1, I gave a +.5 you give a +.5 and we have +2 :) 03:34 paramat yeah 03:35 paramat i feel it has enough review and approval now 03:36 garywhite ? 03:37 garywhite paramat what did I miss? 03:37 paramat #2827 03:37 ShadowBot https://github.com/minetest/minetest/issues/2827 -- Allow overriding the hand by TeTpaAka 03:40 paramat so i'll merge that in a moment 03:50 paramat merging 03:57 paramat finally merged. no more naked hands in space 03:58 paramat i can add a glove to my moonrealm mod 03:59 paramat i hit 200 commits in game. so-far hit 100, congratulations 04:01 garywhite congratulations paramat & sofar 04:02 paramat amazing how neglected game used to be, the rate of commits has increased over the last 2 years 04:09 paramat minetest repo has a new branch called 'update credits'? 04:10 paramat seems to be an error 04:13 paramat something to mention to zeno 16:44 sfan5 Fixer: https://github.com/minetest/minetest/pull/4812 16:47 Fixer sfan5: nice 16:48 Fixer but that needs server and mods to update before fixing this? 16:48 Foghrye4 As I remember Nrzkt suggest to everyone to use UNORDERED_MAP everywhere if possible. 16:48 sfan5 Fixer: yes 16:49 sfan5 Foghrye4: m_detached_inventories is not an unordered map yet, maybe it's not because it will break otherwise; i dunno so i left it that way 16:49 Fixer that means it will take some time before I can actually test it 16:50 sfan5 both rc and ve-creative use -dev 16:50 sfan5 should be no problem for them to upgrade a little 16:51 sfan5 and fixing mods is literally just one added line https://github.com/minetest/minetest_game/pull/1413/files 16:51 Fixer ok 16:51 sfan5 s/rc/xanadu/ 16:52 Fixer i will not test on xanadu, admin is not adequate 16:53 Fixer kicks me out 16:53 Fixer Vanessa/Shara's server are enough 16:54 sfan5 red cat creative is using the release version, upgrading might not be the best idea 16:54 sfan5 though you could apply this patch to the release tree 16:55 Fixer https://github.com/search?utf8=%E2%9C%93&q=create_detached_inventory%28&type=Code&ref=searchresults 16:55 sfan5 lots of mods in theory 16:56 sfan5 but on the servers i tested on it was just 3 or 4: mt_game creative, armor, bags, unified_inventory 16:57 Fixer yes 16:57 Fixer just those 4 will probably fix 80% of problems 16:59 Fixer also this https://github.com/search?utf8=%E2%9C%93&q=org%3Aminetest-mods+create_detached_inventory&type=Code&ref=searchresults 16:59 Fixer just 4 items 16:59 Fixer minetest-mods/global_exchange – exchange_machine.lua 16:59 Shara sfan5: RC does not use dev. I have stressed this a few times, but no one seems to hear it. 17:00 Shara But other servers that experience it do 17:00 Fixer vanessa and xanadu uses 17:00 Shara Yes. 17:00 Shara I don't want to run dev on RC, because I don't have time to update my mods often enough 17:01 Shara If I knew it would guarantee a fix I would, but while those on dev are experiencing the issue, that seems unlikely. 17:02 Fixer Shara: 0.4.15 release (i hope) will be at the end of the year, I hope it will include this fix 17:03 Shara Fixer: I will hope too. 17:03 Shara Thanks (to others here as well) for looking into this. 17:04 Fixer thanks sfan5 for looking into this 17:05 sfan5 Shara: yes i confused the two 17:06 Fixer i hope it gets positive reviews, should not be hard to update the mods, server owners will rejoice 17:07 Shara Players will too. 17:07 sfan5 oh also there seems to be a second unrelated bug: 17:07 sfan5 if you disconnect before stage two init happens you will be at 0,0,0 next time you join correctly 17:07 sfan5 (possibly only when you join for the first time or something) 17:41 est sfan5: in which way will your PR fix the issue? 17:41 est any explanation on the theory you came up with? 17:42 sfan5 start reading here http://irc.minetest.net/minetest-dev/2016-11-25#i_4749185 17:44 est ahh I see 17:44 est makes total sense 17:45 est sfan5: and by " when someone decides to allow player-attached invs from the script api " you mean a function on the player SAO, right? 17:45 sfan5 something like that 17:46 est I doubt that that system would be better, because it only works on logged in players 17:46 sfan5 why 17:46 est but either way, what the PR does concept wise is good 17:47 * est looks at the implementation 17:47 sfan5 for example the creative inventory can only be looked at by the player when he is actually logged in 17:47 sfan5 creating detached invs for each player is feature misuse actually 17:47 sfan5 ideally you could create player-attached invs like i said 17:48 est ah you mean that its stored in the player files 17:48 est yeah thats a better place 17:49 est (even though we have tons of issues with player files already and storing _more_ stuff in them makes it even worse) 17:49 est but hmmm doesnt like dbs so its broken... 17:50 est yeah the PR is a total hack 17:51 est but whatever, if it fixes the issue, its okay 17:51 sfan5 concept or code wise? 17:51 est concept wise 17:51 sfan5 i agree on that 17:51 est code wise its okay 17:52 sfan5 est: about the storage issue: in the case of the "creative" inv it would not need to be permanent 17:52 est yes 17:52 sfan5 the "refill" inv from unified_inv wouldn't need storage either 17:53 sfan5 only stuff like "armor" or "bags" would need that 17:53 est yup 17:53 est so probably a bool param would be a good idea 17:53 est to work on the SAO 17:53 est I guess an SAO only method would be enough for almost all use cases 17:54 est can't think of one where you want to modify the detached inv of a non emerged player 17:54 sfan5 something like player:create_attached_inventory(name[string], callbacks[table like w/ detached], permanent[bool]) 17:55 est yes 17:55 sfan5 actually doesn't create_detached_inventory only create an invlist? 17:55 sfan5 since there's no "main", "craft" and stuff? 17:56 sfan5 or do i have my terminology wrong 17:56 est what I wonder about: why do they actually use detached inventories for things like armour? 17:57 est why can't they use player inventories for that 17:57 sfan5 callbacks 17:57 est err... so might it be a better solution to make callbacks for player inventories then? 17:58 est and some sort of per player non-permanent inventory 17:58 sfan5 yes 17:58 sfan5 but i don't think anyone will get that done (not that it wouldn't be possible) before the feature freeze 17:59 sfan5 so i opted for the simpler solution 17:59 est yes 17:59 est simple solution is okay, after the release someone can tackle that issue 17:59 sfan5 and deprecating what my PR adds won't even break anything 18:00 est can you still add a note that its only a temporary thing to resolve a temporary issue and will be removed in a future release? 18:00 sfan5 in the docs? code? both? 18:01 est lua_api.txt is enough 18:03 sfan5 added 18:03 est thanks 18:04 est can the server owners of affected servers merge PRs to test them? 18:04 est if yes, we should wait for their results before a merge 18:04 est if no, we should merge it so that they can test it 18:05 sfan5 they probably can 18:05 sfan5 ping VanessaE 18:06 celeron55 4403 is quite obscure, i don't think 4812 should be merged without testing if the purpose is to fix 4403 18:07 est I was just thinking about non technical server owners who dont know how to compile minetest, and who have installed the ppa 18:07 celeron55 it seems like a useful workaround, but for a problem that could be completely irrelevant 18:07 * VanessaE hides 18:07 est but as VanessaE can reproduce the issue, it should be no problem :) 18:08 est if we just knew where she was .... :) 18:08 sfan5 well the actual problem might be: why does the network stack break with slow clients(?) in a way that sending 100 packets is super slow 18:08 juhdanad Is there anybody whose client doesn't crash at the waterfall? 18:08 sfan5 me 18:09 juhdanad Could you please place a 'Lamp' node on the server, please? 18:09 juhdanad It has light level 15. 18:09 juhdanad For me, this causes the chrash. 18:09 sfan5 not now 18:11 sfan5 VanessaE: you'll need to apply https://github.com/minetest/minetest/pull/4812 & https://github.com/minetest/minetest_game/pull/1413 and do the same fix to unified_inventory/bags.lua, unified_inventory/callbacks.lua and invsaw/init.lua 18:12 VanessaE I'll let someone else test that, my servers and mods are somewhat out of date. 18:13 sfan5 that doesnt matter 18:13 est https://github.com/minetest/minetest/issues/4813 18:13 est ^ long term fix for the issue 18:18 VanessaE er, do I apply 4812 or 4813? :P 18:18 sfan5 4813 is an issue not a PR 18:18 sfan5 good luck applying that ;) 18:19 VanessaE oh yeah, so it is :) 18:23 paramat juhdanad dev version limits light sources to 14 https://github.com/minetest/minetest/commit/5091cb5ecdac39a49bce2317d7054d69ed43f2c1 15 caused lighting bugs 18:24 paramat .. but not a crash 18:24 Krock and with that commit the super glow glass kinda lost its purpose 18:25 juhdanad But on Red Cat Creative there is a node named 'Lamp'. 18:25 juhdanad And its light_source is 18. 18:25 juhdanad This causes the crash. 18:26 Krock lol. How's the possible? is builtin not up-to-adte? 18:26 Krock *date 18:26 paramat RCC uses 0.4.14 stable 18:26 VanessaE even if it is, it shouldn't crash the client.. 18:26 VanessaE isn't* 18:26 Krock and 18 wouldn't even fit in that 4-bit space where the day light is stored 18:27 paramat how would it overflow? go negative? 18:27 juhdanad There are two options on client side IMO: 1- check received content features, 2- let the light propagating algorithm check the light_source property of the nodedef. 18:28 juhdanad I would prefer the first. 18:29 Krock depends how it's programmed paramat. It also could overwrite the night light table if these are the lower 4 bits 18:29 Krock but I think there are many checks to prevent from that 18:29 paramat well it's best limit the value in register as in my commit 18:29 Krock even better would be C++ side, just to make sure 18:30 paramat does such a lamp crash dev version? 18:31 juhdanad Yes, it can cause crash, this is the secret of the waterfall. 18:32 VanessaE sfan5: http://pastebin.ubuntu.com/23538916/ does this look about right for unified_inventory? 18:36 juhdanad There it is: https://github.com/davisonio/darkage/blob/master/nodes.lua#L226 18:40 paramat i'll try to fix this in register later then 18:42 VanessaE well in any case, both pull requests applied, both mods patched per the above, and all of it uploaded/staged for the morning reboot in ~11 hours. 18:42 juhdanad But the client should also check this, so servers can not make clients chrash. 18:42 paramat ok 18:45 Krock paramat, would you have some time to put your opinion about game#1412 18:45 ShadowBot https://github.com/minetest/minetest_game/issues/1412 -- Carts tweak by SmallJoker 18:45 Krock + question mark 18:46 paramat sure i'm reading the controversy now 18:46 sfan5 VanessaE: yup looks correct 18:46 VanessaE ok. 18:47 paramat Krock can you comment on game#1408 ? 18:47 ShadowBot https://github.com/minetest/minetest_game/issues/1408 -- Carts: Increase power rail acceleration to help climbs by paramat 18:47 VanessaE I see a variant of boost cart went into mt_game.. safe to discontinue using the upstream version? 18:47 Krock I'll try to write some words. Followed the discussion but somehow it feels like it's going ino nowhere 18:48 paramat no merged carts have a bug 18:48 VanessaE oh. 18:49 Krock boost_cart yet overwrites the behaviour of MTG when activated 18:49 VanessaE ok then I'll just leave it in place. 18:49 paramat 1408 helps a lot but i need support for it 18:49 paramat rubenwardy ^ 18:49 paramat it's controversial 18:50 kaeza FWIW, :+1: from me to acceleration 18:50 paramat ^ sfan5 nore 18:50 kaeza I love doing that in... that other voxel game :) 18:50 sfan5 haven't worked with cards yet, so i can't comment right now 18:51 sfan5 carts* 18:52 rubenwardy I'm doing coursework atm, will be able to look later 18:53 paramat merged carts should be good at release though 18:54 paramat if people support my approach 18:58 sfan5 juhdanad: that seems unlikely https://github.com/minetest/minetest/blob/cc0b3c1cd14f036d6336ae0fb2329fd245b6b4b4/builtin/game/register.lua#L130L133 18:59 sfan5 builtin limits the possible light level 18:59 Krock ^ repost alert 18:59 sfan5 oh didnt see that 19:02 juhdanad sfan5: yes, but with gdb I suspended the execution when I placed a darkage:lamp, and in the nodedef manager it had light_source=18. 19:03 Fixer xanadu is installed from ppa 19:04 sfan5 that meeeans: red cat creative is at fault 19:04 sfan5 but the client should probably not crash on invalid data 19:06 garywhite What are you guys talking about? 19:07 paramat read logs 19:07 garywhite I've only been here for 5 mins 19:07 Krock pls http://irc.minetest.net/minetest-dev 19:08 sfan5 http://irc.minetest.net/minetest-dev/today 19:08 Krock It's even in the topic 19:09 Krock paramat, 1412 does not contain the acceleration increase of powerrails 19:10 paramat ok 19:24 sofar #4729 I'm unable to even start minetest, bad, bad bug 19:24 ShadowBot https://github.com/minetest/minetest/issues/4729 -- Crash on tab_multiplayer.lua 19:27 sofar where does the engine store favorites? 19:36 sfan5 client/serverlist/favoriteservers.txt 19:36 sofar yeah, I found it, I have duplicates in it 19:36 sofar that's causing the nil crash somehow 19:40 juhdanad paramat: I made a quick fix - #4814 19:40 ShadowBot https://github.com/minetest/minetest/issues/4814 -- Limit light_source in the engine by juhdanad 19:45 sfan5 the change in nodedef.cpp should be enough 19:45 sfan5 also you can remove the { and } 19:50 est ohh players won't like 4814... it will destroy their farms 19:51 VanessaE ? 19:52 paramat heh too bad >14 causes light bugs 19:52 VanessaE why would it break farms? 19:58 est it will break underground farms that work with artificial light sources 19:58 sfan5 those are already broken 19:58 VanessaE yeah but why would that matter? 19:58 est but you can simply adjust mtgame to support it again (changing the light level where growth is posisble) 19:58 VanessaE no light source is supposed to exceed 14 anyway 19:58 sfan5 builtin already does this 19:58 VanessaE oops. 19:59 sfan5 what is new is that the client enforces this too 19:59 est i see. 19:59 sfan5 this is to prevent breaking with old servers that don't enforce this limit 19:59 juhdanad Underground farms? 19:59 VanessaE so if you're breaking some farm that relies on >15 light sources, oh well. let it break, that's undefined behavior anyway. 20:00 juhdanad Light can't spread farther than 15 units because it has only 4 bytes. 20:00 VanessaE bits. 20:00 Krock sofar, updated the last carts commit. When punching the cart, it won't jump around anymore 20:00 sfan5 VanessaE: btw did you restart ve-creative yet or will it do that anyway in a few hours for backups? 20:00 Krock before it always teleported back to the last position 20:00 VanessaE sfan5:it'll restart on its own at 1am EST 20:00 est meh, why do i get pinged 20:01 Krock lol 20:01 sofar Krock: can I push the cart speed to your pr? 20:01 sofar the 6/7/8 change? 20:01 juhdanad Even if the light_source is 18, old servers made light 13 blocks distant. 20:01 Krock sofar, no thing. I can do that right now ;) 20:01 paramat Krock i noticed that punch teleport, thanks 20:01 juhdanad Just see Red Cat Creative - lamps are not 18 unit bright. 20:02 sofar Krock: already done, lol 20:02 sofar check your branch :D 20:02 Krock uh, you're fast 20:02 sofar I was testing your change for a while and had already ported it 20:02 sofar I have about 100 carts riding the 'coaster 20:04 Krock honestly I'd like to have this velocity behaviour in a seperate pull to move the discussion there 20:04 sofar oh, easy enough to split it out 20:04 sofar but we should merge your change first 20:04 Krock also, you've put the whole code into the loop, it calculates the same thing three times xP 20:05 sofar doh 20:05 sofar aight lemme fix that and take it out of this pr then 20:06 Krock ^^ 20:06 sofar can you force push it out? 20:06 Krock yes I acn 20:06 Krock done 20:07 paramat yeah i'd prefer game1412 stays separate, i'll test it later 20:08 sofar well crud 20:08 Krock time to bring the punch-fix into my original mod too :3 20:08 sofar that's a damn merge conflict due to the file change 20:08 Krock eww right 20:08 sofar FINE 20:08 sofar I'll just push your changes under it as well, your change should be merged first anyway 20:12 sofar game#1414 then 20:12 ShadowBot https://github.com/minetest/minetest_game/issues/1414 -- Carts speed by sofar 20:12 sofar one thing is sure 20:12 sofar I'm fairly happy with the amount of attention and testing this code is getting 20:14 Krock carts are lovely, that's why ^^ 20:22 Shara Sorry, have been away. 20:22 Shara Can I just confirm - waterfall crash you were all getting is just because a mod had the wrong light value? 20:24 paramat possibly, not sure anyone has tested the fix #4814 20:24 ShadowBot https://github.com/minetest/minetest/issues/4814 -- Limit light_source in the engine by juhdanad 20:24 Shara Well, I can simply edit the mod on the server. 20:24 Shara I never crashed there either way... but then I never get sent to 0,0,0 for more than a second, and rarely that. 20:24 paramat that won't help, it's a light level stored in the world 20:25 Shara Hmm, okay. Well, either way, the mod shouldn't be setting it that high. 20:25 paramat yeah 20:25 Shara I thought I had fixed weird lighting values from some mods a while back, but I guess I missed it 20:29 Fixer missed soo much text here 20:53 AcidNinjaFWHR does TenPlus1 comes on irc sometimes ? 20:54 Shara Yes, but never saw him on this channel 20:54 AcidNinjaFWHR oh okay 20:54 rubenwardy he doesn't come into this channel 20:55 rubenwardy he's also not online atm 20:55 AcidNinjaFWHR okay thanks rubenwardy. 21:03 juhdanad I made a large waterfall on Red Cat Creative next to the original one, which makes the client crash in some seconds. You can test now #4814 easily, if you have time. 21:03 ShadowBot https://github.com/minetest/minetest/issues/4814 -- Limit light_source in the engine by juhdanad 21:04 Shara I have changed the light value in the mod now as well. For some reason that was set to 18. I assume this won't stop the crashes for now, but it should not have been 18 to begin with. 21:04 Shara Change will be present on next restart. 21:07 juhdanad Shara: this will stop crashes. No nodes on your server have too high light level actually, just the light spreading thinks so. 21:08 Shara Light spreading? 21:08 juhdanad I mean the algorithm that decides how bright a node will be, when you place a light. 21:15 rubenwardy sofar: squashed game#1179 21:15 ShadowBot https://github.com/minetest/minetest_game/issues/1179 -- Add simple fast inventory by rubenwardy 21:16 rubenwardy I also checked that I didn't drop any other elements from the creative formspec 21:17 rubenwardy commit message debatable 21:18 rubenwardy VanessaE, please may you merge https://github.com/minetest-technic/unified_inventory/pull/58 ? 21:19 rubenwardy ~title 21:19 ShadowBot Disable sfinv if installed by rubenwardy · Pull Request #58 · minetest-technic/unified_inventory · GitHub 21:19 VanessaE sfinv? 21:19 VanessaE oh the above simple fast inv 21:19 VanessaE sure. 21:20 VanessaE er.. 21:20 VanessaE I don't have write access to the repo 21:20 rubenwardy I swear you were a maintainer 21:20 rubenwardy ShadowNinja or nore then, please :) 21:21 VanessaE I was at one time 21:27 ShadowNinja rubenwardy: depends.txt conflicts. 21:28 rubenwardy balls :P 21:31 rubenwardy ShadowNinja: fixed 21:32 rubenwardy amazing how neglected game used to be, the rate of commits has increased over the last 2 years 21:33 rubenwardy late response, but I was pleasantly surprised to see we only had 1 page of PR 21:34 ShadowNinja rubenwardy: Er, sfinv isn't merged yet so it should just be left for now, right? 21:36 rubenwardy sure 23:09 paramat rubenwardy and all, any support for game#1408 ? it doesn't conflict with the other cart PRs. i added some text explaining that this is not just a temporary workaround for the bug, there are other good reasons 23:10 ShadowBot https://github.com/minetest/minetest_game/issues/1408 -- Carts: Increase power rail acceleration to help climbs by paramat 23:28 lhofhansl Hi... Any comments on PR 4811? (only send blocks the client can actually see - based on client's view range and fov). With defaults can save 70% of blocks sent to client. 23:30 sofar can I be honest? I'd like to have blocks immediately behind the player sent too 23:31 sofar it really blows that once in a while you turn around to stare at a blue screen because it's all been unloaded 23:31 paramat hi good to see you here 23:32 lhofhansl sfar: Was thinking that too. Makes 4811 even more important. Could optionally remove the server side fov test. (double the blocks, though) 23:32 lhofhansl Hi paramat :) 23:34 paramat #4811 23:34 ShadowBot https://github.com/minetest/minetest/issues/4811 -- Optimize/adjust blocks/ABM sent at the server based on client settings. by lhofhansl 23:40 paramat added a comment 23:41 paramat oh maybe i misunderstood 23:44 lhofhansl answered on PR. ABMs are not subject to FOV (that'd be ... surprising to players). 23:44 paramat ok 23:45 lhofhansl sofar: Currently server FOV is hardcoded to 72 degrees, the PR doesn't add the FOV check, just corrects it. 23:46 lhofhansl (and removes the weird 4.0/3.0 logic because you can't figure out the client's aspect ratio) 23:46 paramat yeah good 23:49 AcidNinjaFWHR so it will save some bandwidth on servers ? 23:49 paramat possibly a lot 23:49 paramat especially with smartphone clients using low view range 23:50 AcidNinjaFWHR nice, go on, even if people turn around as you said, it will probably use less packets :) 23:51 lhofhansl Yes. 70% with default client's view range of 100 and server's max_block_send_distance at 9. Important thing is that it uses client settings to figure this out, so it'll always be "optimal" for the client. 23:53 AcidNinjaFWHR anyone wanna test that? 23:53 Fixer it is not surprising to me about abm :) 23:53 AcidNinjaFWHR do we need to recompile the client as well with your PR integrated lhofhansl ? 23:53 AcidNinjaFWHR or it's just for the server 23:53 Fixer lhofhansl: lbm and timers will work as expected? not affected? 23:54 lhofhansl Yep... Needs a new client so that the information is sent to the server. 23:55 lhofhansl fixer: even if ABM are not processed behind you? (that's not what this PR does, because I think that would be surprising). No mob would be able to sneak up to you from behind :) 23:56 Fixer lhofhansl: no mob will sneak up on me behind when I afk? feels cheaty 23:57 paramat no that's not what the PR does 23:57 lhofhansl fixer: so PR does not change this... ABMs *are* processed behind you. 23:57 paramat ABMs run within the radius of view range, not affected by FOV