Time Nick Message 00:04 Guest38 Hey can someone point me to where the water logic? 00:04 Guest38 Is the implementation identical to Minecraft Java? 00:04 erlehmann no 00:05 sfan5 https://github.com/minetest/minetest/blob/1b2176a426dd987795f20e9042a8d79f958b7b44/src/map.cpp#L509 00:08 Guest38 Thanks i am interested how multiple water blocks flowing interact. 00:10 MTDiscord They.. don't really 00:10 MTDiscord It's a first come first serve mess AFAIK 00:10 Guest38 Ohhh 00:10 MTDiscord so basically liquids of different types are solid for other liquids 00:11 Guest38 In my case both would be flowing water ... 00:11 MTDiscord flowing river water and normal flowing water? 00:11 MTDiscord or the same flowing water? 00:11 Guest38 same flowing water 00:12 MTDiscord oh lolo 00:12 MTDiscord then everything will work as expected 00:13 Guest38 Yeah i am trying to understand what "expected" means. 00:18 Guest38 I aim to describe the stationary behaviour of water using equations. 00:19 MTDiscord well... flowing water has a level and a "flowing downwards" bit 00:19 MTDiscord the level decreases by one with every metre distance to the source 00:19 MTDiscord if there are multiple sources, it gets hairy 00:20 MTDiscord where distance appears to be the manhattan distance 00:20 Guest38 Not quiet manhattan though. 00:20 MTDiscord ? 00:20 MTDiscord why not? 00:21 Guest38 BBBBBB 00:21 Guest38 B2345B 00:21 Guest38 B1BB6B 00:21 Guest38 B0BB78 00:21 Guest38 Where B is a block and number a num/8 water block 00:21 MTDiscord Yes? 00:22 Guest38 If it was manahattan it would be 00:22 Guest38 BBBBBB 00:22 Guest38 B2345B 00:22 Guest38 B3BB6B 00:22 Guest38 B4BB78 00:22 MTDiscord My point is just being the following: 00:22 MTDiscord S12 00:22 MTDiscord 123 00:22 MTDiscord 233 00:22 MTDiscord where S is the source 00:23 Guest38 Yeah that is clear. 00:24 Guest38 778 00:24 Guest38 B7B 00:35 Guest38 If i am not mistaken the complicated parts is here https://github.com/minetest/minetest/blob/1b2176a426dd987795f20e9042a8d79f958b7b44/src/map.cpp#L684 . 00:52 Guest38 Ok complicated looks simple (if i ignore different heights for now). The level of water is simply max_height of it's NESW neighbours minus 1 which eventually converges. 00:52 Guest38 Thank you everyone. 03:05 MTDiscord Do I have to do something special for the CI builds when adding new headers to lib/? 10:42 sfan5 if you're wondering about android then be aware that it doesn't use cmake 13:15 MTDiscord Ah, the Gradle build. I'll ask about it again once I've fixed the MinGW builds. I'm PRing a CMakeLists.txt to the THST repo and I expect that to help with MinGW. 17:05 appguru sfan5: I assume MTG pull request merging policy is "squash and merge"? 17:05 sfan5 always squash unless there's a reason not to 17:07 sfan5 if it's a single commit with a sane commit message like game#2774 you can also rebase 17:07 ShadowBot https://github.com/minetest/minetest_game/issues/2774 -- Balance sounds, remove silent water footstep by paramat 17:55 JordanL2 Does anyone know how I could get in contact with Wuzzy? I have a couple of PRs for two of his mods, and not seeing how I can submit them via the site they're hosted on 17:58 Krock JordanL2: click on any of the "commit" links and you'll see an email in the author row 17:58 JordanL2 Thank you 17:59 Krock I don't think there's a place on repo.or.cz to propose PRs, hence send them a link or the patch file directly 18:01 JordanL2 Yeah, I tried sending a PR on the command line using "git request-pull" but seems it didn't work 18:03 sfan5 FYI 'git request-pull' just prints some formatted text and doesn't do anything, the concept "pull request" you're thinking of exists on Github and comparable sites but is not part of git itself 18:03 Krock also this topic should better be discussed in #minetest or #minetest-hub. this channel is for the minetest organisation projects only 18:08 MTDiscord Would one of the Android build devs update the gradle build on #11973 after I fix MinGW (pushing soon)? 18:08 ShadowBot https://github.com/minetest/minetest/issues/11973 -- WIP: Performance improvements for entity collision by JosiahWI 18:09 erlehmann JordanL2 you can format your patch files using “git format-patch” and send them to wuzzy 18:09 erlehmann they can be applied using “git am” 18:09 erlehmann linux kernel is developed that way partially i think? 18:26 MTDiscord For those interested: you can generate compile instructions for clang tidy by adding -DCMAKE_EXPORT_COMPILE_INSTRUCTIONS to the build. 19:33 sfan5 rubenwardy: I just found out that we could have set android:preserveLegacyExternalStorage="true" to migrate user data without risk or hurry 19:34 sfan5 (https://developer.android.com/training/data-storage/use-cases#migrate-legacy-storage) 20:39 rubenwardy sfan5: I was reading this page which doesn't mention it https://developer.android.com/about/versions/11/privacy/storage#migrate-data-for-scoped-storage 20:39 sfan5 yes 20:40 sfan5 google's documentation is terrible 20:41 sfan5 hm, I guess we could still set it for next release? 21:03 MTDiscord Regarding 11976: I added it to the mainserver, rebuilt, restarted and set postgres_synchronous_commit = false. With ~30 people doing things, no difference between sync and async. 21:15 sfan5 that's weird 21:15 sfan5 maybe the C++ internal profiler can be of help 21:16 sfan5 there's even already a ScopeProfiler sp(g_profiler, "Server: map saving (sum)"); in the interesting part 21:20 MTDiscord I'm not familiar with the engine internals. If you tell me where and what I should add, I'll do so ? I only can't lag out all those people again, the effect will be more subtle when I set the map saving interval to 30, I guess 21:22 sfan5 it's profiler_print_interval = 60 in the config I think 21:45 MTDiscord This prints the profiler data to a preset folder every 60 seconds? Isn't that entirely lua stuff? 21:46 MTDiscord The database lag seems to not show up there. 21:50 sfan5 that's separate from the Lua profiler 21:51 sfan5 though I can't get it to do anything locally 21:52 sfan5 ah 21:52 sfan5 it prints to infostream so you need --verbose too 21:54 sfan5 2022-01-23 22:53:01: INFO[Main]: Server: map saving (sum) [ms] . . . . . . . 1x 26 21:54 sfan5 ^ you get something like that 21:54 sfan5 and also other detailed debug info about how many blocks MT saved and for what reasons 22:06 MTDiscord So I don't need to add the line above (ScopeProfiler sp(g_profiler, "Server: map saving (sum)");), it's already in. I only need to turn on verbose mode and watch out for that line? 22:07 sfan5 yes 22:07 sfan5 and the setting obviously 22:08 MTDiscord Can both be set from a running server? 22:10 sfan5 profiler_print_interval definitely can't 22:15 MTDiscord Ok, thanks. I'll do that upon next server restart and report back. Should we somehow document chatter somewhere or solutions and if so: where?