Time Nick Message 06:36 Krock will merge #12540, #12632 and #12682 in 15 minutes 06:36 ShadowBot https://github.com/minetest/minetest/issues/12540 -- Textures: introduce world-align overrides by SmallJoker 06:36 ShadowBot https://github.com/minetest/minetest/issues/12632 -- Remove default keybind for range select by fluxionary 06:36 ShadowBot https://github.com/minetest/minetest/issues/12682 -- Allow buffer argument to VoxelManip:get_light_data by TurkeyMcMac 06:37 Krock celeron55: 1,691 src/session.vim.meshgen_cache this file should probably not be there, right? 06:51 Krock merging 09:21 celeron55 Krock: yeah i need to clean it all up, there's all kind of crap still 09:22 celeron55 kinds* 09:25 celeron55 literally the first line shown in "files changed" on github is trash 09:27 celeron55 one thing i'm not sure about is whether this fits the current roadmap. I guess it's close enough given it's meant to support a reduced-level-of-detail rendering method 10:06 Zughy[m] I didn't know that if people block a PR with a discussion, that is a valid reason to remove an approval and also put a possible close, when no reasons are provided: #12417. This is fucking ridiculous 10:06 Zughy[m] That PR is a compromise to make everyone happy, so why exactly I have to lose all my ground because of that? 10:06 ShadowBot https://github.com/minetest/minetest/issues/12417 -- Docs: add "flip moon texture" into breakage file by Zughy 10:07 Zughy[m] I already had to deal with a lot of bikeshedding/erle-like arguments in the prior PR, I don't think I deserve this treatment 10:09 Zughy[m] also because I am the one who improved the sky API, because of the fact I actually I had had to work on it. Contrary to them 10:10 Zughy[m] *I actually had 10:12 celeron55 i don't see the big deal, isn't it just rubenwardy adding an approval and then removing his approval. the possible close is weird without more comments though. but rubenwardy is right in that too much time has been used for that issue already 10:13 celeron55 s/\./?/ 10:13 Zughy[m] yes, and it's been used because of erle poisoning discussions with arguments that made no sense. Plus two more users jumping on the like-wagon 10:14 Zughy[m] This passes the message that if I act like erle on whatever PR I don't like, there is actually a chance to see it closed because of me 10:15 celeron55 i'm not thinking this in terms of erle's arguments or user likes. i barely read those to begin with 10:16 Zughy[m] lucky you 10:17 celeron55 i could make an argument about this from the "i made it this way and it made sense" point of view, but that would mean i'd continue the discussion i don't want to be continued to save time from everyone 10:17 celeron55 save? i mean waste 10:17 celeron55 wait 10:17 celeron55 well you get the point 10:18 sfan5 it will be reassesed anyway when the time actually comes so we can just add it to the file 10:18 celeron55 https://github.com/minetest/minetest/pull/11902#issuecomment-1008064540 10:19 Zughy[m] what sfan has said 10:19 celeron55 i think this comment from me should enable you to look it from a far enough standpoint 10:19 celeron55 i'm fine with adding it to the file. didn't mention that yet. but also fine with not adding 8) 10:21 celeron55 anyway, for that reason i'll go cleaning up my rebase instead 10:22 celeron55 i have an unhealthy relationship with files that have lists of things to do 10:23 celeron55 in my use, those get completely out of hand and become almost useless as there is not enough time in a human lifetime to do everything that ends up there 10:26 Zughy[m] yes, that's why I've opted for 6.0. It's a silly thing so there is no hurry to add it now (the PR you've linked shows that we've already had enough chatting about compatibility). But I think that, for the sake of modder experience, having the two celestial bodies acting in the same way is an improvement (why exactly if I apply the same texture to both the bodies, when these rise one of them is upside down?). And I rest my case, I don't 10:26 Zughy[m] wanna fill the chat nor make anyone waste more time 10:30 celeron55 because you asked it as a question, my short explanation as the one who originally wrote it is: because when both of them are visible at the same time, they are oriented the same way when you yaw around and look at them. but this has been already discussed in the long PR so whatever 10:32 Zughy[m] 16 one approval PRs with 87 PRs open, we're potentially facing a new record of least PRs open in years. N I C E 10:33 Zughy[m] issues are doomed though 10:36 celeron55 sfan5: do you want my approval to merge 12417? 10:36 celeron55 because if i comment "i'm fine either way" that's considered an approval 10:37 celeron55 well whatever, did it anyway 11:08 Zughy[m] was there an issue about freeing more AUX keys? I can't find it 11:13 Krock yes I remember there's one 11:13 Krock #11446 Zughy[m] 11:13 ShadowBot https://github.com/minetest/minetest/issues/11446 -- Add Aux2 key for games to use freely, and maybe Aux3, Aux4, etc.? 11:51 appguru Zughy: The problem with your PR still is that there are no good reasons for the change (vs. the reasons against the change) - as ruben said, he's fine "either way" - and making the change is always more effort than leaving it as is, thus ruben should have closed the PR. 11:56 Baytuch gday, MineTest developers 12:04 celeron55 Minetest 12:04 celeron55 Copyright (C) 2010-2022 celeron55, Perttu Ahola 12:04 celeron55 holy shit that's a long time 12:05 celeron55 i'm just making a new file and using the opportunity 12:05 celeron55 new file based partially on old files 13:30 Zughy[m] Is #1414 still true after the huge particle PR we've seen in 5.6? 13:30 ShadowBot https://github.com/minetest/minetest/issues/1414 -- Particles are very slow 13:39 Krock paricles were made fancier, not faster. 13:49 rubenwardy unfortunately 14:13 Fixer I reproduce https://github.com/minetest/minetest/issues/10844 shader related error, ask me anything 14:13 Fixer another question: how taxing are shadows? I've turned them to very low and got 3 fps o_o 14:15 rubenwardy merging #12678 and #12667 14:15 ShadowBot https://github.com/minetest/minetest/issues/12678 -- Reassure previous `nil` behaviour for `tiles` and `special_tiles` by Zughy 14:15 ShadowBot https://github.com/minetest/minetest/issues/12667 -- Check hp_max > 0 for entities by appgurueu 14:15 rubenwardy in 10 14:15 rubenwardy Fixer: quite, you'll need a recent GPU 14:15 Fixer i have to say minetest is waaaaay moooore smoother these days, silky smooth 14:15 Fixer ah, my Intel HD gives me 3 fps, but very solid fps with it turned off 14:15 rubenwardy shadows are an optional feature very much not for lower end or older hardware 14:16 rubenwardy I get decent FPS with shadows on my Intel iGPU 14:16 Fixer i5-3450 (almost 10 yr old cpu) 14:18 Krock pretty sure this is limited to Intel HD every time 14:19 Krock send a memo to their headquarters to finally fix their buggy drivers 14:23 Fixer i'm still waiting for proper rx6400 price 14:25 Krock you could as well get a used GTX 16XX or RX4XX / RX5XX card which still hold well nowadays 14:26 Zughy[m] is it random that all the people who reported the bug are on Windows? 14:28 Krock give my laptop a few minutes and I'll tell you if the same issue happens on Linux 14:32 Krock 9 FPS on Intel HD 4000 with highest shadow setting. no shader error, this is the Windows driver's fault 14:33 Krock correction: it's just set to "High" and not "Very High" 14:34 Zughy[m] I'll label it 14:34 Krock 60 FPS with all shaders enabled, and shadows to "Very Low" 14:36 Krock that's at range 100. 61 FPS without vsync. that's the most I can get out from this 3rd gen i7 14:38 Krock updated my post accordingly too, for the sake of completeness. 15:13 Krock hmm. interestingly on Linux it reports OpenGL 3.0 whereas TechPowerUp lists 4.0 as supported version 15:13 Krock presumably the additional OpenGL extensions are flawed 15:14 Krock Fixer: if you can figure out how to force OpenGL 2.1 or 3.0 on your Intel HD iGPU. please report if anything changes 15:15 Fixer I don't know if this is possible, maybe some irrlicht option? 15:15 Krock on Linux it can be done using an environment variable... but that's Mesa3d-specific 15:16 Krock no idea though how to do that on Windows. hopefully Google knows more 15:28 Fixer it probably requires editing mt sources 15:30 Fixer interestingly i don't have this error on 5.5.1 with same confug 15:32 Fixer updated my post: so something was changed in those 3 months that causes this error on my system, funny enough original user had this same error on 5.4.0 15:33 Fixer 5.4.0-6d7067f to be precise 16:08 MTDiscord Any objections to fixing android build for x86? 16:17 sfan5 the x86 build probably already works just isn't enabled by default 16:18 MTDiscord nope, it doesn't, it has linker problems 16:18 sfan5 we can fix that then 16:18 MTDiscord okay, I almost have a fix 18:44 MTDiscord Just added #12692 which improves performance of shadow-enabled worlds significantly. It's tiny. 18:44 ShadowBot https://github.com/minetest/minetest/issues/12692 -- Limit force shadow update to urgent blocks by x2048 18:45 rubenwardy do you observe an FPS improvement? 18:46 MTDiscord from 25 to 50+ fps. especially with longer view ranges. 18:46 rubenwardy nice 18:47 MTDiscord the problem was, I force full SM update every time a mesh arrived, but 99% of them are distance meshes, and it happens on every move. 18:47 MTDiscord the change is to force only when an 'urgent' mesh arrives, like digging or building. 20:31 MTDiscord Merging #12679 in a min 20:31 ShadowBot https://github.com/minetest/minetest/issues/12679 -- Reduce the use of porting::getTimeMs() when rendering frames by x2048 20:34 MTDiscord Merged 21:24 Zughy[m] sfan5: should I close #12680 ? 21:24 ShadowBot https://github.com/minetest/minetest/issues/12680 -- fix: update Gitlab CI by cat-master21 21:29 sfan5 eh, we can wait for them to do it themselves 21:35 sfan5 can anyone look at #12616 21:35 ShadowBot https://github.com/minetest/minetest/issues/12616 -- Move some CI jobs to newer compiler versions by sfan5 21:38 rubenwardy what's our minimum compiler requirements? 21:38 schwarzwald[m] They're listed in the README.md 21:39 rubenwardy surely that's outdated 21:40 sfan5 no its accurate 21:40 sfan5 just not regularly tested 21:41 rubenwardy well, given that the compilers aren't the min already then +1. But it's probably worth having CI checks on the minimum compiler versions as most devs will be using something newer 21:41 rubenwardy for building packages, using the most recent compilers makes the most sense 21:42 schwarzwald[m] The last time I tried to test the minimum GCC version there was a dependency with a version I couldn't get IIRC. I had to install the old version of the compiler on a newer system before it worked. I think we may have bumped the version since then, though. 22:12 sfan5 merging #12616 in 5m 22:12 ShadowBot https://github.com/minetest/minetest/issues/12616 -- Move some CI jobs to newer compiler versions by sfan5