Time Nick Message 13:29 paramat game#2047 13:29 ShadowBot https://github.com/minetest/minetest_game/issues/2047 -- Ores: Remove overlapping regions. Make some regions deeper by paramat 14:40 paramat #7064 #7065 are trivial will merge in 30 mins 14:40 ShadowBot https://github.com/minetest/minetest/issues/7064 -- Settingtypes.txt: Fix valleys dungeon ymax error by paramat 14:40 ShadowBot https://github.com/minetest/minetest/issues/7065 -- Find nodes in area (under air): Raise volume limit to 4,096,000 by paramat 15:51 paramat now merging those 2 15:59 paramat merged 18:08 paramat Krock and MTG devs game#2047 is updated and ready 18:08 ShadowBot https://github.com/minetest/minetest_game/issues/2047 -- Ores: Remove overlapping regions. Make some regions deeper by paramat 18:11 Krock concept is fine. will test 18:42 Krock paramat, increasing the chance for the ores to be generated in exchange for no overlaps works quite good. but it also removes a bit variety of different ore "sizes" 18:43 Krock the amount of the ores seems to be about the same as before for each depth 18:43 paramat yes and yes, calculated very carefully to not change ore amount in the lower region 18:44 paramat no other ores have overlapping regions 18:44 Shara To me it seems like one minor plus, one minor negative 18:45 paramat with the amount of ores we now have i don't think it's a loss 18:45 Shara I'm a tiny bit against it, because I do like the current variety, and it does make a difference when playing 18:45 Shara But not strongly against it 18:46 paramat it's partly for consistency, also for simplicity and performance 18:46 Shara How big a performance difference does it really make? 18:46 paramat small 18:47 Shara I'm just struggling with that as a justification because I can't imagine it being enough to matter 18:47 paramat maybe enough to allow adding a new ore sometime 18:47 Shara There's no use for the current ones 18:47 Shara not enough uses* 18:48 paramat hm well, as i often argue, small performance gains add up to the lightweight MT we have now, so small is still significant 18:48 paramat yeah i'm not suggesting adding a new one, needs a lot of justification 18:49 Shara Underground needs different stone types, decorations and so on, to stop it being a repetitive land of bordom :) 18:50 Shara And I don't see that happening soon, which is why I don't really like removing that minute bit of variation from... well, anything 18:52 paramat more underground variation will happen, probably before 0.5 18:52 Shara Well, sofar seems to be suggesting that needs an engine change 18:53 paramat oh the underground biome shape thing, don't worry about that 18:54 Shara And if we want weirdly shaped wobbley-edged underground biome blobs, we do so... 18:54 sofar it would be a mapgen variation 18:54 Shara Just need someone to figure out what approach we do want 18:54 sofar I wouldn't make it biome based, though 18:54 Shara Idealy they wouldn't function like surface biomes at all 18:54 sofar climate and structural geology bear no relationship 18:55 Shara Of course not, but it's also about what's feasible. 18:55 paramat underground biomes have the shape of biomes, biomes have y limits plus a new vertical biome blend. the shape of underground biomes depends on what we do with biome y limits 18:55 sofar underground projection of surface climate biomes is ridiculous imho 18:55 Shara I am assuming the blending on y limit can be utilised somewhat here 18:55 sofar from a geological perspective, at least 18:56 sofar it makes sense in MC, where max depth = 64 nodes 18:56 sofar but it makes no sense in Minetest 18:56 Shara sofar: how much attention have you paid to... let's say v7 mapgen? 18:56 Shara :D 18:56 Shara I think we lost geological realism long ago 18:56 sofar not too late to fix it 18:56 sofar I don't want realism per se 18:57 sofar I'd make a lua mapgen if I wanted realism 18:57 sofar it would just be nice if the underground could use 3d noises for variation 18:57 sofar nothing complex 18:57 Shara I'd love the ability to do this the way you are thinking, but also to set... not sure how best to describe. Some specific underground biomes that would be locked to match surface ones 18:57 Shara But that would mainly be for ideas I have for non-MTG things 18:57 sofar the wouldn't be biomes 18:58 sofar just make underground rock organized to some 3d noise and put ores in specific stone types 18:58 sofar that would already suffice 18:58 sofar it can be really simple 18:58 Shara Yes, but then you lose the ability to set actual "biomes" underground completely? 18:59 sofar or just adjust chances for each ore so they are rare in one type of base rock, and more abundant in another 18:59 Shara areas of different base rock would be my current main aim to be honest 18:59 sofar effectively you'd make geological biomes? it just wouldn't use the biome system 18:59 Shara Then link decoration to rock type 18:59 sofar that's all I'm saying 18:59 paramat 3d noise is complex, and very intensive, more than 1 is not practical 19:00 sofar well, try it first, then decide based on the performance data 19:00 sofar it could be entirely reasonable for new hardware in the future 19:00 sofar or acceptable to people who can affoard a relatively new pc 19:00 paramat i know how intensive 3d noise is 19:00 sofar dismissing something based on projected performance characteristics is always premature 19:01 sofar maybe one day we'll have opensimplex and it will be much faster 19:01 sofar we know you're terrified of breaking people's 2005 CPUs 19:01 Shara My ideal would be that we get a mix of stone types unless something specific is set for some particular depth (example: fire caverns deep beneath a volcano biome) 19:01 paramat we have to think about low power devices and being lightweight in MTG. other games can do the intensive stuff 19:01 sofar make it a new mapgen 19:02 sofar problem solved 19:02 sofar your concern is noted. stop roadblocking it. 19:02 Shara New mapgen doesn't really solve the actual issue. :) 19:02 Shara Being more underground content for MTG... as in the game itself, not some specific mapgen 19:02 sofar enjoy your flat underground rock layers? 19:02 paramat as i said, i know how intensive 3d noise is, you can't say it's gueswork 19:03 Shara Technically we already have flat underground layers.. just only one. 19:03 paramat MTG has to run in every mapgen, it can't be a new mapgen 19:03 Shara Surface does have different stone types 19:03 Shara Then they just stop 19:04 paramat however .. 19:06 paramat using biomes for underground stone is preferable as biomes allow definable stone types, and biomes can dictate ores and decorations 19:07 Shara In all honesty... Are players going to care more about "oh no, this is geologically inaccurate!" or about "wow, we have actual variation now!" 19:07 Shara ? 19:07 paramat however we could add a 2d noise to vary all the y limits to avoid flat tops and bases. it's something i've been thinking about 19:08 Shara I'd love to get both right, but if he just parts the channel after being so negative, without offering to help... 19:08 paramat so possibly we can use biomes but with some added feature 19:08 Shara Yes, that's why I kept on about good blending on y for so long 19:09 Shara I do wonder how many extra biome registration would be needed though if you want lot sof depth variation 19:11 paramat also, such a 2d noise biome y limit variation would be useful for surface biomes too 19:13 Shara No doubt some will disagree with me, but I very much feel underground content is one of th ebiggest failings of MTG 19:13 paramat (hm, and i'm not roadblocking anything, just having suitable concern about the performance of an idealistic noise-heavy underground) 19:13 paramat yeah 19:14 Shara I don't know how we can even keep "mine" in the name of the game given how little reason there is to do it 19:14 Shara everything worth seeing is on the surface 19:15 paramat we have a 'stratum' ore type that can create noise-varied strata as you know, i coded it. problem is it needs a 1 or 2 2D noises for every strata, and underground there could be very many 19:16 Shara Maybe first thing is to work out how many 19:16 paramat then there is the problem of the stone volumes themselves being essentially 'ores', and placing ores in ores. so using the biome system is preferable, it' s designed for placing different stone types 19:17 Shara The onlt time sudden cut offs on y will be visible really is in caves 19:17 Shara only* 19:17 Shara So IF the y blending is good enough, it might be a non-issue 19:18 paramat actually i need to test if vertical blend workd underground, i hope so 19:18 Shara It's actually one of the things I had in mind when we first talked about it :) 19:19 paramat however, horizontal borders wouldn't bother me, with everything else being crazy 3d underground horizontal lines could actually be a nice balance 19:20 Shara Hmm, not so sure there 19:20 Shara It looked terrible when CRL started splitting biomes on mapblock edges 19:21 Shara (not that they were really biomes... but you get the meaning) 19:23 paramat yes, hopefully we can somehow avoid straightness in some simple lightweight way 19:24 Shara Straight edges in general are bad :) 19:28 paramat testing now 19:32 paramat nah doesn't work underground 19:32 Shara Ugh 19:33 Shara Any easy way to make it work underground? :) 19:33 paramat however the dithered type of blend probably wouldn't look right, a clean edge is better 19:33 paramat not sure 19:33 paramat i think i'd prefer to add a 2d noise to vary the border 19:33 Shara It could be better 19:34 paramat so will think on what can be done for underground biomes 19:36 Shara Okay. If you want to talk through ideas or need help with any testing, let me know 19:37 paramat certainly 19:38 paramat i'll try the 2d noise thing 19:40 Shara :) 19:40 paramat anyone know where the background RGBA for main menu tabs is set? 19:41 Shara Heh, was just reading issue and wondering that myself 19:41 Shara With all the little changes I made there, you'd think I'd know 19:42 p_gimeno isn't it where it says to bug sfan5 to add more colours or something like that? 19:45 paramat ok i know why vertical blend doesn't work underground, and it's difficult and intensive to enable. however, it would be unsuitable for stone borders anyway, will think about something new for that 19:49 rubenwardy paramat, irrlicht 19:49 rubenwardy you'll need to change the skin in GUIFormspecMenu 19:49 rubenwardy unless we're using our own tabs implementation 19:53 paramat ok looking, i did look at that file but could have missed it