I quit Minecraft about 6 years ago because of some computer overuse injuries. Yes, Minecraft was a major contributor. By the time I was recovered enough to be able to play Minecraft. there were other demands on my time so I just left it aside. I could have still done modding, but without being able to play the motivation was totally gone. Plus, of course, people change over time, along with their interests. I'm not really Zeno410 anymore - more like Zeno410 plus 10.
However, right now I'm largely confined to the couch for a few days to weeks due to some knee surgery so I installed Minecraft to have a go at it. I'm going to write about a few brief worlds and then do a journal of sorts on a 1.12 modded world (I'll talk about that choice later).
"Why a journal?" I hear you ask. "There are very few users on the forums anymore and nobody does journals now." Well, yeah. Probably almost nobody will read this. But -
A long time ago, (in what now seems like a faraway galaxy) I told somebody here to journal for yourself, because someday in the future you'll look at your old adventures and enjoy. And, oh boy, that is *SOOOOOOO* true. I had some downtime visiting family recently and went through the journals I wrote when I was active here and I had a BLAST. It brought back all the excitement, the satisfaction, the amazement, and the ironies from (sometimes) almost a decade ago. And, in the process, I realized from some offhand comments I made that there were other worlds I had almost completely forgotten, now existing only in those offhand comments. I remember just snippets from earlier worlds and sometimes I don't even remember which went together. So the journal worlds live on in my head, entertaining and inspiring me, while the non-journaled worlds are "Gone with the End" and might as well have never happened.
So that's why I'm journaling. For me - or more precisely, the me of the the future, Zeno410 plus 20, who hopefully will enjoy this as much as I have Zeno410's journals. Anybody who wants to follow along, welcome! I hope you enjoy it too, and maybe get inspired to write, record, or whatever for *your* plus 10 someday. But even if nobody does, that's fine, because we Zenos-of-different-times are going to stick together.
The last time I played Minecraft was is 1.12 (and my last journal is from 1.10). There have been a lot of changes since then, to put it mildly, including a complete overhaul of world generation, my main modding interest, in 1.18. So what does Minecraft generation look like these days? Time to find out!
It had been so long I'd even forgotten the keybindings and basic recipes. So, since I was primarily playing just to get some maps and look at the generation, I decided to play on Peaceful. An overreaction, but oh well.
I started in a Birch Forest in a hilly area. My screenshotting choice skills are as gone as my playing skills so I completely forgot the establishing shot, so you'll have to settle for this much later pic while returning from an exploration:
I had started in the hilly area on the right shoulder of that big mountain. And when I started, I didn't even realize I was next to a big mountain, because I couldn't see it. I chopped some trees, and then looked for a starting base area. I *did* remember you want your first base as low as possible so there is less to mine through to get to the low areas where the good ores are, so I went downslope until I came to a "river" although it was actually just a mix of land and water you couldn't boat through. There I set up my standard starter base: a small house, 3 blocks wide on the interior with 2 high wall, basic starter stuff built *into* the wall, a door with a hole in front of the door, and a "roof" of just a 1x1 rafter in the middle.
Again, no contemporary pic, so here's one from much later:
I find this a super-practical starter house; you can build it from dirt, saving valuable starter wood; no mob can get in; zombies can't break in the door due to the drop; if a spider goes up on the roof you can kill it from below; you can pillar up in the middle and kill creepers at the wall in the morning by swatting their heads; crafting table, furnace, etc., can be accessed from inside or outside the house, and if you put a dropshaft under the open parts of the roof skylight goes down so you can mine (not too far because you only suppress spawns in a 24 block radius) even without torches.
Then I did all the routine starter stuff: get the farms going; ranch some animals; mine down to Y11 for resources (I only later found it's not so simple anymore); and in my case grow a lot of cane because I want mega maps. I'll skip the details because it's not relevant - plus it's often embarrassing what I had forgotten to do.
Once I had a maxed out map, I did a circuit of it as best as I could, then grew some more cane, made another map, and did a circuit of *that*.
To establish what I'm looking for, I have 4 main complaints with the old vanilla minecraft generation from 1.7 to 1.17:
1) Search distances are absurd. Obscenely absurd. If you are searching honestly, (no AMIDST cheats, etc.) it normally takes TENS of THOUSANDS of blocks searched to find both hot and cold terrain. Almost nobody does that much searching. Even I don't, because on pre-1.18 vanilla generation the shortage of water and flat plains makes it hard to travel long distances. For rare biomes it's even crazier; in my first 1.7 world I later ran AMIDST and found the nearest Ice Spikes was 9,000 blocks from start, which would have taken two hundred and fifty THOUSAND blocks of an efficient but honest square spiral search. Ab-surd.
2) No large scale structure. Big maps are just a big mess. The "oceans" are just large lakes, the mountains don't come in ranges, and the hot and cold biomes are rare in land areas.
3) You can't see much. Most terrain is closed forest, and there aren't many clearings or water areas to provide relief. Plus the trees are mostly short and shrubby which puts the foliage in your line of sight.
4) There isn't much *to* see. Most terrain is fairly flat. Hills mostly occur as sub-biomes restricted to the interior of the mostly-forested biomes so you can't see them (Forest Hills in plains are sometimes a delightful exception). Trees are mostly smallish and bland. In the terrains with relief like Extreme Hills there is often seriously crazified terrain which I know some people like but which I find ridiculous.
So, how does the newer system do on these? Well, here's the resulting map:
1) Search distances:
Basically fixed! The first map I did (on the right) had *both* hot and cold terrain on it. Sample size of 1, etc., but on this map a typical player *would* find both kinds of terrain. This would be virtually impossible on the old system. Plus, even in this search I found several of the new biomes, including Cold Ocean, Sparse Jungle, Badlands, and Mangrove Swamp. This is about as good a result as I got with Geographicraft/Climate Control.
2) Large Scale structure:
Not great, but substantially better. There's a hot to cold transition on this area - hot on the left, and cold on the right. Plus, although you can't see it on the map, the ocean wraps around on the right so the entire right map is of a penisula. So there's a north cold penisula, a south cold peninsula, an ice bay inbetween, the hot desert to the right; different regions are actually *different* in a meaningful way and traveling around can feel like you're going somewhere and not just "another forest, another forest, another forest".
I also spotted ocean on all 3 outside sides of the left map so there's probably some sea/land structure there as well, although you'd need more maps to see it.
3) You can't see anything:
Better, but still needs some work in the temperate zone. Hot zone is mostly open terrain (desert/badlands/sparse jungle) so you can see pretty well there and I got a number of great vistas. The temperate zones have these "old growth" areas with bigger and (sparser?) trees and I found it somewhat easier to get views there. And the new wider rivers really help too:
but, not as much as I'd like. I still spent a lot of time in the temperate zone and other forests looking at trees and unable to see anything much. Some of this is good, because contrast is good; when you spend *some* time struggling through a tree maze it makes a grand vista even sweeter when you get it; but at least in the temperate zone it's too much. The fact that I didn't notice that grand mountain when I started just proves the point.
4) Not much to see:
Also much better:
The new terrain system generates a good number of nice attractive hills and mountains *without* the impossible overhangs and floating dirt that used to be common in rugged terrain. The trees are a bit better too, with the larger trees in old growth, although still no match for trees from the mods like BoP, Highlands, and RTG.
There is one newish issue: excessive small-scale ruggedness. I found it surprisingly difficult to get around in my river base because of 2-high jumps; this became particularly noticeable when I was struggling to lead animals into my pens. This is even worse in mountains, which sometimes I found nearly impossible to cross quickly (without digging passes, etc.); that might be a design intention to make mountains more of a barrier, while ruggedness in flatish areas is just a pain.
So, overall, I'd say a B+ result. Not everything is fixed to my taste but everything is substantially improved. I get a sense of a professionalism that was lacking in the 1.7 revisions; they looked at complaints, devised way to address them, and then made sure they actually worked without creating (big) new problems. I also get a sense they've been looking at mods: I see stuff that smells like Alternate Terrain Generation, Realistic Terrain Generation, Geographicraft, and Biomes O Plenty - not in the sense of stealing stuff, but looking at possibilities, design options, etc.
The current professionalism contrasts to the amateur hour of the 1.7 changes. My (un)favorite example was the land-sea code. The #1 land-sea complaint for pre-1.7 was ridiculously long ocean trips; it was not unusual that if you just set off in a random direction from a continent you would spend hours of real time travelling before you saw land. So they repeated a step that doubled the size of the continents. But said step also DOUBLED THE SIZE OF THE OCEANS! Yargh! Look at the code, fellas! So then, apparently in desperation, they filled the ocean with large islands, so thickly they all join together and form a universal continent. This would have worked pretty well, actually, if they'd just had about half as many islands, but that would have taken a few hours of thought and a couple days of testing and I guess that was too much? I get the impression they were horribly understaffed and the changes were super-rushed. It would all be rather funny if it hadn't become the standard generation system for 8 years.
If I were just a player, I'd probably carry on in 1.19. It's not to my taste in everything, but it's pretty good. But, I'm not just a player, I'm a player/modder, and I miss the extreme customizability I had designed into my worldgen mods. I'm also on something of a nostalgia kick too, so I want to experience the playstyle I had designed. So, I decided to put this world aside and start up a modded 1.12 world.
Next: some experiments don't come out too well.
Rollback Post to RevisionRollBack
Geographicraft (formerly Climate Control) - Control climate, ocean, and land sizes; stop chunk walls; put modded biomes into Default worlds, and more!
I'm wondering if you got a good particular world. While I want to say the biome variety is better now than it was since 1.7, and perhaps it is, the climate zone still feels like it could be less strict.
As per your mention of the views and forests, I've felt for a LONG time that the forests, and specifically the trees that make them up, need overhauled. I feel like all the oak trees except for large oak trees should perhaps go away, and that they should also get larger at the a maximum. That would be a start to making some better forests. Keep some (maybe 5%?) of the smaller ones (still with the canopy starting a minimum of 5 or 6 blocks off the ground) for a bit of variety and to add some lower coverage here and there, but not enough to make forests like they are now. Right now, most "trees" are actually overgrown large bushes that start one to three blocks from the ground... those are NOT trees and make for terrible forests!
The downside to such a change might be people complaining they can't just swiftly gather wood without leaving ground level, and those who want a "natural" looking forest might find they need to do more manual work after clearing some trees (I would fall into this myself). And the whole two or three block tall tree is somewhat iconic and people might complain. But I'd still find it worth it. These old trees/numerous forests are a big detractor to modern Minecraft world/biome generation in my opinion.
As you found out, the variants of birch forests with taller trees are an improvement, but instead of adding variants that differed in just one or two ways, they should have just overhauled all of them. Birch forests possibly were going to get that treatment in 1.19 but they didn't in the end, and many of us were left sad. On the other hand, maybe it's for the best as what forests need aren't one being touched up, but the bigger subject of forests as a whole being looked at. I feel like there's too many forest variants and it's part of why forests in particular are so common. Like, don't get me wrong, I love finding a forest full of flowers. But you have probably dozens of types of forests, some just being minor differences (birch forest and "old growth" birch forest, etc.), when instead they should have narrowed them down to a few main types. Maybe get rid of flower forests entirely and just have the "flower biome" a separate overlay that can happen rarely in forest or plains.
But yeah, world generation is vastly improved (and if it hasn't happened yet, wait until you find mountains and/or start going underground). But there's still so much that could be done.
Also, wondering if you found any of the new animals yet. Llamas are the best.
As per your mention of the views and forests, I've felt for a LONG time that the forests, and specifically the trees that make them up, need overhauled. I feel like all the oak trees except for large oak trees should perhaps go away, and that they should also get larger at the a maximum. That would be a start to making some better forests. Keep some (maybe 5%?) of the smaller ones (still with the canopy starting a minimum of 5 or 6 blocks off the ground) for a bit of variety and to add some lower coverage here and there, but not enough to make forests like they are now. Right now, most "trees" are actually overgrown large bushes that start one to three blocks from the ground... those are NOT trees and make for terrible forests!
The downside to such a change might be people complaining they can't just swiftly gather wood without leaving ground level, and those who want a "natural" looking forest might find they need to do more manual work after clearing some trees (I would fall into this myself). And the whole two or three block tall tree is somewhat iconic and people might complain. But I'd still find it worth it. These old trees/numerous forests are a big detractor to modern Minecraft world/biome generation in my opinion.
But yeah, world generation is vastly improved (and if it hasn't happened yet, wait until you find mountains and/or start going underground). But there's still so much that could be done.
Also, wondering if you found any of the new animals yet. Llamas are the best.
I've played a lot on modded worldgen, where bigger trees are the norm, and I've got quite a bag of tricks for chopping bigger trees. I find it part of the fun to come up with ways to chop these trickier trees. We all talk about zillions of ways to get down to mining level and lay out our mines, deal with lava and cave breakouts, etc. and we all think *that's* part of the fun. Why couldn't Mojang treat forestry the same?
But even if Mojang doesn't want people to work to chop trees, there's a lot of room for improvements in your direction. My lowest level trick is to chop out the 2nd and 3rd trunk block, stand on the stump, and chop straight up from there. That will take out a 9 high trunk. In addition, foliage could be higher relative to the trees (another trick the mods use a lot). A 9 high trunk with foliage going from 5 to 11 would be no problem at all visually, allow much better views, and yet still allow players to chop down trees without any more work than stepping on stumps. A very workable compromise, IMO.
One last thing to add is that having a wide range of trees would allow more mixing. There could be old growth forest where they're all 9 high trunks, mixed growth with some high and some low, and they could still have some "scrub" forests like the current ones.
Also; yes, I've seen some of the new caves and I think they're great. I just don't have much to add on that because I didn't do that kind of modding. The one thing I don't like about them is that there's a lot more smallish holes in the ground leading to long falls, and that's a problem for Hardcore, my favorite kind of play. But they're interesting and look nice.
I did see some llamas with a traveling merchant and they are indeed cute! Haven't interacted with the yet; will try to in my next go-round.
Rollback Post to RevisionRollBack
Geographicraft (formerly Climate Control) - Control climate, ocean, and land sizes; stop chunk walls; put modded biomes into Default worlds, and more!
You're referring to your past self as a different person? I think that happens when we spend too much online/on computers indeed, I've caught myself doing it a couple times in the last couple years and I sure know I didn't before covid.
Also, forestry is cool in vanilla with how I do various spirals to get up to trees, but I agree more could be done which is why so many mods revise trees and chopping. Although some plug ins just do a cop out and make the whole tree fall when hitting its base.
I had been wondering what happened to you (and so many others); I'd always been partial to modded survival journals, especially given that you'd been making your own mods to shape the game to your own likings, because that is exactly what I've been doing (I'd always wondered, was your "Climate Control" mod inspired by my own mods? I know that you made a Forge version of my "Random Biomes" mod).
Also, as far as 1.18 goes while I haven't played on it (or even touched anything past 1.13, when I stopped updating a mod that reverted the changes to caves in 1.7 due to a combination of lack of interest/demand and difficulty in updating it after code changes) from what I've seen it is no better than 1.7 in terms of local-scale biome variety, where to me local means on the order of hundreds of blocks (even a level 3 map takes about 1 1/2 months to explore, of course, as always, by exploring interconnected caves underground, which gives a rate of about 100 chunks explored per play session/day).
For example, this is a rendering of my current modded world, which I've been playing on for more than a year, overlayed on top of a random seed in 1.18 (ChunkBase); the size of some of the climate areas is just crazy, biome borders also look way too smooth, compared to the random shapes in the old generation:
The same goes for the underground; I have the impression that it is dominated by giant open caves with a sprinkling of the "old" caves, probably still based on 1.7 generation, and ChunkBase shows that ravines, which had the changes in 1.7, were actually made rarer, with mineshafts being the same as 1.7 and later (1.13 did remove a restriction that made them less common closer to the origin) - I;d expect the opposite given that the ground is twice as deep. For comparison, I've added over a dozen variants of caves, along with more variation in the size and structure of mineshafts and ravines, and new types of dungeons, and actual biome-specific underground (e.g. deserts replace stone-based blocks with sandstone and sand, Ice Plains Spikes has snow and packed ice, etc. Mobs like husks and strays also naturally spawn underground, including more variants, like red husks and silverfish in Mesa and Badlands):
This is the largest cave near the bottom, with a volume of more than a million blocks:
Another cave I found which is even larger:
Another large cave under a desert, showing the biome-specific underground:
A ravine with a volume of more than half a million blocks:
"ice" biomes have water instead of lava in low-level caves:
The wood type of mineshafts depends on the biome, mainly based on the wood used in the trees (spruce in this case); all pieces, not just corridors and intersections, also generate wooden platforms over air:
A "double dungeon", with 2 spawners, each spawning a different type of mob, including creepers, witches, and endermen, and 2-3 chests; chest loot includes loot specific to the mobs and/or biome, and the walls may include stone bricks and "compressed cobblestone" (double dungeons always use chiseled stone bricks in the floors, making it another source besides the 3 in jungle temples):
Giant caves aside, the majority of the underground is vanilla 1.6.4-like "swiss cheese" caves (dense clusters of tunnels), which are my favorite type of cave to explore, with various modifications to tunnels/rooms over 12x12 chunk regions; the largest caves and ravines (volume >= 100000) each occur 2-3 times per level 3 map and giant cave regions occur about once every three maps:
Chunks per underground feature:
Dungeons: 39
Mineshafts: 181
Large ravines (>=25000): 309
Large caves (>=25000): 419
Double dungeons: 633
Large circular rooms: 707
Large ravines (>=50000): 792
Large caves (>=50000): 926
Toroidal caves: 1007
Combination cave systems: 1194
Large caves (>=100000): 1487
Large mineshafts (10-11): 1528
Ribbed tunnel cave systems: 1715
Zigzag cave systems: 1715
Large ravines (>=100000): 1749
Random cave systems: 1951
Spiral cave systems: 2066
Large cave cave systems: 2224
Vertical cave systems: 2429
Circular room cave systems: 2447
Ravine cave systems: 2447
Maze cave systems: 2572
Large ravines (>=200000): 3527
Large caves (>=200000): 3539
Large cave clusters: 4351
Network cave regions: 5656
Colossal cave systems: 6554
Large caves (>=300000): 6951
Large ravines (>=300000): 7580
Strongholds: 8192
Giant cave regions: 13107
Large caves (>=400000): 14838
Large ravines (>=400000): 21478
This map covers a 2048x2048 block area (a level 4 map):
This is the same map without "normal" caves, aside from larger tunnels and rooms, as well as larger ravines and mineshafts, and strongholds:
As far as harvesting large trees goes, I use ladders to scale the trunk, breaking leaves in the way, until I reach the top, then I mine downwards, or mine up in a spiral staircase; this is also helped by my own "vein miner" enchantment, which enables axes to mine 3 logs at a time, although branched trees are still more trouble to harvest (in particular, my "mega trees" have numerous branches, but less than in earlier versions due to increasing the leaf decay radius from 4 to 6, similar to newer versions). As it doesn't matter what wood type you use to make torches I just grow 2x2 spruces for this purpose (I use 2x2 jungle trees in vanilla as they are the only such tree type; TMCW has 2x2 variants of every vanilla sapling, plus a 3x3 spruce variant, with 2x2 spruce being the only one without branches). Otherwise, I always build my bases in relatively flat and treeless biomes to minimize the amount of terraforming I need to do (my main bases are always near spawn, which is guaranteed to be a suitable biome; secondary bases are placed near the middle of each level 3 map surrounding the central map, offset as needed for the most suitable locations).
I know what you mean about journals though. When I started in 2010 I had an online blog noting my adventures, builds and stuff, never really thought about the future but i'm so glad I did now. Blog is totally dead now we few (Hardly any) posts in the last 3/4 years at least because i too tgought "Who reads blog anymore?", so it's been largley inactive for a few years.
I used to do the same with trees, standing on stumps but haven't gotten lazy and just use a timber datapack to chop all the wood down in one fell swoop. That said, I never had automatic tree farms - ever as I found it quite relaxing chopping down trees and satisfying. Guess that's just me.
Like Minecraft forums or interested in my world? Try My message board, it's better moderated because I run it directly and have run Internet message boards for 21+ years! Better software and I have much more control to keep the content more up to date. Free to join, 13 years+.
You're referring to your past self as a different person? I think that happens when we spend too much online/on computers indeed, I've caught myself doing it a couple times in the last couple years and I sure know I didn't before covid.
I've always thought of myself as constantly changing over time and sometimes referred to older versions of myself as a different person - even decades ago, before there even was an online. I really feel the Buddhist aphorism "you can never step in the same river twice". I do seem to change more than most people FWIW.
I had been wondering what happened to you (and so many others); I'd always been partial to modded survival journals, especially given that you'd been making your own mods to shape the game to your own likings, because that is exactly what I've been doing (I'd always wondered, was your "Climate Control" mod inspired by my own mods? I know that you made a Forge version of my "Random Biomes" mod).
Also, as far as 1.18 goes while I haven't played on it (or even touched anything past 1.13, when I stopped updating a mod that reverted the changes to caves in 1.7 due to a combination of lack of interest/demand and difficulty in updating it after code changes) from what I've seen it is no better than 1.7 in terms of local-scale biome variety, where to me local means on the order of hundreds of blocks (even a level 3 map takes about 1 1/2 months to explore, of course, as always, by exploring interconnected caves underground, which gives a rate of about 100 chunks explored per play session/day).
For example, this is a rendering of my current modded world, which I've been playing on for more than a year, overlayed on top of a random seed in 1.18 (ChunkBase); the size of some of the climate areas is just crazy, biome borders also look way too smooth, compared to the random shapes in the old generation:
The same goes for the underground; I have the impression that it is dominated by giant open caves with a sprinkling of the "old" caves, probably still based on 1.7 generation, and ChunkBase shows that ravines, which had the changes in 1.7, were actually made rarer, with mineshafts being the same as 1.7 and later (1.13 did remove a restriction that made them less common closer to the origin) - I;d expect the opposite given that the ground is twice as deep. For comparison, I've added over a dozen variants of caves, along with more variation in the size and structure of mineshafts and ravines, and new types of dungeons, and actual biome-specific underground (e.g. deserts replace stone-based blocks with sandstone and sand, Ice Plains Spikes has snow and packed ice, etc. Mobs like husks and strays also naturally spawn underground, including more variants, like red husks and silverfish in Mesa and Badlands):
This is the largest cave near the bottom, with a volume of more than a million blocks:
Another cave I found which is even larger:
Another large cave under a desert, showing the biome-specific underground:
A ravine with a volume of more than half a million blocks:
"ice" biomes have water instead of lava in low-level caves:
The wood type of mineshafts depends on the biome, mainly based on the wood used in the trees (spruce in this case); all pieces, not just corridors and intersections, also generate wooden platforms over air:
A "double dungeon", with 2 spawners, each spawning a different type of mob, including creepers, witches, and endermen, and 2-3 chests; chest loot includes loot specific to the mobs and/or biome, and the walls may include stone bricks and "compressed cobblestone" (double dungeons always use chiseled stone bricks in the floors, making it another source besides the 3 in jungle temples):
As far as harvesting large trees goes, I use ladders to scale the trunk, breaking leaves in the way, until I reach the top, then I mine downwards, or mine up in a spiral staircase; this is also helped by my own "vein miner" enchantment, which enables axes to mine 3 logs at a time, although branched trees are still more trouble to harvest (in particular, my "mega trees" have numerous branches, but less than in earlier versions due to increasing the leaf decay radius from 4 to 6, similar to newer versions). As it doesn't matter what wood type you use to make torches I just grow 2x2 spruces for this purpose (I use 2x2 jungle trees in vanilla as they are the only such tree type; TMCW has 2x2 variants of every vanilla sapling, plus a 3x3 spruce variant, with 2x2 spruce being the only one without branches). Otherwise, I always build my bases in relatively flat and treeless biomes to minimize the amount of terraforming I need to do (my main bases are always near spawn, which is guaranteed to be a suitable biome; secondary bases are placed near the middle of each level 3 map surrounding the central map, offset as needed for the most suitable locations).
Your random biomes was a part of the motivation for Climate Control/Geographicraft, because Forgifying it made me realize how easy it can be to change the biome code. Unfortunately what CC does is pretty complicated so it's still a lot of work.
The old fractal biomes look a lot better on a false color AMIDST-style map but on a minecraft map or on the ground the new ones (which I'm guessing are based on Perlinesque gradients a la ATG or early Highlands) look fine.
To you, the new climate system isn't varied enough, but I think you realize you are on the very extreme end of a low-travel player and Mojang does have to aim for the average. The area you have covered looks to be literally only a few dozen blocks from *both* hot and cold on that map and for most players it would be super-fine - maybe even *too* close for large area builders. If they are indeed using a Perlinesque system they can't compress it *too* much because then you start getting "sliver" biomes on steep gradients - this was something of a problem even in the old Highlands system, with transition biomes getting thin enough to be a little unsightly.
Your cave mods are really amazing.
Rollback Post to RevisionRollBack
Geographicraft (formerly Climate Control) - Control climate, ocean, and land sizes; stop chunk walls; put modded biomes into Default worlds, and more!
Playing modded, I need a modlist. I have a couple of categories:
First, worldgen mods:
RTG: To get a pretty world (plus partly my mod)
Geographicraft: for control of the kind of world I get (plus my mod)
Biomes o Plenty: for biome variety
Underground Biomes: Good looking rocks, in landscapes and building (plus used to be my mod)
Second, mods to make puttering around the base more interesting:
Bibliocraft: GREAT furniture.
Harvestcraft: Interesting things to do with food.
Hunger Overhaul: So I don't have to continually snack.
Spice of Life: A reason to use Harvestcraft.
(I meant to put in a backpack mod and thought I had but didn't)
Third, a couple of mods to give some long-term goals.
I looked at some tech mods - I've done lots of magic mods in the past and figured I should do some tech mods. But when I looked at them I didn't find one that immediately grabbed me, so back to magic and artificer's mods.
Astral Sorcery - interesting magic mod
Tinker's Construct - fun with complex tools
Psi - a spell mod with a mathematical bent sounds interesting
I was a big fan of Thaumcraft and still am; but I wanted to try some new stuff. Ditto for Botania.
Fourth, some mods for the crazy offhand chance I will go to the end of the world someday.
DaVinci's Vessels
Waystones
Fifth, a couple that sounded fun
Serene Seasons: Finally, a seasons mod, cool!
Better Animals Plus: I wanted to a wide variety of animals
Villager Names: Cool, individualized villagers.
And Sixth, put in some challenge.
With multiple mods souping up me (at least down the line) I figured I needed some challenge. I looked at a couple of monster mods but couldn't pick one, and in the end I used Lycanite's, which I had played with on a server run by Dulciphi oh so many years ago.
So once everything, was set up, New World!
Uhoh, Ice Mountains start. Not a good place; no food or timber. There are numerous Lycanites about; I hope they're not hostile.
Head down the mountain, past a lake (fortunately the locals weren't hostile) through a pass and -
Ocean.
Back up, take another pass
More Ice Mountains.
Keep going (how much of the day have I already used?)
Ah! finally! some trees!
I get there and it's Grove, a pretty BoP biome. I can live with this.
Chop some wood, put up a basic starter shelter, mine stone for a furnace, roast up a few charcoal for torches, and start a dropshaft. It turns dark but I keep going and then:
A thunderstorm starts. Ehh, just an annoyance.
The cognoscenti will notice I've got vanilla stone. I'd forgetten to put Underground Biomes in my mod folder. Foo.
But, the mining is going well! I find a coal deposit, solving my very limited charcoal supply. Pickpickpickpick.
FWOOMM!
I get Geonached. The Geonach is a Lycanite's monster that can appear as you're mining resources. It's very tough, and afflicts you with a Weight effect that makes you move slow and unable to climb. I have - a wooden pick, and no armor. Yeah, CRUNCH!
So, now it's night, thunderstorming, and since I never made a bed (never even saw sheep) my restart is hundreds of blocks away on an Ice Mountain.
Yeah, forget it. Bye, world.
So, put UBC in the mod folder, alter the Lycanite's config to shut off Geonachs, and New World!
BoP Marsh. No trees in sight. Could be a problem. But wait! In the distance, I spot ONE small oak tree on an island. Hey, it's a survival island with a twist! I am on for it!
I swim and walk over, noting a couple of exposed ores in water channels that might come in useful. Chop it down, plant the ONE sapling I get (close!) and set up my starter base with loam soil:
It's longer than usual because I don't have sticks for a dropshaft ladder so I'll have to staircase down. Long mean I can go back and forth right underneath and maximize my spawning exclusion zone.
I spend the evening staircasing down, until I run out of torches. Then I wait for dawn.
Come dawn there are two creepers outside. I lure them to the walls and kill them by hitting their heads over the wall, while standing on some raised blocks in the center. Satisfying!
Out to mine those resources.
And, as I'm mining these, I get - attacked by a lamprey. No weapons, no armor. Whatever, I'll just climb out and -
I can't? Trapped by a Geonach is one thing but this is ridiculous. So I fight as well as I can, but still lose.
Well, not as bad because it's day and my spawn is nearby. Run over, grab my stuff, whack a few tall grass because I need some seeds and -
A Spriggan materializes. A Spriggan is a Lycanite's monster that can appear where you're destroying greenery (seeing a pattern here?) It then flies up where I can't attack it, and kills me. (seeing a pattern here?)
OK, that's it. I'm done with Lycanites for now. Maybe when I'm souped up.
Next: New World! (but this one sticks).
Rollback Post to RevisionRollBack
Geographicraft (formerly Climate Control) - Control climate, ocean, and land sizes; stop chunk walls; put modded biomes into Default worlds, and more!
I've always thought of myself as constantly changing over time and sometimes referred to older versions of myself as a different person - even decades ago, before there even was an online. I really feel the Buddhist aphorism "you can never step in the same river twice". I do seem to change more than most people FWIW.
I change somewhat often too and can relate on the Buddhist concept; I guess I just say 'the old (my name here)' rather than re-name or uh re-number myself lol.
This is pretty great - I'd like to think maybe we're returning to that 2014-2015 era where journals were quite trendy, but I suppose time will tell. The forum community is certainly much more like a small town today, which is not necessarily a bad thing at all - but comparatively, activity is much more muted than it once was.
Personally, I haven't ventured into a post-1.18 world as of yet, and I don't know if I will. It's not because I take issue with the terrain, but more so because I think updating my world beyond that point will create significant performance issues amongst other things.
On the topic of hot and cold biome distances - it took me years to find a mesa biome initially, and when I finally did so, guess where it was? A solid 10,000 blocks from spawn! Later, I found a closer one about 6,000 blocks away - how much more convenient (sarcasm). On the other side of that spectrum, I have sections of ice spikes across bodies of water from a bamboo jungle even in 1.15-generated terrain. I didn't mind that they kept some realistic blending among biome temperatures, but the distances were exhaustive in the beginning (and continue to be, though now I have access to every biome in the game within a 10,000 block radius).
Rollback Post to RevisionRollBack
LP series? Not my style! Video series? Closer, but not quite. Survival journal, maybe? That's better. Now published is the Season 3 FINALE of the Legends of Quintropolis Journal (<< click to view)!! World download and more can be found there.
Dude this is actually really cool! I've been trying to get back into Minecraft too, and this does give me a few ideas. Maybe I should try documenting everything too
Personally, I haven't ventured into a post-1.18 world as of yet, and I don't know if I will. It's not because I take issue with the terrain, but more so because I think updating my world beyond that point will create significant performance issues amongst other things.
On the topic of hot and cold biome distances - it took me years to find a mesa biome initially, and when I finally did so, guess where it was? A solid 10,000 blocks from spawn! Later, I found a closer one about 6,000 blocks away - how much more convenient (sarcasm). On the other side of that spectrum, I have sections of ice spikes across bodies of water from a bamboo jungle even in 1.15-generated terrain. I didn't mind that they kept some realistic blending among biome temperatures, but the distances were exhaustive in the beginning (and continue to be, though now I have access to every biome in the game within a 10,000 block radius).
I would NEVER update a world from pre-1.18 to 1.18+. TOTALLY different worldgen. New world or no update.
(originally typo'd 1.8, thx for the correction, Princess Garnett)
The excessive travel distances and crazy excessive search distances were just as unnecessary as erasing the oceans, and even more problematic for most players (although not as funny to point out.) It took me about 2 weeks of thought and testing to get CC so it almost always had both within 2000, and without even changing the BiomeLayer code much, just reordering (although I did have to write a new climate smoothing layer). And it was a big enough problem (along with the other assorted biome issues like Jungle clumping) to justify a chunk wall generating update long before 2021. That was a ball dropped from orbit.
And, yes, the fact that Oceans didn't have temperature could create some real odd situations with narrow oceans. Although, with the 1.7 code, if Oceans had temperatures there would be almost no hot or cold zones anywhere. I have no idea whether they tried that.
I would NEVER update a world from pre-1.8 to 1.8+. TOTALLY different worldgen. New world or no update.
The change to terrain generation around that time was between 1.6 and 1.7, not 1.7 and 1.8 (all that changed there was the new stone types and water temples, but biome placement/generation remain largely unchanged).
But they said 1.18 not 1.8, and 1.18 introduced a feature you may find neat if you're unaware of it, as it gets around that old problem finally; terrain blending.
That's 1.2.5 terrain on the bottom, and 1.19 terrain on the top.
I felt the same way as you before. Terrain transitions were unacceptable to me. When 1.7 came, I pre-generated a large area of land and then used nether tunnels to access "the new lands" of 1.7. But largely the world felt stuck in 1.6, with access to some 1.7+ blocks and features. 1.18 allowed it to be more seamlessly brought up to date.
But they said 1.18 not 1.8, and 1.18 introduced a feature you may find neat if you're unaware of it, as it gets around that old problem finally; terrain blending.
That's 1.2.5 terrain on the bottom, and 1.19 terrain on the top.
That's very impressive. One of the RTG members was trying to do that between vanilla and RTG but couldn't get the performance acceptable. How do the junctions look when they're not covered by water? RTG actually does that between different biomes (it's a much easier task with RTG generation) and I had to put a lot of work into making the junctions nice.
Rollback Post to RevisionRollBack
Geographicraft (formerly Climate Control) - Control climate, ocean, and land sizes; stop chunk walls; put modded biomes into Default worlds, and more!
How do the junctions look when they're not covered by water?
It varies, but typically, quite well in my opinion. The example I showed there is actually one that would normally be a a bad example (at least if the original terrain is from before 1.13), because water is where transitions can tend to be more apparent, not less, because the underwater block pattern differs, and lack of kelp in pre-1.13 oceans.
Here's a good and probably typical example, with a straight transition from one biome (extreme hills, I think?) to another (jungle). Old on the bottom; new on the top. In-game, this would look rather natural outside of the transition being a literal line (but even that looks a bit less out of place since biome transitions in 1.18+ seem to be more gradual shaped than before anyway).
Here's more of the opposite; one showing one of the tell-tale signs (also, a transition from desert... to snow... in MY 1.19 world!?). Old on the left; new on the right.
You can see where the river stops and there's a straight line.
This is most common with rivers and swamps.
I know NOTHING about terrain generation, but from what I can gather, it gives the game four chunks worth of space to transition from old to new. At least, those "lines" are always exactly that long. It's apparent with swamps due to the coloring they have.
But stuff like that is more of an imperfection of a feature that otherwise works very well (in my opinion) for FINALLY letting us bring old worlds up to new versions. The only drawback is this wasn't backported, so if you want to update to a specific version before 1.18, you're still left dealing with the transitions of course.
On maps it looks rather apparent. Here's my oldest world originally, and then after updating.
The "pale" chunks are the 1.6 and prior generation, the rest are new ones. I trimmed my world on a region basis so it gave it plenty of space to work with.
And I can't comment much on performance. 1.18+ is more performance heavy already when generating terrain, largely since the underground is deeper. It also does updates to old chunks to extend their depth (checks for bedrock, turns it to deep slate, and then generates new deeper terrain below that). So I can't say how much of that performance impact is from terrain blending alone. Whatever it is, it's worth it in my opinion.
I'm really impressed Mojang could do the junctions so well. The problem with junctions is that they're the average of two different systems, so the variance drops and they look "flat". But either they cooked up a solution or it's not as much of a problem with 3D perlin. They probably should have "fractalized" the boundaries by generating additional terrain under old rules in varying widths along the edge. But still nice.
If, as i suspect, they're doing the terrains with some kind of 2D perlinesque noise for climate parameters, they could have smoothed out those awkward climate junctions by inferring climate parameters from the old biomes and averaging *those* over longer distances. Wouldn't help with rivers though.
Rollback Post to RevisionRollBack
Geographicraft (formerly Climate Control) - Control climate, ocean, and land sizes; stop chunk walls; put modded biomes into Default worlds, and more!
Yeah, rivers and swamps (due to coloring) are the ones that stand out most. It's definitely not perfect beyond that, but it's very good regardless. I'd even say the uniqueness of it adds to some variety during these transition areas versus the old terrain alone or new terrain alone. When I originally updated, I cut out a lot of land. I then later went back and cut more regions out, and it was interesting to see how the blending adapted in different circumstances.
I don't have it anymore as it's one of the regions I later removed, but there was originally a spot that had a VERY interesting transition. Namely, forests going into plains right at a cutoff that is now ocean in new generation leads to some nice waterfront terrain IMO, maybe more realistic than what the game ever does.
And if the game goes to put a village or structure on the edge, you can get some VERY interesting results. I have a desert temple partially underwater next to a mangrove swamp (old ocean went to blend into mangrove swamp, but old ocean would be desert in new terrain generator, so that four chunk border must have been a valid place for the desert temple despite being shallow ocean).
It's interesting to watch and see what it does at times. The imperfections have their own character, or as I call them, "scars of an old world".
Oh, and caveat, this is all referring to above ground. I don't think caves tend to line up as well (I've found they just end in a sudden flat wall at times), but I don't think most of the people who fell into the spot of objecting from updating due to unsightly borders will find that to be something that keeps them from updating.
Anyway, looking forward to your world updates. I just wanted to mentioned that, and if you do update your world now that you know of this, it'd be interesting to see your impressions (and perhaps what "scars" the terrain blending may result in with your world).
And I can't comment much on performance. 1.18+ is more performance heavy already when generating terrain, largely since the underground is deeper. It also does updates to old chunks to extend their depth (checks for bedrock, turns it to deep slate, and then generates new deeper terrain below that). So I can't say how much of that performance impact is from terrain blending alone. Whatever it is, it's worth it in my opinion.
Most of the increase in resource usage in newer versions has to do with how Mojang decided to start coding the game in 1.8, with absolutely no regard to code complexity or object creation; there is absolutely no reason why the default memory allocation needs to be 2 GB, and mods are unimaginably worse (I see people saying to allocate up to 8 GB(!) on modern modpacks) - I've added at least a thousand new features to TMCW yet haven't seen any increase in memory or CPU usage at all, if anything I've decreased them with optimizations and bug fixes. I haven't actually run newer versions but I've had people make statements like this:
Amazing work falls very short, I'm completely impressed at what you accomplished with TMCW, the game runs at 180+ fps on max settings (1.18 barely hits 40 on modest settings, 1.16.5 with a lot of performance mods gets close to that but with choppier feel) with all that extra content of incredible quality.
A comment made by the creator of Optifine about 1.8; I can only imagine how bad it must be by now:
Maybe the biggest and the ugliest problem is the memory allocation. Currently the game allocates (and throws away immediately) 50 MB/sec when standing still and up to 200 MB/sec when moving. That is just crazy.
This comment left on a bug report claims to have doubled performance by caching blockstate/blockpos objects (which is still slower than not having them at all; my own "blockstate" system uses ints which correspond to a block ID + data value * 256):
I noticed in 1.12.x that getBlockState (in World, Chunk, and ChunkCache) accounted for substantial amount of CPU overhead. I developed a block state cache (write-through direct-mapped cache using a specially tuned hash to map from coordinates to cache entries), which made a HUGE difference. That plus a BlockPos neighbor cache literally doubled Minecraft performance for the test cases we tried.
For comparison, this is a memory/CPU profiler chart while running at max settings (Vsync at 75 FPS); memory usage slowly craws upwards by about 100 MB over about 40 seconds; also shown is a listing of the top objects by retained size (an approximation of the memory they use, including anything they reference); chunks dominate by far, as I'd expect for a game like Minecraft, and why I simply cannot believe that even the largest modpacks would need so much more memory (I'm also of the opinion that if a mod is adding like 20 different types of e.g. tools and armor it is doing something very wrong; enchantments and attributes let you have dozens to hundreds or more variants of a single item, and how do you manage dozens of types of ores?
Also, a deeper underground should not place such a heavy demand on the game; my first computer had hardware from the mid-2000s and 1.8 was crippling, yet I played with modding the underground to be up to 128 layers deeper ("triple height terrain") along with caves scaled up to match, and didn't have much issue playing ("Advanced OpenGL" / hardware occlusion queries worked quite well on my old system, notably, the example of 1.8, which replaced AOGL with a custom CPU-based culling method, seems to show not much culling is being done with 130 sections still being rendered despite being underground, while AOGL was much more aggressive), and it didn't have the optimizations I've since made to the game (TMCW actually generates terrain twice as fast as vanilla 1.6.4 despite being more complex than 1.19, if not as deep (I much prefer horizontal exploration while caving), and only runs on a single thread (server-side, plus a file-I/O thread, but only chunk saving is threaded):
Vanilla 1.6.4 (5 seconds from starting server to player joined, when the world loads on the client):
2023-02-24 17:36:36 [SERVER] [INFO] Starting integrated minecraft server version 1.6.4
2023-02-24 17:36:36 [SERVER] [INFO] Generating keypair
2023-02-24 17:36:36 [SERVER] [INFO] Converting map!
2023-02-24 17:36:36 [SERVER] [INFO] Scanning folders...
2023-02-24 17:36:36 [SERVER] [INFO] Total conversion count is 0
2023-02-24 17:36:36 [SERVER] [INFO] Preparing start region for level 0
2023-02-24 17:36:37 [SERVER] [INFO] Preparing spawn area: 10%
2023-02-24 17:36:38 [SERVER] [INFO] Preparing spawn area: 32%
2023-02-24 17:36:39 [SERVER] [INFO] Preparing spawn area: 56%
2023-02-24 17:36:40 [SERVER] [INFO] Preparing spawn area: 81%
2023-02-24 17:36:41 [SERVER] [INFO] TheMasterCaver[/127.0.0.1:0] logged in with entity id 266 at (-84.5, 64.0, 244.5)
2023-02-24 17:36:41 [SERVER] [INFO] TheMasterCaver joined the game
1.13.2 (the latest version I've run); took 15 seconds, a lot of it on something other than terrain generation (preparing spawn area):
2023-04-03 19:05:09 [SERVER] [INFO] Starting integrated minecraft server version 1.6.4
2023-04-03 19:05:09 [SERVER] [INFO] Generating keypair
2023-04-03 19:05:09 [SERVER] [INFO] Preparing start region for level 0
2023-04-03 19:05:10 [SERVER] [INFO] Preparing spawn area: 23%
2023-04-03 19:05:11 [SERVER] [INFO] Preparing spawn area: 70%
2023-04-03 19:05:12 [SERVER] [INFO] TheMasterCaver[/127.0.0.1:0] logged in with entity id 416 at (-248.5, 69.0, -244.5)
2023-04-03 19:05:12 [SERVER] [INFO] TheMasterCaver joined the game
Also, the game is highly wasteful by generating noise with a precision far beyond what is necessary - I removed the first 6 octaves (or about 1/3) of noise with virtually no effect on terrain, while giving a significant performance improvement, furthered by not generating 3D noise for the lower 40 layers (the deepest oceans are around y=45) - in all, TMCW is about 6 times faster at generating noise than vanilla 1.6.4:
// Benchmarks:
// Vanilla: 3.578 ms per chunk
// TMCWv4: 1.465 ms per chunk = 2.44 times faster than vanilla
// TMCWv5: 0.600 ms per chunk = 5.96 times faster than vanilla
private static void testNoiseGenerator()
// First number is octaves, second is octaves skipped; sum is the effective number of octaves
// Last parameter is set to 5 to skip lowest levels when generating 3D noise to reduce generation time
this.noiseGen1 = new TerrainNoiseGenerator(this.rand, 10, 6, 5);
this.noiseGen2 = new TerrainNoiseGenerator(this.rand, 10, 6, 5);
this.noiseGen3 = new TerrainNoiseGenerator(this.rand, 6, 2, 5);
this.noiseGen4 = new TerrainNoiseGenerator(this.rand, 10, 6, 5);
this.noiseGen5 = new TerrainNoiseGenerator(this.rand, 4);
this.noiseGen6 = new TerrainNoiseGenerator(this.rand, 1);
// Minimum ocean and river/other biomes
public static final int RIVER_DEPTH = 55;
public static final int OCEAN_DEPTH = 45;
// Returns minimum depth of surface terrain (is always solid stone at or below this)
int minHeight = (Biomes.isOcean(biomeType) ? this.getOceanDepth(blockX, blockZ, groundNoise) : this.getRiverDepth(groundNoise));
// Fills in ground below minHeight with stone and bedrock at y=0
for (int y = minHeight; y > 0; --y)
{
this.terrainData[index | y] = (short)BlockStates.stone;
}
this.terrainData[index] = biome.getBedrockBlock();
// getOceanDepth similarly returns returns 45-47 (includes code to smooth height difference from non-ocean biomes)
// Returns minimum depth of river and other non-ocean biomes, ranging from 55 to 56
private int getRiverDepth(float groundNoise)
{
// variation is between 0 and 1
return RIVER_DEPTH + ((int)(groundNoise + 16.0F) & 1);
}
I also did something similar in double/triple height terrain, which raised normal terrain up by 64 or 128 layers and filled in the space below with solid stone and a layer of bedrock (actually, the lower 160 layers in THT, since again, oceans don't go all the way down to y=0; the minimum depth in vanilla seems to be around y=40, in TMCW it is a bit less, 45, with less height variation so it doesn't become exposed so the average depth is similar, if not greater. The limit of y=55 for rivers is because I decreased their "minHeight" (base height) to ensure they can carve though even the most extreme terrain; most do reach their minimum depth, with a 1 block variation added so they aren't uniform, this also makes them much wider than vanilla rivers, which are often a sorry sight in 1.6.4, even in a plains biome).
Also, the slowest part of cave generation is all the calls it makes to "Random", which is extremely slow, and this was responsible for most of the impact of my earlier mods, especially DHT/THT - simply subclassing it to not use "AtomicLong" halved the time taken by vanilla cave generation, with my own true 64 bit RNG being even faster by a factor of about 8, if with some limitations on the range of "nextInt(n)" which don't matter for my needs (Random also only uses the lower 48 bits, which caused seeds with the same lower 48 bits to match, except where features were biome-dependent as it used a custom 64 bit RNG; only since 1.18 has vanilla used a custom 64 bit RNG for all world generation (or almost all). I even went to far as to use a very basic 32 bit LCG for the RNG in individual features, where it is not necessary to have such a long sequence or that many unique states.
Regarding 1.8, it is also remarkable how much larger it became compared to older versions - and even with all the content I've added to 1.6.4 it is still significantly smaller (deleting META-INF does reduce the size, on the other hand, a lot of my code completely replaces vanilla classes with custom ones so the originals are sitting there doing nothing; in terms of the source code my source is about as large as vanilla 1.6.4, but only has 1/3 as many classes, so on average they are 3 times larger):
I've also added a lot of new textures and sounds but not as much as if every block/variant has a unique texture; there are 168 variants of stalagmites/stalactites but they only need 48 textures since they are recolored and/or flipped (all 17 hardened/stained clay variants are recolored to match the block they are placed on, defaulting to hardened clay; stalactites are flipped vertically. There are 7 "real" variants of each of 3 sizes, as defined by the obtainable items, and I count them as 21 blocks/items):
Episode 0a: The Return
I quit Minecraft about 6 years ago because of some computer overuse injuries. Yes, Minecraft was a major contributor. By the time I was recovered enough to be able to play Minecraft. there were other demands on my time so I just left it aside. I could have still done modding, but without being able to play the motivation was totally gone. Plus, of course, people change over time, along with their interests. I'm not really Zeno410 anymore - more like Zeno410 plus 10.
However, right now I'm largely confined to the couch for a few days to weeks due to some knee surgery so I installed Minecraft to have a go at it. I'm going to write about a few brief worlds and then do a journal of sorts on a 1.12 modded world (I'll talk about that choice later).
"Why a journal?" I hear you ask. "There are very few users on the forums anymore and nobody does journals now." Well, yeah. Probably almost nobody will read this. But -
A long time ago, (in what now seems like a faraway galaxy) I told somebody here to journal for yourself, because someday in the future you'll look at your old adventures and enjoy. And, oh boy, that is *SOOOOOOO* true. I had some downtime visiting family recently and went through the journals I wrote when I was active here and I had a BLAST. It brought back all the excitement, the satisfaction, the amazement, and the ironies from (sometimes) almost a decade ago. And, in the process, I realized from some offhand comments I made that there were other worlds I had almost completely forgotten, now existing only in those offhand comments. I remember just snippets from earlier worlds and sometimes I don't even remember which went together. So the journal worlds live on in my head, entertaining and inspiring me, while the non-journaled worlds are "Gone with the End" and might as well have never happened.
So that's why I'm journaling. For me - or more precisely, the me of the the future, Zeno410 plus 20, who hopefully will enjoy this as much as I have Zeno410's journals. Anybody who wants to follow along, welcome! I hope you enjoy it too, and maybe get inspired to write, record, or whatever for *your* plus 10 someday. But even if nobody does, that's fine, because we Zenos-of-different-times are going to stick together.
Geographicraft (formerly Climate Control) - Control climate, ocean, and land sizes; stop chunk walls; put modded biomes into Default worlds, and more!
Episode 00b - So what's happened to Minecraft?





The last time I played Minecraft was is 1.12 (and my last journal is from 1.10). There have been a lot of changes since then, to put it mildly, including a complete overhaul of world generation, my main modding interest, in 1.18. So what does Minecraft generation look like these days? Time to find out!
It had been so long I'd even forgotten the keybindings and basic recipes. So, since I was primarily playing just to get some maps and look at the generation, I decided to play on Peaceful. An overreaction, but oh well.
I started in a Birch Forest in a hilly area. My screenshotting choice skills are as gone as my playing skills so I completely forgot the establishing shot, so you'll have to settle for this much later pic while returning from an exploration:
I had started in the hilly area on the right shoulder of that big mountain. And when I started, I didn't even realize I was next to a big mountain, because I couldn't see it. I chopped some trees, and then looked for a starting base area. I *did* remember you want your first base as low as possible so there is less to mine through to get to the low areas where the good ores are, so I went downslope until I came to a "river" although it was actually just a mix of land and water you couldn't boat through. There I set up my standard starter base: a small house, 3 blocks wide on the interior with 2 high wall, basic starter stuff built *into* the wall, a door with a hole in front of the door, and a "roof" of just a 1x1 rafter in the middle.
Again, no contemporary pic, so here's one from much later:
I find this a super-practical starter house; you can build it from dirt, saving valuable starter wood; no mob can get in; zombies can't break in the door due to the drop; if a spider goes up on the roof you can kill it from below; you can pillar up in the middle and kill creepers at the wall in the morning by swatting their heads; crafting table, furnace, etc., can be accessed from inside or outside the house, and if you put a dropshaft under the open parts of the roof skylight goes down so you can mine (not too far because you only suppress spawns in a 24 block radius) even without torches.
Then I did all the routine starter stuff: get the farms going; ranch some animals; mine down to Y11 for resources (I only later found it's not so simple anymore); and in my case grow a lot of cane because I want mega maps. I'll skip the details because it's not relevant - plus it's often embarrassing what I had forgotten to do.
Once I had a maxed out map, I did a circuit of it as best as I could, then grew some more cane, made another map, and did a circuit of *that*.
To establish what I'm looking for, I have 4 main complaints with the old vanilla minecraft generation from 1.7 to 1.17:
1) Search distances are absurd. Obscenely absurd. If you are searching honestly, (no AMIDST cheats, etc.) it normally takes TENS of THOUSANDS of blocks searched to find both hot and cold terrain. Almost nobody does that much searching. Even I don't, because on pre-1.18 vanilla generation the shortage of water and flat plains makes it hard to travel long distances. For rare biomes it's even crazier; in my first 1.7 world I later ran AMIDST and found the nearest Ice Spikes was 9,000 blocks from start, which would have taken two hundred and fifty THOUSAND blocks of an efficient but honest square spiral search. Ab-surd.
2) No large scale structure. Big maps are just a big mess. The "oceans" are just large lakes, the mountains don't come in ranges, and the hot and cold biomes are rare in land areas.
3) You can't see much. Most terrain is closed forest, and there aren't many clearings or water areas to provide relief. Plus the trees are mostly short and shrubby which puts the foliage in your line of sight.
4) There isn't much *to* see. Most terrain is fairly flat. Hills mostly occur as sub-biomes restricted to the interior of the mostly-forested biomes so you can't see them (Forest Hills in plains are sometimes a delightful exception). Trees are mostly smallish and bland. In the terrains with relief like Extreme Hills there is often seriously crazified terrain which I know some people like but which I find ridiculous.
So, how does the newer system do on these? Well, here's the resulting map:
1) Search distances:
Basically fixed! The first map I did (on the right) had *both* hot and cold terrain on it. Sample size of 1, etc., but on this map a typical player *would* find both kinds of terrain. This would be virtually impossible on the old system. Plus, even in this search I found several of the new biomes, including Cold Ocean, Sparse Jungle, Badlands, and Mangrove Swamp. This is about as good a result as I got with Geographicraft/Climate Control.
2) Large Scale structure:
Not great, but substantially better. There's a hot to cold transition on this area - hot on the left, and cold on the right. Plus, although you can't see it on the map, the ocean wraps around on the right so the entire right map is of a penisula. So there's a north cold penisula, a south cold peninsula, an ice bay inbetween, the hot desert to the right; different regions are actually *different* in a meaningful way and traveling around can feel like you're going somewhere and not just "another forest, another forest, another forest".
I also spotted ocean on all 3 outside sides of the left map so there's probably some sea/land structure there as well, although you'd need more maps to see it.
3) You can't see anything:
Better, but still needs some work in the temperate zone. Hot zone is mostly open terrain (desert/badlands/sparse jungle) so you can see pretty well there and I got a number of great vistas. The temperate zones have these "old growth" areas with bigger and (sparser?) trees and I found it somewhat easier to get views there. And the new wider rivers really help too:
but, not as much as I'd like. I still spent a lot of time in the temperate zone and other forests looking at trees and unable to see anything much. Some of this is good, because contrast is good; when you spend *some* time struggling through a tree maze it makes a grand vista even sweeter when you get it; but at least in the temperate zone it's too much. The fact that I didn't notice that grand mountain when I started just proves the point.
4) Not much to see:
Also much better:
The new terrain system generates a good number of nice attractive hills and mountains *without* the impossible overhangs and floating dirt that used to be common in rugged terrain. The trees are a bit better too, with the larger trees in old growth, although still no match for trees from the mods like BoP, Highlands, and RTG.
There is one newish issue: excessive small-scale ruggedness. I found it surprisingly difficult to get around in my river base because of 2-high jumps; this became particularly noticeable when I was struggling to lead animals into my pens. This is even worse in mountains, which sometimes I found nearly impossible to cross quickly (without digging passes, etc.); that might be a design intention to make mountains more of a barrier, while ruggedness in flatish areas is just a pain.
So, overall, I'd say a B+ result. Not everything is fixed to my taste but everything is substantially improved. I get a sense of a professionalism that was lacking in the 1.7 revisions; they looked at complaints, devised way to address them, and then made sure they actually worked without creating (big) new problems. I also get a sense they've been looking at mods: I see stuff that smells like Alternate Terrain Generation, Realistic Terrain Generation, Geographicraft, and Biomes O Plenty - not in the sense of stealing stuff, but looking at possibilities, design options, etc.
The current professionalism contrasts to the amateur hour of the 1.7 changes. My (un)favorite example was the land-sea code. The #1 land-sea complaint for pre-1.7 was ridiculously long ocean trips; it was not unusual that if you just set off in a random direction from a continent you would spend hours of real time travelling before you saw land. So they repeated a step that doubled the size of the continents. But said step also DOUBLED THE SIZE OF THE OCEANS! Yargh! Look at the code, fellas! So then, apparently in desperation, they filled the ocean with large islands, so thickly they all join together and form a universal continent. This would have worked pretty well, actually, if they'd just had about half as many islands, but that would have taken a few hours of thought and a couple days of testing and I guess that was too much? I get the impression they were horribly understaffed and the changes were super-rushed. It would all be rather funny if it hadn't become the standard generation system for 8 years.
If I were just a player, I'd probably carry on in 1.19. It's not to my taste in everything, but it's pretty good. But, I'm not just a player, I'm a player/modder, and I miss the extreme customizability I had designed into my worldgen mods. I'm also on something of a nostalgia kick too, so I want to experience the playstyle I had designed. So, I decided to put this world aside and start up a modded 1.12 world.
Next: some experiments don't come out too well.
Geographicraft (formerly Climate Control) - Control climate, ocean, and land sizes; stop chunk walls; put modded biomes into Default worlds, and more!
I'm wondering if you got a good particular world. While I want to say the biome variety is better now than it was since 1.7, and perhaps it is, the climate zone still feels like it could be less strict.
As per your mention of the views and forests, I've felt for a LONG time that the forests, and specifically the trees that make them up, need overhauled. I feel like all the oak trees except for large oak trees should perhaps go away, and that they should also get larger at the a maximum. That would be a start to making some better forests. Keep some (maybe 5%?) of the smaller ones (still with the canopy starting a minimum of 5 or 6 blocks off the ground) for a bit of variety and to add some lower coverage here and there, but not enough to make forests like they are now. Right now, most "trees" are actually overgrown large bushes that start one to three blocks from the ground... those are NOT trees and make for terrible forests!
The downside to such a change might be people complaining they can't just swiftly gather wood without leaving ground level, and those who want a "natural" looking forest might find they need to do more manual work after clearing some trees (I would fall into this myself). And the whole two or three block tall tree is somewhat iconic and people might complain. But I'd still find it worth it. These old trees/numerous forests are a big detractor to modern Minecraft world/biome generation in my opinion.
As you found out, the variants of birch forests with taller trees are an improvement, but instead of adding variants that differed in just one or two ways, they should have just overhauled all of them. Birch forests possibly were going to get that treatment in 1.19 but they didn't in the end, and many of us were left sad. On the other hand, maybe it's for the best as what forests need aren't one being touched up, but the bigger subject of forests as a whole being looked at. I feel like there's too many forest variants and it's part of why forests in particular are so common. Like, don't get me wrong, I love finding a forest full of flowers. But you have probably dozens of types of forests, some just being minor differences (birch forest and "old growth" birch forest, etc.), when instead they should have narrowed them down to a few main types. Maybe get rid of flower forests entirely and just have the "flower biome" a separate overlay that can happen rarely in forest or plains.
But yeah, world generation is vastly improved (and if it hasn't happened yet, wait until you find mountains and/or start going underground). But there's still so much that could be done.
Also, wondering if you found any of the new animals yet. Llamas are the best.
I've played a lot on modded worldgen, where bigger trees are the norm, and I've got quite a bag of tricks for chopping bigger trees. I find it part of the fun to come up with ways to chop these trickier trees. We all talk about zillions of ways to get down to mining level and lay out our mines, deal with lava and cave breakouts, etc. and we all think *that's* part of the fun. Why couldn't Mojang treat forestry the same?
But even if Mojang doesn't want people to work to chop trees, there's a lot of room for improvements in your direction. My lowest level trick is to chop out the 2nd and 3rd trunk block, stand on the stump, and chop straight up from there. That will take out a 9 high trunk. In addition, foliage could be higher relative to the trees (another trick the mods use a lot). A 9 high trunk with foliage going from 5 to 11 would be no problem at all visually, allow much better views, and yet still allow players to chop down trees without any more work than stepping on stumps. A very workable compromise, IMO.
One last thing to add is that having a wide range of trees would allow more mixing. There could be old growth forest where they're all 9 high trunks, mixed growth with some high and some low, and they could still have some "scrub" forests like the current ones.
Also; yes, I've seen some of the new caves and I think they're great. I just don't have much to add on that because I didn't do that kind of modding. The one thing I don't like about them is that there's a lot more smallish holes in the ground leading to long falls, and that's a problem for Hardcore, my favorite kind of play. But they're interesting and look nice.
I did see some llamas with a traveling merchant and they are indeed cute! Haven't interacted with the yet; will try to in my next go-round.
Geographicraft (formerly Climate Control) - Control climate, ocean, and land sizes; stop chunk walls; put modded biomes into Default worlds, and more!
Thank you for coming back and sharing your: (playthrough) "journal"
There are more Eyes here than you think, most prefer to quietly observe rather than overtly contributing.
These Journal type Threads are my favourite posts to read, I love: Shared Creativity... I whish more members would contribute this way.
All the best.
ᛁᚨᚾ ᚷᛖᛞᛞᛖᛋ
You're referring to your past self as a different person? I think that happens when we spend too much online/on computers indeed, I've caught myself doing it a couple times in the last couple years and I sure know I didn't before covid.
Also, forestry is cool in vanilla with how I do various spirals to get up to trees, but I agree more could be done which is why so many mods revise trees and chopping. Although some plug ins just do a cop out and make the whole tree fall when hitting its base.
Yeah, I noticed.
Also, as far as 1.18 goes while I haven't played on it (or even touched anything past 1.13, when I stopped updating a mod that reverted the changes to caves in 1.7 due to a combination of lack of interest/demand and difficulty in updating it after code changes) from what I've seen it is no better than 1.7 in terms of local-scale biome variety, where to me local means on the order of hundreds of blocks (even a level 3 map takes about 1 1/2 months to explore, of course, as always, by exploring interconnected caves underground, which gives a rate of about 100 chunks explored per play session/day).
For example, this is a rendering of my current modded world, which I've been playing on for more than a year, overlayed on top of a random seed in 1.18 (ChunkBase); the size of some of the climate areas is just crazy, biome borders also look way too smooth, compared to the random shapes in the old generation:
The same goes for the underground; I have the impression that it is dominated by giant open caves with a sprinkling of the "old" caves, probably still based on 1.7 generation, and ChunkBase shows that ravines, which had the changes in 1.7, were actually made rarer, with mineshafts being the same as 1.7 and later (1.13 did remove a restriction that made them less common closer to the origin) - I;d expect the opposite given that the ground is twice as deep. For comparison, I've added over a dozen variants of caves, along with more variation in the size and structure of mineshafts and ravines, and new types of dungeons, and actual biome-specific underground (e.g. deserts replace stone-based blocks with sandstone and sand, Ice Plains Spikes has snow and packed ice, etc. Mobs like husks and strays also naturally spawn underground, including more variants, like red husks and silverfish in Mesa and Badlands):
This is the largest cave near the bottom, with a volume of more than a million blocks:
Another cave I found which is even larger:
Another large cave under a desert, showing the biome-specific underground:
A ravine with a volume of more than half a million blocks:
"ice" biomes have water instead of lava in low-level caves:
The wood type of mineshafts depends on the biome, mainly based on the wood used in the trees (spruce in this case); all pieces, not just corridors and intersections, also generate wooden platforms over air:
A "double dungeon", with 2 spawners, each spawning a different type of mob, including creepers, witches, and endermen, and 2-3 chests; chest loot includes loot specific to the mobs and/or biome, and the walls may include stone bricks and "compressed cobblestone" (double dungeons always use chiseled stone bricks in the floors, making it another source besides the 3 in jungle temples):
Giant caves aside, the majority of the underground is vanilla 1.6.4-like "swiss cheese" caves (dense clusters of tunnels), which are my favorite type of cave to explore, with various modifications to tunnels/rooms over 12x12 chunk regions; the largest caves and ravines (volume >= 100000) each occur 2-3 times per level 3 map and giant cave regions occur about once every three maps:
This map covers a 2048x2048 block area (a level 4 map):
This is the same map without "normal" caves, aside from larger tunnels and rooms, as well as larger ravines and mineshafts, and strongholds:
As far as harvesting large trees goes, I use ladders to scale the trunk, breaking leaves in the way, until I reach the top, then I mine downwards, or mine up in a spiral staircase; this is also helped by my own "vein miner" enchantment, which enables axes to mine 3 logs at a time, although branched trees are still more trouble to harvest (in particular, my "mega trees" have numerous branches, but less than in earlier versions due to increasing the leaf decay radius from 4 to 6, similar to newer versions). As it doesn't matter what wood type you use to make torches I just grow 2x2 spruces for this purpose (I use 2x2 jungle trees in vanilla as they are the only such tree type; TMCW has 2x2 variants of every vanilla sapling, plus a 3x3 spruce variant, with 2x2 spruce being the only one without branches). Otherwise, I always build my bases in relatively flat and treeless biomes to minimize the amount of terraforming I need to do (my main bases are always near spawn, which is guaranteed to be a suitable biome; secondary bases are placed near the middle of each level 3 map surrounding the central map, offset as needed for the most suitable locations).
TheMasterCaver's First World - possibly the most caved-out world in Minecraft history - includes world download.
TheMasterCaver's World - my own version of Minecraft largely based on my views of how the game should have evolved since 1.6.4.
Why do I still play in 1.6.4?
I'll be reading!
I know what you mean about journals though. When I started in 2010 I had an online blog noting my adventures, builds and stuff, never really thought about the future but i'm so glad I did now. Blog is totally dead now we few (Hardly any) posts in the last 3/4 years at least because i too tgought "Who reads blog anymore?", so it's been largley inactive for a few years.
I used to do the same with trees, standing on stumps but haven't gotten lazy and just use a timber datapack to chop all the wood down in one fell swoop. That said, I never had automatic tree farms - ever as I found it quite relaxing chopping down trees and satisfying. Guess that's just me.
I'll be interested to see where this goes though.
Closed old thread
Like Minecraft forums or interested in my world? Try My message board, it's better moderated because I run it directly and have run Internet message boards for 21+ years! Better software and I have much more control to keep the content more up to date. Free to join, 13 years+.
16yrs+ only
I've always thought of myself as constantly changing over time and sometimes referred to older versions of myself as a different person - even decades ago, before there even was an online. I really feel the Buddhist aphorism "you can never step in the same river twice". I do seem to change more than most people FWIW.
Your random biomes was a part of the motivation for Climate Control/Geographicraft, because Forgifying it made me realize how easy it can be to change the biome code. Unfortunately what CC does is pretty complicated so it's still a lot of work.
The old fractal biomes look a lot better on a false color AMIDST-style map but on a minecraft map or on the ground the new ones (which I'm guessing are based on Perlinesque gradients a la ATG or early Highlands) look fine.
To you, the new climate system isn't varied enough, but I think you realize you are on the very extreme end of a low-travel player and Mojang does have to aim for the average. The area you have covered looks to be literally only a few dozen blocks from *both* hot and cold on that map and for most players it would be super-fine - maybe even *too* close for large area builders. If they are indeed using a Perlinesque system they can't compress it *too* much because then you start getting "sliver" biomes on steep gradients - this was something of a problem even in the old Highlands system, with transition biomes getting thin enough to be a little unsightly.
Your cave mods are really amazing.
Geographicraft (formerly Climate Control) - Control climate, ocean, and land sizes; stop chunk walls; put modded biomes into Default worlds, and more!
Playing modded, I need a modlist. I have a couple of categories:








First, worldgen mods:
RTG: To get a pretty world (plus partly my mod)
Geographicraft: for control of the kind of world I get (plus my mod)
Biomes o Plenty: for biome variety
Underground Biomes: Good looking rocks, in landscapes and building (plus used to be my mod)
Second, mods to make puttering around the base more interesting:
Bibliocraft: GREAT furniture.
Harvestcraft: Interesting things to do with food.
Hunger Overhaul: So I don't have to continually snack.
Spice of Life: A reason to use Harvestcraft.
(I meant to put in a backpack mod and thought I had but didn't)
Third, a couple of mods to give some long-term goals.
I looked at some tech mods - I've done lots of magic mods in the past and figured I should do some tech mods. But when I looked at them I didn't find one that immediately grabbed me, so back to magic and artificer's mods.
Astral Sorcery - interesting magic mod
Tinker's Construct - fun with complex tools
Psi - a spell mod with a mathematical bent sounds interesting
I was a big fan of Thaumcraft and still am; but I wanted to try some new stuff. Ditto for Botania.
Fourth, some mods for the crazy offhand chance I will go to the end of the world someday.
DaVinci's Vessels
Waystones
Fifth, a couple that sounded fun
Serene Seasons: Finally, a seasons mod, cool!
Better Animals Plus: I wanted to a wide variety of animals
Villager Names: Cool, individualized villagers.
And Sixth, put in some challenge.
With multiple mods souping up me (at least down the line) I figured I needed some challenge. I looked at a couple of monster mods but couldn't pick one, and in the end I used Lycanite's, which I had played with on a server run by Dulciphi oh so many years ago.
So once everything, was set up, New World!
Uhoh, Ice Mountains start. Not a good place; no food or timber. There are numerous Lycanites about; I hope they're not hostile.
Head down the mountain, past a lake (fortunately the locals weren't hostile) through a pass and -
Ocean.
Back up, take another pass
More Ice Mountains.
Keep going (how much of the day have I already used?)
Ah! finally! some trees!
I get there and it's Grove, a pretty BoP biome. I can live with this.
Chop some wood, put up a basic starter shelter, mine stone for a furnace, roast up a few charcoal for torches, and start a dropshaft. It turns dark but I keep going and then:
A thunderstorm starts. Ehh, just an annoyance.
The cognoscenti will notice I've got vanilla stone. I'd forgetten to put Underground Biomes in my mod folder. Foo.
But, the mining is going well! I find a coal deposit, solving my very limited charcoal supply. Pickpickpickpick.
FWOOMM!
I get Geonached. The Geonach is a Lycanite's monster that can appear as you're mining resources. It's very tough, and afflicts you with a Weight effect that makes you move slow and unable to climb. I have - a wooden pick, and no armor. Yeah, CRUNCH!
So, now it's night, thunderstorming, and since I never made a bed (never even saw sheep) my restart is hundreds of blocks away on an Ice Mountain.
Yeah, forget it. Bye, world.
So, put UBC in the mod folder, alter the Lycanite's config to shut off Geonachs, and New World!
BoP Marsh. No trees in sight. Could be a problem. But wait! In the distance, I spot ONE small oak tree on an island. Hey, it's a survival island with a twist! I am on for it!
I swim and walk over, noting a couple of exposed ores in water channels that might come in useful. Chop it down, plant the ONE sapling I get (close!) and set up my starter base with loam soil:
It's longer than usual because I don't have sticks for a dropshaft ladder so I'll have to staircase down. Long mean I can go back and forth right underneath and maximize my spawning exclusion zone.
I spend the evening staircasing down, until I run out of torches. Then I wait for dawn.
Come dawn there are two creepers outside. I lure them to the walls and kill them by hitting their heads over the wall, while standing on some raised blocks in the center. Satisfying!
Out to mine those resources.
And, as I'm mining these, I get - attacked by a lamprey. No weapons, no armor. Whatever, I'll just climb out and -
I can't? Trapped by a Geonach is one thing but this is ridiculous. So I fight as well as I can, but still lose.
Well, not as bad because it's day and my spawn is nearby. Run over, grab my stuff, whack a few tall grass because I need some seeds and -
A Spriggan materializes. A Spriggan is a Lycanite's monster that can appear where you're destroying greenery (seeing a pattern here?) It then flies up where I can't attack it, and kills me. (seeing a pattern here?)
OK, that's it. I'm done with Lycanites for now. Maybe when I'm souped up.
Next: New World! (but this one sticks).
Geographicraft (formerly Climate Control) - Control climate, ocean, and land sizes; stop chunk walls; put modded biomes into Default worlds, and more!
I change somewhat often too and can relate on the Buddhist concept; I guess I just say 'the old (my name here)' rather than re-name or uh re-number myself lol.
This is pretty great - I'd like to think maybe we're returning to that 2014-2015 era where journals were quite trendy, but I suppose time will tell. The forum community is certainly much more like a small town today, which is not necessarily a bad thing at all - but comparatively, activity is much more muted than it once was.
Personally, I haven't ventured into a post-1.18 world as of yet, and I don't know if I will. It's not because I take issue with the terrain, but more so because I think updating my world beyond that point will create significant performance issues amongst other things.
On the topic of hot and cold biome distances - it took me years to find a mesa biome initially, and when I finally did so, guess where it was? A solid 10,000 blocks from spawn! Later, I found a closer one about 6,000 blocks away - how much more convenient (sarcasm). On the other side of that spectrum, I have sections of ice spikes across bodies of water from a bamboo jungle even in 1.15-generated terrain. I didn't mind that they kept some realistic blending among biome temperatures, but the distances were exhaustive in the beginning (and continue to be, though now I have access to every biome in the game within a 10,000 block radius).
LP series? Not my style! Video series? Closer, but not quite. Survival journal, maybe? That's better. Now published is the Season 3 FINALE of the Legends of Quintropolis Journal (<< click to view)!! World download and more can be found there.
Dude this is actually really cool! I've been trying to get back into Minecraft too, and this does give me a few ideas. Maybe I should try documenting everything too
I would NEVER update a world from pre-1.18 to 1.18+. TOTALLY different worldgen. New world or no update.
(originally typo'd 1.8, thx for the correction, Princess Garnett)
The excessive travel distances and crazy excessive search distances were just as unnecessary as erasing the oceans, and even more problematic for most players (although not as funny to point out.) It took me about 2 weeks of thought and testing to get CC so it almost always had both within 2000, and without even changing the BiomeLayer code much, just reordering (although I did have to write a new climate smoothing layer). And it was a big enough problem (along with the other assorted biome issues like Jungle clumping) to justify a chunk wall generating update long before 2021. That was a ball dropped from orbit.
And, yes, the fact that Oceans didn't have temperature could create some real odd situations with narrow oceans. Although, with the 1.7 code, if Oceans had temperatures there would be almost no hot or cold zones anywhere. I have no idea whether they tried that.
Geographicraft (formerly Climate Control) - Control climate, ocean, and land sizes; stop chunk walls; put modded biomes into Default worlds, and more!
The change to terrain generation around that time was between 1.6 and 1.7, not 1.7 and 1.8 (all that changed there was the new stone types and water temples, but biome placement/generation remain largely unchanged).
But they said 1.18 not 1.8, and 1.18 introduced a feature you may find neat if you're unaware of it, as it gets around that old problem finally; terrain blending.
That's 1.2.5 terrain on the bottom, and 1.19 terrain on the top.
I felt the same way as you before. Terrain transitions were unacceptable to me. When 1.7 came, I pre-generated a large area of land and then used nether tunnels to access "the new lands" of 1.7. But largely the world felt stuck in 1.6, with access to some 1.7+ blocks and features. 1.18 allowed it to be more seamlessly brought up to date.
That's very impressive. One of the RTG members was trying to do that between vanilla and RTG but couldn't get the performance acceptable. How do the junctions look when they're not covered by water? RTG actually does that between different biomes (it's a much easier task with RTG generation) and I had to put a lot of work into making the junctions nice.
Geographicraft (formerly Climate Control) - Control climate, ocean, and land sizes; stop chunk walls; put modded biomes into Default worlds, and more!
It varies, but typically, quite well in my opinion. The example I showed there is actually one that would normally be a a bad example (at least if the original terrain is from before 1.13), because water is where transitions can tend to be more apparent, not less, because the underwater block pattern differs, and lack of kelp in pre-1.13 oceans.
Here's more of the opposite; one showing one of the tell-tale signs (also, a transition from desert... to snow... in MY 1.19 world!?). Old on the left; new on the right.
You can see where the river stops and there's a straight line.
This is most common with rivers and swamps.
I know NOTHING about terrain generation, but from what I can gather, it gives the game four chunks worth of space to transition from old to new. At least, those "lines" are always exactly that long. It's apparent with swamps due to the coloring they have.
But stuff like that is more of an imperfection of a feature that otherwise works very well (in my opinion) for FINALLY letting us bring old worlds up to new versions. The only drawback is this wasn't backported, so if you want to update to a specific version before 1.18, you're still left dealing with the transitions of course.
On maps it looks rather apparent. Here's my oldest world originally, and then after updating.
The "pale" chunks are the 1.6 and prior generation, the rest are new ones. I trimmed my world on a region basis so it gave it plenty of space to work with.
And I can't comment much on performance. 1.18+ is more performance heavy already when generating terrain, largely since the underground is deeper. It also does updates to old chunks to extend their depth (checks for bedrock, turns it to deep slate, and then generates new deeper terrain below that). So I can't say how much of that performance impact is from terrain blending alone. Whatever it is, it's worth it in my opinion.
I'm really impressed Mojang could do the junctions so well. The problem with junctions is that they're the average of two different systems, so the variance drops and they look "flat". But either they cooked up a solution or it's not as much of a problem with 3D perlin. They probably should have "fractalized" the boundaries by generating additional terrain under old rules in varying widths along the edge. But still nice.
If, as i suspect, they're doing the terrains with some kind of 2D perlinesque noise for climate parameters, they could have smoothed out those awkward climate junctions by inferring climate parameters from the old biomes and averaging *those* over longer distances. Wouldn't help with rivers though.
Geographicraft (formerly Climate Control) - Control climate, ocean, and land sizes; stop chunk walls; put modded biomes into Default worlds, and more!
Yeah, rivers and swamps (due to coloring) are the ones that stand out most. It's definitely not perfect beyond that, but it's very good regardless. I'd even say the uniqueness of it adds to some variety during these transition areas versus the old terrain alone or new terrain alone. When I originally updated, I cut out a lot of land. I then later went back and cut more regions out, and it was interesting to see how the blending adapted in different circumstances.
And if the game goes to put a village or structure on the edge, you can get some VERY interesting results. I have a desert temple partially underwater next to a mangrove swamp (old ocean went to blend into mangrove swamp, but old ocean would be desert in new terrain generator, so that four chunk border must have been a valid place for the desert temple despite being shallow ocean).
It's interesting to watch and see what it does at times. The imperfections have their own character, or as I call them, "scars of an old world".
Oh, and caveat, this is all referring to above ground. I don't think caves tend to line up as well (I've found they just end in a sudden flat wall at times), but I don't think most of the people who fell into the spot of objecting from updating due to unsightly borders will find that to be something that keeps them from updating.
Anyway, looking forward to your world updates. I just wanted to mentioned that, and if you do update your world now that you know of this, it'd be interesting to see your impressions (and perhaps what "scars" the terrain blending may result in with your world).
Most of the increase in resource usage in newer versions has to do with how Mojang decided to start coding the game in 1.8, with absolutely no regard to code complexity or object creation; there is absolutely no reason why the default memory allocation needs to be 2 GB, and mods are unimaginably worse (I see people saying to allocate up to 8 GB(!) on modern modpacks) - I've added at least a thousand new features to TMCW yet haven't seen any increase in memory or CPU usage at all, if anything I've decreased them with optimizations and bug fixes. I haven't actually run newer versions but I've had people make statements like this:
A comment made by the creator of Optifine about 1.8; I can only imagine how bad it must be by now:
This comment left on a bug report claims to have doubled performance by caching blockstate/blockpos objects (which is still slower than not having them at all; my own "blockstate" system uses ints which correspond to a block ID + data value * 256):
For comparison, this is a memory/CPU profiler chart while running at max settings (Vsync at 75 FPS); memory usage slowly craws upwards by about 100 MB over about 40 seconds; also shown is a listing of the top objects by retained size (an approximation of the memory they use, including anything they reference); chunks dominate by far, as I'd expect for a game like Minecraft, and why I simply cannot believe that even the largest modpacks would need so much more memory (I'm also of the opinion that if a mod is adding like 20 different types of e.g. tools and armor it is doing something very wrong; enchantments and attributes let you have dozens to hundreds or more variants of a single item, and how do you manage dozens of types of ores?
Also, a deeper underground should not place such a heavy demand on the game; my first computer had hardware from the mid-2000s and 1.8 was crippling, yet I played with modding the underground to be up to 128 layers deeper ("triple height terrain") along with caves scaled up to match, and didn't have much issue playing ("Advanced OpenGL" / hardware occlusion queries worked quite well on my old system, notably, the example of 1.8, which replaced AOGL with a custom CPU-based culling method, seems to show not much culling is being done with 130 sections still being rendered despite being underground, while AOGL was much more aggressive), and it didn't have the optimizations I've since made to the game (TMCW actually generates terrain twice as fast as vanilla 1.6.4 despite being more complex than 1.19, if not as deep (I much prefer horizontal exploration while caving), and only runs on a single thread (server-side, plus a file-I/O thread, but only chunk saving is threaded):
1.13.2 (the latest version I've run); took 15 seconds, a lot of it on something other than terrain generation (preparing spawn area):
TMCWv5; took 3 seconds:
Also, the game is highly wasteful by generating noise with a precision far beyond what is necessary - I removed the first 6 octaves (or about 1/3) of noise with virtually no effect on terrain, while giving a significant performance improvement, furthered by not generating 3D noise for the lower 40 layers (the deepest oceans are around y=45) - in all, TMCW is about 6 times faster at generating noise than vanilla 1.6.4:
I also did something similar in double/triple height terrain, which raised normal terrain up by 64 or 128 layers and filled in the space below with solid stone and a layer of bedrock (actually, the lower 160 layers in THT, since again, oceans don't go all the way down to y=0; the minimum depth in vanilla seems to be around y=40, in TMCW it is a bit less, 45, with less height variation so it doesn't become exposed so the average depth is similar, if not greater. The limit of y=55 for rivers is because I decreased their "minHeight" (base height) to ensure they can carve though even the most extreme terrain; most do reach their minimum depth, with a 1 block variation added so they aren't uniform, this also makes them much wider than vanilla rivers, which are often a sorry sight in 1.6.4, even in a plains biome).
Also, the slowest part of cave generation is all the calls it makes to "Random", which is extremely slow, and this was responsible for most of the impact of my earlier mods, especially DHT/THT - simply subclassing it to not use "AtomicLong" halved the time taken by vanilla cave generation, with my own true 64 bit RNG being even faster by a factor of about 8, if with some limitations on the range of "nextInt(n)" which don't matter for my needs (Random also only uses the lower 48 bits, which caused seeds with the same lower 48 bits to match, except where features were biome-dependent as it used a custom 64 bit RNG; only since 1.18 has vanilla used a custom 64 bit RNG for all world generation (or almost all). I even went to far as to use a very basic 32 bit LCG for the RNG in individual features, where it is not necessary to have such a long sequence or that many unique states.
Regarding 1.8, it is also remarkable how much larger it became compared to older versions - and even with all the content I've added to 1.6.4 it is still significantly smaller (deleting META-INF does reduce the size, on the other hand, a lot of my code completely replaces vanilla classes with custom ones so the originals are sitting there doing nothing; in terms of the source code my source is about as large as vanilla 1.6.4, but only has 1/3 as many classes, so on average they are 3 times larger):
I've also added a lot of new textures and sounds but not as much as if every block/variant has a unique texture; there are 168 variants of stalagmites/stalactites but they only need 48 textures since they are recolored and/or flipped (all 17 hardened/stained clay variants are recolored to match the block they are placed on, defaulting to hardened clay; stalactites are flipped vertically. There are 7 "real" variants of each of 3 sizes, as defined by the obtainable items, and I count them as 21 blocks/items):
TheMasterCaver's First World - possibly the most caved-out world in Minecraft history - includes world download.
TheMasterCaver's World - my own version of Minecraft largely based on my views of how the game should have evolved since 1.6.4.
Why do I still play in 1.6.4?