DeanF Posted Wednesday at 06:57 PM Report Posted Wednesday at 06:57 PM I find that the deserts in VS aren't really deserts. They are too much like a random biome from that other game, just "an area with sand" or "an area with gravel". Maybe I just haven't traveled far enough? If so, let me know. Anyway, since VS has such huge maps I think it would be interesting (and preferable) to instead have HUGE deserts across the desert latitudes, approximately 15-35 degrees. They would be actual obstacles, creating challenges to cross. 2
williams_482 Posted Wednesday at 06:58 PM Report Posted Wednesday at 06:58 PM What do you mean by them not "really" being deserts? What are you expecting which isn't there, or vice versa?
LadyWYT Posted Thursday at 01:46 AM Report Posted Thursday at 01:46 AM What exactly do you consider a "desert latitude"? Because those large swaths of dry land filled with sand and gravel are deserts, and to my knowledge it's not the latitude that qualifies a region as a desert, but rather the annual rainfall. If you're referring to the stereotypical desert that's presented in pop culture--ie, lots of sand with a few cacti and cow skulls/bones lying around--you'll need to travel much further south to the warmer tropical zones in order to encounter those. 6 hours ago, DeanF said: Anyway, since VS has such huge maps I think it would be interesting (and preferable) to instead have HUGE deserts across the desert latitudes, approximately 15-35 degrees. They would be actual obstacles, creating challenges to cross. You'd need to change world generation entirely for this to happen, since the current settings don't allow any land feature to be so massive that it takes days to cross. At least to my knowledge. Also keep in mind that a land feature/biome that big sounds interesting...until you get spawned in the middle of such a biome, or otherwise have to actually deal with them on the regular. For particular kinds of games, that kind of challenge could be rather fun, but I don't think that kind of game is what the average player prefers to play, and likely why the devs have balanced the current world generation as they have.
Mac Mcleod Posted Friday at 08:39 PM Report Posted Friday at 08:39 PM On 10/21/2026 at 1:57 PM, DeanF said: I find that the deserts in VS aren't really deserts. They are too much like a random biome from that other game, just "an area with sand" or "an area with gravel". Maybe I just haven't traveled far enough? If so, let me know. Anyway, since VS has such huge maps I think it would be interesting (and preferable) to instead have HUGE deserts across the desert latitudes, approximately 15-35 degrees. They would be actual obstacles, creating challenges to cross. I understand you to mean that on a 640,000 block map the two desert bands at +20 to +30 and -20 to -30 degrees would be 64k blocks from the northern to southern edge. So if I understand you... you are suggesting roughly... Size Climate Zone / Region 32k North pole 40k Cold northern <- where we seem to start by default 96k Temperate northern 64k Desert northern 176k Tropical / equatorial 64k Desert southern 96kTemperate southern 40k Cold southern 32k South pole 640k Total These bands roughly match Earth but don't account for things like the AMOC (Gulf Stream). VS does seem to have a north and south poles and a hot equatorial region along with realistic geology generation. Warm, moist air rises near the equator, loses moisture as rain, then moves poleward and sinks around 20°–30° latitude. This descending air warms, suppressing clouds and rainfall, creating deserts. It appears the VS world is a sphere. So it doesn't seem unreasonable to me to have the dry desert zones. Do I understand your position?
DeanF Posted 6 hours ago Author Report Posted 6 hours ago (edited) Sonoran Desert: On 10/21/2026 at 12:58 PM, williams_482 said: What do you mean by them not "really" being deserts? What are you expecting which isn't there, or vice versa? Size, and thus challenge. It would be nice if deserts were large, and thus real obstacles to contend with- you would have to go through or around them (probably by sea). Think of the Sahara. And for a real challenge, start in one. It would be nice if rivers- when those get implemented- might have lush floodplains in deserts, and there were occasional oases, and sky islands, etc. And detailed desert flora and fauna. I absolutely love sky islands- the real ones, I don't just mean floating blocks. (And Minecraft does have the real ones, nowadays...) Davis Mountain, a sky island in Texas: Again, I have only played in temperate zones and am just now getting around to wide exploring, so if I'm wrong let me know. But the current small patches of sand or gravel seem more like the very unrealistic Minecraft biome system. But VS seems to at least make motions towards a bit of realism. The map is 1000x1000 km, with four climate bands: arctic, temperate, warm, and tropical. (Right?) Mind you, we could still have those gravel and sand patches- many of the gravel ones are covered in pine trees, similar to real environments in temperate areas, and things like the Great Sand Dunes in Colorado do exist. But the real deserst should be a band between 15 to 35 degrees, and those huge dunes in other areas should be relatively uncommon. Great Sand Dunes National Park: GSDNP is one of my favorites. I used to live nearby, and it and the San Luis Valley and Sangre de Crusto mountain range were sort of my stomping grounds. On 10/21/2026 at 7:46 PM, LadyWYT said: What exactly do you consider a "desert latitude"? Because those large swaths of dry land filled with sand and gravel are deserts, and to my knowledge it's not the latitude that qualifies a region as a desert, but rather the annual rainfall. Ineed, and if you look at a globe you will see IRL that those areas of low annual rainfall tend to happen from 15 to 35 degrees (both north and south), usually starting on the west side of continents and spreading eastwards until it either hits a weather-driving mountain range or the eastern shore. (There are also arctic deserts, but that's not what I'm talking about.) This is because of Hadley cells, which are an effect of the Earth's size and rotation: The equator heats the air. Strong sunlight warms the tropical oceans and air. Warm, moisture-laden air rises, creating a belt of low pressure. As it rises, it expands and cools, and water vapor condenses into clouds and heavy rainfall. This is why equatorial regions tend to be wet. The air moves poleward high in the atmosphere. Having lost much of its moisture through rainfall, the air spreads north and south, away from the equator. It cools as it moves, eventually becoming dense enough to descend, typically around 20–35° latitude. The air sinks and warms. Descending air is compressed by increasing atmospheric pressure and warms. Its relative humidity falls, clouds struggle to form, and rainfall is suppressed. This creates the subtropical high-pressure belts in which many of the world's great deserts occur. The air travelling along the surface back towards the equator- skewed a bit by the Ciroiolis efect- is what causes the Trade Winds. And air travelling towards the poles causes the Prevailing Westerlies. (And at about 30 degrees are the Doldrums, with little wind.) That's why deserts tend to form on west coasts- the east coast at least gets a little moisture that the winds have picked up over the oceans, and those east-coast oceans tend to be warm currents, which release more moisture. Contrarily, there tend to be cold ocean currents along west coasts centered around 30 degrees, and cold water releases less moisture, thus deserts form on west coasts. These currents are also an effet of the Earth's rotation (See: Humboldt Current, Benguela Current, California Current, etc.) These effects- wind and currents- reinforce one another to cause these deserts. The Sahara-Arabian Desert, the Desert Southwest of North America, the Atacama of South America, Western Australia, the Namib-Kalahari of southern Africa- they all follow this model. The Sahara-Arabian desert ends at the Himalayas, and the Atacama at the Andes. The others just peter out as they approach the east coast. Namib: There are other factors, of course. The Earth's rotation and prevailing surface winds also affect this, and can nudge the desert band around a bit. And large mountains can also cause deserts- the Gobi and Taklamakan are almost entirely due to the rainshadow effect of the Himalayas, for instance, and the North American interior is at least semi-arid because of the very long span of the Rockies and Sierra Nevada, which also creates a rainshadow. )And Eurasia is just so immensely huge that it's interior gets little moisture.) But the 15 to 35 degree arid band due to currents and Hadley cells is the general trend. Explanation- I got seriously into realistic climate models during a worldmaking phase a couple of decades ago. (Cartographersguild.com has great resources.) It's a rabbithole, and I'm not proposing anything like that for VS. But the devs do seem to like the idea of large climate zones, so a desert could be one of them. On 10/21/2026 at 7:46 PM, LadyWYT said: You'd need to change world generation entirely for this to happen, since the current settings don't allow any land feature to be so massive that it takes days to cross. At least to my knowledge. I had actually thought of this already, and it may not be as involved as it first seems. As I mentioned, the devs already have large climate zones where temperature and rainfall vary. And this then drives the local flora. The desert band would just have rainfall curtailed, then it would probably generate automagically. And decent desert flora and fauna would be nice. On 10/21/2026 at 7:46 PM, LadyWYT said: Also keep in mind that a land feature/biome that big sounds interesting...until you get spawned in the middle of such a biome, or otherwise have to actually deal with them on the regular. For particular kinds of games, that kind of challenge could be rather fun, but I don't think that kind of game is what the average player prefers to play, and likely why the devs have balanced the current world generation as they have. Well, we get to select where we spawn. I would propose five climate bands instead of four, since the desert lattitudes would act as one: Arctic Temperate Warm Desert Tropical I think right now we have four, right? Arctic, temperate, warm, and tropical? I could be mis-remembering- I have never spawned anywhere but temperate. And yes, the challenge is one feature that I already mentioned above. If you want a real challenge then spawn in the desert climate band! Everyone likes messing about in deserts- they seem "exotic". Unless you live in one, of course, as I do... And if you do want to go down the climate rabbithole then the desert band should have a narrow band of savannah around it- north, south, and east. The Maasai Mara, a savannah: On 10/23/2026 at 2:39 PM, Mac Mcleod said: I understand you to mean that on a 640,000 block map the two desert bands at +20 to +30 and -20 to -30 degrees would be 64k blocks from the northern to southern edge. No, not unless I am misunderstanding the VS map. I think it only covers the northern hemisphere, right? With arctic at the northern end (pole) and tropical at the southern end (equator)? So we would only see the 15 to 35 degree north desert band, between the Warm and Tropical climate bands. Am I wrong? Does the VS map go to both poles? I could easily be wrong- as I said, I am just now starting serious exploring. Edited 2 hours ago by DeanF 1
MKMoose Posted 2 hours ago Report Posted 2 hours ago (edited) On 10/23/2026 at 10:39 PM, Mac Mcleod said: I understand you to mean that on a 640,000 block map the two desert bands at +20 to +30 and -20 to -30 degrees would be 64k blocks from the northern to southern edge. A 640k world is absolutely humongous. The default polar-equator distance is 100k, so I see no reason to change that. Aside from this, the proportion between these climatic bands seems generally reasonable. On 10/22/2026 at 3:46 AM, LadyWYT said: Also keep in mind that a land feature/biome that big sounds interesting...until you get spawned in the middle of such a biome, or otherwise have to actually deal with them on the regular. For particular kinds of games, that kind of challenge could be rather fun, but I don't think that kind of game is what the average player prefers to play, and likely why the devs have balanced the current world generation as they have. It's already a huge dice toss in terms of where the game spawns you. I had one 200x200 island spawn where I had clay, chalk, copper, cassiterite, iron, goats, sheep, pigs, chickens, four fruit tree species and six crops including flax, among other things, all on that tiny piece of land. And I had plenty of spawns in the middle of a granite wasteland with very little in terms of food or valuable resources within some thousand or more blocks in each direction - not to mention the empty island spawns. I've seen a person wander for two hours without finding any cattails. The solution is to control the spawn area better or to introduce alternative resource sources and less linear progression, not to just avoid features which could make the already-existing problem potentially worse. Not to mention, this change may actually help to avoid frustrating or difficult spawn locations, because the assumption that the player could just randomly get spawned in a huge desert is not necessarily justified. If this desert is only confined to a specific rough climate band, then the player would only spawn in the desert proper if they choose that spawn location, which could be appropriately advertised as difficult during world creation. 3 hours ago, DeanF said: Again, I have only played in temperate zones and am just now getting around to wide exploring, so if I'm wrong let me know. But the current small patches of sand or gravel seem more like the very unrealistic Minecraft biome system. But VS seems to at least make motions towards a bit of realisms. The map is 100x100 km, with three climate bands: arctic, temperate, and tropical. Overall correct, with the caveat that there aren't really any strict climate bands and it's more of a continuous gradient which ends up mostly producing five rough regions: polar (a lot of glaciers), cold (basically temperate but worse), temperate (the baseline), warm (redwoods, bald cypress and sandy areas), and hot (acacias, tropical forests). Besides this, it's just a repetitive small-scale patchwork of dry, grassy and forested areas. Looking at something like the TOPS map can give you a good idea about how the world actually looks at large scale. I would note that excessively rigid climatic strips (and especially large, uniform strips) would also likely be quite boring. There is a number of additional factors like ocean currents, continentality, air circulation, water systems and mountain ranges which can greatly influence climate in much more varied and interesting ways than just latitude. Even relatively simplistic climatic approximation could be much better than . I tend to refer to the Köppen climate classification and OneEarth ecoregions for a general overview of Earth's climates. I don't really have any good references for more details regarding climate-shaping mechanisms at the moment, though. 3 hours ago, DeanF said: But the current small patches of sand or gravel seem more like the very unrealistic Minecraft biome system. On this specifically, there is a kind of debate between Minecraft-like and VS-like climate and biome generation, and I would actually argue that Minecraft gets closer to reality in principle in certain aspects. Minecraft defines unique biomes based on climatic conditions first, which then determines the biological makeup of the region, whereas VS only allows defining broad matching conditions for each plant individually and the emergent patterns can be described as biomes, so the causality in the two approaches is directed in opposite ways. Realistically, very similar climatic conditions can still produce different biomes based on a large variety of additional factors - Minecraft largely reflects this by adjusting the entire world generation depending on biome, so that each small region always follows a cohesive formula, whereas in VS a lot of world parameters are independent, which can feel much more samey and indistinct as well as sometimes contradictory. Neither is perfect and both can work fine, naturally, and in-between solution may provide the best results. In that vein, some things which I personally would like to see in the long term. Spoiler because it's a bit of a long tangent to an already excessively long post. Spoiler Increase the variety of environmental conditions to also include factors like: water retention and runoff, continentality and coastal proximity (would likely benefit from oceans being reworked), air circulation (broad zones based on latitude), potentially also longitude. Allow environmental conditions to influence terrain in more ways, for example: sedimentary rocks => flatter terrain, limestone => more caves, high water runoff => reduced soil thickness (this could be used to make slopes and mountains more rocky in a natural way). Create secondary environmental parameters which depend on other conditions, for example: mountain ranges and wind => rain shadow, oceans => water current temperature, air circulation, rain shadow and water current temperature => annual precipitation and rainfall seasonality, Finally, create parameters which roughly correspond to regions a lot like it would be done realistically. This is not strict biome classification like in Minecraft, but instead an easer way to define broad regions without having to use hyperspecific parameter sets for all plants which are supposed to occur in roughly the same areas. There's, of course, a lot of ways to make this more flexible or varied. This could include parameters like: high precipitation, no runoff => bog, low precipitation, high temperature => arid/desert (this is your large hot desert - since air circulation has a lot of influence over precipitation, this naturally results in deserts at appropriate latitudes), low precipitation, low temperature => tundra. This would be a huge overhaul, and could easily be argued to be overkill, but I do regardless think that having a comprehensive system like this would be so much more convenient for designing worlds. A plant's spawn conditions could change from something like this: MinTemp: 28, MaxTemp: 40, MinRain: 80, MaxRain: 140, MinFert: 50, MaxFert: 200, MinForest: 50, MaxForest: 255, to a simple: MinSavanna: 80, Crucially, though, all of the lower-level underlying parameters would be still available - this is intended as additional tooling for creating distinct ecoregions and large-scale variety, not a full replacement for the current system. Edited 2 hours ago by MKMoose 1
DeanF Posted 2 hours ago Author Report Posted 2 hours ago (edited) 59 minutes ago, MKMoose said: Not to mention, this change may actually help to avoid frustrating or difficult spawn locations, because the assumption that the player could just randomly get spawned in a huge desert is not necessarily justified. If this desert is only confined to a specific rough climate band, then the player would only spawn in the desert proper if they choose that spawn location, which could be appropriately advertised as difficult during world creation. Reveal hidden contents Increase the variety of environmental conditions to also include factors like: water retention and runoff, continentality and coastal proximity (would likely benefit from oceans being reworked), air circulation (broad zones based on latitude), potentially also longitude. Allow environmental conditions to influence terrain in more ways, for example: sedimentary rocks => flatter terrain, limestone => more caves, high water runoff => reduced soil thickness (this could be used to make slopes and mountains more rocky in a natural way). Create secondary environmental parameters which depend on other conditions, for example: mountain ranges and wind => rain shadow, oceans => water current temperature, air circulation, rain shadow and water current temperature => annual precipitation and rainfall seasonality, Finally, create parameters which roughly correspond to regions a lot like it would be done realistically. This is not strict biome classification like in Minecraft, but instead an easer way to define broad regions without having to use hyperspecific parameter sets for all plants which are supposed to occur in roughly the same areas. There's, of course, a lot of ways to make this more flexible or varied. This could include parameters like: high precipitation, no runoff => bog, low precipitation, high temperature => arid/desert (this is your large hot desert - since air circulation has a lot of influence over precipitation, this naturally results in deserts at appropriate latitudes), low precipitation, low temperature => tundra. This would be a huge overhaul, and could easily be argued to be overkill, but I do regardless think that having a comprehensive system like this would be so much more convenient for designing worlds. A plant's spawn conditions could change from something like this: MinTemp: 28, MaxTemp: 40, MinRain: 80, MaxRain: 140, MinFert: 50, MaxFert: 200, MinForest: 50, MaxForest: 255, to a simple: MinSavanna: 80, Crucially, though, all of the lower-level underlying parameters would be still available - this is intended as additional tooling for creating distinct ecoregions and large-scale variety, not a full replacement for the current system. A fair point. I would agree. 59 minutes ago, MKMoose said: Overall correct, with the caveat that there aren't really any strict climate bands and it's more of a continuous gradient which ends up mostly producing five rough regions: polar (a lot of glaciers), cold (basically temperate but worse), temperate (the baseline), warm (redwoods, bald cypress and sandy areas), and hot (acacias, tropical forests). Besides this, it's just a repetitive small-scale patchwork of dry, grassy and forested areas. Looking at something like the TOPS map can give you a good idea about how the world actually looks at large scale. Reveal hidden contents Increase the variety of environmental conditions to also include factors like: water retention and runoff, continentality and coastal proximity (would likely benefit from oceans being reworked), air circulation (broad zones based on latitude), potentially also longitude. Allow environmental conditions to influence terrain in more ways, for example: sedimentary rocks => flatter terrain, limestone => more caves, high water runoff => reduced soil thickness (this could be used to make slopes and mountains more rocky in a natural way). Create secondary environmental parameters which depend on other conditions, for example: mountain ranges and wind => rain shadow, oceans => water current temperature, air circulation, rain shadow and water current temperature => annual precipitation and rainfall seasonality, Finally, create parameters which roughly correspond to regions a lot like it would be done realistically. This is not strict biome classification like in Minecraft, but instead an easer way to define broad regions without having to use hyperspecific parameter sets for all plants which are supposed to occur in roughly the same areas. There's, of course, a lot of ways to make this more flexible or varied. This could include parameters like: high precipitation, no runoff => bog, low precipitation, high temperature => arid/desert (this is your large hot desert - since air circulation has a lot of influence over precipitation, this naturally results in deserts at appropriate latitudes), low precipitation, low temperature => tundra. This would be a huge overhaul, and could easily be argued to be overkill, but I do regardless think that having a comprehensive system like this would be so much more convenient for designing worlds. A plant's spawn conditions could change from something like this: MinTemp: 28, MaxTemp: 40, MinRain: 80, MaxRain: 140, MinFert: 50, MaxFert: 200, MinForest: 50, MaxForest: 255, to a simple: MinSavanna: 80, Crucially, though, all of the lower-level underlying parameters would be still available - this is intended as additional tooling for creating distinct ecoregions and large-scale variety, not a full replacement for the current system. Good to know. TOPS looks like it's a map of someone's long-standing server? That's nice-looking. I had found a different VS map utility, which looks terrible but it does let you generate any seed that you like. 59 minutes ago, MKMoose said: I would note that excessively rigid climatic strips (and especially large, uniform strips) would also likely be quite boring. There is a number of additional factors like ocean currents, continentality, air circulation, water systems and mountain ranges which can greatly influence climate in much more varied and interesting ways than just latitude. Even relatively simplistic climatic approximation could be much better than . Reveal hidden contents Increase the variety of environmental conditions to also include factors like: water retention and runoff, continentality and coastal proximity (would likely benefit from oceans being reworked), air circulation (broad zones based on latitude), potentially also longitude. Allow environmental conditions to influence terrain in more ways, for example: sedimentary rocks => flatter terrain, limestone => more caves, high water runoff => reduced soil thickness (this could be used to make slopes and mountains more rocky in a natural way). Create secondary environmental parameters which depend on other conditions, for example: mountain ranges and wind => rain shadow, oceans => water current temperature, air circulation, rain shadow and water current temperature => annual precipitation and rainfall seasonality, Finally, create parameters which roughly correspond to regions a lot like it would be done realistically. This is not strict biome classification like in Minecraft, but instead an easer way to define broad regions without having to use hyperspecific parameter sets for all plants which are supposed to occur in roughly the same areas. There's, of course, a lot of ways to make this more flexible or varied. This could include parameters like: high precipitation, no runoff => bog, low precipitation, high temperature => arid/desert (this is your large hot desert - since air circulation has a lot of influence over precipitation, this naturally results in deserts at appropriate latitudes), low precipitation, low temperature => tundra. This would be a huge overhaul, and could easily be argued to be overkill, but I do regardless think that having a comprehensive system like this would be so much more convenient for designing worlds. A plant's spawn conditions could change from something like this: MinTemp: 28, MaxTemp: 40, MinRain: 80, MaxRain: 140, MinFert: 50, MaxFert: 200, MinForest: 50, MaxForest: 255, to a simple: MinSavanna: 80, Crucially, though, all of the lower-level underlying parameters would be still available - this is intended as additional tooling for creating distinct ecoregions and large-scale variety, not a full replacement for the current system. I did mention all of that, unless I was editing while you were posting. But nonetheless, the 15 to 35 degree trend stands. A lot of what you are referring to is smaller-scale. I did mention Nile-like floodplains and sky islands, right? Love those. A large desert would still be varied- areas of stone, areas of dune, mountain ranges (some with sky islands), rivers (some day), differing vegetation depending upon just how arid it is, etc. 59 minutes ago, MKMoose said: I tend to refer to the Köppen climate classification and OneEarth ecoregions for a general overview of Earth's climates. I don't really have any good references for more details regarding climate-shaping mechanisms at the moment, though. Reveal hidden contents Increase the variety of environmental conditions to also include factors like: water retention and runoff, continentality and coastal proximity (would likely benefit from oceans being reworked), air circulation (broad zones based on latitude), potentially also longitude. Allow environmental conditions to influence terrain in more ways, for example: sedimentary rocks => flatter terrain, limestone => more caves, high water runoff => reduced soil thickness (this could be used to make slopes and mountains more rocky in a natural way). Create secondary environmental parameters which depend on other conditions, for example: mountain ranges and wind => rain shadow, oceans => water current temperature, air circulation, rain shadow and water current temperature => annual precipitation and rainfall seasonality, Finally, create parameters which roughly correspond to regions a lot like it would be done realistically. This is not strict biome classification like in Minecraft, but instead an easer way to define broad regions without having to use hyperspecific parameter sets for all plants which are supposed to occur in roughly the same areas. There's, of course, a lot of ways to make this more flexible or varied. This could include parameters like: high precipitation, no runoff => bog, low precipitation, high temperature => arid/desert (this is your large hot desert - since air circulation has a lot of influence over precipitation, this naturally results in deserts at appropriate latitudes), low precipitation, low temperature => tundra. This would be a huge overhaul, and could easily be argued to be overkill, but I do regardless think that having a comprehensive system like this would be so much more convenient for designing worlds. A plant's spawn conditions could change from something like this: MinTemp: 28, MaxTemp: 40, MinRain: 80, MaxRain: 140, MinFert: 50, MaxFert: 200, MinForest: 50, MaxForest: 255, to a simple: MinSavanna: 80, Crucially, though, all of the lower-level underlying parameters would be still available - this is intended as additional tooling for creating distinct ecoregions and large-scale variety, not a full replacement for the current system. Yes, I'm familiar. As I mentioned, I went through climate-pedant worldbuilding phase a while ago. I got far too obsessive about it. I tend to prefer Holdridge life zone classifications rather than Koppen. It gives you more of a "ground reality" than the mere temp/rainfall of Koppen- it tells you what sort of land you are notionally walking through. But a lot of software tools still rely upon Koppen, unfortunately. That even looks good as a color scheme for a map! 59 minutes ago, MKMoose said: On this specifically, there is a kind of debate between Minecraft-like and VS-like climate and biome generation, and I would actually argue that Minecraft gets closer to reality in principle in certain aspects. Reveal hidden contents Increase the variety of environmental conditions to also include factors like: water retention and runoff, continentality and coastal proximity (would likely benefit from oceans being reworked), air circulation (broad zones based on latitude), potentially also longitude. Allow environmental conditions to influence terrain in more ways, for example: sedimentary rocks => flatter terrain, limestone => more caves, high water runoff => reduced soil thickness (this could be used to make slopes and mountains more rocky in a natural way). Create secondary environmental parameters which depend on other conditions, for example: mountain ranges and wind => rain shadow, oceans => water current temperature, air circulation, rain shadow and water current temperature => annual precipitation and rainfall seasonality, Finally, create parameters which roughly correspond to regions a lot like it would be done realistically. This is not strict biome classification like in Minecraft, but instead an easer way to define broad regions without having to use hyperspecific parameter sets for all plants which are supposed to occur in roughly the same areas. There's, of course, a lot of ways to make this more flexible or varied. This could include parameters like: high precipitation, no runoff => bog, low precipitation, high temperature => arid/desert (this is your large hot desert - since air circulation has a lot of influence over precipitation, this naturally results in deserts at appropriate latitudes), low precipitation, low temperature => tundra. This would be a huge overhaul, and could easily be argued to be overkill, but I do regardless think that having a comprehensive system like this would be so much more convenient for designing worlds. A plant's spawn conditions could change from something like this: MinTemp: 28, MaxTemp: 40, MinRain: 80, MaxRain: 140, MinFert: 50, MaxFert: 200, MinForest: 50, MaxForest: 255, to a simple: MinSavanna: 80, Crucially, though, all of the lower-level underlying parameters would be still available - this is intended as additional tooling for creating distinct ecoregions and large-scale variety, not a full replacement for the current system. I would definitely contest that. You say: 59 minutes ago, MKMoose said: Minecraft defines unique biomes based on climatic conditions first, which then determines the biological makeup of the region[...] Reveal hidden contents Increase the variety of environmental conditions to also include factors like: water retention and runoff, continentality and coastal proximity (would likely benefit from oceans being reworked), air circulation (broad zones based on latitude), potentially also longitude. Allow environmental conditions to influence terrain in more ways, for example: sedimentary rocks => flatter terrain, limestone => more caves, high water runoff => reduced soil thickness (this could be used to make slopes and mountains more rocky in a natural way). Create secondary environmental parameters which depend on other conditions, for example: mountain ranges and wind => rain shadow, oceans => water current temperature, air circulation, rain shadow and water current temperature => annual precipitation and rainfall seasonality, Finally, create parameters which roughly correspond to regions a lot like it would be done realistically. This is not strict biome classification like in Minecraft, but instead an easer way to define broad regions without having to use hyperspecific parameter sets for all plants which are supposed to occur in roughly the same areas. There's, of course, a lot of ways to make this more flexible or varied. This could include parameters like: high precipitation, no runoff => bog, low precipitation, high temperature => arid/desert (this is your large hot desert - since air circulation has a lot of influence over precipitation, this naturally results in deserts at appropriate latitudes), low precipitation, low temperature => tundra. This would be a huge overhaul, and could easily be argued to be overkill, but I do regardless think that having a comprehensive system like this would be so much more convenient for designing worlds. A plant's spawn conditions could change from something like this: MinTemp: 28, MaxTemp: 40, MinRain: 80, MaxRain: 140, MinFert: 50, MaxFert: 200, MinForest: 50, MaxForest: 255, to a simple: MinSavanna: 80, Crucially, though, all of the lower-level underlying parameters would be still available - this is intended as additional tooling for creating distinct ecoregions and large-scale variety, not a full replacement for the current system. But the very problem is that Minecraft sets those "climactic conditions" in a wholely arbitrary fashion. The way that VS has climate bands north-to-south is still simplistic but nonetheless MUCH better than what Minecraft does. Edited 1 hour ago by DeanF 1
MKMoose Posted 51 minutes ago Report Posted 51 minutes ago (edited) 2 hours ago, DeanF said: I had found a different VS map utility, but TOPS looks very good. I didn't think of a map viewer, frankly. That's useful. 2 hours ago, DeanF said: But nonetheless, the 15 to 35 degree trend stands. A lot of what you are referring to is smaller-scale. Fair, that's true for most of what I described, though air circulation specifically is very large-scale and is probably the most important cause aside from directly latitude for why climate is arranged roughly into stripes, and it's also why those hot deserts are where they are. Otherwise, yes, I'm mostly talking about relatively small-scale features (though still generally significantly larger than the current patches), because while having a huge desert is something I would certainly also be interested in, I would also want that huge desert to be more varied and interesting - if you were to just upscale the current deserts, then even increasing the biodiversity wouldn't really keep them from being quite samey. When rainfall, fertility, and forestation are kept to a minimum, there's not much variety that you can squeeze out of the current system, so iIn that case I'd love to see sources of variety based on more than just the current random maps. 2 hours ago, DeanF said: But the very problem is that Minecraft sets those "climactic conditions" in a wholely arbitrary fashion. The way that VS has climate bands north-to-south is still simplistic but nonetheless MUCH better than what Minecraft does. A different temperature distribution is the only difference between "realistic" and "patchy" climate setting in VS, but if you look at a small region in isolation, then that temperature variation with latitude basically doesn't matter. At a smaller scale, VS still just uses simple noise maps and based on that creating simple patches not too dissimilar from Minecraft's biomes. Adjusting Minecraft's temperature map to be similar to VS and vary primarily along the North-South axis would be a fairly trivial change and it would produce a more realistic climate distribution, but it has no impact on how this temperature and other parameters are translated into actual world generation. I'm primarily talking about how those worldgen parameters are interpreted and used later on when deciding what to place where. Minecraft assigns distinct biome labels which are then used to determine the terrain, fauna, flora and whatnot. This system can produce more distinct and varied biomes, so it's well-suited to Minecraft's more gamified world. This makes biome blending and smooth transitions significantly more difficult. This also means that each biome can be significantly more cohesive, because they are designed more explicitly instead of being a combination of largely independent factors. Increasing biodiversity by adding more biomes can end up very messy, but just improving the existing selection of biomes still keeps them cohesive internally and neatly self-contained. VS has to individually define suitable conditions for every single feature in the world. Terrain elevation in VS is also one of the input parameters and is mostly independent from all other factors (higher elevations have lower temperature, but some things only consider temperature at fixed elevation), and rock strata are completely independent of everything. This is a much more freeform system, which can be very flexible in some ways. In some respects, this is more realistic just because of the easy blending and smoothing between different biomes and regions. However, this also leads to many regions of the world being very indistinct - almost all of the VS world can be boiled down to just a couple biomes, with added large-scale variety to fauna and flora depending on latitude. In the current state, it is nearly impossible (or at least extraordinarily inconvenient) to create small regions of unique biodiversity (e.g. bogs or swamps), and it's even worse if you want to make them only occur in specific terrain landforms. Unique vegetation in oases and river deltas may also be very difficult unless new worldgen parameters are added. Increasing biodiversity with mods like Floral Zones or Fauna of the Stone Age, past a certain point, often just creates a mess out of the huge variety of plants and animals which get up mixed together pretty haphazardly in the same regions with no rhyme or reason. While this system can work just fine, what I'm suggesting is that it primarily just needs a greater variety of more advanced parameters to really bring out its strengths, not just a couple random maps plus the Y level. The reason why I say that Minecraft's biomes are more realistic in principle in certain respects primarily has to do with the fact that real-life biomes are not just created by semi-arbitrarily adding together this stone type, this soil fertility, these trees, these bushes, and adding other roughly matching plants and animals - biomes are distinct environments which are more than the sum of their individual parts. The presence of many species of fauna and flora is dictated not just by roughly matching climate conditions, but by how the overall region evolved over time into a distinct ecosystem. Compare for example the tropical forests of South America, Africa and Asia - they are similar at high level, but one is flooded, one is relatively dry, one is famous for its monsoons and isolated islands, and there are very few specific animal or plant species which occur in any two of these regions at the same time. Minecraft could trivially implement multiple distinct biomes easily, while VS would struggle to differentiate them unless a bunch of additional parameters are added to allow separating islands, monsoon-heavy regions and flooded forests from drier tropical regions. Granted, having multiple types of rainforests could be considered too granular, and this matters less when only using rough climate bands. Edited 22 minutes ago by MKMoose
Recommended Posts