I think old Minecraft is better, and that's what I play every day. 1.20 is bad, 1.19 is bad, 1.18 is bad, 1.17 is bad, 1.16 is ok, 1.15 is ok, 1.14 is ok, 1.13 is ok. These new updates take the simpleness of Minecraft out of Minecraft. Last update I would start a world on would be 1.12. Right now I am playing on 1.6 on the Hermitcraft II seed. Mojang is making Minecraft too easy, these hardcore players are getting diamond gear from villagers super easy. Back in the day you would have too go mining for diamonds but now you can get diamond gear and tools from villagers.
You can also get diamond gear from villagers in 1.6, and the trades are much cheaper - I've bought thousands of diamonds worth of gear in my first world, for the low cost of just 10 emeralds per diamond pickaxe, or 180 wheat (18 per emerald), exploiting the fact that if the final trade is something like wheat for emeralds you can trade it indefinitely without ever needing to trade something else to unlock it again (or the final trade, if not the main trade, is also something that gives you emeralds; such villagers were called "perfect villagers" (note the date - this was when 1.6 was the current version) and were removed in 1.8, and since 1.14 you can only trade a limited number of times per day - I've traded a double chest of wheat in one go before and only have two active villagers, a blacksmith and a farmer, plus backups for both).
The only possible downside is that more heavily enchanted items are impractical to repair with items, if at all, due to the way repair costs are calculated, but you can get around this by wearing down the sacrifice a bit, as I do for swords (a Sharpness V, Knockback II, Unbreaking III diamond sword costs 40 levels to repair with a brand new sword but by damaging it a bit, e.g. killing 3 stacks of chickens, the cost is lowered to 38 levels, enabling a full repair after adding a 12% durability bonus. Even better, those chickens can be traded for enough emeralds to buy a new sword, otherwise, you'd have to use 1-2 diamonds for 29-35 levels. Other items, like a maxed-out Fortune pickaxe, are much costlier, 37 levels for a single diamond, so this method is less practical (you have to wear the sacrifice down a lot more, something like 500-600 left will cost as much as one diamond but restore closer to twice the durability), but still avoids any need to mine diamonds, except for an enchantment table).
Note that I only trade for the items I need to repair my gear for fun and so I can hoard every diamond I mine (closing in on a double chest of diamond blocks, mined without Fortune), as even without using Fortune I get many times the diamonds I'd need thanks to my playstyle of nonstop caving (you could say that trading is a diversion from just caving all the time, and I literally do mean all the time as I only build basic bases to act as temporary storage and food/wood farms).
On the other hand, it is definitely much harder to get enchanted books, with a probability of only 1.75% per reroll of trades to get a single enchantment, which will also eventually be replaced with a lower-priced offer (usually meaning worse, unless it something like Silk Touch and the cost has bottomed out at 5 emeralds. I used this fact to more easily get Mending before I fixed it (by "Mending" I mean my own enchantment which works like renaming an item to keep the cost from increasing; the official enchantment is overpowered and broken because it lets you repair any item, no matter how heavily enchanted, for a flat rate and only requires XP. This is the only time I've ever traded to get an enchantment as it is just not worth it, given that anything else can be gotten from the table (you do need books to get Unbreaking on weapons and armor though so it can be very worth trading it, as opposed to say, mining thousands of quartz and enchanting hundreds of books, as I do, but I also use it to build my main base anyway).
Any version with beds, stackable food, sprinting or the ability to switch to Creative mode is a disgrace to the memes that where running around back when "Survival" meant something (so anything above beta 1.2).
The latest versions added so much quality of life and new things to do, don't get me wrong, but the feeling or living in a world that wanted you dead and you had to mine/craft your way to survival is totally gone now. It's like back then Minecraft had a proper early game but no late game (once you have diamond gear and a defendable base, you're done), and now it's the opposite because you go straight to late game in a matter of hours. Kind of bizarre from a game design perspective.
If you're trying to define eras, the groupings seem a bit strange.
Anyway, I would find it hard to ever go back to pre-1.18 unless it was on the side. Like maybe I could have fun starting a modded world on an older version for a different experience, or doing a beta 1.7.3 world for a similar reason, but going back to a version before 1.18 as the highest I'd play on and staying there? I wouldn't enjoy it.
Sure, I have nostalgic love for the older versions. I also have qualms with the new versions in some regards, but I think the game is objectively in a better place overall.
I'm choosing 1.18 or above every time. It's the "foundation" I now need. I can't go back to the missing content and other changes from versions before that now. Even 1.13 through 1.16 feel aged and that is saying something (especially since 1.16 isn't that old and is still a fine version, and perhaps what most would consider the best single update Mojang has done so far), and 1.12 and older feel downright ancient. 1.18 is the foundation version I must have. It's my personal single favorite update (in contrast to 1.8, being my least favorite). And 1.19 has everything 1.18 does, and then some. And 1.20 has everything 1.19 does, and then some. Therefore, 1.20 is my current favorite version by extension, but 1.18 is my favorite update and the minimum I could live with. Going back to before that now would be rough.
If you're trying to define eras, the groupings seem a bit strange.
I'm choosing 1.18 or above every time. It's the "foundation" I now need. I can't go back to the missing content and other changes from versions before that now.
If you think current is best, you should vote that way. Though I agree the groupings are strange - 1.16 is by far the most popular version between 1.12 and "current as possible", both in servers and mod supply, so that category should be 1.13 - 1.16 or 1.17.
How important is the terrain generation in your evaluation? How would you like an older version with greatly improved terrain generation?
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 important is the terrain generation in your evaluation? How would you like an older version with greatly improved terrain generation?
Rather important, I would say. So much so that whenever one has happened, it's when I feel one era ends and another begins. That's not to say it would always make me feel an era shift has occurred (as always, it's all in the details), but thus far it always has.
These are what I feel were "major era shifts" (there's some minor sub-eras within these too, namely between the 1.7 to 1.12 era which can honestly be broken into three sub-eras).
Beta 1.7.3 to beta 1.8.
Release 1.6 to release 1.7.
Release 1.12 to release 1.13.
Release 1.16 (1.17) to release 1.18.
Terrain generation changed markedly in three of those four splits I see as era changing. Release 1.12 to release 1.13 is the exception, and that sort of served as the foundation for modern updates and what would collectively be the era 1.16 served as a capstone to.
There's other things I think that warrant changes in eras. An important one would be something that radically changes the "meta" of the game. Think all of the changes between beta 1.7.3 and beta 1.8 (hunger system, sprinting, enchanting, potions, etc.), the new combat of 1.9, and addition of elytra and mending (these formally happened in 1.9, but it was more the addition of rocket boosting in 1.11 that made elytra meta changing, and 1.14s villager changes that made mending more common, so those are more muddy to pin down and I'd say actually happened after 1.9 itself).
The biggest collective shift seems to have been beta 1.7.3 to beta 1.8. New terrain generation, and many new systems.
1.16 is by far the most popular version between 1.12 and "current as possible", both in servers and mod supply, so that category should be 1.13 - 1.16 or 1.17.
https://minecraft-statistic.net/en/global_statistic.html has always shown the latest version to be the most popular by far, at least as far as the number of servers goes (and of this, I assume those tracked by the site); currently 1.20.1 is in the lead with 36.4% of all servers; 1.16.5 is far behind with only 1/10 as many servers, with 1.12.2 still being about even with it, and the once mighty modding version of 1.7.10 has been left in the dust (even recently I've seen people arguing that it was still the most popular modded version, caveats with this data aside, including the fact it only tracks major(?) multiplayer servers, not singleplayer or small home servers(?), as also reflected by its claimed active player counts vs what https://activeplayer.io/minecraft/ shows).
Also, this seems to show that the chat reporting added in 1.19.1 had no impact on updating to newer versions (I've seen mods/plugins that disable it and many 1.19.1+ servers may be using one, at the same time, 1.9 seemed so hard to work around in a similar manner). Likewise, the upcoming changes to villager trading is still only an optional feature so there should be no reluctance to update to 1.20.2.
(wondering how many servers it lists for 1.6? None, not even a single one, but that doesn't mean there aren't any, nor does it list Beta versions like 1.7.3, which I know have a sizeable playerbase, much greater than 1.6)
I would expect that most of the 1.12.2 and 1.16.5 servers on that report are modded; I can't see why they'd be so much more popular than 1.13 or 1.17 other than for the mods. I agree it seems impossible there are NO Beta servers whatsoever, but I'd want some data to back up a claim than 1.7.10 is more popular for mods *now* (as opposed to the most popular modded version over history, which I'd find plausible.) Both the base game and mods tend to add features over time and you give up a lot to go with older version. There is a huge supply of 1.7.10 mods, of course, but many have been abandonware for a very long time and that introduces its own problems in the form of unfixed and unfixable bugs.
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!
Both the base game and mods tend to add features over time and you give up a lot to go with older version. There is a huge supply of 1.7.10 mods, of course, but many have been abandonware for a very long time and that introduces its own problems in the form of unfixed and unfixable bugs.
There is no reason though why somebody can't make a mod that backports newer content (looks at myself), and in fact, I know of at least one such mod that has been continued across multiple developers (I even used some of its open-source code to implement newer features):
This mod mostly focuses on backporting content while replicating the original as closely as possible and not adding much non-vanilla content, as the majority of my own mod is (nor are the features I backported always accurate to fit with my views of how they should work, e.g. Mending as a way to make the "renaming negates prior work penalty" bug legitimate and not so easy to get; my "attack cooldown" specifically penalizes spam-clicking without nerfing attack speed/DPS for "proper" attacks. I've also backported developmental features, like endermen spawning endermites when they teleport and naturally spawning killer rabbits,which also use their original texture and name, and nerfing "mineral" drops to require a player kill (especially iron golems), all of which Mojang actually implemented in 1.8 snapshots).
There is no reason though why somebody can't make a mod that backports newer content (looks at myself), and in fact, I know of at least one such mod that has been continued across multiple developers (I even used some of its open-source code to implement newer features):
In the abstract, no, but there's just too much content for one person to backport. Note said project isn't done. If somebody could organize a team that would work; it seems kinda worth it because such a project would make it possible for ALL modders to work on that version and not face the endless Treadmill Of DOOOM! of trying to keep up the endless stream of updates coming out. If every modder who'd done something substantive on any of the "refuge" mod versions (1.7.10, 1.12.2, 16.5) had backported just *one* current feature we'd be done.
A technical problem is that Forge doesn't support old versions, which caused me quite a bit of trouble trying to get a modding setup working again. Dunno if it would be possible to get LexManos in on a "backport everything" project.
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 would you like an older version with greatly improved terrain generation?
I forgot to answer this when answering the part before it.
Without knowing what "greatly improved terrain generation" means, or what "old version" we're talking about, it's hard to say. I feel the terrain generation ranks roughly as follows...
Current (1.18+) >>>>> Beta 1.7.3 and prior >>> Beta 1.8 to Release 1.6 ~= Release 1.7 to release 1.16 (1.17)
...So if I presume that means giving the current terrain generation to any older version, it might be an improvement but it doesn't outweigh the loss of content and other changes. In other words, it closes the gap, but fails to make them equal, let alone better.
Something like early release versions probably stand to gain the most, since I find them to be about the worst period in terrain generation to begin with. Those versions really don't have terrain generation so much as they have a random jigsaw biome placement system, and the biome determines terrain. That's not good for my style of play. I dislike retivieness over a larger scale, and I like some order more than pure randomness. 1.7 was therefore theoretically better than the immediately prior generation (hence the "about equals" above) but it ended up slightly worse despite it. It took one step forward and two back because of some of the choices it made. The oceans needed shrunk but not that much, and the climate system was too strict and there was a lot of repetitiveness you don't see even today (though there's definitely still some of that). Some biomes got really rare (jungles, for example, which can be rather large today). And generation was of course still set by the biome itself.
The only version I might consider interest in with modern (or better) terrain generation might be beta 1.7.3 because it was different enough from modern versions (no sprinting, no hunger, no potions, no enchantments, no flight, etc.). The problem is, whereas modern versions might be more imbalanced with the additional things, they're also greatly served by them as quality of life changes for a sandbox survival game, so I feel like beta 1.7.3 might be more suited to a more pure survival first, sandbox second experience for a world you might want to "beat" as opposed to spend time in and built up really long term, but it's also a version that lacks the end and many other things to accomplish.
So a hypothetical beta 1.7.3 with modern terrain generation and "biomes" and the end added (and maybe some other new things) would be a version that sounds interesting... but you could go the other way and say that's basically modern versions minus sprinting, hunger, potions, enchantments, and flight. You may recall my recent hardcore world has limits against some of those, so I'd say yeah, there might be some interest from me in such a thing. But would I prefer it over modern versions? Not a chance.
In the abstract, no, but there's just too much content for one person to backport. Note said project isn't done. If somebody could organize a team that would work; it seems kinda worth it because such a project would make it possible for ALL modders to work on that version and not face the endless Treadmill Of DOOOM! of trying to keep up the endless stream of updates coming out. If every modder who'd done something substantive on any of the "refuge" mod versions (1.7.10, 1.12.2, 16.5) had backported just *one* current feature we'd be done.
A technical problem is that Forge doesn't support old versions, which caused me quite a bit of trouble trying to get a modding setup working again. Dunno if it would be possible to get LexManos in on a "backport everything" project.
Backports aside, I see mods as a way to create your own unique experience, rather than just being "add-ons" to the base game, as I noted in this post on Reddit; future vanilla updates are completely irrelevant and may even clash with your modded content, or require major requites (which in my case means of the vanilla game; instead of rewriting my code I'd rewrite vanilla. e.g. add methods that use xyz variables or an "index" instead of BlockPos, a "block state" as a simple int instead of an object, arrays storing a numerical block ID instead of references to block/blockstate objects, and so on):
Mods like TMCW are actually fairly popular for older versions (note that these are all different from your usual "add-on" mod; many completely change the game and purport to be "updates" with some even calling themselves e.g. Beta 1.7.4 (as if Mojang had continued the Beta 1.7.x series). TMCW itself just started as a collection of various changes to world generation, hence its name):
Also, referring to the discussion about terrain being independent of biomes, most biomes in TMCW actually have the same height variation, aside from "special" biomes like Plains, which are flatter, with "hilly" and "mountainous" biomes generating in all sorts of combinations (not just your old "hills in the middle of a biome, but you can also find the opposite, or half-normal/half hills; and yes, I have "hills" biomes generating as full-sized biomes, not just sub-biomes; note the various shades of orange in this "desert"):
// Sets min/max heights of most biomes; REGIONAL_HEIGHT is regional height variation added to min/max
protected static final float BIOME_MIN_HEIGHT = 0.2F;
protected static final float BIOME_MAX_HEIGHT = 0.7F;
protected static final float BIOME_REGIONAL_HEIGHT = 0.3F;
protected static final float HILLS_MIN_HEIGHT = 0.75F;
protected static final float HILLS_MAX_HEIGHT = 0.9F;
protected static final float HILLS_REGIONAL_HEIGHT = 0.4F;
// Example of biomes which do not use the standard heights (min, max, regional)
public static final BiomeGenBase plains ... .setHeight(0.1F, 0.15F, 0.1F);
public static final BiomeGenBase hillyPlains ... .setHeight(0.2F, 0.6F, 0.2F);
public static final BiomeGenBase hillyPlainsHills ... .setHeight(0.8F, 0.9F, 0.4F);
"regional height variation" refers to the use of a large-scale noise field (wavelength about 1000 blocks) to modulate the heights, which e.g. range from 0.7 to 1.0 for the "max height"; for comparison, vanilla 1.6.4 Extreme Hills has a min/max height of 0.3 and 1.5 and most other biomes are 0.1 and 0.3, and 0.3 and 0.7-0.8 for hills, so this gives most biomes a similar or greater height variation than vanilla's "hills".
I also apply special treatment to the noise field for various biomes, most notably rivers and beaches, fixing the issues of rivers drying up (except where it is intentional) and beaches "crawling" up hillsides; in an upcoming update I've added an additional manipulation of biome heights when deserts are adjacent to other biomes to reduce the same thing happening with them (the height is forced downwards / adjacent hills are pushed back from deserts; overall my terrain generator is extremely complex and it is impressive that worlds generate 2-3 times faster than vanilla 1.6.4 despite all this).
I think old Minecraft is better, and that's what I play every day. 1.20 is bad, 1.19 is bad, 1.18 is bad, 1.17 is bad, 1.16 is ok, 1.15 is ok, 1.14 is ok, 1.13 is ok. These new updates take the simpleness of Minecraft out of Minecraft. Last update I would start a world on would be 1.12. Right now I am playing on 1.6 on the Hermitcraft II seed. Mojang is making Minecraft too easy, these hardcore players are getting diamond gear from villagers super easy. Back in the day you would have too go mining for diamonds but now you can get diamond gear and tools from villagers.
You can also get diamond gear from villagers in 1.6, and the trades are much cheaper - I've bought thousands of diamonds worth of gear in my first world, for the low cost of just 10 emeralds per diamond pickaxe, or 180 wheat (18 per emerald), exploiting the fact that if the final trade is something like wheat for emeralds you can trade it indefinitely without ever needing to trade something else to unlock it again (or the final trade, if not the main trade, is also something that gives you emeralds; such villagers were called "perfect villagers" (note the date - this was when 1.6 was the current version) and were removed in 1.8, and since 1.14 you can only trade a limited number of times per day - I've traded a double chest of wheat in one go before and only have two active villagers, a blacksmith and a farmer, plus backups for both).
The only possible downside is that more heavily enchanted items are impractical to repair with items, if at all, due to the way repair costs are calculated, but you can get around this by wearing down the sacrifice a bit, as I do for swords (a Sharpness V, Knockback II, Unbreaking III diamond sword costs 40 levels to repair with a brand new sword but by damaging it a bit, e.g. killing 3 stacks of chickens, the cost is lowered to 38 levels, enabling a full repair after adding a 12% durability bonus. Even better, those chickens can be traded for enough emeralds to buy a new sword, otherwise, you'd have to use 1-2 diamonds for 29-35 levels. Other items, like a maxed-out Fortune pickaxe, are much costlier, 37 levels for a single diamond, so this method is less practical (you have to wear the sacrifice down a lot more, something like 500-600 left will cost as much as one diamond but restore closer to twice the durability), but still avoids any need to mine diamonds, except for an enchantment table).
Note that I only trade for the items I need to repair my gear for fun and so I can hoard every diamond I mine (closing in on a double chest of diamond blocks, mined without Fortune), as even without using Fortune I get many times the diamonds I'd need thanks to my playstyle of nonstop caving (you could say that trading is a diversion from just caving all the time, and I literally do mean all the time as I only build basic bases to act as temporary storage and food/wood farms).
On the other hand, it is definitely much harder to get enchanted books, with a probability of only 1.75% per reroll of trades to get a single enchantment, which will also eventually be replaced with a lower-priced offer (usually meaning worse, unless it something like Silk Touch and the cost has bottomed out at 5 emeralds. I used this fact to more easily get Mending before I fixed it (by "Mending" I mean my own enchantment which works like renaming an item to keep the cost from increasing; the official enchantment is overpowered and broken because it lets you repair any item, no matter how heavily enchanted, for a flat rate and only requires XP. This is the only time I've ever traded to get an enchantment as it is just not worth it, given that anything else can be gotten from the table (you do need books to get Unbreaking on weapons and armor though so it can be very worth trading it, as opposed to say, mining thousands of quartz and enchanting hundreds of books, as I do, but I also use it to build my main base anyway).
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?
Any version with beds, stackable food, sprinting or the ability to switch to Creative mode is a disgrace to the memes that where running around back when "Survival" meant something (so anything above beta 1.2).
The latest versions added so much quality of life and new things to do, don't get me wrong, but the feeling or living in a world that wanted you dead and you had to mine/craft your way to survival is totally gone now. It's like back then Minecraft had a proper early game but no late game (once you have diamond gear and a defendable base, you're done), and now it's the opposite because you go straight to late game in a matter of hours. Kind of bizarre from a game design perspective.
If you're trying to define eras, the groupings seem a bit strange.
Anyway, I would find it hard to ever go back to pre-1.18 unless it was on the side. Like maybe I could have fun starting a modded world on an older version for a different experience, or doing a beta 1.7.3 world for a similar reason, but going back to a version before 1.18 as the highest I'd play on and staying there? I wouldn't enjoy it.
Sure, I have nostalgic love for the older versions. I also have qualms with the new versions in some regards, but I think the game is objectively in a better place overall.
I'm choosing 1.18 or above every time. It's the "foundation" I now need. I can't go back to the missing content and other changes from versions before that now. Even 1.13 through 1.16 feel aged and that is saying something (especially since 1.16 isn't that old and is still a fine version, and perhaps what most would consider the best single update Mojang has done so far), and 1.12 and older feel downright ancient. 1.18 is the foundation version I must have. It's my personal single favorite update (in contrast to 1.8, being my least favorite). And 1.19 has everything 1.18 does, and then some. And 1.20 has everything 1.19 does, and then some. Therefore, 1.20 is my current favorite version by extension, but 1.18 is my favorite update and the minimum I could live with. Going back to before that now would be rough.
If you think current is best, you should vote that way. Though I agree the groupings are strange - 1.16 is by far the most popular version between 1.12 and "current as possible", both in servers and mod supply, so that category should be 1.13 - 1.16 or 1.17.
How important is the terrain generation in your evaluation? How would you like an older version with greatly improved terrain generation?
Geographicraft (formerly Climate Control) - Control climate, ocean, and land sizes; stop chunk walls; put modded biomes into Default worlds, and more!
Rather important, I would say. So much so that whenever one has happened, it's when I feel one era ends and another begins. That's not to say it would always make me feel an era shift has occurred (as always, it's all in the details), but thus far it always has.
These are what I feel were "major era shifts" (there's some minor sub-eras within these too, namely between the 1.7 to 1.12 era which can honestly be broken into three sub-eras).
Beta 1.7.3 to beta 1.8.
Release 1.6 to release 1.7.
Release 1.12 to release 1.13.
Release 1.16 (1.17) to release 1.18.
Terrain generation changed markedly in three of those four splits I see as era changing. Release 1.12 to release 1.13 is the exception, and that sort of served as the foundation for modern updates and what would collectively be the era 1.16 served as a capstone to.
There's other things I think that warrant changes in eras. An important one would be something that radically changes the "meta" of the game. Think all of the changes between beta 1.7.3 and beta 1.8 (hunger system, sprinting, enchanting, potions, etc.), the new combat of 1.9, and addition of elytra and mending (these formally happened in 1.9, but it was more the addition of rocket boosting in 1.11 that made elytra meta changing, and 1.14s villager changes that made mending more common, so those are more muddy to pin down and I'd say actually happened after 1.9 itself).
The biggest collective shift seems to have been beta 1.7.3 to beta 1.8. New terrain generation, and many new systems.
https://minecraft-statistic.net/en/global_statistic.html has always shown the latest version to be the most popular by far, at least as far as the number of servers goes (and of this, I assume those tracked by the site); currently 1.20.1 is in the lead with 36.4% of all servers; 1.16.5 is far behind with only 1/10 as many servers, with 1.12.2 still being about even with it, and the once mighty modding version of 1.7.10 has been left in the dust (even recently I've seen people arguing that it was still the most popular modded version, caveats with this data aside, including the fact it only tracks major(?) multiplayer servers, not singleplayer or small home servers(?), as also reflected by its claimed active player counts vs what https://activeplayer.io/minecraft/ shows).
Also, this seems to show that the chat reporting added in 1.19.1 had no impact on updating to newer versions (I've seen mods/plugins that disable it and many 1.19.1+ servers may be using one, at the same time, 1.9 seemed so hard to work around in a similar manner). Likewise, the upcoming changes to villager trading is still only an optional feature so there should be no reluctance to update to 1.20.2.
(wondering how many servers it lists for 1.6? None, not even a single one, but that doesn't mean there aren't any, nor does it list Beta versions like 1.7.3, which I know have a sizeable playerbase, much greater than 1.6)
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 would expect that most of the 1.12.2 and 1.16.5 servers on that report are modded; I can't see why they'd be so much more popular than 1.13 or 1.17 other than for the mods. I agree it seems impossible there are NO Beta servers whatsoever, but I'd want some data to back up a claim than 1.7.10 is more popular for mods *now* (as opposed to the most popular modded version over history, which I'd find plausible.) Both the base game and mods tend to add features over time and you give up a lot to go with older version. There is a huge supply of 1.7.10 mods, of course, but many have been abandonware for a very long time and that introduces its own problems in the form of unfixed and unfixable bugs.
Geographicraft (formerly Climate Control) - Control climate, ocean, and land sizes; stop chunk walls; put modded biomes into Default worlds, and more!
There is no reason though why somebody can't make a mod that backports newer content (looks at myself), and in fact, I know of at least one such mod that has been continued across multiple developers (I even used some of its open-source code to implement newer features):
https://www.minecraftforum.net/forums/mapping-and-modding-java-edition/minecraft-mods/3187078-et-futurum-requiem-a-revival-of-ganymedes01s
This mod mostly focuses on backporting content while replicating the original as closely as possible and not adding much non-vanilla content, as the majority of my own mod is (nor are the features I backported always accurate to fit with my views of how they should work, e.g. Mending as a way to make the "renaming negates prior work penalty" bug legitimate and not so easy to get; my "attack cooldown" specifically penalizes spam-clicking without nerfing attack speed/DPS for "proper" attacks. I've also backported developmental features, like endermen spawning endermites when they teleport and naturally spawning killer rabbits,which also use their original texture and name, and nerfing "mineral" drops to require a player kill (especially iron golems), all of which Mojang actually implemented in 1.8 snapshots).
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?
In the abstract, no, but there's just too much content for one person to backport. Note said project isn't done. If somebody could organize a team that would work; it seems kinda worth it because such a project would make it possible for ALL modders to work on that version and not face the endless Treadmill Of DOOOM! of trying to keep up the endless stream of updates coming out. If every modder who'd done something substantive on any of the "refuge" mod versions (1.7.10, 1.12.2, 16.5) had backported just *one* current feature we'd be done.
A technical problem is that Forge doesn't support old versions, which caused me quite a bit of trouble trying to get a modding setup working again. Dunno if it would be possible to get LexManos in on a "backport everything" project.
Geographicraft (formerly Climate Control) - Control climate, ocean, and land sizes; stop chunk walls; put modded biomes into Default worlds, and more!
I forgot to answer this when answering the part before it.
Without knowing what "greatly improved terrain generation" means, or what "old version" we're talking about, it's hard to say. I feel the terrain generation ranks roughly as follows...
Current (1.18+) >>>>> Beta 1.7.3 and prior >>> Beta 1.8 to Release 1.6 ~= Release 1.7 to release 1.16 (1.17)
...So if I presume that means giving the current terrain generation to any older version, it might be an improvement but it doesn't outweigh the loss of content and other changes. In other words, it closes the gap, but fails to make them equal, let alone better.
Something like early release versions probably stand to gain the most, since I find them to be about the worst period in terrain generation to begin with. Those versions really don't have terrain generation so much as they have a random jigsaw biome placement system, and the biome determines terrain. That's not good for my style of play. I dislike retivieness over a larger scale, and I like some order more than pure randomness. 1.7 was therefore theoretically better than the immediately prior generation (hence the "about equals" above) but it ended up slightly worse despite it. It took one step forward and two back because of some of the choices it made. The oceans needed shrunk but not that much, and the climate system was too strict and there was a lot of repetitiveness you don't see even today (though there's definitely still some of that). Some biomes got really rare (jungles, for example, which can be rather large today). And generation was of course still set by the biome itself.
The only version I might consider interest in with modern (or better) terrain generation might be beta 1.7.3 because it was different enough from modern versions (no sprinting, no hunger, no potions, no enchantments, no flight, etc.). The problem is, whereas modern versions might be more imbalanced with the additional things, they're also greatly served by them as quality of life changes for a sandbox survival game, so I feel like beta 1.7.3 might be more suited to a more pure survival first, sandbox second experience for a world you might want to "beat" as opposed to spend time in and built up really long term, but it's also a version that lacks the end and many other things to accomplish.
So a hypothetical beta 1.7.3 with modern terrain generation and "biomes" and the end added (and maybe some other new things) would be a version that sounds interesting... but you could go the other way and say that's basically modern versions minus sprinting, hunger, potions, enchantments, and flight. You may recall my recent hardcore world has limits against some of those, so I'd say yeah, there might be some interest from me in such a thing. But would I prefer it over modern versions? Not a chance.
Backports aside, I see mods as a way to create your own unique experience, rather than just being "add-ons" to the base game, as I noted in this post on Reddit; future vanilla updates are completely irrelevant and may even clash with your modded content, or require major requites (which in my case means of the vanilla game; instead of rewriting my code I'd rewrite vanilla. e.g. add methods that use xyz variables or an "index" instead of BlockPos, a "block state" as a simple int instead of an object, arrays storing a numerical block ID instead of references to block/blockstate objects, and so on):
https://www.reddit.com/r/feedthebeast/comments/tc60yc/do_you_see_mods_as_updates_or_addons_to_vanilla/
Mods like TMCW are actually fairly popular for older versions (note that these are all different from your usual "add-on" mod; many completely change the game and purport to be "updates" with some even calling themselves e.g. Beta 1.7.4 (as if Mojang had continued the Beta 1.7.x series). TMCW itself just started as a collection of various changes to world generation, hence its name):
https://www.reddit.com/r/GoldenAgeMinecraft/comments/wd1k0o/any_obscure_overhaul_mods/
Also, referring to the discussion about terrain being independent of biomes, most biomes in TMCW actually have the same height variation, aside from "special" biomes like Plains, which are flatter, with "hilly" and "mountainous" biomes generating in all sorts of combinations (not just your old "hills in the middle of a biome, but you can also find the opposite, or half-normal/half hills; and yes, I have "hills" biomes generating as full-sized biomes, not just sub-biomes; note the various shades of orange in this "desert"):
"regional height variation" refers to the use of a large-scale noise field (wavelength about 1000 blocks) to modulate the heights, which e.g. range from 0.7 to 1.0 for the "max height"; for comparison, vanilla 1.6.4 Extreme Hills has a min/max height of 0.3 and 1.5 and most other biomes are 0.1 and 0.3, and 0.3 and 0.7-0.8 for hills, so this gives most biomes a similar or greater height variation than vanilla's "hills".
I also apply special treatment to the noise field for various biomes, most notably rivers and beaches, fixing the issues of rivers drying up (except where it is intentional) and beaches "crawling" up hillsides; in an upcoming update I've added an additional manipulation of biome heights when deserts are adjacent to other biomes to reduce the same thing happening with them (the height is forced downwards / adjacent hills are pushed back from deserts; overall my terrain generator is extremely complex and it is impressive that worlds generate 2-3 times faster than vanilla 1.6.4 despite all this).
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?