TL;DR - Do you think there are more caves than before? Almost every cave I run into is sickeningly enormous!
It seems to me that almost every time I attempt to build an underground mine network or structure, I run into a GIGANOURMOUS MONSTER cave network. I'm talking large enough to spend half an IRL day on, fuel the world's furnace demands with coal, and fill chests with iron BLOCKS just by spelunking alone
I used to be stoked like Indiana Jones when I found a cavern that took me more than a half hour to explore, but since around Beta 1.5, and especially after 1.8, that feeling has pretty much reversed in itself
Running Cartograph and Minecraft X-Ray scans for the 10 experimental worlds I generated recently shows that in numerous chunks, the frequency of Air blocks actually somehow outnumbers the Stone. Around the 30 altitude mark lies vast expanses of Abandoned Mineshafts and interconnected caves, spanning in some cases over 3 kilometers long in a continuous network. There seems to be little middle ground for cave size. A few are little pockets, or 5 minute stops, but an overwhelming majority put the Japanese transportation network to shame!
I know that things are likely not to change because of the code freeze and official release. Using terrain gen mods / different versions of Minecraft are 2 ways of mitigating the issue. Some of us like the aspect of adventure, and some like to mine their own tunnels, some yet in between. Minecraft is different for everyone, but in the case of level generation, this shoe isn't one size fits all. Realistically, it is what it is - might as well keep an open hopeful mind :smile.gif:
Something I would welcome seeing in the future sometime is an option for cave frequency (ex. Many, Some, Few, None) just like the "Generate Structures" option.
As for Samsonguy920's mob opportunity argument, I believe there are ways to adjust play style for more efficient mob outcomes. A natural or anthropogenic system can be made more efficient, albeit at the cost of resources from the player(s), being item blocks and/or time invested. Each way has it's pros and cons!
Interesting stuff to think about.. I'd love to hear your opinion!
I was originally gonna ask why you where hating but I do agree with you...
Iron has definitely gotten easier to obtain, that I actually like since iron has the most practical uses. But I wish not every cave was a massive network... I find myself grudgingly going along and lighting the system in a systematic manner. I am annoyed when I find cave systems now...
Please make most of the systems smaller Minecraft, the sheer number of these systems and likeliness is too much to remain inserting.
Short of ravines being added, I really haven't seen any sort of increase in caves, just a couple changes to how they work. Right now I have been able to make my usual 50m mineshafts without any interruption, which is on par with every other beta version.
I don't see the cave system changing anytime, and as you posted at the end, you did find an area without significant cave presence. I don't really see the problem. I would prefer they stay as they are, as the alternative is making the caves ourselves to find the resources we can get access to now, as well as the mob experience. Don't like hostile mobs? Mine in Peaceful mode.
Rollback Post to RevisionRollBack
In the real world, you stick your head in the dirt to hide from problems.
In Minecraft, you stick your head in the dirt to find problems.
Well personally i prefer spending HOURS of IRL time exploring caverns. Im a splunker what can i say :biggrin.gif: but i do like the idea of beingable to turn down the size of caverns in the menu.
Definition I use here:
UN (Underground Network): Everything underground physically connected. You might have to break blocks to make the ceiling high enough so that you can pass, but you can see there something directly. This includes overlapping ravines, tunnels, caves, abandoned mineshafts, dungeons, etc.
I use a "strict" definition:
- Anything so tiny that I can explore completely in less than a minute, count as simple Underground Features, not a true UN.
- Non-connected underground areas that are close enough to be heard from (ambiance, mobs, water, lava, etc.) count as separate UN.
I play over 20 hours per week, sometimes >80. That is a LOT. I ALSO prefer spending days exploring caves.
But a UN is so big that after real-life MONTHS of intensive mining I still am nowhere near finished clearing it, that is where I draw the line. And currently in 1.7 that seems like the NORM, not the exception !
1.6.4 underground = "infinite swiss-chesse in all directions forming a single world-wide UN with also a few unconnected and always always quite small UN popping out here and there"
1.7 underground = "UN are mainly mega-networks that can eventually be finished but only after an insane amount of mining, with some but not many UN going from tiny to huge spread out in the gaps"
I'm not asking to go to the other extreme, aka the following :
"I searched for 100 hours and I found only 3 measly cave systems that I finished in less than 5 minutes each".
Which is what almost all the "I like having lots of caves" players think ANY reduction in amount and size of cave networks will inevitably end up doing to the game.
I just want something reasonable! Something that would appeal to MOST players not just to the infinite-tunnels-lovers group.
A typical network should, on average, be doable in a couple good sized gaming sessions. I'm talking time sent "fully mining" here, lighting almost all of it up and extracting most of the resources (probably not all the gravel), with the time computed using good tools. Not "time required to just sprint through it using night vision gettintg only the best stuff". And not "time needed to also pave all tunnels with nice-looking stairs slabs so it looks super-nice or is super-easy to walk through", either.
In any case, the way tunnels are generated should make it so a "mega-network" should still have some kind of "maximum upper limit" put to it so it doesn't become ridiculously "impossible to complete within a reasonable amount or effort".
Possible solutions:
--> Way #1 : Region row/column numbers used to determine amount of tunnels.
Make the world grid of regions (32x32 chunks each = 512x512 blocks each) look like a chessboard:
If the region is a...
...White square: normal "full features" underground.
...Black square : Zero abandoned mineshaft spawn there, and much smaller and less connected networks.
Basically, mega-networks are still possible, but much less numerous and would not tend to be humongously insanely huge, because while severla networks inside the same region would easily "connect", only a FEW networks would connect from one "high tunnels" ara to another such area.
- Make the underground have "underground biomes (not linked to surface biome, but hints of what biome is underground could appear at or near the surfiace). The "borders" between such underbiomes would only have have connections. Thus, you could have huge tunnel networks, but it would only not tend to be uber-mega-way-too-big-for-the-average-player.
- Offer better world creation tweaking options that allows players to control underground stuff, like tunnels spacing, size, and density, of tunnel networks, cave networks, ravines, abandoned mineshafts, etc. Very important: also a slider for the amount of "connector tunnels", those long tunnels that currently together underground tunnels that would otherwise stay totally disconnected.
I'm all for SOME areas of the underground to be wide and complex maze-like semi-open expanses, SOME places where the underground becomes more made up of "air" than "stone", but that should be the exception (not rare, just uncommon) instead of being the norm over freaking thousands and thousands of blocks of distance spent mining. Special structures make the underground look way cooler only if they are special, and there is a "baseline" to compare the weird structures to. Because when there are extreme-looking cave structures EVERYWHERE, then it kind of turns the whole thing into a "meh" experience very quickly.
What the game should be:
Finding a new cave network from exploring the surface should be an common (not rare) but still a-bit-special find. Not something that makes you nearly constantly sidestep as explore to avoid falling in them.
Finding a new cave network from already mining underground should be a moment of joy. Not something that makes you wince in a "Sigh I'll never be able to finish this mine now" kind of way.
Note that I play very regularly. usually well, well way over the 20 hours per week mark. Sometimes above 80 hours, even.
So let me say that I also prefer spending HOURS exploring caverns. DAYS, even. Or I could even b extreme and say WEEKS.
It is when I spent MONTHS in a SINGLE cave network and SITLL haven't finihed at all, that I draw the line.
And 1.6.4 undergrand *is* "infinite swiss chesse in all directions". OR every nearly so.
1.7 has LOTS less, but still a whopping lot. Still way too much., IMHO.
I am not saying we should go from one extreme to another, and suddenly it takes you months of exploration to find a cave network that, if you're lucky, if finished in more than 5 minutes. But a single network should, ON AVERAGE, be able to be finished in a "reasonable" number of game sessions.
And when I say "average", I use total time, excluding all "false positives" (things that seem like cave networks structures, but that are so tiny that they kind of "fail" to deserve the term "cave network"). So if you have an area with say 2 tunnel networks and one is finishable in say 10 hours and the other requires a 990 hours, then the average is 500 hours per cave network,. and that is just wayyyyy too much for an average value.
Either the amount of caves should be dropped again by a noticeable factor (not A HUGe factor - no need to go overboard and do something just as unbalanced but "in the other direction"'), or a screen specifically for a few custom world options strictly for determining how tunnel networks get generated. Sliders for things like : node size, node spacing, node density, average tunnel length, average cave vs tunnel ratio, overall "maze vs smooth" factor.
I'm all for SOME areas of the underground to be wide and complex maze-like semi-open expanses, such places where the underground becomes more "air" than "stone", but that should be the exception, not the norm.
And no, they should NOT reduce the amount of caves, no, not without adding something like this:
In 1.6.4 I can easily blow through an average size cave system in a few hours - plus a mineshaft and a few ravines. In fact, even the largest cave system I've ever found (18 separate abandoned mineshafts embedded in and around it!) only took about a week (around one IRL day of playtime) to fully explore - and that was huge by any 1.7 standards (I laugh when I see a seed for an "insane", "crazy", etc cave system in 1.7-1.8+):
Here's a rendering of the whole world, the giant cave system is located near the left side, under a taiga biome (more explored around it since the first image; these maps only show caves that have been explored):
There's also an extremely dense cave system, the largest and densest (at the same time) cave system I've found, centered around -800, -1050, a third of the way down on the left side (seed -123775873255737467) - and I should note, extremely rare with nothing even close to it (and completely impossible since 1.7) on this map (outdated, but I haven't found anything else like it yet) of the lowest 10 layers above lava, where denser caves stand out better:
See also: Customizable cave/structure generation, plus other suggestions for "customized" settings (that's my own suggestion, originally posted before 1.7 was released. At the very least, there should be two settings, "size" and "frequency", which control the number of caves per cave system (randomized) and the frequency of cave systems; higher values for both increase variation, 1.6.4 actually has less frequent cave systems than 1.7, if much larger, and more variation in local cave density as a result, if also higher on average):
Of course, with this attitude we can't expect Mojang to do much; then again, they did fire TheMogMiner:
I had asked MogMiner back when the customized options came out if we could get some settings to tweak the cave generation and he said it wasn't possible currently.
(yeah, when I can mod SNAPSHOTS to have 1.6.4 cave generation in a few minutes as long as they do not significantly change the code, and I'm not a professional programmer)
Also, most of the complaints of "too many caves" starting in Beta 1.8 are in fact due to the addition of ravines and mineshafts; I decompiled Beta 1.7.3 and believe it or not it had the EXACT same number and size of cave systems as up to 1.6.4. They did fix a bug that caused caves to generate inconsistently across chunk borders but that did not increase the volume of caves. Even more fun, back in InfDev and Alpha caves (even back then the code was practically the same) were actually 50% more common (that's right - 1.7 is actually the second time they have reduced cave frequency, though cave system size was not altered until then); for example, here is a rendering of Alpha I found (you can see some inconsistencies in some areas due to the aforementioned bug but overall that didn't really cause much disruption. Note also that some caves have small breaks (1-2 blocks), which are still present today and are intentional; of course, I can usually recognize these "fake dead ends" by their appearance and mine through them):
Speaking of mineshafts, they can definitely use some tweaks to how they generate (not just adding a slider to adjust their frequency); I've modded them so they do not generate if there are more than a certain number of caves around them (this works better with 1.6.4 cave generation due to more local-scale variation), as well as having them generate with a minimum spacing like other structures, which is much better than having big cave systems full of wooden planks or mineshafts overlapping other mineshafts and forming ridiculous messes (of course, they were 2.5 times more common in 1.6.4 - and twice as common as that when first introduced in Beta 1.8 - Notch must of had something for the underground; then again, he did call the game "Cave Game" at first).
Damned me, I was sure I had read the "previous posts in topic's" date as Nov 11 2015, instead of Nov 15 2011. I must be borderline "sydlexic" ! Sorry everybody my bad.
"They shouldn't reduce caves unless adding something like this"
Now THAT (your 1st image showing world options, let's call it image A) would be really nice and is exactly what I'm talking about ! Is that from a "fake screen" you did (nice work there!) or is it from real mod (me want!) ?
Well, if they reduce the size without such an options screen, then the reduction shouldn't be as big as "potetially ideal", which means that the change would not actually all that big, and if you're going to break old maps continuity, then you might as well not do it at all. So yeah I agree: reduce ok, only if ALSO adding add the configuration options at the same time.
You did a network consisting of 18 Abandoned Mineshafts (plus all ravines and caves connected to it) in less than a month ? Wow you work fast !
I must admit I personally am on the slow side of things. until last year I kind of "200%" mined my caves. I mean I not only took ALL resources (even if i already had several chestsfulls of say mere Coal !), but I even placed back smooth stone after digging my ores so everything would look nice and perfect. All my tunnels (well, those I used more frequently at least) were rounded just so and slabbed so that I could sprint without having to jump or without risking hitting a corner or any obstacle. My mineshafts ended as perfectly smooth stone tunnels, 3x3 size and empty. And so on. It takes me MUCH less time now, but still nowhere near your speed. You're dang fast you know lol !
The next 3 images (images B C D) you give are very interesting. you talk about these networks as if they are sepadrayte, vbut you say '"not all tunnels were explored", wehich seems contradictory. either a network is "done" or it is not, in wh9ch case odds are good that it will connect to another networek ad in fact you don't have two huge networks, but a single even bigger one. Can you say make "circles" around a few of the biggest networks you explored, and write down the values defining your exploration as best you can ? Say, 2 exits, 20 hours. I know I played some Tekkit pack 2 years ago and eventually got to the best end-game stuff. My armor made me move much faster and see in the dark perfectly. My laser cut down long tunnels like a searing hot sword through butter. MY special dimensional bags allowed me to carry megatons of stuff with me. Mining was like 20 times faster than using the best vanilla tools. So if you did your stuff in pure vanilla, you deserve some kind of medal !
On the server I played most, and a coupl other places, even AFTER 1.7 came out, the underground felt much more like image D than image E. One mine complex I did I spent a full year on and it ended up being bigger than 2000x4000 blocks (with some "holes" inside it, as it wasn't totally chock-full of tunnels in there) and was still "open" in 6 pathas to more sub-networks nodes, before the server resetted on me. I agree that image E seems... actually way too sparse for vmy tastes ! The only network standing up a bit more than the others is near the top, and I would call a large network, not even a huge one unless the unrelevealed pasrt is even bigger than what can be seen.
Thanks a lot for the link to your post ! Very interesting !!!
You are right that ravines/mineshafts connecting different sets of cave networks together makes a HUGE difference. IMGO a caves network should take into account EVERYTHING (exclusing openings to the surface or ocean) that connect.
Yeah Mineshafts need tweaks too. They're really big and "twistedly noodly" and expansive. Tweak ooptions would allow a player to make an infinite-minehshaft underground... or tiny mineshafts that also have extra dungeons. Loot Chest frequency should also be a tweak.
Finally, thanks again for the wonderful post and... just how do we recognize the false dead ends ?
On the server I played most, and a coupl other places, even AFTER 1.7 came out, the underground felt much more like image D than image E. One mine complex I did I spent a full year on and it ended up being bigger than 2000x4000 blocks (with some "holes" inside it, as it wasn't totally chock-full of tunnels in there) and was still "open" in 6 pathas to more sub-networks nodes, before the server resetted on me. I agree that image E seems... actually way too sparse for vmy tastes ! The only network standing up a bit more than the others is near the top, and I would call a large network, not even a huge one unless the unrelevealed pasrt is even bigger than what can be seen.
Really? It sounds like you are even more obsessed than I am at caving; by any chance can you run MCMap on your world (check the underground option; it will only render areas around torches, note that this includes unexplored mineshafts and strongholds since they have naturally generated torches, although I wouldn't say it is a big spoiler since it is hard to tell exactly where they are due to the rendering perspective).
You also don't seem to realize the scale of the cave map I posted, if you are referring to this one - that covers an area close to 5,000 blocks long and contains, using the 1.6.4 parameters, some 15,000 individual caves, 700 ravines and some 200-250 abandoned mineshafts (the last is more of a guess since they are less common within 1280 blocks of the origin; I've taken close to 50,000 rails from them though).
Or were you referring to this map when you said "sparse" - That's only caves near lava level, which I used to show the density of a particular cave system; otherwise, it doesn't stand out so well; here is what the same map, but up to sea level, looks like:
In addition, here is a comparison of 1.6.4 to 1.7, with 1.6.4 on the left, both over 4000x4000 block areas using the same seed (only caves and ravines are shown):
Also, when I mentioned "not fully explored" when talking about the huge cave/mineshaft network I meant that when I rendered the cave system by itself I hadn't explored everything else around it - the map mentioned above (second link) was rendered months afterwards, after I'd explored all the other caves surrounding it. I've used MCEdit (like Spectator mode) to look around at areas I've explored and I'd say that at least 95%, if not more, of all caves are explored (excluding those that aren't connected to anything); for example, here are a few different areas; the last one shows some unexplored caves in a relatively empty area but otherwise I've explored most of everything else:
My method of of finding new caves, etc is mainly through interconnections with new areas - for example, just the last time I played I discovered caves after mining through a cave that went just below the lava level (the ceiling dropped down to y=10), which then lead to three ravines and another mineshaft. I've also found new areas before after breaking into them while mining ore (I once found a mineshaft that wasn't connected to anything else after mining a vein of coal). I'll also explore caves around a "waypoint" (a staircase dug to the surface with a cobblestone pillar where I last returned home) until I've fully exhausted the area, before moving on to the next marker(s) I left.
This should give you an idea of how much caving I do, this was from the last time I played - that's not even my single play session record either, I've mined more than 5,000 ores on several occasions and mine 2-3,000 almost every time I play, the last time I didn't mine that much was when I built a new base at a village I found, and extended my railway to it (note the stack of logs as well - that was a full stack of 64 when I started out, so I made about 1,248 torches):
Here is a full-size rendering of the region I've been caving in recently, which is at the extreme lower-right of the full-size map in the second link in this post; circled in red is the approximate area I explored the last time I played (the remainder of the mass of caves on the right and all the caves going to the left) and my location is near the green circle (I can tell because you can see cobwebs, which are associated with a cave spider spawner I'd just destroyed when I took the screenshot, but hadn't mined the cobwebs yet; the mineshaft appears to extend to the south, or towards the lower-right on this map; normally MCMap will show parts of unexplored mineshafts due to the torches that naturally generate in them but I modded them out so they only render after I've explored them). Crossing the map from lower-left to upper-right is a railway I made to my latest base (I only rendered up to y=63 so the village/base doesn't render here; only a few caves above sea level get cut off):
Also, here is a surface rendering showing two of the markers that I use to mark where I left off; I've made three return trips so far to the one towards the lower-right and likely will make a new one further south the next time I leave off since I'm getting further away, plus there isn't such a well-defined path underground to where I currently am. Some of these markers I make have their coordinates written down since I plan to come back in the far future after moving on to other areas; I don't just explore caves in one direction and the size of my world makes it hard to remember where they are (that said, one such marker recently become rather useless after I ended up exploring caves back to it after completing a large circuit underground; I had planned to return to it after I made a new fully zoomed map to the south of the one around spawn - but ended up reaching it from the south after exploring outwards from a different area):
You do a LOT of caving. Way more than me in terms of world area explored in any case. Maybe we could start an OCD club or something lol !
I fully realize the scale of the maps -- I didn't know how many thousands of blocks that was exaxtly, but when I saw the biomes map I guessed between 3000 and 10000 approximately. You say it's 5000 so yeah my guesstimate was ok. Yeah that is a HUGE area.
What I wonder is what you mean by "most of the caves".
If you mean "there are cave networks, especially smaller ones that don't connect to anything, that I didn't cover, it's more of a coverage vs the map" Then we have different definitions.
My definition is more: take N cave networks, with N a relatively small number, or maybe even only 1. Even if "my" networks cover only a small part of the map, and there are a lot of other networks on the same map, when I say I covered most of the networks, it means there are still paths I haven't lighted up yet inside "my" networks that I did mine in. I don't care at all about all those other still unexplored networks, elsewhere on the same map, unless they actually connect to the networks I started mining, meaning that those would not be "other" networks, but just other areas of the networks I already started".
Basically, from my own point of view, I don't care much about unexplored networks, I care only about limiting onbeself to see if an explored networks WAS fully explored, THAT network was it FULLY done vs MOSTLY done ? And adding up for al networks that were partially or fully done, but ignored those that weren't even touched.
Time to mine is nice too but that can be grossly extrapolated from total network area, and anyway varies even more from player to player. So yeah, exploring in spectator mode can quicken things up a lot.
Maybe using MCEdit, to do this:
- Remove everything aboveground
- Replace all water with lapis blocks
- Replace all lava with redstone blocks
- Replace all stone with netherrack
- Replace all air with stone
- Replace all netherrack with Air
Then you can just fly in the map and see the networks directly :-)
Also, we seem to differ a wee bit on our definition of what forms a "connecting part of the same cave network".
- First, you treat lava as just an obstacle that in the way, kind of like when I am in a tunnel and the path is blocked by gravel I just dig it out of the way, I don't treat that as "forming an actual dead-end" like a few gravel-hating players tend to do. I treat lava as a total obstacle, like stone, so I do not bother digging "alongside/underneath" lava tunnels or swimming in hard-to-see lava (using Respiration III helmet plus fire resistance potions), to try to determine if the tunnel/cave leads somewhere else in the same cave network.
Apart from Respiration III, anything which helps vision in lava ? I feel even Respiration III ain't such a good visibility in there.
Vanilla maximum Respiration III immersion survival time from max hearts: 60 seconds until suffocation then another 70 seconds until death.
Breaking a stone block with an Efficiency V Diamond Pickaxe, even with the slowdown from being immersed, is quick enough to create air pockets in the ceiling as needed, every minute or so that shouldn't be a problem. Or just using the "empty bucket" trick. The only real problem is the inventory space for several Fire Resist potions, but I guess once you exit the lava part, you just dig a simple tunnel out of the lava exit back to the lava area entrance, thus it's only a minor problem, right ? In any case, I just never bothered with it / thought much about, until now lol.
- Second, we both treat ores as if they were air physically connecting tunnels into a single bigger network. Iexclude vnaiulla gravel as the pockets are way too big and numerous and usually NOT something the player bothers to "mine." On this issue we seem to act the same.
- Third, I treat "physically unconnected tunnels" that are close enough that you can hear sounds coming from unseen parts (water, lava, ambiance, or mobs) as actually being part of the same network MAYBE.
I say MAYBE because there are just too much unknowns. Without using mapping mods, it's really hard to use a "reliable" 3D mental map of the local twisting maze of tunnels. Some sounds like ambiance are too rare or short lived. others like mobs stay longer, but move around and they too will stop. the only reliable ones are flowing liquids, really. But these sounds might come from already lit up parts of the network too.
Ideally, I just ignore most of them. But when I "feel" the local tunnels are mostly finished, yet still hear some sound, I might try digging to reach the dark area. So yeah I only end up catching "some" of them and also end up getting a lot of "false positives".
However, if I had a mod that gives you an compass-like item that points in a reliable way to still dark areas, then heck yeah I would consider "nearby" tunnels as being part of the same overall bigger network. i.e. each cave network requires at least 15 blocks of "inter-distance" to really count as being a separate network "outside of hearing range". Presently, my definition is "meh sometimes it is, sometimes it isn't, other times I just don't bother because I have really crappy hearing and don't really like trying to home-in on a sound to find the "missing"' tunnel or cave.
How do you do that ?
Is there a program that can do this:
- Take some overworld world map.
- Treat still-not-genned chunks bordering that world's already genned chunks as if they were total obstacles.
- Replace EVERYTHING with Air, except the most basic natural blocks that occur under the surface: Stone, Dirt, gravel, Sandstone, Bedrock. And I do mean everything from torches to lava to bricks or what have you !
Replace all blocks above ground level with let's say Glass blocks, Basically, to block the entire surface. If getting ground-elevation is impossible, then just use a world-wide plane of Glass blocks around Y=60 in order to isolate the surface from the usual underground stone level.
- For each Air block, check if there is a path lading outside a 5x5x5 area centered on it. If no, replace that Air block with Black Coal block. Basically, this will "plug off" tiny defective areas of air that isn't really a cave or tunnel.
- Once that is done, for each smooth Stone block, check in a 31x31x31 area centered on it, if there is at least one Air block in the area If yes, replace that Stone block with Netherrack. Basically, this tags the stony "skin" of a cave network. Note that this 31x31x31 volume assumes sound travels 16 blocks away even through stone. the netherrack indicate "heardable-sound-thickness-skind" between nearby networks that are separated by sound alone, that should count as only one bigger network.
- Once that is done, do a for XYZ through the entire underground volume to search for Air blocks:
--- When an Air block is found:
--- Select the next non-natural block in an arbitrary long list of non-natural blocks, chosen for their color constrast Gold blocks lapis blocks etc.
--- 3D-fill algorithm to tag that entire Air and Netherrack but Not-Stone volume with that specific arbritratry block, replacing all Air (and only Air) with the abritrary tagging block.
--- Increment counter for each air block replaced.
--- Once the 3D fill is done, that entire network is "done", store number of air blocks in an "NetworkTotalSize[]" vector, increment the network counter, and resume doing the for XYZ loop.
- Finally, replace all the netherrack and all the stone with Air.
You end up with a world showing all the cave networks in all their glory.
Possible Tweaks:
- Ignore the netherrack step. This, treating networks separated by 16 stone or less as "separate".
- Don't replace lava. Same principle.
- Replace gravel and/or dirt, too, just to see the effect on the number of networks.
Sparse map: yeah I kinda figured something weird was going on. The side-to-side maps is cool to see. It shows clearly how 1.6.4 is "nearly all one single worldwide cave system", while 1.7 seems to feature a LOT of humongous cave systems.
I'd be really curious about the automated cave-analysis algorithm above, outputting different colors clearly showing what touches what or not, and maybe one predefined color of the "worldwide" network, and another predefined color or two for "small minor networks", and get a graph showing distribution of networks according to total size of said networks. sigh that probably would require hundreds of hours of programming involved !
I like a LOT the image showing a red line circling how much you did in 1 play session. And this is exactly what I'm talking about: Take the area you circled, and double it. That should IMHO be the TOTAL size of a TYPICAL network. By tpical I mean as in "you mine down a stairs to lava layer then dig a long tunnel and the odds are quite good that is the kind of network you will find".
I any case, you really deserve your forum name ! :-)
- First, you treat lava as just an obstacle that in the way, kind of like when I am in a tunnel and the path is blocked by gravel I just dig it out of the way, I don't treat that as "forming an actual dead-end" like a few gravel-hating players tend to do. I treat lava as a total obstacle, like stone, so I do not bother digging "alongside/underneath" lava tunnels or swimming in hard-to-see lava (using Respiration III helmet plus fire resistance potions), to try to determine if the tunnel/cave leads somewhere else in the same cave network.
What I meant by that is I'll mine at y=11 until there isn't any more lava under my feet; here ire several examples; the first one is the place where I broke into a new cave system last time (no idea if there are any other interconnections yet; I have not looked around since I don't want to see what I haven't explored yet; however I do have a marker nearby to the east that I haven't gone to yet). In the second example I just formed a loop between two caves extending from the same cave system, while in the third I stopped mining (mining a vein of ore at the end), which would have just come back to the same cave anyway:
Otherwise, I don't try swimming in lava or anything like that (or use any potions aside from curing villagers and in the Nether).
Also, if I see water or lava dripping down from the ceiling I'll dig up to see what it is (sometimes it is just an underground lake) and I do go down any surface cave openings I come across; for example, while ago I found a previously unexplored ravine after coming across a 1x2 hole in sand next to water, which was not connected to anything else (and extremely infested with mobs since I'd explored most of the caves in the area; mobs are quite common on the surface at night, as are ocelots in jungles during the day, in explored regions). Many unexplored cave openings in areas I've explored out are usually just short dead ends, or small cave systems that only take 10-20 minutes to explore).
In addition, here is an example of how close caves can come to intersecting; there isn't really even one block separating them with air touching corner-corner. Actually, this is also an example of what I meant by "fake dead ends", which are due to the cave generating code having a 25% chance of not generating a "segment" (a spherical section similar to circular rooms, which are just one such segment by themselves; usually this does not cause complete breaks unless it happens several times in a row or crosses a chunk border); one way to identify these is that they are often more than 2-3 blocks wide, have unnaturally flat ends (not here) or the cave is too short; one of the two caves going towards the top are likely the same cave continuing on after the break (the other end connects to a ravine, which then connects to two more ravines in a line, plus a medium size cave system, all surrounded by caves I've explored):
Many of the "empty areas" on the explored cave maps I've shown are likely not actually empty, just contain caves that aren't connected to other caves or the surface, or I haven't come across their surface openings, which I don't actively search for.
Here is another example; it is hard to tell but the cave where the cursor (white box) is only has a connection to the surface; if the ore I'd mined (it appears to be a depression where there was ore, perhaps a 2x2x2 vein that generated halfway in the cave above) had been one block lower and a 2x2x2 vein I'd have found the cave after mining it in the ceiling of the cave below it, which I'd explored first (I only recently found this cave after coming across an unlit surface opening in an area I'd explored long ago):
Yeah ok I turn lava into obsidian using water too. That is what I meant by "treating lava as stone" obstacle.
IMHO all those things you showed me, count as part of a single network. It may be possible to have a bit of difficulty reaching all areas of a single network because of all such things : false dead ends, 100% lava-filled paths, thick stone wall but close enough to hear sounds from the other side.
Basically I think we have to treat tunnel network like this:
Local node: One sub-section of the tunnels, probably maze-like looking, with most tunnels ending up back into the same general area.
Local network: Everywhere you can reach easily (even if you need to dig a bit to make room for Steve because there is only a 1x1 opening).
Supernetwork: This is everywhere reached by digging through obvious obstacles and obvious hints: False Dead-ends, ore veins, lava sounds.
Meganetwork: The fusioning of supernetworks with less than 16 blocks of stone thickness between them, or that you need to swim through lava to reach. Meganetworks are thus effectively "isolated from each other for real" instead of by a mere obstacle.
I tend to want to know about the meganetworks-level coverage, but when I play I am relatively happy when I manage to complete a supernetwork.
A meganetwork is a like a pomegranate: the air parts of caves are like the juice inside the pomegranate seeds (local network). The 1st layer of stone around the tunnels's air blocks, along with all "obvious to cross" obstacles such as ore veins forming thin walls between two areas, are like the "skin" of the pomegranate seeds. Thus, an entire cluster of pomegranate seeds separated only by their pure skins = one supernetwork. And finally the 16-blocks-thickness of pure stone around the caves and tunnels, the "maximum reach for sound", along with "need to swim under lava" paths, would correspond to the springy white inedible membrane stuff separating and isolating the clusters of pomegranate seeds from each other. (1 whole pomegranate = 1 meganetwork).
I am mainly interested in the supernetwork and meganetwork here. I'm curious about their size spread. How much % of the map they cover.
I'm guessing 1.6.4 worlds undergeound are ALL made up of a single gigantic meganetwork covering the entire world, with a number of much, much smaller meganetworks spread out in the few gaps that remain. This is because the "expansiveness" and omni-presence of the main meganetwork effectively puts a size limit on all other meganetworks that aren't connected to it.
I would wager that only say 2% of 1.6.4 world surface area is NOT part of any meganetwork (don't forget counting the 16 blocks of stone wall thickness around the supernetworks as part and parcel of the meganetworks). Then maybe 8% of the underground is tiny to medium meganetwork (most tiny), and a whopping 90% of the world area forming the main meganetwork. I'd say that in those conditions a player just trying to make an underground base hoping to not constantly run into tunnels will have a VERY hard time, meeting new tunnels every few scores of blocks or so. And when he hits some tunnel, there will be solid odds that what he found will be the main meganetwork.
What is your feeling on my values here ?
Now, my experience for 1.7 is different. I'd say about 10% of the underground is not part of any network so a player can dig one or two hundreds blocks before hitting tunnels. Still a bit on the small side of things, but not as bad. And a new discovered area would have fair about equal odds of leading either to a meganetwork ranging anywhere from tiny to reasonably huge, vs hitting a super-huge meganetwork spanning thousands and thousands of blocks.
What is interesting is that in 1.7 the overall density, as seen at the scale of connections between local nodes (instead of the maze of tunnels inside each local node), is much lower, seemingly just enough so that the networks feel more like a web or net with nearly half it's segments broken, so when you reach a new node, it then spreads out to only a few other nodes, and often only one. In 1.6.4 you have a totally dense lattice always going in almost all directions , but in 1.7 you almost feel that there are big long "lines" of underground structures local nodes before you meet an actual "true fork" in the nodes that leads in more than one direction of "big long lines of nodes", and there are more or less big areas that can become "enclosed" inside these lines, possible with smaller networks of their own. For example: a ravine (node A), then lots of tunnels with a few caves (node B), then a big open cave area with some tunnels (node C), then a mineshaft (node D), then another tunnels and cave area (node E), and then another ravine (node F).
Meanwhile in 1.6.4 that example would be more like: a ravine (node A), which opens to four new areas, the one in the north is lots of tunnels with a few caves (node B), which also opens up to 3 new directions. Thus, 3 new nodes, but the 2 on the east and west of those seem to go to new nodes that also connect back to node A, or other nodes that connect back to A, but amongst several other connections leading further out to the east or west AND north and south. Let's take one node going even more away from A on the north, and we reach a big open cave area with some tunnel, which opens to YET more directions, and as we continue moving away north we will reach node D a mineshaft, again opening up on almost all sides, and so on, ad vitam eternam. If I could put all the nodes as chess pieces on a table and link them with elastic strings to represent the connections, I could almost moveeach pice a bit to align it all on an infite grid, with only a FEW spaces in that matrix having zero or more than 1 local node, and with most nodes connecting to most of their local neighbours. Infinite swiss cheese.
I think we need to go one more step further down towards less density, however that step should not nowhere near as big as the step between 1.6.4 and 1.7. 1.7. feels like not even half the total cave density of 1.6.4. So maybe another reduction as small as 20% could do the trick and give us lots of networks but uber-sized meganetworks would be incredibly rare.
However, I'd make ocean biomes have only a tiny amount of tunnels, forming a kind of "isolation zone" between the continental tunnel networks, and the deep oceans tunnels networks. Deep oceans would have the 1.6.4 network density. Thus, "infinite tunnel lovers", who don't spend that much time on the surface anyway, could go find a deep ocean biome, setup their base either on an island (or a minecart tunnel to the shore of the nearest continent) (or just directly underground!), and they would get a truly humongous nigh-infinite mega-ultra-network to play with. While the other players, the majority of players in fact, those who tend to like to be able to "complete" a network without spending an insane amount of time on it, and who play a lot more on the surface, would find the kind of tunnel networks they prefer easily, anywhere under the continent !. We'd get the best of both worlds !
This separation using the deep oceans would also protect the casual miners from the mega-miners on multiplayer servers. Presently, I see a LOT of players complaining every so often that some unknown uber-miner came and "totally mined up" everything both in *and* around their relatively small mined area. They never saw the guy come and go ! Heavy miners work faster and work intensively, while miner coulds can play for weeks before goin to their mine again. ENTIRE neighbourhood of players had almost ALL of their underground mines "ruined" by a single super-mining player.
Having the default mines networks a bit smaller and less super-connected would encourage casual miners: they could feel they "won" over a network before reaching real-life old age. It would also serve to discourage heavy miners from mining there. Meanwhile at the SAME time having a totally separate environment that offers truly expansive mines would encourage naturally the super-miners to go "under the oceans", leaving the server's underground usability-time-before-being-used-up much more ok for the more casual players. Basically, the server underground wouldn't start to look "old" as fast, because the "mining damage", which is mainly done by the uber miners, would be in parts of the world that most players would not be in (they want to play on the surface too, so going to deep oceans is not their main choice here). Finally, uber-miners usually do not care as much about tunnels with less valuable ores (those nearer the surface). Per region, they would get less "thickness" elevation of tunnels to mine, but on average of "hour spent mining", this means much richer tunnels. in lots of worlds I saw, most uber miners simply tend to skip anything above the best ore layers anyway, so using the deep oceans to create the uber miner's dreamy humongous supernetwork mining areas makes even MORE sense here.
This is a bit like having separate areas for PVE lovers vs PVP lovers, but applied to mining style.
What do you think ?
You are right that the amount of "node" connection is greatly affected by the expansiveness and number of Abandoned Mineshafts (Abbreviation: AM). Making mineshafts less "dense" would mean tunnels separated by only 1 or 2 blocks of wall thickness would not be the norm, and would get rid of the "more air than stone" around here" feeling. AM much smaller in horizontal spread (maybe they could cover a taller distance, though !) and with less of a "maximum" amount of AM in an area, would make multi-central-dirt-room mineshafts uncommon, not the norm. BIG multi-central-room AM would be much rarer (like say with a whopping 6 central dirt rooms), instead of more-or-less common.
A fixed density won't do any good. You'll get people who are like TheMasterCaver and the OP who want massive caves, and then you'll get people like Ouatcheur who want generally infrequent caves. Further, you won't always want the same number of caves all the time; some days you might want to explore caves endlessly, and others you'll have burned yourself out from exploration and just want a nice expanse of mostly solid rock.
Point being, people aren't going to all want the same thing all the time. A fixed cave density is a horrible idea as it only benefits people who want a that specific cave density throughout the entire time they play through the game, which is obviously a small portion of the community.
It's much better to implement a varying cave density, wherein one part of a world might have scattered, unconnected tunnels and another is swiss-cheese, along with everything in-between. This should be coupled with customized world settings to control the minimum, maximum, and average cave size, the amount of space it takes to go from one cave density to a different density, et cetera. This would please almost everyone; ordinary worlds would have enough variety to offer something for every preference of density, while people who either insist on having one density or crave a specific density at a given time have ways to force that cave density to primarily or always occur in a given world (to force a constant cave density, you'd simply set the minimum, maximum, and average to equal levels).
Rollback Post to RevisionRollBack
Did something happen to you in your childhood to give you this unreasonable fear of rutabaga?
I wouldn't say that the current cave density, at least not in 1.6.4, is constant; true, over large scales or from world to world it doesn't vary but on a local scale it does; for example, consider the following maps, from 1.6.4:
There's definitely variation in the size and density of cave systems, and even some nice large cave-free areas (for a sense of scale the map is about 1800 blocks across); as I mentioned earlier 1.7 significantly reduced this local-scale variation. Here is a map from 1.7 for comparison (only about 900 blocks across but I've seen more or less the same in every 1.7 world I've looked at):
Also, the overall number of caves isn't actually that much lower in 1.7 - they are just much more spread out rather than in clumps, with individual cave systems being close to 3 times smaller (0-39 caves in 1.6.4, 0-14 in 1.7. This number is randomized in a way that gives an average that is 1/8 of the maximum or 4.875 for 1.6.4 and 1.75 for 1.7) but about 2 times more common (1/15 chance per chunk in 1.6.4, 1/7 in 1.7. This results in 0.325 caves per chunk in 1.6.4 and 0.25 in 1.7, thus 1.7 has about 77% as many caves overall). I suspect that cave systems in 1.7 are much more likely to have disconnected caves within them, whereas in 1.6.4 the majority of caves are interconnected. Note that I'm referring to caves within cave systems - the dense clusters of caves seen on these maps, not between such clusters. Also, a cave "system" is really just a collection of caves randomly generated in the same area - there is no guarantee that any will connect so a minimum density is desired; for example, if you want separate cave systems to be less likely to intersect instead of making them smaller make them less common, such as using the 1.6.4 size but with half the frequency (so a cave system with 0-39 caves every 30 chunks).
As seen on the first map, 1.6.4 also has larger areas free of caves; I've been able to make mines over 200 blocks long without much disruption:
This was from one of my modded worlds, but cave generation is basically the same as 1.6.4 with a larger size variation of individual caves and ravines. Also, I mined below the lava level due to a mod ore, so avoiding caves was more important.
In addition, if you look at the 1.7 map you might notice a lack of mineshafts - these were reduced by a factor of 2.5 fold in 1.7, plus they are less common closer to the origin in either case. For example, here is a comparison of mineshafts using the same seed (also the same seed used to make the cave maps) in 1.6.4 and 1.7, first centered around the origin then at a point further away (using ChunkBase's Mineshaft Finder app):
Near the origin; 1.6.4 has 3-4 areas where mineshafts potentially overlap (if they are at similar elevations) while 1.7 has only one; in either case there are no big networks of mineshafts:
Further away; in 1.6.4 the majority of mineshafts are now parts of huge clusters, some with 20 or more mineshafts potentially overlapping. Even 1.7 often has clusters but this is more due to the fact that thy randomly generate in any chunk regardless of nearby mineshafts (the 1.6.4 density is really too high even for a uniform distribution to work well; a mineshaft can extend up to 5 chunks from their center, or a 10x10 chunk area - which is also how frequently they generate in 1.6.4, a 1% chance per chunk):
To further demonstrate how common mineshafts are in 1.6.4, I found this chain of mineshafts which is well over 1000 blocks long (continuing past both sides) and contains as many as 50 separate mineshafts (if not directly interconnected caves and ravines likely connect them together):
It should be noted that 1.7 did not change the frequency of ravines, which have a 2% chance of generating in any chunk, and are much more common than what you might think from finding them on the surface since they mostly generate deeper down; I'd say that one in 10 ravines are exposed at the surface.
Also, here is a comparison of Beta 1.7.3 to release 1.5; note the different air distribution below sea level, which increases (nonlinearly, note the log scale) down to layer 12 in Beta 1.7.3 but in release 1.5 all the mineshafts (most common around y=30) and ravines (biased deeper like caves but their vertical center point does not go below y=20) give a significant upward bias with a broad peak around layer 24:
Beta 1.7.3:
Release 1.5:
(the amount of lava below y=11 is erroneous on the second graph; I believe they only looked at stationary lava as caves generate with flowing lava, at least more recently, and generating a world with a utility like Minecraft Land Generator will not cause them to update and convert to stationary lava unless you actually play on it)
TL;DR - Do you think there are more caves than before? Almost every cave I run into is sickeningly enormous!
It seems to me that almost every time I attempt to build an underground mine network or structure, I run into a GIGANOURMOUS MONSTER cave network. I'm talking large enough to spend half an IRL day on, fuel the world's furnace demands with coal, and fill chests with iron BLOCKS just by spelunking alone
Honestly this must largely depend on your seed, my worlds are NOT like this at all, lots of caves, but little snaky caves.
That said, more world generation toggle options would be amazing;
Colder world
Warmer world
Fewer/More caves
etc
Honestly this must largely depend on your seed, my worlds are NOT like this at all, lots of caves, but little snaky caves.
That said, more world generation toggle options would be amazing;
Colder world
Warmer world
Fewer/More caves
etc
Check out my posts, and look at the date of this thread. You are obviously playing on 1.7 or later, and no, the seed does not affect how many caves generate, except on a local scale (and that is much less varied since 1.7).
It seems to me that almost every time I attempt to build an underground mine network or structure, I run into a GIGANOURMOUS MONSTER cave network. I'm talking large enough to spend half an IRL day on, fuel the world's furnace demands with coal, and fill chests with iron BLOCKS just by spelunking alone
I used to be stoked like Indiana Jones when I found a cavern that took me more than a half hour to explore, but since around Beta 1.5, and especially after 1.8, that feeling has pretty much reversed in itself
Running Cartograph and Minecraft X-Ray scans for the 10 experimental worlds I generated recently shows that in numerous chunks, the frequency of Air blocks actually somehow outnumbers the Stone. Around the 30 altitude mark lies vast expanses of Abandoned Mineshafts and interconnected caves, spanning in some cases over 3 kilometers long in a continuous network. There seems to be little middle ground for cave size. A few are little pockets, or 5 minute stops, but an overwhelming majority put the Japanese transportation network to shame!
I know that things are likely not to change because of the code freeze and official release. Using terrain gen mods / different versions of Minecraft are 2 ways of mitigating the issue. Some of us like the aspect of adventure, and some like to mine their own tunnels, some yet in between. Minecraft is different for everyone, but in the case of level generation, this shoe isn't one size fits all. Realistically, it is what it is - might as well keep an open hopeful mind :smile.gif:
Something I would welcome seeing in the future sometime is an option for cave frequency (ex. Many, Some, Few, None) just like the "Generate Structures" option.
As for Samsonguy920's mob opportunity argument, I believe there are ways to adjust play style for more efficient mob outcomes. A natural or anthropogenic system can be made more efficient, albeit at the cost of resources from the player(s), being item blocks and/or time invested. Each way has it's pros and cons!
Interesting stuff to think about.. I'd love to hear your opinion!
Iron has definitely gotten easier to obtain, that I actually like since iron has the most practical uses. But I wish not every cave was a massive network... I find myself grudgingly going along and lighting the system in a systematic manner. I am annoyed when I find cave systems now...
Please make most of the systems smaller Minecraft, the sheer number of these systems and likeliness is too much to remain inserting.
I don't see the cave system changing anytime, and as you posted at the end, you did find an area without significant cave presence. I don't really see the problem. I would prefer they stay as they are, as the alternative is making the caves ourselves to find the resources we can get access to now, as well as the mob experience. Don't like hostile mobs? Mine in Peaceful mode.
In Minecraft, you stick your head in the dirt to find problems.
Definition I use here:
UN (Underground Network): Everything underground physically connected. You might have to break blocks to make the ceiling high enough so that you can pass, but you can see there something directly. This includes overlapping ravines, tunnels, caves, abandoned mineshafts, dungeons, etc.
I use a "strict" definition:
- Anything so tiny that I can explore completely in less than a minute, count as simple Underground Features, not a true UN.
- Non-connected underground areas that are close enough to be heard from (ambiance, mobs, water, lava, etc.) count as separate UN.
I play over 20 hours per week, sometimes >80. That is a LOT. I ALSO prefer spending days exploring caves.
But a UN is so big that after real-life MONTHS of intensive mining I still am nowhere near finished clearing it, that is where I draw the line. And currently in 1.7 that seems like the NORM, not the exception !
1.6.4 underground = "infinite swiss-chesse in all directions forming a single world-wide UN with also a few unconnected and always always quite small UN popping out here and there"
1.7 underground = "UN are mainly mega-networks that can eventually be finished but only after an insane amount of mining, with some but not many UN going from tiny to huge spread out in the gaps"
I'm not asking to go to the other extreme, aka the following :
"I searched for 100 hours and I found only 3 measly cave systems that I finished in less than 5 minutes each".
Which is what almost all the "I like having lots of caves" players think ANY reduction in amount and size of cave networks will inevitably end up doing to the game.
I just want something reasonable! Something that would appeal to MOST players not just to the infinite-tunnels-lovers group.
A typical network should, on average, be doable in a couple good sized gaming sessions. I'm talking time sent "fully mining" here, lighting almost all of it up and extracting most of the resources (probably not all the gravel), with the time computed using good tools. Not "time required to just sprint through it using night vision gettintg only the best stuff". And not "time needed to also pave all tunnels with nice-looking stairs slabs so it looks super-nice or is super-easy to walk through", either.
In any case, the way tunnels are generated should make it so a "mega-network" should still have some kind of "maximum upper limit" put to it so it doesn't become ridiculously "impossible to complete within a reasonable amount or effort".
Possible solutions:
--> Way #1 : Region row/column numbers used to determine amount of tunnels.
Make the world grid of regions (32x32 chunks each = 512x512 blocks each) look like a chessboard:
If the region is a...
...White square: normal "full features" underground.
...Black square : Zero abandoned mineshaft spawn there, and much smaller and less connected networks.
Basically, mega-networks are still possible, but much less numerous and would not tend to be humongously insanely huge, because while severla networks inside the same region would easily "connect", only a FEW networks would connect from one "high tunnels" ara to another such area.
- Make the underground have "underground biomes (not linked to surface biome, but hints of what biome is underground could appear at or near the surfiace). The "borders" between such underbiomes would only have have connections. Thus, you could have huge tunnel networks, but it would only not tend to be uber-mega-way-too-big-for-the-average-player.
- Offer better world creation tweaking options that allows players to control underground stuff, like tunnels spacing, size, and density, of tunnel networks, cave networks, ravines, abandoned mineshafts, etc. Very important: also a slider for the amount of "connector tunnels", those long tunnels that currently together underground tunnels that would otherwise stay totally disconnected.
I'm all for SOME areas of the underground to be wide and complex maze-like semi-open expanses, SOME places where the underground becomes more made up of "air" than "stone", but that should be the exception (not rare, just uncommon) instead of being the norm over freaking thousands and thousands of blocks of distance spent mining. Special structures make the underground look way cooler only if they are special, and there is a "baseline" to compare the weird structures to. Because when there are extreme-looking cave structures EVERYWHERE, then it kind of turns the whole thing into a "meh" experience very quickly.
What the game should be:
Finding a new cave network from exploring the surface should be an common (not rare) but still a-bit-special find. Not something that makes you nearly constantly sidestep as explore to avoid falling in them.
Finding a new cave network from already mining underground should be a moment of joy. Not something that makes you wince in a "Sigh I'll never be able to finish this mine now" kind of way.
Necro much?
And no, they should NOT reduce the amount of caves, no, not without adding something like this:
In 1.6.4 I can easily blow through an average size cave system in a few hours - plus a mineshaft and a few ravines. In fact, even the largest cave system I've ever found (18 separate abandoned mineshafts embedded in and around it!) only took about a week (around one IRL day of playtime) to fully explore - and that was huge by any 1.7 standards (I laugh when I see a seed for an "insane", "crazy", etc cave system in 1.7-1.8+):
Here's a rendering of the whole world, the giant cave system is located near the left side, under a taiga biome (more explored around it since the first image; these maps only show caves that have been explored):
There's also an extremely dense cave system, the largest and densest (at the same time) cave system I've found, centered around -800, -1050, a third of the way down on the left side (seed -123775873255737467) - and I should note, extremely rare with nothing even close to it (and completely impossible since 1.7) on this map (outdated, but I haven't found anything else like it yet) of the lowest 10 layers above lava, where denser caves stand out better:
See also: Customizable cave/structure generation, plus other suggestions for "customized" settings (that's my own suggestion, originally posted before 1.7 was released. At the very least, there should be two settings, "size" and "frequency", which control the number of caves per cave system (randomized) and the frequency of cave systems; higher values for both increase variation, 1.6.4 actually has less frequent cave systems than 1.7, if much larger, and more variation in local cave density as a result, if also higher on average):
Of course, with this attitude we can't expect Mojang to do much; then again, they did fire TheMogMiner:
(yeah, when I can mod SNAPSHOTS to have 1.6.4 cave generation in a few minutes as long as they do not significantly change the code, and I'm not a professional programmer)
Also, most of the complaints of "too many caves" starting in Beta 1.8 are in fact due to the addition of ravines and mineshafts; I decompiled Beta 1.7.3 and believe it or not it had the EXACT same number and size of cave systems as up to 1.6.4. They did fix a bug that caused caves to generate inconsistently across chunk borders but that did not increase the volume of caves. Even more fun, back in InfDev and Alpha caves (even back then the code was practically the same) were actually 50% more common (that's right - 1.7 is actually the second time they have reduced cave frequency, though cave system size was not altered until then); for example, here is a rendering of Alpha I found (you can see some inconsistencies in some areas due to the aforementioned bug but overall that didn't really cause much disruption. Note also that some caves have small breaks (1-2 blocks), which are still present today and are intentional; of course, I can usually recognize these "fake dead ends" by their appearance and mine through them):
http://www.minecraftforum.net/forums/mapping-and-modding/minecraft-tools/1260523-cartograph-map-your-alpha
Speaking of mineshafts, they can definitely use some tweaks to how they generate (not just adding a slider to adjust their frequency); I've modded them so they do not generate if there are more than a certain number of caves around them (this works better with 1.6.4 cave generation due to more local-scale variation), as well as having them generate with a minimum spacing like other structures, which is much better than having big cave systems full of wooden planks or mineshafts overlapping other mineshafts and forming ridiculous messes (of course, they were 2.5 times more common in 1.6.4 - and twice as common as that when first introduced in Beta 1.8 - Notch must of had something for the underground; then again, he did call the game "Cave Game" at first).
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?
Damned me, I was sure I had read the "previous posts in topic's" date as Nov 11 2015, instead of Nov 15 2011. I must be borderline "sydlexic" ! Sorry everybody my bad.
"They shouldn't reduce caves unless adding something like this"
Now THAT (your 1st image showing world options, let's call it image A) would be really nice and is exactly what I'm talking about ! Is that from a "fake screen" you did (nice work there!) or is it from real mod (me want!) ?
Well, if they reduce the size without such an options screen, then the reduction shouldn't be as big as "potetially ideal", which means that the change would not actually all that big, and if you're going to break old maps continuity, then you might as well not do it at all. So yeah I agree: reduce ok, only if ALSO adding add the configuration options at the same time.
You did a network consisting of 18 Abandoned Mineshafts (plus all ravines and caves connected to it) in less than a month ? Wow you work fast !
I must admit I personally am on the slow side of things. until last year I kind of "200%" mined my caves. I mean I not only took ALL resources (even if i already had several chestsfulls of say mere Coal !), but I even placed back smooth stone after digging my ores so everything would look nice and perfect. All my tunnels (well, those I used more frequently at least) were rounded just so and slabbed so that I could sprint without having to jump or without risking hitting a corner or any obstacle. My mineshafts ended as perfectly smooth stone tunnels, 3x3 size and empty. And so on. It takes me MUCH less time now, but still nowhere near your speed. You're dang fast you know lol !
The next 3 images (images B C D) you give are very interesting. you talk about these networks as if they are sepadrayte, vbut you say '"not all tunnels were explored", wehich seems contradictory. either a network is "done" or it is not, in wh9ch case odds are good that it will connect to another networek ad in fact you don't have two huge networks, but a single even bigger one. Can you say make "circles" around a few of the biggest networks you explored, and write down the values defining your exploration as best you can ? Say, 2 exits, 20 hours. I know I played some Tekkit pack 2 years ago and eventually got to the best end-game stuff. My armor made me move much faster and see in the dark perfectly. My laser cut down long tunnels like a searing hot sword through butter. MY special dimensional bags allowed me to carry megatons of stuff with me. Mining was like 20 times faster than using the best vanilla tools. So if you did your stuff in pure vanilla, you deserve some kind of medal !
On the server I played most, and a coupl other places, even AFTER 1.7 came out, the underground felt much more like image D than image E. One mine complex I did I spent a full year on and it ended up being bigger than 2000x4000 blocks (with some "holes" inside it, as it wasn't totally chock-full of tunnels in there) and was still "open" in 6 pathas to more sub-networks nodes, before the server resetted on me. I agree that image E seems... actually way too sparse for vmy tastes ! The only network standing up a bit more than the others is near the top, and I would call a large network, not even a huge one unless the unrelevealed pasrt is even bigger than what can be seen.
Thanks a lot for the link to your post ! Very interesting !!!
You are right that ravines/mineshafts connecting different sets of cave networks together makes a HUGE difference. IMGO a caves network should take into account EVERYTHING (exclusing openings to the surface or ocean) that connect.
Yeah Mineshafts need tweaks too. They're really big and "twistedly noodly" and expansive. Tweak ooptions would allow a player to make an infinite-minehshaft underground... or tiny mineshafts that also have extra dungeons. Loot Chest frequency should also be a tweak.
Finally, thanks again for the wonderful post and... just how do we recognize the false dead ends ?
Really? It sounds like you are even more obsessed than I am at caving; by any chance can you run MCMap on your world (check the underground option; it will only render areas around torches, note that this includes unexplored mineshafts and strongholds since they have naturally generated torches, although I wouldn't say it is a big spoiler since it is hard to tell exactly where they are due to the rendering perspective).
You also don't seem to realize the scale of the cave map I posted, if you are referring to this one - that covers an area close to 5,000 blocks long and contains, using the 1.6.4 parameters, some 15,000 individual caves, 700 ravines and some 200-250 abandoned mineshafts (the last is more of a guess since they are less common within 1280 blocks of the origin; I've taken close to 50,000 rails from them though).
Or were you referring to this map when you said "sparse" - That's only caves near lava level, which I used to show the density of a particular cave system; otherwise, it doesn't stand out so well; here is what the same map, but up to sea level, looks like:
In addition, here is a comparison of 1.6.4 to 1.7, with 1.6.4 on the left, both over 4000x4000 block areas using the same seed (only caves and ravines are shown):
Also, when I mentioned "not fully explored" when talking about the huge cave/mineshaft network I meant that when I rendered the cave system by itself I hadn't explored everything else around it - the map mentioned above (second link) was rendered months afterwards, after I'd explored all the other caves surrounding it. I've used MCEdit (like Spectator mode) to look around at areas I've explored and I'd say that at least 95%, if not more, of all caves are explored (excluding those that aren't connected to anything); for example, here are a few different areas; the last one shows some unexplored caves in a relatively empty area but otherwise I've explored most of everything else:
My method of of finding new caves, etc is mainly through interconnections with new areas - for example, just the last time I played I discovered caves after mining through a cave that went just below the lava level (the ceiling dropped down to y=10), which then lead to three ravines and another mineshaft. I've also found new areas before after breaking into them while mining ore (I once found a mineshaft that wasn't connected to anything else after mining a vein of coal). I'll also explore caves around a "waypoint" (a staircase dug to the surface with a cobblestone pillar where I last returned home) until I've fully exhausted the area, before moving on to the next marker(s) I left.
This should give you an idea of how much caving I do, this was from the last time I played - that's not even my single play session record either, I've mined more than 5,000 ores on several occasions and mine 2-3,000 almost every time I play, the last time I didn't mine that much was when I built a new base at a village I found, and extended my railway to it (note the stack of logs as well - that was a full stack of 64 when I started out, so I made about 1,248 torches):
Here is a full-size rendering of the region I've been caving in recently, which is at the extreme lower-right of the full-size map in the second link in this post; circled in red is the approximate area I explored the last time I played (the remainder of the mass of caves on the right and all the caves going to the left) and my location is near the green circle (I can tell because you can see cobwebs, which are associated with a cave spider spawner I'd just destroyed when I took the screenshot, but hadn't mined the cobwebs yet; the mineshaft appears to extend to the south, or towards the lower-right on this map; normally MCMap will show parts of unexplored mineshafts due to the torches that naturally generate in them but I modded them out so they only render after I've explored them). Crossing the map from lower-left to upper-right is a railway I made to my latest base (I only rendered up to y=63 so the village/base doesn't render here; only a few caves above sea level get cut off):
Also, here is a surface rendering showing two of the markers that I use to mark where I left off; I've made three return trips so far to the one towards the lower-right and likely will make a new one further south the next time I leave off since I'm getting further away, plus there isn't such a well-defined path underground to where I currently am. Some of these markers I make have their coordinates written down since I plan to come back in the far future after moving on to other areas; I don't just explore caves in one direction and the size of my world makes it hard to remember where they are (that said, one such marker recently become rather useless after I ended up exploring caves back to it after completing a large circuit underground; I had planned to return to it after I made a new fully zoomed map to the south of the one around spawn - but ended up reaching it from the south after exploring outwards from a different area):
See also this thread I posted: Does anybody else come close to this level of caving?
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?
You do a LOT of caving. Way more than me in terms of world area explored in any case. Maybe we could start an OCD club or something lol !
I fully realize the scale of the maps -- I didn't know how many thousands of blocks that was exaxtly, but when I saw the biomes map I guessed between 3000 and 10000 approximately. You say it's 5000 so yeah my guesstimate was ok. Yeah that is a HUGE area.
What I wonder is what you mean by "most of the caves".
If you mean "there are cave networks, especially smaller ones that don't connect to anything, that I didn't cover, it's more of a coverage vs the map" Then we have different definitions.
My definition is more: take N cave networks, with N a relatively small number, or maybe even only 1. Even if "my" networks cover only a small part of the map, and there are a lot of other networks on the same map, when I say I covered most of the networks, it means there are still paths I haven't lighted up yet inside "my" networks that I did mine in. I don't care at all about all those other still unexplored networks, elsewhere on the same map, unless they actually connect to the networks I started mining, meaning that those would not be "other" networks, but just other areas of the networks I already started".
Basically, from my own point of view, I don't care much about unexplored networks, I care only about limiting onbeself to see if an explored networks WAS fully explored, THAT network was it FULLY done vs MOSTLY done ? And adding up for al networks that were partially or fully done, but ignored those that weren't even touched.
Time to mine is nice too but that can be grossly extrapolated from total network area, and anyway varies even more from player to player. So yeah, exploring in spectator mode can quicken things up a lot.
Maybe using MCEdit, to do this:
- Remove everything aboveground
- Replace all water with lapis blocks
- Replace all lava with redstone blocks
- Replace all stone with netherrack
- Replace all air with stone
- Replace all netherrack with Air
Then you can just fly in the map and see the networks directly :-)
Also, we seem to differ a wee bit on our definition of what forms a "connecting part of the same cave network".
- First, you treat lava as just an obstacle that in the way, kind of like when I am in a tunnel and the path is blocked by gravel I just dig it out of the way, I don't treat that as "forming an actual dead-end" like a few gravel-hating players tend to do. I treat lava as a total obstacle, like stone, so I do not bother digging "alongside/underneath" lava tunnels or swimming in hard-to-see lava (using Respiration III helmet plus fire resistance potions), to try to determine if the tunnel/cave leads somewhere else in the same cave network.
Apart from Respiration III, anything which helps vision in lava ? I feel even Respiration III ain't such a good visibility in there.
Vanilla maximum Respiration III immersion survival time from max hearts: 60 seconds until suffocation then another 70 seconds until death.
Breaking a stone block with an Efficiency V Diamond Pickaxe, even with the slowdown from being immersed, is quick enough to create air pockets in the ceiling as needed, every minute or so that shouldn't be a problem. Or just using the "empty bucket" trick. The only real problem is the inventory space for several Fire Resist potions, but I guess once you exit the lava part, you just dig a simple tunnel out of the lava exit back to the lava area entrance, thus it's only a minor problem, right ? In any case, I just never bothered with it / thought much about, until now lol.
- Second, we both treat ores as if they were air physically connecting tunnels into a single bigger network. Iexclude vnaiulla gravel as the pockets are way too big and numerous and usually NOT something the player bothers to "mine." On this issue we seem to act the same.
- Third, I treat "physically unconnected tunnels" that are close enough that you can hear sounds coming from unseen parts (water, lava, ambiance, or mobs) as actually being part of the same network MAYBE.
I say MAYBE because there are just too much unknowns. Without using mapping mods, it's really hard to use a "reliable" 3D mental map of the local twisting maze of tunnels. Some sounds like ambiance are too rare or short lived. others like mobs stay longer, but move around and they too will stop. the only reliable ones are flowing liquids, really. But these sounds might come from already lit up parts of the network too.
Ideally, I just ignore most of them. But when I "feel" the local tunnels are mostly finished, yet still hear some sound, I might try digging to reach the dark area. So yeah I only end up catching "some" of them and also end up getting a lot of "false positives".
However, if I had a mod that gives you an compass-like item that points in a reliable way to still dark areas, then heck yeah I would consider "nearby" tunnels as being part of the same overall bigger network. i.e. each cave network requires at least 15 blocks of "inter-distance" to really count as being a separate network "outside of hearing range". Presently, my definition is "meh sometimes it is, sometimes it isn't, other times I just don't bother because I have really crappy hearing and don't really like trying to home-in on a sound to find the "missing"' tunnel or cave.
How do you do that ?
Is there a program that can do this:
- Take some overworld world map.
- Treat still-not-genned chunks bordering that world's already genned chunks as if they were total obstacles.
- Replace EVERYTHING with Air, except the most basic natural blocks that occur under the surface: Stone, Dirt, gravel, Sandstone, Bedrock. And I do mean everything from torches to lava to bricks or what have you !
Replace all blocks above ground level with let's say Glass blocks, Basically, to block the entire surface. If getting ground-elevation is impossible, then just use a world-wide plane of Glass blocks around Y=60 in order to isolate the surface from the usual underground stone level.
- For each Air block, check if there is a path lading outside a 5x5x5 area centered on it. If no, replace that Air block with Black Coal block. Basically, this will "plug off" tiny defective areas of air that isn't really a cave or tunnel.
- Once that is done, for each smooth Stone block, check in a 31x31x31 area centered on it, if there is at least one Air block in the area If yes, replace that Stone block with Netherrack. Basically, this tags the stony "skin" of a cave network. Note that this 31x31x31 volume assumes sound travels 16 blocks away even through stone. the netherrack indicate "heardable-sound-thickness-skind" between nearby networks that are separated by sound alone, that should count as only one bigger network.
- Once that is done, do a for XYZ through the entire underground volume to search for Air blocks:
--- When an Air block is found:
--- Select the next non-natural block in an arbitrary long list of non-natural blocks, chosen for their color constrast Gold blocks lapis blocks etc.
--- 3D-fill algorithm to tag that entire Air and Netherrack but Not-Stone volume with that specific arbritratry block, replacing all Air (and only Air) with the abritrary tagging block.
--- Increment counter for each air block replaced.
--- Once the 3D fill is done, that entire network is "done", store number of air blocks in an "NetworkTotalSize[]" vector, increment the network counter, and resume doing the for XYZ loop.
- Finally, replace all the netherrack and all the stone with Air.
You end up with a world showing all the cave networks in all their glory.
Possible Tweaks:
- Ignore the netherrack step. This, treating networks separated by 16 stone or less as "separate".
- Don't replace lava. Same principle.
- Replace gravel and/or dirt, too, just to see the effect on the number of networks.
Sparse map: yeah I kinda figured something weird was going on. The side-to-side maps is cool to see. It shows clearly how 1.6.4 is "nearly all one single worldwide cave system", while 1.7 seems to feature a LOT of humongous cave systems.
I'd be really curious about the automated cave-analysis algorithm above, outputting different colors clearly showing what touches what or not, and maybe one predefined color of the "worldwide" network, and another predefined color or two for "small minor networks", and get a graph showing distribution of networks according to total size of said networks. sigh that probably would require hundreds of hours of programming involved !
I like a LOT the image showing a red line circling how much you did in 1 play session. And this is exactly what I'm talking about: Take the area you circled, and double it. That should IMHO be the TOTAL size of a TYPICAL network. By tpical I mean as in "you mine down a stairs to lava layer then dig a long tunnel and the odds are quite good that is the kind of network you will find".
I any case, you really deserve your forum name ! :-)
What I meant by that is I'll mine at y=11 until there isn't any more lava under my feet; here ire several examples; the first one is the place where I broke into a new cave system last time (no idea if there are any other interconnections yet; I have not looked around since I don't want to see what I haven't explored yet; however I do have a marker nearby to the east that I haven't gone to yet). In the second example I just formed a loop between two caves extending from the same cave system, while in the third I stopped mining (mining a vein of ore at the end), which would have just come back to the same cave anyway:
Otherwise, I don't try swimming in lava or anything like that (or use any potions aside from curing villagers and in the Nether).
Also, if I see water or lava dripping down from the ceiling I'll dig up to see what it is (sometimes it is just an underground lake) and I do go down any surface cave openings I come across; for example, while ago I found a previously unexplored ravine after coming across a 1x2 hole in sand next to water, which was not connected to anything else (and extremely infested with mobs since I'd explored most of the caves in the area; mobs are quite common on the surface at night, as are ocelots in jungles during the day, in explored regions). Many unexplored cave openings in areas I've explored out are usually just short dead ends, or small cave systems that only take 10-20 minutes to explore).
In addition, here is an example of how close caves can come to intersecting; there isn't really even one block separating them with air touching corner-corner. Actually, this is also an example of what I meant by "fake dead ends", which are due to the cave generating code having a 25% chance of not generating a "segment" (a spherical section similar to circular rooms, which are just one such segment by themselves; usually this does not cause complete breaks unless it happens several times in a row or crosses a chunk border); one way to identify these is that they are often more than 2-3 blocks wide, have unnaturally flat ends (not here) or the cave is too short; one of the two caves going towards the top are likely the same cave continuing on after the break (the other end connects to a ravine, which then connects to two more ravines in a line, plus a medium size cave system, all surrounded by caves I've explored):
Many of the "empty areas" on the explored cave maps I've shown are likely not actually empty, just contain caves that aren't connected to other caves or the surface, or I haven't come across their surface openings, which I don't actively search for.
Here is another example; it is hard to tell but the cave where the cursor (white box) is only has a connection to the surface; if the ore I'd mined (it appears to be a depression where there was ore, perhaps a 2x2x2 vein that generated halfway in the cave above) had been one block lower and a 2x2x2 vein I'd have found the cave after mining it in the ceiling of the cave below it, which I'd explored first (I only recently found this cave after coming across an unlit surface opening in an area I'd explored long ago):
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?
Yeah ok I turn lava into obsidian using water too. That is what I meant by "treating lava as stone" obstacle.
IMHO all those things you showed me, count as part of a single network. It may be possible to have a bit of difficulty reaching all areas of a single network because of all such things : false dead ends, 100% lava-filled paths, thick stone wall but close enough to hear sounds from the other side.
Basically I think we have to treat tunnel network like this:
Local node: One sub-section of the tunnels, probably maze-like looking, with most tunnels ending up back into the same general area.
Local network: Everywhere you can reach easily (even if you need to dig a bit to make room for Steve because there is only a 1x1 opening).
Supernetwork: This is everywhere reached by digging through obvious obstacles and obvious hints: False Dead-ends, ore veins, lava sounds.
Meganetwork: The fusioning of supernetworks with less than 16 blocks of stone thickness between them, or that you need to swim through lava to reach. Meganetworks are thus effectively "isolated from each other for real" instead of by a mere obstacle.
I tend to want to know about the meganetworks-level coverage, but when I play I am relatively happy when I manage to complete a supernetwork.
A meganetwork is a like a pomegranate: the air parts of caves are like the juice inside the pomegranate seeds (local network). The 1st layer of stone around the tunnels's air blocks, along with all "obvious to cross" obstacles such as ore veins forming thin walls between two areas, are like the "skin" of the pomegranate seeds. Thus, an entire cluster of pomegranate seeds separated only by their pure skins = one supernetwork. And finally the 16-blocks-thickness of pure stone around the caves and tunnels, the "maximum reach for sound", along with "need to swim under lava" paths, would correspond to the springy white inedible membrane stuff separating and isolating the clusters of pomegranate seeds from each other. (1 whole pomegranate = 1 meganetwork).
I am mainly interested in the supernetwork and meganetwork here. I'm curious about their size spread. How much % of the map they cover.
I'm guessing 1.6.4 worlds undergeound are ALL made up of a single gigantic meganetwork covering the entire world, with a number of much, much smaller meganetworks spread out in the few gaps that remain. This is because the "expansiveness" and omni-presence of the main meganetwork effectively puts a size limit on all other meganetworks that aren't connected to it.
I would wager that only say 2% of 1.6.4 world surface area is NOT part of any meganetwork (don't forget counting the 16 blocks of stone wall thickness around the supernetworks as part and parcel of the meganetworks). Then maybe 8% of the underground is tiny to medium meganetwork (most tiny), and a whopping 90% of the world area forming the main meganetwork. I'd say that in those conditions a player just trying to make an underground base hoping to not constantly run into tunnels will have a VERY hard time, meeting new tunnels every few scores of blocks or so. And when he hits some tunnel, there will be solid odds that what he found will be the main meganetwork.
What is your feeling on my values here ?
Now, my experience for 1.7 is different. I'd say about 10% of the underground is not part of any network so a player can dig one or two hundreds blocks before hitting tunnels. Still a bit on the small side of things, but not as bad. And a new discovered area would have fair about equal odds of leading either to a meganetwork ranging anywhere from tiny to reasonably huge, vs hitting a super-huge meganetwork spanning thousands and thousands of blocks.
What is interesting is that in 1.7 the overall density, as seen at the scale of connections between local nodes (instead of the maze of tunnels inside each local node), is much lower, seemingly just enough so that the networks feel more like a web or net with nearly half it's segments broken, so when you reach a new node, it then spreads out to only a few other nodes, and often only one. In 1.6.4 you have a totally dense lattice always going in almost all directions , but in 1.7 you almost feel that there are big long "lines" of underground structures local nodes before you meet an actual "true fork" in the nodes that leads in more than one direction of "big long lines of nodes", and there are more or less big areas that can become "enclosed" inside these lines, possible with smaller networks of their own. For example: a ravine (node A), then lots of tunnels with a few caves (node B), then a big open cave area with some tunnels (node C), then a mineshaft (node D), then another tunnels and cave area (node E), and then another ravine (node F).
Meanwhile in 1.6.4 that example would be more like: a ravine (node A), which opens to four new areas, the one in the north is lots of tunnels with a few caves (node B), which also opens up to 3 new directions. Thus, 3 new nodes, but the 2 on the east and west of those seem to go to new nodes that also connect back to node A, or other nodes that connect back to A, but amongst several other connections leading further out to the east or west AND north and south. Let's take one node going even more away from A on the north, and we reach a big open cave area with some tunnel, which opens to YET more directions, and as we continue moving away north we will reach node D a mineshaft, again opening up on almost all sides, and so on, ad vitam eternam. If I could put all the nodes as chess pieces on a table and link them with elastic strings to represent the connections, I could almost moveeach pice a bit to align it all on an infite grid, with only a FEW spaces in that matrix having zero or more than 1 local node, and with most nodes connecting to most of their local neighbours. Infinite swiss cheese.
I think we need to go one more step further down towards less density, however that step should not nowhere near as big as the step between 1.6.4 and 1.7. 1.7. feels like not even half the total cave density of 1.6.4. So maybe another reduction as small as 20% could do the trick and give us lots of networks but uber-sized meganetworks would be incredibly rare.
However, I'd make ocean biomes have only a tiny amount of tunnels, forming a kind of "isolation zone" between the continental tunnel networks, and the deep oceans tunnels networks. Deep oceans would have the 1.6.4 network density. Thus, "infinite tunnel lovers", who don't spend that much time on the surface anyway, could go find a deep ocean biome, setup their base either on an island (or a minecart tunnel to the shore of the nearest continent) (or just directly underground!), and they would get a truly humongous nigh-infinite mega-ultra-network to play with. While the other players, the majority of players in fact, those who tend to like to be able to "complete" a network without spending an insane amount of time on it, and who play a lot more on the surface, would find the kind of tunnel networks they prefer easily, anywhere under the continent !. We'd get the best of both worlds !
This separation using the deep oceans would also protect the casual miners from the mega-miners on multiplayer servers. Presently, I see a LOT of players complaining every so often that some unknown uber-miner came and "totally mined up" everything both in *and* around their relatively small mined area. They never saw the guy come and go ! Heavy miners work faster and work intensively, while miner coulds can play for weeks before goin to their mine again. ENTIRE neighbourhood of players had almost ALL of their underground mines "ruined" by a single super-mining player.
Having the default mines networks a bit smaller and less super-connected would encourage casual miners: they could feel they "won" over a network before reaching real-life old age. It would also serve to discourage heavy miners from mining there. Meanwhile at the SAME time having a totally separate environment that offers truly expansive mines would encourage naturally the super-miners to go "under the oceans", leaving the server's underground usability-time-before-being-used-up much more ok for the more casual players. Basically, the server underground wouldn't start to look "old" as fast, because the "mining damage", which is mainly done by the uber miners, would be in parts of the world that most players would not be in (they want to play on the surface too, so going to deep oceans is not their main choice here). Finally, uber-miners usually do not care as much about tunnels with less valuable ores (those nearer the surface). Per region, they would get less "thickness" elevation of tunnels to mine, but on average of "hour spent mining", this means much richer tunnels. in lots of worlds I saw, most uber miners simply tend to skip anything above the best ore layers anyway, so using the deep oceans to create the uber miner's dreamy humongous supernetwork mining areas makes even MORE sense here.
This is a bit like having separate areas for PVE lovers vs PVP lovers, but applied to mining style.
What do you think ?
You are right that the amount of "node" connection is greatly affected by the expansiveness and number of Abandoned Mineshafts (Abbreviation: AM). Making mineshafts less "dense" would mean tunnels separated by only 1 or 2 blocks of wall thickness would not be the norm, and would get rid of the "more air than stone" around here" feeling. AM much smaller in horizontal spread (maybe they could cover a taller distance, though !) and with less of a "maximum" amount of AM in an area, would make multi-central-dirt-room mineshafts uncommon, not the norm. BIG multi-central-room AM would be much rarer (like say with a whopping 6 central dirt rooms), instead of more-or-less common.
A fixed density won't do any good. You'll get people who are like TheMasterCaver and the OP who want massive caves, and then you'll get people like Ouatcheur who want generally infrequent caves. Further, you won't always want the same number of caves all the time; some days you might want to explore caves endlessly, and others you'll have burned yourself out from exploration and just want a nice expanse of mostly solid rock.
Point being, people aren't going to all want the same thing all the time. A fixed cave density is a horrible idea as it only benefits people who want a that specific cave density throughout the entire time they play through the game, which is obviously a small portion of the community.
It's much better to implement a varying cave density, wherein one part of a world might have scattered, unconnected tunnels and another is swiss-cheese, along with everything in-between. This should be coupled with customized world settings to control the minimum, maximum, and average cave size, the amount of space it takes to go from one cave density to a different density, et cetera. This would please almost everyone; ordinary worlds would have enough variety to offer something for every preference of density, while people who either insist on having one density or crave a specific density at a given time have ways to force that cave density to primarily or always occur in a given world (to force a constant cave density, you'd simply set the minimum, maximum, and average to equal levels).
I wouldn't say that the current cave density, at least not in 1.6.4, is constant; true, over large scales or from world to world it doesn't vary but on a local scale it does; for example, consider the following maps, from 1.6.4:
There's definitely variation in the size and density of cave systems, and even some nice large cave-free areas (for a sense of scale the map is about 1800 blocks across); as I mentioned earlier 1.7 significantly reduced this local-scale variation. Here is a map from 1.7 for comparison (only about 900 blocks across but I've seen more or less the same in every 1.7 world I've looked at):
Also, the overall number of caves isn't actually that much lower in 1.7 - they are just much more spread out rather than in clumps, with individual cave systems being close to 3 times smaller (0-39 caves in 1.6.4, 0-14 in 1.7. This number is randomized in a way that gives an average that is 1/8 of the maximum or 4.875 for 1.6.4 and 1.75 for 1.7) but about 2 times more common (1/15 chance per chunk in 1.6.4, 1/7 in 1.7. This results in 0.325 caves per chunk in 1.6.4 and 0.25 in 1.7, thus 1.7 has about 77% as many caves overall). I suspect that cave systems in 1.7 are much more likely to have disconnected caves within them, whereas in 1.6.4 the majority of caves are interconnected. Note that I'm referring to caves within cave systems - the dense clusters of caves seen on these maps, not between such clusters. Also, a cave "system" is really just a collection of caves randomly generated in the same area - there is no guarantee that any will connect so a minimum density is desired; for example, if you want separate cave systems to be less likely to intersect instead of making them smaller make them less common, such as using the 1.6.4 size but with half the frequency (so a cave system with 0-39 caves every 30 chunks).
As seen on the first map, 1.6.4 also has larger areas free of caves; I've been able to make mines over 200 blocks long without much disruption:
This was from one of my modded worlds, but cave generation is basically the same as 1.6.4 with a larger size variation of individual caves and ravines. Also, I mined below the lava level due to a mod ore, so avoiding caves was more important.
In addition, if you look at the 1.7 map you might notice a lack of mineshafts - these were reduced by a factor of 2.5 fold in 1.7, plus they are less common closer to the origin in either case. For example, here is a comparison of mineshafts using the same seed (also the same seed used to make the cave maps) in 1.6.4 and 1.7, first centered around the origin then at a point further away (using ChunkBase's Mineshaft Finder app):
Further away; in 1.6.4 the majority of mineshafts are now parts of huge clusters, some with 20 or more mineshafts potentially overlapping. Even 1.7 often has clusters but this is more due to the fact that thy randomly generate in any chunk regardless of nearby mineshafts (the 1.6.4 density is really too high even for a uniform distribution to work well; a mineshaft can extend up to 5 chunks from their center, or a 10x10 chunk area - which is also how frequently they generate in 1.6.4, a 1% chance per chunk):
To further demonstrate how common mineshafts are in 1.6.4, I found this chain of mineshafts which is well over 1000 blocks long (continuing past both sides) and contains as many as 50 separate mineshafts (if not directly interconnected caves and ravines likely connect them together):
It should be noted that 1.7 did not change the frequency of ravines, which have a 2% chance of generating in any chunk, and are much more common than what you might think from finding them on the surface since they mostly generate deeper down; I'd say that one in 10 ravines are exposed at the surface.
Also, here is a comparison of Beta 1.7.3 to release 1.5; note the different air distribution below sea level, which increases (nonlinearly, note the log scale) down to layer 12 in Beta 1.7.3 but in release 1.5 all the mineshafts (most common around y=30) and ravines (biased deeper like caves but their vertical center point does not go below y=20) give a significant upward bias with a broad peak around layer 24:
Release 1.5:
(the amount of lava below y=11 is erroneous on the second graph; I believe they only looked at stationary lava as caves generate with flowing lava, at least more recently, and generating a world with a utility like Minecraft Land Generator will not cause them to update and convert to stationary lava unless you actually play on it)
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?
Honestly this must largely depend on your seed, my worlds are NOT like this at all, lots of caves, but little snaky caves.
That said, more world generation toggle options would be amazing;
Colder world
Warmer world
Fewer/More caves
etc
Your signature sucks.
Check out my posts, and look at the date of this thread. You are obviously playing on 1.7 or later, and no, the seed does not affect how many caves generate, except on a local scale (and that is much less varied since 1.7).
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?