It was a server i had with a friend and I liked the world gen. It was version 1.18. I dont have contact with my friend anymore (they became kinda toxic and made me feel back if I didnt play with them) so i cant just message him on discord to see if he knew what the seed was. (i doubt he would still have the server)
You should try put the details into a seed finder of some sort like AMIDST.
AMIDST doesn't have any such capabilities, it only shows a basic display of the biomes and with markers for the locations of structures, nothing about the appearance of the actual world (I haven't used it in years but a search suggests it still works like this; example map).
In order to find a world from a screenshot you need a supercomputer, or a large-scale distributed computing project, like the ones used to find the title screen panorama:
Minecraft@Home is a volunteer distributed computing project powered by BOINC allowing users to volunteer their idle computers to help advance Minecraft-related research, one of which was this panorama project. In less than 24 hours after launching the panorama application; a volunteer host on Minecraft@Home found this seed. This was approximately 93 days of processing time at a total of 54.5 exaFLOPs compressed into 24 hours.
In other words, it would take the average computer months to find a seed in this manner - it really is about brute-forcing every one of up to 18,446,744,073,709,551,616 seeds, especially since 1.18 no longer uses Java's "Random" for most world generation (the significance of this is that Random only used 48 out of 64 bits, meaning there are 65536 times less seeds to search. As I understand it, Minecraft@Home also uses other methods to reduce the number of seeds it needs to search through, but it is still far more than the average user can reasonably search - in fact, they only performed 3 FLOPS per seed to find the panorama when it would take thousands if not millions per seed as the terrain needs to actually be generated, maybe they were just very lucky to hit it very early).
For another example, this page estimated that it would take 14 years to search all 2^64 seeds (the "catch" they later mention no longer applies to 1.18 since again it uses a proper 64 bit RNG):
Minecraft world seeds are 64-bit numbers, which means there are 264 (over 18 quintillion) possible values. When this tool was originally developed in early 2014, it could test about 10 billion seed values per second on a GTX 460; a GTX 760 could do ~20 billion values per second, and today a GTX 1070 (or other similar modern GPU) can test about 40 billion seed values per second. That means even with GPUs continuing to get faster and faster, it still should take over 14 years to reverse engineer a Minecraft world seed by trying all 18 quintillion possible values.
This was also a much simpler search which only checks the locations of a list of known structures for matches, which is much simpler than having to actually generate terrain (for most seeds only one location needs to be checked, aborting as soon as a mismatch is found, and the code that determines where a structure can spawn is only a few lines*; I'm surprised they didn't include mineshafts since they are the most common and have the simplest algorithm).
*e.g. mineshafts; no matter how simple, or fast the implementation is it still adds up given how many seeds need to be checked, traditional CPU-based computing is simply not an option:
// External to canSpawnStructureAtCoords
this.rand.setSeeed(worldSeed);
long xMultiplier = this.rand.nextLong();
long zMultiplier = this.rand.nextLong();
// Only these lines need to be executed for every location checked in a seed (only one in 100-250 or more seeds need to
// check more than one location so they could all be placed in the same method with a negligible loss in performance)
this.rand.setSeed(((long)chunkX * xMultiplier) ^ ((long)chunkZ * zMultiplier) ^ worldSeed);
this.rand.nextInt();
protected boolean canSpawnStructureAtCoords(int chunkX, int chunkZ)
{
// chance is 0.01 prior to 1.7 and 0.004 since, with 1.13 omitting the second check
return this.rand.nextDouble() < chance && this.rand.nextInt(80) < Math.max(Math.abs(chunkX), Math.abs(chunkZ));
}
It was a server i had with a friend and I liked the world gen. It was version 1.18. I dont have contact with my friend anymore (they became kinda toxic and made me feel back if I didnt play with them) so i cant just message him on discord to see if he knew what the seed was. (i doubt he would still have the server)
How would this help find the seed? Clouds are completely unrelated to the seed, indeed, you can change them with a resource pack, and they constantly move around (they would only be very useful if they were completely static and you still wouldn't be able to tell exactly where they might be, near 0,0, or near some multiple of their repeat distance? They do only move independently of the players' position from east-west but that would only narrow down the possible location to bands running east-west, which may as well still be an infinite area to search).
Besides, this would only be attempted (by those who actually have the resources, i.e. supercomputer level) if the server or world was actually extremely important / historical to the general community (like the title screen panorama), otherwise, your options are more limited and you need more than just a few screenshots which don't show much (the tools I've seen which can be run on a regular PC in a reasonable amount of time require knowing the locations of multiple structures; in order to use a screenshot of terrain you need to actually generate it and check for a block-by-block match. This also means nothing that was possibly altered by a player (the perfectly flat ground in the screenshots suggest that the area was leveled), even an enderman taking a block could cause a mismatch unless missing blocks can easily be seen).
It was a server i had with a friend and I liked the world gen. It was version 1.18. I dont have contact with my friend anymore (they became kinda toxic and made me feel back if I didnt play with them) so i cant just message him on discord to see if he knew what the seed was. (i doubt he would still have the server)
You should try put the details into a seed finder of some sort like AMIDST.
AMIDST doesn't have any such capabilities, it only shows a basic display of the biomes and with markers for the locations of structures, nothing about the appearance of the actual world (I haven't used it in years but a search suggests it still works like this; example map).
In order to find a world from a screenshot you need a supercomputer, or a large-scale distributed computing project, like the ones used to find the title screen panorama:
In other words, it would take the average computer months to find a seed in this manner - it really is about brute-forcing every one of up to 18,446,744,073,709,551,616 seeds, especially since 1.18 no longer uses Java's "Random" for most world generation (the significance of this is that Random only used 48 out of 64 bits, meaning there are 65536 times less seeds to search. As I understand it, Minecraft@Home also uses other methods to reduce the number of seeds it needs to search through, but it is still far more than the average user can reasonably search - in fact, they only performed 3 FLOPS per seed to find the panorama when it would take thousands if not millions per seed as the terrain needs to actually be generated, maybe they were just very lucky to hit it very early).
For another example, this page estimated that it would take 14 years to search all 2^64 seeds (the "catch" they later mention no longer applies to 1.18 since again it uses a proper 64 bit RNG):
This was also a much simpler search which only checks the locations of a list of known structures for matches, which is much simpler than having to actually generate terrain (for most seeds only one location needs to be checked, aborting as soon as a mismatch is found, and the code that determines where a structure can spawn is only a few lines*; I'm surprised they didn't include mineshafts since they are the most common and have the simplest algorithm).
*e.g. mineshafts; no matter how simple, or fast the implementation is it still adds up given how many seeds need to be checked, traditional CPU-based computing is simply not an option:
// External to canSpawnStructureAtCoords this.rand.setSeeed(worldSeed); long xMultiplier = this.rand.nextLong(); long zMultiplier = this.rand.nextLong(); // Only these lines need to be executed for every location checked in a seed (only one in 100-250 or more seeds need to // check more than one location so they could all be placed in the same method with a negligible loss in performance) this.rand.setSeed(((long)chunkX * xMultiplier) ^ ((long)chunkZ * zMultiplier) ^ worldSeed); this.rand.nextInt(); protected boolean canSpawnStructureAtCoords(int chunkX, int chunkZ) { // chance is 0.01 prior to 1.7 and 0.004 since, with 1.13 omitting the second check return this.rand.nextDouble() < chance && this.rand.nextInt(80) < Math.max(Math.abs(chunkX), Math.abs(chunkZ)); }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?
Do u have images of the clouds?
How would this help find the seed? Clouds are completely unrelated to the seed, indeed, you can change them with a resource pack, and they constantly move around (they would only be very useful if they were completely static and you still wouldn't be able to tell exactly where they might be, near 0,0, or near some multiple of their repeat distance? They do only move independently of the players' position from east-west but that would only narrow down the possible location to bands running east-west, which may as well still be an infinite area to search).
Besides, this would only be attempted (by those who actually have the resources, i.e. supercomputer level) if the server or world was actually extremely important / historical to the general community (like the title screen panorama), otherwise, your options are more limited and you need more than just a few screenshots which don't show much (the tools I've seen which can be run on a regular PC in a reasonable amount of time require knowing the locations of multiple structures; in order to use a screenshot of terrain you need to actually generate it and check for a block-by-block match. This also means nothing that was possibly altered by a player (the perfectly flat ground in the screenshots suggest that the area was leveled), even an enderman taking a block could cause a mismatch unless missing blocks can easily be seen).
TheMasterCaver's First World - possibly the most caved-out world in Minecraft history - includes world download.
TheMasterCaver's World - my own version of Minecraft largely based on my views of how the game should have evolved since 1.6.4.
Why do I still play in 1.6.4?
i mean yeah man
Is it possible to find a seed based off of an map image of the world map?
yes it is possible if you want to click the "Random" button 18,446,744,073,709,551,616 times
geez thats a lot of math, try minecraft@home to find the seed