I was testing out a port I made of Resource Overhaul (my mod) to 1.15.2. Though it does technically effect World Generation, it only adds new ores in that respect and should not be what is causing the strange world generation pictured. To confirm this, I recreated several worlds, including the exact seed, with the mod and loaded many chunks without this happening.
What happened:
I loaded into a world and well... let the picture speak for itself.
Again I am fairly certain it is not my mod causing this.
I recreated the same seed (same location) and it looks like this:
Any time you see 16 x 16 block of land that doesn't fit you are looking at a chunk error.
Possible causes can be turning off or losing power to your computer while Minecraft is trying to save the world but more common is creating a world with the same name as the one just deleted.
There may be other causes.
Rollback Post to RevisionRollBack
There are no dangerous weapons. There are only dangerous people. R.A. Heinlein
If you aren't part of the solution, then you obviously weren't properly dissolved.
This is a fairly common(?) issue affecting newer versions, with the greatest number of reports since 1.15, while some go as far back as 1.8, possibly earlier (I've never seen this happen in 1.6.4, modded or otherwise, even with 100+ crash reports in my MCP folder; the worst that has ever happened is level.dat becoming corrupted/deleted. part of this may be due to my own optimizations.fixes to the chunk loading process, as I haven't played strictly "vanilla" for years, my chunk loading code also throws away wrong-located chunks instead of attempting to fix them (which just causes them to overwrite other chunks) but I've never seen it report any errors after a crash, only partially decorated chunks when it happened during world generation):
Note that many reports do not involve improperly shutting down/quitting a world just normal gameplay, including going to another dimension and returning to a corrupted Overworld/Nether/End within the same session, though the cause of at least some cases could still be related; singleplayer runs an internal server on a separate thread and the way the client shuts it down can cause it to be terminated in mid-save, although I imagine if you are using Forge it still has the "shutting down internal server" feature that I added in my own mods from back when I used Forge in 1.6.2, which is a simple loop that forces the client to wait for the internal server to fully shut down, making it impossible to reproduce various other world corruption issues, such as worlds failing to delete properly (caused by the client trying to delete files that are still in use, and/or the internal server saving chunks after the client deleted the original files, causing new ones to be created):
I suspect that a more significant factor is the multithreading of world generation and chunk loading in newer versions, which requires cooperation between threads to ensure that they are fully aware of what other threads are doing and any changes to data are synchronized between them (by contrast, 1.6.4 has a single main server thread that does everything, plus a separate file I/O thread; despite this, performance is far superior to newer versions, and actually, excessive multithreading can slow things down if there are too many threads trying to run at once).
In general, newer versions are also far more fragile when it comes to bad chunk data; you can create a world in 1.6.4 and load it in 1.5.2 and nothing really bad will happen, other than missing blocks.items/entities, while if you do the same in 1.16-1.15 the entire world (or loaded chunks) will be totally reset as the game simply can't handle chunks with unknown data. This may also apply to using mods and changing or removing them, although I wouldn't expect this to cause dislocated chunks, only reset chunks, unless changes to world generation occurred, in which case a similar effect would occur if the biome/terrain changed (your screenshots appear to show a forest in the first one but not in the second, otherwise, the shoreline seems to be the same so the chunks in the water are likely dislocated, not reset. It is also possible that you are actually experiencing MC-315, which is caused by the shutdown issue mentioned above; the best way to avoid this is to delete a world after reloading the game, or manually delete it outside of the game).
Some background:
I was testing out a port I made of Resource Overhaul (my mod) to 1.15.2. Though it does technically effect World Generation, it only adds new ores in that respect and should not be what is causing the strange world generation pictured. To confirm this, I recreated several worlds, including the exact seed, with the mod and loaded many chunks without this happening.
What happened:
I loaded into a world and well... let the picture speak for itself.
Again I am fairly certain it is not my mod causing this.
I recreated the same seed (same location) and it looks like this:
Has anyone else had this happen to them?
Any time you see 16 x 16 block of land that doesn't fit you are looking at a chunk error.
Possible causes can be turning off or losing power to your computer while Minecraft is trying to save the world but more common is creating a world with the same name as the one just deleted.
There may be other causes.
There are no dangerous weapons. There are only dangerous people. R.A. Heinlein
If you aren't part of the solution, then you obviously weren't properly dissolved.
The latest release of Amidst, version 4.6 can be found here:
https://github.com/toolbox4minecraft/amidst/releases
You should probably also read this:
https://www.minecraftforum.net/forums/mapping-and-modding-java-edition/minecraft-tools/2970854-amidst-map-explorer-for-minecraft-1-14
You can find me on the Minecraft Forums Discord server.
https://discord.gg/wGrQNKX
This is a fairly common(?) issue affecting newer versions, with the greatest number of reports since 1.15, while some go as far back as 1.8, possibly earlier (I've never seen this happen in 1.6.4, modded or otherwise, even with 100+ crash reports in my MCP folder; the worst that has ever happened is level.dat becoming corrupted/deleted. part of this may be due to my own optimizations.fixes to the chunk loading process, as I haven't played strictly "vanilla" for years, my chunk loading code also throws away wrong-located chunks instead of attempting to fix them (which just causes them to overwrite other chunks) but I've never seen it report any errors after a crash, only partially decorated chunks when it happened during world generation):
MC-161823 Chunks can occasionally be misplaced upon loading a world (there are many older reports which were marked as fixed, only for more reports to be opened indicating that this wasn't the case)
Note that many reports do not involve improperly shutting down/quitting a world just normal gameplay, including going to another dimension and returning to a corrupted Overworld/Nether/End within the same session, though the cause of at least some cases could still be related; singleplayer runs an internal server on a separate thread and the way the client shuts it down can cause it to be terminated in mid-save, although I imagine if you are using Forge it still has the "shutting down internal server" feature that I added in my own mods from back when I used Forge in 1.6.2, which is a simple loop that forces the client to wait for the internal server to fully shut down, making it impossible to reproduce various other world corruption issues, such as worlds failing to delete properly (caused by the client trying to delete files that are still in use, and/or the internal server saving chunks after the client deleted the original files, causing new ones to be created):
https://www.minecraftforum.net/forums/minecraft-java-edition/discussion/2579965-how-to-get-an-error-level-save-conflict-in?comment=3
I suspect that a more significant factor is the multithreading of world generation and chunk loading in newer versions, which requires cooperation between threads to ensure that they are fully aware of what other threads are doing and any changes to data are synchronized between them (by contrast, 1.6.4 has a single main server thread that does everything, plus a separate file I/O thread; despite this, performance is far superior to newer versions, and actually, excessive multithreading can slow things down if there are too many threads trying to run at once).
In general, newer versions are also far more fragile when it comes to bad chunk data; you can create a world in 1.6.4 and load it in 1.5.2 and nothing really bad will happen, other than missing blocks.items/entities, while if you do the same in 1.16-1.15 the entire world (or loaded chunks) will be totally reset as the game simply can't handle chunks with unknown data. This may also apply to using mods and changing or removing them, although I wouldn't expect this to cause dislocated chunks, only reset chunks, unless changes to world generation occurred, in which case a similar effect would occur if the biome/terrain changed (your screenshots appear to show a forest in the first one but not in the second, otherwise, the shoreline seems to be the same so the chunks in the water are likely dislocated, not reset. It is also possible that you are actually experiencing MC-315, which is caused by the shutdown issue mentioned above; the best way to avoid this is to delete a world after reloading the game, or manually delete it outside of the game).
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?
This comment was very insightful, thank you for the information. I've never seen this before so I was really confused.