Steps to reproduce
Player 1 relogs in the nether.
Player 2 loads the overworld, either by relogging there (saved & quit before player 1 relogged) or by relogging in the nether and travelling to the overworld.
Player 1 goes to the overworld at the same location as player 2.
Player 2 travels away from player 1 to a distance greater than player 2's render distance + 5 chunks.
Expected result
The game continues to simulate chunks around player 1.
Observed result
The game stops simulation in chunks that are > player 2's render distance + 5 chunks away from player 2, for all entities that existed when player 2 loaded the overworld, except for player-launched projectiles which fall through the world into the void (MCPE-141539). Player 1 is still able to move and interact in some ways, but many entity updates get postponed until player 2 returns to within the threshold distance.
The attached video
[media]demonstrates this. White and blue blocks mark the Z = 47-48 chunk border. Steve's render distance is 6, and when he crosses the Z = 223-224 chunk border the entities on the left stop and tridents there fall, then when he crosses Z = 239-240 entities on the right stop and tridents there fall.
[media]is the test world used in the video.
Simulation distance reaches to 14 chunks, but only for a single player. Crossing into the 15th chunk for the one player while two are on Realms activates simulation. That player backs out of the chunk by one block, other player witnesses simulation pausing. Villagers stop bouncing in water, animals stop moving, other player without simulation cannot push minecarts, hoppers stop transferring items. Sounds remain active for nearby player while simulation is suspended. Simulation-active player leaves Realm, simulation resumes for sim-locked player, and effect is transferred to the newly logged in player that previously activated the simulation at significantly greater range than the 4 chunks applied by Realms worlds. Root cause unknown. Both players logging out simultaneously resolves the issue, but can reoccur.
Related issues
is duplicated by
MCPE-142206
Mobs freeze
MCPE-143275
All Entities Freeze When Another Player Moves Away From The Area
MCPE-143503
Invincible Mobs
MCPE-144655
Passenger Minecarts get stuck for guests only
MCPE-145165
Villagers, iron golems and horses jamming
MCPE-145348
Spawn in mobs still freeze if the person who spawns them in walks away. Multi-player
MCPE-145962
minecart with hopper disappeared
MCPE-148892
Mobs Freeze Blocks Away!
REALMS-9488
Frozen Villagers (full village frozen)
REALMS-9828
Dungeon Monster Spawners are broken
relates to
Attachments
Comments
migrated
migrated
GoldenHelmet
Thank you for your report. It sounds like something odd is happening, but I want to be clear on the terms. "Simulation distance" in Minecraft refers to the chunks around each player where blocks, entities, and redstone power continuously update. On Realms, simulation distance is locked to 4 chunks. By contrast, render distance controls how far you can see blocks. (Entities like mobs and items are visible only within simulation distance, up to a maximum of 70 blocks.) Render distance on Realms is capped at 12, but blocks are actually loaded 2 chunks beyond render distance so the threshold of 14 to 15 chunks makes sense that way. I think what you are describing is that when player A on a realm moves to where player B's chunk would not be rendered (15 chunks away from player B), it stops simulation around player B.
I have some questions about this:
Can player B interact with the world at all?
Is the "locked" player always the one that has been on the Realm a shorter time? or longer time?
Is the not-locked player always the Realm owner?
Does this happen in multiplayer outside of realms? (Don't worry if you're not sure, we can test that later.)
Can you make a video of this happening?
GoldenHelmet
Thank you for your report. It sounds like something odd is happening, but I want to be clear on the terms. "Simulation distance" in Minecraft refers to the chunks around each player where blocks, entities, and redstone power continuously update. On Realms, simulation distance is locked to 4 chunks. By contrast, render distance controls how far you can see blocks. (Entities like mobs and items are visible only within simulation distance, up to a maximum of 70 blocks.) Render distance on Realms is capped at 12, but blocks are actually loaded 2 chunks beyond render distance so the threshold of 14 to 15 chunks makes sense that way. I think what you are describing is that when player A on a realm moves to where player B's chunk would not be rendered (15 chunks away from player B), it stops simulation around player B.
I have some questions about this:
Can player B interact with the world at all?
Is the "locked" player always the one that has been on the Realm a shorter time? or longer time?
Is the not-locked player always the Realm owner?
Does this happen in multiplayer outside of realms? (Don't worry if you're not sure, we can test that later.)
Can you make a video of this happening?
migrated
Player B can interact with some things in the world in a limited capacity. Buttons and levers activate with a sound, but do not initiate redstone signals. Villagers can have a trade window opened, but cannot be traded with even given the appropriate items for the trade. Entities cannot be “pushed” by the player including mobs/villagers and minecarts. Mining still seems possible in all instances of this happening. (Edit: chests can be opened while “locked” and villagers can be traded with but exp** for the villager does not increase. I misunderstood how it had been communicated to me while we were testing)
The “locked” player is the player who has been on the realm the least. This effect transfers from one player to the other when the “locked” player logs off and rejoins the Realm. It transfers again when the newly “unlocked” player leaves and rejoins without the other player leaving.
The “locked” player is usually the Realm owner when we notice the issue is occurring. Today we encountered this again, and the issue was identical.
We have not encountered this issue in multiplayer games off of realms, but we also do not spend significant time doing things in a testing world far apart from each other - and even then the Simulation and Render distances are different than on Realms so there are too many variables for me to confirm similarities off of realms.
Unfortunately neither of us have a PC to render any video. Given that a video recorded demonstration of the issue would need a side by side view of two different players views, I apologize that this is beyond our technical ability to provide.
I am aware of the simulation and render distance limitations of Realms, and believe I was using the terms correctly. Your statement that once the “unlocked” player approaches within Render Distance of the “locked” player, simulation resumes - but it does not resume in the next chunk farther away from the “unlocked” player from the “locked” player’s perspective. This was tested with various passive mobs being in visual range of the “locked” player in another chunk away from the “unlocked” player still being frozen as though they are rendered from the “locked” players perspective but not simulating. This issue occurred again today, and we were both doing different activities than when it has previously manifested. The locked player in this case was near our base while the other was quite distant clearing a desert. Repeating the distance tests once noticed yielded identical results to simulation activation. We were not able to test if chunks that would be out of simulation distance between our independent locations were simulating, given that 4+4-15 should leave 7 chunks where we wouldn’t have any way of observing for simulation status.
migrated
Player B can interact with some things in the world in a limited capacity. Buttons and levers activate with a sound, but do not initiate redstone signals. Villagers can have a trade window opened, but cannot be traded with even given the appropriate items for the trade. Entities cannot be “pushed” by the player including mobs/villagers and minecarts. Mining still seems possible in all instances of this happening. (Edit: chests can be opened while “locked” and villagers can be traded with but exp** for the villager does not increase. I misunderstood how it had been communicated to me while we were testing)
The “locked” player is the player who has been on the realm the least. This effect transfers from one player to the other when the “locked” player logs off and rejoins the Realm. It transfers again when the newly “unlocked” player leaves and rejoins without the other player leaving.
The “locked” player is usually the Realm owner when we notice the issue is occurring. Today we encountered this again, and the issue was identical.
We have not encountered this issue in multiplayer games off of realms, but we also do not spend significant time doing things in a testing world far apart from each other - and even then the Simulation and Render distances are different than on Realms so there are too many variables for me to confirm similarities off of realms.
Unfortunately neither of us have a PC to render any video. Given that a video recorded demonstration of the issue would need a side by side view of two different players views, I apologize that this is beyond our technical ability to provide.
I am aware of the simulation and render distance limitations of Realms, and believe I was using the terms correctly. Your statement that once the “unlocked” player approaches within Render Distance of the “locked” player, simulation resumes - but it does not resume in the next chunk farther away from the “unlocked” player from the “locked” player’s perspective. This was tested with various passive mobs being in visual range of the “locked” player in another chunk away from the “unlocked” player still being frozen as though they are rendered from the “locked” players perspective but not simulating. This issue occurred again today, and we were both doing different activities than when it has previously manifested. The locked player in this case was near our base while the other was quite distant clearing a desert. Repeating the distance tests once noticed yielded identical results to simulation activation. We were not able to test if chunks that would be out of simulation distance between our independent locations were simulating, given that 4+4-15 should leave 7 chunks where we wouldn’t have any way of observing for simulation status.
GoldenHelmet
Moving this report back to the MCPE project because I have reproduced the bug in local multiplayer (both LAN and Microsoft account connected multiplayer).
GoldenHelmet
Moving this report back to the MCPE project because I have reproduced the bug in local multiplayer (both LAN and Microsoft account connected multiplayer).
GoldenHelmet
In 1.18.0 the tridents no longer fall through blocks into the void (MCPE-141539 is fixed) but everything else happens as before the update.
GoldenHelmet
In 1.18.0 the tridents no longer fall through blocks into the void (MCPE-141539 is fixed) but everything else happens as before the update.