Mark Means

I have been testing the attached world in 1.19. I have to say that out of 5 spawns, there was ONE TIME a trader spawned just a few platforms below the bed. All others were very deep underground. But now it's not always ALWAYS deep underground as far as I can tell. 

I've confirmed that this is still a problem in 1.18.2.

These tests were performed with a new copy of Windows 10 Pro version 21H1 installed in a Hyper-V Virtual Machine with BDS 1.18.2.03 and TrapDoor BDS addin to profile server performance and display the TPS rates using the TrapDoor command "/prof". (It's highly recommended to use a VM to do this test because using the TrapDoor plugin triggers malware warnings).

I tested the turtle farm using the JCPlayz farm design. Blocking the path to the ocean and placing and array of 24 four egg spots and setting time via the BDS console to 21000-ish to keep it night so the eggs continue to hatch. When babies hatched they stayed stationary. As the clusters of eggs hatched, TPS dropped. After all were hatched I unblocked access to the ocean-based kill chamber and TPS immediately increased to normal.

On a second test, I allowed all baby turtles that are prevented from pathfinding to the ocean to grow up to adult. As they grew to adult TPS improved even though ocean access was still blocked.

So, I have confirmed that the issue applies only to farms that prevent baby turtles from pathfinding upon hatching. This would apply to initial setup of farms like the JCPlayz design only when the player is creating the initial breeding stock.

I have dozens of screen shots documenting the tests and can provide a configured Hyper-V VM (about 20 GB file) with the BDS and TrapDoor installation upon request by Mojang.

I've tracked the lag effect on a nearby furnace piston feed tape. With the ceiling maps in place the four pistons would cycle once every 5.5 seconds. After removing the 418 maps, but leaving the item frames in place, the piston feed tape would cycle every 1.7 seconds. So the maps were causing 3.5x slower piston operation as well as other overall lag. It should be noted that two other players on the Realm, hundreds to thousands of blocks away reported bad lag when the maps were in place. One reported that he couldn't even fire a bow.

I am experiencing this in a realm with version 1.14.60. I built a 28 block diameter round building with multiple floors. The ceiling of the lower floor is a lit up white drop ceiling created by placing Jack o'Lanterns on top of slabs, then removing the slabs and placing an item frame and map on the bottom of each Jack o'Lantern. The maps were created by levelling an entire 128x128 (map size) tundra field to create an all white map. Then used an anvil to change the map name to a space character, then made enough copies of the map to cover the ceiling. The top of the Jack o'Lanterns is covered with carpet and is the floor of the second story. Now I see lag when eating food or breaking blocks. When breaking blocks the item breaks but jumps around for several seconds before it finally falls to the ground. The attachment Minecraft 4_19_2020 10_10_22 shows the room with the large map filled ceiling.

Also happens in Win10 v1.14.60. Tested In a snowy tundra in survival on realm. Tundra levelled flat to y67, While standing in the middle and watching a level 0 map while snowing, map slowly fills with white but suddenly the entire map changes and reverts to just before the snowfall. Tried placing snow top layers on all uncovered blocks and as well. Even my manually placed top snow layers revert. No lighting anywhere. Tried placing double top layers and it didn't seem to make a difference; they would revert as well. I haven't tried full snow blocks.

in addition, snow layers are melting DURING a snow storm.

Since this is a parity issue between Java and Bedrock I think the issue should be reopened; or at least a reason provided why it is "working as intended" and why it is intended (i.e. what benefit exists for it) to be working differently than Java.

So looking at the videos posted 24/jan/20 I notice the + cursor movement against the terrain. It looks like it moves in a U shape on Bedrock and just up and down in Java. It appears that the center of the tilt rotation (vertical rotation plane perpendicular to the movement vector) and the forward focal point angle (translation and rotation in the horizontal plane centered on a point far ahead) looks a little off to me. Not that it's bad. I think 1.14.1 looks fine on a HD PC monitor, just a little different. But for parity purposes I think they should be identical as long as that doesn't cause other problems with smaller screens.

However, when looking that the left edges of the screens for total apparent movement for any a given rotation and translation algorithm, the amount of "bobbing" at the edges will be different for a screens depending on the aspect ratio (4:3, 16:9, 18:9) and on the FOV setting. Maybe an optional compensation setting to account for screen width and FOV is desired in this case.

A better suggestion may be to replace the bobbing toggle to a slider or two and let users decide how much bobbing they want from none to ridiculously nauseating. That way it negates any future complaint if the user can set it how they like it.

To help the developers, I've attached small world K1 that demonstrates the problem. Just place minecarts as shown in the video the OP included and you can easily test this.

This bug also affects bedrock 1.10.1

This is also happening in 1.10.1. Here are the repro steps:

Create a small wheat farm. Wall the farm in on three sides with fence. In the forth side place a row of hoppers above the ground leading to a chest. Place a row of blocks two blocks above the hoppers so any villagers can't get in or leave the farm once it is closed. Make sure there is one brown robed villager in the farm. Place other villagers outside of the farm next to the row of hoppers, fence them against the hoppers so they continually beg for food. When the farmer tries to throw bread to the other villagers, the row of hoppers will intercept the bread. Once that starts, cut down all the wheat. The farmer villager should have 8 inventory slots, so at most you should get 8 stacks of bread, but it keep coming, overfilling a double chest.

My guess is that the farmer villagers wheat inventory is not being decremented when the bread is thrown unless another villager actually picks it up. This issue doesn't occur with carrots, potatoes, or beetroot; so it is something unique to bread that is causing endless bread generation.

In Windows 10 version 1.8.1 survival mode it also has this problem. After about 10 hours of play of hours of play I explored to find a village. There were cats in the village when I found it. This was after not sleeping for many days. I didn't tame any of the stray cats then. The next day, no stray cats. After several hours, stray cats appeared one day. This was also after several days not sleeping and phantoms appeared the previous night. That day I tamed two. I didn't count the days that had passed but it must have been 30+ days. Since the second occurrence of stray cats the village was extensively upgraded, fences and light and the village grown to the point of self spawning an iron golem, but I have never seen another stray cat. I tried not sleeping for 5 days, thinking that phantoms might need to exist first, since cats repel phantoms, but this tactic didn't pan out. Still no more stray cats. I hope this heps devs track the causes or helps players figure out a workaround.