RimaNari

Ah, sorry - you are able to reproduce the bug, but you are not able to reproduce what you observed in your comment three days ago. That's what you meant, okay. Everything fine 🙂

I assume you forgot a negation in your comment, actually meaning that you are not able to reproduce your own report. You wrote that you were able to reproduce again, but apparently that's a typo.

I can imagine that maybe you didn't use a fresh copy of the test world, but one that was opened already. If one has several copies of a world in the saves folder, maybe one original and one that was opened in another snapshot already, it certainly gets confusing quickly. It doesn't help that Minecraft displays the world's name (which is the same for two copies of a world) as the bold-faced largest piece of text of a save, and the folder name only smaller and less prominent. So that would be one possible explanation.

Anyways, glad that I could help 🙂 

I cannot confirm what you are saying @carpet0928 .

When I load in 14w45a and enter the end, the exit portal is at its usual position slightly offset from 0,0 and there is no ender dragon. I cannot respawn the dragon by the expected process, but once I break a bedrock block from the exit portal and place a single end crystal on it, it triggers the creation of another exit portal at 0,0. Then I can respawn the dragn on the new portal at 0,0.

So it's the same as it always was for me, and the bug persists... not sure what you did?

I was able to workaround MC-137300 by loading the world in 1.13.1 first and then loading it in 1.13.2-pre1. This should not interfere with reproducing this bug.

I can confirm for 1.13.2-pre1.

Confirmed for 1.13.1 (I opened another, now duplicate-marked, report, not having found this one).

I created two separate bug reports.

Layout issues: MCL-9774

Missing server download: MCL-9775

This can be closed then 🙂

I'm sorry, I was imprecise: I couldn't find the "Download server" option for any version at all anymore. However this was on another machine than the one I am on usually and there the launcher windows was smaller in horizontal size than on my other machine. Apparently if one makes the launcher window horizontally smaller beyond a certain point both the buttons next to the game directory text box as well as the "Download server" link aren't displayed.

I find this quite a weird behaviour and would report it as its own bug report. Or is it intended to work this way @peterix?

Anyway, now that we are on the same page again, I agree of course - if you don't have those server JARs anyways then it all works correctly now. Just wondering why you don't have those JARs anymore - but maybe you never had them to begin with. I can't remember anymore so I guess you just didn't build server JARs before 1.2.5.

So yea, I guess this is fixed then.

I am confused, as I cannot find the download option for the server JAR anymore. Am I blind, or has this option been removed altogether?

Please take my report with a grain of salt as I am running Minecraft on a potato. Due to this I had low FPS the whole time anyways, however I still noticed the GC spikes easily due to the F3 statistics freezing and because it created a system-wide noticeable (outside of the game) lag spike. I also observed the memory usage approaching 100 %.

Test: Buffet world with generator type "Surface" and biome "Dark Forest" and structure generation turned off. After world load no movement, but I did look around once in order to trigger the chunks getting loaded.

Minecraft version: 1.13-pre6

Launcher version: 2.1.1216

JVM arguments:

-Xmx1G -XX:+UnlockExperimentalVMOptions -XX:+UseG1GC -XX:G1NewSizePercent=20 -XX:G1ReservePercent=20 -XX:MaxGCPauseMillis=50 -XX:G1HeapRegionSize=16M

Results: Loading the world with render distance 16 chunks went fine (tested for several minutes). Allocated memory was at ~ 60 %. Increasing the render distance to 32 chunks quickly allocated 100 % memory and within seconds the used memory approached 100 % and the first lag spike occurred. After this the memory usage was at 70 - 90 % and every minute or so suddenly went towards 100 % with a lag spike occurring. The results are identical for tests performed with low and high graphics settings ("Fancy" graphics, "Maximum" smooth lighting and "4" Mipmap levels vs. "Fast", "OFF" and "OFF").

Umm... why is this supposed to be fixed? Literally the first and obvious thing I tried - teleporting into a block and trying to break it - didn't work. Apparently, I am not looking at the block I am inside of.

I added "2.9.1003" to the affected versions, however my launcher shows me version "2.0.1003". I assume the version number in the launcher is incorrect.

So now I face the difficult decision: does one report a wrong version display as bug? If so, what would be the correct "affected version"? 😛

Wouldn't it be a good idea to update the description of the bug then? Currently it only explains the issue of the chat being shared across two servers. If this bug is to represent all symptoms that are caused by the same underlying bug of the connection not getting cancelled, a more general description would be appropriate.

Maybe a new description could contain both the current version explaining the issue with chat and my description of the teleporting. Then it would be clear that this is not only a single symptom (or even only a "minor" issue with the chat - the teleporting seems more important to me), but in fact an issue with a broad spectrum of symptoms.

This appears to be similar to a bug Xisumavoid showcased on Youtube. He demonstrates how this can be exploited to change the actual player position in a singleplayer world:

• In a singleplayer world you are at coordinates x, y, z.

• Log into a server (or possibly also another singleplayer world?) where you are at coordinates x', y', z'.

• Before loading has finished, quickly cancel joining the server and instead open the singleplayer world again.

• The client renders the server you were trying to join before canceling, however the actual block positions are different.

• Leaving and re-opening the singleplayer world removes render glitches - you see the singleplayer world again, but you remain at the coordinates of the server, effectively teleporting you from x, y, z to x', y', z' in the singleplayer world.

This bug can not only be exploited but can happen by chance easily. Additionally there may be further things that can be exploited about this.

@unknown, @unknown for discussions of the nature of "the developers should" / "the developers should not" please refer to the Mojira subreddit. You can create a text post there to discuss this matter.

The issue with entity controlling what they are riding is historical; Minecraft never underwent a proper design and planning phase prior to implementation (or at least not one that could have foreseen its future success). (When Notch made Minecraft and added minecarts your argument was the very other way around.)

The apparent cause - two falling sand entities at the same location for once and falling sand entity in lazy chunks - might be different but there are a lot of parallels. Falling sand entities in lazy chunks also loose their data value so there might be a common part to both under the hood.

And also, I guess this boosts both this and MC-81746, and TBH I think it's quite a significant bug now with concrete in the game.

Ah, my bad, sorry. Delete the entire folders region and DIM-1, these store the data for overworld and nether, respectively. Folder DIM-1 contains the data from the End, so delete any files with a number 2 or higher in the filename from there to reduce overall file size. These will contain the outer end regions which should not interfere in any way with the issues of interest here. Thanks for making the effort 🙂

@@unknown: Would you mind uploading your world anyways - in a state before you got the new portal to spawn? Strip the world of overworld and nether first though by removing the folders DIM and DIM-1 (only leaving DIM1) and if you explored a lot of the end already also delete all region files within DIM1 that have a number 2 or higher number in their file name. Then ZIP the world folder and attach it here.

@@unknown: This is the same situation as for @unknown and the same situation that I re-created for testing. The workaround I posted several comments up should apply, as the game will get triggered to create the correct exit portal and create the ExitPortalLocation entry in the NBT data correctly if it is missing. If the trick explained in the linked video doesn't work try the other method using creative mode (removing the bedrock block and placing the end crystal one block lower). This should create a new exit portal at coordinates 0, 0. If this does not work for you, can you upload your world (stripped of overworld and nether) for me and others to examine?

Deleting the gateway entries was probably contra-productive, however it won't interfere with the dragon respawn issue. I believe with only one gateway entry left the game will only create a single gateway now, no matter how often you kill the dragon. (You should probably use a made backup of your level.dat file to get the gateway entries back. Maybe deleting the entire DragonFight entry works as well, as the game should re-create it due to the legacy scanning once you enter the end.)

@@unknown: Please refer to MC-104736, as apparently any further discussion of this issue is supposed to happen there. (Although I believe the title of MC-104736 is not correct, as someone updating from 1.8 to e.g. 1.11.2 will experience an issue, but the world never got corrupted by MC-103497...)

You should find a workaround for situation 3 there.

I was able to re-create your situation by loading a pre-1.9 world in 1.11.2. The same holds true when in 1.10.2, so it's nothing new in the 1.11 version. It appears as if the above linked comment describing the four possible situations is incorrect: situation three is not fixed by the legacy scanning that kicks in when entering the end. My server log shows

[Server thread/INFO]: Scanning for legacy world dragon fight...
[Server thread/INFO]: Found that the dragon has been killed in this world already.

directly after entering the end, which I believe indicates that the legacy scanning indeed happens.

This issue is certainly not yet fixed - and also persists in 1.11.2. Can someone update the affected version list?

To workaround the issue: Switch top creative mode and break one of the bedrock blocks on the exit portal on which you would place your end crystals on top. Place a new bedrock block below the broken one and place an end crystal on top. This should trigger the creation of the correct exit portal (may take a couple of seconds) on which you can respawn the ender dragon as expected. To make sure the old exit portal will be disabled, return it to its original state (remove the end crystal and the placed bedrock block and add the bedrock block to the rim again) and spawn the dragon once (using the new portal). The game will then correctly find the old portal and replace it with end stone.