[MC-342] Positions of entities become offset to clients in SMP Created: 25/Oct/12  Updated: 19/Feb/21  Resolved: 23/Apr/19

Status: Resolved
Project: Minecraft: Java Edition
Component/s: None
Affects Version/s: Minecraft 1.4.2, Minecraft 1.4.6, Minecraft 1.4.7, Snapshot 13w04a, Snapshot 13w10b, Minecraft 1.5, Minecraft 1.5.2, Snapshot 13w23b, Minecraft 1.6.1, Minecraft 1.6.2, Minecraft 1.6.4, Minecraft 1.7.4, Minecraft 1.8.9, Minecraft 15w51b, Minecraft 16w06a, Minecraft 1.10.2, Minecraft 17w06a, Minecraft 1.13.2
Fix Version/s: None

Type: Bug
Reporter: WolfieMario Assignee: Unassigned
Resolution: Awaiting Response Votes: 23
Labels: multiplayer, rendering
Environment:

Windows 7 64-bit, 4GB RAM


Issue Links:
Duplicate
is duplicated by MC-8451 Players in Multiplayer will "Sink" in... Resolved
is duplicated by MC-11289 Lan game bug Resolved
is duplicated by MC-16782 multi player floaters Resolved
is duplicated by MC-96000 Blocks that can't be placed will not ... Resolved
is duplicated by MC-197 Walking through Solid Resolved
Relates
relates to MC-2783 Minecarts sometimes appear off the track Reopened
CHK:
Confirmation Status: Confirmed
Category:
(Unassigned)

 Description   

In an adventure map in SMP, I was experiencing strange issues where the locations of other players seemed to be offset. This would occur mainly in cramped places (on the map in question, in the dungeon at the top of the mountain).

I would notice players passing through walls, and sometimes floating or "swimming" in the ground. At some times, the offset seemed to be so bad that players would be rendered several blocks away from their true locations.

Logging out and back in would sometimes correct this issue, but it would frequently return.

I noticed this issue with:

  • Players who were far away and made a long journey
  • Players who were teleported to me
  • Players who I was with during a long journey (namely, up the mountain in the map)

I was running the server on my own computer, and I was connected via localhost. The other players were connected from various places in different states. As my computer is not very good, I noticed a few "<Playername> moved wrongly!" messages in the console. This may be related, but I did not notice any direct connection between the messages and the phenomenon.

This issue has occurred before, on the 1.4.0 prerelease, as well, and is not an isolated incident.

EDIT: I should also note that this never happens to me on Bukkit servers, whether or not I am the host. A server host I spoke with said he has noticed this extensively and exclusively on vanilla servers when he hosts them.

Video footage can be seen in this Sethbling livestream, starting after a series of teleports at 1:38:10. Note that frequent complaints of players in the wrong position (or "invisible") have occurred throughout the video. This server is vanilla 1.4.6 (or possibly 1.4.7, but the two are interchangeable as only one class differs between the two). Sethbling may or may not be the server host; I'm sure Marc Watson of Mojang can clarify (he was present).

EDIT: Since then, I have realized this effect can be reproduced reliably on minecart entities. Steps to reproduce:

  1. Run your own server
  2. Set up a minecart rail loop with powered rails and a two block drop
  3. Add a minecart, and make sure it goes around the loop
  4. Observe that, after some time (this seems random), the minecart gets misaligned: although it tracks the same path, it may appear floating or offset by several blocks.

In my experience, lag on my system makes this effect more likely, as does moving around.



 Comments   
Comment by [Bot] Arisa [ 09/Feb/21 ]

This report is currently missing crucial information. Please take a look at the other comments to find out what we are looking for.
If you added the required information and a moderator sees your comment, they will reopen and update the report. However, if you think your update to this report has been overlooked or you want to make sure that this report is reopened, you can contact the Mojira staff on Discord or Reddit.
– I am a bot. This action was performed automatically! Please report any issues on Discord or Reddit

Comment by [Bot] Arisa [ 11/Jan/21 ]

This report is currently missing crucial information. Please take a look at the other comments to find out what we are looking for.
If you added the required information and a moderator sees your comment, they will reopen and update the report. However, if you think your update to this report has been overlooked or you want to make sure that this report is reopened, you can contact the Mojira staff on Discord or Reddit.
– I am a bot. This action was performed automatically! Please report any issues on Discord or Reddit

Comment by [Mod] Pokechu22 [ 23/Apr/19 ]

It also happened in Etho's LP, some time in episode 518... but I don't recall exactly when. It's worth noting that that's a singleplayer vanilla world, and again it's a minecart. But it doesn't help too much with reproducing :/

Comment by Galaxy_2Alex [ 23/Apr/19 ]

I've tried to reproduce the issue by having a setup where a villager in a cart goes to chunks that are constantly loaded and unloaded, but I am unable reproduce the issue. Does anyone have a reliable way of reproducing this issue? If not, I will close this within a few weeks.

The video provided doesn't really tell us much about the server itself, it is possible the server had modificiations installed.

Comment by [Mod] Pokechu22 [ 17/Feb/19 ]

Shown to still exist in this video (at 1:15:36), which I'd say is sufficient evidence that this still exists (even though they're using optifine) when combined with the lack of any protocol changes regarding entity motion. It's not exactly the same setup as in the description, but it is a card going some distance and then getting desync'd.

Comment by Fabian Röling [ 05/Apr/18 ]

Does the way to reproduce in the description still work?

Comment by [Mod] Pokechu22 [ 10/Dec/16 ]

I haven't reproduced it, but relative movement still is used, so it can hypothetically happen. I haven't seen it actually happen though, so if anyone has seen it, that would be useful information.

Comment by CDES5 (Inactive) [ 10/Dec/16 ]

Is this still a problem in Minecraft 1.11?

Comment by [Mod] Pokechu22 [ 06/Aug/16 ]

I personally haven't reproduced this in 1.10.2 but I did ask other players on a server I play on and they had. So it still seems to be present.

Comment by user-f2760 (Inactive) [ 04/Aug/16 ]

Is this still an issue in 1.10.2?

Comment by Jess Edward [ 12/Feb/16 ]

16W06A Player falls onto active redstone then sinks into block. Thus far cannot reproduce purposely altho occurred several times. Possibly related to early Elytra flight.

Comment by Robin T. [ 18/Dec/15 ]

Can experience this with Minecarts after a while in 15w51b

Comment by Galaxy_2Alex [ 23/Oct/14 ]

Is this still a concern in the current Minecraft version 1.8.1 Prerelease 3 / Launcher version 1.5.3 or later? If so, please update the affected versions in order to best aid Mojang ensuring bugs are still valid in the latest releases/pre-releases.

Comment by [Mod] Ezekiel (ezfe) [ 22/Mar/14 ]

Is this still a concern in the latest Minecraft version 14w11b? If so, please update the affected versions in order to best aid Mojang ensuring bugs are still valid in the latest releases/pre-releases.

Comment by stuartmorse [ 27/Nov/13 ]

I have also seen the player bug on that same 1.6.4 server

Comment by stuartmorse [ 27/Nov/13 ]

i'm pretty sure i have seen the minecart bug on a 1.6.4 server.

Comment by Alexander Gundermann [ 12/Aug/13 ]

I recently made a bugfix mod for glitching mobs and I think this should be covered as well: https://github.com/taurose/Unglitch (4).

In the tracker code there's a flag that determines whether an entity has moved far enough to send a position update. Several lines later, if the flag is true, the tracker stores the entity's location that is supposedly sent to the players for later updates. However, it seems that over time Mojang added several conditions to circumvent that flag and send or not send position updates regardless, which causes the "last sent" locations to be off. And once that happened, the clients receive false position updates and the entity appears offset.
Luckily, this can be fixed quite easily with a few lines once you know the problem.

Comment by fake fakes [ 11/Jun/13 ]

Still in 13w23b

Comment by WolfieMario [ 11/May/13 ]

I found a video where the bug is extensively demonstrated and tested . Also, note at the end the client-side offset arrows. Considering this bug pertains to entity offset in SMP, it may be related, but probably isn't. But all of the teleportation bugs are textbook demonstrations of this bug.

Comment by Aleksi Postari [ 03/Apr/13 ]

So far I've mostly managed to notice this with the player's not appearing to others after doing teleports. Been there for ages, as far as from beta 1.8. Players mostly appear back when you drop down a block below. Disappearing part seems to have something to do with the entitytracker, in MCP it looked like that player was not in the server's entity tracker after doing a teleport and player positions where therefore not syncronized to other players (until the player dropped a block below and then was added back into the entity tracker).

Comment by WolfieMario [ 18/Mar/13 ]

I've managed to get this now in 1.5 on minecarts, so it's not limited to players. Which is pretty bad, to be honest, but at least it makes it easier to test.

Comment by Tails [ 15/Jan/13 ]

Reopened.

Comment by WolfieMario [ 15/Jan/13 ]

Oh, sorry, I had forgotten about this, as it's been a while since I've dealt with vanilla servers (I will note that I haven't experienced it in Bukkit whatsoever).

However, I recently saw Sethbling livestreaming on a vanilla 1.4.6 server: http://www.twitch.tv/sethbling/b/355261121

At 1:38:10, you can clearly observe that, after a series of teleports, players appear in the wrong positions for Sethbling (including the blaze-suit player sinking into the ground at 1:38:30, and the flying cactus-suit player just before that). Sethbling and other players were also complaining about "invisible" players, and frequent relogs were needed to correct the issue.

So, can this be reopened?

Generated at Sun Jan 12 11:50:42 UTC 2025 using Jira 9.12.2#9120002-sha1:301bf498dd45d800842af0b84230f1bb58606c13.