Ben Lewis

Similar to this, I got a pigman riding a baby strider riding an adult strider:

The comment about hitting blocks causing damage is incorrect.

From my observations, a ghost/duplicate of the player is created/not properly removed where the player died. Thus, mobs will attack that ghost player, or if in somewhere like lava, it will continue to take damage and kill the still living (newer) player.

I've also had this, but I died in the Nether. I then switched to creative and tried to fly through a portal but it would not let me. Every time I subsequentially log in, I die, click respawn, and respawn trapped inside my character model unable to move or break blocks.

I noticed it didn't happen on my SMP server, but did happen on two worlds I'd upgraded from 20w12a.

I've attached debug-report-2020-02-29_19.39.40, a 5 minute profile said out TPS was 10.99 at the time. We had around 7 users online.

Cows like many other passive mobs require grass to spawn on even when spawning from a mob spawner, this is like skeletons require a light level of 7 or below even when spawning from a mob spawner. So this behaviour is intentional.

The problem with detecting dead players which led to this 'bug' is that they are at the location that they died a solution to this would be for dead players to have no location in the context of commands. For example /testfor @a[r=10] would never select a dead player.

After upgrading to OS X 10.9 Mavericks(update released today) this bug seems to be fixed for me, the window resizes as expected without restarting my computer as it did before the upgrade.

This bug still occurs in the latest snapshot 13w41a, i tried expanding the minecraft window (when in windowed mode). My screen went black and then my macbook restarted.

@Patrick There is no longer any limit on command block length but, the /tellraw "run_command" command executes the command as if the player typed the command into chat. So as there is a limit on the length a chat message can be, so this bug occurs.

How the /tellraw command (I assume) is implemented is the JSON bit of the command is sent to the applicable clients where they do all the logic from there. A possible solution is to split the logic between the server and client so the client sends the server a click event where the server executes the command(maybe on behalf of player). This would also mean non-Op issue will also be fixed. But it might be hard to split up the logic and Dinnerbone intended it to be all client side.