Stop overwriting block locations, root issue looks to be the TE fixer itself

Fixes GH-352

Spigot-Server-Patches/Don-t-try-and-fix-TileEntities-as-they-are-removed.patch

@@ -1,13 +1,16 @@
 From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
 From: Zach Brown <1254957+zachbr@users.noreply.github.com>
 Date: Sat, 25 Jun 2016 23:55:56 -0500
-Subject: [PATCH] Move setBlock tile->tile transition workaround into TE
- removal code and out of setblock
+Subject: [PATCH] Don't try and fix TileEntities as they are removed
 
-Works around an issue in which running it in setBlock for all block changes could cause sign (and other?) data corruption.
-Also tells CB's TE fixer to ignore removals, as it would end up in a state in which it'd set the block, then
-call TE removal, which would get the TE to remove, then making CB's fixer run and get confused as to the current state of
-the block, but we don't care because that's the whole reason we're removing it.
+Currently, CraftBukkit tries to fix TEs that do not match the present block at the location. This is normally good,
+however, this same fixer runs when the TE removal functions go through to remove a TE after its block has been changed.
+So a block will be changed, the server will attempt to remove the TE present, but will then get caught up in CB's overzealous
+TE fixer. That fixer checks the block against the TE present, and throws a fit because it doesn't match. Which is why we're
+removing it in the first place.
+
+The 'fix' to this issue is to skip the fixer entirely when we're removing the TE, as it shouldn't ever need to run
+then anyway, we're removing it.
 
 diff --git a/src/main/java/net/minecraft/server/BlockChest.java b/src/main/java/net/minecraft/server/BlockChest.java
 index 0000000000000000000000000000000000000000..0000000000000000000000000000000000000000 100644
@@ -61,18 +64,6 @@ index 0000000000000000000000000000000000000000..00000000000000000000000000000000
  
                  if (tileentity instanceof TileEntitySkull) {
                      TileEntitySkull tileentityskull = (TileEntitySkull) tileentity;
-diff --git a/src/main/java/net/minecraft/server/BlockTileEntity.java b/src/main/java/net/minecraft/server/BlockTileEntity.java
-index 0000000000000000000000000000000000000000..0000000000000000000000000000000000000000 100644
---- a/src/main/java/net/minecraft/server/BlockTileEntity.java
-+++ b/src/main/java/net/minecraft/server/BlockTileEntity.java
-@@ -0,0 +0,0 @@ public abstract class BlockTileEntity extends Block implements ITileEntity {
-     public void remove(World world, BlockPosition blockposition, IBlockData iblockdata) {
-         super.remove(world, blockposition, iblockdata);
-         world.s(blockposition);
-+        world.setTypeAndData(blockposition, Blocks.AIR.getBlockData(), 0); // Paper - SPIGOT-611 workaround moved from setBlock
-     }
- 
-     public boolean a(IBlockData iblockdata, World world, BlockPosition blockposition, int i, int j) {
 diff --git a/src/main/java/net/minecraft/server/World.java b/src/main/java/net/minecraft/server/World.java
 index 0000000000000000000000000000000000000000..0000000000000000000000000000000000000000 100644
 --- a/src/main/java/net/minecraft/server/World.java
@@ -131,7 +122,7 @@ index 0000000000000000000000000000000000000000..00000000000000000000000000000000
          IBlockData blockData = getNMSBlock(type).fromLegacyData(data);
          BlockPosition position = new BlockPosition(x, y, z);
  
-+        // Paper start - Moved to TileEntity removal
++        // Paper start - Unnecessary, root issue is CB TE's fixer getting too involved. Also potential data issues.
 +        /*
          // SPIGOT-611: need to do this to prevent glitchiness. Easier to handle this here (like /setblock) than to fix weirdness in tile entity cleanup
          if (type != 0) {

scripts/importmcdev.sh

@@ -44,7 +44,6 @@ import BlockFluids
 import BlockFurnace
 import BlockIceFrost
 import BlockPosition
-import BlockTileEntity
 import ChunkCache
 import ChunkProviderFlat
 import ChunkProviderGenerate