1.19 update #638

DarkLexFirst · 2023-10-23T10:18:24Z

No description provided.

1.19 update fixes.

kennyvv · 2023-10-23T10:55:25Z

Thank you for submitting! I'll need a bit to go through all of the changes, as said on discord.

kennyvv

Looks good! I still need to test it, but it looks very promising. Could you check my feedback? It's mainly a few small things

kennyvv · 2023-10-23T10:48:47Z

.github/workflows/dotnetcore.yml

@@ -4,6 +4,7 @@ on:
  push:
    branches:
      - master
+      - 1.19-update


Can you remove this?

kennyvv · 2023-10-23T15:56:22Z

src/MiNET/MiNET.Console/log4net.xml

@@ -107,7 +107,7 @@
 	</appender>

 	<root>
-		<level value="INFO" />
+		<level value="TRACE" />


Could you put the default back to INFO?

kennyvv · 2023-10-23T15:57:11Z

src/MiNET/MiNET.Test/GenerateBlocksTests.cs

+			// TODO - 1.19-update
+
+			//foreach (KeyValuePair<string, int> keyValuePair in ItemFactory.IdToRuntimeId)
+			//{
+			//	if(keyValuePair.Key.Equals("sapling")) Console.WriteLine(keyValuePair.Key);
+			//}
+			//var itemId = ItemFactory.GetItemIdByName("minecraft:sapling");
+			//Assert.AreEqual(6, itemId);
+
+			//Item item = ItemFactory.GetItem("minecraft:sapling");
+			//Assert.AreEqual("minecraft:sapling", item.Name);
+			//Assert.IsNotNull(item as ItemBlock);


Could you fix this? Could change it into ID based

I'm not sure it makes any sense now...

speaking in general, as a testing automation engineer in the past, I can say that the MINET auto tests look bad.

When I'm done with the features, I could write some good tests.

A lot of the current tests are... Not great, i agree with that. It would be nice to get some proper ones at some point!

I made it so that it would at least test something.

src/MiNET/MiNET/Blocks/Block.cs

src/MiNET/MiNET/Plugins/PluginManager.cs

src/MiNET/MiNET/Utils/IO/HighPrecisionTimer.cs

src/MiNET/MiNET/Worlds/BlockDataCompressionUtils.cs

kennyvv · 2023-10-23T16:43:44Z

src/MiNET/MiNET/Worlds/LevelDbProvider.cs

@@ -189,7 +189,9 @@ public ChunkColumn GetChunk(ChunkCoordinates coordinates, IWorldGenerator genera
 				if (flatDataBytes != null)
 				{
 					Buffer.BlockCopy(flatDataBytes.AsSpan().Slice(0, 512).ToArray(), 0, chunkColumn.height, 0, 512);
-					chunkColumn.biomeId = flatDataBytes.AsSpan().Slice(512, 256).ToArray();
+
+					// TODO - 1.20 - update


Any chance you could look into loading the biome id's in still?

This should be supported, since the client is currently working with 3d chunks.

should it be active?

kennyvv · 2023-10-23T16:45:03Z

src/MiNET/TestPlugin/CoreCommands.cs

+			// TODO - 1.20 - update
+			//for (int i = 0; i < chunk.biomeId.Length; i++)
+			//{
+			//	chunk.biomeId[i] = biomeId;


Any chance you could keep this working?

Why not? But before that, I need to rework data storage in chunks, since the current one is outdated and inconvenient.

When I'm done, it will turn into something like:

subChunk.Biomes.Fill(id);

should it be active?

1.19 update fixes.

src/MiNET/MiNET/Player.cs

kennyvv · 2023-10-24T17:32:32Z

src/MiNET/MiNET/Player.cs

@@ -1189,7 +1317,7 @@ public virtual void Teleport(PlayerLocation newPosition)
 				Monitor.Exit(_teleportSync);
 			}

-			MiNetServer.FastThreadPool.QueueUserWorkItem(SendChunksForKnownPosition);
+			//MiNetServer.FastThreadPool.QueueUserWorkItem(SendChunksForKnownPosition);


Is there a reason for not doing this anymore?

Chunks are often sent before the client can process them, because of this they are lost.
The most correct solution is to send them only after teleportation. After that, they will not go, since MINET thinks that the player has already received them.

kennyvv · 2023-10-24T17:32:35Z

src/MiNET/MiNET/Player.cs

@@ -1176,7 +1303,8 @@ public virtual void Teleport(PlayerLocation newPosition)
 						HeadYaw = 91,
 					});

-					ForcedSendChunk(newPosition);
+					//ForcedSendChunk(newPosition);


Is there a reason for not doing this anymore?

I think this particular case can be returned, but I'm afraid to cache it, since the player may lose it again.

kennyvv · 2023-10-24T17:32:42Z

src/MiNET/MiNET/Player.cs

@@ -1670,14 +1799,14 @@ public virtual void SpawnLevel(Level toLevel, PlayerLocation spawnPoint, bool us

 					SetNoAi(oldNoAi);

-					ForcedSendChunks(() =>
-					{
+					//ForcedSendChunks(() =>


Is there a reason for not doing this anymore?

Chunks are often sent before the client can process them, because of this they are lost.
The most correct solution is to send them only after teleportation. After that, they will not go, since MINET thinks that the player has already received them.

kennyvv · 2023-10-24T17:32:45Z

src/MiNET/MiNET/Player.cs

@@ -1659,7 +1787,8 @@ public virtual void SpawnLevel(Level toLevel, PlayerLocation spawnPoint, bool us

 				CleanCache();

-				ForcedSendChunk(SpawnPosition);
+				//ForcedSendChunk(SpawnPosition);


Is there a reason for not doing this anymore?

I think this particular case can be returned, but I'm afraid to cache it, since the player may lose it again.

kennyvv · 2023-10-24T17:42:10Z

src/MiNET/MiNET/Player.cs

@@ -2236,6 +2294,23 @@ public void HandleMcpeSubChunkRequestPacket(McpeSubChunkRequestPacket message)
 			SendPacket(response);*/
 		}

+		public virtual void HandleMcpeRequestAbility(McpeRequestAbility message)
+		{
+			if (message.ability == 18) // flying??


I have doubts this is correct

Is it important to fix this as part of this PR?

kennyvv · 2023-10-24T17:44:43Z

src/MiNET/MiNET/Utils/IO/Compression.cs

@@ -32,26 +32,46 @@

 namespace MiNET.Utils.IO
 {
+	//TODO - rework on NONE and ZLib instances


What is it you want to rework exactly?

Split into two implementations: without compression and with ZLib compression.
Since now before sending McpeNetworkSettings we should not use compression. And it's not the same as ZLib "NoCompression"

This reverts commit cfd80d7.

kennyvv · 2023-10-29T16:29:14Z

src/MiNET/MiNET/Net/RakNet/RakConnection.cs

@@ -673,7 +676,7 @@ internal async Task UpdateAsync(RakSession session)

 					long elapsedTime = datagram.Timer.ElapsedMilliseconds;
 					long datagramTimeout = rto * (datagram.TransmissionCount + session.ResendCount + 1);
-					datagramTimeout = Math.Min(datagramTimeout, 3000);
+					//datagramTimeout = Math.Min(datagramTimeout, 3000);


Hmm, i still feel like it'd be good to atleast have A timeout to avoid filling memory with crap. Maybe make it configurable using the config.

kennyvv · 2023-10-29T16:29:36Z

src/MiNET/MiNET/Net/RakNet/RakOfflineHandler.cs

@@ -84,7 +84,7 @@ internal void HandleOfflineRakMessage(ReadOnlyMemory<byte> receiveBytes, IPEndPo
 			// Shortcut to reply fast, and no parsing
 			if (messageType == DefaultMessageIdTypes.ID_OPEN_CONNECTION_REQUEST_1)
 			{
-				if (!_greyListManager.AcceptConnection(senderEndpoint.Address))


I'd still like to see this be put back as it was.

Still waiting for an explanation why this shouldn't be, since the old implementation doesn't allow whitelist clients by Endpoint and not Ip but at least for me, this feature is critically important. It doesn't affect performance or anything else in any way.

kennyvv · 2023-10-29T20:53:10Z

src/MiNET/MiNET/Utils/IO/HighPrecisionTimer.cs

@@ -68,7 +68,7 @@ public class HighPrecisionTimer : IDisposable
 		public long Avarage = 0;


-		public HighPrecisionTimer(int interval, Action<object> action, bool useSignaling = false, bool skipTicks = true)
+		public HighPrecisionTimer(int interval, Action<object> action, bool useSignaling = false, bool skipTicks = true, bool highPrecision = false)


Highprecision should default to true

Yes, I think you're right

but then the question is whether we need high precision in SkyLightCalculations.Calculate
https://github.com/NiclasOlofsson/MiNET/blob/df0d4ffb8c008654271e0c94187cada0abf6a30b/src/MiNET/MiNET/Worlds/SkyLightCalculations.cs#L182C4-L182C4

and may I set the highPrecision parameter from Config in the same way as it is done in Level for RakConnection SendTick timer? or does it need a different configuration name?

kennyvv · 2023-10-29T20:57:49Z

src/MiNET/MiNET/GreyListManager.cs

I would still like to see this reverted back.

kennyvv · 2023-10-29T21:00:57Z

Overall looking good at this point, just a few more things. Mainly the greylistmanager change from "ipaddress" to "ipendpoint" that one makes no sense to me. Could you answer my remaining questions?

… updated.

kennyvv and others added 30 commits July 23, 2022 00:03

Start updating to 1.19.10

55e57dc

Update packets

6fec3cf

Add support for SByte reading/writing

e219c3e

Update Ability handling

9d6ea5e

Fix build

4881e1a

Changes 1.19.20 without blockdata

ab60d3f

Changes 1.19.30

8057ec6

Changes 1.19.40

0592f40

Changes 1.19.50

8a02476

Changes 1.19.60 and 1.19.62

292e764

fix

aa0ef88

Fix

79db8ff

1.19.70

938256b

1.19.80

0ff7f07

Anvil world 1.19-1.20 support.

712ae20

Blocks update par1 and items update started.

67c38e1

Items update part 1.

6e96ee7

Items update part 2 + inventory transaction fix.

10e0a0a

Items update part 3 and some fixes.

3397ac0

Creative inventory update.

e1f4920

Crafting and Inventory transaction update.

ff35525

Items update pre final.

b3a1098

Smelting recipes added.

4c600ac

Partial blocks generator update.

82c0718

Factories update.

3a8a9e7

1.19 pre final update.

ac5017b

Biomes calculation from amvil noise.

de3ae46

Commands extension.

de39535

Disconnect message fix.

56d838e

Some fixes and optimizations.

ecd28cf

DarkLexFirst added 14 commits October 12, 2023 05:04

Anvil palette updated, chunks reading and sending optimized.

23a4e58

Block destroy particle update.

5324c89

Bow using times for shoot reduced.

11cc3d4

Forms fixed.

76de481

Chunks loading little optimization.

7a00697

Chunks on teleportation rendering fix.

daa44f4

Anvil mapping update.

dd8cfff

Tests fix.

da7a399

Chunks network serialization fix.

672d59c

Anvil map and palette update.

01d8e20

Minor chunks and PlayerLocation fixes.

23d4bb4

Players move broadcast fix.

f5b7ed1

Fix for fundamental packets order indexing.

a080fdb

Merge pull request #3 from CRPE-Team/1.19-update-fixes

07ade84

1.19 update fixes.

kennyvv added the enhancement label Oct 23, 2023

kennyvv requested changes Oct 23, 2023

View reviewed changes

Merge pull request #4 from CRPE-Team/master

f081655

1.19 update fixes.

kennyvv requested changes Oct 24, 2023

View reviewed changes

DarkLexFirst added 4 commits October 25, 2023 13:46

Review fixes.

32c9c02

Revert "Nuget publish test."

5f87c16

This reverts commit cfd80d7.

GreyListManager update.

e2c9cfc

Revert log levels.

0c7426f

kennyvv requested changes Oct 29, 2023

View reviewed changes

DarkLexFirst requested a review from kennyvv October 29, 2023 19:55

kennyvv requested changes Oct 29, 2023

View reviewed changes

EnableHighPrecision and DatagramSendTimeout configurations. 1.19 todo…

a2f6cd8

… updated.

DarkLexFirst requested a review from kennyvv November 5, 2023 01:18

Chunk sending for new location revert.

47cca18

1.19 update #638

Are you sure you want to change the base?

1.19 update #638

Conversation

DarkLexFirst commented Oct 23, 2023

kennyvv commented Oct 23, 2023

kennyvv left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

DarkLexFirst Nov 5, 2023 • edited

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

DarkLexFirst Oct 24, 2023 • edited

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

DarkLexFirst Oct 29, 2023 • edited

Choose a reason for hiding this comment

Choose a reason for hiding this comment

kennyvv commented Oct 29, 2023

DarkLexFirst Nov 5, 2023 •

edited

DarkLexFirst Oct 24, 2023 •

edited

DarkLexFirst Oct 29, 2023 •

edited