Noticed a bug where a custom vec2 action is not synced if x or y is 0

the vector2 is set on the client but either not sent to the server or the server is decoding it incorrectly.

Steps to reproduce:

• Set a custom Vector2 action
• Change the x value of the Vector2 every physics step but keep y at 0
• Input will be out of sync

The repository currently doesn't seem to contain any license. Could you add one, please? See Choose a License for more information 🙂

For reference, most open source Godot add-ons use the MIT license.

How do you assign the inputs to the player nodes on the server side?

Your example is a bit too complex to understand where you actually bind these things together.

Also: It would be nice if you would share a demo where the server does not participate actively in the game (has no player that he controls) since in server authoritative games the server usually is just a spectator.

Allow clients to specify the rate in which snapshots will be sent, which could be at a lower pace than that of the simulation itself.

I'm new to godot, so of course I do the sensible thing and jump right into making a persistent network game!

I'm making a sci-fi ship-crew game where players are person-like avatars (Kinematic) inside a ship (Ridgedbody). The server is always running so players and ships will be spawning/despawning as they log in and out. There will be an arbitrary number of ships and players.

Ideally I would like the player avatar to be a child node of the ship, however the register_spawner function seems to requires a hard-coded node to spawn in.

If I have a PlayerCharacter spawn node, will the network addon allow me to reparent without the system breaking? Or is there a better way for me to go about this?

While I tried to follow some of your Inventory addon tutorials I stumbled upon some issues, I managed to rewrite the "_generate_item()" function as "_get_item(id, category)" so that its being used to load valid (I think?) item entries that gets added to a dictionary variable, now when I use this "_get_item(id, category)" function to add a item to the InventoryBag node I get this error: Invalid call. Nonexistant function 'disabled_slots_occupied' in base 'Nil'.

When I use the "add items" code example and try to add the item that way I get the same error.

Debugger output:

0 - res://addons/keh_ui/inventory/bag.gd:948 - at function: _get_colliding_data
1 - res://addons/keh_ui/inventory/bag.gd:1029 - at function: _find_fitting_spot
2 - res://addons/keh_ui/inventory/bag.gd:164 - at function: add_item 
3 - res://System/Inventory/Init.gd:19 - at function: add_items
4 - res://System/Inventory/Init.gd:14 - at function: ready 

Node2D, InvDemo.gd script: https://ghostbin.com/paste/km8Ek

Node, Init.gd script: https://ghostbin.com/paste/8yRFk

This is more of a question and possibly feature request:

In FPS type games it's common that the Server goes back in time (for a maximum of N ticks) to check if someone's bullet hit another player (if a client says he hit someone but the server did not see it on his tick, due to network delay).

Most FPS games favor the shooter (which means you can run around the corner and get hit like a second afterwards because the player shooting you had a latency of ~500ms), while others favor the one running away.

Is it currently possible to implement something like this with this library?
I know that Godot does not allow rewinding a scene and doing so would cause a lot of other problems (all clients would rewind their states too during that tick)

However it'd be already "good enough" to get history data for that action:

• Where did the shooter aim at ? (at the tick we acknowledged the shot)
• Where was the player that got shot? (at the (tick + travel time of the bullet))

If i was able to know this i could reconstruct an invisible bullet + hitbox of the player and use smth like raycasting to check it.

Edit: Assuming i have Aim and position of all players stored in previous state snapshots(!)

Question:

Is it possible to smooth the rotation (and just the rotation) of a Sprite with the Smooth Addon?
As far as i can see it is only possible to sync the whole Transform.

We could extend the Plugin to be able to selectively update transform properties (position, rotation, scale) and select the Node that is referenced (instead of taking the parent).
This would allow to create invisible entities that are changed and bind the visuals (which might have different anchor points).

For example:
A Sprite with an Anchor point (Changeable with V Key) outside the sprite boundaries
A Node2D (lets call it SpriteRotation) which represents the data that we want to bind
Binding rotation of Node2D to the Sprite would result in the sprite being rotated correctly

Right now making the Smooth2D a child of SpriteRotation and the Sprite a child of Smooth2D
So: SpriteRotation -> Smooth2D -> Sprite
Would make the Sprite rotate different than what one would want to accomplish

Suggestion:
Smooth2D Configuration (in the Editor)
From Node: Any Node2D (currently the parent)
To Node: Any Node2D (currently the child of Smooth2D)
Properties: Array<(PropertyName, PropertyType)>

So we could configure it to take any Node2D as a reference and apply the Properties to any Node (given that both have the property and the type is the same on both)

PR possible? I could give it a try

Edit: I can see that you decided against linking the nodes manually. Maybe they could just be initially selected as a trade off?
If you don't want this to be part of the Addon to keep it simple, i'd create a fork of it

Provide means to selectively eliminate certain objects from the encoded snapshot, which could save some bandwidth.

Hey me again,

could we set up on the projects tab a feature roadmap or wishlist?
I'd like to contribute to this project because i want to use it on my own game.

On http://kehomsforge.com/tutorials/multi/GodotAddonPack/part02/ the extra signal table isn't formatted as a table.

Is there any reason why EncDecBuffer does not support strings by default? I have come up with this:

static func write_string_utf8(encdecbuffer: EncDecBuffer, string: String) -> void:
	var bytes := string.to_utf8()
	encdecbuffer.write_uint(bytes.size())
	for i in bytes.size():
		encdecbuffer.write_byte(bytes[i])


static func read_string_utf8(encdecbuffer: EncDecBuffer) -> String:
	var bytes := PoolByteArray()
	var size := encdecbuffer.read_uint()
	for i in size:
		bytes.append(encdecbuffer.read_byte())
	return bytes.get_string_from_utf8()

(I am not using strings in snapshots but in custom player_data properties to optimize and send things like player name, skin etc. to be send in one packet)

There's line OverlayDebugInfo.set_label("dbg1", "Received delta snapshot") in network.gd line 768. I think it should not be there.

It would be great if one could simply swap data layer - let's say you want to use steam networking API layer for sending and receiving packets.
Currently everything is based on rpc calls but it would be good to use some abstract layer between sending and receiving data in network.

For example there's readP2PPacket and sendP2PPacket in steam library:
https://gramps.github.io/GodotSteam/functions-module.html

At one point I will need this feature so it's possible I will take look on this future and create PR.

It seems your Inventory addon is missing facilities to check if a item got swapped by another item, I need this form of check since with my current implementation of the shop menu that disables certain bags so that the player cannot get free items gets bypassed when the player is dragging a item and swapping it with another item. Trying to use the signal "item_clicked" on my end only registers right clicks and not left clicks so I cannot really use that one for saving item datacode and making sure the bags stays disabled on the shop menu.

Hey,

is it possible to get a just_pressed function for input events?
For example in a 2d platformer game there's no need to continously trigger jump events even if the player keeps holding the jump key down.

thx for the addon btw, looks great so far!

Or: croxis needs to commit more often so breaks are easier to revert!

I did... something which is causing errors to be thrown and I am not sure where to start looking to fix it. I used the dedicated server tutorial as the template.

E 0:00:32.735 get_node: Node not found: @@8. <C++ Error> Condition "!node" is true. Returned: __null <C++ Source> scene/main/node.cpp:1381 @ get_node()

E 0:00:32.735 _process_get_node: Failed to get path from RPC: @@8. <C++ Source> core/io/multiplayer_api.cpp:256 @ _process_get_node()

E 0:00:32.735 _process_packet: Invalid packet received. Requested node was not found. <C++ Error> Condition "node == __null" is true. <C++ Source> core/io/multiplayer_api.cpp:204 @ _process_packet()

When I spinup the server node @@8 does exist with the player_1 node under it. On the client side player_1 is under node @@14

Hi, i just recently discovered your addon pack while looking for an grid-based inventory system. and it matches all i need!, thanks you very much!.

So, for trying to implement an sort of S.T.A.L.K.E.R-like inventory system where the grid is infinite as long the weight capacity limit is not reached, i cannot find an way to grow the grid as needed (like grow the grid after 90% of the grid was used in items), tried using scrolling bars and some hackery but with no luck. do you have an better idea or code how could be achieved?

As well would be nice of selecting the item by dragging them until drop to an special slot and not just clicking it once and then clicking it again to putting in an special slot.

Thanks.

I have written a crude script for a manual drag & drop operation and I managed so far to access the currently held Item of the cursor but right now I'm stuck with my implementation. The reason I want to do the manual drag & drop operation is that I want to implement a S.T.A.L.K.E.R like shop system which your addon almost provides everything for it, the shop system in the before mentioned game has 4 elements where items appear under certain conditions.

With this call I can get the "drag" operation going:

{InventoryNode}.pick_item_from(Column, Row, -1) 
ItemDragged = InvEvent

Then with the following call I get a stack overflow (1024) error which it seems the add_item() part gets stuck in a loop and gets called multiple times yet when I do a basic print() call then the print() function only gets called once per mouse click which is odd.

	match (event.get_class()):
		"InputEventMouseButton":
			#Drop the item.
			if Input.is_action_just_pressed("mouse_left") && ItemDragged:

				if Bag._shared_data.is_dragging():
#					print(ItemDragged.get("item_data"))
#					var NewItem = ItemDragged.get("item_data")
#					emit_signal("Self_AddItem", NewItem)
#					print("True:", Bag)
					pass

I did read your blog documentation page but I couldn't see a signal function or anything similar which lets me retrieve and manipulate the mouse during the "dragging" part so that it behaves almost same as if the Inventory mouse setting is not in "none" mode. I apologize in advance if I have missed something.

InventoryManager_DragnDrop.txt

player_added signal works fine and it's called on all clients but player_removed is called only on server.

Godot 4 changes things. This issue is for a Godot 4 compatible pack!

github has a newish feature called discussions, basically an embedded forum. I feel a bit awkward opening an issue when I mostly have questions or need clarification on something.

After the latest update (i believe) i noticed a strange bug on my demo project:
If a player disconnects, the despawning on the server works perfectly, however on clients the node is despawned, and on the next tick respawned (and it stays there forever).

Is this the only way despawning players on the clients should be done:
func on_player_left(pid: int) -> void: network.snapshot_data.despawn_node(NetPlayer, pid)

NetPlayer is my SnapEntity Object and i'm directly assigning the pid of the player to the node (players only have one controllable character)

is it possible to send the tickrate of the server without having to change the physics fps ?
Because i'd like to still run on 60fps on server but sync with clients on a different ticks/second

In case it is relevant: Because I am exporting to HTML5 for the client I replaced enet with websocket (would love webrtc instead but I have yet to figure out how to get that bundle of joy implemented, but I digress.)

Client calls network.join_server() at the end of the ready function of the initial scene.
Client listens for "join_accepted" before calling network.notify_ready()
But no snapshots propagate to client, regular godot rpc calls do work however.

Relevant code snips from gameworld.gd

func _on_server_join() -> void:
    if (!network.has_authority()):
        debug("Notifying server that I am ready")
        network.notify_ready()

# Called when the node enters the scene tree for the first time.
func _ready():
    var is_client: bool = false
    
    if OS.get_name() == 'HTML5':
        is_client = true
    if Array(OS.get_cmdline_args()).has("-c"):
        is_client = true
    
    if is_client:
        print("Debug Gameworld: Is_client")
        _setup_network_input()
        network.connect("join_accepted", self, "_on_server_join")
        network.join_server('127.0.0.1', gamestate.server_info.port)
        var lobbyUI = load("res://lobby.tscn")
        var lobby_node = lobbyUI.instance()
        $HUD.add_child(lobby_node)
    else:
        print("Debug Gameworld: Is_server")
        network.create_server(gamestate.server_info.port, gamestate.server_info.name, gamestate.server_info.max_players)
        $HUD/PanelServerInfo.hide()    
    
    # Spawn the ship
    if (!get_tree().is_network_server()):
        #rpc_id(1, "spawn_player_ship", gamestate.ship_name)
        print('Debug Gameworld: I am a client attempting to call server spawning player info')

In network.gd I threw in some debug print statements. The remote_player dict is empty so nothing is set to ready.

# Server will only send snapshot data to a client when this function is called
# by that peer.
remote func server_client_is_ready() -> void:
    assert(has_authority())
    # Retrieve the unique ID of the player that called this
    var pid: int = get_tree().get_rpc_sender_id()
    var player: NetPlayerNode = player_data.remote_player.get(pid)
    print("Debug Network: server_client_is_ready " + str(pid) + " " + str(player_data.remote_player) + " " + str(player))
    if (player):       
        player.set_ready(true)
        print("player ready")

Ok, so when you create a table you have 2 options: save it as "embedded" or save it as a resource in your db's folder
so i used the latter and created a table. then i removed the table i created with the little minus icon here:

and when i added the same table back (by dragging it from the filesystem) it showed up in the editor but when i ran the code and printed the list of tables it wasnt there.
so basically the bug is that if you save a table as a resource it stops working if you remove it and add it back again
the embedded ones work fine though so there's a workaround
(btw i really appreciate your addons ty 👉 👈 )

I am working on moveable platform in 2D which has constant velocity and uses StaticBody. The issue is that corrections keep incrementing even though I applied new corrected position. I am running this code on both server and client.

The logic is following:

func _ready() -> void:
 uid = name.hash()
 network.snapshot_data.add_pre_spawned_node(MovingPlatformSnapshot, uid, self)
 # load platform points


func _physics_process(delta: float) -> void:
 ####
 if correction_needed:
  position = correction.position
  next_stop_index = correction.next_stop_index
  ...
  for _i in network.snapshot_data.get_prediction_count(uid, MovingPlatformSnapshot):
   update_platform(delta)
 #### 
update_platform(delta)
network.snapshot_entity(MovingPlatformSnapshot.new(uid, 0).from_node(self))


func update_platform(delta: float) -> void:
    # velocity is constant calculated from next_stop_index which is next point where should platform stop but it is same on server and client
    position += velocity * delta

MovingPlatformSnapshot contains position and next_stop_index

The position in corrected always differ in very similar numbers like 3 pixels difference. Also I have observed same issue with your demo project - glowprojectile.gd it keeps correcting. Maybe there's something missing in the addon like we needed for player entities (correct_in_snapshot) ? I don't even use anything advanced with physics it's just simple position and once it is corrected it should be sync with server.

Could you please add option to enable/disable debug logs from network addon (Registering network input logs etc.)?

In your mega demo i noticed that jumping (and predicting it) works fine as long as no other button is pressed.

However if you jump while moving with 2 keys (lets say W and D) the prediction seems to cause jittering.
In my own game prototype i can see that replaying is triggered many times when the server has to process many inputs while it is only called 1-2 times when i only press jump (and nothing else)
Could this be related to move_and_slide?

You can see it in the attached video:
Addons (DEBUG) 09.07.2020 14_12_59.mp4.zip

Hello everyone.

I'm opening this issue to let you know that I'm aware of a problem with resource leaks, more specifically with the Inventory, Network and DebugHelper plugins.

Specifically for the Network and the DebugHelpers the error message will appear only if at least one of those plugins is enabled. I have tested forcefully freeing the two automatically created singletons at the plugin's _exit_tree() function and the error message went away. I will further investigate to identify if this is a problem with the code or a bug in the EditorPlugin.add_autoload_singleton() function.

As for the Inventory, the problem is a bit more complicated. On my preliminary tests it seems the root cause of the problem is in any script class that static types "itself", something like this:

extends whatever_that_is_a_reference
class_name some_class

func some_func(some_arg: some_class)....

Just having a script with code like this will result in resource leak error message.

Based on this information, I'm not sure about the severity of the problem on the Inventory plugin, however it's probably harmless on the Network and the DebugHelpers plugins, as the problem seems to be that the singletons are not freed in time during the exit process.

Is it possible to use arrays with SnapEntityBase? I would like to use PoolByteArray to synchronize array of bytes.

Hello,

Say you have an action registered:

network.register_action("attack", false)

in the physics_process() of the player script, the code input = network.get_input(_uid) is only executed on the local machine and on the server:

network.gd:

func get_input(player_id: int) -> InputData:
	var is_authority: bool = has_authority()

	if (!is_id_local(player_id) && !is_authority):
		return null
	
	var pnode: NetPlayerNode = player_data.get_pnode(player_id)
	var retval: InputData = pnode.get_input(_update_control.sig) if pnode else null
	
	return retval

meaning a remote player scene will not be able to check and execute the input.is_pressed("attack") to play, for example, an attack animation.

For now, the attack animation is played if the player is local and on the machine that has authority, but a client that doesn't have authority will not be able to execute actions of other players.

How would you go about synchronizing the actions of all players on all machines?
Same question goes for an NPC entity.

Would the solution be to put new booleans in the snapshot script extending SnapEntityBase, and set them as action = true when receiving a new action from the snapshot ?

Thanks for the addons and for your help.

Is it possible to retrieve custom property for all players? Does server sync custom properties of all players? Because I can't get this working. I have something like this:

network.player_data.add_custom_property("character_skin", "default", NetCustomProperty.ReplicationMode.ServerBroadcast)

in game menu I am setting the property type as string:

network.player_data.local_player.set_custom_property("character_skin", skin)

in player script I have this:

func _physics_process(delta: float) -> void:
if corrected_state != null:
 if !initialized:
  var skin = network.player_data.get_pnode(meta_uid).get_custom_property("character_skin")
  # set skin
  initialized = true

I can set skin as another variable for syncing in snapshot class but I don't want to do this because it doesn't make sense to sync skin everytime but only once.

1. add a column of type "color" to the database, godot has that built in very nicely
2. add an option to sort rows by a column (for example if i have a table called "hats" with a column that stores the hats' color and i want to see all the red hats or all the hats of type "rare" etc)
3. maybe add an option for storing long strings? right now it just stores them as one line, but if i want to store a paragraph for example its hard to read it as one line
   (wasn't sure if this is the right place for posting this so correct me if im wrong)

No check is done if WebSockets module has been installed.

SCRIPT ERROR: GDScript::reload: Parse Error: The identifier "WebSocketServer" isn't a valid type (not a script or class), or couldn't be found on base "self". At: res://addons/keh_network/network.gd:403. ERROR: reload: Method failed. Returning: ERR_PARSE_ERROR At: modules/gdscript/gdscript.cpp:599. ERROR: start: Script does not inherit a Node: res://addons/keh_network/network.gd At: main/main.cpp:1763.

I have module_websocket_enabled = "no" in my godot build.

The only way I can think to fix would be checking if ClassDB.can_instance() the WebSocketServer prior to attempting to ref it.

Not a major issue if no clean fix can be found.

Can the network plugin add a solution for serialization and deserialization using protobuf, currently it's directly using Godot's own PoolByteArray right? Maybe protobuf is better for passing large amounts of data.

and also could be helpful to show number of corrections and time of last correction for specific objects