Hey guys,

So, I was looking over the source and there seems to be a large use of raw pointers which could be avoided through the use of smart pointers. These will define ownership for every object in the game while also avoiding manual deallocation and possible memory leaks or double frees.

For example, in Game.h:131 a map object is stored as a raw pointer and passed to objects. This could be converted to shared_ptr's or even better stored as a unique_ptr in the Game class and delegated as references. In this case, this would work quite well, as the objects that require it ensure that it is passed through the constructor, so it is allowed to be stored as a reference in the class instance.

If you're interested in making this switch, which I personally think will improve the projects architecture and remove memory issues later down the line, let me know. I'm more than willing to do the all initial work.

Thanks.

as of 9dea0d9 this happens: http://i.imgur.com/HsFWIGg.jpg
Minions from all lanes teleport and fight on the blue shrine. if you try to attack them you walk to their real location.

Sup guys.
I got interested in this project since it seems to be pretty active and would love to implement some champions within the next times.
but sadly it seems that the compile for vs 12 is broken (127 errors total)

used software:
win7
cmake 2.8.12.1
vs 2013 express

revision of the report: 02b3df3
(updated issue)

In all the places I had traveled I continued to be able to see minions as soon as they entered those areas.
Version tested on: 4a892ba

When targetting something new, minions teleport to a location. Needs a fix. This is happening probably because of the rework.

All tabs are spaces, why don't we change them using search and replace? it shouldn't affect linux/mingw compilation and would be better for people using visual studio, we all win.

Loaded 417 RAF files
Games startedBefore mapterminate called after throwing an instance of 'sol::error'
what<>: lua: error: cannot open ../../lua/config.lua: Too many open files

I'm getting a lot of trouble by the fact that IntWars can't detect the LoL 4.20 backup that I have, and because of that I can't test neither help the project.

Is this project still alive? It is planned to get support to the newer versions of League of Legends?

I used .ch and .model to change my model and then later my champion and my friend told me that he did not see any change. It used to work in the past for everyone on the server.

Haven't found any instructions yet, how is this done on say a ubuntu x64 installation?

Pretty sure it should be X,Y,Z if IntWars uses the DirectX axis system (pretty sure it does). In L# the y and z axis are swapped.

After vigorous testing I have discovered there is nothing wrong with the placement of the shopkeepers and yonkeys on the server, the client just refuses to display them in their place due to the fact that they are out of bounds. This may be because they inherit from the player object, and as such the client refuses to place them outside the playable area.

Item stats don't carry over, but stats from commands carry over client side and not server side.

When I was implementing a change model system, I noticed that the RAFManager does not see a lot of inibin files. The ones I was trying to load (Nidalee's) were located at DATA/Characters/Nidalee/Spells, but the current loading system only seems to find spells in DATA/Spells.

Commit 75d23ae breaks client connection on all versions, tested with versions 4.12, 4.13, 4.16 and 4.14 Garena.

It's probably my fault somehow, but I get nothing but:

WARNING 22:03:47 D:\IntWars-master\gamed\src\PacketHandler.cpp handlePacket:208 Unhandled OpCode 00
INFO 22:03:47 D:\IntWars-master\gamed\src\PacketHandler.cpp printPacket:105 Printing packet with size 6 0000-0005 00 00 00 00 00 00 ...... \x00\x00\x00\x00\x00\x00

I haven't seen anything else related to this, suggestions?

@Querijn There's an issue with this wall here.

Action:

Running .ch command with no arguments.

Expected:

".ch" will show message in chat of usage.

Actual:

Running the .ch command alone causes error
ERR : couldn't find champion stats for
Champion Levelup to 1
Then causes client crash.

The firing destination vector of Ezreal's abilities doesn't seem to match the indicator path.

Champion will change but the entire game gets bugged after than, and the console is filled with 'received nilllua: error: expected function.'

I get this error when im opening CompileServer.bat :

No spell can deal damage...
Can someone fix it? 👍

Despite the changes in 2273bfd trying to start the server results in the error:
"what(): lua: error: cannot open ../../lua/config.lua: Too many open files"
I built 1fe70da on Windows 7 x64 with MinGW

After minions clear an enemy minion wave near a tower or immediately after destroying a tower, they 'freeze' for a second then teleport forward along their path to the next target (the tower or the next minion wave)

If the value of the .model command does not exist the server still asks the client to load it, breaking both the server and the client.

I compiled & everything but it's not working since league updated to 4.18, It just bugsplats when it fully loads.

Just a heads up

The client disconnects from the server (attempting to reconnect message) after ~35 seconds have passed. The Server does not crash and no message prints.

Honestly, a lot of the code being developed needs to take vision into account, but there's really no support for this. For example, when minions spawn, I shouldn't be notified about the enemy minions unless I have vision, and yet there's simply a broadcastPacket call for it.

Turrets don't generate vision, and I'm guessing the only reason anything looks like it's generating vision is because of the client. This will become a problem when trying to work with things that stop working when you lose vision, for a basic situation, as well as how a minion can interact with a player in a bush. Right now they're heat seeking and will know you're there because you're the closest thing to them. I consider vision to be one of the top priorities right now, instead of improving the scripting hooks to continue to do things wrong. (no offense meant, they're right as far as the code can be right)

Purple team players are properly placed in the right team, but they spawn at the blue team spawn.

EIREXE forgot to redo packets.h when he fixed the indentation in his pull request, currently the MapID is still hardcoded in packets.h (line 116).

No worries, i already did the hard work: the outgoing enet header.peerId (the first byte of every packet) is wrong.
Packet dump:

Client -> Server
A7 FF 00 15 82 FF 00 01 00 4B 05 78 00 00 80 00 00 00 00 08 00 00 00 00 00 00 00 00 00 00 13 88 00 00 00 02 00 00 00 02 A7 C6 8F 04

Yorick -> Client
29 80 25 6D 83 FF 00 01 00 7F 05 78 00 00 80 00 00 00 00 08 00 00 00 00 00 00 00 00 00 00 13 88 00 00 00 02 00 00 00 02

Riot -> Client
A7 80 07 60 83 FF 00 01 00 41 03 E4 00 00 80 00 00 00 00 08 00 00 00 00 00 00 00 00 00 00 13 88 00 00 00 02 00 00 00 02

It gets set here:

header.peerID = currentPeer->outgoingPeerID + 0x29;

If i change it to A7 there it works

The auto-attacks seem to be destroyed incorrectly, resulting in breaks in GDB when minions attack, it points to functionality inside the update function of CollisionHandler.

Objects get removed from the CollisionHandler when they're being deleted, how are they still in the CollisionHandler? Is delete not working? Are autoattacks not being removed correctly?

Hey,
it seems that all lua script doesn't work. I tried to use cait E and nothing happened, the animation works but it doesn't teleport.
Anyway, thank you for your work on it !! I would like to help but I don't know where to start ..

Hi, as of the last patch the structure of multiple send (from client to server) and receive packets have changed.

Send:
Cast,
ChargedCast,
Move,
Ping
(all have extra byte added)

Recv:
Damage new struct:

There may be others.

Btw if you want more information about the packet(s) 0xFE then let me know.

Here are some other headers you might have missed:
PKT_C2S_HeartBeat = 0x08,
PKT_C2S_RemoveItem = 0x09,
PKT_C2S_InteractObject = 0x3A, // thresh lantern and dominion towers capping
PKT_C2S_GameFinished = 0x8C,
PKT_C2S_Disconnect = 0x9B,
PKT_C2S_ChargedCast = 0xE6,
PKT_C2S_Refund = 0xFE, //AKA 0x110

a1a816b seems to have broken GCC build on Win64. Build fails with the error:
F:/Intwars/gamed/include/Buff.h:22:17: error: 'unint64' has not been declared
void update(uint64 diff){

Ever since I updated to the 4.14 client with Riot's new sound engine, sounds don't play when I try to start the client using IntWars as the server. I checked the game logs and the client is attempting to use FMOD when it's supposed to use Wwise. This issue doesn't happen when I play on Riot's actual servers. Does anyone else experience this issue?

As of latest commit from dotdeath (bd109b1), vision is not working, you can't see people from the other team, pic here: