DDraceNetwork - developer

Hey everyone !

[ddnet/ddnet] Pull request closed: #8229 Add optional name argument to start_video command, improve descriptions of demo-related console commands, refactoring

00:58

github-merge-queue[bot]

[ddnet:master] 3 new commits

93d5a8b Improve descriptions of demo-related console commands - Robyt3 d2c8132 Add optional name argument to start_video command, refactoring - Robyt3 e5e48a8 Merge pull request #8229 from Robyt3/Client-StartVideo-Refactoring - def-

01:03

def-

[ddnet/ddnet] Pull request opened: #8232 Translations for upcoming 18.2

Checklist

[ ] Tested the change ingame
[ ] Provided screenshots if it is a visual change
[ ] Tested in combination with possibly related configuration options
[ ] Written a unit test (especially base/) or added coverage to integration test
[ ] Considered possible null pointers and out of bounds array indexing
[ ] Changed no physics that affect existing maps
[ ] Tested the change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-ad...

01:06

def-

[ddnet:18.2-rc1] 2 new commits

0494878 Translations for upcoming 18.2 - def- c66886b Version 18.2 - def-

01:08

def-

[ddnet/ddnet] Pull request opened: #8233 DDNet 18.2

DDNet 18.2 is scheduled to release in 12 days, assuming no bad bugs are found. Please test the Release Candidate to prevent problems being only discovered after release. Report bugs in the #bugs channel on DDNet Discord or directly on Github:

Windows 32bit: https://ddnet.tw/downloads/DDNet-18.2-rc1-win32.zip
Windows 64bit: https://ddnet.tw/downloads/DDNet-18.2-rc1-win64.zip
Linux x86: https://ddnet.tw/downloads/DDNet-18.2-rc1-linux_x86.tar.xz
Linux x86-64: https://ddnet.tw/downl...

01:12

github-merge-queue[bot]

[ddnet:master] 3 new commits

8f7055f Change PNG in function names to Png - Robyt3 a743962 Pass CImageInfo by reference to IGraphics::LoadPng - Robyt3 173e675 Merge pull request #8203 from Robyt3/Graphics-LoadPng-Cleanup - def-

01:12

def-

[ddnet/ddnet] Pull request closed: #8203 Change PNG in function names to Png, pass CImageInfo by reference to IGraphics::LoadPng

yea

def-

[ddnet-web:master] 1 new commit

5f8ca9f New RUS server - def-

ChillerDragon

[ddnet/ddnet] Pull request opened: #8234 Fix reckoning typo in comment

def-

[ddnet-web:master] 1 new commit

18ff1b3 Update translation stats - def-

def-

[ddnet/ddnet] Pull request closed: #8234 Fix reckoning typo in comment

02:36

github-merge-queue[bot]

[ddnet:master] 2 new commits

db62797 Fix reckoning typo in comment - ChillerDragon b355032 Merge pull request #8234 from ChillerDragon/pr_typo_reckoning - def-

<ChillerDragon> endianess is such a pain but i managed to understand ddnet snap uuids :D ... i know im late to the party

02:57

<ChillerDragon> so teeworlds uses types starting from 1 to max teeworlds snap items and ddnet uses types starting from max snap type (32767) going down to 32703

02:58

<ChillerDragon> the teeworlds types are known at compile time and can be looked up in a table. The ddnet ex items are not known at compile time and change even while a tee is connected

02:59

<ChillerDragon> There can only be 64 unique ddnet ex items in a single snapshot but there can be as many ddnet ex items during one play session as there are uuids

03:00

<ChillerDragon> So 128 player support that snaps 128 ddnet ex characters in one snap would be impossible?

04:54

i own rtx 3060 now

04:54

i have 4 more gigs of vram

github-merge-queue[bot]

[ddnet:master] 2 new commits

0494878 Translations for upcoming 18.2 - def- 4b3035d Merge pull request #8232 from def-/pr-18.2-transl - Robyt3

09:00

Robyt3

[ddnet/ddnet] Pull request closed: #8232 Translations for upcoming 18.2

Ewan

Click to see attachment 🖼️

wtf dude

09:06

2 GPUs 64GB RAM, did u get a paycheck or what did i miss xD

one is integrated gpu

09:06

misleading

ah yeah it even says so

i have had multi gpu setups before

09:07

it's great for VMs

nice PC, bad OS xd

yea

Ewan

i have had multi gpu setups before

i could even rn but i would have to sacrifice my third nvme drive

09:32

and that’s a part of a non redundant raid array with important os stuff on it (edited)

09:32

so i would needa linux

09:33

https://youtu.be/AmW2v4XWV-0?si=NDW5koIfj4XSgQug

LOFTmusic

Chick Corea Akoustic Band - On Green Dolphin Street (1991)

furo321

[ddnet/ddnet] Pull request opened: #8235 Update Swedish translations for 18.2

Checklist

[ ] Tested the change ingame
[ ] Provided screenshots if it is a visual change
[ ] Tested in combination with possibly related configuration options
[ ] Written a unit test (especially base/) or added coverage to integration test
[ ] Considered possible null pointers and out of bounds array indexing
[ ] Changed no physics that affect existing maps
[ ] Tested the change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-ad...

ChillerDragon

[ddnet/ddnet] Pull request opened: #8236 Add delta debug dump

There is already a DebugDump() method for snapshots but there was none for deltas yet. I wrote this to debug an issue I was facing that only happend after the delta was unpacked so I wanted to see what was in the delta. The method is unused and supposed to be patched in while debugging. For example like this ```diff diff --git a/src/engine/client/client.cpp b/src/engine/client/client.cpp index 79a19a65b..ce0be00b8 100644 --- a/src/engine/client/client.cpp +++ b/src/engine/client/cli...

chillerdragon: do you have stats for delta sizes?

10:51

i'm interested in the size of deltas after network compression

<ChillerDragon> LEEERATOOOOOOOOO

11:21

<ChillerDragon> @Jupstar ✪ you mean the size thats sent over the network and how big it is unpacked?

yes and no i want the packed size i want the size what is actually sent

11:22

but only from snapshot deltas, not other traffic

<ChillerDragon> i guess

11:23

<ChillerDragon> lemme check

and ofc it would be nice to know how many characters were in the game at that moment

<ChillerDragon> just me :D

perfect

<ChillerDragon> but ye i would need too much time to properly look it up i think

<ChillerDragon> im in the middle of a debugging sessions

yeah np

11:26

if u some day can just look up the packet size if u stand still and do nothing

11:26

that would already help

11:27

but i will otherwise just do it

<ChillerDragon> my first guess would be 83 bytes get unpacked to 88 bytes :D

i am just bad and finding it in cpp code xd

<ChillerDragon> but thats 99% totally off

11:28

<ChillerDragon> https://zillyhuhn.com/cs/.1713612473.png

11:28

<ChillerDragon> 83 bytes are sent over the network

that's a vanilla server?

11:28

or ddnet server

<ChillerDragon> for a fresh 0.7 connection

11:28

<ChillerDragon> no thats ddnet with ddnet ex items too

ok

<ChillerDragon> https://paste.zillyhuhn.com/f7

11:29

<ChillerDragon> on connect it updates the character items

11:29

<ChillerDragon> duuudde i am THIS close to fixing the last bug

11:30

<ChillerDragon> where is lerato!?

def-

[ddnet/ddnet] Pull request closed: #8235 Update Swedish translations for 18.2

11:30

github-merge-queue[bot]

[ddnet:master] 2 new commits

06d6e3c Update Swedish translations for 18.2 - furo321 a00d6a3 Merge pull request #8235 from furo321/swedish-18.2 - def-

<ChillerDragon> implicit cast of bool sixup to size_t size !?

11:31

<ChillerDragon> lerato watafak were you thinking

11:31

<ChillerDragon> https://zillyhuhn.com/cs/.1713612625.png

11:31

<ChillerDragon> https://zillyhuhn.com/cs/.1713612681.png

11:32

<ChillerDragon> @Learath2 i feel like this code could use a comment :D

chillerdragon: ah interesting

11:32

is that a dumb of single character?

<ChillerDragon> thats me fresh after connect ye

i'm suprised it is so small

<ChillerDragon> it creates a delta for both vanilla char and ddnet char

11:33

<ChillerDragon> 83 bytes?

well my diff results in 56 bytes, after compression

11:33

but the initial diff size is insanely much bigger

11:33

https://pastebin.com/euRNUVyJ

value: [ 98, 2, 0, 0, -...

Pastebin.com is the number one paste tool since 2002. Pastebin is a website where you can store text online for a set period of time.

<ChillerDragon> pls dont benchmark my numbers

11:34

<ChillerDragon> i told u i didnt do measurements just a random screenshot xd

11:34

<ChillerDragon> idk what that number means

11:34

<ChillerDragon> its probably mostly ip udp headers

the biggest difference in my approach is, that it is a binary diff so it's not aware of the snapshot content

<ChillerDragon> the snapshot data is 44 bytes

that surprises me too

11:35

ah, but 0.7 removes player info from snapshot right?

<ChillerDragon> ye pls do ur own benchmark :D

11:35

<ChillerDragon> ye it does

nah i totally trust your benchmarks

<ChillerDragon> omg i should have said nothing

def-

[ddnet-web:pr-18.2] 1 new commit

b9ebdae Version 18.2 - def-

11:36

def-

[ddnet/ddnet-web] New branch created: pr-18.2

<ChillerDragon> instead of gaslighting you

def-

[ddnet/ddnet-web] Pull request opened: #292 Version 18.2

it's ok chiller. why should your dumb tool not show the truth xd

<ChillerDragon> its heinrichs dissector

see

<ChillerDragon> its OG tool but idk how to use it

11:38

<ChillerDragon> dude lerato

11:39

<ChillerDragon> the item size of sound world is 0 for 0.6 and 1 for 0.7?!

11:39

<ChillerDragon> i cant

wtf is a sound world

<ChillerDragon> ok watever im pretty sure whatever this satanistic code is doing i need it too because fakin sound world was breaking my demo deltas

11:39

<ChillerDragon> https://chillerdragon.github.io/teeworlds-protocol/07/snap_items.html#event_sound_world

ah lol

11:40

ah right

11:40

in teeworlds events are part of the snapshot

11:40

i don't understand that decision tbh

11:40

aren't events a one time thing

<ChillerDragon> ye ye you convinced me on that one

ah did we even talk about that

<ChillerDragon> yes

well defs was fun

<ChillerDragon> i liked your point of some events being important

we should repeat that

<ChillerDragon> and snapshots might be lost including all their events and they will never be resend

11:42

<ChillerDragon> ye we should

but i'd like to have heinrich in too next time

11:42

but he afk ignores me

11:42

<ChillerDragon> ye that be fun

jupstar vs heinrich the best buddies ever

<ChillerDragon> epic

but i also want to talk about other things anyway

<ChillerDragon> ok i gotta get back to debugging before bed time

11:43

<ChillerDragon> bye

good night

ws-client

<ChillerDragon> @Learath2 i feel like this code could use a comment :D

It may also be a copy paste mistake somewhere along the way. It's been so long now I don't remember the specifics

Jupstar ✪

aren't events a one time thing

It's a workaround for the fact that the snapshots aren't synced to netmsgs in any way

i dont understand anything xd netmsg = simply a new packet?

<ChillerDragon> https://zillyhuhn.com/cs/.1713617218.png

12:47

<ChillerDragon> that was a weird one after switching branch back and forth i can not reproduce anymore

12:47

<ChillerDragon> i think ccache starts to troll me

12:48

<ChillerDragon> i am surprised that this is the first weird thing i see given that for quite some time ccache was hiding compile errors from me for quite some time

12:48

<ChillerDragon> cant fail to build a file if its loaded from cache lmao

12:49

<ChillerDragon> @Jupstar ✪ if you ask me leratos message does not make much sense xd

CHILLER, wazzup, u awake already? or does the debugging session haunt your sleep

<ChillerDragon> i just rq

12:50

<ChillerDragon> i fixed the thing which i thought was the last bug and finish all features of ddnet7 with leratos weirdest code ever

12:50

<ChillerDragon> only to realize that demos do not work when the dummy is connected and that seems to be a nasty new quest

ah good thing this will never happen with dd-pg

12:51

bcs dummies dont really exist

12:51

i mean they exist, but not as a new network entity

<ChillerDragon> https://github.com/ddnet/ddnet/blob/master/src/engine/client/client.cpp#L2300

ddnet/src/engine/client/client.cpp at master · ddnet/ddnet

DDraceNetwork, a free cooperative platformer game. Contribute to ddnet/ddnet development by creating an account on GitHub.

12:52

<ChillerDragon> void CClient::ProcessServerPacket(CNetChunk *pPacket, int Conn, bool Dummy)

12:52

<ChillerDragon> if Conn is 0 Dummy is true and vice versa

12:52

<ChillerDragon> so confusing :D i thought since there are two variables they have two different informations such as wether the packet is sent to the active tee or not

12:53

<ChillerDragon> but both arguments Conn and Dummy have the same information

12:53

<ChillerDragon> Dummy is false*

?

12:53

Conn is an enum isnt it

<ChillerDragon> yes

Dummy is true if the message is not from the active connection

12:53

pretty confusing, yes

<ChillerDragon> i dont think so

12:54

<ChillerDragon> https://github.com/ddnet/ddnet/blob/master/src/engine/client/client.cpp#L2300

ddnet/src/engine/client/client.cpp at master · ddnet/ddnet

DDraceNetwork, a free cooperative platformer game. Contribute to ddnet/ddnet development by creating an account on GitHub.

12:54

<ChillerDragon> oh wait

12:54

<ChillerDragon> aaaa

12:54

<ChillerDragon> because cl_dummy changes

12:54

<ChillerDragon> ah right nvm then

12:54

<ChillerDragon> ah no then its totally fine xd

12:55

<ChillerDragon> But i still have no idea how leratos code works

it's still confusing

<ChillerDragon> by setting sound world to the wrong size of 0.7 connections

12:56

<ChillerDragon> for*

sound world

12:56

world sound

12:56

rename it

<ChillerDragon> lerato be like true == 3

12:57

<ChillerDragon> sound in da world

12:57

<ChillerDragon> im ok with that

12:58

<ChillerDragon> but im not native i dont use the saying "oh thats a sound thing"

12:58

<ChillerDragon> like "a sound plan"

xd

<ChillerDragon> sound world

you use that library unsound

<ChillerDragon> wowowo

yeah i agree, world sound is better

13:00

i agree with myself

<ChillerDragon> did you know clang built a C repl?

13:00

<ChillerDragon> they copied my project

hillerDragon

13:01

hashset uses 48 bytes xDD sadly tinyset is broken

13:01

now i use Option<Box<HashSet>> epyc totally not over complicated to use

13:02

but at least only uses 8 bytes, which is important in my case

13:03

@Robyt3 i have an important question and you need to answer honestly:

do you think tee-ish UI is something desirable?
Do you think that a new user that never seen ddnet, likes the home menu page?

Like imagine u open any game and it looks like ddnet's first impression

13:04

<ChillerDragon> https://zillyhuhn.com/cs/.1713618254.png

13:05

<ChillerDragon> they dont have readline support yet so no history and emacs mode thats pretty cring

13:05

<ChillerDragon> all in all my repl still feels supirior to clangs but im sure they will catch up

Jupstar ✪

@Robyt3 i have an important question and you need to answer honestly:

do you think tee-ish UI is something desirable?
Do you think that a new user that never seen ddnet, likes the home menu page?

Like imagine u open any game and it looks like ddnet's first impression

I don't feel anything particular wrong with it, I think it fits the game. Certainly better than others modern games that immediately open 3 popups telling you about sales and updates while you just want to start playing. Though the first thing you see if you open the game as a new player would be the language selection, name selection and tutorial (unless you skip it).

ok

<ChillerDragon> jopsti u like dis menu?

13:23

<ChillerDragon> https://www.youtube.com/watch?v=TN0zLNKQsGg

DoXto1234

Dr.Fetus Doing Stuff In The Pause Menu

i mean for a pause menu it looks pretty good xd

<ChillerDragon> a bit bloated but ye i think super meat boy managed to fit their 2d cartoon style while also being applealing to mainstream

ChillerDragon: Could I translate your Teeworlds programming videos and upload them to chinese video website?

<ChillerDragon> @fred axaxax

13:27

<ChillerDragon> i dont speak chinese ._.

13:28

<ChillerDragon> you mean like bilibili? sounds annoying to sign up there but i assume its a big market

Yes

13:30

Ah so, I think the best way to translate them is to add subtitles

<ChillerDragon> i still have trauma from when i tried to create a email on chinese yahoo or something like that

13:30

<ChillerDragon> so google translate subtitles?

he said, he translate it

13:30

not you

<ChillerDragon> ah lol thanks jopsti

13:31

<ChillerDragon> he also does the uploading then i assume

13:32

<ChillerDragon> @Bamcane ah you are just asking for permissions? ye sure

I was asking the premissions

<ChillerDragon> i licensed the videos on youtube under creative commons they are free to use and redistribute

13:32

<ChillerDragon> credits are appreciated

Thank you for your permission. Also, there is someone had upload them 4 years ago

13:36

And he only upload 7 videos in them.

13:36

So that's why i asked just now (edited)

<ChillerDragon> do you have a link? :)

What's the link?

<ChillerDragon> to the old 7 videos

13:40

<ChillerDragon> i would like to see how it looks like c:

https://www.bilibili.com/video/BV1da4y177Tc

[Teeworlds]Teeworlds programming 模組製作_哔哩哔哩_bilibili

<ChillerDragon> ty

he didn't do any translate

<ChillerDragon> it even links youtube

13:41

<ChillerDragon> is that a common practice? to reupload youtube videos to bilibili and then link the original video?

13:41

<ChillerDragon> almost like trying to advertise youtube on bilibili :D

Yes

13:41

if someone don't link the original video?

chiller: but nobody reads description xdd

<ChillerDragon> yeye ofc

13:42

<ChillerDragon> i store my passwords in youtube descriptions

13:42

<ChillerDragon> safest place on earth

Bamcane

if someone don't link the original video?

The uploader will be accused

<ChillerDragon> oh by whom?

13:43

<ChillerDragon> by the community? bilibili? or youtube?

ws-client

<ChillerDragon> by the community? bilibili? or youtube?

community

<ChillerDragon> ah i see

13:44

<ChillerDragon> https://zillyhuhn.com/cs/.chinadragon.png

13:44

<ChillerDragon> easy captcha

Hm

13:45

There will be a problem for you(probably)

<ChillerDragon> yea

you need a mobile phone number to receive the captcha

13:46

and more

<ChillerDragon> i see

13:46

<ChillerDragon> ok whatever bed time now thanks for sharing my content with china c:

Jupstar ✪

i dont understand anything xd netmsg = simply a new packet?

It's not always a packet, without flushing netmsgs are grouped into netchunks which are packets

Learath2

It's not always a packet, without flushing netmsgs are grouped into netchunks which are packets

ah right there clicks something in my head

The issue in teeworlds is that we have both state and events. So the source of truth for things is mixed, but they need to stay in sync

14:38

e.g. grenade explosion is an event, but it's effects on tees are only visible in the new state in the snapshot

14:38

if you send the explosion in a netmsg but it's effects of pushing tees around in the snapshot, now it is possible the netmsg arrives late, and since it has no facility to sync to a tick it's not possible to apply retroactively either

i'd claim on a stable connection events come in almost when snapshots come in

I'm guessing matricks just didn't feel like thinking about all the complexity of moving everything to events or everything to state and just had events that affect state visibly in the snapshot as a simple solution

Jupstar ✪

i'd claim on a stable connection events come in almost when snapshots come in

but we don't have a guaranteed way to make that happen in the current form of network we have without flushing all the time

Learath2

but we don't have a guaranteed way to make that happen in the current form of network we have without flushing all the time

are netchunks per connection?

We have a sendqueue of cnetchunks per cconnection iirc

ah our queue sytem is not multiplex?

If a chunk with a flush flag is added to the sendqueue, the entire queue is sent without <I don't remember what the regular forced flush event is> event happening

Learath2

If a chunk with a flush flag is added to the sendqueue, the entire queue is sent without <I don't remember what the regular forced flush event is> event happening

if you ask me, that should be done anyway ideally the server only sends snapshots and events to a single connection

14:45

at least i don't see a huge reason to not directly flush. maybe the implementation is not built for it but locking the mutex to put a packet on a send queue doesn't sound too expensive

14:46

well ok the implementation is also single threaded

14:47

that could defs be a problem

14:50

mh interesting do our netchunks contain some kind of timestamp? (edited)

bring back rainbow team colors this randomized one suck

15:14

who decided to change even

Ryu

bring back rainbow team colors this randomized one suck

it's not random it's just different order

15:25

looks uglier on tab tho :(

why was the order changed btw?

apparently to make it easier to distinguish between the teams

15:51

@MilkeeyCat what happens if you skill necromancer on max level?

wot

15:52

xd

MilkeeyCat

wot

correct answer

pub fn new(op: BinOp, left: Option<Box<Self>>, right: Option<Box<Self>>) -> Self {
  Self { op, left, right }
}

would you use Self as type or write type name? thonk

i'd write Self

Jupstar ✪

correct answer

now make binds system for 10+ dummies

lmao

and add vim motions to chat

Hello, I've got a question about @DDNet, how is it able to find people online? I am trying to make an npm package which should facilitate all things about ddnet and ddnet data and seeing that I want to also implement some sort of PlayerFinder class, but I am not sure how it's done (edited)

Sans

Hello, I've got a question about @DDNet, how is it able to find people online? I am trying to make an npm package which should facilitate all things about ddnet and ddnet data and seeing that I want to also implement some sort of PlayerFinder class, but I am not sure how it's done (edited)

u mean in server browser?

no, i meant this

15:58

$find Sans3108

Found 1 server(s) with "Sans3108" currently playing: 1. Server: DDNet RUS - Moderate — Link: https://ddnet.org/connect-to/?addr=37.230.210.231:8340/

Sans

$find Sans3108

but that is probably using the server browser information

15:58

https://master1.ddnet.org/ddnet/15/servers.json

15:59

^ here are all currently playing players

15:59

all server etc,

all of them from 1 single master server?

15:59

that makes stuff easier, nice

yep

1

def-

[ddnet/ddnet] Pull request opened: #8237 Update Estonian translations (by Cammo)

Checklist

[ ] Tested the change ingame
[ ] Provided screenshots if it is a visual change
[ ] Tested in combination with possibly related configuration options
[ ] Written a unit test (especially base/) or added coverage to integration test
[ ] Considered possible null pointers and out of bounds array indexing
[ ] Changed no physics that affect existing maps
[ ] Tested the change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-ad...

Jupstar ✪

yep

I'm having a bit of trouble finding the script/part of the code that sends the final response, I want to inspect it so I know what types I'm working with as well as construct a schema that I can use to validate the data

16:34

Do you know where can I find it?

Sans

I'm having a bit of trouble finding the script/part of the code that sends the final response, I want to inspect it so I know what types I'm working with as well as construct a schema that I can use to validate the data

i think every server can send what they want to a certain extend

16:38

but you could look how the client parses it

Jupstar ✪

i think every server can send what they want to a certain extend

I've noticed that with some other parts of the api

Still, there should be some code that constructs that json, and I think it might help me better understand it

16:39

but I'm not sure where to find it

https://github.com/ddnet/ddnet/blob/a00d6a311971cafe96f2ec7baf9637a9c5989be4/src/engine/client/serverbrowser_http.cpp#L258-L269

ddnet/src/engine/client/serverbrowser_http.cpp at a00d6a311971cafe9...

DDraceNetwork, a free cooperative platformer game. Contribute to ddnet/ddnet development by creating an account on GitHub.

Sans

but I'm not sure where to find it

sadly i cannot tell you where in the server it happens. simply search for http_register or similar

16:40

i mostly work client side

alright, thank you! tip

Sans

Hello, I've got a question about @DDNet, how is it able to find people online? I am trying to make an npm package which should facilitate all things about ddnet and ddnet data and seeing that I want to also implement some sort of PlayerFinder class, but I am not sure how it's done (edited)

https://github.com/ddnet/ddnet-discordbot/blob/master/cogs/playerfinder.py

16:43

the entire bot is open source

I've seen that, hence my question, thanks

Chairn

[ddnet/ddnet] Issue opened: #8238 Team saving message is confusing

Team save says to use /load mysecret if save is successful or /load word1 word2 word3 if it fails, then says the save is successful and to use /load word1 word2 word3 which is the opposite of the first message. !image Save is always successful, otherwise it wouldn't be possible to reload it. However, secret password might collide and a new one is generated, but the save is still successful.

Zwelf

[ddnet/ddnet] Issue opened: #8239 ASAN game/client/components/players.cpp:420 index -2 out of bounds m_aClients[ClientId]

Found in https://github.com/ddnet/ddnet/actions/runs/8766935096/job/24059631486?pr=8231 https://github.com/ddnet/ddnet/blob/a00d6a311971cafe96f2ec7baf9637a9c5989be4/src/game/client/components/players.cpp#L420 ClientId seems to be -2 ``` [-] Error: ASAN has detected the following errors: /home/runner/work/ddnet/ddnet/src/game/client/components/players.cpp:420:18: runtime error: index -2 out of bounds for type 'CGameClient::CClientData [64]' #0 0x2003748 in CPlayers::RenderPlaye...

Robyt3

[ddnet/ddnet] Pull request opened: #8240 Fix OOB accesses rendering ghosts, fix ghost hook collision line

Fix out-of-bounds accesses when rendering ghost players, which use the client ID -2. This was causing the hook collision line of ghost players to be affected by real players, which is not correct, as ghosts do not interact with other characters. Closes #8239.

Checklist

[X] Tested the change ingame
[ ] Provided screenshots if it is a visual change
[X] Tested in combination with possibly related configuration options
[ ] Written a unit test (especially base/) or added c...

What's this about? maxTime isn't being used where I'm assuming it should be, instead minTime is used for both keys, is this intended? right here: maps.py (edited)

Jupstar ✪

correct answer

should have naded down

Zwelf

[ddnet/ddnet] Pull request closed: #8240 Fix OOB accesses rendering ghosts, fix ghost hook collision line

22:19

github-merge-queue[bot]

[ddnet:master] 2 new commits

9be40e1 Fix OOB accesses rendering ghosts, fix ghost hook collision line - Robyt3 1b6ad6b Merge pull request #8240 from Robyt3/Ghost-Asan-Fixes - Zwelf

22:19

Zwelf

[ddnet/ddnet] Issue closed: #8239 ASAN game/client/components/players.cpp:420 index -2 out of bounds m_aClients[ClientId]

Sans

What's this about? maxTime isn't being used where I'm assuming it should be, instead minTime is used for both keys, is this intended? right here: maps.py (edited)

Doesn't look right

murpi

Doesn't look right

you're welcome

22:38

@murpi oh btw, I found a map which has an entry in releases, but doesnt actually exist

22:39

Mapper link, https://ddnet.org/mappers/Blue/ the map is called Lunar Base

22:39

it redirects to https://ddnet.org/maps/Lunar-32-Base/ but it says not found

AFAIK the map was removed from the map pool because it became impossible to complete

22:42

(due to a physics change I believe)

ohh

murpi

(due to a physics change I believe)

i heard about that also

Sans

you're welcome