DDraceNetwork - developer

@Voxel: honestly not sure what your images mean. You gotta use words to explain it. Btw I placed one really ugly hack as Easter egg for you to find in some of the tiles

chillerdragon

? You mean it’s not symmetrical? Should it be? Symmetric is boring and from what I understood the original isn’t either.

>symmetric is boring boring =/= visually pleasing >the original isnt either

Uhm

12:04AM

Okay yea maybe you have a point there :D

how many tees do u have ingame lmao

Go send fix pr mr artist

12:04AM

yes

Replying to @ReiTW how many tees do u have ingame lmao

ok one second

which matrix client do u use on phone ?

Element and FluffyChat

12:05AM

iOS

why both

Idk

12:06AM

To enjoy all the bugs

ok for one im going to steal your bolts because i dont know how to make it myself

Wat is a bolt

12:09AM

12:09AM

T_T

@Voxel: ?

12:10AM

Wat is it

The whole thing is a bolt?

12:11AM

Or only the smaller ones inside

12:11AM

Aren’t you the artist? You don’t know how to make those? o.0

im not skilled enough to make utter jank

Woah

12:19AM

Guess I’m ArtistDragon now then

12:20AM

More skilled than voxel

12:20AM

Check out my artworks https://www.deviantart.com/chillerdragon

chiller do you want the tiles or not

12:21AM

the fucking what now

JK it’s not my deviantart profile xd

cyberFighter

Click to see attachment 🖼️

im fixing chiller's unhooks

chillerdragon

JK it’s not my deviantart profile xd

ok i can sleep

Yea please if you think you can pixel perfecelize my work please adapt it and pr your improvement:)

Replying to @Voxel chiller do you want the tiles or not

halfway done btw

Nice I’ll sleep now. I will check your pr tomorrow :)

12:32AM

Thanks for doing it

VoxelDoesCode

[ddnet/ddnet-data-svg] Pull request opened: #52 Rework generic_unhookable_0.7 SVG

I saw the recent SVG recreation of generic_unhookable_0.7; and I do congratulate @ChillerDragon for the hard work put into it! However, there were some glaring visual bugs I wanted to try and tackle. I mostly retouched the proportions of the blocks and made it much more accurate to the original block set. I did reuse the tinier bolts/squares inside the tiles, so I do want to credit us both for the set. A closeup of before and after: ![image](https://github.com/ddnet/ddnet-data-svg/asset...

@Voxel tbh the vanilla ones look buggy with those broken edges

6:46AM

Do you think it was intended?

Put me as reviewer by chance

What broken edges?

7:34AM

I don’t even see the changes haha

Sry i meant corners

7:47AM

The corners of the inner quads

7:47AM

Top left corner to be precise

7:47AM

It looks bit cut off

7:47AM

But to me chillers look more correct

ChillerDragon

[ddnet/ddnet-data-svg] Pull request closed: #52 Rework generic_unhookable_0.7 SVG

8:39AM

ChillerDragon

[ddnet-data-svg:main] 2 new commits

ac96ee8 Rework generic_unhookable_0.7 - VoxelDoesCode 69f85b5 Merge pull request #52 from VoxelDoesCode/unhooks_07_redo - ChillerDragon

8:41AM

ChillerDragon

[ddnet/ddnet-data-svg] Pull request opened: #53 Fix fix in filename

diff
@@ -9,7 +9,7 @@
   id="svg5"
   inkscape:version="1.3 (0e150ed6c4, 2023-07-21)"
   xml:space="preserve"
   sodipodi:docname="generic_unhookable_0.7_fix.svg"
   sodipodi:docname="generic_unhookable_0.7.svg"
   inkscape:export-xdpi="96"
   inkscape:export-ydpi="96"
   xmlns:inkscape="http://www.inkscape.org/namespaces/inkscape"

trollerdragon chillerdragon: we actually have a rule that, if something fixes vanilla stuff we can keep it

@Jupstar ✪ opsi didnt see ur message and alr merged

8:43AM

i still do not know what you mean

ok wait

overall voxels thing is way better checkout my comparison gif

8:44AM

but yea maybe you are onto some little something

8:45AM

i'd say both versions have some stuff better

<ChillerDragon> ok when jopsti pr with best of both worlds

lmao

<ChillerDragon> ok so which edge you like better?

i like your corners more. but your big quad is bit worse than voxels

8:46AM

it has some weird offset between inner and outer quad

I like voxels corner more

8:47AM

his is more correctly matching and id arge it is not a bug

but why is there a random rounding?

8:47AM

yeah but why has vanilla this rounding xDD

you can also see it while tilting those in editor

to me that looks completely random tbh

they are not symmetric

8:47AM

makes them swag

yeah i understand that vanilla has those. but why xD

you wouldnt understand you are not a artist

ok, so only for swag, i c

8:48AM

yeah true

8:48AM

my brain only knows symmetry

symmetry is lazy

and perfection

perfection is not best looking

german engineering

8:48AM

:D

8:49AM

makes it more organic

but unhook is metal

you are mental

it's meant to look unorganic

8:51AM

get one of those in da editor and press the rotate key :D

8:51AM

its pog

8:52AM

i dont see how a corner like that happens by accident but im pretty sure there is some genius artwork behind

8:52AM

who even made the tileset? landil?

8:52AM

its masterpiece

LMAO

8:54AM

@Voxel read all this, i need your art brain to judge. but try to be objective in a sense that you don't look as vanilla to be the best one

@Voxel i take all my bolt bullying back you are truely one of the artist. I u fan now

8:56AM

jopsti thing is it looks good like a dynamic shadow

yeah he is really good

8:56AM

almost as good as an AI

creating such a corner is more work than spanning a perfect rectangle

8:56AM

it was to be placed intentionally

ChillerDragon

jopsti thing is it looks good like a dynamic shadow

yeah dunno. brain = symmetry

<ChillerDragon> https://user-images.githubusercontent.com/20344300/205356641-7c205d8f-4751-4248-9806-0cd15ed89aa4.png

8:56AM

<ChillerDragon> this is a bug

uff

wat now

which texture is that

cloud9

8:58AM

idk one of the clouds

8:58AM

i fixed it in my submission dont worry

ok, maybe it was a artist genius moment

8:58AM

maybe you broke it

omg trol

ok i rq now, cya later.always good to talk about art

Hello

9:25AM

From quacknet

9:25AM

quakenet*

helo

@Jupstar ✪ what are good resources to learn vulkan? :D

MilkeeyCat

@Jupstar ✪ what are good resources to learn vulkan? :D

i have 0 idea i read https://vulkan-tutorial.com/ especially for the setup code. but once i had all this stuff used once, i got a feeling for how vk works.. then i mostly used the vulkan docs indeed

10:33AM

what 100% helps are validation layers. and always fix the errors it says. because that makes some stuff clearer in your head

10:35AM

also what is important to note is. Knowing OpenGL 3.3 doesn't really help Opengl 4.4+ probably more. vulkan is not a global state machine like opengl 3.3

thanks

Jupstar ✪

also what is important to note is. Knowing OpenGL 3.3 doesn't really help Opengl 4.4+ probably more. vulkan is not a global state machine like opengl 3.3

yee.. vulkan aint for grugs pepeW

yeah, stuff like glEnable(BLEND) or smth in vulkan (at least if you don't use dynamic rendering) is really setup code already

10:36AM

u need to know what your app will use in advance

i never know it

rip anyway, it's mostly just tryhard

10:37AM

u should not give up if you debug a single thing for a whole day

i feel like it will be a fun journey

it is xd

10:38AM

i went full tryhard back then xDD

10:38AM

i stood up, worked on it, went to sleep

10:39AM

i miss those times ;~;

wanna forget everything and relearn again justatest

i'd also recommend to use the c headers to learn but later switch cpp headers or in rust to ash it's not fundamental different, but it helps you a bit to make fewer stupid mistakes you make because c sucks to work in xdd. be it ptrs or stuff like that imo that helps to make you understand the higher level header files a bit more and get a feeling how vulkan returns error codes etc.

10:50AM

and also, don't copy paste too much xd if you do an own project you are forced to actually understand what you are doing.. that's worth more than just reading code

i just want to understand how graphics work justatest

10:51AM

all these fancy ahh words

10:51AM

and such

graphics work like this: u have triangles u rasterize triangles u have pixels u colorize pixels

Fair

i dunno what exactly you want to learn

10:52AM

vulkan removes the driver overhead. it wont make your GPU run faster

I just want to understand ddnet code

that's very important to understand

In 3 weeks.. i didnt understand much justatest

e.g. for ddnet it's much faster bcs the memory model is MUCH more explicit i can control when to flush memory. i can say if i want coherent memory etc.

10:53AM

but GPU wise vulkan doesn't run any faster. it's really mostly CPU wise

MilkeeyCat

In 3 weeks.. i didnt understand much justatest

see the problem is, reading ddnet vk code is probably not a good idea to learn. I can give you a very quick summary what the idea of the backend is:

streamed memory is always batched, flushed at once before work is submitted to the GPU
in-use memory is always bound to the frames, so cleanup is delayed until the frame gets in use again
render threads are just there to collect draw calls. they don't mutate vulkan state
there is a memory heap that classifies between buffers, staging buffers (buffers that are used to copy content to images or vertex buffers), textures (which have sub classes, depending on the texture requirements & streamed buffers (per frame filled buffers)

if you see these concept in the code, u understand what's different compared to opengl backend

Jupstar ✪

see the problem is, reading ddnet vk code is probably not a good idea to learn. I can give you a very quick summary what the idea of the backend is:

streamed memory is always batched, flushed at once before work is submitted to the GPU
in-use memory is always bound to the frames, so cleanup is delayed until the frame gets in use again
render threads are just there to collect draw calls. they don't mutate vulkan state
there is a memory heap that classifies between buffers, staging buffers (buffers that are used to copy content to images or vertex buffers), textures (which have sub classes, depending on the texture requirements & streamed buffers (per frame filled buffers)

if you see these concept in the code, u understand what's different compared to opengl backend

Im not reading ddnet code to understand vk code, i wanna take a look at vk to understand ddnet vk code XD

MilkeeyCat

Im not reading ddnet code to understand vk code, i wanna take a look at vk to understand ddnet vk code XD

ok, anyway, i'd recommend to not look too much into the setup code. most important is simply to understand that there is no real global state. so you will not find something like changing texture wrap mode on fly it's all done in setup code.. later it's only like an array access pipelines[MY_BLEND_MODE_INDEX][MY_WRAP_MODE_INDEX]

hi sometimes i get this error in the console and server shuts down

11:10AM

someone help?

kahvebardagi

hi sometimes i get this error in the console and server shuts down

is that windows?

Jupstar ✪

is that windows?

yes

1. dump from %appdata%/DDNet/dumps would be nice 2. is this latest DDNet-Server or custom server?

MilkeeyCat

@Jupstar ✪ what are good resources to learn vulkan? :D

read the src justatest

Teero

read the src justatest

fair

Jupstar ✪

1. dump from %appdata%/DDNet/dumps would be nice 2. is this latest DDNet-Server or custom server?

i installed it on ddnet.org

11:19AM

and

11:19AM

can i share dump

11:19AM

here?

yea. try not to leak ips. (edited)

DDNet-Server_win64_crash_log_2023-11-13_11-32-31_4552_5032e0ee7a9f4406defc14873506356dde476d0c.RTP

6.44 KB

kahvebardagi

Click to see attachment 🖼️

thanks, if you have a github account you could create an issue on gh repo. Else i do it later if i don't forget it

Jupstar ✪

thanks, if you have a github account you could create an issue on gh repo. Else i do it later if i don't forget it

okay thanks

Jupstar ✪

@Voxel tbh the vanilla ones look buggy with those broken edges

honestly probably, but i kinda like it a tiny bit. it was like that in the original so i just had to keep it. (edited)

GitHub

Click to see attachment 🖼️

oops

Jupstar ✪

@Voxel read all this, i need your art brain to judge. but try to be objective in a sense that you don't look as vanilla to be the best one

I can try to see what it would look like without rounded corners

nightlies be trolling me again

cyberFighter

nightlies be trolling me again

What?

i get nightly update but it still says 17.3

That's intended

3:05PM

17.4 not yet released

ok but release candidate says 17.4

Because that's a release candidate xd

3:06PM

A candidate to be the next release

god deen hasnt allowed u 17.4

Voxel

anyways i made another example for what i see color correction in ddnet as

@Ryozuki did u see this

You are such a team mate voxel

Voxel

anyways i made another example for what i see color correction in ddnet as

accidental discord logo

xD

Jupstar ✪

You are such a team mate voxel

@Jupstar ✪

4:07PM

https://arxiv.org/abs/2310.01882

Fortran performance optimisation and auto-parallelisation by levera...

MLIR has become popular since it was open sourced in 2019. A sub-project of LLVM, the flexibility provided by MLIR to represent Intermediate Representations (IR) as dialects at different abstraction levels, to mix these, and to leverage transformations between dialects provides opportunities for automated program optimisation and parallelisation...

4:07PM

MLIR

4:07PM

https://arxiv.org/pdf/2310.01882.pdf

4:08PM

In this paper we explore complimenting the Flang MLIR general purpose compiler by combining with the domain specific Open Earth Compiler's MLIR stencil dialect. Developing transformations to discover and extracts stencils from Fortran, this specialisation delivers between a 2 and 10 times performance improvement for our benchmarks on a Cray supercomputer compared to using Flang alone. Furthermore, by leveraging existing MLIR transformations we develop an auto-parallelisation approach targeting multi-threaded and distributed memory parallelism, and optimised execution on GPUs, without any modifications to the serial Fortran source code.

4:08PM

by leveraging existing MLIR transformations we develop an auto-parallelisation approach targeting multi-threaded and distributed memory parallelism, and optimised execution on GPUs, without any modifications to the serial Fortran source code.

4:08PM

epyc

Interesting interesting

i told u

4:10PM

MLIR is probably the future of compilers rn

Ryozuki

MLIR is probably the future of compilers rn

just like the rust parallel frontend

4:34PM

we'll see how well it will be used

its way diferent lol

yeah but, what i mean is their claims doesn't need to reflect the real state

like the rust frontend can use MLIR

4:35PM

its totally unrelated

4:35PM

well this is a paper

4:35PM

not a blog post

LK99 is also a super conductor

u can read it

lmao

ok

4:35PM

why i bother even sharing

chill bro

then dont be like this bro

all i am saying is, wait and see how good it actually is

4:36PM

don't get overhyped about a product that isnt even finished

u can shame the The School of Informatics, University of Edinburgh, Edinburgh, UK

4:36PM

if u think their results are wrong

did i say that?

the paper is showing how good it actually is

4:37PM

using a supercomputer in the uk

4:37PM

This work has been funded by the xDSL ExCALIBUR EPSRC project. This work used the ARCHER2 UK National Supercomputing Ser- vice (https://www.archer2.ac.uk). This work used the Cirrus UK National Tier-2 HPC Service at EPCC (http://www.cirrus.ac.uk) funded by the University of Edinburgh and EPSRC (EP/P020267/1)

Cirrus

supercomputer for ddnet when

@Voxel this language is made for u https://www.modular.com/mojo

Mojo

: Programming language for all of AI

Mojo combines the usability of Python with the performance of C, unlocking unparalleled programmability of AI hardware and extensibility of AI models.

Ryozuki

the paper is showing how good it actually is

yes, but the question is just bcs this workload profits from it, doesn't mean it directly a replacement for anything

4:39PM

but as said, we'll see

4:39PM

thanks for sharing

4:39PM

intersting first analysis

4:39PM

it clearly has huge potential

mojo also uses MLIR

4:39PM

but i havent tried it yet

Ryozuki

@Voxel this language is made for u https://www.modular.com/mojo

ew

its made by the author of LLVM, clang and swift

4:41PM

Christopher Arthur Lattner (born 1978) is an American computer scientist, former Google and Tesla employee[1] and co-founder of LLVM, Clang compiler, MLIR compiler infrastructure[2] and the Swift programming language. As of 2022, he is the co-founder and CEO of Modular Inc, an artificial intelligence platform for developers.[3][4] Before founding Modular AI, he worked as the President of Platform Engineering, SiFive[5][6][7] after two years at Google Brain.[8] Prior to that, he briefly served as Vice President of Autopilot Software[9] at Tesla, Inc. and worked at Apple Inc. as Senior Director of the Developer Tools department, leading the Xcode, Instruments, and compiler teams.

4:41PM

thats one hell of a career xd

4:42PM

@Jupstar ✪ we should make this https://en.wikipedia.org/wiki/OpenCola_(drink)

OpenCola (drink)

OpenCola is a brand of open-source cola whose list of ingredients and preparation instructions are freely available and modifiable. Anybody can make the drink, and anyone can modify and improve on the recipe. It was launched in 2001 by free software P2P company Opencola, to promote their company.

4:42PM

at DDNetCon

4:42PM

if it ever happens

Ryozuki

@Voxel this language is made for u https://www.modular.com/mojo

jesus christ- it uses an emoji as a filetype??

open source coke

lmao

Voxel

jesus christ- it uses an emoji as a filetype??

bruv

4:44PM

@Jupstar ✪ !!

4:44PM

we got a new nvidia toy

4:44PM

https://www.nvidia.com/en-gb/data-center/h200/

NVIDIA H200 Tensor Core GPU

For supercharging AI and HPC workloads.

Ryozuki

bruv

what

Based on the NVIDIA Hopper architecture, the NVIDIA H200 is the first GPU to offer 141 gigabytes (GB) of HBM3e memory at 4.8 terabytes per second (TB/s) —that’s nearly double the capacity of the NVIDIA H100 Tensor Core GPU with 1.4X more memory bandwidth. The H200’s larger and faster memory accelerates generative AI and LLMs, while advancing scientific computing for HPC workloads with better energy efficiency and lower total cost of ownership.

4:44PM

need more ram

Ryozuki

https://www.nvidia.com/en-gb/data-center/h200/

nobody says supercharging except heartless executives

übercharge ready

Ryozuki

@Jupstar ✪ we should make this https://en.wikipedia.org/wiki/OpenCola_(drink)

LOL

Ryozuki

https://www.nvidia.com/en-gb/data-center/h200/

nice, i wonder if openai and comps already had these beforehand xd

this gpu costs more than

4:46PM

45k$

let's buy it

4:47PM

and in 2 years every consumer GPU can do the same xd

https://peergos.org/posts/blake3

4:47PM

when blake3?

4:47PM

Reasons to prefer blake3 over sha256

4:47PM

the q is, when blake3 over sha3

4:47PM

perf is insane

4:48PM

q is if in future sha3 will be more common or some blake version

gentoo uses blake3

4:49PM

The flavouring formula for OpenCola is:[4][5] 10.0 g food-grade gum arabic 3.5 mL orange oil 3.0 mL water 2.75 mL lime oil 1.25 mL cassia oil 1.0 mL lemon oil 1.0 mL nutmeg oil 0.25 mL coriander oil 0.25 mL neroli oil 0.25 mL lavender oil

i dont even drink coke xd

i drink coke zero sometimes

4:50PM

coke is too sweet

4:50PM

also makes u fat

sweeteners often make u hungry

4:50PM

so i dunno if that helps

whats blake3

the successor of blake2

Jupstar ✪

sweeteners often make u hungry

ye but technically then u can just eat healthier food

louis

ye but technically then u can just eat healthier food

exactly.. sweet drinks are always a problem

diet coke > coke zero though

bcs drinks don't make your body fed

Jupstar ✪

the successor of blake2

what does blake3 do

Voxel

what does blake3 do

u could say it's basically alternative to sha256

4:52PM

if u ever heard of that

it is a cryptographic hash function

i think? ddnet uses it maybe

louis

it is a cryptographic hash function

oh ok

Voxel

i think? ddnet uses it maybe

ddnet uses it for file identifying

4:53PM

e.g. your downloaded map

when replace with blake3

Jupstar ✪

e.g. your downloaded map

wasn't there that one time zombietroll used the map scraper to upload a funny map

louis

diet coke > coke zero though

whats the diff

idk diet tastes better imo

4:55PM

better aftertaste

louis

wasn't there that one time zombietroll used the map scraper to upload a funny map

when we still used CRC instead of sha?

Jupeyy

[ddnet/ddnet] Issue opened: #7450 Assert in network_server on windows

Reported by kahvebardagi on discord ``` 0x00000001401b9347: dbg_break() at src/base/system.cpp:223 (inlined by) dbg_assert_imp(char const, int, int, char const) at src/base/system.cpp:216 0x00000001400b9b9b: dbg_assert_imp(char const, int, int, char const) at src/base/system.cpp:218 0x00000001400a8338: CNetServer::Send(CNetChunk) at src/engine/shared/network_server.cpp:748 0x0000000140009d48: CServer::SendMsg(CMsgPacker, int, int) at src/engine/server/server.cpp:893 0x00000001...

@Robyt3 could I maybe move the https download URL into community json?

5:41PM

that way you don't have to do string concatenation to extract the actual URL

heinrich5991

that way you don't have to do string concatenation to extract the actual URL

So you specify the full URL for every community object individually? Fine for me, though I don't think the string concatenation is a problem in itself.

yes, that's what I mean

5:43PM

the problem I see with not doing it is that every consumer of this JSON has to implement it on its own

Hm, yeah, I suppose it's better if the json is standalone

5:44PM

Would also be easier to change URLs in the future then

yes

5:44PM

only one nightly version with this has shipped so far?

5:44PM

no official releases?

yeah

5:44PM

I don't think we need additional compatibility for those few

yes

5:46PM

while I'm at it, I'd also remove the "ranks-key": "" if it's empty

5:46PM

to distiguish between empty string and non-existent value

5:46PM

would that also be okay?

5:47PM

thanks for the amazing PR btw ^^

5:47PM

was a lot of good stuff and I'm only nitpicking here

Seems fine to use default "" if it's missing, but you also have to adjust the validation because the entry would otherwise be considered invalid

I'd consider the empty string a valid key then (no longer treating it special)

You mean differentiate between missing key and empty key? What would the empty key be useful for?

5:49PM

Right now the community entry is only considered valid if the attribute exists and is of type string, is what I mean with validation

we're not going to use it, but I find it weird to special-case the empty string

5:49PM

theoreticlaly (we're not going to do it), you could use the empty string as the key for the community ranks

Yeah, conceptually leaving the key out or using null would be cleaner

any admin up for dms?

do tiles behind clipping get rendered and then hidden or does the game not bother rendering them at all

cyberFighter

do tiles behind clipping get rendered and then hidden or does the game not bother rendering them at all

they get clipped by the GPU

7:57PM

so they vertices are called

7:57PM

but they fragments are filtered

jup isnt there a way to fake rotation in a fragment shader

8:30PM

just realizing

Voxel

jup isnt there a way to fake rotation in a fragment shader

u either do a rotation or u dont xd

8:32PM

or what kind of fake rotation does come into your mind

apelsin21

[ddnet/ddnet] Pull request opened: #7451 Appending in editor checks for duplicate images

To fix issue #7399 that appending maps in editor may duplicate images

Checklist

[x] 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
[x] Considered possible null pointers and out of bounds array indexing
[ ] Changed no physics that affect existing maps
[ ] Tested the change with [ASa...

aaaaaa

9:29PM

src/game/editor/editor.cpp:8064:69: error: code should be clang-formatted [-Wclang-format-violations]
                auto nameIsTaken = [pNewImage](const std::shared_ptr<CEditorImage>& otherImage) { return str_comp(pNewImage->m_aName, otherImage->m_aName) == 0; };

(edited)

9:30PM

whats the issue pepeW

& -> & is one clang-format issue, we also capitalize variable names (but not prefixes), so NameIsTaken and OtherImage

yuck

If you use Visual Studio then you can also run Clang-Format directly in there

im trying out clion :^)

Should be possible to use clang-format in CLion, but maybe you need to install/compile clang-format separately

Do All Variable Names Really Need To Be Like This?

https://wiki.ddnet.org/wiki/Development#Code_conventions

Development

9:36PM

Single letter variables can be lower case, e.g. i, x, y (edited)

axel

Do All Variable Names Really Need To Be Like This?

even if it's not the best casing for variables (which is completely subjective) it's still best to have consistent casing schemes for different keywords

ive seen worse is ok

wdym

i mean that its very ugly but ive seen worse

Jupstar ✪

or what kind of fake rotation does come into your mind

you said you can't rotate things inside a fragment shader, right?

Voxel

you said you can't rotate things inside a fragment shader, right?

AFAIR you can but then there would have to be another stage in the shader for it. Having the images pre-rotated is more efficient

Hm, I guess so.

Robyt3

[ddnet/ddnet] Pull request opened: #7452 Fix server crash on DDTeam message with lower sv_max_clients

When the CGameContext::SendChatTeam function is used to send a chat message to all members of a DDTeam (e.g. on completed /swap chat command), it was not checked whether the target players are online but only whether they belong to the specified team according to the teams core. However, the default team for unconnected/cleared players is 0 in the teams core, which is the same for players in team 0, so chat messages were being sent to client IDs not corresponding to connected clients when c...

@axel You should consider configuring your git user name and email so they are linked with your GitHub user, it currently doesn't link your profile on your commits (edited)

aaaaaaaa

You can squash all your commits in #7451 and then rewrite the author of the current HEAD commit with git commit --amend --author="Author Name <email@address.com>" --no-edit

https://github.com/ddnet/ddnet/pull/7451

Appending in editor checks for duplicate images by apelsin21 · Pull...

To fix issue #7399 that appending maps in editor may duplicate images Checklist Tested the change ingame Provided screenshots if it is a visual change Tested in combination with possibly rela...

ChillerDragon

[ddnet/ddnet] Pull request opened: #7453 Move teams out of ddrace gamecontroller (#7127)

Getting rid of some of the casts. Road to moddable ddnet with gamecontroller support.

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 ...

Robyt3

You can squash all your commits in #7451 and then rewrite the author of the current HEAD commit with git commit --amend --author="Author Name <email@address.com>" --no-edit

github-merge-queue[bot]

[ddnet:master] 2 new commits

b8d38f7 Fix server crash on DDTeam message with lower sv_max_clients - Robyt3 90fb33a Merge pull request #7452 from Robyt3/Server-Max-Clients-Crash - heinrich5991

11:01PM

heinrich5991

[ddnet/ddnet] Pull request closed: #7452 Fix server crash on DDTeam message with lower sv_max_clients

11:01PM

heinrich5991

[ddnet/ddnet] Issue closed: #7450 Assert in network_server on windows