DDraceNetwork - developer

DDraceNetwork

Development / developer

Development discussion. Logged to https://ddnet.tw/irclogs/ Connected with DDNet's IRC channel, Matrix room and GitHub repositories — IRC: #ddnet on Quakenet | Matrix: #ddnet-developer:matrix.org GitHub: https://github.com/ddnet

Between 2022-03-24 00:00:00Z and 2022-03-25 00:00:00Z

that's weird, clang and gcc disagree on the shadowing/shadowed variables, yet they read the same headers

[ddnet/ddnet] Pull request opened: #4881 Renamed all shadowed/shadowing variables

Get rid of all the shadowing that was reported by both gcc and clang. How both compiler can not see the same though they include same headers is a mystery.

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 if it works standalone, system.c especially
[ ] Considered possible null pointers and out of bounds array indexing
[ ] Changed no ph...

that should be my last one for the moment

[ddnet/ddnet] Pull request opened: #4882 Updated french translation

Updated french translation in agreement with T tee

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 if it works standalone, system.c especially
[ ] 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:/...

[ddnet-maps:master] 1 new commit

eb9b78e A Short And Precise 5 - ddnet-maps

[ddnet/ddnet] Pull request closed: #4875 Update Chinese translations by cheeser0613

07:29

[ddnet:master] 2 new commits

ab33372 Update Chinese translations by cheeser0613 - def- eb3e472 Merge pull request #4875 from def-/pr-chn0322 - Jupeyy

[ddnet-libs:master] 1 new commit

44a5db0 Update drmingw from https://github.com/Jupeyy/drmingw/actions/runs/2032792037 - def-

07:50

[ddnet/ddnet] Pull request opened: #4883 Update ddnet-libs with drmingw

from https://github.com/Jupeyy/drmingw/actions/runs/2032792037 Thanks Jupeyy

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 if it works standalone, system.c especially
[ ] Considered possible null pointers and out of bounds array indexing
[ ] Changed no physics that affect existing maps
[ ] Tested the change with [ASan+UBSa...

i made the best skin renderer (requires firefox & only tested on linux xd) (with dom.postMessage.sharedArrayBuffer.bypassCOOP_COEP.insecure.enabled true) <-- disable this again after testing https://aliveclan.de/ddnet/DDNet.html (edited)

Uncaught TypeError: Module.printErr is not a function

try firefox developer/nightly edition

ye im using that

webgpu on?

Image attachment

07:55

oh webgpu is off

i have webgl.disable-angle true webgpu true

Ryozuki

Click to see attachment 🖼️

i use the same version

still same error

restarted?

07:56

ok but your setup is also broken as we seen yesterday xddd

07:56

https://webglreport.com/?v=2

07:56

does this work?

my setup is fixed

07:57

Image attachment

07:57

hm

07:57

it reports the wrong renderer

07:57

r9 200 series

07:57

xd

mine too

07:57

mhh what do i do different xd

module.printerr is missing

07:57

xd

07:57

fix it

i dont have that error xD

u should have made it in rust

07:58

im sure it would work

07:58

already

07:58

greenthing

it works

Uncaught TypeError: Module.printErr is not a function setStatus https://aliveclan.de/ddnet/DDNet.html:1288 onprogress https://aliveclan.de/ddnet/DDNet.js:119

yeah

07:59

but thats bcs we have a difference in our browser

07:59

konqeror (kde browser) also works for me

07:59

dom.postMessage.sharedArrayBuffer.bypassCOOP_COEP.insecure.enabled true

07:59

try that

btw

08:00

do u have webrender

08:00

enabled?

08:00

gfx.webrender.all

08:00

i got it off

me2

08:00

i bet its dom.postMessage.sharedArrayBuffer.bypassCOOP_COEP.insecure.enabled

08:01

xd

[ddnet:master] 2 new commits

bf00e73 Updated french translation - Chairn 4be3fb9 Merge #4882 - bors[bot]

08:01

[ddnet/ddnet] Pull request closed: #4882 Updated french translation

gotta see, SDL2 is still a bit buggy with pthreads

it is

08:01

it worksn ow

08:01

well

nice

"downloading data"!

thats ok

08:01

my vps is slow

im playing ddnet on the browser

08:01

xd

xd

Image attachment

yeah i disabled internet for now

08:02

u know, websockets

08:02

annoying

08:02

but best skin renderer xD

XD

08:02

dom.postMessage.sharedArrayBuffer.bypassCOOP_COEP.insecure.enabled

08:02

ths doesnt sound secure

08:02

xd

yeah just disable again

08:02

i only use firefox dev edition for dev xD

08:03

i wonder why konqueror works tho xD

08:03

chrome also doesnt work

chrome bad

but its nice that emscripten works so well

08:03

the code changes to ddnet are very small

08:05

https://github.com/Jupeyy/ddnet/commit/28731ed886e3f6716d3cd64e8c4115059073fa1f

Emscripten support · Jupeyy/ddnet@28731ed

pog

(if u ignore the scripts, the actual source code changes are small)

does emscripten us web workers

08:06

https://developer.mozilla.org/en-US/docs/Web/API/Web_Workers_API/Using_web_workers

Using Web Workers - Web APIs | MDN

Web Workers are a simple means for web content to run scripts in background threads. The worker thread can perform tasks without interfering with the user interface. In addition, they can perform I/O using XMLHttpRequest (although the responseXML and channel attributes are always null) or fetch (with no such restrictions). Once created, a worker...

08:06

for threading?

08:06

xd

yeah

nice

but without sharedarraybuffers it currently wont work

rip

i think the option can stay disable when i enable PROXY_TO_PTHREAD but then SDL2 is broken xD

[ddnet:master] 2 new commits

c4d0a7a Update ddnet-libs with drmingw - def- 92aeddd Merge #4883 - bors[bot]

08:17

[ddnet/ddnet] Issue closed: #4877 drmingw creates directory on D:/

08:17

[ddnet/ddnet] Pull request closed: #4883 Update ddnet-libs with drmingw (fixes #4877)

@Ryozuki now it always works

https://aliveclan.de/ddnet/DDNet.html just had to change the apache config to send a header to disallow cross origin sharedarraybuffers

xd nice

I have an Intel i5-6200U CPU and that should have Vulkan support https://www.techsavvy.de/prozessoren/intel_core_i5-6200u/. I am on Arch Linux 5.16.12-arch1-1 and have installed vulkan-intel. In the nightly DDNet version I still have only OpenGL 3.0.0 to choose from. Does anyone have any idea why this is the case?

Intel Core i5-6200U im Detail

Hier finden Sie alle wichtigen Daten & Fakten zum Intel Core i5-6200U Prozessor. Spezifikationen ✓ Funktionen ✓ Spieletauglichkeit ✓

self compiled?

c0d3d3v

I have an Intel i5-6200U CPU and that should have Vulkan support https://www.techsavvy.de/prozessoren/intel_core_i5-6200u/. I am on Arch Linux 5.16.12-arch1-1 and have installed vulkan-intel. In the nightly DDNet version I still have only OpenGL 3.0.0 to choose from. Does anyone have any idea why this is the case?

also isnt there atleast opengl 3.3 too?

I tested both self compiled and the binary from the PR

11:46

Image attachment

11:49

Actually installed is opengl 4.6

11:49

❯ glxinfo | grep OpenGL OpenGL vendor string: Intel OpenGL renderer string: Mesa Intel(R) HD Graphics 520 (SKL GT2) OpenGL core profile version string: 4.6 (Core Profile) Mesa 21.3.7 OpenGL core profile shading language version string: 4.60 OpenGL core profile context flags: (none) OpenGL core profile profile mask: core profile OpenGL core profile extensions: OpenGL version string: 4.6 (Compatibility Profile) Mesa 21.3.7 OpenGL shading language version string: 4.60 OpenGL context flags: (none) OpenGL profile mask: compatibility profile OpenGL extensions: OpenGL ES profile version string: OpenGL ES 3.2 Mesa 21.3.7 OpenGL ES profile shading language version string: OpenGL ES GLSL ES 3.20 OpenGL ES profile extensions:

Jupstar ✪

@Ryozuki now it always works

https://aliveclan.de/ddnet/DDNet.html just had to change the apache config to send a header to disallow cross origin sharedarraybuffers

nice

c0d3d3v

I tested both self compiled and the binary from the PR

and have u tried to click on opengl 3.0.0?

11:52

it should open a popup

11:54

maybe we need a little popup icons like website do it, so its less confusing xDD

11:55

dropdown*

[ddnet/ddnet] Issue opened: #4884 Player in "Your team has 1 player that has not started yet" is wrong

As reported by cheeser0613. Seems to happen with large teams !unknown

Jupstar ✪

dropdown*

lol O.o I thought from the screenshot the list is always visible. xD yes if I click it vulkan is also visible

xd

I still only get 60 fps (no matter if vulkan or opengl) seems to be some limitation on my system. I used to have more fps, no idea what happened then. Probably a wrong configuration in my system.

[ddnet/ddnet] Pull request opened: #4885 Add dropdown icon

Checklist

[x] Tested the change ingame
[x] Provided screenshots if it is a visual change
[ ] Tested in combination with possibly related configuration options
[ ] Written a unit test if it works standalone, system.c especially
[ ] Considered possible null pointers and out of bounds array indexing
[ ] Changed no physics that affect existing maps
[ ] Tested...

c0d3d3v

I still only get 60 fps (no matter if vulkan or opengl) seems to be some limitation on my system. I used to have more fps, no idea what happened then. Probably a wrong configuration in my system.

turn off vsync xd

Ah I see

thanks

12:11

at least above 100 fps (I have set the update rate to infinite)

If I just close ddnet abruptly with Alt+F4 then it takes a long time to close and I get the following output: /usr/include/c++/11.2.0/bits/stl_vector.h:1045: std::vector<_Tp, _Alloc>::reference std::vector<_Tp, _Alloc>::operator[](std::vector<_Tp, _Alloc>::size_type) [with _Tp = std::vector<CCommandProcessorFragment_Vulkan::SRenderCommandExecuteBuffer>; _Alloc = std::allocator<std::vector<CCommandProcessorFragment_Vulkan::SRenderCommandExecuteBuffer> >; std::vector<_Tp, _Alloc>::reference = std::vector<CCommandProcessorFragment_Vulkan::SRenderCommandExecuteBuffer>&; std::vector<_Tp, _Alloc>::size_type = long unsigned int]: Assertion '__n < this->size()' failed. [1] 24548 IOT instruction (core dumped) DDNet If that was not yet known I can also look with UBSAN what that outputs.

do you do smth special?

12:12

i tested stuff like that with tsan, asan etc. so its not yet known

Nope also happens If I do it not in game but in menu.

Hey, why did u guys made that "Desktop fullscreen" option in the gfx_fullscreen ? Is it for mac users that were blocked without that option, or other important reasons ?

bcs multi monitor users requested a mode that doesnt minimize

12:18

windowed fullscreen e.g. is different, bcs it forces windows internal vsync

@Jupstar ✪ if one is always better than the other should we maybe merge them under the same name?

windowed fullscreen was explicitly requested

1

12:19

https://github.com/ddnet/ddnet/issues/4597

Windowed mode · Issue #4597 · ddnet/ddnet

OS: Windows 10 Pro 20H2 Client version: 15.8.1-win64 Radeon software version: 22.1.1 Every time I launch client in windowed mode: gfx_screen_height resets to 1421, despite the value 1440 in setting...

12:20

its also not comparable to bordleress window that uses the fullscreen

12:20

windowed fullscreen uses Windows' dxgi component somehow(dont ask me why and how, but thats the state of art)

c0d3d3v

Nope also happens If I do it not in game but in menu.

mh weird stuff

12:27

maybe a stacktrace for all threads can help understanding the problem

[ddnet:master] 2 new commits

252110a Support uppercase commands for /r and /w - def- b1e88dd Merge #4872 - bors[bot]

12:31

[ddnet/ddnet] Pull request closed: #4872 Support uppercase commands for /r and /w

12:37

[ddnet/ddnet] Issue opened: #4886 Graphics mode hotswapping inconsistencies

Starting the game in "Fullscreen" or "Desktop fullscreen" and switching to "Windowed" will result in a window with no decorations, no border. Starting the game in "Windowed borderless" and switching to "Windowed" will give borders but the new title will lack the icon for the client. !image vs ![image](https://user-images.githubusercontent.com/490500/159917660-acb73e5f-cc4c-4bd0-9ddb-58...

Jupstar ✪

mh weird stuff

Happens even with normal close

had not tried that. I will debug it later.

@Jupstar ✪ if the windowed borderless vs windowed fullscreen vs desktop fullscreen is confusing to me, I think it's overwhelming for a normal person just used to seeing "Windowed", "Windowed borderless" and "Fullscreen" in their options from other games

Learath2

@Jupstar ✪ if the windowed borderless vs windowed fullscreen vs desktop fullscreen is confusing to me, I think it's overwhelming for a normal person just used to seeing "Windowed", "Windowed borderless" and "Fullscreen" in their options from other games

i didnt invent windows weirdness

Could you maybe summarize all options so we can atleast document them somewhere and maybe give some guidance on best practice?

i did that a few times yesterday and today

What is the option I should pick if I want what I'm used to as "Borderless" from other games?

that depends on if u want smooth fps or not

12:44

smooth fps, then desktop fullscreen else windowed fullscreen

Can windows render above "desktop fullscreen"? That is is it suitable for IME users?

no

12:45

windowed fullscreen maybe, i cannot test it tho

12:46

lmao

Image attachment

12:46

windows vm xd

server browser counts dummy as player,Is it intended or is it a bug?

12:48

like, you see serv with 8 players, but there are only 4 actually, the rest are dummys

intended

[ddnet:master] 2 new commits

747dc1b Add note about gmock version (fixes #4853) - def- 07a95c2 Merge #4874 - bors[bot]

12:49

[ddnet/ddnet] Issue closed: #4853 Cannot compile test suite

12:49

[ddnet/ddnet] Pull request closed: #4874 Add note about gmock version (fixes #4853)

@Learath2 Windowed = A window that is part of the WM eco system Borderless = A window that is either part of WM eco system, or with fullscreen resolution implementation dependent (e.g. Windows does fullscreen optimizations for screens that block the full window(might be disabled by rightclick->properties->compability->fullscreen optimizations), KDE does not allow this in borderless) Windowed fullscreen = A special mode to allow fullscreen, but act like "Windowed" Desktop fullscreen = use desktop video mode(this should normally act like fullscreen just without minimize). This is also nativly supported by SDL for easier window managment if you dont care to request video modes etc. Fullscreen = exclusive fullscreen. for true gamers xdd

/

12:54

ops

rm -R /

12:55

xd

Jupstar ✪

@Learath2 Windowed = A window that is part of the WM eco system Borderless = A window that is either part of WM eco system, or with fullscreen resolution implementation dependent (e.g. Windows does fullscreen optimizations for screens that block the full window(might be disabled by rightclick->properties->compability->fullscreen optimizations), KDE does not allow this in borderless) Windowed fullscreen = A special mode to allow fullscreen, but act like "Windowed" Desktop fullscreen = use desktop video mode(this should normally act like fullscreen just without minimize). This is also nativly supported by SDL for easier window managment if you dont care to request video modes etc. Fullscreen = exclusive fullscreen. for true gamers xdd

Want to add those as tooltips with Ryozuki's change?

yeah should probs do that

[ddnet/ddnet] Issue opened: #4887 add new components to gameclient

Hello, I'm getting an error while I going to add new file to client/components folder so I added my file and also add that to gameclient.h but i'm getting error like blow, object of abstract class type "CDropweapon" is not allowed pure virtual function "CComponent::Sizeof" has no overrider

Seems IMEs can only render on top of "Windowed fullscreen". Thanks for explaining them

13:10

Borderless at full desktop resolution seems to have some windows special sauce indeed which prevents things from being rendered on top of it

13:10

but at any other resolution IMEs can render on top of borderless aswell

13:11

And our IME support seems quite nice now, I'm nicely surprised

yeah, its kinda weird

windows needs some way to just disable the compositors vsync, that would solve all our issues xd

Learath2

And our IME support seems quite nice now, I'm nicely surprised

nvm backspace absolutely breaks it

in menu or also ingame?

only ingame, seems a buffer is left unclean

13:13

in menu it works fine

ok, then should defs fix it

13:13

other way around would be more acceptable

same issue in f1, f1 and chat are broken

13:13

menu is completely fine

13:14

I'll create an issue

we have nobody to fix it sadly

@TsFreddie will fix it :3

robyte and tsfreddie dont really seem to be intereseted

I might give it a go myself if no one wants it

nice

Maybe an issue with text input length not matching actually, anyway, I'll leave it up for a bit see if maybe any of our resident IME people take it up

hehe

Also seems we misreport the position of our text box, but I bet most asian people have already gotten used to the IME just appearing in a random corner

[ddnet/ddnet] Issue closed: #4887 add new components to gameclient

xd

is it possible that turning on vulkan causes crashes or is that my bot client?

i could swear u already asked that yesterday

its defs possible

yes similar wonderings new crashes tho

i only tested the backend for like 1 month, and the only other person also testing it was mind

oke lemme alpha test for u babe

just show crashlog and we'll see

its not release yet is it?

13:50

lets see if i can reproduce

yeah not released yet

only RC

i can

the game runs weird when I use vulkan with the visual studio debugger but seems fine in release mode

https://paste.zillyhuhn.com/H3

ui

I use Arch BTW

xd

13:52

u use nvidia thats also good to know

Smetanolub

server browser counts dummy as player,Is it intended or is it a bug?

because a dummy takes a slot

13:52

so you can consider it as a player

chillerdragon, are you resizing or smth?

ReiTW

because a dummy takes a slot

why it takes a slot btw?

my dummy has feelings so its considered humanlike

it is like if you have two players connected tho

what do you do before it crashes?

13:53

random?

@Jupstar ✪ i jon multimapless and walk 10 tiles

for the server

zoom

13:53

tab

ok thanks

shoot

13:53

usualy vibin stuff

u have modern nvidia?

I use Arch BTW

13:53

or wat u mean

what nvidia

13:53

u have

13:53

rtx 4090

13:53

700 watt

Embedded image

13:53

super old

980 ok thanks

13:54

but will defs look into it

thanks babe

13:54

if u need anything tell me

ok

seems to happen in game only (not in menu or browser) after about 20-40 seconds of doing whatever

weird af

and i can reproduce with vanilla ddnet :D

i dunno if nvidia ships validation layers else try ./DDNet "dbg_gfx 4"

14:01

oh

14:01

normally nvidia drivers are really good

14:01

but gtx 980 is really old

14:01

so hard to say what could be the problem xd

https://paste.zillyhuhn.com/Jq

14:02

want me to buy new gpu?

xd

only true people uses freebsd pepoBagar

pepoBagar

i'd rather understand the problem, but sadly no validation layers enabled

vali wat

14:03

the ddoser?

? xd

idk wat a vali layer is

validation layer

i dont speak vulkan

xd

nerd

to validate your code

is this some scifi language?

yes

from starship enterprise?

btw what is your driver version?

the vulkans?

do u install yourself?

how to check

from nvidia.com?

i compiled my self

14:04

xd

14:04

nah jk

the driver?

i just installed some nvidia from pacman

ah ok

14:04

is arch using latest nvidia drivers always?

5 extra/nvidia 510.54-9 (27.5 MiB 27.5 MiB) (Installed)

sounds relativly new

it can do rendering already

14:04

that i can tell

can u try gfx_render_thread_count 1 -> restart -> to force single threaded vulkan

https://archlinux.org/packages/extra/x86_64/nvidia/

14:05

can i also start with gfx_render_thread_count 1 ?

14:05

i dont get the restart part

yes

14:06

bcs most ppl arent cool terminal kids xd

now my screen turned pink

its ok

prank

14:06

it didnt

14:06

same crash

ok MHHHHHHHHHHH

14:07

can u easily downgrade on arch?

want some coredump?

14:07

downgrade nvidia

14:07

no thats a pain xd

14:07

last time i checked it was a pain

14:08

but i think i can switch to lts kenrel plus nvidia lts

14:08

ou boi im gonna brick my system

then dont do it xd

@Jupstar ✪ What is the difference between WRAP_CLAMP and WRAP_REPEAT. When does this become noticeable?

Lmao tell pacman to think its 2014 https://wiki.archlinux.org/title/Arch_Linux_Archive#How_to_restore_all_packages_to_a_specific_date

Arch Linux Archive

i think i found the issue

14:10

thanks chillerdragon

pro

14:10

how u so smart

fucking copy&paste errors always kill me xD

do you wear coding socks?

[ddnet/ddnet] Pull request opened: #4889 Fix text texture name Texture -> TextureOutline

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 if it works standalone, system.c especially
[ ] 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-addresssanitizer--u...

can u test this chillerdragon https://github.com/Jupeyy/ddnet/tree/pr_text_texture_bug

14:15

u on 4k chillerdragon?

yes

ah

42 sec in

14:15

lgtm

thats why u got it so ez

ALMOST 1 MINUTE

14:15

omagawd i think u fixed it pro

yeah thanks for finding these stupid bugs xD

thanks for fixing those!!!!!!!!!!!!!!!

14:16

pog

14:17

have you considered rewriting the graphics backend in shell to avoid crashbugs?

maybe this also fixes the crash for @bencie and others (hopefully it all was this crash xD)

that would be pog

good idea

14:18

i'd love to see a translation library for vulkan to webgpu

14:19

then we have epic fps in browser xddd

isnt step 0 to have a browser client?

we have xdd

14:21

but it doesnt support UDP

14:21

and curl doesnt support websockets

ah that one

so currently no internet

isnt it super deprecated and old

https://aliveclan.de/ddnet/DDNet.html

14:21

here test out

the one by east eee something like that?

no

14:21

its current client

powered by emscripten

14:22

cringe watermark

the one from eee also uses emscripten, but since then it evolved

14:22

i didnt add any javascript myself

so its activley maintained?

14:22

by who where

thats the point

and why is it so ugly xd

its almost upstream DDNet code

14:23

since we have GLES 3 code which is alsmost webgl2 we dont need any hack

14:23

SDL2 works out of the box

14:23

how is it ugly

fullscreen is black xd

chrome?

14:23

firefox works for me

omg i lost my 3rd monitor xd

chrome is black for me too

im on firefox

strange

aaa ctrl+t

14:24

to gain back control xd

14:24

are you associated with the alive clan?

14:24

isnt that the fng gang?

14:25

omg its german website

yes i am leader xd

so where ddnet web code

14:28

and why is there this watermark -.-

u can also use custom html

14:28

i just chose the one from emscripten bcs i was too lazy xD

so why u dont

and it has some features

i noticed

14:29

but not worth the overallfeeling

14:29

Embedded image

14:29

i feel humiliated

u wont be able to compile rn: https://github.com/Jupeyy/ddnet/commit/1813b08fb06d34bcbf847957f99174aa542b6d91

ou

14:29

5hrs ago

14:29

its brand new!?

u need to compile all libs but freetype requires zlib and i havent finished that yet

14:30

yes i did it yesterday mostly

it still runs tho

14:30

when pr against ddnet

its nice for map preview

14:30

when i finish the zlib module

how do u load maps

not decided yet

u make my web editor look bad xd

xd

14:31

editor can already be used btw

14:31

but no way to save permament rn xD

14:31

i really just wanted it to compile and run

when i save

14:32

do i get code execution on ur vps?

no

14:32

its webasm

so its in my downloads folder?

14:32

xd

i dunno if webasm even has anything like a filesystem

/home/web_user

quite possible it just gets emulated

honeypot

14:33

but cool

14:33

gj

as said its not really made for IO, a map preview that isnt buggy like the one on ddnet.tw would already be nice

why delete adroid things tho

i didnt delete

14:34

i moved them

14:34

and merged in 1 script

14:34

i'll reduce the diff when i pr it

14:35

the script is genreally nice to compile the libs

ofc its nice

ez to extend it for windows too

its shell

then i can compile libs myself ez xd

why do you capitalize all your local variables?

i dunno

capitalized is for global env vars

feels more correct to me in script languages xd

exported crap

14:36

no

we do it in cmake too

its not script lang

why not

its pro lang

ez

[ddnet/ddnet] Issue closed: #4879 texts in editor also scaled

btw /register command not found

14:40

scam server

yo jupstar wat renderer you recommend btw?

14:49

reason i tested vulkan at all is i occasionally get some glitchy lines wondering my horizontals

14:49

is opengl 3.3 better than 3.0 if it runs?

14:51

https://www.youtube.com/watch?v=38WplkkBVGc

Old TV Static | Overlay (Download)

[ddnet:master] 4 new commits

54e8b64 Don't render tees or nameplates offscreen - sjrc6 abaf216 Seperate hook collision line from player rendering - sjrc6 5c29456 Improve nameplate boundary and clarify comments - sjrc6 309283d Merge #4870 - bors[bot]

15:03

[ddnet/ddnet] Pull request closed: #4870 Don't render tees or nameplates offscreen

Jupstar ✪

normally nvidia drivers are really good

kappa

kappa

rekappler

[quakenet] ChillerDragon

is opengl 3.3 better than 3.0 if it runs?

3.3 should perform better than 3.0, VK performs better than both iirc, all except 1.x render correctly

so I have a question about one bug (feature?). If you press A/D and go in spec, when you unspec you're still 'pressing' A/D, but if you press enter (to enter the chat) and then enter again to close the chat, you'll 'release' A/D. My fellow tee friend found, that if you remove this ResetInput call in src/game/client/components/controls.cpp, line 262, the bug get fixed. So the question is: can we safely remove this call? Or somehow modify it? Or was it intentional to have that kind of behaviour?

Image attachment

Ryozuki

kappa

kappa

rekappler

at least they are relativly stable xd mesa is best ofc xd

15:45

chillerdragon i dunno, nvidia on linux is gambling anyway, just choose what works best

Lumpy ◐ω◑

so I have a question about one bug (feature?). If you press A/D and go in spec, when you unspec you're still 'pressing' A/D, but if you press enter (to enter the chat) and then enter again to close the chat, you'll 'release' A/D. My fellow tee friend found, that if you remove this ResetInput call in src/game/client/components/controls.cpp, line 262, the bug get fixed. So the question is: can we safely remove this call? Or somehow modify it? Or was it intentional to have that kind of behaviour?

Looks like that is very intentionally being emulated there. You'd have to ask whoever added that

Learath2

Looks like that is very intentionally being emulated there. You'd have to ask whoever added that

how can I find who added that?

Git blame \O/

16:44

If it'a hard to find there maybe check when BugDDRaceInput was added

I'm not really familiar with git, is it a command 'blame'? or a program?

It is a command, but it's easier on githuv

16:50

There should be a button towards the top of the page that says blame

@deen do i need to rebase https://github.com/ddnet/ddnet/pull/4881 ?

Renamed all shadowed/shadowing variables by Chairn · Pull Request #...

Get rid of all the shadowing that was reported by both gcc and clang. How both compiler can not see the same though they include same headers is a mystery. Checklist Tested the change ingame Pro...

yes

Learath2

Looks like that is very intentionally being emulated there. You'd have to ask whoever added that

so I guess, deen added it? but then 'm_PlayerFlags != PLAYERFLAG_PLAYING' got changed to 'm_PlayerFlags == PLAYERFLAG_CHATTING' with comment 'Don't stop movement after opening player list' so I think A/D snaps in all cases except chatting? still not sure if it's a feature @deen can you look into it?

Image attachment

Image attachment

yeah, that's by design

can you explain a bit? what's the point?

I'm not sure, I think players always complained when something changed there

17:31

So it's a bit dangerous to change the code

[ddnet:master] 5 new commits

6086893 Renamed all shadowing/shadowed variables - Chairn dd7dfe1 Set Wshadow in CMakeLists - Chairn 5fe3e9a clang-format - Chairn b51957f Fix storage File shadowing - Chairn ebc714f Merge #4881 - bors[bot]

17:35

[ddnet/ddnet] Pull request closed: #4881 Renamed all shadowed/shadowing variables

deen

I'm not sure, I think players always complained when something changed there

are the complaints on the github or no? Just want to test the cases, coz I have this fix for a week now I guess and haven't seen any weird behaviour

lmao i just looked up how to install programms on suse and first hit shows teeworlds xd https://youtu.be/O3bgZEBI2DE?t=63

How to install "stuff" on OpenSuSE

[ddnet/ddnet] Pull request opened: #4890 Format member variable's name correctly

Still training my regex fu

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 if it works standalone, system.c especially
[ ] 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/ddne...

[ddnet/ddnet] Issue opened: #4891 Switching to vulkan and restarting will crash

As the game crash on start, once people switched, they can't change the renderer option back without twiddling in the cfg file. It should at least revert back to a base graphic renderer. ```shell [2022-03-24 19:48:31][client]: starting... [2022-03-24 19:48:31][sdl]: SDL version 2.0.8 (compiled = 2.0.8) [2022-03-24 19:48:31][gfx]: Created Vulkan 1.0 context. [2022-03-24 19:48:32][vulkan]: vulkan error: No vulkan compatible devices found. [2022-03-24 19:48:32][assert]: ../src/engine/cli...

18:55

[ddnet/ddnet] Issue opened: #4892 Firefox overlay stays on

Not sure if we can do anything about this. !Screenshot_2022-03-24_19-45-09

wat about add cmd for check teamrank/rank points ingame

what do you mean?

20:14

/rankteam & /rank & /points

he means the points you get for finishing top10 in team / in t0

20:15

Image attachment

20:15

2nd 3rd column

ah there https://ddnet.tw/ranks/

yeah

c0d3d3v

at least above 100 fps (I have set the update rate to infinite)

If I just close ddnet abruptly with Alt+F4 then it takes a long time to close and I get the following output: /usr/include/c++/11.2.0/bits/stl_vector.h:1045: std::vector<_Tp, _Alloc>::reference std::vector<_Tp, _Alloc>::operator[](std::vector<_Tp, _Alloc>::size_type) [with _Tp = std::vector<CCommandProcessorFragment_Vulkan::SRenderCommandExecuteBuffer>; _Alloc = std::allocator<std::vector<CCommandProcessorFragment_Vulkan::SRenderCommandExecuteBuffer> >; std::vector<_Tp, _Alloc>::reference = std::vector<CCommandProcessorFragment_Vulkan::SRenderCommandExecuteBuffer>&; std::vector<_Tp, _Alloc>::size_type = long unsigned int]: Assertion '__n < this->size()' failed. [1] 24548 IOT instruction (core dumped) DDNet If that was not yet known I can also look with UBSAN what that outputs.

any news on that?

Jupstar ✪

any news on that?

xD I have in my free time today only worked on the ninja bar that I just posted in #showroom. I'll take care of it tomorrow.

ah nice

20:56

looks good already ;D

[ddnet:master] 2 new commits

3778e0f Format member variable's name correctly - Chairn dbac9b7 Merge #4890 - bors[bot]

21:26

[ddnet/ddnet] Pull request closed: #4890 Format member variable's name correctly

[ddnet:master] 3 new commits

5459895 Add dropdown icon - Jupeyy 7c7fe59 Fix text texture name Texture -> TextureOutline - Jupeyy e8fc175 Merge #4885 #4889 - bors[bot]

21:45

[ddnet/ddnet] Pull request closed: #4889 Fix text texture name Texture -> TextureOutline

21:45

[ddnet/ddnet] Pull request closed: #4885 Add dropdown icon

Lumpy ◐ω◑

are the complaints on the github or no? Just want to test the cases, coz I have this fix for a week now I guess and haven't seen any weird behaviour

probably not, they were many years ago. We could try changing it and see if anyone complains

[ddnet/ddnet] Pull request opened: #4893 Update Chinese translations again

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 if it works standalone, system.c especially
[ ] 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-addresssanitizer--u...

[ddnet/ddnet] Issue closed: #4892 Firefox overlay stays on

Exported 473 message(s)