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 2022-09-06 00:00:00Z and 2022-09-07 00:00:00Z

ping (edited)

[ddnet-maps:master] 1 new commit

9f8e283 M Strategic Paraplegic - ddnet-maps

is there any settings in ddnet server that allow me to set only ips can join the server?

why dont you just set a password lmao

The current state is that https://zillyhuhn.com/cs/.1660902889.pngI was working on it every day in vacation where I had no mouse or internet to distract me. Now back home I’m playing minecraft 24/7 :DSo the current state is that at first glance when looking at it in the dissector all is fine. Client and server send all the packets. The ddnet Client can properly join a 0.7 server and stay online. But I haven’t gotten to parsing the snap yet. So the client is not even rendering tees. Just the world.Now I need a new way of approaching this. Because until here I could always check which packets should be send and which I was missing and then search the code.I have no proper plan yet on how to tackle understanding the snapshot. (@heinrich5991)

heinrich5991#9827

ChillerDragon: cool that the 0.7 dissector helped you

what's the current state of 0.7 support? (reading backlog 2022-08-19, not sure if there's newer info available later) (<@749222324980416602_=5bquakenet=5d=20=43hiller=44ragon>)

chillerdragon

The current state is that https://zillyhuhn.com/cs/.1660902889.pngI was working on it every day in vacation where I had no mouse or internet to distract me. Now back home I’m playing minecraft 24/7 :DSo the current state is that at first glance when looking at it in the dissector all is fine. Client and server send all the packets. The ddnet Client can properly join a 0.7 server and stay online. But I haven’t gotten to parsing the snap yet. So the client is not even rendering tees. Just the world.Now I need a new way of approaching this. Because until here I could always check which packets should be send and which I was missing and then search the code.I have no proper plan yet on how to tackle understanding the snapshot. (@heinrich5991)

I could help

07:58

Maybe

08:00

I do the same basically in the server already

08:01

chillerdragon: do you have a public repo for this already? I would give it a try later

08:08

When it comes to implementing this to the browser, there should be a separated 0.7 tab which also does not show servers that are already on the main 0.6 tab. Maybe a manual connection as 0.7 client should be possible tho via command

I'd simply show all 0.6 and all 0.7 servers in the same tab

08:30

and prefer connecting to 0.6 for now

[ddnet/ddnet] Issue opened: #5811 Sometimes stuck in "getting game info"?

I've seen people on Discord and elsewhere saying that they're stuck on "getting game info". Given our flaky server connections, I assumed this would happen due to network filters, but I managed to accidentally reproduce it while connecting to localhost as well. So it appears there's a problem there.

chillerdragon: so what you're saying is that you need snap support in the wireshark dissector?

i keep confusing heinrich with fokko now LOL

Yea it’s a mess tho. I think it’s a branch In my ddnet fork. But it also needs another repo it’s a bit of a duck tape project as pf right now. But sure let’s have a call some day and look into it :) (@fokkonaut)

chillerdragon: do you have a public repo for this already? I would give it a try later

09:36

Yea we’ll see. I also though about a tab. My current implementation added a optional argument to the „connect“ console command to switch to 0.7 mode. Which is nice for testing (@fokkonaut)

When it comes to implementing this to the browser, there should be a separated 0.7 tab which also does not show servers that are already on the main 0.6 tab. Maybe a manual connection as 0.7 client should be possible tho via command

09:36

Hehe yea that could help :p (@heinrich5991)

heinrich5991#9827

chillerdragon: so what you're saying is that you need snap support in the wireshark dissector?

chillerdragon

Yea we’ll see. I also though about a tab. My current implementation added a optional argument to the „connect“ console command to switch to 0.7 mode. Which is nice for testing (@fokkonaut)

why the different tab though?

09:38

we could use the url scheme used by the master to distinguish between addresses of 0.7 and 0.6 servers

09:39

it calls 0.7 servers something like tw-0.7+udp://127.0.0.1:8303

I think especially in the beginning when it’s still alpha it makes sense to warn users what they get into. But we can also signalize it in one united tab or even apply filters

Oh wait, I remember someone doing a 0.7 skin renderer in DDNET

10:01

Do you think that would still apply to this whole 0.7 thing?

Wot?

10:01

Pls send

It was a PR, lemme see if it still exists

Maybe mine :D

10:02

Yea 0.7 needs more than just the protocol I’m working on. It needs a new skin menu as well

#5660

https://github.com/ddnet/ddnet/pull/5660

[DRAFT] Start working on 0.7 skins menu by ChillerDragon · Pull Req...

I started working on this a while ago and it is still far from being finished. Just wanted to share that I am working on this and get some feedback if it even has a chance to be merged when finishe...

And new textures and binds for ready mode

10:02

Ye that

Yea, makes sense. How would it work though? Would you automatically be set as a 0.7 tee in the respecting servers?

Yes

hm. it might make more sense to only define one skin

10:04

and make ddnet servers support 0.7 skins as well

10:04

even for 0.6 clients

They do already

10:04

To some degree

also for 0.6 clients?

10:04

I don't think so

But there are more 0.6 skins and the new databases etc

10:04

Yes

10:04

It translates

10:04

Not too good but a bit

ah. but I mean sending 0.7 skins to 0.6 clients as well

10:05

without translation

O

10:05

I see

10:05

Only works with the 0.7 supporting clients then tho

ye, would need to be guarded by a version check I guess

Yes that sounds neat

10:07

But then still players would lose access to all the 0.6 skins which is a reason to not updated

no, keep supporting 0.6 skins

So I would wait with that one release at minimum

support both 0.6 and 0.7 skins

Ah i thought remove the 0.6 skin menu

10:07

Then I understood you wrong

no, provide a toggle between both of them, basically

10:08

you can either select a 0.7 skin or a 0.6 skin

10:08

maybe find nicer names for them; maybe 'skin builder' and 'complete skin'

I see yes

10:08

That’s my plan

chillerdragon

Ah i thought remove the 0.6 skin menu

if it did then i would cry because i designed my tees with passion...

it'd be throwing away soo many years of creativity that it's not worth it

ima_6a3b909.jpeg

Image attachment

10:09

Like two settings entries that’s my plan atm

ddnet is all about creativity imo

I also wanted to look into porting all the 0.6 skins to a 0.7 format

10:10

But protocol first

10:11

Also no idea how to not render the 0.7 in black :D I might need some help on that

i stated this before but if we add in the 0.7 skin handler, will there be a way to make 0.6 tees with custom colors less... muddy?

what do 0.7 skins have to do with 0.6 skins?

good point. i was just bringing it up in terms of loose relevancy

okay

i'm trying to disect it myself, even if its low priority

heinrich5991

you can either select a 0.7 skin or a 0.6 skin

when you join a plain 0.7 server it means it picks the 0.7 skin from behind, even if you want a 0.6 sin usually?

no, you'd get a default skin then

Not a good design, imo

11:46

Having to switch manually to 0.7 skin settings when you join a 0.7 server seems a bit odd

11:49

Also, when connecting to a 0.7 server the client needs to tell it's skin information. Why would we pass standard settings here, when we have some selected ones available?

We can try to map atleast the vanilla set of skins like we already do, so if you’ve selected a “full skin” a similar-ish vanilla skin could be used in 0.7 servers

12:19

UI/UX is very difficult to get right in this case because “0.7 skin” doesn’t mean anything to a new player, but similarly if they pick a “full skin” it not working on 0.7 servers is also not very intuitive

Voxel

ddnet is all about creativity imo

cope

chillerdragon: you can't really port 0.6 skins to 0.7 because 0.7 skins are a bit higher res and upscaling won't look good

Aw shit. Yea I guess it requires a lot of manual work then. But should work better with skins where the svg files are available. If not one can recreate them manually or even automated a bit. (@Ravie)

chillerdragon: you can't really port 0.6 skins to 0.7 because 0.7 skins are a bit higher res and upscaling won't look good

DDNet

https://github.com/ddnet/ddnet/pull/5660

PLS NO

13:48

break backward compatibility already, focus on developing the game and not breeding crutches, refuse support for 0.7

13:48

Embedded image

is 0.7 support even "backward" compatibility?

14:02

I guess it can be thought of as being backward compatibility if you consider ddnet to be an evolution of teeworlds thus forwards of any teeworlds version

sideways compatibility

multiway compatibility

Matodor

Click to see attachment 🖼️

Yes

14:27

New map format

14:27

New protocol

14:27

Rust

14:27

What else?

14:28

A editor with networking

Ryozuki

A editor with networking

how would editor benefit from networking? xd

ddnet secretly spying editor!?!??!!??!!??

https://github.com/RoaringBitmap/roaring-rs

GitHub - RoaringBitmap/roaring-rs: A better compressed bitset in Rust

A better compressed bitset in Rust. Contribute to RoaringBitmap/roaring-rs development by creating an account on GitHub.

Ravie

how would editor benefit from networking? xd

collaborative editing

14:42

https://github.com/RoaringBitmap/roaring-rs/releases/tag/v0.10.0

14:42

impressive

why would multiple people edit a map at once

14:42

that's not very convenient

ive been working on that for the last 4 months XD

it is convenient lol

14:43

specially if u are prototyping parts

still one person working on a part at a time

a pixel art editor inside ddnet

14:45

specialized for making skins

+ you have to host the map on some server

well it's similar to how collaborative google docs work, ppl are not editing the same "part" at the same time but someone can do a part, someone else another and get feedback and do minor changes on others parts

or your friend just can destroy the whole map troll

troll

remove the gui and use imgui

14:46

add better ingame debugging tools

bomab

or your friend just can destroy the whole map troll

troll

not with versioned edits

with imgui its so ez

k2d222

not with versioned edits

oh damn

Ryozuki

remove the gui and use imgui

yes

https://github.com/jakobhellermann/bevy-inspector-egui

GitHub - jakobhellermann/bevy-inspector-egui: Inspector plugin for ...

Inspector plugin for the bevy game engine. Contribute to jakobhellermann/bevy-inspector-egui development by creating an account on GitHub.

14:47

this is such a convenient feature

14:47

you select a entity in the world

14:47

and yo uget a nice gui with all its properties

14:47

editable

14:48

Image attachment

14:48

game dev 101

Ryozuki

What else?

accounts

aPES2_Sweat

k2d222

well it's similar to how collaborative google docs work, ppl are not editing the same "part" at the same time but someone can do a part, someone else another and get feedback and do minor changes on others parts

you can already do that tho

14:50

just communicate and split work in a way that won't suck to connect together later

You can always find a way to make anything, it's about improving the tools available Personally I cant live without collab gdocs now even though sure you technically already could do it before just not the same workflow

it's not even inconvenient to do currently

14:54

maybe multiplayer editing is "cool" to some people but I don't think it really solves any problems

so what's your workflow when creating a collab map? you create each a few parts and exchange the map files and then one of you merge them together?

we make parts in our own files and test them together and when it's done we paste them into the "real file"

do you decide in advance where the parts go in the final file? do you also collaborate on the aestetics?

make some parts in advance and then put them into the real map in a fitting way

I think ur being a bit stubborn about smth that may benefit others, you can always collab ur way even with this

I just don't understand how it would benefit others

I think its clear

when do you ever need multiple people to edit the same map file at once?

Live editing

15:01

Thinking about ideas together? Mapping together

15:01

Idk to me its crystal clear

15:01

That this is rly useful

you need to load the map in game and play it together anyway

@Ravie ok well its good to know how you work, i would prefer to make a useful tool so I appreciate your feedback. I see how for you it wouldn't change much, although I would be satisfied if it could help you connect the parts together at the end and making final touches (edited)

15:03

but my primary goal I envisioned was to do the parts together not just a patchwork of individual parts

so make everything already in place?

yes and no, mostly prototyping

I think even with multiplayer everyone will be working on his own part in some separate place and then have to decide together how to paste them

1

maybe, idk, probable

15:05

but we wouldn't know until we try

sure

[ddnet/ddnet] Pull request closed: #5771 Fix weak bounce in move, improve vel change of bounce(when other char…

64bit Linux keeps such a surprisingly massive write buffer

16:35

20% of available system memory it seems

16:35

I was surprised at how quickly a copy to usb was done

windows does that better by default IMO

16:37

it doesn't keep a (big?) write buffer for externally attached media

I wonder if there is some kernel setting I can tweak to achieve that here

https://unix.stackexchange.com/questions/292024/how-to-reduce-linux-write-buffer-for-removable-devices

How to reduce Linux' write buffer for removable devices?

When writing content to removable devices on Linux (USB sticks/HDDs, SD cards, etc), I often see incredible write speeds in the first few seconds, sometimes in the order of GB/s (filling of the write

16:38

But why do you want to reduce it?

just buy 20% more RAM

deen

But why do you want to reduce it?

Because it's almost as if my tools are lying to me when they say the write is complete

16:38

copy*

I want to reduce it for external media so that file copying processes can show accurate process

If your tools don't sync afterwards, they don't tell you when the write is complete

cp doesn't sync

But my tools can't keep track of how far the sync is along anyway

at that point you're telling me that cp lies to me?

What the tools are saying is that the source files can be removed

fwiw gnome will just get stuck at 100% iirc

But yeah, no progress bar for the unmount is annoying

I'd be okay with that too

16:40

You basically get a umount stuck in an uninterruptable sleep with no information as to what's happening

Or use tools that skip the write buffer altogether like dd

Another weird result is mkfs.ntfs, it will 0 the partition by default, it will progress extremely quickly at the start. Then get stuck in uninterruptable sleep for a huge duration because it'll have filled the entire buffer

dd doesn't skip the write buffer

16:42

dd is just a fancy cp (unless you give it more parameters)

I usually dd straight to the usb stick without a file system. Otherwise I guess oflag=direct should skip

Anyway, if you are ever stuck waiting on a sync/umount you can monitor /proc/meminfo

16:43

grep Dirty:

heinrich5991

dd is just a fancy cp (unless you give it more parameters)

The further parameters are what makes it interesting of course. And that you can kill -SIGUSR1 to get a progress update

I use the tool progress to get progress reports for arbitrary tools operating on a single file. it also works for stuff like teehistorian2demo

hello there. can somebody help me to edit few thing in ddnet server. im kinda dumb for that :)

what do you want to edit? i cant help but im curious

Voxel

what do you want to edit? i cant help but im curious

xD

17:22

im creating uuh

17:22

https://www.figma.com/file/IptFiCs3mJyJKTVMS0CQUr/TW-Teams-(maybe)?node-id=202%3A3

TW Teams (maybe)

Created with Figma

17:22

can u check this :D

sure!

your eyes can start bleed. im sorry

:0 another figma user

sick burn

Image attachment

WTFFFF I JUST NOTICED THAT

Voxel

:0 another figma user

figma enjoyer

figma balls

creating a windows boot drive is surprisingly annoying to do on linux

17:30

mostly because windows installer is shit, but still

Image attachment

this whole concept page is just a ploy to insult voxel

Voxel

Click to see attachment 🖼️

i was lazy to put unique tees

Ravie

this whole concept page is just a ploy to insult voxel

why D:

17:32

i was asking about skins

17:32

but nobody answered me

E tier finally finished sunny side up etc

its ok i think its rly funny

Ravie

E tier finally finished sunny side up etc

coincidence

Voxel

its ok i think its rly funny

it is

it was the skin i rly liked

brownbear

1

17:33

:D

so many little voxels

contanimation

i wanted to make different tees

17:33

but

anyways i think my one complaint is that some of the text is rly hard to read

i gave up after i rendered 2

especially here

Image attachment

done

17:35

ez

grrh i wanna help with designing this but i think you need a paid figma account

17:35

ah cool

Voxel

grrh i wanna help with designing this but i think you need a paid figma account

why?

design every button painfully in inkscape like a chad

yea but you cant do it with other ppl

17:36

i always wanted to design shit with other ppl

17:37

closest thing i found to a free alternative is a thing called Penpot

cant u do it in figma?

idk can you?

u just need an account

17:38

do u have one? (edited)

yea

17:38

example of something i did in figma (its messy) https://www.figma.com/file/FC6e6BmlsAHePkmUJtWf3g/DDNet-Layouts

Created with Figma

wtf big effort for a teams finder lol

louis

wtf big effort for a teams finder lol

i sent this link in teeams server and nobody respond

17:39

cool

17:39

feelsbadman

Voxel

yea

are you hecta vox?

yea

@Voxel what's the auto tee thingy supposed to be? looks cool

Voxel

yea

now u can edit that layouts

Ravie

@Voxel what's the auto tee thingy supposed to be? looks cool

its been a plan in my head for quite a while, inspired by urrh.. sylbot i wanted a tee "animation" program thats not linked to any server at all, where you can have tees do certain actions on certain frames

ah replay bot editor

Voxel

its been a plan in my head for quite a while, inspired by urrh.. sylbot i wanted a tee "animation" program thats not linked to any server at all, where you can have tees do certain actions on certain frames

i need that D:

so you can do rly funny bugs that arent humanly possible

MilkeeyCat

now u can edit that layouts

thx!

Voxel

thx!

i think if u want help me u should do everything from 0

17:43

D:

Ravie

ah replay bot editor

yea. but again i want it to be abundantly clear i do not accept this as botting real plays. i just want a sandbox with a bunch of tees i can control

I've thought of something similar to program little ai tees for a tutorial map so they actually help you through parts

that'd be cute

but in my idea you would painfully write it in .cfg files or server settings xd

AI is getting better and better

17:46

Would be great if it was accessible for normal people without nasa pcs

today i spent 2 hours to find out how to stop ddrace server :D

click X on the server window

nnoono

17:47

from code

std::exit()

i did some other thing. dunno is it ok. but at least works

https://docs.microsoft.com/en-us/cpp/cpp/program-termination?view=msvc-170

C++ program termination

Learn about the standard ways to exit a C++-language program.

ok. next question. how can i... send messages in chat with delay?

17:49

i tried sleep() but it broke game xD

yeah sleep pauses the whole thread

yep

u probably need a separate counter

:|

you need some timer or count ticks

make a counter that decrements every tick and then print when counter=0

peepoNoted

Why do you need a delay?

50 ticks per second

Ravie

yeah sleep pauses the whole thread

ravie programmer

17:51

but thats wrong

17:51

it pauses the current thread

???

pilonpl

Why do you need a delay?

i just want try write 3 2 1 with delay

sleep pauses the thread it is called from

well he still said thread not threads

its not wrong just ambiguous

Ryozuki

it pauses the current thread

Isn't this literally what he has said?

pilonpl

Isn't this literally what he has said?

you need to say "current thread"

17:52

imho

he said "the" thread, so its clear which one is named

or yeah its ambigous

17:52

cuz ppl usuallyrefer to the main thread

17:52

when they say it like this

17:53

meh just nitpicking BASED

BASED

Ryozuki

you need to say "current thread"

Why? I am not a programmer but it seems like the exact same thing "whole thread" and "current thread"

it doesnt to me

17:54

whole means nothing at all

17:54

a thread can not be partial

Well

17:54

Not necessarily

17:54

Async? I guess

its rly important that it pauses the current thread

isnt async just a wrapper for spawning a new thread

I don't think so

anyway, the c++ call is called this_thread::sleep i guess it's clear from context xD

noby

isnt async just a wrapper for spawning a new thread

async can have nothing to do with threads at all

17:55

it depends on the runtime

create a whole new thread to use sleep() troll

troll

there are single threaded async runtimes

17:55

and multi threaded async runtimes

17:55

multi threaded distribute tasks on a pool of threads

17:55

e.g by using a work-stealing queue

17:56

like tokio does from rust

so can anybody do that for me feelsamazingman

feelsamazingman

learn rust to know more

17:56

actually

17:56

to learn about async more

17:56

make your own OS

17:56

do it in rust

17:56

https://os.phil-opp.com/

Writing an OS in Rust

This blog series creates a small operating system in the Rust programming language. Each post is a small tutorial and includes all needed code.

17:57

you can learn a lot

learn vulkan to understand synchronizing, vulkan forces you to know about memory layouts, coherent/non-coherent memory, you need to sync shader stages sometimes since GPUs are highly multithreaded, just learn vulkan, vulkan vulkan vulkan vulkan vulkan (edited)

yeah

17:57

learn vulkan too

Anyways, you could also wait before sending the command to the server so the server doesn't have to delay anything

if only i had more time

17:57

i just have an excuse to use rust at work

17:58

but hard to use vulkan

17:58

i dont do gpu stuff

17:58

@Jupstar ✪ i literally didnt know what interrupts were before i read and did this os-dev tutorial

17:59

mayeb i knew about it vaguely

MilkeeyCat

so can anybody do that for me feelsamazingman

feelsamazingman

if u really lazy, make smth like static std:.vector<std::pair<std::chrono::microseconds, std::string>> delayedmsg iterate through all instances, check if saved time is smaller than the wanted delay add your chats with current time

also rust comes rly handy in that lower level

17:59

at preventing races

wow rust is very woke, doesn't see race

Jupstar ✪

if u really lazy, make smth like static std:.vector<std::pair<std::chrono::microseconds, std::string>> delayedmsg iterate through all instances, check if saved time is smaller than the wanted delay add your chats with current time

Vec<(Duration, String)>

18:00

atleast rust got better naming sense

Ryozuki

mayeb i knew about it vaguely

would the perfect os even need interrupts xd

BASED

@MilkeeyCat Does this look good?

Image attachment

Jupstar ✪

if u really lazy, make smth like static std:.vector<std::pair<std::chrono::microseconds, std::string>> delayedmsg iterate through all instances, check if saved time is smaller than the wanted delay add your chats with current time

my javascript brain didnt understand anything. but huuuge thanks

Voxel

@MilkeeyCat Does this look good?

im looking at it rn

18:01

xD

MilkeeyCat

my javascript brain didnt understand anything. but huuuge thanks

its basically like in typescript

Jupstar ✪

would the perfect os even need interrupts xd

xd

array = vector

a pair is a tuple

Pair = {first:any, second:any}

but js doesnt have tuples

tru but typescript has generics

18:02

u could probs build the same datastructure

not truly

18:03

hmm

18:03

well maybe but its not first class

18:03

you cannot get a syntax like in rust (a, b)

@Ryozuki I have a question about rust, if you're rusty at rust does that mean you're better at it?

tee_thinking

1

we need philosopho tee meme

Ravie

@Ryozuki I have a question about rust, if you're rusty at rust does that mean you're better at it?

idk

18:06

i just know the more you use rust the more you like it

stockholm syndrome

BASED

chill, he really spams rust, i cannot keep up with this guy vulkan vulkan vulkan vulkan

everyone is promoting their own favorite thing

its a contest

MilkeeyCat

everyone is promoting their own favorite thing

https://discord.com/channels/252358080522747904/252358080522747904/1014947719078477874

kek

2

Image attachment

yeah, he is too fast

bluestripe

18:29

i dont even do it on purpose

When someone does Rust for the 1st time he gets converted ez

18:34

Rust is rly cool

ReiTW

Rust is rly cool

yes, you are one of my priests now

18:53

see you even make it SOUND like a cult LOL

it's like converting to islam

18:56

in a rly good way

thou must trust rust

@Ryozuki btw just saw that there's no raw socket implementation in the std

https://doc.rust-lang.org/std/net/index.html

std::net - Rust

Networking primitives for TCP/UDP communication.

19:16

there is udpsocket and tcpsocket

19:17

are raw sockets unix specific?

raw sockets are very much platform dependent

i guess its better to be left outside

had fun making stupid things like WoL using raw sockets

19:17

heard it was planned to be implemented

Voxel

see you even make it SOUND like a cult LOL

ofc i follow the joke

I think I still have my ugly wol code xD

https://docs.rs/pnet/latest/pnet/

libpnet

19:18

this looks like it could be used for raw sockets

damn I don't have my cleanest version of wol.c (rip, did alot of editafter) but was this (edited)

19:19

1.86 KB

heinrich5991

https://docs.rs/pnet/latest/pnet/

have to take a look

is there a c++ flag/lint that tells whether a header is used?

include what you use, but built in no

Ryozuki

is there a c++ flag/lint that tells whether a header is used?

probably undecidable :p

rust can tell it :p

since when does rust use headers?

it doesn't

19:31

rust can tell what?

19:31

I think rust doesn't even yell at you when you have an unused dependency

19:31

unfortunately. but in the case of rust, this would be decidable

is there some c++ tool that lets you just write rust (edited)

is that tool called rust?

19:32

what do you mean by "c++ tool"?

some early rust compiler probs was written in c++ xd

19:32

maybe u can use that

no

19:32

it was written in ocaml

heinrich5991

is that tool called rust?

rust does it again

there's a bootstrapping compiler for rust written in c++

Ryozuki

is there a c++ flag/lint that tells whether a header is used?

anyway as said https://github.com/include-what-you-use/include-what-you-use it was even used in ddnet already

GitHub - include-what-you-use/include-what-you-use: A tool for use ...

A tool for use with clang to analyze #includes in C and C++ source files - GitHub - include-what-you-use/include-what-you-use: A tool for use with clang to analyze #includes in C and C++ source files

hacky

19:37

i just do cargo c

19:37

in rust

if u want to compare it to rust, u'd need to compare it to c++ modules

BASED

bcs headers are much more powerful

no

yes

you can have conditional stuff too

19:38

with cfg!

19:38

https://doc.rust-lang.org/rust-by-example/attribute/cfg.html

cfg - Rust By Example

Rust by Example (RBE) is a collection of runnable examples that illustrate various Rust concepts and standard libraries.

19:38

(there is more than there)

how poweful is it?

as powerful as ast is

can u change the cfg like a preprocessor macro?

19:38

include the same file twice

nah

see

its more like conditional stuff for oses

19:39

you dotn want to include modules twice

19:39

you cant i think

19:39

anyway headers are a thing that new languages dont even think about

Jupstar ✪

include the same file twice

i mean u can probs make a macro to include source code

19:40

actually

19:40

sqlx has a macro that validates queries at compile time

anyway, include in c is really just text replacement

@noby i found a nice minecraft server for you! It has inverse anti bot. It is called anti human and basically kicks all players without a hacked client :D

kek

1

hi can anyone help me test a game im developing

kek

is that a yes

DESTROYER

hi can anyone help me test a game im developing

is it teeworlds related?

yes

what kind of game is it, a mod?

no its like a twist on teeworlds

is the source code on github?

yes

just share the github link

its ok someone joined

20:36

thanks tho

20:36

use wasd

20:38

whoever joined use wasd lol

20:40

and scrollwheel to zoom, and q to spectate

20:41

bruh why r u hacking lol

20:41

Image attachment

20:43

gg

20:43

who was it

Me

howd u hack lol

Huh

does it look likethis for u?

20:44

Image attachment

20:44

ur hammer streatched super far>?

Nope

oh what must be a bug

Zoom rate should be much higher, painful trying to zoom out

yea

20:45

alright

20:45

thanks for testing

No problem, everything else felt fine

@c0d3d3v I just tested this. Hmm...

20:52

All fixed!

20:58

also changed this

old:                    bool Stationary = Player.m_VelX <= 1 && Player.m_VelX >= -1;
new:                    bool Stationary = absolute(Player.m_VelX) <= 1;
ravie's reccomendation: bool Stationary = in_range(Player.m_VelX, -1, 1);

(edited)

21:03

didnt do ravie's thing but should i

@heinrich5991 what's the delay for a server to disappear from the https list?

I wouldn't, the 2nd one optimizes the best from what I can see

22:14

With ints they are all equivalent, but with floats abs optimizes the best

@Learath2 I suggested it just because it does the exact same thing as the first line but is the nicest to read imo

Exported 477 message(s)