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 2021-01-18 00:00:00Z and 2021-01-19 00:00:00Z

[ddnet-maps:master] 1 new commit

8e52331 A Eventide - ddnet-maps

[ddnet-web:master] 1 new commit

b920081 Add donation by Songoku - def-

09:17

[ddnet-web:master] 1 new commit

577118b Remove defunct link - def-

Why does DDNet client hold hook if you hook, then go to /pause with a bind, release hook and unpause? I want that behaviour for my server aswell, but I dont know how

09:43

right now it holds the hook even if you release it in /pause, but after you resume the hook gets destroyed, only if you hold it again while resuming it will be held

hello, my name is apollo and i'm going through some crash problems, some maps are making my client stop responding like: baby aim 1 and caventure, please someone help me

@fokkonaut i think thats client side

10:01

since u can do that with f1 too

@Ryozuki yes, but something has to be serverside, i dont know what tho

10:02

it is clientside indeed, but it wont work on my server

u tried with ddnet client

10:02

?

yes ofc

idk then

ApolloZzcc

hello, my name is apollo and i'm going through some crash problems, some maps are making my client stop responding like: baby aim 1 and caventure, please someone help me

@deen any idea?

@ApolloZzcc are you on the latest version can you consitantly reproduce it?

and do you use Intel Graphics on Windows or something else?

wrapping errors in rust feels so nice

11:25

https://doc.rust-lang.org/rust-by-example/error/multiple_error_types/wrap_error.html

Wrapping errors - Rust By Example

A description

11:25

went from comented code to this

Image attachment

11:26

this was how i wrapped it

Image attachment

11:26

epic

11:26

feelsgoodman

11:27

@heinrich5991

11:27

implementing From<> on errors allows u to use ? which is rly nice

11:27

epic

[quakenet] ChillerDragon

@ApolloZzcc are you on the latest version can you consitantly reproduce it?

I'm using the steam version

bro nothing is more random than the vscode extentions marketplace all extensions appear and disappear or change author or are duped I can not find a single c++ extension currently :/

11:57

oh well ok "clangd" seems to be language support a weird name tho

11:57

maybe i should fully join the vim masterrace but for cpp its so incovnienent idk

12:00

is it me being color blind or did clangd decide to use same colors for variables and functions? :/ https://zillyhuhn.com/cs/.1610971212.png

Embedded image

irctroll

clangd is not the C++ plugin that’s the best on vscode. Microsofts offering is much better imo

clangd best

As for coloring, I think it depends on your color scheme, some color schemes only have so many colors so some get reused, or some don’t define different colors for everything

do u use vscode or vsodium or whathever it is?

12:26

and ye extensions disappear randomly

12:26

another reason iswitched to vim

12:26

it was rly annoying

Ah the open source version is a bit restrictive w.r.t extensions

12:26

Maybe that’s why you are missing some?

12:31

@Ryozuki are you a vim wizard yet?

i consider vim like c++

I never got as fast as I wanted with vim, too used to the mouse

u cant fully master it

12:32

i dont use the mouse tho

Ryozuki

i consider vim like c++

You should watch geohot’s stream if he ever streams coding again

12:33

That man is a true vim-wizard

xd

12:33

well there will always be super dope ppl

12:33

im on the level of someone who uses it on a daily basis i guess

Mastering is one thing, I just wasnt fast enough. Maybe I should give it another try one day

@Learath2 to make most profit from vim

12:34

u have to personalize it a lot

12:34

create ur own macros commands etc

Does vim have proper multiple cursors?

u can edit multiple lines

12:35

i did it once

I use multiple cursor editing quite a lot since refactoring is quite shit with most every ide

How to install the one by microsoft tho @Learath2

but i dont know if multiple cursos is the best solution

12:35

this sounds like the XY problem

12:35

u can use replace or make a macro

12:35

depends on the problem i guess

oh there is a open source version?

ChillerDragon: Honestly, no idea, it suggested it to me when I first opened a C++ file

ah nice i think ur right

brb watchign snk

Embedded image

12:36

seems like i have to install some ms binary blobs xd

12:36

from a sketchy aur

Binary version only has branding, telemetry and store settings

12:37

Telemetry you can disable(tm)

and backdoor to break linux and promote windows

https://vscodium.com/

VSCodium - Open Source Binaries of VSCode

Free/Libre Open Source Software Binaries of VSCode

i think i have that open src thing

12:37

i would sell my first born to get proper color so ms it is

The open source version has the store set to some open source clone of the ms store

12:38

Which is a nice idea. But it’s always so out of date and is missing a lot of extensions

12:39

Chiller I really think your color issue is your colorscheme not the c++ extension

doubt

12:40

the mobile age introduced the "there is a app for everything" slogan

12:40

and ppl always say linux has no software

12:40

but aur is bae xd

12:40

Embedded image

12:42

love you @Learath2 now i got OSS vscode and a ms marketplace and ms exnetions and colors are golden :3 thanks vscode por

12:42

pro*

12:42

good colors to fixe whitespaces and typos in comments :)

@Ryozuki there are some crates that generate these From implementations for you

:o

https://github.com/dtolnay/thiserror for libraries, https://github.com/dtolnay/anyhow for applications. for examples

dtolnay/thiserror

derive(Error) for struct and enum error types. Contribute to dtolnay/thiserror development by creating an account on GitHub.

Flexible concrete Error type built on std::error::Error - dtolnay/anyhow

13:47

*for example

13:47

there are a couple of others

@heinrich5991 this is nice, but only from a end user perspective

13:48

a library should wrap the errors

13:48

and not use this

13:48

imho

check the first link for libraries

oh

13:49

i only checked the second sry xd

13:49

thiserror looks like it

13:49

i dont like anyhow, and i used it once

13:50

its ok if ur lazy or early in the development i guess

13:50

idk

13:50

Use Anyhow if you don't care what error type your functions return, you just want it to be easy. This is common in application code. Use thiserror if you are a library that wants to design your own dedicated error type(s) so that on failures the caller gets exactly the information that you choose.

13:50

yep

13:52

@heinrich5991 why do u think thiserror is not part of the std?

13:52

to keep things minimal?

it doesn't need to be part of std in any case, since it doesn't appear in your public API

yeah

13:52

from my experience rust likes to have lot of things in external crates

13:53

u dont have to fear dependencies

13:53

regex is a crate, rand is a crate

13:53

etc

13:53

i saw this in a discussio

13:53

n

13:54

c++ developers are always afraid of using a library

13:54

but in rust case its not the same

13:54

xd

14:00

nice

Real c++ devs aren’t scared of any library. They all use the mother of bloat that is boost even

this crate reduced my code by 100 lines

14:00

more or less

14:00

gg

Image attachment

14:01

rust macros dope

Ryozuki

u dont have to fear dependencies

This is a consequence of static linking, the state of dynamic linking is so bad that people really don’t want to mess around with libraries if not necessary

@Learath2 my biggest problem with c++ is usually adding a library

I think it's a consequence of having a default way to build dependencies

not even using a library

14:01

u spend shitload of time just to get a library

14:01

specially in non windows

14:02

non linux

14:02

idk what im talking

14:02

xdd

14:02

*

@heinrich5991 eh, building all libraries yourself and statically linking them in removes all complexity from libraries on almost any language I can think of

14:02

Cargo just makes it trivial for rust

im making a url shortener for fun in rust

14:03

soon ill make it public

E.g. look how we never have any issue with the handful of libraries we build and statically link in ddnet either. Even though our build system is very questionable

we did have problems with the statically linked ones

sdl?

linking order, getting them to work for the first time

14:04

all not trivial

@Ryozuki using an in memory key-value store?

sqlite

14:04

im doing it rly simple

14:04

Image attachment

14:04

im nearly done

14:04

248 lines of code only

[ddnet-web:master] 1 new commit

68c384f Add donation by David*°Villa - def-

@heinrich5991 ah, linking order can get a little interesting, but it’s still usually trivial to figure out what is out of order. Iirc nowadays lld even tells you which one needs moving using some heuristic check

@Learath2 dynamic linking can be a bit of a shitfest on linux

14:06

one example i found is

14:06

haskell

14:06

arch maintainers rly want to use dynamic libraries on haskell

14:06

u should see the massive amounts of haskell libraries

14:06

i update

Dynamic linking is not “a bit of a shitfest” it is a shitfest on linux

xddd

Sadly it’s not much better on other platforms either. Windows almost never uses system libraries so every application brings their own libraries and they use very crude image signature techniques to not load the library a billion times. The really shared ones have to ship with 50 different versions and the loader needs to load the correct version, like msvcrt or the vc++ redis

14:09

macOS has frameworks the size of an entire linux distro to handle the versioning issue for the system shared libraries, but almost always applications ship every other lib/framework they use as the system versions can’t be relied on as they are ancient most of the time

14:10

The most recent version of appkit is around 40M

14:11

Linux is the only one that doesn’t waste memory and require an overly intelligent loader by default. But that comes at the cost of the pain of mismatching libraries

14:15

I once envisioned a runtime flatpak crafter, you run the executable and before the actual loader, I wanted to populate the “environment” with the correct dependencies. It would work on all platforms and not waste space, but it requires a huge repository with compatibility information

@Ryozuki I like the way arch packages haskell packages

[ddnet-web:master] 1 new commit

d1ef619 Wiki2 - def-

wtf with wiki now

14:46

where is my encodings page I need

14:47

@deen help

I will stop answering your questions if you ping deen

how can I unping deen

if you tell me that you're not going to ping deen for every single issue, I can tell you the answer

I wont

14:48

If you will answer my issues

What encodings page?

I believe it was wiki.ddnet.tw/encodings

@Ryozuki Any idea what happened with https://webcache.googleusercontent.com/search?q=cache:gSXqwOzoa0IJ:https://wiki.ddnet.tw/encodings/+ ?

Encodings :: DDNet Wiki

i guess it wasnt ported, it should be easily ported since it was just a mirror

14:50

embrace the wiki spirit and create the page

14:50

add content

14:50

and dont complain

@deen do you find it okay to be pinged randomly? if so, I rescind my comment above

Evgesha probably assumed that I was involved in the wiki project since I posted the news, so not really a random ping

@deen Zwelf told me he didnt want to port it

I've seen him ping you on quite a lot of issues. hm okay

On steam I get "An error occured while updating (app configuration unavailable)" any ideas?

14:53

Trying to play ddnet via steam for the first time

Nope, haven't seen that

@timakro rerun

doesn't work

hmm

@Дядя Женя the content was mirrored from https://github.com/heinrich5991/libtw2/tree/d3788aeb947fbecb33e6ebf6ef740cdc08040158/doc

heinrich5991/libtw2

Some Teeworlds stuff in Rust. Contribute to heinrich5991/libtw2 development by creating an account on GitHub.

"They are automatically generated. Better suited to just include it on ddnet.tw on a sub page"

nvm, was not a ddnet problem, can't install anything ^^

are u on arch

14:56

?

fresh windows install

oh windows

having issues with my vm

oh vm

14:56

xd

actually gpu

maybe missing some net framework redistribuibles

14:56

or smth like that

i thought it was a vm issue, had blackscreens and video glitches. then just said fuck it go back to dual boot and got the same issues ^^

14:57

now it magically works agin

monkaS

Zwelf: Can you make a PR on https://github.com/ddnet/ddnet-web to include them?

ddnet/ddnet-web

Static part of DDNet's website: https://ddnet.tw/. Contribute to ddnet/ddnet-web development by creating an account on GitHub.

Not sure what to do now. I had these issues a year ago. I could still rma my gpu but if they actually test it it might work for them, idk what happens then

u got the drivers right

Yes, I'll make a pr

what gpu btw

rtx 2060

15:00

I just told geforce experience to install the drivers

15:00

you can't do anything wrong there

did u download from nvidia web

15:00

windows installs automatically old drivers

15:00

its always better to redownload from the web

i got black screens with both

feelsbadman

also with half a year old drivers

15:01

if i get a 3070 or 3080 i'll get that and then try to rma ^^

15:01

but for now i prefer a half working gpu over no gpu

@timakro got another PSU to try? maybe it delivers instable power

15:03

Or putting the GPU into another system

never heard of instable power?

15:04

it's definitely enough if you mean that

i woud suspect the psu if it shutdown for no reason

15:04

but idk if black screens could be caused by the psu

but i can't really try any of that

15:04

Image attachment

15:04

Or this :D

i would rma the gpu

15:05

u rly should try to get it fixed

15:05

cuz buying a gpu noawadays

15:05

is a nono

15:06

my 200€ gpu costs 500€ now

15:06

the marketi s crazy

I can live without a gpu for a month

15:06

dedicated gpu

idk if its just a month monkalaugh

monkalaugh

yeah me neither -.-

15:06

But maybe I'll get rma money xD

xD

I'd be okay with paying 100 € more than the recommended retail price for an nvidia card

15:07

And you can get that if you look for it

i wonder

I had an eye on gpu prices because I was looking to upgrade anyways

15:08

I saw it twice the last months

15:08

xDD

the 3070 costs 920€ its msrp price is 500€

15:08

xd

like ~100 more

15:08

never for the "real" price

15:09

I see a 3070 for 730 here

my 180€ gpu

Image attachment

https://tease-shop.de/pc-notebook-zubehoer/pc-komponenten/12221064/gainward-rtx3070-phoenix-8gb-gddr6-hdmi-3xdp-grafikkarte-pci-express

german

monkaS

I thought availability of the 30 series would get better but it got worse haha

15:10

much worse

i heard this is due to only 2 companies being able to make the gpu silicon

15:10

and covid

Have you seen those monster machines they use to make cpu dies?

15:11

Must be similar to gpus

maybe with gpu is more insane

Can you do memory tests on vram?

[ddnet-web:master] 1 new commit

dcb1ea4 Add donation by SP | Someone - def-

@timakro maybe something like occt and there seem to be some cuda-based vram memtests

so should I sell my gpu now and upgrade to sli when the prices dump again? :D

SLI makes no sense

oof why

16:49

then the better gpu model i guess

16:50

i heard sli is not getting twice the performance but "no sense" sounds even worse

most games don't support it and even if they do, you don't get much more performance, but have much more power draw

so more eletricity costs hm

16:51

well my gpu is 5yrs old i can also upgrade it without going to sli ;D

16:51

invest in gpu market now :D

u dont invest when its in a all time high

16:55

i think

yea invest might be the wrong word but sell ur gpu now to rebuy later

16:59

its not investing its going short on gpu stocks xd

16:59

sounds like a save bet to me i doubt they stay high like this

sell your computer now

17:27

it's going to be cheaper in the future

troll

1

vram seems fine, this is the weirdest shit, i can't even reliably reproduce it (edited)

tried a different hdmi cable xd

17:39

or display port or whatever

would a broken display port cable crash the system though?

hard to say, probs not

17:50

tried unplugging from PCIe and clean the contacts?

17:51

and maybe try a ubuntu live usb

[ddnet/ddnet-web] Pull request opened: #111 Mirror libtw2 technical documentation

1

[ddnet/ddnet] Pull request opened: #3520 Fix median calculation

as reported by I.K.U on Discord

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/ddn...

[ddnet/ddnet] Pull request opened: #3521 Don't predict input from the dummy when it's paused

Fixes #3481

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-addresss...

[ddnet/ddnet] Issue closed: #3481 prediction bug on dummy spec while holding hook and cl_dummy_resetonswitch 1

23:26

[ddnet:master] 2 new commits

34d80e6 Don't predict input from the dummy when it is paused - trml aebe115 Merge #3521 - bors[bot]

23:26

[ddnet/ddnet] Pull request closed: #3521 Don't predict input from the dummy when it's paused

Exported 298 message(s)