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 2023-06-04 00:00:00Z and 2023-06-05 00:00:00Z

[ddnet-maps:master] 2 new commits

8d7fada A Nouisification - ddnet-maps e291155 A Unbalanced 3 - ddnet-maps

[ddnet/ddnet] Issue opened: #6701 Add file autocomplete for console ``exec`` command

I have a bunch of nested folders for all my client and server configs. And most of the time I have to look up the paths before running exec myfolder/somesubfolder/config.cfg because there is no way to list files from the console or even better auto complete paths. It should autocomplete all active storage locations. This feature will leak the file structure but exec already leaks most file contents so thats not much worse from a security perspective i guess.

i just went through a shitton of work writing code for ddnet that uses the STL without using exceptions because my stupid IDE said some one thing about it

06:35

and i automatically assumed exceptions were off

06:35

and then

terminate called after throwing an instance of 'std::bad_function_call'
  what():  bad_function_call

06:35

06:37

i wrote stuff like this

#define TRY_CALLBACK(a, b, c) { dbg_assert(bool(b), a " callback has no candidate."); b c; }
...
TRY_CALLBACK("Operation failed", fnOperationFailedCallback, (LOAD_ERROR_DIRECTORY_UNREADABLE, &it))
// instead of fnOperationFailedCallback(LOADLOAD_ERROR_DIRECTORY_UNREADABLE, &it);

so i could handle this error without catching an exception

06:37

Image attachment

06:37

oh well

cool that const char *GetPath(int Type, const char *pDir, char *pBuffer, unsigned BufferSize) in CStorage is not a part of the interface (edited)

07:18

so i can't get to it when i need to

07:20

oh nvm i see GetCompletePath

[ddnet-web:master] 1 new commit

ed26f23 Version 17.0.3 - def-

<ChillerDragon> @heinrich5991 do you remember why you marked sys.info client_version as optional?

08:24

<ChillerDragon> https://github.com/heinrich5991/libtw2/blob/master/gamenet/generate/spec/teeworlds-0.7.5.json

libtw2/teeworlds-0.7.5.json at master · heinrich5991/libtw2

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

08:24

<ChillerDragon> https://github.com/teeworlds/teeworlds/blob/26d24ec061d44e6084b2d77a9b8a0a48e354eba6/src/engine/client/client.cpp#L351

teeworlds/client.cpp at 26d24ec061d44e6084b2d77a9b8a0a48e354eba6 · ...

A retro multiplayer shooter. Contribute to teeworlds/teeworlds development by creating an account on GitHub.

08:24

<ChillerDragon> seems to be sent everytime on latest master

08:24

<ChillerDragon> was there a prior version?

08:24

<ChillerDragon> https://zillyhuhn.com/cs/.1685867077.png

Embedded image

08:25

<ChillerDragon> Its there since 5 years isnt that all 0.7 versions already?

08:25

<ChillerDragon> its already on the 0.7.0 git tag

08:26

<ChillerDragon> ops i messed up the spec link

08:26

<ChillerDragon> i mean this https://github.com/heinrich5991/libtw2/blob/09144b0ced40782b7a939a7d2a88c74c0e02db45/gamenet/generate/spec/teeworlds-0.7.5.json#L909

libtw2/teeworlds-0.7.5.json at 09144b0ced40782b7a939a7d2a88c74c0e02...

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

08:27

<ChillerDragon> @Jupstar ✪ should links open in a new tab without having to ctrl+click?

they should always open in new tab

<ChillerDragon> alr

Mr.Gh0s7

is this https://github.com/ddnet/ddnet/issues/2892 being worked on? I am interested on trying to implement it :) (edited)

Not that I know. You can try it, but I wouldn't say that it's a beginner friendly issue

1

1

1

Ewan

is ddnet against use of std strings in new code?

In the game loop yes, outside std::string is okay if the alternative would be to malloc/free some buffer anyway

Ewan

it's not a vector but it's similar

We only use v prefix for std::vector and not for other collections like std::set

i'll use a (edited)

a is array

initializer list is an array

08:35

you iterate over it

08:35

unless you mean a raw array

08:35

seems weird to not classify other containers

what are u even working on xd

08:36

i didnt read

not important

08:36

also i didn't say

08:37

basically batch file loader that i can use to easily asynchronously load skins and assets (edited)

you know whats great

08:37

it's june

08:37

and that means c++20

08:37

so u can use variadic template arguments and use constraints with it

08:37

https://en.cppreference.com/w/cpp/language/constraints

cpp/language/constraints

yea it's dope

08:38

but unnecessary

08:38

for this

Ewan

initializer list is an array

not really, we just use a for normal arrays and std::array and v for std::vector, we don't use prefixes for other iterable collections like std::set and std::list

silly

08:41

buy better IDE i'd say

why

then it tells u what type the variable is

it does already

08:41

what's that got to do with it

!rust good thing this discussion doesnt exist here

@Ryozuki

everything is simply snake case and good is

Ewan

what's that got to do with it

u say its silly

using a ONLY for T[] arrays and v ONLY for std::vector specifically is weird

08:43

so specific and not general

a is also for std::array

08:44

why does initializer list even exist at all

08:44

what problem does it solve over std::array

easy replacement for variadic arguments

yeah but how is that better than an std::array

template <typename T>
void doStuff(const std::initializer_list<T> &args) {
  for(const auto &it : args) {
    // stuff
  }
}

doStuff<int>({1, 5, 2, 7, 12});

08:45

idk

08:45

that's just what it's for

yeah, but to me it looks like this is nothing a std::array couldn't do or a constexpr std::vector

i don't know why it exists or why it's better

08:46

i just know that this is its use case

08:46

and if it exists alongside std::array and is for this specific purpose, i have reason to believe it is better suited to this purpose

08:46

otherwise it would not exist

i guess they wanted to save the size template argument

Using std::vector would be better here because you can just create a vector from an initializer list IIRC

no need for dynamic size

maybe they didnt like the fact that an vector can be dynamic

08:48

and forced compile time here (edited)

It just looks like some testing code so in that case it's okay

but since u need a template anyway, u can probs also simply use auto

But for other code I'd avoid using different types so we don't need to convert as much between layers

Ewan

basically batch file loader that i can use to easily asynchronously load skins and assets (edited)

btw sounds really useful, esp. the assets are so dead ass slow

1

Jupstar ✪

why does initializer list even exist at all

It's mostly a language construct. It's the glue that allows you to initialize vectors/std::arrays with the pretty brace syntax

hack

it feels hacky indeed

10:40

a new syntax but then a template struct

It also has some small guarantees that makes it lighter and trivial to optimize

i mean i often use the brace syntax, but i never used the resulting template definition

is there a client can join ddnet for cmd only?

10:41

i just want to use cmds to join a ddnet server

<ChillerDragon> what is cmd only?

10:42

<ChillerDragon> yes

10:42

<ChillerDragon> DDNet "connect ger.ddnet.tw"

i want to add bot to my server

<ChillerDragon>

botmorning

<ChillerDragon> ok what should the bot do?

10:43

<ChillerDragon> should it be server or client side?

move and fire

<ChillerDragon> based on what

10:43

<ChillerDragon> move where fire on what

players

<ChillerDragon> client or server side?

10:44

<ChillerDragon> client side this sounds kinda sus

10:44

<ChillerDragon> anyways it sounds like a technical thing to do

i mean is there a project like https://www.npmjs.com/package/teeworlds?activeTab=readme

Library for (ingame) teeworlds bots.. Latest version: 2.4.7, last published: 7 months ago. Start using teeworlds in your project by running npm i teeworlds. There are no other projects in the npm registry using teeworlds.

<ChillerDragon> do you know C++?

yes

<ChillerDragon> what is the problem with the project you referenced?

10:45

<ChillerDragon> or do you want it server side?

Jupstar ✪

i mean i often use the brace syntax, but i never used the resulting template definition

If you want your function to take a brace initializer list, if you take a initializer_list even with O0, it'll be zero copy. If you take a vector instead you are up to the mercy of the compiler

10:45

That's the only use

i just want to test connect a bot to a server with out modifiy the server itself (edited)

Learath2

If you want your function to take a brace initializer list, if you take a initializer_list even with O0, it'll be zero copy. If you take a vector instead you are up to the mercy of the compiler

but what if i simply use auto

10:45

which is allowed since c++20

<ChillerDragon> ok

10:46

<ChillerDragon> maybe dbg dummies is what you want

10:46

<ChillerDragon> im not really fllowing what execatly you need

Jupstar ✪

but what if i simply use auto

My inbrain compiler doesn't support templates yet, buuut I think it'll end up being an initializer list

yeah, but at least i dont use this weird initializer list struct that seems like bloat in newer cpp versions

<ChillerDragon> @Vinson do you want to start a client and connect it to a server? or do you want a server that connects clients to it self?

well it has it sense

[ddnet/ddnet] Issue opened: #6702 Add ``dbg_dummies`` to CI

It is currently crashing. And is nice for reaching code coverage in integration tests ``` 2023-06-04 12:48:10 I server: ClientID=0 rcon='dbg_dummies 10' Thread 1 "DDNet-Server" received signal SIGSEGV, Segmentation fault. 0x000055555561d373 in CPlayer::OnPredictedInput (this=0x0, pNewInput=0x7fffffffc010) at /home/chiller/Desktop/git/ddnet/src/game/server/player.cpp:514 514 if((m_PlayerFlags & PLAYERFLAG_CHATTING) && (pNewInput->m_PlayerFlags & PLAYERFLAG_CHATTING)) (gdb...

btw why nobody did a Docker image for ddnet? (I mean an official one) (edited)

Jupstar ✪

yeah, but at least i dont use this weird initializer list struct that seems like bloat in newer cpp versions

Mh, it only is sort of replaced by a constexpr array in newer cpp. So not sure how bloat it is

ws-client

<ChillerDragon> @Vinson do you want to start a client and connect it to a server? or do you want a server that connects clients to it self?

i want a client can connect to a server

10:51

but without graphics

<ChillerDragon> @Vinson okay then the npm package your referenced works fine

10:51

<ChillerDragon> have you tried it?

ok

10:51

nope

10:51

i am install npm

<ChillerDragon> ye its good

wow it works fine

Image attachment

<ChillerDragon> told ya

10:57

<ChillerDragon> enjoy botting axaxaxax

10:58

<ChillerDragon> note that if you automate walking towards players and shooting at them you need the permission of the server owner to run that

ReiTW

btw why nobody did a Docker image for ddnet? (I mean an official one) (edited)

^

there was one irrc

11:02

IIRC

<ChillerDragon> https://github.com/ddnet/ddnet/blob/master/Dockerfile

ddnet/Dockerfile at master · ddnet/ddnet

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

11:02

<ChillerDragon> or you mean pushed?

11:02

<ChillerDragon> because official ddnet servers run bare metal

this one doesn't run anything

<ChillerDragon> and kog is full on private

I meant a dockerfile that can be used by anyone

<ChillerDragon> i guess from the active ddnet devs there is not enough interest

11:03

<ChillerDragon> if you want one go pr one

will prob later make one

<ChillerDragon> do you run tw servers in docker?

If wg21 ever decides to accept P2447 std::span will replace all brace params so in C++26 explicit initializer_list constructors might become obsolete

yes I do

<ChillerDragon> which

currently only blockworlds

<ChillerDragon> u also rn kubernetes?

11:18

<ChillerDragon> do you use docker docker? or podman?

nah cuz not useful rn lol

<ChillerDragon> and why xd

11:18

<ChillerDragon> not useful ever*

11:18

<ChillerDragon> axaxax

well docker is still shit to run teeworlds, but well when you know how to fix stupid things that docker does, it works fine

<ChillerDragon> what stupid things?

11:20

<ChillerDragon> why is it shit?

For example I use a failover IP at OVH, but docker kinda doesn't support that

<ChillerDragon> i see

11:22

<ChillerDragon> and why do you want to use docker anyways?

11:22

<ChillerDragon> why not bare metal?

I mean I'd still be able to contact the container with that IP, but the response will be NATed with the main IP lol

11:23

So had to add custom iptables rules

11:23

Cuz it's easier to deploy things

11:23

Once you know how to use

<ChillerDragon> how is docker easier to deploy?

11:23

<ChillerDragon> could you name a example what was harder specifically and now got easier?

Pre-built images, you don't need to do everything by yourself

<ChillerDragon> like what?

Eg mariadb image (edited)

<ChillerDragon> ah so you do not have to run the oneliner to setup a fresh mariadb server?

11:24

<ChillerDragon> like creating a user etc

Yes

<ChillerDragon> https://zillyhuhn.com/cs/.1685877905.png

Embedded image

U just specify env vars, also I can easily isolate stuff

<ChillerDragon> what stuff tho?

For eg my mysql container will only be on the same docker network as my game server, but won't be visible by the rest

<ChillerDragon> seems like additional complexity to me

11:25

<ChillerDragon> which rest?

11:25

<ChillerDragon> what else is there

If I have other stuff to run

11:26

Wait

<ChillerDragon> like conflicting dbs?

11:26

<ChillerDragon> hm

Image attachment

11:31

this is an example of what u can do for eg

<ChillerDragon22> are you doing this?

11:32

<ChillerDragon22> and why?

you can have a network for all your websites, all your www containers will be in 2 networks, a private one to access their database, and the 2nd network is for traefik

11:32

yes I do that, I just restrict at maximum the visibility of each databases

<ChillerDragon22> and thats for what?

11:33

<ChillerDragon22> security?

11:33

<ChillerDragon22> cant you just make your db not expose public ports and done?

also traefik uses the docker socket, I use https://hub.docker.com/r/tecnativa/docker-socket-proxy to restrict the usage of it

<ChillerDragon22> seems so complicated

ws-client

<ChillerDragon22> cant you just make your db not expose public ports and done?

it is not exposed publicly, I meant to isolate it from other containers

ws-client

<ChillerDragon22> seems so complicated

it's not that hard lol

<ChillerDragon22> so other tw servers can not reach the db?

11:34

<ChillerDragon22> do you have a public infrastructure repo?

11:35

<ChillerDragon22> also TIL javascript can crash firefox xd

ws-client

<ChillerDragon22> so other tw servers can not reach the db?

for example yes

<ChillerDragon22> no for your infra

I don't think I have

<ChillerDragon22> sad

u mean u want to see how I setup all of that

<ChillerDragon22> i want to see every config file and the setup command

11:36

<ChillerDragon22> too evaluate how easy it really is

11:36

<ChillerDragon22> do you have the infrastructure tracked as code?

chillerevaluater

<ChillerDragon22> all the traefik and dockerfiles?

traefik is the most annoying to configure lmao

<ChillerDragon> i knew it xd

but the rest is easy

<ChillerDragon> if you get a new server what do you do to replicate the setup?

11:37

<ChillerDragon> how long does it take you?

takes me less than 30min

<ChillerDragon> thats pog

I just backup all my docker-compose stuff

<ChillerDragon> but how

11:38

<ChillerDragon> so you have not tracked it yet?

wdym by tracked

<ChillerDragon> check it in git

ah it's not in a git repo

<ChillerDragon> if you decide to do so and publish it let me know im curious

11:39

<ChillerDragon> im yet to be sold on docker

xdd

11:39

docker-compose makes your life easier tho

<ChillerDragon> but if you show me a simple repo where the config files make sense and there is a readme with 3 commands to set it up and then i have a nice tw network with dbs etc id buy it

I can make an example & show you

11:40

but later

<ChillerDragon> but until then i use my 13k lines of bash and bare metal to deploy tw servers

justatest

1

lmao

<ChillerDragon> minimal dependencies blazingly fast highly customizable

I made a docker image for ddnet but never put it anywhere

11:41

Maybe I can polish it up and put it on the repo?

<ChillerDragon> lerato wants to join kog dev team

well I removed alot

11:43

version: '3'
services:
  db:
    container_name: db
    image: mariadb
    restart: always
    volumes:
      - /opt/volumes/bombay/mariadb:/var/lib/mysql
    environment:
      MYSQL_ROOT_PASSWORD: pw
      MYSQL_USER: tw
      MYSQL_PASSWORD: pw
      MYSQL_DATABASE: bw
    networks:
      - bombay

  v3royal:
    container_name: v3royal
    depends_on:
      - db
    image: bombay:latest
    restart: always

    volumes:
      - /opt/volumes/bombay/v3royal/data:/blockworlds/data
    networks:
      - bombay
    ports:
      - 149.202.19.227:8404:8404/udp
  
  phpmyadmin:
    container_name: phpmyadmin
    image: phpmyadmin
    restart: always
    networks:
      - traefik
      - bombay
    labels:
      traefik.enable: true
      traefik.http.routers.phpmyadmin.entrypoints: websecure
      traefik.http.routers.phpmyadmin.rule: Host(`url`)
      traefik.http.services.phpmyadmin.loadbalancer.server.port: 80   
    environment:
      - PMA_ARBITRARY=1
networks:
  bombay:
    external: true
    name: bombay
  traefik:
    external: true
    name: traefik

but here's an example

<ChillerDragon> oh no

it sets up bombay, a phpmyadmin that is on the bombay network & traefik

11:44

and from the bombay network only the v3royal server is exposed

pls dont kill irc thank

xD

11:44

I can send a pastebin if it's better

Ew phpmyadmin user

@Learath2 yes, well it's only cuz I have some mates that are noobs

11:45

overwise wouldn't have used it lol

Make them learn sql

xD they don't have time for it ig

<ChillerDragon> thanks for the docker compose leak

11:45

<ChillerDragon> pretty boring tho xd

well ig easier than a bash script

11:45

if I want another srv I could easily duplicate the v3royal stuff & edit some lines

<ChillerDragon> i see

11:46

<ChillerDragon> yea not too bad i guess

11:46

<ChillerDragon> i just cp -r my 13k lines of bash

11:46

<ChillerDragon> and edit my config file

yesterday I've added the support for env vars

We scp some stuff over, edit some configs then hammer until it works

<ChillerDragon> HAMMER

Very manual workflow here at ddnet inc.

for eg:

  tw_srv:
    container_name: tw_srv
    image: bombay:latest
    restart: always
    environment:
      TW_sv_name: "This is a test server"
  ...

<ChillerDragon> isnt deen running his magic shell script on a new vps and its up and running after a cup of tea?

11:47

<ChillerDragon> i think id have to setup some tw servers with docker one day just for the memes

Ideally I would write ansible playbooks for deploying these

<ChillerDragon> ew ansible

yes that's the best option for you

<ChillerDragon> i somehow really dislike ansible idk why

as you are a docker hater if I remember correctly

11:48

xD

11:48

or was it heinrich, forgot

<ChillerDragon> docker is propriatariy

yes

I'm disgusted by docker

<ChillerDragon> proprietary*

11:49

<ChillerDragon> so it bad

But we have rootless podman to rescue

cammostripes

<ChillerDragon> @Ryozuki leak some emotes

well luckily podman uses the same set of commands

at docker

<ChillerDragon> like cammostripes

Works + not proprietary + rootless

idk if they have a docker-compose alternative

<ChillerDragon> pogman

ReiTW

idk if they have a docker-compose alternative

They do, it's podman-compose

<ChillerDragon> but little ecosys compared to docker

ah, damn

ws-client

<ChillerDragon> but little ecosys compared to docker

They share the same ecosys, wym?

<ChillerDragon> its not as drop in as they claim

It is very very drop in after you get through the setup struggles

<ChillerDragon> ..

If you are ok with running it as root it's even better

<ChillerDragon> lerato how to debug js?

spam console.log everywhere

don't use nodejs

11:51

use deno

<ChillerDragon> i cant open console

11:51

<ChillerDragon> my browser crashes if i open page

soon new laptop, idk if I stay on bloatdows 11 or I switch to a linux distro

<ChillerDragon> buy gnome pro

What sort of unholy bug did you encounter that it directly crashes your browser? Damn

<ChillerDragon> ikr

11:54

<ChillerDragon> idk

At that point your only hope is to attach gdb to the browser

<ChillerDragon> lmao

he tried to mine btc or wut

<ChillerDragon> that sounds horrible

use gentoo

11:54

be in control

I was thinking about manjaro

<ChillerDragon> majaro is better than windows

https://hannahmontana.sourceforge.net/Site/Home.html

11:55

this is better than windows

<ChillerDragon> its like ubuntu of arch users

wtf lmaoo

<ChillerDragon> lmao hannamonatana

maybe I could use templeOS

11:55

for my laptop

<ChillerDragon> i dunno bout that

u dunno templeOS?

ReiTW

maybe I could use templeOS

I don't know if it can even boot on real hardware

https://templeos.org/

Learath2

I don't know if it can even boot on real hardware

idk lmao, worth a try

No network drivers or networking layer. You'll just have to do pcie commands by hand to send packets

justatest

<ChillerDragon> xd

11:57

<ChillerDragon> i mean i dunno if it is a good idea

11:57

<ChillerDragon> stick with linux

11:57

<ChillerDragon> or maybe some bsd

Can you read words of god in linux?

<ChillerDragon> no but..

11:59

<ChillerDragon> i mean

11:59

<ChillerDragon> hm

12:03

<ChillerDragon> aaa leratoo!

12:03

<ChillerDragon> i found bugo

12:03

<ChillerDragon> i confused = and +=

12:04

<ChillerDragon> and i guess maffs fucked my ass due to it growing too big

Oh btw there is one other way to debug that kind of thing. You can use the debugger statement

<ChillerDragon> yea i wasnt sure where to put it

12:05

<ChillerDragon> so i used binary search with commenting out until it stops hang

12:05

<ChillerDragon> and then i knew which lines to look closer at

Smort

ws-client

<ChillerDragon> or maybe some bsd

bsd damn good

12:39

I've used freebsd in the past, it's amazing

discord is dead

13:16

rip

why?

bcs its offline

13:17

completely

13:17

i bet he wrote the messages half hour ago or smth xD

hmm? the #developer ?

13:18

ah yes ic now my msgs don't come trough either

ReiTW

I've used freebsd in the past, it's amazing

i used it too, and it lagged like shit

every 10 seconds a complete freeze for at least 1 second

Jupstar ✪

i used it too, and it lagged like shit

every 10 seconds a complete freeze for at least 1 second

never had issues

13:19

+ bsd was way better than linux in terms of network performance

13:19

prob why it's alot used by companies like Juniper etc..

how did u measure that? lmao

I didn't, I've read some blogs about it

netflix invested into the network code afaik

hmm? the #developer ?

yeah i read it too before

13:21

but i guess on normal human network it makes no real difference

13:21

discord hello

13:22

stop lag pls

13:22

i wrote my msgs at 15:20

Jupstar ✪

how did u measure that? lmao

this was also good some years ago

13:22

but I think since linux 5.x they optimized a lot

lel

13:23

lel

13:23

yeah i read it too before

<Jupstar> irc more stable

13:24

<Jupstar> why does it send my messages twice now

yeah i read it too before

<Jupstar> great

yeah i read it too before

bb discord spam = big ban discord spam

<Jupstar> yeah i also read it too before

13:26

<ChillerDragon22> hellow

13:26

<ChillerDragon22> oh no NaN

yaml
yaml:
  - sucks
  - so bad

<ChillerDragon> @Jupstar finally multi line code blocks

13:27

<Jupstar2> ah missed the chance

fn no() -> i8 {}

<ChillerDragon> oh tru

13:27

<ChillerDragon> lang is same line as ``` not next? right xd

13:27

<ChillerDragon> aaaaaaaaaaaaaaaaa

lel

ws-client

<ChillerDragon> @Jupstar finally multi line code blocks

couldn't you just write ``` directly in your chat?

Jupstar ✪

yeah i read it too before

spammer

giga_chad

oh god, dc ded

<Jupstar> don't need to talk anymore

ws-client

<ChillerDragon> @Jupstar finally multi line code blocks

cool add replies and I'm gonna switch too ;)

Jupstar ✪

lel

this is also not from me

13:29

what happens if discord resends a message that was a reply to another one xDD

let's communicate with morse code

today is the day to find it out

<Jupstar> yeah sounds great

from the typing

<Jupstar> new features, im in

13:30

<ChillerDragon> @Mr.Gh0s7 not sure how to make replies with a single line message protocol under the hood xd

13:30

<Jupstar2> our*

13:31

<ChillerDragon22> xd

13:31

<Jupstar> discord does it for me

13:31

<ChillerDragon> jupsti want me to deploy web update?

Jupstar ✪

yeah i read it too before

spammer

giga_chad

everything is so out of context on discord side xDDD

13:32

Image attachment

Jupstar ✪

yeah i read it too before

spammer

giga_chad

ReiTW

spammer

giga_chad

13:32

and you call me spammer

ws-client

<ChillerDragon> @Jupstar finally multi line code blocks

cool add replies and I'm gonna switch too ;)

<ChillerDragon> this happens when you code with specs from da head

ws-client

<ChillerDragon> @Jupstar finally multi line code blocks

couldn't you just write ``` directly in your chat?

Jupstar ✪

lel

this is also not from me

<Jupstar2> lmao, discord still auto sends some msgs

13:34

<ChillerDragon> @Chairn the problem was multiple lines and code highlight because irc only has single line so i had to virtually merge them to apply synatx to multi line code blocks

ws-client

<ChillerDragon> @Jupstar finally multi line code blocks

cool add replies and I'm gonna switch too ;)

13:34

from the typing

today is the day to find it out

Discord implemented ghosts! Ghost typers are out don't miss them

<Jupstar2> lets compress out messages to save bandwidth for discord

everything is so out of context on discord side xDDD

13:36

today is the day to find it out

13:36

everything is so out of context on discord side xDDD

13:36

@Mr.Gh0s7 discord trolls u

13:37

Image attachment

Image attachment

<ChillerDragon> lol discord as bugged as matrix xd

Image attachment

Image attachment

<Jupstar2> chillerdragon: irc has problems if i upload 2 images at once

13:38

<ChillerDragon> im sure they use kubernetes at discord and it bugged axaxax

Image attachment

Image attachment

U cannOt StoP the Gh0stς

13:39

U cannOt StoP the Gh0stς

13:39

U cannOt StoP the Gh0stς

nice it even dublicates images

U cannOt StoP the Gh0stς

test

feelsbadman

4 times now

18000ms

justatest

From the lags from disc I feel like vscode lags too and now I double check that what I did actually happened kek

kek

13:43

oh its fast again

cool

13:45

Is it only #developer or discord in general?

<Jupstar2> i assume in general

13:45

<Jupstar2> at least some servers

looks like cloudflare is in pain too

13:45

looks like cloudflare is in pain too

cool

13:48

cool

looks like cloudflare is in pain too

<ChillerDragon> what problems @Jupstar2

13:50

<Jupstar2> i uploaded 2 images

13:50

<Jupstar2> but only shows one

13:51

<Jupstar2> but maybe its the irc bridge

it shows 2 to me

<Jupstar2> then u probs lot of content already :D

yeah on discord its 2

<ChillerDragon> ah no i think i know what yo mean

13:52

<ChillerDragon> https://zillyhuhn.com/cs/.1685886724.png

Embedded image

13:53

<ChillerDragon> i guess never merge images then

13:53

<ChillerDragon> ah yea because of the message merge hmm

guys let's join ALIVE fng 64 players so we can communicate ;)

<ChillerDragon> just ask jupsti for webchat pw

!jupsti plz

13:57

!jupsti plz

<Mr.Gh0s7> noice

14:01

<ChillerDragon> ok nice images fixed

14:03

<ChillerDragon> the backlog is so nice to test stuff because you can reload the page without having to resend messages

14:04

<Mr.Gh0s7> pog

<ChillerDragon> @Ryozuki can you send a smol edlang snippet pls? annotated with ```edlang

14:18

<ChillerDragon22> https://zillyhuhn.com/cs/.1685888301.png

Embedded image

14:18

<ChillerDragon22> o yikes > html moment

hey, it's the first weekend i'm getting a full night sleep

15:11

love it

15:11

good night everyone

<Mr.Gh0s7> good night!

15:12

<ChillerDragon> gn

[ddnet/ddnet] Pull request opened: #6703 Only highlight hovered edit boxes when they are the hot item

Prevent highlighting when edit box hovered but behind a popup menu.

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

15:30

[ddnet/ddnet] Pull request opened: #6704 Extract `CMenus::UpdateColors` and `CUI::DebugRender` functions to improve readability of `CMenus::OnRender`

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
[ ] 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-addres...

15:35

[ddnet/ddnet] Pull request opened: #6705 Handle escape to go to start menu after popup menus

Popup menus should be closed before escape is used to navigate back to the start menu.

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

[ddnet:master] 2 new commits

455d2e5 Handle escape to go to start menu after popup menus - Robyt3 cdfb24f Merge #6705 - bors[bot]

18:27

[ddnet/ddnet] Pull request closed: #6705 Handle escape to go to start menu after popup menus

[ddnet:master] 4 new commits

efea2de Only highlight hovered edit boxes when they are the hot item - Robyt3 322e06e Extract CMenus::UpdateColors function - Robyt3 1c16d0e Extract CUI::DebugRender function - Robyt3 72108c4 Merge #6703 #6704 - bors[bot]

18:56

[ddnet/ddnet] Pull request closed: #6704 Extract `CMenus::UpdateColors` and `CUI::DebugRender` functions to improve readability of `CMenus::OnRender`

18:56

[ddnet/ddnet] Pull request closed: #6703 Only highlight hovered edit boxes when they are the hot item

Why is this like this

    for(auto &pLayer : m_vpLayers)
    {
        if(pLayer->m_Visible)
        {
            if(pLayer->m_Type == LAYERTYPE_TILES)
            {
                CLayerTiles *pTiles = static_cast<CLayerTiles *>(pLayer);
                if(pTiles->m_Game || pTiles->m_Front || pTiles->m_Tele || pTiles->m_Speedup || pTiles->m_Tune || pTiles->m_Switch)
                    continue;
            }
            if(m_pMap->m_pEditor->m_ShowDetail || !(pLayer->m_Flags & LAYERFLAG_DETAIL))
                pLayer->Render();
        }
    }

    for(auto &pLayer : m_vpLayers)
    {
        if(pLayer->m_Visible && pLayer->m_Type == LAYERTYPE_TILES && pLayer != m_pMap->m_pGameLayer && pLayer != m_pMap->m_pFrontLayer && pLayer != m_pMap->m_pTeleLayer && pLayer != m_pMap->m_pSpeedupLayer && pLayer != m_pMap->m_pSwitchLayer && pLayer != m_pMap->m_pTuneLayer)
        {
            CLayerTiles *pTiles = static_cast<CLayerTiles *>(pLayer);
            if(pTiles->m_Game || pTiles->m_Front || pTiles->m_Tele || pTiles->m_Speedup || pTiles->m_Tune || pTiles->m_Switch)
            {
                pLayer->Render();
            }
        }
    }

at CLayerGroup::Render() on game/editor/editor.cpp

what is your question? XD

more specifically why is there two if under for which check for the same thing

first loop render tile layers that are not gameplay related, mostly backgrounds and graphics one (edited)

19:17

second loop render tile layers that are gameplay related

but it checks if they are not game layers

19:18

in the first if

19:20

well besides the if check which i dont understand chairn is right tho i guess

I mean why are the if seperate. Wouldn't it better it checked (pLayer->m_Visible && pLayer->m_Type == LAYERTYPE_TILES) and then

(pLayer != m_pMap->m_pGameLayer && pLayer != m_pMap->m_pFrontLayer && pLayer != m_pMap->m_pTeleLayer && pLayer != m_pMap->m_pSpeedupLayer && pLayer != m_pMap->m_pSwitchLayer && pLayer != m_pMap->m_pTuneLayer)

under the same if?

wow thats your question? xd

19:23

answer: who the fuck cares xD

xD it's confusing though

better tell me why

(pLayer != m_pMap->m_pGameLayer && pLayer != m_pMap->m_pFrontLayer && pLayer != m_pMap->m_pTeleLayer && pLayer != m_pMap->m_pSpeedupLayer && pLayer != m_pMap->m_pSwitchLayer && pLayer != m_pMap->m_pTuneLayer)

evaluates to true

I am working on something that will probably use CLayerGroup::Render() should I merge those to for loops into one?

the for loops cant be merged

19:25

as chairn said, it purposely renders the physics layers after design layers

19:25

so they are always visisble in editor

ah so it's because of how pLayer->Render() works?

mh?

19:26

it's so they are rendered last

19:26

which makes them visible

yeah that's what I meant pepeH

pepeH

Jupstar ✪

better tell me why

(pLayer != m_pMap->m_pGameLayer && pLayer != m_pMap->m_pFrontLayer && pLayer != m_pMap->m_pTeleLayer && pLayer != m_pMap->m_pSpeedupLayer && pLayer != m_pMap->m_pSwitchLayer && pLayer != m_pMap->m_pTuneLayer)

evaluates to true

even tho i still don't get this check

19:27

if the pLayer is no physic layer, and the layer is a physic layer, render (edited)

19:27

wtf

Jupstar ✪

even tho i still don't get this check

I think it checks for everything that's not in a "predefined" layer

sry my iq is too low for this code

19:39

so it basically makes a copy of all layers

19:39

and checks against the loadedd version?

19:39

then the question would be, why would this ever be false

when the layer in question is GameLayer, FrontLayer etc

yeah, nvm editor code is weird xD

That condition almost makes me want to refactor it into a function

1

I'm pretty sure such a function already exists

i miss ryozuki where is the daily: with rust you could use this and that

Mr.Gh0s7

I am working on something that will probably use CLayerGroup::Render() should I merge those to for loops into one?

no, because we have no choice but to use 2 loops to keep rendering order

20:53

another way could be to sort layers once and then iterate over sorted layers

this

1

!c https://www.phoronix.com/news/Google-Chrome-Maglev-Compiler

Google Chrome Begins Rollout Of New "Maglev" Mid-Tier Compiler

With Chrome 114 is the start of Google beginning to roll-out Maglev as their new mid-tier compiler for further enhancing the JavaScript browser performance.

ChillerDragon

Firefox dead

[ddnet/ddnet] Pull request opened: #6706 Fix `color_cast` between `ColorHSLA` and `ColorHSVA` losing alpha, add test

Checklist

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

[ddnet/ddnet] Pull request opened: #6707 Fix UI element check with `m_HintCanChangePositionOrSize`

It was incorrectly checked for UIElement.Rect(0)->m_Y != pRect->h, so any UI rect using m_HintCanChangePositionOrSize would be updated every frame (although, no UI element currently uses this hint). Additionally, checks for changed rounding size and corners are added so UI elements are updated when those are changed.

Checklist

[X] Tested the change ingame
[ ] Provided screenshots if it is a visual change
[ ] Tested in combination with possibly related configuration opt...

[ddnet:master] 2 new commits

265f16e Fix UI element check with m_HintCanChangePositionOrSize - Robyt3 384a356 Merge #6707 - bors[bot]

22:51

[ddnet/ddnet] Pull request closed: #6707 Fix UI element check with `m_HintCanChangePositionOrSize`

[ddnet:master] 3 new commits

4032745 Fix color_cast between ColorHSLA and ColorHSVA losing alpha - Robyt3 85bfe19 Add test to ensure that color_casts keep the same alpha value - Robyt3 f178bdf Merge #6706 - bors[bot]

23:18

[ddnet/ddnet] Pull request closed: #6706 Fix `color_cast` between `ColorHSLA` and `ColorHSVA` losing alpha, add test

I'll continue working on it tommorow.

imagine having a huge ass map only to have to load it to preview so you can load it to edit

1

Mr.Gh0s7

I'll continue working on it tommorow.

oh rly nice feature

23:47

im a map hoarder so that would be helpful for me

Mr.Gh0s7

I'll continue working on it tommorow.

@Ravie we need an executice decision

Exported 546 message(s)