Teeworlds - bridge

Teeworlds

IRC / bridge

One-way IRC channel bridge. If you want to be able to send messages to IRC, contact @Dune or @heinrich5991. https://www.teeworlds.com/?page=docs&wiki=rules/irc_rules

Between 2020-02-10 00:00:00Z and 2020-02-11 00:00:00Z

ok, I posted that fucking fairness post in the forum, in general discussions.

05:17

sorry if that's the wrong place for it. :P

that's the right place :)

jewZeus: not bad, but your chiller server review forgot to mention the most vital site seeing attraction, the ultimate noob slayer -HS- Machine!

jewzeus: your proposal violates the unwriten rule of teeworlds game design: as little rng as possible.

15:27

players spawning with unknown/ for the player untrackable amounts of hp would imho fall into that category. Then again you don't know how much hp a player at any given time has anyways so maybe not :?

well, this is just a proposal of an idea of something that might be a useful step to try in terms of balancing, it's not intended to be 'the ultimate perfect solution', just the justification for a first step towards a solution, and a direction to step in

16:07

the "hide how much your hp has changed" thing, that was a last minute idea and one that I haven't thought through as much as the rest of it

i do not think that fairness is the issue of people not playing vanilla

when I've been thinking about this, I've been thinking back to playing "thief: The dark project", because when you changed the game's difficulty it would have your health icons in the same place but more or fewer of them depending on the difficulty, sort of stuff

16:08

I think it was actually 10 shield/heart things for the basic difficulty there too, almost like everybody finds that an intuitive number

16:09

ChillerDragon, yes exactly? this whole thing is about fairness in vanilla, not mods

vanilla is rather easy compared to gametypes

16:10

playing vanilla, not winning at vanilla

i can not think of any mod that helps noobs to fight pros

16:13

"My ideal use-case of this, even trying to break it, would be to be able to set up a lan party with my kids and some adult friends all playing"

this is possible in vailla

16:14

just spin up a lan party and play with a bunch of noobs no need to change the mechanics

16:14

skilled players should get more kills. And not tactically suicide to then collect hp and rush the flag with a op amount of hp and armor.

16:15

noobs should die fast and not struggle a long slow death while their 20/20 protection slowly shrinks

To me its just a Matter of active people

16:28

The less the players the less its fair, as more people play at the same time the more they will gather in “categorized” groups

ChillerDragon the point is to give both the kids and the adults a serious fighting chance, because I'm sure you've observed and probably nerfed yourself down too

16:29

the point is, the game has to be fun for everybody

Now with just one full server per day its obvious that both noobs and pros will want to play cause its the only one active.. and the crying starts

eh, if it were a balanced game, there would be more vanilla servers

16:31

*populated vanilla servers

16:32

please don't try to tell me that in a 1v1 where one player is getting 10 kills at a streak and the other flat out isn't, that there can't be some kind of balancing put in so that both people actually want to keep playing for a bit because it's challenging but also actually fun for both

16:32

the point is to do this kind of balancing intervention as little as possible

16:32

leave it as vanilla as possible

16:33

but actually balance it so that it stops this ragequit/borequit stuff for noobs or pros

16:37

I understand that the intent is to leave this as a side scrolling FPS as much as possible, and FPS's don't classically do this, but the speed at which some players because of their skill with the rocket jumps and the grappling shit, is faster than the weapons can actually shoot at them, other than the laser, which has a limited range and can't actually hit them by the time they are aimed at with human reaction speeds

16:38

as a result, much of the more intense fighting in the game at higher levels seem to consist of high velocity speculative grenade launchings, which new players are simply unequipped to participate in and need a way to weather while they learn to navigate

16:39

if they catch on fast? the mechanic stops helping them fast

16:40

even if people try to abuse this system, eg, advanced player kills an oponent and then commits suicide, that still robs their inventory and their time to get it and gives the noob a chance against them next time

16:41

I don't think the way to do this is 'every kill adjust every players hp in rank', but rather, only for advanced players on streaks of kills, or noobs on streaks of deaths

16:42

so yeah a pro could theoretically kill themselves 20 times and miss the first 1/2-2/3 of the game, to get a full 15/15 and do some special rocket jump, type thing

i dont think teeworlds vanilla should be the type of game that allows for a equal fight between adults and children

18:27

if you don't want your kids to cry then just play worse thats how to do it and how it has always been done on board games

18:30

you are new right JewZeus? I do not want to keep talking against you :/ Nice that you joined us and care about the game. I consider my self a conservative person and I would strongly fight against any change in teeworlds hehe.

18:35

Omg can we please get the old chat system :/

18:36

Back in the days grepping logs for errors that show up in the chat was easy nowerdays they are clientside only

18:36

Image attachment

18:37

Half server leaves the game because (Error sending data) and the looks is like https://external-content.duckduckgo.com/iu/?u=https%3A%2F%2Fstatic01.nyt.com%2Fimages%2F2016%2F08%2F05%2Fus%2F05onfire1_xp%2F05onfire1_xp-articleLarge-v2.jpg%3Fquality%3D75%26auto%3Dwebp%26disable%3Dupscale&f=1&nofb=1

18:37

logs not looks

child/adult is an example to illustrate newbie/pro

ye ikr

ChillerDragon, the point is to do this as minimally as possible, but some kind of measure like this clearly needs to exist, that whole ragequit/borequit thing isn't just kids vs adults or noobs vs pro, I don't think it's rocket science to say 'some matchups are just uneven and need some kind of intervention, ideally as small as possible, to make the game even enough for both to play"

21:18

the more I think about this, the more I think there should be a 'balancing threshold' ie 'you have to do a kill streak of x many kills' (or maybe even 'x many y times' type thing) as a minimum threshold to turn anything on, and then it should increment with the smallest steps possible that it can be effective with

21:19

like the server specifies 5 as the minimum threshold, that means the player has to get 5 kills in a streak without getting killed to have one increment of this, ie, lose 1 heart of total hp

Ye I get the idea

this doesn't seem unreasonable, I have certainly been killed 5 times in a row in a 1 on 1, and quit in frustration, because what the fuck is there for me to do in the game besides die at that point?

21:20

and I've also done that to other players and watched them quit too

But I agree with @Sonix balance should come with bigger amount of players and skill groups

that's the anthropic form of balancing

when there were more servers it was still newcomers getting rekt until they quit the game (or switch to more casual game modes)

that already exists but is insufficient especially since so many matches are so small and so skill-disparate

I would say we currently have 2 skill groups and it works okayish

eh which part works?

I agree with JewZeus sentiment but I don't find the proposed changes pretty / simple

what is being done that I'm not aware of

21:21

petty? lol this is going to bite my ass as much as anybody's

Most pros play on their servers and public servers are noobs

this isn't a "beef me up", I'd predict to be playing at 8 hearts most of the time on it

But pros also occasionally rip noobs that’s true

unfortunately this is an issue that plagues old games with no matchmaking, and is not specific to teeworlds :/

Ye but the playerbase is too small for matchmaking :/

yeah this is an attempt at matchmaking in terms of skill to difficulty of opponent

21:22

based on score

21:22

in-game data shit

21:22

no accounts, none of that

not enough vanilla players for matchmaking unfortunately

that's why this is my proposal rather than matchmaking :P

21:23

"oh I can maybe get 5 people on a good day to play when I want them to"

21:24

accounting for the fact, 2 of my own kids "you can play THIS game if you like" sort of "lets take a break from other games today" arm twisting, + their friends if I'm lucky

21:24

fortunately my kids like and are pretty competent at this, my daughter especially likes ddrace :P

21:24

this is a GOOD game

Wait you have kids? And you play tw with them?

I don't want to change that

21:25

yep

That’s nice o.0

it's a good game, I'll say it as many times as I need to

21:25

change as little as possible

21:25

but I'm serious about that 'adults vs kids' lan party as a literal case

21:25

"make this rock for birthday parties"

21:26

ideally even "make a version that runs on pis

Pls invite me thanks

21:26

Tw runs fine on pi

you've played against them, they've been on your server :D

21:26

eh tw SERVER runs fine on a pi

21:26

but not a playable client

Oh

not that I've gotten running

Isn’t pi 4 dual 4K 4core nasa rig?

yeah, I have pi 3's, lol

21:27

it can technically start the client, and get like 1 frame per second

Idk I need hilarious strong hardware to have acceptable teeworlds feeling

it seems simple enough that a 1.4 ghz quad core should be fine

What is your ingame name btw?

eh most recently "god of butts" :D

Oh your still in the Name searching Phase?

21:28

It took me almost a year I think

eh I have a rotation of trolly names (like this one too)

I recommend sticking with one so people won’t forget who you are

21:29

It improves social interaction

I don't change often, but "god of butts" was a more kid-friendly name than 'painus', so I'm sticking with that change

Can I ask how old you are?

I am Painis!

can someone pls plot a graph of irc actvity since the bridge was activated :D

Teeworlds is in my opinion not multithreaded enough to make use of multiple cores. At least the serversode code is completely single threaded, the main stuff I have seen so far.

yea the good old multithread topic

21:35

every tw source i touch gets unstable and crashes weirdley

21:36

i do not want to think about how it would be when there were more threads

21:36

but for sure it is a good future proof idea since the hardware seems to go the more cores way rather than more single core power

21:37

but maybe rather rewrite tw in rust then so its less likley we break everything :D

21:37

im sure @heinrich5991 would love to rewrite tw in rust hehe

that would be interesting.

i saw a basic functional 0.6 tw server by heinrich once

21:38

written in rust but it had not all features yet

Yeah, be basic concept of multithreading serverside is: "don't fucking block the main thread, never period."

could help against dos

21:39

iirc there were some multithreading patches in the block ddos war that seemed to be effective

I think teeworlds might need to use "unsafe rust". I'd really love to learn rust soon :D

it probably never will happen

21:41

i mean replacing decades of C/C++ code is crazy

https://github.com/heinrich5991/libtw2

heinrich5991/libtw2

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

I guess compared to other languages go and rust are the emerging languages to have great tooling.

ye we all know famous libtw2

poc 0.6 server in there

oh there it is

syntax looks as cryptic as c++

i hardly see any projects written in rust

21:42

sometimes go

at least it has generics

i have the feeling heinrich is the only person on planet who uses rust

x)

21:42

I know one highly capabke person that also loves rust.

ripgrep is also written in rust

21:43

firefox has important rust components

the security aspect is great

ye

21:43

i guess idk baut rust but i thought its stable af

21:43

like c without segfaults

21:43

or c++

perfect language for os development and microcontroller development

i used all kinds of grep but never ripgrep

there is ssl stuff rewritten in rust that has seemingly better performance than c/c++ implementation.

the ripgrep is also a much faster grep

21:46

https://blog.burntsushi.net/ripgrep/

ripgrep is faster than {grep, ag, git grep, ucg, pt, sift} - Andrew Ga...

I blog mostly about my own programming projects.

21:46

cool blog post about performance

kinda sad that one does not see many job opportunities for rust or go.

is that the only reason to go through the struggle of changing the tool?

21:46

i used egrep and pcreegeprep for better regex

performance is a good reason

but speed never seemed to be a bottleneck for me

if it is a bottleneck :D

21:47

like parsing terrabytes of log files?

rg also respects .gitignore etc.

i was about to write that

21:48

could be seen as feature

do you know any great beginner books in rust? or is the "necronomicon" good enough?

21:48

for rust

but if you dont know or forget about that it can be very annoying when you cant find stuff listed in the gitignore

alias rg=rg -u

21:49

now it only disregards binaries

imagine grepping for the definition of some tw network code which is generated and ignored

imagine using vs code search instead of grep x)

ok i would switch if it supports multiline search

21:50

does rg 'foo.\n.bar' work?

rg -U for multiline search IIRC

21:51

but why multiline?

twlogs

21:51

all the time i search something in tw i need multiline search

21:51

for example earlier today i had a username and a disconnect error message

21:52

thats split up in two lines

21:52

all the stuff is always split

21:53

r u telling me that im the only person who always searches strings split over multiple lines?

I've never done it before, I think

lol

21:54

i have to do it once a month min

21:55

either i go with some -C context heads and tails and double grepping crap or i use pcregrep

21:59

how differenct are the arguments?

21:59

re learning stuff is annoying

rg -C also for context

ye just realized it seems to be kinda similar

22:01

doesnt look to bad i might switch

22:01

but grep is pre installed everywhere huh?

yes

thats a big pro

is grep even coreutils?

22:02

bridges were a mistake, everyone should just connect to quakenet like a proper gamer

i can not remember installing grep anywhere

it's not like you'll unlearn grep

quakenet doesnt work for me

22:02

but when i pick rg as preference i will end up using it in my scripts and then my scripts will break on other systems

I'd use grep for scripts

yes

22:03

thats what i mean

rg is for interactive use

i do not want to distinguish interactive and script

22:03

because my scripts start interactive

then don't use rg I guess

idk

22:04

i will keep it in mind it will probably replace my pcregrep

22:05

so minus you would say quakenet irc is more gamer than discord? xd

retro gamer lol

hi from qnet :)

i wish i could join you

it's actually better because it needs less ram :^)

the place where the cool kids hang :D

22:07

well im sure there are more bloated irc clients than cordless discord client

22:07

for example irc over browser is a bloat

certainly

22:08

mine's just a terminal tho

cordless is also just terminal

"nice"

ikr

22:09

or are u being sarcastic? :D

22:09

idk what those quotes mean

inofficially using some internal api?

22:09

yes

ye

22:10

probably borderline legal

so it's only a question of time until it gets fucked over

wait internal?

22:10

nah public api

IRC lacks some modern features unfortunately. would gladly go js-free otherwise :p

oh?

22:10

didn't know discord had public apis

its opensource project that wraps around the public discord api

Dune: like what?

https://github.com/Bios-Marcel/cordless

Bios-Marcel/cordless

The Discord terminal client you never knew you wanted. - Bios-Marcel/cordless

minus, formatting standards

one of the few go projects i know

22:11

no rust :p

WARNING: Third party clients are discouraged and against the Discord TOS.

22:11

from that repo

ye borderline legal

22:11

:D

not borderline legal

well it's disallowed

but bannable by discord

22:11

legal is about being put into prison

hm

lol, codecoverage 13%

better than teeworlds

Dune: interesting; i haven't heard that complaint about irc before. fwiw it does have colors, italic, bold, underline

true xD

nobody but me writes tests :>

22:12

:< *

so cordless is even more gangster than irc cuz its agianst TOS

minus, is that in the standard? TIL

i was more expecting things like image upload, etc

go test are ultra easy to write

it lacks code formatting though

nah i don't think that's standardized, Dune

I dunno abouz C++ ._.

out of convenience, the need for a bouncer and uploads yeah

ah yeah code snippets is something else

@jxsl13 look into the tests, not hard to write IMO

how would you code format in a terminal based program @Dune ?

theoretically it also lacks UTF8 lol

22:13

@ChillerDragon syntax coloring

ah

idk if the encoding for irc is standardized, i don't think so

In regard to testing tw feels like trying noot to shoot into your foot every time you change something.

22:14

-o

yeah

22:14

it'd be nice if the IRC standards evolved

that's why you try to add tests

modern software development needs modern tooling :D for test.

for me its hard to unit test chunks of tw becuase i usually include literally every file to get all my global vars interacting with each other :D

wdym @jxsl13

22:16

I think LLVM uses gtest

oh btw any suggestions how i get the new Config() stuff in IClient?

heinrich5991, i think so yeah

I think around 2002 there was no such test driven development mwntality

22:16

mentality

I still don't like it :p

22:16

tests tend to lack flexibility

22:16

you change your code, you break the tests

22:16

meh

Kernel()->RequestInterface<IConfigManager>()->Values()

22:17

the same way you get it from anywhere, @ChillerDragon

in engine/client.h ?

@heinrich5991 isn't that expensive btw

you do that once, on initialization

is it even initialized anywhere?

well, I think a proper testing of individual component should be done. Integration test might be hard if not impossible to realize properly.

ChillerDragon: src/engine/client.h is not a code file

test-driven development is more than a few tests, jxsjsl13

it's an interface file

yes

you shouldn't put complex code there

and i need configs there

22:18

ddnet did

no, do it in src/engine/client/client.cpp

22:19

this is the wrong place

https://github.com/ddnet/ddnet/blob/master/src/engine/client.h#L81

DDraceNetwork, a cooperative racing mod of Teeworlds - ddnet/ddnet

I think rust also provides proper tooling for testing :0?

yes, it does

22:19

but you don't need it

22:19

gtest works. otherwise use some other testing framework

so you say that should be refactored?

22:19

the code looks good to me

infinite compile times already tell you your bugs :D?

you don't need a built-in testing framework in your language to do testing

22:20

is what I meant

22:20

ChillerDragon: that code looks horrible to me

yeah, I guss, but using external tools sucks imo.

why does that static accessor function return something different depending on some global config variable

22:21

@jxsl13 why?

external dependencies.

yes. gtest is stable

22:21

it doesn't have most of the downsides of external dependencies

could you please refactor that for me heinrich? <3

no

meh

@heinrich5991 I saw you wrote parts of the backends, rhetorically, is a complete rust rewrite a good idea or are some parts shitty?

try using the double <3<3, Chiller

any hint on how i could refactor it?

I would go for a gradual rewrite if I were trying to do it

ofc, but is that a good idea even

22:22

or does rust suck for some parts

e.g. first rewrite some low-level horrendously unsafe parts in rust

rewriting the masterserver would make sense

nah, http master server incoming™

22:23

some people would probably say that rust limits them too much

ha

Dune: ^

I think the browser protocol should be rewritten to tcp

gradual adjective: moving, changing, or developing by fine or often imperceptible degrees

the http master server looks like a lot more work than I thought it'd be

22:23

I'm surprised

monthly word teached by heinrich :D

@jxsl13 that's garbage for latency

22:24

that's what stops teeworlds from being properly implemented in a browser iirc?

not really

is it stopped?

@redix

there's a websocket port of teeworlds by eeeee

werent there multiple fully functional browser clients already?

you oculd also do teeworlds via webrtc

I thought they had lag?

not due to network

22:24

it's udp (webrtc)

oh ok

shouldnt webassembly be near native?

I think within 2x of native, for most parts

I think it's ~2

22:25

yeah

22:25

pretty good compared to anything else tho

Maybe gRPC

what's that?

I'm not into that topic tho

and in which direction would your gradual rewrite go? @heinrich5991

stackoverflow suggested it being reasonably better than udp

ChillerDragon: first some low-level stuff

22:27

maybe network, maybe file parsing, etc.

what?

22:27

what are you talking about?

rust rewrite

ah

22:27

oh i mean moving the config out of IClient

please redesign VarInt x)

its a somewhat accute problem for me

22:28

is accute even a word?

22:28

xd

acute is

calls the doctor

ah ye to many c

22:28

too

._. I don't like the VarInt design! please discuss

wdym?

it's a class without an internal state, basically functions

oh that one, yea

basically forcing oop where no oop is needed

probably left-over from the oop rewrite

sounds like a large part of oop code

also, it has buffer overflow problems

varint or do you mean Packer?

varint, not packer

why that Oo?

Pack(0,1)

22:32

hello there

hey rand

I guess heinrich5991 would like some size parameter or end pointer to Pack/Unpack

ah no, it seems it has been fixed

redix did that?

22:34

I guess

gradual rewrite

22:34

*ea6d380f7f1a7333db4470c60562b49827cfd13a

22:34

wrong clipboard

ye, redix

oh, Pack calls look safe

22:38

wait, is UnPack safely used ?

+1 for rewriting in rust ^^

ye, it assumes a max length of 5 bytes

22:40

and aborts

22:40

cannot soeak for every us case, tho.

redix but like client and server and stop maintaining the C++ version?

speak

22:41

some security critical stuff, leider ke masterserver, I'd goess

22:41

like

22:41

rip my fingers

are u on phone?

22:41

xd

si

beerlang autocomplete detected

this is confusing

22:42

because we don't know if you type 'si' or mistype 'no'

xD

a rust rewrite would be cool but a lot of work, so i doubt it will happen at all

lol

I'd love to learn some rust :o

rand: I'm willing to do partial rewrites in rust; but I think Oy spoke negatively about that, so I'm not really hyped

do you remember reasons?

only gradually as heinrich said

go rewrite.

like actually against rust or agianst the amount of work and duplicated source bases?

probably boils down to 'it's not c++'

oh

with partial rewrites you don't have duplicate sorces

A rust rewrite would also alienate the 5 or 6 developers left here, might need to consider that

base.c -> not c++

hi Learath :)

Learath2: maybe also get new ones ;)

Yeah I don't think most codebase changes are worth it at this point

tw uses python, C, C++, lua

and a complete rewrite might be a hurdle for mod developers

I'd say the energy is better spent writing features

rust would probably make it easier to spend time writing features

game's feature comolete with mostly minor tweaks and bot stuff left?

you have a built-in usable data structures, for example

pls be slices

wym built-in data structures? We aren't even allowed to use the C++ ones :P

and you have a usable package manager

slices exist in rust, but they're slightly different from go

heinrich5991, surely that would be outweighted by the time spent learning a new language / how to debug pitfalls of said new language

22:46

though I don't have anything against rust ;)

rust is basically the language designed against pitfalls

slices != std::vector?

that is, if it compiles

no need to debug or pitfall

22:46

ye

22:46

isnt that the reasom to switch to rust to avoid pitfalls and debugging

22:47

that could be seen as a pro for mod developers

there are still plenty of reasons to fail

well, it does not stop the need to debug

less instable mods

you also get easier multi-threading

and you have to set up new dev environment etc.

22:47

there's always a cost

yes, there's a cost

a lot

22:47

especially in the beginning

there's also the cost that new developers trying to get into teeworlds have right now

the build process is most likely less of a oain in the ass comoared to c++ :D?

switching to rust could be seen as a expensive investment in the future

yes. 'cargo build'

22:48

yes

no header files

22:49

infinite compile times.

in the order of c++, I think

22:49

a lot slower than go

Go is basically python in better, my current opinion

faster. better depends on the use case

i hardly know anything about go but i would never compare it with python lol

sidetrack detected

Go has a scriotin language like workflow

22:50

scripting

22:50

as it compiles so fast

22:51

also go run

22:51

like interpreter

that is obviously one aspect of a scripting language. you see that there are others, I guess? ^^

https://www.youtube.com/watch?v=k7oosn5HrKk

Go As a Scripting Language in Linux

well, atleast I wouldn't be coding much if we do make a move to rust, I just never got used to coding in rust

15 mins talk for hello world in go

22:51

best scripting language :D

jsxl13: don't you know about “tcc -run main.c” ?

no :(

oh tcc is nice

22:53

i used that back when i was in school on usb drive :D

chiller, well that talk makes go literally a scrioting language

no that talk shows how useless go is :D

is not :(

useless as scripting language

well, yeah in that regard maybe.

https://www.youtube.com/watch?v=bD5k0V08wRY

Bash Programmierung #1 "Hello World!" und nano - ITWorks4U

go run main.go

22:54

is what I meant

xxxxxxxxxxD

22:54

dont hate on bash heinrich

22:54

oh cmn u fucked me

also the deployment of gobis better than pythin :D

where did u get that hahaha

fat fingers

"hello world bash !youtube"

searched bash beginner ordered by length and views?

although I admit I didn't pick the first one

55 clicks xd

https://www.youtube.com/watch?v=xv5_-rEmJ-E

Introduction to Bash Scripting 02 - Hello World

22:55

that was the first one

22:55

2nd tutorial, already got to hello world

ur a big :troll:

22:56

no language can compete with bash

in regard to ripgrep the syntax is literally python regex

and in helloworld its beast

@jxsl13 you can have monolithic binaries for python, too

22:56

they're more annoying to do thought

22:56

cx_freeze

yeah, I already did use pyinstaller

eh, that's what I meant, maybe

on windows

why do ppl even use nano?

I used something like that to packa pyqt app

nit fun, but works

22:57

same also did pyqt

22:57

for some workflow automation

22:58

nano > vim

22:58

I'd love to have a parallel official rust project with hopefully lots of code comments.

22:59

that might slowly port some critical stuff, maybe

22:59

might get some rust devs interested in teeworlds

lol i tried '$ nano > vim' and it didnt end well :D

operating system coukd not handle the truth and crashed?

^

23:01

ok ima try the other way around

23:01

at least vim threw a warning :p

23:02

oh jesus its harder to kill

xD

23:02

nano the way to go

https://github.com/hakluke/how-to-exit-vim/blob/master/README.md

hakluke/how-to-exit-vim

Below are some simple methods for exiting vim. Contribute to hakluke/how-to-exit-vim development by creating an account on GitHub.

lool

xD

i created a file when i cat it it writes stuff to my terminal on mouse click

23:03

bois vim > nano is crazy compared to nano > vim

23:03

i can highly recommend trying it :D

I bet one way is: unplug your pc

it literally highjacks my mouse

https://github.com/hakluke/how-to-exit-vim/blob/master/README.md#the-hardware-way

after i killed the process

this is gold

vimkiller :D

the ine above

23:04

one

not that nano is easy to quit^

can't you use vim to do buffer overflows or forbidden accesses to memory

ctrl+x ?

idk about forbidden access to memory i use vim to support kids in uganda

23:06

:D

23:06

ctrl+x+y+enter

23:06

vs

23:06

:x

alias vim=/bin/false

dfq

23:09

alias fuck=sudo

23:09

fuck !!

23:09

runs last command as sudo :D

I would call the "Docker Way" the Overengineered way

23:09

there is a fuck tool

23:09

that corrects typos

lol didnt know

23:10

could you send a link my research only showed porn lol

https://github.com/nvbn/thefuck

https://github.com/nvbn/thefuck

Magnificent app which corrects your previous console command. - nvbn/thefuck

too slow

"https://github.com/nvbn/thefuck"

23:11

eh

23:11

"fuck bash cli"

>:D

first search returned it

idk how to search

that's one of the skills one truly needs

23:11

learn it

ye

yeah

my search literally returned 100% porn

23:12

"fuck tool typos"

that's basically the only true skill a programmer needs

or maybe its my biased search :D

maybe google knows your search history :0

lmao google

duckduck

23:13

aw, man

:)

23:13

but is it useful?

same story, different name

i mean is it faster to type fuck then look at the suggestion instead of uparrow and fixing it?

check out the gif in the readme

well duckduckgo shouldnt give me a biased search result as far as i know

dunno, I am not using it

google bubble

porn bubble

23:14

xD

:D

rust

23:15

back on track

https://www.factorio.com/blog/post/fff-333 could apply this kind of "background" rendering to layers in teeworlds

Friday Facts #333 - Terrain scrolling | Factorio

Hello, We released 0.18.4 this week, same old same old, more bugfixes, more bugs, more changes. At this stage of development, not many interesting things are happening, we are just polishing what we have. Minor terrain render optimization posila Just a couple days befo...

I think there are some heavily good people working on factorio :0 and their blog is super interesting

23:23

sadly don't understand it :'(

i think for tw it might be enough to cache the vertices on the gpu instead of a texture

won't that cost some vram though?

23:24

if gpu performance is a concern at this point, so is vram?

well not the whole map but only some chunks around the current view

chunks for maps could be a thing too huh?

performance is not a problem ingame. but in the editor

23:24

not for the maps itself but for rendering maybe

i have the feeling that very big maps can make trouble

23:25

iirc last time i tried making a big map it crashed clients

right now we need to calc all vertices per frame and send them to the gpu

23:26

we could instead cache the vertices for chunks

the background scrolling seems to help with weak and integrated cpus

23:28

that might fix the raspi issue JewZeus mentioned

anybody knows the state of the raspberry gpu drivers?

23:29

they should have enough power to run tw but maybe the driver is the problem

Exported 614 message(s)