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 2022-04-25 00:00:00Z and 2022-04-26 00:00:00Z

Jupstar whenever someone enjoys their frames untorn aquaTriggered

aquaTriggered

monkaS

07:47

learath when C is badmouthed blue_rage

blue_rage

07:49

Me trying to justify the existence of UB ^^

@heinrich5991 nah i had an continue before the recv code :D so it just was never run. It works now. Thanks for checking in with me <3

Any opinions on having some branch patterns ignored from CI? Like tmp_* or test-* so one can test and still push commits without getting email spammed

why do u even get emails from ci

If one of my pipelines fails i get a email from github. Which is actually nice imo.

i dont get why u always push all ur personal config onto upstream

08:41

it feels weird to me xd

08:41

e.g gitignore, u can just use git/info/exclude

ChillerDragon: You could set up email rules to instantly move them into trash folder.

08:43

Or maybe branches-ignore: works with *

btw the + thing exists on emails

08:44

youremail+something@domain.com

Yes, works: https://docs.github.com/en/actions/using-workflows/workflow-syntax-for-github-actions#onpushbranchestagsbranches-ignoretags-ignore

08:44

@Ryozuki that's not standard, I think gmail started it and some other providers copied it

rly?

08:45

sounds like a good feature

https://en.wikipedia.org/wiki/Email_address#Address_tags

An email address identifies an email box to which messages are delivered. While early messaging systems used a variety of formats for addressing, today, email addresses follow a set of specific rules originally standardized by the Internet Engineering Task Force (IETF) in the 1980s, and updated by RFC 5322 and 6854. The term email address in thi...

RFC 5233,[13] refers to this convention as sub-addressing, but it is also known as plus addressing, tagged addressing or mail extensions.

08:46

well looks like most email providers implement it

08:47

postfix too

[ddnet-discordbot:master] 2 new commits

6ffd2ff Add Mexican flag - def- 1b4d34e Fix players cache format - def-

yes deen it works thats why i am wondering if we should add it

[ddnet-discordbot:master] 1 new commit

8a3c9fe +x shell script - def-

[quakenet] ChillerDragon

yes deen it works thats why i am wondering if we should add it

Do we have any temporary branches like that? I think we shouldn't use them. Such experiments should be done in private forks, not on the main ddnet/ddnet

Yes i would use it in a private fork. But i do not want to edit the workflow in my fork for that.

10:29

But its not too dramatic just wondering if we could do it

If you find a second person that would use that, ok

10:38

Or use a very speaking name, for example *_noci

https://github.com/ddnet/ddnet/pull/4986#issuecomment-1099951400 @Jupstar ✪ no idea what that means. Where to put the comment? what even is OpenGL is that OpenGL1 ?

Free some opengl2 pointers by ChillerDragon · Pull Request #4986 · ...

Fixes these memory leaks on my system (default opengl 3.0.3 renderer debian 11) Direct leak of 72 byte(s) in 1 object(s) allocated from: #0 0x4eac0d in operator new(unsigned long) (/home/chille...

10:44

i can not reproduce additional memory leaks when switching to opengl1

[ddnet-scripts:master] 1 new commit

6fa651b Update teehistorian scripts - def-

since no one has ever noticed this bug: https://discord.com/channels/252358080522747904/757720336274948198/967069481279045633 should we just remove cl_predict_ddrace or should I fix the bug? (bug was added in May 2020 see https://discord.com/channels/252358080522747904/293493549758939136/967133955230928986) (edited)

I vote for remove

11:45

there are too many options for prediction/antiping

I would also prefer to remove cl_predict_ddrace.

[ddnet/ddnet] Pull request opened: #5018 fix dummy hammer interval, this became incorrect after a few days

Reported by Konsti here: https://discord.com/channels/252358080522747904/293493549758939136/967073772836376586 I have tested the change on Justfly with my dummy :D since it is exactly the same interval as before it should not change any Physics.

Checklist

[x] Tested the change ingame
[ ] Provided screenshots if it is a visual change
[ ] Tested in combination with possibly related configuration options
[ ] Written a unit test if it works standalone, system.c especially

-...

@Jupstar ✪ I can not reproduce https://github.com/libsdl-org/SDL/issues/5572 on debian 11 with vulkan or opengl i tried all fullscreen settings and tabbing in and out it all worked fine

Restoring from minimized fullscreen broken · Issue #5572 · libsdl-o...

apparently (git bisected) 8bae343 (1) and 0249df9 (2) (i hope i did the bisect correct, bcs the (1) issue i only discovered while bisecting but not sure if i can reproduce it on the main code) brea...

12:16

I tried latest SDL https://github.com/libsdl-org/SDL/commit/981e1e3c4489add5bf6d4df5415af3cf1ef2773d

Fixed logical size synchronization issue on macOS · libsdl-org/SDL@...

https://discourse.libsdl.org/t/sdl-2-0-22-prerelease/35306/6

[ddnet:master] 2 new commits

2f4ccb9 fix dummy hammer interval, this became incorrect after a few days - C0D3D3V 1dea02d Merge #5018 - bors[bot]

13:37

[ddnet/ddnet] Pull request closed: #5018 fix dummy hammer interval, this became incorrect after a few days

[quakenet] ChillerDragon

https://github.com/ddnet/ddnet/pull/4986#issuecomment-1099951400 @Jupstar ✪ no idea what that means. Where to put the comment? what even is OpenGL is that OpenGL1 ?

just put it in shutdown function

15:01

who moved #bugs over #questions FBI!

[quakenet] ChillerDragon

@Jupstar ✪ I can not reproduce https://github.com/libsdl-org/SDL/issues/5572 on debian 11 with vulkan or opengl i tried all fullscreen settings and tabbing in and out it all worked fine

gnome doesnt seem to be affected as ryan said

15:04

" but I can reproduce the alt-tab issue with Gnome, too." so at least one of the issues is reproducable on gnome

15:04

for him

[ddnet/ddnet] Pull request opened: #5019 remove cl_predict_ddrace

Since most of the predict code is only ddrace relevant anyway, I would suggest removing the cl_predict_ddrace option instead of checking if m_WorldConfig.m_PredictDDRace is set everywhere ddrace stuff is predicted. in case of conflicts with physics from other mods, the check on m_WorldConfig.m_PredictDDRace must however be implemented in the appropriate places (for example for the freeze and unfreeze functions). But since there are currently no conflicts apparently, this is sufficient for...

Jupstar ✪

who moved #bugs over #questions FBI!

Fixed. This happens super easily, I wish we could lock the channels somehow

moving a channel triggers the OCD of everyone

15:29

monkalaugh

[ddnet/ddnet] Issue opened: #5020 Can't launch the game after enabling Vulkan renderer.

[2022-04-25 18:13:57][client]: starting...
[2022-04-25 18:13:57][sdl]: SDL version 2.0.16 (compiled = 2.0.16)
[2022-04-25 18:13:57][gfx]: Created Vulkan 1.0 context.
WARNING: radv is not a conformant Vulkan implementation, testing use only.
[2022-04-25 18:13:58][assert]: /run/build/ddnet/src/engine/client/backend/vulkan/backend_vulkan.cpp(7077): Either use 1 main thread or at least 2 extra rendering threads.

I get this before the game closes by itself after trying to run it. Happens ...

@Jupstar ✪ ^^

15:31

lol

lol

15:33

how did he do that

[ddnet/ddnet] Pull request opened: #5021 fix clamp of m_ThreadCount

fixes #5020

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...

Ryozuki

moving a channel triggers the OCD of everyone

true xd

decltype(m_ThreadCount) returns the type of the variable?

15:44

nice to know

15:44

rust has this thing, where it can infer the type without u needed to type all this

15:44

u just need to use _

15:44

or nothing at all

I wonder if Rust can do this one where the types don't match

cant u rly not use clamp(m_ThreadCount)

15:46

without having to type the type?

15:46

m_ThreadCount = clamp<decltype(m_ThreadCount)>(m_ThreadCount, 3, std::max<decltype(m_ThreadCount)>(3, std::thread::hardware_concurrency()));

15:46

so verbose

Hm, I don't see why the type is needed in clamp

m_ThreadCount = m_ThreadCount.clamp(3, m_ThreadCount.max(hardware_concurrency()));

15:48

this would be

15:48

the rust equivalent

15:48

quickly typed

15:49

oh wait

15:49

the max is wrong probs

15:49

xd

Learath2

Hm, I don't see why the type is needed in clamp

i excplicity added it to make clear that the cast from int to size_t is correct

15:49

we have pretty low compiler warnings

15:49

but with stronger ones it would fail

15:50

also template deduction between int and size_t might fail to compile at all

15:50

3 is int

we have low compiler warnings?

while the other is size-t

15:51

we have pretty weak warnings yeah

15:51

we cant cast around as we want basically

one might thing -Wall means all warnings but c++ is special

15:51

monkalaugh

Weverything is best

15:51

just disable the few warnings u dont need

15:52

set(S "${S} -Weverything -stdlib=libc++")
set(S "${S} -Wno-c++20-compat -Wno-unused-command-line-argument")
set(S "${S} -Wno-cast-align -Wno-reserved-id-macro -Wno-weak-vtables -Wno-padded -Wno-format-nonliteral -Wno-documentation-unknown-command -Wno-extra-semi-stmt -Wno-switch-enum -Wno-missing-prototypes -Wno-covered-switch-default")
set(S "${S} -Wc++98-compat-extra-semi -Wnewline-eof -Wdouble-promotion -pedantic -pedantic-errors -Wmissing-variable-declarations -Wextra-semi -Wundef -Wunreachable-code-break")
set(S "${S} -Wno-gnu-anonymous-struct -Wno-nested-anon-types -Wno-old-style-cast -Wno-c++98-compat-pedantic")
set(S "${S} -Wno-reserved-identifier")

(edited)

15:52

this is what i usually use

Jupstar ✪

also template deduction between int and size_t might fail to compile at all

I actually am not sure about the rules when it comes to template subs here. I've never really looked into the technicalities of metaprogramming-magic

yeah if u dont add a template arg it has to deduct it based on the params

15:53

so the mismatch can create errors

15:55

i bet in c++20 u can do like auto(3) xd

15:55

just spam auto everywhere until it breaks xdd

awesome i didnt know Weverything is a thing! sounds cool!

16:03

i thought ddnet warnings are as strict as it gets :D

yeah u need clang tho

16:03

clang even has wdocument

clang masterrace

perfect for @Ryozuki

wats document

16:04

oh for docs?

it can parse doxygen

16:04

yes

or format?

16:04

a

16:04

col

16:04

but who needs docs when u got sourcecode?

who needs games if u have brain

who needs brain if you have drugs

16:05

wat?

fix ur pr

16:05

u added a space

oh

16:05

wetefek

copy & paste magician chillerdragon

fakof

16:06

fixxored

rust has

16:06

#![deny(missing_docs)]

Ouu i love cargo fmt btw xd

16:07

its so hot

xd

any daytraders from europe here?

16:09

i need financial advice

invest in amd

thanks

if nvidia really launches gtx 4090 with 700 watts, then its bcs amd does smth similar xd

16:10

rtx

don't invest in anything, refuse to take part in this corrupt neolib market structure

xd

16:10

tru story

wetefek learath

16:11

how to get piss rich then?

Learath2

don't invest in anything, refuse to take part in this corrupt neolib market structure

and lose money to inflation

just stop caring about money and enjoy your life

"you gotta play the system to beat it"

16:11

is that a saying?

16:11

xd

don't get piss rich, money is not important

yeah

but i need shelter and food

i enjoy rust lang

i enjoy ryo os

BASEDHALT

as long as you have enough to enjoy the things you like then you don't need more

i do not have enought to buy shelter and food when i quit my job

I didn't say quit your job, did I?

you sayid enjoy life

16:13

i dont see how working everyday of my life is enjoying life

xd

16:13

then quit it if u got the money

16:13

u rich we all know that

yes

open secret

so i need stoks

16:14

to get rich

16:14

then quit job

16:14

-> amd

u cant fool me

16:14

im detective

?

rich af boi

i swear im not

don't get rich, find work you enjoy, work gives your life meaning

ok sure

xd

where?

preferably on earth

freeze ur body and wait 100 years

i feel like losing money on the stonks market is more fun than applying for jobs

then u dont need to work anymore xd

i feel like tw wont survive 100 yrs

wat?

no point in having free time if there is no tw

what u talkn bout bro (edited)

taiwan

16:16

clima crisis will swallow taiwan

16:16

wat u dont get

ppl still play all these weird commendore 64 games

16:17

and u tell me tw wont survive 100 years

oh ur optemistic

ofc

nice

i will live around 60 years from now on

16:17

so 60 years new backends

16:17

xd

nah

16:17

u rq next year

16:18

when clubpenguin source code gets leaked

quite possible xd

oy will stay for another 20 yrs

yes, but will only join 20 times in these 20 years

deen gets pregnant this fall

16:18

and china will delete the internet

well the chance deen gets a child soon is quite huge actually

europa gets stuck in 0.6 for 10 yrs

16:19

and then around year 2080 we have 100 players max

16:19

and those players all use VR gear to play tw

do u bet your amd stocks against it?

but playing 2d games in VR will be very retro

16:20

yes i would short teeworlds

16:20

eventho it hurts me

[ddnet/ddnet] Pull request opened: #5022 Use std::clamp

Not sure if we want to, there are some easy functions like this we could take from stdlib now

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 ...

@Learath2 wat if trading stoks is the job i enjoy?

gigachad

2

Ew

xd

[ddnet:master] 2 new commits

d364c63 Free some opengl2 pointers - ChillerDragon 9e22700 Merge #4986 - bors[bot]

16:32

[ddnet/ddnet] Pull request closed: #4986 Free some opengl2 pointers

[ddnet/ddnet] Pull request opened: #5023 fix WalkTime for negativ positions

I am not sure if there is a more elegant solution, the problem is that fmod in cpp does not return a positive value. maybe we can just use a mod function that always returns positive values.

Checklist

[x] Tested the change ingame
[ ] Provided screenshots if it is a visual change
[ ] Tested in combination with possibly related configuration options
[ ] Written a unit test if it works standalone, system.c especially
[x] Considered possible null pointers and out of bounds ...

16:55

[ddnet:master] 2 new commits

428c1ca fix clamp of m_ThreadCount - Jupeyy 6270b1c Merge #5021 - bors[bot]

16:55

[ddnet/ddnet] Issue closed: #5020 Can't launch the game after enabling Vulkan renderer.

16:55

[ddnet/ddnet] Pull request closed: #5021 fix clamp of m_ThreadCount

@moderator Absolwit blocking on ger2 8308

Hey I am learning C++ so that I can stop writing C code in a C++ codebase cause apparently that's bad (and the more I learn the more it seems true). Question is why does the client code use C-style strings? Aren't C++ strings 'better'? Is it just that no one has gotten around to converting C-style strings to C++ strings or is there some other reason? It seems weird to me because C's strings are probably its worst feature

Cuz c strings are used everywhere

18:14

So its a lot of effort to port it all

18:14

Not worty, or is it?

18:14

Worth*

18:15

If its for me i would use string and string views

I don't think I'm qualified yet to know how worth it is xd

18:16

So there's no problem other than the volume of C-style strings?

18:16

No other catch?

For me? No

c++ has no flat strings

18:16

so they are heap strings

I would even make it utf8 only like rust

Don't desktop programs relly heavily on the heap anyway?

use boost static string xd

@Jupstar ✪ is it possible to make a buffer and string view it?

stack is still faster tho

Idk i never tried c++

Ryozuki

@Jupstar ✪ is it possible to make a buffer and string view it?

string views are just for viewing

Xd

its simply bcs c++ standard is slow af

18:18

we need boost

18:18

just accept it everyone xd

what's boost xdd

an alternative std

18:20

well

18:20

probs more like a patch over std

18:20

xd

So is Bjarne lying to me or what? He talks about std like it's crack cocaine xd

well the standard also adds language features

18:25

while boost is basically just advanced STL

18:25

so its not 100% comparable

18:25

but boost adds nice to haves

If only it's niceties didn't come at insane compilation times

18:33

@Cipy29 what do you think is bad about C strings that isn't shared by C++?

Learath2

If only it's niceties didn't come at insane compilation times

where does this information even come from, heinrich said the same

18:34

boost is relativly modular

Jupstar ✪

where does this information even come from, heinrich said the same

It's what everyone has been saying for the last 2 decades. I haven't verified it myself in almost a decade but it shouldn't be too difficult to find sources on the claim

mh ok but pretty much all STL is slow compared to c code

18:35

its simply bcs templates arent the fastest

18:36

all TL*

18:36

when i see how fast SDL2 compiles with 400 files xD

18:37

but templates are best

18:37

xd

For now I really like how strings play with vectors, how I can iterate over words in a string easily, or split the string into words more easily (instead of the C way of using the 'not-good way' of strtok), how I can do str1 == str2 instead of strcmp

I guess we could look into it if we want a concrete answer, but heinrich mentioned factorio ripped it out aswell and for the same reason. I really doubt anything changed since I last researched a decade ago

I'm reading a big C++ book tho and I'm just 100 pages in or so, so I expect I'll discover more good things about std::string

Cipy29

For now I really like how strings play with vectors, how I can iterate over words in a string easily, or split the string into words more easily (instead of the C way of using the 'not-good way' of strtok), how I can do str1 == str2 instead of strcmp

Ah, the convenience stuff. C++ strings indeed are much more ergonomic than C style strings

18:39

However, they pretty much share all the performance downsides, which is why I was a bit confused

And I didn't even mention the null termination of C-style strings xd

C++ strings are also null terminated

Why? Is it so they fit better to legacy C code?

18:40

they have .size() so I don't see other reason

Cipy29

Why? Is it so they fit better to legacy C code?

Yep, and they were always like that, C++ almost never breaks backwards compatibility

good guy C++ poggers

poggers

Fwiw I find null terminated strings much easier to work with on the lower level. I'd guess that and the extra space needed to store the length is the reason C ended up with null terminated strings

strings on low level? like what kind of application?

I think I also read that ritchie was concerned about the size of the length variable limiting future systems ability to handle larger strings

Cipy29

strings on low level? like what kind of application?

Like actually implementing algorithms that operate on them rather than using primitives made by other people. E.g. writing your own append, or working with unicode without a library

well yea but who does that xd

18:49

those things are already written

18:49

you only do them if you want to know how they work

They didn't write themselves lol. If everyone thought like that we'd never have new algorithms

18:52

Dragon4 already solved float to string conversion. If everyone just decided that was good enough we'd never have Ryu

Cipy29

And I didn't even mention the null termination of C-style strings xd

null termination itself is nothing bad, that one byte extra wont make an insane diff, std::string still saves the string size other than c strings

well those algorithms such as append or working with unicode aren't new

18:54

and coming up with new ideas doesn't depend on re-creating all old ideas

Jupstar ✪

null termination itself is nothing bad, that one byte extra wont make an insane diff, std::string still saves the string size other than c strings

saving the size together with the string 'feels' better

New ideas are built on top of old ones. There is immense value in doing these things yourself, especially if you are interested in things like embedded programming

so you don't need smth like strlen() to run when u need a string

18:55

i know haha, i work in embedded C programming actually xd

18:56

all im saying is that C strings are inferior to C++ strings

18:56

(just started learning C++ tho)

Literally any languages strings are probably better equipped than C strings. But heap allocation for every string is very awful for performance

18:58

So awful that even with modern arena allocators, every C++ compiler does small string optimization to keep them in stack

so for example switching from C strings to C++ strings in the whole client codebase would have impact performance?

18:59

like noticeably?

yes

18:59

heaps

like how bad could it be? xd (im asking cause i actually don't know)

but static strings are probs faster

19:01

and streaming is faster than printf

@Cipy29 learn c++ and then learn rust

19:12

the best way to know rust greatness

19:12

its like eating the least food u like first

19:12

to enjoy some good shit at the end

19:12

BASED

19:12

BASEDDEPT

xd so why's Rust better than cpp

learn cpp first

Bjarne seems to know his shit xd

haha

19:14

@Cipy29 did u learn about pointers yet

19:14

well u should use references

19:14

most of the time

19:14

in c++

safe rust has potential

I know about pointers yea

do you know about allocating memory on the heap

ye

in c its malloc in c++ its new

19:15

where are u at

19:15

whats ur progress

i know c pretty well, in cpp im reading this huge freaking book Programming Principles and Practice using C++

19:16

and im on page 120/1k+

19:16

chapter 4.smth

oh reading a book

19:16

well have fun

u have better way?

19:16

xd

Do you think its good? I am considering picking up a cpp book

i like this book yea

Cipy29

u have better way?

i find most books boring, i enjoy programming by doing stuff

i picked it up for free from google tho

ddnet is such stuff

19:17

but each has their own way to learn

Oh lol, you're just reading it from the pdf?

yea

Ryozuki

ddnet is such stuff

i wanna see what's the "most principled cpp" way first before xd

19:18

i also listened to Bjarne's talks on yt and he seemed to know what's up xdd

Ryozuki

i find most books boring, i enjoy programming by doing stuff

only sane way tbh

19:18

u read books and forget 99%

well i'm not only reading

19:19

also doing stuff ofc, only did small programs tho in c++ for now

19:19

like printing out the frequency of each word from stdin

@Jupstar ✪ i would only read books for truly complicated stuff

19:20

like cryptography

19:20

https://joyofcryptography.com/

The Joy of Cryptography

A free undergraduate-level textbook that introduces students to the fundamentals of provable security

19:21

i also have another

19:21

about perfomance

19:21

if only i could find it

19:21

https://en.algorithmica.org/hpc/

19:21

this

19:21

this one is a really good source

19:21

for high perfomance computing

Ryozuki

@Jupstar ✪ i would only read books for truly complicated stuff

ok thats more math than coding imo

19:22

i mean when do u implement a crypto lib yourself xD

Cipy29

like printing out the frequency of each word from stdin

https://www.codewars.com/

Codewars - Achieve mastery through coding practice and developer me...

Coding practice for all programming levels – Join a community of over 3 million developers and improve your coding skills in over 55 programming languages!

19:22

do problems from here

19:22

good website

19:23

user driven

Jupstar ✪

i mean when do u implement a crypto lib yourself xD

monkalaugh

19:23

feel free to add me as friend https://www.codewars.com/users/Ryozuki

Are those different from just leetcode algorithm memorization tests?

i guess

19:24

they are made by users

19:24

from that website

Ryozuki

https://www.codewars.com/

is this kinda like hackerrank.com idk if you heard of it

u can also make problems

19:25

codewars is better than those imho

19:25

more friendly

Ryozuki

feel free to add me as friend https://www.codewars.com/users/Ryozuki

ah nice

19:25

i love battles

also i love the japanese aethetic

i liked doing stuff from hackerrank and https://projecteuler.net/

About - Project Euler

A website dedicated to the fascinating world of mathematics and programming

but not today i have headache

I find books to be the only way to learn something in-depth. Mostly because other resources are usually very lacking. Especially the case with C and C++

projecteuler is more math-y but it's nice

Learath2

I find books to be the only way to learn something in-depth. Mostly because other resources are usually very lacking. Especially the case with C and C++

yeah u probs can only find the obscure things of c++ in books

High quality modern learning resources just don't exist for these dinasaur languages

19:26

Dinosaur*

specially c++

19:26

c++ is so complicated

19:26

its insane

19:27

c is simple

19:27

well

19:27

ub makes it not so simple

C has few quirks. C++ is just one big quirk

tbh i find all languages kinda hard, depends really what u do

yeah

when i use bind in javascript and have no template placeholders its hard

because in programming, simplicity makes things complicated

19:31

nice

19:31

i invented a quote

19:31

ill save it

nice

19:31

now get fame

19:31

and ppl will quote u in 200 years

aPES3_ChadShirtRip

but srsly sometimes i find simple library so insanly hard xD bcs they are not always very intuitive even for simple problems

sometimes u cant have simple stuff

19:32

because technology is complicated

xd

xd

thats probs why x86 sucks

19:33

bcs its so bloated and complicated xD

true

19:33

invent new arch pls

yes pls

https://xkcd.com/927/

xkcd: Standards

https://doc.rust-lang.org/book/

The Rust Programming Language - The Rust Programming Language

19:35

this is a pretty neat book

19:35

BASED

ctrl+T -> ctrl+F "rust" 'no matches' -> ctrl+W

Cipy29

https://xkcd.com/927/

or u have java with 10000000000000000000000000000 ways to solve the same problem xd

19:37

just that every new version is 0.0001% faster than the one before

19:37

and gets hyped xD

xd

19:37

@Jupstar ✪ there are some perf improvements

19:38

with new gc research

19:38

better paralelization

yeah java is decent

19:38

python or some other stuff is slower, and what do they do to fix it?

19:38

write c libraries

https://javamana.com/2021/01/20210115131703791G.html

What is ZGC, a Java garbage collector that claims to kill STW for 1...

ZGC Introduce ZGC（The Z Garbage Collector） yes JDK 11 Launched in th

and bind them into the langauge xD

The Z Garbage Collector (ZGC) is a scalable low latency garbage collector. ZGC performs all expensive work concurrently, without stopping the execution of application threads for more than 10ms, which makes is suitable for applications which require low latency and/or use a very large heap (multi-terabytes).

19:39

10ms tho

19:39

xd

19:39

just use rust 4head

technology actually so complicated

19:40

i wanna be in a world without computers feelsbadman

feelsbadman

@Jupstar ✪ do u think its true cpus are developed with C in mind

ofc

sad

i mean what purpose would rings have if software is 100% proven

19:41

there is no security issue if there is no untrusty software

monkalaugh

but tbf i cannot say when the time will come where quality stands over quantity again

19:42

and developing in higher lvl languages is simply often faster

19:42

and making everything a string xdddd i swear when i am in javascript everything suddenly a string

xd

Jupstar ✪

and making everything a string xdddd i swear when i am in javascript everything suddenly a string

https://wtfjs.com/

19:45

    typeof NaN === 'number' // true
    Infinity === 1/0        // true
    0.1 + 0.2 === 0.3       // false

the last one is a classic about floating point

the 2nd one is completly logic & true

Ryozuki

invent new arch pls

RISCV revolution when?

it has already begun for small devices in IOT

23:31

+ Linux support

Exported 447 message(s)