00:31 < bridge> <cyberfighter> hell no how much fps would i have on that crap
00:31 < bridge> <cyberfighter> seconds per frame*
01:02 < bridge> <Mr.Gh0s7> Now that I have a little more experience with programming I thought about giving rust another shot xDD. It went pretty well, I found fasterthanli again started to read his day 1 2022 AoC and well... this is what I ended up thinking.
01:02 < bridge> <Mr.Gh0s7> https://cdn.discordapp.com/attachments/293493549758939136/1175949264560599090/CodersVsProgrammers.png?ex=656d1718&is=655aa218&hm=7d10a8a80d95d015bed7d7933d120bde474378429de16dd29d11717d738eed3c&
01:07 < bridge> <Mr.Gh0s7> I don't know how to make rust feel like I am programming :/. Maybe it's skill issue but rust feels like it's for coders
01:07 < bridge> <Mr.Gh0s7> (I myself make the distiction from programmers like this: 
01:07 < bridge> <Mr.Gh0s7> Programmer is someone who thinks and then researches and tries to make code work to his desires (engineer of a sort). Coder is someone who researches and then writes code but doesn't necessarily understands what and why it's working but is happy :happy: ).
01:08 < bridge> <Mr.Gh0s7> Well I guess I'll stick with learning OpenGL and C. That seems to work for me :)
01:54 < bridge> <TsFreddie> https://youtu.be/xaMkMmHGjpI?si=WOtuMPw8NFBH3_M_
05:02 < bridge> <_voxeldoesart> whoever made this meme does not know how this meme format works
05:02 < bridge> <_voxeldoesart> rust == c?
05:29 < bridge> <ocacam> join baby ❤️
05:29 < bridge> <ocacam> 
05:29 < bridge> <ocacam> https://discord.gg/sexygood
05:29 < bridge> <ocacam> 
05:29 < bridge> <ocacam> @everyone
08:27 < bridge> <ryozuki> i dont understand this take
08:27 < bridge> <ryozuki> it just seems like u dont understand rust and pass it as rust coders not knowing what they do? weird take
08:52 < bridge> <teero777> Coding is the writing part. Programming is researching and then coding it up
08:55 < bridge> <ryozuki> yeah, in C you can be a coder easily, and do it wrong outright in some places, e.g memory safetiness, in Rust, to be a coder you need to pass some programming skills, which the compiler aids you, mainly understanding why some code wont work, because it is not memory safe
08:56 < bridge> <ryozuki> it puts the mental burden on you at the start, rather than at the end when ur program breaks and u need to debug and understand why
08:56 < bridge> <ryozuki> which is why u are giving up early
09:40 < bridge> <Jupstar ✪> i'd say if you never really programmed before c is not the worst language to start with. for the simple reason that it's easy to shoot yourself, which makes you understand what higher abstractions do for you
11:19 < bridge> <ryozuki> just use assembly like me
11:19 < bridge> <ryozuki> :NekoEvil:
11:19 < bridge> <ryozuki> but rust errors teach you a lot on why you cant do stuff
11:20 < bridge> <ryozuki> well i guess coming from C it makes these errors have way more sense
11:20 < bridge> <ryozuki> oh i need to update gentoo guru ddnet
11:21 < bridge> <ryozuki> looks like ill be the first distro
11:21 < bridge> <ryozuki> again
11:21 < bridge> <ryozuki> :BASEDDEPT:
11:42 < bridge> <Jupstar ✪> yeah, i'd say learning rust is indeed very useful too, bcs it's still pretty low level, even if it has many high level concepts. e.g. the borrow checker makes it defs clearer what references are than e.g. Java (IMO)
12:10 < bridge> <ryozuki> $rax   : 0xfffffffffffffff7
12:10 < bridge> <ryozuki> my mmap call returns this void* xD
12:20 < bridge> <ryozuki> oh i was using rcx instead of r10
12:21 < bridge> <アリヂン> People say that GC is a minor tradeoff for automatic memory management, and you might think the same until you realize your C\# app does not clean memory until your swap is 7.9/8 and ram 15/16gb
12:23 < bridge> <アリヂン> GC is cool but people have to stop relying on it so much, personally C\# was one of my first languages and when i decided to dig deeper and started learning C++ and rust i just couldn't wrap my head around it, i got too used to automatic cleanup
12:29 < bridge> <learath2> That's not how I ever thought about the words, but that is exactly how rust makes me a feel. When I write code in Rust or any other newfangled language I feel more like a stenographer than a programmer
12:32 < bridge> <learath2> Memory being managed for you automagically (whether statically or through runtime gc) is easy to get addicted to, which is why I'm a firm advocate of C first. Learn your C then you can go learn whatever other new language with cool space age features you like
12:43 < bridge> <ryozuki> @learath2 do u know if linux syscalls (not using libc) require the stack pointer aligned to 16 too?
12:43 < bridge> <learath2> Not sure
12:44 < bridge> <ryozuki> i think if you dont know how to do manual memory management you dont really know programming or computers
12:45 < bridge> <ryozuki> but well my bar is set to high
12:45 < bridge> <ryozuki> xd
12:48 < bridge> <ryozuki> >  The Linux Kernel does not require stack alignment. i.e. syscalls don't.
12:48 < bridge> <ryozuki> > 
12:48 < bridge> <ryozuki> > However, shared libraries do, since they follow the stupid AMD ABI.
12:48 < bridge> <ryozuki> @learath2 why does the abi require the stack aligned to 16, isnt it more natural if it is aligned to 8? then u dont need to align the stack between calls if u dont modify the stack
12:49 < bridge> <ryozuki> call pushes the 8 byte return addr to the stack
12:49 < bridge> <ryozuki> so the requirement of 16 is stupid
12:49 < bridge> <ryozuki> u will alwyas need to align
12:49 < bridge> <ryozuki> i guess it was done because u always push rbp to setup the stack frame?=
12:49 < bridge> <ryozuki> xd
12:49 < bridge> <ryozuki> but im sure the 16byte thing is before the 64bit era
12:50 < bridge> <ryozuki> @chairn maybe u know more about this
12:50 < bridge> <ryozuki> > But my take on this is to just follow the ABI alignment requirement to avoid future problems. I have this hunch that the alignment requirement is actually meant for syscalls that make use of SSE instructions. I don't see any other practical reasons for enforcing aligned stack other than SSE. The alignment is not actually for the function's code, but for use by the SSE save and restore inside the red zone of one particular syscall. 128 bytes is en
12:50 < bridge> <ryozuki> https://board.flatassembler.net/topic.php?t=19874
12:51 < bridge> <ryozuki> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=53383
12:52 < bridge> <ryozuki> >  H. Peter Anvin 2012-05-17 00:12:57 UTC
12:52 < bridge> <ryozuki> > 
12:52 < bridge> <ryozuki> > In the Linux kernel, we do not use SSE or floating-point of any kind (except in very particular highly controlled places); furthermore, stack space is at an extreme premium.  As such, it makes absolutely no sense to align the stack to 16 bytes in x86-64.
12:52 < bridge> <ryozuki> > 
12:52 < bridge> <ryozuki> > gcc, however, refuses to allow setting the stack alignment to 8 bytes:
12:52 < bridge> <ryozuki> > 
12:52 < bridge> <ryozuki> > : anacreon 105 ; gcc -mpreferred-stack-boundary=3 -mincoming-stack-boundary=3 -O2 -mno-sse -c /tmp/foo.c
12:52 < bridge> <ryozuki> > /tmp/foo.c:1:0: error: -mpreferred-stack-boundary=3 is not between 4 and 12
12:52 < bridge> <ryozuki> > /tmp/foo.c:1:0: error: -mincoming-stack-boundary=3 is not between 4 and 12
12:52 < bridge> <ryozuki> > 
12:52 < bridge> <ryozuki> > This is particularly ironic since we have found out that, in fact, ALL kernel stacks are actually 16-byte misaligned on x86-64 as the entry code pushes a total of 88 bytes onto the stack.  As such, gcc trying to keep a 16-byte alignment actually makes the code do worse, not better.
12:52 < bridge> <ryozuki> interesting loophole xd
12:55 < bridge> <learath2> Hm interesting. Let me have a think on why the abi requires 16
12:55 < bridge> <ryozuki> its most probs for SSE code
12:57 < bridge> <ryozuki> ```x86asm
12:57 < bridge> <ryozuki> %ifndef INCLUDE_ALLOC
12:57 < bridge> <ryozuki> %define INCLUDE_ALLOC
12:57 < bridge> <ryozuki> 
12:57 < bridge> <ryozuki> %include "lib/syscalls.asm"
12:57 < bridge> <ryozuki> 
12:57 < bridge> <ryozuki> %define PROT_READ       0x1
12:57 < bridge> <ryozuki> %define PROT_WRITE      0x2
12:57 < bridge> <ryozuki> %define MAP_PRIVATE     0x02
12:57 < bridge> <ryozuki> %define MAP_ANONYMOUS   0x20
12:57 < bridge> <ryozuki> 
12:57 < bridge> <ryozuki>     section .bss
12:57 < bridge> <ryozuki> 
12:57 < bridge> <ryozuki>     section .text
12:57 < bridge> <ryozuki> 
12:57 < bridge> <ryozuki> ; void* alloc(size_t bytes)
12:57 < bridge> <ryozuki> ; returns a ptr to the allocated bytes on rax
12:57 < bridge> <ryozuki> alloc:
12:57 < bridge> <ryozuki>     mov rsi, rdi ; rdi has the bytes
12:57 < bridge> <ryozuki>     xor rdi, rdi ; addr 0
12:57 < bridge> <ryozuki>     mov rdx, PROT_READ|PROT_WRITE
12:57 < bridge> <ryozuki>     mov r10, MAP_PRIVATE|MAP_ANONYMOUS
12:57 < bridge> <ryozuki>     mov r8, -1
12:57 < bridge> <ryozuki>     xor r9, r9
12:57 < bridge> <ryozuki>     call mmap
12:57 < bridge> <ryozuki>     ret
12:57 < bridge> <ryozuki> 
12:57 < bridge> <ryozuki> %endif
12:57 < bridge> <ryozuki> 
12:57 < bridge> <ryozuki> ```
12:58 < bridge> <ryozuki> i switched to mmap
12:58 < bridge> <ryozuki> xd
12:58 < bridge> <ryozuki> now i just need to do a small optimization
12:58 < bridge> <ryozuki> to allocate in bigger batches
12:59 < bridge> <learath2> Hm, I guess it's common enough that there is an sse array at the start and 8 bytes of waste isn't that important to get common code for restoring rsp
13:01 < bridge> <ryozuki> anyway just making this stuff i learnt a lot
13:01 < bridge> <ryozuki> stack frames and such
13:03 < bridge> <learath2> When will you code an arena allocator in asm? 😄
13:16 < bridge> <ryozuki> xd
13:17 < bridge> <ryozuki> @learath2 thats what i was thinking of doing
13:17 < bridge> <ryozuki> a bump allocator
13:40 < bridge> <stormaxd> Doped
13:40 < bridge> <stormaxd> https://cdn.discordapp.com/attachments/293493549758939136/1176139975864090704/image.png?ex=656dc8b5&is=655b53b5&hm=9dc00832a4ad55a66bd34bccae4d95bbee770dd78695aa0ba43be46ebd6d2a9b&
13:40 < bridge> <stormaxd> https://cdn.discordapp.com/attachments/293493549758939136/1176139976430333962/image.png?ex=656dc8b6&is=655b53b6&hm=98ae599109273378bf05bcc7a2aa5beeefb3657eb565c2468aaa6a91eb5fb85a&
13:40 < bridge> <stormaxd> https://cdn.discordapp.com/attachments/293493549758939136/1176139977021726790/image.png?ex=656dc8b6&is=655b53b6&hm=a45829d3de34d5eba3e1cf8d7f8aff330736b3d9fe0792f7288fbf9be77b0ff8&
13:57 < bridge> <Mr.Gh0s7> It was I.
13:59 < bridge> <Mr.Gh0s7> hmm is Chiller here?
14:00 < bridge> <Mr.Gh0s7> :monkalaugh: I wanna paste a code snippet of rust and then of C
14:28 < bridge> <logik_1499> hi after update i have so many micro lag how its fix fps is stable
14:28 < bridge> <logik_1499> ping too
14:30 < bridge> <Jupstar ✪> `shutdown /r /p` (in cmd)
14:30 < bridge> <Jupstar ✪> Else test old version of it's really the client
14:30 < bridge> <アリヂン> Hmm, I also had this issue after the update, but seems like it resolved itself after i switched backend to vulkan (it switched to opengl on its own for some reason)
14:35 < bridge> <logik_1499> it helped a little but the lags did not disappear
14:39 < bridge> <Jupstar ✪> Test https://ddnet.org/downloads/DDNet-17.4-win64.zip
14:39 < bridge> <Jupstar ✪> UPS
14:39 < bridge> <Jupstar ✪> https://ddnet.org/downloads/DDNet-17.3-win64.zip
14:39 < bridge> <Jupstar ✪> Test
14:39 < bridge> <Jupstar ✪> This is the previous version
14:40 < bridge> <ryozuki> @jupeyy_keks did u change smth?
14:40 < bridge> <ryozuki> why so many ppl complain
14:41 < bridge> <ryozuki> :justatest:
14:41 < bridge> <logik_1499> i
14:41 < bridge> <logik_1499> I tried the lags did not pass
14:50 < bridge> <Jupstar ✪> As you can see i didn't change anything. It's Windows always annoying^^
14:50 < bridge> <Jupstar ✪> That's sad but then it's either Internet lags or you changed something in your system
14:56 < bridge> <logik_1499> most likely it's the internet because it's been crashing lately
15:22 < bridge> <.krain.> if any dev want a feature to develop:
15:22 < bridge> <.krain.> It would be very nice in practice mod, that spikes execute /r instead of kill
15:28 < bridge> <.krain.> ```
15:28 < bridge> <.krain.> [Client] Don't disable practice mode on death [furo321]
15:28 < bridge> <.krain.> ```
15:28 < bridge> <.krain.> is it what I just said? x)
15:37 < bridge> <furo321> Not exactly, it won't automatically do `/r` for you. But if you die by a spike (and haven't touched the start line) you will still be in practice. Then once you died, you could run `/lasttp` to get back to your previous spot you used `/tp` to.
15:42 < bridge> <devinci24> But isn't that not only overcomplicating it + not even solving the issue either? Will be far from as smooth as just running /r.
15:42 < bridge> <devinci24> Or is there some sort of complication code wise that wouldn't let that happen?
15:43 < bridge> <.krain.> So we can die, then /r ? :o
15:43 < bridge> <.krain.> oh yeh we can't cauz we hit ground on spawn
16:23 < bridge> <yekrm> How can I configure a workflow to create an arm build?
16:23 < bridge> <yekrm> https://cdn.discordapp.com/attachments/293493549758939136/1176181142987218974/image.png?ex=656def0c&is=655b7a0c&hm=1dbd0c7cf2fa3b92de8cd588ffb4b2783054db50bc673c4f29ecee13f8194123&
16:40 < bridge> <ryozuki> if u mean a macos silicon u cant
16:40 < bridge> <ryozuki> actually i think they dont have  arm either
16:40 < bridge> <ryozuki> https://github.com/marketplace/actions/arm-runner
16:50 < bridge> <yekrm> It's ubuntu minimal arm
17:28 < bridge> <louis> openai drama goes hard lmfao
17:30 < bridge> <Jupstar ✪> ^
17:30 < bridge> <Jupstar ✪> 
17:30 < bridge> <Jupstar ✪> i'd say that prediction was quite good xd
17:31 < bridge> <Jupstar ✪> they might have not predicted the CEO gets fired.
17:31 < bridge> <Jupstar ✪> But they probably wanted more to say about OpenAI
18:26 < bridge> <Jupstar ✪> u can cross compile
18:26 < bridge> <Jupstar ✪> 
18:26 < bridge> <Jupstar ✪> u could see how the mingw cross compile is done
18:27 < bridge> <Jupstar ✪> it will work similar, u just need to install the correct packages
18:27 < bridge> <Jupstar ✪> at least if you meant with "create", that u want artifacts
18:27 < bridge> <Jupstar ✪> if u meant to actually run ARM on github, i guess that's not possible
18:28 < bridge> <yekrm> Sure I want the artifacts
18:28 < bridge> <Jupstar ✪> https://github.com/ddnet/ddnet/blob/375dd7da8b4bb6bc530d5726874bd0c8de8e7fa4/cmake/toolchains/mingw64.toolchain
18:29 < bridge> <Jupstar ✪> https://github.com/ddnet/ddnet/blob/375dd7da8b4bb6bc530d5726874bd0c8de8e7fa4/cmake/toolchains/mingw64.toolchain
18:29 < bridge> <Jupstar ✪> lol
18:29 < bridge> <Jupstar ✪> https://github.com/ddnet/ddnet#cross-compiling-on-linux-to-windows-x86x86_64
18:30 < bridge> <Jupstar ✪> but i bet u find lots of resources online how to cross compile to arm 😉
18:32 < bridge> <yekrm> I want the build to be compiled by the workflow :Paimon_DeadInside:
18:32 < bridge> <Jupstar ✪> yes
18:32 < bridge> <Jupstar ✪> https://github.com/ddnet/ddnet/blob/375dd7da8b4bb6bc530d5726874bd0c8de8e7fa4/.github/workflows/build.yml
18:33 < bridge> <Jupstar ✪> here is the workflow build file
18:33 < bridge> <Jupstar ✪> u gotta compile all skills i mentioned xdd
18:37 < bridge> <yekrm> After configuring so many servers I have no skills now
18:37 < bridge> <yekrm> :dead_pb:
18:38 < bridge> <Jupstar ✪> ask chatgpt
18:43 < bridge> <Jupstar ✪> here. no guarantee that this does smth useful:
18:43 < bridge> <Jupstar ✪> ```yml
18:43 < bridge> <Jupstar ✪> name: Cross-Compile for ARM
18:43 < bridge> <Jupstar ✪> 
18:43 < bridge> <Jupstar ✪> on:
18:43 < bridge> <Jupstar ✪>   push:
18:43 < bridge> <Jupstar ✪>     branches: 
18:43 < bridge> <Jupstar ✪>       - main
18:43 < bridge> <Jupstar ✪>   pull_request:
18:43 < bridge> <Jupstar ✪>     branches: 
18:43 < bridge> <Jupstar ✪>       - main
18:43 < bridge> <Jupstar ✪> 
18:43 < bridge> <Jupstar ✪> jobs:
18:43 < bridge> <Jupstar ✪>   build:
18:44 < bridge> <Jupstar ✪>     runs-on: ubuntu-latest
18:44 < bridge> <Jupstar ✪> 
18:44 < bridge> <Jupstar ✪>     steps:
18:44 < bridge> <Jupstar ✪>     - name: Checkout Repository
18:44 < bridge> <Jupstar ✪>       uses: actions/checkout@v2
18:44 < bridge> <Jupstar ✪> 
18:44 < bridge> <Jupstar ✪>     - name: Install ARM Cross Compiler
18:44 < bridge> <Jupstar ✪>       run: |
18:44 < bridge> <Jupstar ✪>         sudo apt-get update
18:44 < bridge> <Jupstar ✪>         sudo apt-get install -y gcc-arm-linux-gnueabihf g++-arm-linux-gnueabihf
18:44 < bridge> <Jupstar ✪> 
18:44 < bridge> <Jupstar ✪>     - name: Create Build Directory
18:44 < bridge> <Jupstar ✪>       run: mkdir build && cd build
18:44 < bridge> <Jupstar ✪> 
18:44 < bridge> <Jupstar ✪>     - name: Run CMake
18:44 < bridge> <Jupstar ✪>       run: |
18:44 < bridge> <Jupstar ✪>         cd build
18:44 < bridge> <Jupstar ✪> ```cmake
18:44 < bridge> <Jupstar ✪> # Set the system name to cross compile for
18:44 < bridge> <Jupstar ✪> set(CMAKE_SYSTEM_NAME Linux)
18:44 < bridge> <Jupstar ✪> set(CMAKE_SYSTEM_PROCESSOR arm)
18:44 < bridge> <Jupstar ✪> 
18:44 < bridge> <Jupstar ✪> # Specify the cross compiler
18:44 < bridge> <Jupstar ✪> set(CMAKE_C_COMPILER arm-linux-gnueabihf-gcc)
18:44 < bridge> <yekrm> <a:amehype:814855437865844746>
18:44 < bridge> <yekrm> I will try
19:01 < bridge> <_voxeldoesart> thats not really learning, is it
19:04 < bridge> <heinrich5991> did someone ask about learning?
19:23 < bridge> <ryozuki> ```x86asm
19:23 < bridge> <ryozuki> next_pow_2:
19:23 < bridge> <ryozuki>     dec eax
19:23 < bridge> <ryozuki>     bsr rcx, rax
19:23 < bridge> <ryozuki>     inc ecx
19:23 < bridge> <ryozuki>     mov eax, 1
19:23 < bridge> <ryozuki>     shl rax, cl
19:23 < bridge> <ryozuki>     ret
19:23 < bridge> <ryozuki> ```
19:23 < bridge> <ryozuki> i found this
19:23 < bridge> <ryozuki> looks op
20:20 < bridge> <heinrich5991> ChillerDragon: muzzle → muscle
21:10 < bridge> <Jupstar ✪> ```
21:10 < bridge> <Jupstar ✪>  $ cargo clean
21:10 < bridge> <Jupstar ✪>      Removed 90180 files, 48.7GiB total
21:10 < bridge> <Jupstar ✪> ```
21:10 < bridge> <Jupstar ✪> :twinbop:
21:11 < bridge> <Jupstar ✪> that's almost the size of my first ssd 😂
21:12 < bridge> <heinrich5991> since when does it display stats? 😮
21:13 < bridge> <Jupstar ✪> good question
21:16 < bridge> <chillerdragon> lol
21:17 < bridge> <chillerdragon> Chiller English be like
21:17 < bridge> <chillerdragon> https://youtu.be/wtRvTHUO6Z4?si=fwuVUbzpGKl40Gh9
21:45 < bridge> <chillerdragon> What are the chances of this getting merged this week? I would like to take a look at it before.\_.
21:47 < bridge> <heinrich5991> then say that on the PR
21:50 < bridge> <chillerdragon> Ok I did the review on ma phone :D
22:01 < bridge> <_voxeldoesart> i mean, the person did "I have no skills now"
22:01 < bridge> <_voxeldoesart> using ai doesnt really *give* you skill, does it?
22:13 < bridge> <yekrm> <a:RageJK:662741101983301642>
22:14 < bridge> <heinrich5991> if you can use LLMs to help you, then that's a skill right there 🙂
22:15 < bridge> <Jupstar ✪> IMO it's not a skill to setup some boring CIs
22:15 < bridge> <Jupstar ✪> i've done that many times
22:15 < bridge> <Jupstar ✪> i still forget their shitty syntax xD
22:16 < bridge> <yekrm> I said skills with irony, the truth is I'm busy with web development and other things
22:17 < bridge> <_voxeldoesart> sometimes i think about using ai to assist me with things but then i remember that its better to just. do it. instead of having to rely on a machine to do the work for you
22:17 < bridge> <_voxeldoesart> because most of the time the language models dont even do what you ask for
22:17 < bridge> <_voxeldoesart> bard and gpt suck
22:18 < bridge> <Jupstar ✪> so you also convert all english source to binary by hand? 😄
22:18 < bridge> <Jupstar ✪> u just hate AI
22:18 < bridge> <Jupstar ✪> they do it to like 80% depending on the complexity
22:18 < bridge> <Jupstar ✪> and they only give a limited response length
22:18 < bridge> <Jupstar ✪> that's also a problem
22:19 < bridge> <_voxeldoesart> https://lingojam.com/EnglishtoBinary
22:21 < bridge> <stormaxd> got a problem with long as hell render while starting client
22:21 < bridge> <stormaxd> Possible somehow to fix?
22:21 < bridge> <stormaxd> https://cdn.discordapp.com/attachments/293493549758939136/1176271184057339965/image.png?ex=656e42e8&is=655bcde8&hm=db7738b51878b3ef4eca7d6e139f8a08b337da17cfd5bb4d7a7919a0c7583e2f&
22:24 < bridge> <Jupstar ✪> can you please be more precise?
22:24 < bridge> <Jupstar ✪> is that a release build?
22:25 < bridge> <Jupstar ✪> shadow play for example might already make the startup much slower
22:25 < bridge> <stormaxd> debug
22:25 < bridge> <Jupstar ✪> OpenGL?
22:25 < bridge> <stormaxd> yea
22:26 < bridge> <Jupstar ✪> what even is long as hell
22:26 < bridge> <Jupstar ✪> <1s should be ok for the first render call i guess
22:26 < bridge> <Jupstar ✪> if the opengl driver setups stuff on fly
22:26 < bridge> <stormaxd> i
22:26 < bridge> <stormaxd> cracking hate
22:26 < bridge> <stormaxd> openGL
22:27 < bridge> <stormaxd> ty jupey
22:27 < bridge> <stormaxd> i just switch the render
22:27 < bridge> <Jupstar ✪> i blame windows tho xd
22:27 < bridge> <Jupstar ✪> opengl works awesome on linux
22:27 < bridge> <ryozuki> my record is 150gb
22:28 < bridge> <robyt3> Was anyone able to confirm that #7508 fixes the weird lags yet?
22:28 < bridge> <DDNet> https://github.com/ddnet/ddnet/pull/7508
22:30 < bridge> <Jupstar ✪> i asked multiple times in #bugs
22:30 < bridge> <Jupstar ✪> but sadly they hard ignore
22:30 < bridge> <Jupstar ✪> if they have problems, they ur best friends. else they full ego
22:30 < bridge> <Jupstar ✪> 😂
22:30 < bridge> <robyt3> It's possible that the compatibility flags make windows chose a worse option for the game
22:31 < bridge> <robyt3> It's quite a lot of compatiblity in Windows though: https://learn.microsoft.com/en-us/windows/win32/w8cookbook/windows-8-and-windows-server-8-compatibility-cookbook-portal
22:31 < bridge> <Jupstar ✪> it kinda must be
22:31 < bridge> <Jupstar ✪> bcs it makes 0% sense that the opengl windows software driver is used
22:31 < bridge> <robyt3> You can download as PDF though, only 227 pages ::monkaS:
22:31 < bridge> <Jupstar ✪> u cannot control that programatically
22:31 < bridge> <robyt3> You can download as PDF though, only 227 pages :monkaS:
22:32 < bridge> <robyt3> What is this? https://apps.microsoft.com/detail/opencl%E2%84%A2-and-opengl%C2%AE-compatibility-pack/9NQPSL29BFFF?hl=en-US&gl=US
22:33 < bridge> <Jupstar ✪> If a DirectX 12 driver is installed, supported apps will run with hardware acceleration for better performance. This package supports apps that use OpenCL version 1.2 and earlier and OpenGL version 3.3 and earlier.
22:33 < bridge> <Jupstar ✪> ogl on dx12
22:34 < bridge> <Jupstar ✪> not interesting for their potato problems
22:34 < bridge> <robyt3> What if they accidentally installed it?
22:34 < bridge> <robyt3> Would it make things worse?
22:34 < bridge> <Jupstar ✪> nah
22:34 < bridge> <Jupstar ✪> i dont think they installed that
22:35 < bridge> <Jupstar ✪> they all had old PCs
22:35 < bridge> <Jupstar ✪> it's probably bcs the drivers of their GPUs are flagged as win7 compability mode
22:35 < bridge> <robyt3> Which OS did they have btw?
22:35 < bridge> <Jupstar ✪> win10
22:36 < bridge> <robyt3> yeah, could be something like this
22:36 < bridge> <robyt3> right now it doesn't matter I guess, but at some point we might need the compatibility flags with new OS
22:37 < bridge> <robyt3> Are their drivers already end-of-life?
22:37 < bridge> <Jupstar ✪> you want to say drop OGL <2
22:37 < bridge> <Jupstar ✪> i am in
22:37 < bridge> <Jupstar ✪> 1000% xD
22:42 < bridge> <Jupstar ✪> does vanilla master servers use a different port to check open ports?
22:42 < bridge> <Jupstar ✪> 
22:42 < bridge> <Jupstar ✪> aka:
22:42 < bridge> <Jupstar ✪> "ERROR: the master server reports that clients can not connect to this server."
22:42 < bridge> <Jupstar ✪> i read it as this:
22:42 < bridge> <Jupstar ✪> the master server can contact my servers
22:42 < bridge> <heinrich5991> yes
22:42 < bridge> <Jupstar ✪> this is so annoying
22:43 < bridge> <heinrich5991> it's a port forward check ^^
22:43 < bridge> <Jupstar ✪> yeah but i can't cleanly block IPs that way
22:43 < bridge> <heinrich5991> you need to use a different port, otherwise the port forward check doesn't work
22:43 < bridge> <Jupstar ✪> yes
22:43 < bridge> <Jupstar ✪> but it could be a static port
22:45 < bridge> <Jupstar ✪> well let's hope the servers where the master servers don't report me for port scanning 😂
22:45 < bridge> <Jupstar ✪> this is so annoying xD
23:32 < bridge> <chairn> sorry no idea. But i had the same question for a lab work with RISC-V assembly where the stack is advanced by 16 for each function...
23:36 < bridge> <heinrich5991> because then you can easily put 16-byte aligned stuff onto the stack, and since you're likely gonna push rbp, it doesn't matter as much, I guess
23:37 < bridge> <ryozuki> u push if u want a stack frame
23:37 < bridge> <ryozuki> but yeah i guess
23:37 < bridge> <ryozuki> well u always need a stack frame for the abi i think
23:38 < bridge> <heinrich5991> I think windows always requires you to push rbp as well
23:53 < bridge> <ryozuki> u dont need to specifically push rbp tho
23:53 < bridge> <ryozuki> but pushing aligns the pointer
23:54 < bridge> <ryozuki> with -fno-frame-pointer it doesnt push
23:54 < bridge> <ryozuki> i read this today
23:54 < bridge> <ryozuki> it uses relative to rsp somehow
23:54 < bridge> <ryozuki> -fomit frame pointer sry
23:54 < bridge> <ryozuki> well if i got this correctly ofc
23:58 < bridge> <Mr.Gh0s7> This is an example of what I mean.
23:58 < bridge> <Mr.Gh0s7> Rust itertools has this <https://docs.rs/itertools/latest/itertools/trait.Itertools.html#method.k_smallest> which probably from my understading (correct me if I am wrong :)) is this https://en.wikipedia.org/wiki/Sorting_network. 
23:58 < bridge> <Mr.Gh0s7> In C you gotta implement/make it yourself
23:58 < bridge> <Mr.Gh0s7> btw I am talking about day1 part2 of the 2022 AoC
23:59 < bridge> <Mr.Gh0s7> so I feel like you learn a lot more about engeering in general but you also have to re-implement the wheel (since rust prob has it already).