01:24 < bridge> I'll test it out, see if I can get over the 5k fps xd 01:24 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1291178966714941451/image.png?ex=66ff279e&is=66fdd61e&hm=1ab0c4e213a80a16aa7e5054bf9bfba0c1cb330237abba6dfdc4469adac0919d& 01:45 < bridge> frame generation for a 2D game lol 01:46 < bridge> it's almost certainly more expensive than rendering an entire ddnet frame 02:25 < bridge> huge 02:25 < bridge> when will u be compiler developer making $1,000,000,000 a year like ryo 04:54 < bridge> what& 04:54 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1291231959099506802/image.png?ex=66ff58f8&is=66fe0778&hm=eb7633047d4b2b3d7c9d69f41ebb7989911302bad69da06e890918e0d95fd444& 05:38 < bridge> try not to post duplicate questions 05:38 < bridge> i would consult your IDE... obviously it knows that something is wrong 05:38 < bridge> what is the errer 05:38 < bridge> what is the error 06:09 < bridge> dont u need hashtags? 06:15 < bridge> that’s comment in cmake 06:34 < bridge> всем привет 06:35 < bridge> #off-topic 06:39 < bridge> please help me, how do I change my server type from TestDD to another one? 06:41 < bridge> ahh 06:41 < bridge> I think it’s sv_test or smth 06:43 < bridge> this 06:43 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1291259383275323434/image.png?ex=66ff7283&is=66fe2103&hm=d53ffb2c19a45ac5eb8b29961ca954f5d48c9ebff87c170bf9c478de37f5a867& 06:44 < bridge> change TestDD 06:58 < bridge> me when i read 07:04 < bridge> Never :feelsbadman: 07:06 < bridge> y not 07:07 < bridge> Because I am a neanderthal xdd 07:08 < bridge> no u are a student 07:08 < bridge> i am a neanderthal 07:09 < bridge> i wrap my head around the basic concept of writing something like a compiler and then i disregard it like i could do it with ease 07:09 < bridge> but bro 07:09 < bridge> u are doing it 07:52 < bridge> false 07:54 < bridge> https://redd.it/9iu0hj Join us now and share the software 07:54 < bridge> you’ll be free 08:35 < bridge> what's this wall of text for a single division operation :pepeH: is this some kind of optimization because using `idiv` is too bad? 08:35 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1291287388101804092/image.png?ex=66ff8c98&is=66fe3b18&hm=1b433445f03d926fbd13d32e0f32b94d3445f082e856333b2f95098184244d8b& 09:07 < bridge> btw, I checked how rust and C do it and it's interesting that rust just moves results on stack and then just loads them in registers but C puts everything in register from start and handles 3rd argument in the end 09:14 < bridge> i find a aled with Pseudo Fly 09:17 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1291297992963260508/202410031515.mp4?ex=66ff9678&is=66fe44f8&hm=f0d1752e8f559427f2e3cf4c46166617b697bf27f5c4138438a9eacd192b588e& 09:17 < bridge> just need 4tile high 09:27 < bridge> morning fellas 09:27 < bridge> good morning 09:27 < bridge> ``` 09:27 < bridge> [ 3.364712] usb 3-6: new high-speed USB device number 5 using xhci_hcd 09:27 < bridge> [ 8.870773] usb 3-6: device descriptor read/64, error -110 09:27 < bridge> [ 24.741791] usb 3-6: device descriptor read/64, error -110 09:27 < bridge> [ 25.024690] usb 3-6: new high-speed USB device number 6 using xhci_hcd 09:27 < bridge> [ 30.373739] usb 3-6: device descriptor read/64, error -110 09:27 < bridge> [ 46.245946] usb 3-6: device descriptor read/64, error -110 09:27 < bridge> [ 46.352822] usb usb3-port6: attempt power cycle 09:27 < bridge> [ 46.801693] usb 3-6: new high-speed USB device number 7 using xhci_hcd 09:27 < bridge> [ 51.601890] usb 3-6: Device not responding to setup address. 09:27 < bridge> [ 56.614006] usb 3-6: Device not responding to setup address. 09:27 < bridge> [ 56.821514] usb 3-6: device not accepting address 7, error -71 09:27 < bridge> [ 57.000692] usb 3-6: new high-speed USB device number 8 using xhci_hcd 09:27 < bridge> [ 61.801134] usb 3-6: Device not responding to setup address. 09:27 < bridge> [ 66.814256] usb 3-6: Device not responding to setup address. 09:27 < bridge> [ 67.021761] usb 3-6: device not accepting address 8, error -71 09:27 < bridge> [ 67.025083] usb usb3-port6: unable to enumerate USB device 09:27 < bridge> [ 67.200764] usb 3-9: new full-speed USB device number 9 using xhci_hcd 09:28 < bridge> [ 67.578333] usb 3-9: New USB device found, idVendor=31b2, idProduct=0011, bcdDevice= 1.01 09:28 < bridge> [ 67.578543] usb 3-9: New USB device strings: Mfr=1, Product=2, SerialNumber=3 09:28 < bridge> [ 67.578735] usb 3-9: Product: TONOR TD510 Dynamic Mic 09:28 < bridge> [ 67.578925] usb 3-9: SerialNumber: 2023-02-16-0000-0000-0000- 09:28 < bridge> ``` 09:28 < bridge> @learath2 i think i have a faulty usb port (port 6), and it slows down my boot time massively, do u know if its possible to disable this specific port? 09:28 < bridge> tbh idk what i have there since all the devices i know work 10:06 < bridge> how the hell does godot debugger not have a step out 10:07 < bridge> :pepeW: the proposal for adding one was in 2021 with no activity. gamedev just don't use debuggers huh. 10:07 < bridge> Rip it 10:13 < bridge> Yes, divs are shit, everyone uses mul 10:29 < bridge> I don’t think so. Atleast a quick google suggests not 11:22 < bridge> https://doc.rust-lang.org/nomicon/ffi.html#the-nullable-pointer-optimization 11:52 < bridge> https://github.com/godotengine/godot/pull/97758 it's not even hard to implement 11:53 < bridge> cellegen yesterday casually found a bug in godot with compressed textures xd 11:54 < bridge> did you guys fix it 11:55 < bridge> no, i dislike compressed textures anyway xd 11:55 < bridge> they have really weird bug, even back when ddnet had them 11:55 < bridge> they have really weird bugs, even back when ddnet had them 11:55 < bridge> godot is swapping VRAM texture compression lib in 4.4 iirc was it pre lib change or post? 11:55 < bridge> isn't that part of the driver? 11:56 < bridge> but dunno which version he used 11:56 < bridge> oh well 11:56 < bridge> i probably don't care anyway 11:56 < bridge> yeah same 11:56 < bridge> i just wanted to write random stuff 11:57 < bridge> https://godotengine.org/article/dev-snapshot-godot-4-4-dev-1/#betsy-texture-compressor was talking about this 11:58 < bridge> oh wow 11:59 < bridge> ah they are purely lossy it seems that makes sense 11:59 < bridge> i thought they went all out and decode on the fly in the shaders 11:59 < bridge> vram compressed textures are always lossy 11:59 < bridge> yeah and that is also why they suck 12:00 < bridge> try to update 2 pixels on your lossy texture 12:00 < bridge> it's kinda required for running large 3d scenes on old mobile chip tho lol 12:00 < bridge> how so 12:00 < bridge> sounds like the vertex buffers would also cost lot of VRAM 12:01 < bridge> :think_bot: big texture chokes slow mem bandwidth. it's just a snapdragon thing 12:02 < bridge> probably same with vertex buffers i guess. our VR game really chokes on full 4k textures 12:02 < bridge> Generally understanable, but I am suprised still. 12:02 < bridge> 12:02 < bridge> What kind of resolutions do we talk about? 12:02 < bridge> Ah 4k 12:02 < bridge> Why don't you simply give out HQ full hd textures then? 12:02 < bridge> as alternative 12:03 < bridge> honestly it is also a artist problem 12:03 < bridge> our artist only worked with single objects before. 12:04 < bridge> some of our big buildings are single objects that is UV unwrapped onto one texture with substance painter 12:04 < bridge> so is either i remake all the textures and models, or i just crank the resolution 12:04 < bridge> guess which solution i chose 12:05 < bridge> 32k 12:06 < bridge> i also tried secondary details, sampling a repeated texture and splat them onto the model, but the additional sample also tanks the framerate. 12:06 < bridge> solution. i'm just saying i kept 4k res instead of remaking all the texture 12:06 < bridge> :justatest: 12:06 < bridge> Did you manually wrote shaders for Unity? 12:06 < bridge> 12:06 < bridge> Or does unity support so much stuff? 12:06 < bridge> i use some third party node graph thing 12:08 < bridge> free shader. lol 12:08 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1291341077059731519/NVToonBasic.shader?ex=66ffbe98&is=66fe6d18&hm=4e3a447c7fdb7287bf5658a53e09300d22d4625049e755f87346dc924f1bc1f7& 12:08 < bridge> cell shading 12:08 < bridge> free **proprietary** shader leak. lol 12:08 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1291341077059731519/NVToonBasic.shader?ex=66ffbe98&is=66fe6d18&hm=4e3a447c7fdb7287bf5658a53e09300d22d4625049e755f87346dc924f1bc1f7& 12:08 < bridge> BRO WHAT 12:08 < bridge> Don't tell me you wrote this 12:09 < bridge> is that auto generated? 12:09 < bridge> ye 12:09 < bridge> shader graph 12:09 < bridge> by ur graphs? 12:09 < bridge> ok 12:09 < bridge> sick 12:09 < bridge> so much code that has to run on the GPU xd 12:10 < bridge> :think_bot: probably could hand write the whole thing and save a lot 12:10 < bridge> but you know 12:10 < bridge> crunched and stuff 12:11 < bridge> does unity offer debug tools to measure VRAM bandwidth? 12:11 < bridge> to find bottlenecks 12:11 < bridge> maybe 12:12 < bridge> but i just use snapdragons tool or renderdoc on device 12:12 < bridge> I kinda would find it cool to add gpu benchmarks to ddnet to test out what makes it faster or slower 😄 12:12 < bridge> running in editor is very different from on device, unity packs textures differently per target. 12:13 < bridge> these engines are so massive, sometimes i really wonder how they even work at all xd 12:14 < bridge> gamedevs care none of that 12:14 < bridge> it builds for your target. if something doesn't work, you just call unity and get on-site help (supposedly, but we are not unity enterprise licensed) 12:15 < bridge> Ah so then my question is: Does it happen often that smth doesn't work? 😄 12:15 < bridge> the only thing that I can't solve is the physics engine just stopped detecting raycasts 12:16 < bridge> but it was fixed after we disabled like 15 invisible collision shapes from our character. 12:16 < bridge> still don't know why it happens 12:16 < bridge> most other things are solvable or solved really 12:17 < bridge> 12:17 < bridge> we had a rendering issue where our stencil portal effect and outline is flying all over the place with our dynamic viewport 12:18 < bridge> found the render code, forked it, applied proper xr rendering matrix, booted the official packages and rolled our own. 12:18 < bridge> good time 12:20 < bridge> and there was also a third-party package that runs IL codegen which breaks IL2CPP randomly for half a year, but it was eventually fixed by said third-party. we just spend two unpaid weekend migrating the code to the new version since it was breaking changes 12:20 < bridge> and there was also a third-party package that runs IL codegen which breaks IL2CPP randomly for half a year, but it was eventually fixed by said third-party. we just spend two unpaid weekend migrating the code to the new version since it was a breaking change 12:22 < bridge> unpaid pf 12:22 < bridge> that's just a china thing 12:22 < bridge> shall your boss fix it, if he thinks he is so clever 12:22 < bridge> she* 12:23 < bridge> my lead does not make game. so he actually listen to me for everything, other than deadline. which is the product team's lead's job 12:23 < bridge> our CEO is a she* 12:23 < bridge> i thought u quit 12:23 < bridge> i did 12:24 < bridge> i don't work there anymore. just saying it was like that before 12:24 < bridge> funny enough that they made oct 1st deadline which i thought was impossible for two interns to do without me 12:24 < bridge> but they haven't bothered me once since i left 12:25 < bridge> so maybe i'm actually the slow one. they actually move super fast withoute 12:25 < bridge> so maybe i'm actually the slow one. they actually move super fast without me 12:25 < bridge> they just make you think that 12:25 < bridge> they probs cry in pain 12:25 < bridge> "their CEO" 12:26 < bridge> good point 12:26 < bridge> not "out" :p 12:26 < bridge> our 12:26 < bridge> they didn't make any new announcement on their PR channel tho, so i assumed they missed the deadline actually 12:26 < bridge> lol 12:28 < bridge> this is the graph. super simple, see? 12:28 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1291346170370854973/image.png?ex=66ffc357&is=66fe71d7&hm=de9d96a17e89179cc8566f13de880cbe786b6642d4ab9c5d12cac66002f9e398& 12:28 < bridge> :kek: 12:29 < bridge> WAAT that is the shader i just saw? xd 12:29 < bridge> ye 12:29 < bridge> i mean it's cool that it scales so well. 12:29 < bridge> 12:29 < bridge> I never wrote a complex shader like that 12:30 < bridge> even tho it looks so simlple 12:30 < bridge> even tho it looks so simple 12:30 < bridge> i think the hardest i ever did was deferred lights 12:30 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1291346666565275668/image.png?ex=66ffc3cd&is=66fe724d&hm=c023fa3a7d047a92984c1426890e8e688054a37cac39fdc687ec5b0baac43f71& 12:30 < bridge> same shader two layers down 12:31 < bridge> it always says toon shader, is that like cell shading or what is toony about it? 12:31 < bridge> it's just celshade 12:31 < bridge> i gotta call it NVToon to impress our CEO saying that "we made this" 12:32 < bridge> our company name shorten to NV btw 12:32 < bridge> ah lol 12:32 < bridge> nvidia 12:32 < bridge> we do post processing in material shader too 12:32 < bridge> mobile post processing is quite slow, so we just color correct and stuff on the object itself 12:32 < bridge> 12:32 < bridge> it would be cool to have materials in ddnet. but i also know that this isn't trivial 12:33 < bridge> ah man 12:33 < bridge> also in 2d many materials don't really work like that 12:33 < bridge> i forgot to stole my cool sky shader... 12:33 < bridge> xD 12:33 < bridge> when i did sky boxes, i really made a box and instead murphed the texture so, that it looks round 12:33 < bridge> was super fast to render 12:33 < bridge> since it was 6 textures only 😄 12:34 < bridge> when i did sky boxes, i really made a box and instead morphed the texture so, that it looks round 12:34 < bridge> i mean.. wasn't that just a regular skybox 12:34 < bridge> many games so low poly 3d circles 12:34 < bridge> man my english sucks xd 12:34 < bridge> spheres 12:34 < bridge> right 12:34 < bridge> our game had day night cycle and different weathers so i made a dynamic skybox shader lol 12:35 < bridge> ah man, now i want my shader back. i might ask my intern to stole it for me 12:35 < bridge> 12:35 < bridge> mario kart double dash 12:35 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1291347970540961802/image.png?ex=66ffc504&is=66fe7384&hm=900d0439c2411a40cd53819036c7e4f44316204c8757b8862b76cad6253e3369& 12:35 < bridge> for game cube 12:38 < bridge> i kinda like to see how old games rendered the lvl. 12:38 < bridge> 12:38 < bridge> This is dark souls for pc 12:38 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1291348613896863775/image.png?ex=66ffc59d&is=66fe741d&hm=0bb1baa23b20c6021d8ca61b5f8ff5711eabd2c69f6bfeda9930c0d2fdb63fe6& 12:39 < bridge> :think_bot: sky 12:40 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1291349090596294766/image.png?ex=66ffc60f&is=66fe748f&hm=48866c9efd1a2d113a95e7e40de81b46759e88d46f6a1501f6630b325c677bb0& 12:42 < bridge> noclip.website is really the best website ever xdd 12:42 < bridge> it's just cool to fly around all the games u used to play as kiddo 12:43 < bridge> oh great 12:44 < bridge> i didn't do anything today 12:44 < bridge> time to cry in bed 12:55 < bridge> @learath2 I solved the problem with `rdx`, every time when I do mul/div operation it checks if `rdx` is in use, if it it push it to stack, do the operation and pop it back xd 13:22 < bridge> Don't forget to allocate stack for it 13:23 < bridge> Also you can use a register for it too if possible. Moving to memory is expensive 13:23 < bridge> but `push` instruction does it, no? 13:23 < bridge> Yep, that's fine. I didn't think you used the push instruction for whatever reason 13:24 < bridge> is that bad :justatest: 13:24 < bridge> Nonono, it's fine 13:24 < bridge> i just pop it right after idiv 13:24 < bridge> i just pop it right after idiv/imul 13:25 < bridge> Actually now that I think about it. You could have handled all your expression mess using a stack machine. Very inefficient but it makes it trivial to calculate any expression 13:26 < bridge> All you need to do would have been to traverse the expression tree in post-order. For leaves emit stack push. For non leaves, pop, pop, operation. 13:27 < bridge> Oh well 13:37 < bridge> currently it generates ``` 13:37 < bridge> mov r15b, 1 13:37 < bridge> mov r14b, 2 13:37 < bridge> movzx rax, r14b 13:38 < bridge> mov r14b, 3 13:38 < bridge> cqo 13:38 < bridge> idiv r14b 13:38 < bridge> mov r14b, al 13:38 < bridge> movzx rax, r14b 13:38 < bridge> mov r14b, 4 13:38 < bridge> imul r14b 13:38 < bridge> mov r14b, al 13:38 < bridge> add r15b, r14b 13:38 < bridge> movzx r15w, r15b 13:38 < bridge> mov word ptr [rbp - 2], r15w 13:38 < bridge> ``` 13:38 < bridge> for `1 + 6 / 3 * 4` it's not the most readable but also not the worst 13:38 < bridge> currently it generates ``` 13:38 < bridge> mov r15b, 6 13:38 < bridge> movzx rax, r15b 13:38 < bridge> mov r15b, 3 13:38 < bridge> cqo 13:38 < bridge> idiv r15b 13:38 < bridge> mov r15b, al 13:38 < bridge> movzx rax, r15b 13:38 < bridge> mov r15b, 4 13:38 < bridge> imul r15b 13:38 < bridge> mov r15b, al 13:38 < bridge> add r15b, 1 13:39 < bridge> movzx r15w, r15b 13:39 < bridge> 13:39 < bridge> ``` 13:39 < bridge> for `1 + 6 / 3 * 4` it's not the most readable but also not the worst 13:39 < bridge> currently it generates ``` 13:39 < bridge> mov r15b, 1 13:39 < bridge> mov r14b, 6 13:39 < bridge> movzx rax, r14b 13:39 < bridge> mov r14b, 3 13:39 < bridge> cqo 13:39 < bridge> idiv r14b 13:39 < bridge> mov r14b, al 13:39 < bridge> movzx rax, r14b 13:39 < bridge> mov r14b, 4 13:39 < bridge> imul r14b 13:39 < bridge> mov r14b, al 13:39 < bridge> add r15b, r14b 13:39 < bridge> movzx r15w, r15b 13:39 < bridge> ``` 13:39 < bridge> for `1 + 6 / 3 * 4` it's not the most readable but also not the worst 13:39 < bridge> currently it generates ``` 13:39 < bridge> mov r15b, 1 13:39 < bridge> mov r14b, 6 13:39 < bridge> movzx rax, r14b 13:39 < bridge> mov r14b, 3 13:39 < bridge> cqo 13:39 < bridge> idiv r14b 13:39 < bridge> mov r14b, al 13:39 < bridge> movzx rax, r14b 13:40 < bridge> mov r14b, 4 13:40 < bridge> imul r14b 13:40 < bridge> mov r14b, al 13:40 < bridge> add r15b, r14b 13:40 < bridge> movzx r15w, r15b 13:40 < bridge> ``` 13:40 < bridge> for `1 + 6 / 3 * 4` it's not the most readable but also not the worst 13:40 < bridge> now if i also add sethi ullman algorithm here, it will be so goood 13:57 < bridge> is there a C library that is famous and doesnt have bindings in rust? 13:58 < bridge> considering what is famous 13:59 < bridge> like used somewhere 13:59 < bridge> idk 14:00 < bridge> i need project ideas 14:01 < bridge> https://github.com/tsoding/olive.c 14:02 < bridge> nvm they exist 14:02 < bridge> lmao 14:02 < bridge> https://crates.io/crates/olivec-sys 14:02 < bridge> idk go make an interpreter then 14:02 < bridge> its relatively easy 14:03 < bridge> bro i already have my own language 14:03 < bridge> <7ovvv> @everyone 14:03 < bridge> https://ed-lang.org/ 14:03 < bridge> pro 14:03 < bridge> maybe i should refactor it 14:03 < bridge> its compiled tho 14:03 < bridge> compiled is better 14:03 < bridge> make a scripting language 14:03 < bridge> nah 14:03 < bridge> yea 14:03 < bridge> wasm2 14:04 < bridge> i also have a rysk emulator https://github.com/edg-l/rysk 14:04 < bridge> i also have a riscv emulator https://github.com/edg-l/rysk 14:05 < bridge> is edlang written in edlang 14:05 < bridge> nah 14:05 < bridge> in rust 14:05 < bridge> it uses llvm 14:05 < bridge> bootstrap it 14:05 < bridge> rust is written in rust 14:05 < bridge> edlang should be written in edlang 14:07 < bridge> useless idea but still an idea 14:07 < bridge> have u made your own javascript framework? 14:08 < bridge> make new js runtime 14:09 < bridge> great idea 14:09 < bridge> not like there are at least 5 of them 14:09 < bridge> i dont wanna code in js 14:09 < bridge> i wanna code in rust 14:10 < bridge> Please help, how to increase the chances of registration in the community and how to change the server type from TestDD to my own? 14:10 < bridge> maybe i could do a DSL language 14:10 < bridge> nobody does 😬 14:10 < bridge> the language instructions are to draw graphics 14:10 < bridge> jdsl?? 14:10 < bridge> dsl = domain specific lang 14:11 < bridge> https://thedailywtf.com/articles/the-inner-json-effect 14:11 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1291371968133857352/1727922783025.jpg?ex=66ffdb5d&is=66fe89dd&hm=444c14d3d4c1c843ae8f67506f91f00201ca1068d5c3f522ddbb59e58d940a54& 14:11 < bridge> json dsl 14:11 < bridge> Please help, how to increase the chances of registration in the community and how to change the server type from TestDD to my own? 14:11 < bridge> make a json library that emits rust docs as comments 14:13 < bridge> ask in #questions 14:13 < bridge> Thx 14:40 < bridge> @jupeyy_keks nice job on the gpu prs 😄 14:40 < bridge> is it possible to check if vulkan would crash and swap to opengl automatically? 14:55 < bridge> Probably. But probably not in a cross plattform way, easily 14:59 < bridge> Except for android there doesn't seem to be high demand for it tho 14:59 < bridge> 14:59 < bridge> And on Android the driver situation is horrible as expected 15:00 < bridge> would be great so there's no need for so many launch options 15:00 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1291384441750687815/image.png?ex=66ffe6fb&is=66fe957b&hm=9d5e553ea56ab883ff658a91de53048e92ad44f91d8c370f0d7c3bfb2dafc9e2& 15:01 < bridge> Yes, that would be great 15:01 < bridge> But the amount doesn't matter 15:01 < bridge> either None or it doesnt matter 15:08 < bridge> @kebscs How do you want to detect an issue like this tho: 15:08 < bridge> https://discord.com/channels/252358080522747904/757720336274948198/1199056232498086029 15:09 < bridge> The intel OpenGL driver situtation is still really horrible on Windows 15:09 < bridge> well we recently also had a case where it looked very weird 15:10 < bridge> idk, just the one where a game doesnt start with vulkan at all is fine 15:10 < bridge> https://discord.com/channels/252358080522747904/745926398140612678/1281958197841887322 15:10 < bridge> xd 15:11 < bridge> https://discord.com/channels/252358080522747904/757720336274948198/1276507134158442608 15:11 < bridge> this one is funny too 15:11 < bridge> WTF happens there 15:11 < bridge> that is a modern intel gpu 15:12 < bridge> the video is so nice 15:12 < bridge> it works 15:12 < bridge> and then the beat drops 15:12 < bridge> 😂 15:14 < bridge> @learath2 omagad, look at dis, it's code generated for `1 + (2 + (3 + (4 + (5 + 6))))` expression :poggers2: 15:14 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1291387969395097610/image.png?ex=66ffea44&is=66fe98c4&hm=b8f6a10ffa1c611122669fe3e457001dc3d95b1338f106d16c7ca9cea43d21bd& 15:15 < bridge> So hot, is it just the immediates being added? 15:15 < bridge> ye xd 15:20 < bridge> @matodor why do i hear constant moaning from your video in #9099 ? :Susge: 15:20 < bridge> https://github.com/ddnet/ddnet/pull/9099 15:21 < bridge> wtf 15:21 < bridge> enough internet 15:21 < bridge> for this year 15:22 < bridge> here's for `a + (b + (c + (d + (e + f))))` 15:22 < bridge>