00:39 < bridge> pay 1$ for mentioning*that* thing 00:39 < bridge> pay 1$ for mentioning *that* thing 00:39 < bridge> :troll: 09:14 < bridge> https://cdn.discordapp.com/emojis/1084542360375017604.png?quality=lossless&size=48 10:51 < bridge> Has anyone ever come across a Discord embed that contains a functional video? 10:53 < bridge> What is a functional video? 10:53 < bridge> A video that works? 10:55 < bridge> Yes, im trying to upload a video and embed it, but for some reason it doesn't want to load. 10:55 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1095633175398584350/image.png 10:55 < bridge> It looks like it tries tho 10:55 < bridge> Have u tried to press F5 xd/ browser 10:56 < bridge> Yes, maybe it's because the media lacks some META informations so it doesn't embed properly :/ 10:57 < bridge> it tries yeah, but turns into the poop emoji image thing after awhile 😄 10:59 < bridge> Why error in Javascript console? Maybe u can try a different video? 11:00 < bridge> Any error in Javascript console? Maybe u can try a different video? 11:05 < bridge> yoo is that a new report system? 11:06 < bridge> 👀 11:07 < bridge> I've tried multiple videos, different types but all seem to fail. Meh. 11:07 < bridge> Rip and embed by url not an option? 11:08 < bridge> links should work, i'll give it a try 11:12 < bridge> some video codecs dont work in embed 11:12 < bridge> H.264 should work 11:16 < bridge> If he used a video rendered by client it should be h264 14:43 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1095690584930398278/RDT_20230412_1442483230580334376168936.jpg 17:38 < bridge> make ddnet a non profit so donations are tax deductible xd 17:38 < bridge> 😄 17:39 < bridge> https://www.hetzner.com/press-release/arm64-cloud/ 17:39 < bridge> HETZNER INTRODUCES ARM64-BASED CLOUD SERVERS 17:39 < bridge> @deen ddnet on arm? 17:39 < bridge> I thought about that 17:40 < bridge> but we even had problems with physics differences on x86 servers, so arm would be even worse probably 17:40 < bridge> f 17:40 < bridge> and if I make the main server arm, then have to cross-compile which is annoying 17:40 < bridge> and also lots of effort to set it up yet again 17:42 < bridge> and I don't want to know what kind of bugs mariadb or other software have which don't show up on x86-64 because it makes concurrency easier 17:42 < bridge> see for example https://research.swtch.com/hwmm 17:49 < bridge> is arm weakly ordered? 17:51 < bridge> yes 17:52 < bridge> nice link 17:52 < bridge> i think i seen it before 17:56 < bridge> https://deck.of.cards/old/ 18:20 < bridge> omg epic, but sadly more expensive than a vps on other hoster 18:20 < bridge> else i'd switch 18:21 < bridge> starting to render weapons ^^ 18:21 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1095745552211255366/output.mp4 18:21 < bridge> ez 18:22 < bridge> i wanna make my own client with like, OTT graphic related things 18:22 < bridge> hard to explain rn 18:29 < bridge> hrmpf when did I break the tees again 18:29 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1095747442353385572/output.mp4 18:29 < bridge> ah I remember 18:30 < bridge> so this is to spectate tees on a web browser? 18:30 < bridge> sorry im dumb 18:31 < bridge> hm mainly its just my hobby project where I try to replicate the rendering 18:31 < bridge> cursed tees ghosts 18:31 < bridge> how is that dumb, could be 18:32 < bridge> cant wait for webgpu to stabilize ❤️ 18:32 < bridge> but yeah that should be very easily possible, dunno about performance ^^ 18:32 < bridge> I think chrome announced it for second quarter this year o.o 18:32 < bridge> Oo 18:33 < bridge> did you try your twgpu on experimetal chrome? 18:33 < bridge> no :d 18:33 < bridge> then ill try for you XD 18:34 < bridge> https://developer.chrome.com/blog/webgpu-release/ 18:35 < bridge> @Patiga is this server demo rendering? 18:35 < bridge> atm my day job is doing gpgpu on the web (with webgl1) I cant fucking wait for webgpu 18:35 < bridge> this is a normal client demo 18:35 < bridge> boring 18:35 < bridge> but if it helps #reports 18:35 < bridge> hm? 18:35 < bridge> cool 18:36 < bridge> oh interesting, what kind work do you do with webgl? 18:37 < bridge> you bring quite a depressing mood here ^^. what you see is a client demo, the graphic interface just takes a snapshot, so you could input all the other storage formats as well 18:37 < bridge> is a server demo even different from a client demo? 18:38 < bridge> I don't think so 18:38 < bridge> maybe default0 means teehistorian 18:38 < bridge> if webgpu works, try to benchmark against webgl2 18:38 < bridge> and also against ddnet client with webgl2 18:39 < bridge> https://aliveclan.de/ddnettest/load_map.html 18:39 < bridge> 18:39 < bridge> here u can also render demos 18:40 < bridge> what? is there a webgl2 backend for wgpu? 18:40 < bridge> yes :D 18:40 < bridge> no way 18:40 < bridge> https://mapview.patiga.eu/ 18:40 < bridge> omg 18:41 < bridge> mh ah yeah vsync is always enabled for no apparent reason, maybe it was a SDL bug 😄 18:41 < bridge> then why am i still not using twgpu then 18:41 < bridge> that suprises u more than webgpu backend? XD 18:42 < bridge> ok its their main goal apparently 18:42 < bridge> so maybe indeed not surprising xD 18:43 < bridge> @deen could you please enable this again? :) 18:43 < bridge> can u compile it with webgpu? 18:43 < bridge> hmmh i had this very bug using wasm/sdl2 cant remember the workaround but i can look it up if u want 18:43 < bridge> or is it some flag 18:43 < bridge> tbh i dunno if the client still compiles to wasm with the rust bridge 18:43 < bridge> never tested it 18:45 < bridge> "When running in a web browser (by compilation to WebAssembly) without the "webgl" feature enabled, wgpu relies on the browser's own WebGPU implementation. WGSL shaders are simply passed through to the browser, so that determines which WGSL features you can use." 18:45 < bridge> so i assume simply remove the webgl feature 18:45 < bridge> I had hardcoded the backend https://gitlab.com/Patiga/twgpu/-/blob/8bb9c111af10cf85da90ab04ea856530fc2fbf35/map-inspect-web/src/lib.rs#L198 though I don't know if it will run straight away. https://github.com/gfx-rs/wgpu/wiki/Running-on-the-Web-with-WebGPU-and-WebGL this is quite outdated already 😅 18:46 < bridge> gitlab down, sad 18:48 < bridge> weird, works for me 18:49 < bridge> https://aliveclan.de/ddnettest/ is that twgpu or smth else? 18:49 < bridge> its ddnet 18:49 < bridge> literally 18:49 < bridge> oh 18:49 < bridge> i thought why the effort to make the ui match XD 18:50 < bridge> it runs pretty smoothly 18:50 < bridge> i just remove some data and added something that emulates a command parameter to work with the javascript input 18:53 < bridge> ok the loadmap is a bit customized, it wanted to use it for official ddnet, but webgl2 support seems to non existing on macos lel+ 18:54 < bridge> i dunno why but my internet seems broken on many sites rn xD 18:54 < bridge> discord also doesnt work anymore 18:54 < bridge> yes its a bummer, they say they support but its full of garbage 18:56 < bridge> is there a way to load other maps on https://mapview.patiga.eu/ ? 18:56 < bridge> is there a way to load other maps on https://mapview.patiga.eu/ ? @Patiga 19:01 < bridge> @k2d222 https://mapview.patiga.eu/?map=dm2 the `map` parameter for maps that are stored in the server behind the site (I think the teeworlds maps rn), `url` https://mapview.patiga.eu/?url=https://heinrich5991.de/teeworlds/maps/maps/Animations_c2aa2208a9f27fcfd8349f34cf1633c038c17bdedbc225c1cdd14d4ecc19c3c0.map needs cross-origin enabled, and `ddnet` and `ddnet-testing` also exists in theory but cross-origin is disabled there rn 19:02 < bridge> the map renderer doesn't have mipmaps yet 19:03 < bridge> and also no parallax-zoom :p 19:05 < bridge> cant help but notice its a bit laggy 😅 19:06 < bridge> yes, for me too 19:06 < bridge> on chrome its already better 19:06 < bridge> wow yes, so much difference 19:07 < bridge> why firefox is weird af 19:08 < bridge> when i zoom out one step more(in ddnet rendering) for some reason it drops from like 240fps to 15fps xD 19:08 < bridge> otherwise, one of: tilemap rendering method is just less performent, gl2 backend is badly optimized (I think jupstar looked into the commands run in the background once and found a lot of redundant commands), wgpu cpu-intensive checks 19:08 < bridge> always had a couple fps more on chrome on wasm and webgl, but here its more than usual 19:09 < bridge> btw how do you mipmap the tiles? without the texture bleeds in the texture atlas? 19:09 < bridge> 2d texture array 19:09 < bridge> perfect fit :D 19:09 < bridge> (I don't do that yet tho :d) 19:10 < bridge> 1 layer per tile? 19:10 < bridge> yea 19:10 < bridge> 256 layers, actually the minimum of layers supported everywhere according to wgpu limits 19:11 < bridge> I do custom mipmap levels for tiles layer on my mc clone, that works well but hell to implement in js on the fly 19:11 < bridge> I have a gimp script to do that 19:12 < bridge> heh ^^ 19:14 < bridge> dont u just need to make multiple textures and upload them all without generating mipmaps and swap them on the CPU then 19:15 < bridge> if its 2d u can estimate what mipmap would be used 19:15 < bridge> oh, I see u mean based on zoom level I swap the texture? 19:16 < bridge> if u even have a concept like zoom 19:16 < bridge> 19:16 < bridge> but yeah depending on resolution and zoom 19:17 < bridge> how is that better than having all the mip levels attached to the texture and letting gl do mipmap lookup? 19:18 < bridge> my issue is how do I generate the mip levels 19:18 < bridge> i assumed u faced texture bleeding 19:18 < bridge> and in webgl1 u might not have 2d array textures or sampler methods in the shader 19:19 < bridge> well glGenerateMipMaps or a texture flag is it for old OpenGL at least 19:19 < bridge> else search a resize js library and always half the size 19:19 < bridge> else search a resize js library and always half the size 19:21 < bridge> my problem is, glGenMipmaps generates mipmaps by averaging 4 pixels, but for a texture atlas, that doesnt work bcs it will mix with the neighboring tile on the edges, thats what I meant by texture bleed 19:21 < bridge> what u need is to avg 4 pixels but taking into account tile borders 19:21 < bridge> so custom mipmaping 19:22 < bridge> yep but even then it will make the rendering quite inaccurate 19:22 < bridge> for a 512x512 texture u have completly different texel offsets than 1024x1024 19:22 < bridge> 1/1024 vs 1/512 19:22 < bridge> so u basically face the opposite of bleeding still 19:23 < bridge> with your way of rendering 19:23 < bridge> generally u cannot get the pixel perfect rendering without 2d array textures tho 19:23 < bridge> the edge texels are always cut by exactly half a pixel 19:24 < bridge> I dont get it 🥹 19:24 < bridge> ok wait tbf it depends on if u are using linear interpolation for sampling 19:25 < bridge> are you doing it? 😄 19:25 < bridge> oh yea now I think I see what u mean 19:25 < bridge> I cant do the double interpolation thingy 19:26 < bridge> gl_nearest_mipmap_linear 19:26 < bridge> but gl_linear_mipmap_linear would not work is that it? 19:27 < bridge> that too 19:27 < bridge> 19:27 < bridge> but gl_linear 19:27 < bridge> and gl_linear_mipmap_linear 19:27 < bridge> and 19:27 < bridge> gl_linear_mipmap_nearest wont work either 19:27 < bridge> basically only with gl_nearest it might work 19:28 < bridge> right I get it now, still better than nearest 19:34 < bridge> i still dont get the map-inspect to build at all xD 19:34 < bridge> have you tried the steps from the readme? ^^ 19:34 < bridge> wasm ah yeah 19:41 < bridge> cargo is my second package manager xd 19:41 < bridge> when first 19:41 < bridge> zsh: segmentation fault (core dumped) wasm-pack build --target web 19:41 < bridge> rust and segmentation fault 19:41 < bridge> 🤦‍♂️ 19:42 < bridge> yeah rust only os soon™️ 19:42 < bridge> redox? 19:42 < bridge> nah bad design 19:42 < bridge> then, probably not soon XD 19:44 < bridge> wat o.o 19:45 < bridge> mh weird, 0.10.3 also crashes 19:46 < bridge> let me try rust stable 19:47 < bridge> mh no chance, rip 19:47 < bridge> bro how 19:47 < bridge> works for me 19:47 < bridge> how does wasm-pack manage to die 19:48 < bridge> i run it in twgpu and the other dir 19:48 < bridge> 19:48 < bridge> both dont work 19:48 < bridge> where do i need to run it xd 19:48 < bridge> inside of `map-inspect-web` 19:48 < bridge> yeah weird that it crashes for me 😄 19:48 < bridge> but even outside it shouldn't crash 19:48 < bridge> > Error: failed to parse manifest: /home/patiga/git/twgpu/Cargo.toml 19:48 < bridge> > Caused by: missing field `package` 19:49 < bridge> maybe my system is broken 😄 19:49 < bridge> can you run the command at all without crashing, like version or smth 19:49 < bridge> -V, --help, does the command run at all? 19:49 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1095767758031044639/image.png 19:52 < bridge> @Jupeyy_Keks try `cargo install wasm-pack --no-default-features` 19:53 < bridge> https://github.com/rustwasm/wasm-pack/issues/1203 19:53 < bridge> ah nice thx 19:57 < bridge> bro too hard for me xD 19:57 < bridge> [INFO]: found wasm-opt at "/home/jupeyy/.cargo/bin/wasm-opt" 19:57 < bridge> Error: /home/jupeyy/.cargo/bin/bin/wasm-opt binary does not exist 19:57 < bridge> why does it try to find it in /bin/bin now xD 19:58 < bridge> after it found it already xD 19:58 < bridge> lmao `/bin/bin/` I can't 19:58 < bridge> have dinner now tho, gl ^^ 19:59 < bridge> good one, ty xd 21:07 < bridge> @Jupeyy_Keks did you manage to solve it? 😅 21:08 < bridge> yes, but it doesnt run, adapter cannot be created, maybe a newer version of wgpu fixes problems with webgpu? 21:10 < bridge> twgpu is currently on the latest major version of wgpu, so that shouldn't be the issue 21:10 < bridge> https://www.phind.com/search?q=how+to+make+a+curl+request+with+rust&c=&source=searchbox&init=true 21:10 < bridge> but yeah, seems like some incompatibility issue 21:10 < bridge> chatgpt google for devs 21:11 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1095788195754549430/image.png 21:15 < bridge> hm interesting website, that is the exact example from the readme/docs 21:22 < bridge> @Jupeyy_Keks https://www.intel.com/content/www/us/en/newsroom/news/intel-foundry-arm-announce-multigeneration-collaboration-leading-edge-soc-design.html#gs.uk5fnk 21:24 < bridge> "Easy" API 21:24 < bridge> its rly called like this 21:24 < bridge> on the C side too 21:24 < bridge> actually 21:24 < bridge> look at ddnet code 21:24 < bridge> u will find itg 21:24 < bridge> wait wtf i thought that was sarcasm 21:24 < bridge> the easy api is meant to be ez to use 21:24 < bridge> cuz curl has million options 21:24 < bridge> fair 21:24 < bridge> btw curl is everywhere 21:24 < bridge> under ur bed too 21:25 < bridge> i curl under my blanket when i go to sleep does that count 21:25 < bridge> y 21:25 < bridge> its comfy, 22:13 < bridge> i'm impressed how good lapce became in such a small time frame, if it has debugger support it could already replace vscode for rust development 22:13 < bridge> 22:13 < bridge> the performance is superb 22:33 < bridge> lapce you say 👀 22:43 < bridge> Q: Can lapce invoke building of a project? 22:44 < bridge> i guess not yet 22:44 < bridge> yeah, I couldn't find it. Snappy however 22:45 < bridge> i hope they get it done this year, would be nice to have a bit of challenge for vscode 😄 22:46 < bridge> I've tried to like vscode... can't.... 22:46 < bridge> i like it tbh 😄 22:46 < bridge> 22:46 < bridge> i can even use it in a browser and write code on my vm 22:46 < bridge> For me it always feels sluggish 22:46 < bridge> but sometimes it gets slow already yeah 22:46 < bridge> on my desktop its ok, bcs good hardware 22:47 < bridge> but laptop is already little struggle 22:47 < bridge> But it feels very wrong that you have shit-ton of good hardware, and Turbo Pascal back in the 90s was snappier 22:47 < bridge> xD 22:49 < bridge> when it comes to electron and javascript in general, i'd still say vscode is one of a kind. 22:49 < bridge> 22:49 < bridge> most js stuff sucks so hard, vscode is pretty nice compared to it 22:50 < bridge> I rather not touch JS even with a 10 foot pole 22:50 < bridge> i feel like the discord app on my phone is slower than opening it in the browser ^^ 22:50 < bridge> WebTech© 22:51 < bridge> As you might have guessed, I'm not too fond of some of the trends in software development 🙂 22:55 < bridge> u just need for daily dose of ryozuki news in this channel then you are always up to date 22:55 < bridge> 22:55 < bridge> if there ever is a language that can replace rust, ryozuki will let me know 22:55 < bridge> he's the news boss 22:57 < bridge> Is he even a contributer? BAH 😄 23:04 < bridge> Of ddnet? Then yes^^ 23:20 < bridge> I ment of rust 23:43 < bridge> wassup matricks ! 23:45 < bridge> Not much, about to hit the hey