00:20 < bridge> Teelevision 00:20 < bridge> :poggers2: 00:20 < bridge> tv at home: 00:20 < bridge> 😮 00:31 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1203482909563031592/GFbnIq9XMAAZ1CD.png?ex=65d141cd&is=65becccd&hm=77babc90ef14848eda5139ce39fc9211ad9b8db740db8c7a8ef228b42fcba28d& 00:31 < bridge> why do i laugh at such stupid things 00:32 < bridge> 50/50, either he finishes or he doesnt 00:32 < bridge> easy! 00:32 < bridge> 50/50, either I finish it or don't 00:32 < bridge> 😄 00:32 < bridge> DEEN 00:32 < bridge> MY JOKE 00:32 < bridge> two idiots, one thought 00:32 < bridge> :kek: 00:33 < bridge> lmao 01:00 < bridge> That has to be the weirdest flex ever 01:05 < bridge> Donation by teelevision?!!!! Isn’t that the guy who run zCatch in the 90s? I did not know he is still around 01:06 < bridge> just zero the filesize character? 01:09 < bridge> or don't use std::vector for this, you don't need expandable container for reading a file 01:27 < bridge> that is not him 01:38 < bridge> is it possible to allow ddnet client connect deepl api for translating chat? 01:39 < bridge> https://github.com/sienori/simple-translate/wiki/How-to-register-DeepL-API 01:39 < bridge> its half-free to use legally 01:58 < bridge> Yeah but you have to do it lol 02:08 < bridge> `std::fs::read(filename)` in rust 02:09 < bridge> rust doesn't do it with the above-mentioned method 02:10 < bridge> in rust you can write to the not-yet initialized parts of the vector and afterwards set its length 02:11 < bridge> this is guaranteed to be a memcpy for types that support it, I think 02:11 < bridge> read it as you'd read the C standard. something will happen, but the result will be the same as if each element was cloned 02:12 < bridge> btw, even that is very easy to do: https://docs.rs/memmap2/latest/memmap2/ 02:12 < bridge> ```rs 02:12 < bridge> let file = File::open("LICENSE-APACHE")?; 02:12 < bridge> let mmap = unsafe { Mmap::map(&file)? }; 02:12 < bridge> ``` 02:12 < bridge> works on all platforms already 02:13 < bridge> (mmap is inherently unsafe, if the underlying file changes) 02:17 < bridge> I feel you. I also like to write efficient code, which is why I started doing it in C. later I found that I could achieve the same in Rust, and to me, it felt a lot easier to get correct 02:40 < bridge> Imposter? 02:48 < bridge> just someone else 02:59 < ws-client> Its the same spelling isnt it? https://github.com/Teelevision 02:59 < bridge> how to have a number parameter with clap 02:59 < bridge> this seems obvious 03:00 < ws-client> Reusing such a well known name seems a bit problematic to me 03:01 < bridge> nvm 03:02 < bridge> had to use `default_value_t` in the arg() proc macro instead of `default_value` 03:02 < bridge> o_O 03:03 < bridge> sucks that it doesn't come with de/serializers for std::time Duration 03:03 < bridge> i think if u conform to a particular syntax it's easy 03:46 < bridge> well there's my first contribution in rust 🥱 04:26 < ws-client> uhm @avolicious the 2nd hit on duck.com for qshar is some april fool from 2023 about kog shutting down. That seems a bit odd 04:26 < ws-client> https://zillyhuhn.com/cs/.1707017179.png 04:31 < ws-client> also when add a irc bridge to kog so i can stop using ddnet#developer for kog stuff 04:46 < bridge> lol true 04:46 < bridge> idk why you'd wanna talk about kog tho 05:27 < monkyyy> hello i'd like to complain that I joined multimap on usa3, and in 15 seconds batsoup vote kicked me 05:27 < monkyyy> I remember that name being an asshole from eailer so thats probaly not the first time 05:27 < monkyyy> fuck him 06:14 < bridge> dude you should not use the dev channel to complain about a blocker 06:14 < bridge> please create a ticket 06:15 < bridge> mods dont read this chat. too high iq :greenthing: 06:20 < bridge> when creating a ticket from irc? 06:53 < ws-client> when turn off kick votes? 07:48 < bridge> when they hire more mods 09:50 < bridge> omg @robyt3 awesome :3 09:52 < bridge> @deen @learath2 when i start ddnet from steam i get: 09:52 < bridge> ``` 09:52 < bridge> ~/.steam/debian-installation/steamapps/common/DDraceNetwork/ddnet/DDNet: symbol lookup error: ~/.steam/debian-installation/steamapps/common/DDraceNetwork/ddnet/DDNet: undefined symbol: curl_multi_poll 09:52 < bridge> ``` 09:52 < bridge> 09:52 < bridge> if i start the steam game files without steam (./DDNet) it works 09:52 < bridge> 09:52 < bridge> maybe steam has it's own curl library? 09:53 < bridge> ^ nightly 09:53 < bridge> (probably related https://discord.com/channels/252358080522747904/757720336274948198/1203282828310224977) 10:05 < bridge> Aaaaaaaaaaaa 10:07 < bridge> @jupeyy_keks can you find the curl they ship and tell me whether it has `curl_multi_wait`? 10:10 < bridge> sure, which one should i test with? 😬 10:10 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1203628674449145966/image.png?ex=65d1c98e&is=65bf548e&hm=2ce8a7b2ff0adc5f4d5cbc17f0eff1b9b9ec2379e173f37546617dded8ed9876& 10:11 < bridge> how can i look up which one is loaded without starting the game? 10:11 < bridge> Normally I'd say ldd, but that won't work here 10:14 < bridge> @jupeyy_keks how about setting `LD_LIBRARY_PATH` like steam does to this directory and doing the ldd? 10:14 < bridge> That should work 10:15 < bridge> but don't i need the path for that xd 10:16 < bridge> i can just try the one from `.steam/bin32/steam-runtime/usr/lib/x86_64-linux-gnu/` 10:16 < bridge> that sounds like a useful one 10:16 < bridge> That should be it 10:17 < bridge> can't post it for some reason, read pm 10:25 < bridge> »Spam« 10:27 < bridge> "Das kann nicht gepostet werden, da Inhalte enthalten sind, die von diesem Server blockiert werden. Diese Nachricht wird eventuell auch von den Servereigentümern gelesen." 10:31 < bridge> @jupeyy_keks yo get back here :p 10:31 < bridge> https://fosdem.org/2024/schedule/event/fosdem-2024-3615-0-a-d-game-vulkan-api/ 10:32 < bridge> nice they also get vk api now 10:33 < bridge> jup in diguise doing 0ad stuff 10:34 < bridge> AutoMod kicked in ^^ 10:35 < bridge> @ryozuki there is a chain of llvm talks today https://fosdem.org/2024/schedule/track/llvm/ 10:38 < bridge> jup in disguise doing 0ad stuff 11:01 < bridge> nice 11:12 < bridge> When license.txt update to new year? :gigachad: 11:12 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1203644187439013899/image.png?ex=65d1d800&is=65bf6300&hm=384aff36fb444931ca5ce21510994e95109e0eaa1c5d39a9c9fdda7211fbaa22& 11:12 < bridge> make pr get famous for being the change 11:13 < bridge> :poggers2: 11:19 < bridge> @jupeyy_keks done 11:20 < bridge> :gigachad: 11:23 < bridge> @learath2 i discovered ive been using my dualsense a bit buggy 11:23 < bridge> because i didnt enable osme kernel options 11:23 < bridge> now it works nice 11:23 < bridge> https://wiki.gentoo.org/wiki/Sony_DualShock 11:24 < bridge> it also supports bt sync via usb 11:25 < bridge> This os the downside of building your own kernel. You have to be careful 😄 11:25 < bridge> xd 11:25 < bridge> What is a bt sync? 11:26 < bridge> bluetooth auto sync 11:26 < bridge> bluez says wanna accept this device? y/nn 11:26 < bridge> > Connect the DualShock 3 to the system using a USB cable and press the PlayStation button. 11:26 < bridge> > 11:26 < bridge> > Allow the service authorization request: 11:26 < bridge> > [agent]Authorize service service_uuid (yes/no): yes 11:26 < bridge> Ah, I get it 11:26 < bridge> I thought sony did that in a proprietary way 11:27 < bridge> i couldnt know a reason why do that propietary 11:27 < bridge> if ur controller is used for more than ps5 u get more sells 11:27 < bridge> i bought my ps5 dualsense for pc and i dont have a ps5 11:27 < bridge> xd 11:27 < bridge> 70€ btw 11:28 < bridge> oh it was ur first contribution, epic change 😏 11:29 < bridge> :poggers2: 11:30 < bridge> Hey can you add the % rank thing as a filter to ddnet.tw so I can see on what maps I can improve best? 11:31 < bridge> I've always used xbox controllers on my pc. I've always found it more comfy 11:31 < bridge> https://tenor.com/view/ni%C3%B1o-asustado-huh-what-meme-stare-gif-16752413 11:31 < bridge> you are one of them 11:32 < bridge> Seriously. The asymmetric placement of the analog sticks look janky, but it just feels correct in the hand 😄 11:32 < bridge> it never felt good for me xd 11:32 < bridge> I haven't even had an xbox like ever until last year 11:33 < bridge> does the xbox controller also have the pressure thing? 11:33 < bridge> on l2 and r2 11:33 < bridge> its smth dualsense has new 11:33 < bridge> I mean if you tried it without prejudice and didn't like it, then nothing I can say 11:33 < bridge> its epic for some games 11:33 < bridge> resistance* 11:33 < bridge> idk the name 11:33 < bridge> No it doesn't. Are there even pc games that use that? 11:34 < bridge> yeah 11:34 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1203649717800800296/image.png?ex=65d1dd27&is=65bf6827&hm=476675ff1274a55a0ed0eb66820eb52c07eb5bde238b68e374b5c7d8b74e3029& 11:34 < bridge> and more 11:35 < bridge> Cool 11:35 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1203649997426528256/image.png?ex=65d1dd6a&is=65bf686a&hm=1f5b38e3f9187553215c5526a4fbd71147af06c8724b98560da29c61d940483f& 11:35 < bridge> it supports it when the game natively supports dualsense controllers 11:35 < bridge> steam tells u 11:35 < bridge> im playing persona 3 and it supports dualsense 11:35 < bridge> altho its a simple game so no features used xd 11:37 < bridge> @learath2 wanna know smth curious 11:37 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1203650562172657684/image.png?ex=65d1ddf0&is=65bf68f0&hm=abee509b40f91018a5b93aea494720a8c3b1f0e379a7616557b904e9bfea6a43& 11:38 < bridge> Huh, interesting 11:38 < bridge> https://controller.dl.playstation.net/controller/lang/en/fwupdater.html 11:39 < bridge> oh u can update the firmware 11:39 < bridge> sad i dont have winshit 11:39 < bridge> https://linuxgamingcentral.com/posts/how-to-update-dualsense-firmware-on-linux/ 11:39 < bridge> lol 11:39 < bridge> New project time. Reverse engineer and port the updater 11:41 < bridge> Or a different project. Touch grass and find a friend with a ps5 11:41 < bridge> xd 12:02 < bridge> :poggers2: 12:02 < bridge> Developer moment 12:03 < bridge> ok maybe not checked checkboxes in my pull request wasn't the problem 😦 12:03 < bridge> It was personal 12:04 < bridge> yep 12:09 < bridge> yep 12:12 < bridge> yep 12:24 < bridge> @learath2 why is bottles not on gentoo overlay? 12:25 < bridge> It should be available on some overlay 12:25 < bridge> yea 12:26 < bridge> i tried usinh flatpak for bottles for quick n dirty but 12:26 < bridge> sandboxed shit, doesnt detect my controller 12:26 < bridge> So just use an overlay and get it from there? Or steal the ebuild from an overlay and put it in your local repo 12:27 < bridge> yea 12:27 < bridge> will try later 12:27 < bridge> i went to the outside world for a bit 12:27 < bridge> where many windows users lie 12:28 < bridge> @learath2 ever used a iphone 12:28 < bridge> my sister got a new phone and is giving me their iphone 11 12:30 < bridge> I did, back in highschool 12:31 < bridge> well thats long ago ig 12:31 < bridge> They make decent phones, if only it wasn't a walled garden 12:31 < bridge> i forgot ur age 12:31 < bridge> yeah 12:32 < bridge> I remember it being fairly usable after you jailbreak 12:32 < bridge> time to break the jail 12:32 < bridge> Though recently I think there aren't any untethered jailbreaks around 😦 12:32 < bridge> whats untethered 12:33 < bridge> > An untethered jailbreak is a jailbreak wherein a user can reboot their device at will, and have their device start up with the jailbreak automatically applied without the assistance of a computer or a utility on the device. 12:33 < bridge> You can reboot the device out in the wild. With a tethered jailbreak you need a computer when you restart to rejailbreak 12:34 < bridge> Anything else seems unusable unless you only use the jailbreak for playing around 12:34 < bridge> I guess we did get a semi-tethered jailbreak too. An app on your phone that you had to launch to rejailbreak 12:35 < bridge> Though that came with the annoying having to either have a apple developer certificate or having to resign the app every other month 12:55 < ws-client> @ryozuki you are the one who added the doc comments for the CUIRect code right? 12:55 < ws-client> its epic i just managed to vsplit thanks to it <:poggers2:1008007455936094328> 12:55 < bridge> ye 12:55 < ws-client> was smoother than me trying to vsplit something in html for the first time 12:56 < bridge> :poggers2: 12:57 < bridge> https://edgarluque.com/blog/ui-code-ddnet/ smh 12:57 < ws-client> ah ye that true i know it exists that would have been my fallback 12:57 < bridge> :kek: 12:57 < ws-client> but the comments in the code are so good already 12:57 < bridge> true, its great 12:58 < ws-client> I guess its time for server side rendering 12:58 < ws-client> I will ditch html/css for server side generated tw screenshots 12:58 < ws-client> front end in C++ <:poggers:546812233867329556> 13:00 < ws-client> @meloƞ i had to look up smh xd 13:00 < ws-client> meaning, "shaking my head", smh is typically used when something is obvious, plain old stupid, or disappointment. 13:00 < ws-client> https://www.urbandictionary.com/define.php?term=smh 13:00 < ws-client> still dont get it xd 13:00 < bridge> i rather translate it to "somehow" meaning, "how did you somehow miss this" 13:01 < bridge> but yeah shaking my head fits aswell 13:01 < ws-client> i have been reading it as "somehow" so far 13:01 < bridge> same 13:01 < bridge> and thats usually the context i use it in 13:01 < bridge> smh 13:02 < bridge> i use it for disapointment 13:02 < bridge> like here 13:02 < bridge> :kek: 13:02 < ws-client> guuuys time to republish the lgpl blog under freebsd license https://edgarluque.com/blog/ui-code-ddnet/ 13:02 < ws-client> its license expired! 13:03 < bridge> :justatest: oh no 13:03 < ws-client> https://zillyhuhn.com/cs/.1707048192.png 13:03 < ws-client> EZ 13:06 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1203672959575789649/image.png?ex=65d1f2cc&is=65bf7dcc&hm=d4fe9e1eafcfdfcabeb47cf530e674a0cb24d9c5e060a9f230c78d60ef029536& 13:06 < bridge> EZ 13:17 < bridge> I found after i run "sudo pacman -Syuu", my vscode become more green (Color Theme: Dark +). 13:17 < bridge> Are you got the same problem? 14:06 < bridge> is the ddnet bind system touring complete? 14:25 < bridge> lol 14:26 < bridge> i just made bind files and ddnet segfaulted xDDD 14:26 < bridge> It most certainly is not turing complete 14:26 < bridge> ```python 14:26 < bridge> for i in range(1000): 14:26 < bridge> with open(f'e{i}.bind', 'w') as file: 14:26 < bridge> file.write(f'echo {i}; exec e{i+1}.bind') 14:26 < bridge> ``` 14:27 < bridge> executing e0.bind crashes the game 14:27 < bridge> might be a stack overflow 14:27 < bridge> although 1000 seems a little low 14:27 < bridge> maybe we have a large buffer in that function 14:27 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1203693355725496390/image.png?ex=65d205cb&is=65bf90cb&hm=c5fa6d2a5abc8241113990c106d0f1591f12c3aa4e107e6534ea9d91626a76a7& 14:28 < bridge> 110 is a little random tbh 14:28 < bridge> probably a stack overflow 14:28 < bridge> I suppose we could limit recursion to like 10 or so 14:29 < bridge> very annoying limiter 14:29 < bridge> i would suggest just not implementing it recursively if possible 14:29 < bridge> practically, I don't think you need more than 10 14:30 < bridge> on a theoretical level I agree with you. but it doesn't seem to be worth it to me 14:33 < bridge> If you could branch on the value somehow it could be turing complete but I can't think of a way to do that 😄 14:33 < bridge> i have noticed anothe bug 14:33 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1203694809181593700/image.png?ex=65d20726&is=65bf9226&hm=5acbdad5ef9d158ae7854766a6c942598cdeaf9fa093fe0695924843deee89ab& 14:34 < bridge> also i think it would be cool to exec every line of the bind after 1 tick. so the game doesnt hang or lang on executing large binds 14:34 < bridge> and it would allow for cool matchups 14:35 < bridge> Looks like some flavor of memory corruption. Try with asan should be a free bugfix 14:35 < bridge> would allow delayed binds which would enable input delaying which we don't want to allow 14:35 < bridge> this would also fix the recursion bug since it would queue the binds 14:36 < bridge> what kind of input delaying? 14:36 < bridge> fire after 50 ticks 14:36 < bridge> ^^ 14:36 < bridge> we don't want that 14:36 < bridge> you cant do +fire or anything in bind files? 14:36 < bridge> You can, but it'll happen instantly when you press a button 14:37 < bridge> lemme try 14:37 < bridge> anyway. bind delaying is something we'll likely never support 14:37 < bridge> why is that bad? 14:37 < bridge> if every exec happened after a tick, you could spam 50 execs to get 50 ticks of delay 14:37 < bridge> yes. 14:37 < bridge> because it's an input we don't want to support. we don't want to automate players away 14:38 < bridge> you still cant do +right, +left or +hook or look in any kind of direction? how would just firing be useful? maybe for double rockets? but you still need to just jump perfectly and aim 14:39 < bridge> I can't even imagine much of a legitimate use for it, so I'd guess whatever you want to do with it is what we don't want in the game 😄 14:39 < bridge> making weaponswitches better 14:39 < bridge> e.g. double rockets yes, with no recursion limit and the ability to get 1 tick delays, you can automate that away completely 14:40 < bridge> how???? you cant even aim or jump automatically 14:40 < bridge> if the jumpt isnt perfectly tuned to the tick it fires it wont work 14:40 < bridge> if the jump isnt perfectly tuned to the tick it fires it wont work 14:40 < bridge> anyway. we don't want to open that can of worms 14:40 < bridge> delayed binds aren't coming 14:42 < bridge> rust 14:42 < bridge> You should be able to jump in a bind file 14:43 < bridge> you cant. 14:45 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1203697766371688479/2024-02-04_14-44-14.mp4?ex=65d209e7&is=65bf94e7&hm=01e26e9dbe48d262de7c0706963042b8a6560cfa17110dc5917f199262360ff1& 14:46 < bridge> That's nice to see 14:49 < bridge> But I bet you can do it by binding `exec file` to the mousewheel, having `file` bind `+jump` to the mousewheel, then just spam 14:50 < bridge> doesnt work 14:52 < bridge> as long as its in a file it doesnt get treated diffrently when bound to mouse or keybord 14:52 < bridge> as long as its in a file it doesnt get treated diffrently when bound to mouse or keyboard 14:52 < bridge> as long as its in a file it doesnt get treated differently when bound to mouse or keyboard 14:52 < bridge> Actually binding +jump to mousewheeldown just straight up doesn't work 14:53 < bridge> It probably needs an actual stroke with button down + button up 14:54 < bridge> my suggestion is that binds that are chained like `echo "a"; echo "b"` are executed instantly while binds like 14:54 < bridge> ``` 14:54 < bridge> echo "a" 14:54 < bridge> echo "b" 14:54 < bridge> ``` 14:54 < bridge> are delayed by 1 tick each 14:54 < bridge> Won't happen anyway, I was just theorycrafting 14:55 < bridge> i had high hopes for a fellow aw member 😔 14:57 < bridge> btw aimline is still bugged by 1 pixel 14:57 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1203700837080236133/2024-02-04_14-48-41.mp4?ex=65d20cc3&is=65bf97c3&hm=dc752e3fa3c6d4b553d1f5f637f5d7af3d883b186c99506f1a15198419775e9e& 14:59 < bridge> latest master code? 15:00 < bridge> i reverted the commit that caused the last bug 15:00 < bridge> lemme test 15:02 < bridge> yes 15:02 < bridge> ~~lemme test~~ 15:02 < bridge> ~~i reverted the commit that caused the last bug~~ 15:03 < bridge> its just off by 1 tw unit 15:04 < bridge> @furo321 ^ 15:13 < bridge> Eh, I have no clue what could be the issue this time. I can't reproduce the issue either. 15:21 < bridge> in rust, this would panic (abort?) as well 15:26 < bridge> yeah no segfault 15:31 < bridge> really? 15:31 < bridge> if you go on x=15.00 on tutorial and hook straight up? 15:33 < bridge> Ah, now it happens. It happens on older versions as well. 15:33 < bridge> yea i thought so. 15:34 < bridge> doesnt seem related to the new bug 15:34 < bridge> doesnt seem related to the last bug 15:34 < bridge> probably just the intersect line function that is off by 1 15:36 < bridge> I think it's actually done via segfaults ^^ 15:36 < bridge> I think they started abort()ing from the segfault handler because people asked why rust code can segfault ^^ 15:36 < bridge> via reliable segfaults 15:52 < bridge> the only important is that the hookline color is correct 15:52 < bridge> so if the ground would be lower it should be red 15:53 < bridge> if u know what i mean 15:53 < bridge> visually it might be wrong 15:53 < bridge> drawing lines is also a "weird" task for the GPU 15:53 < bridge> subpixels & stuff 15:55 < bridge> kk. color is correct 16:16 < bridge> Not the same person, i wasnt even born in the 90s 17:04 < bridge> Is it hard to make DDNet available for `winget`? 17:53 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1203745160232439909/image.png?ex=65d2360a&is=65bfc10a&hm=559f7196c72ab802a596e6a73b006f530615f49e6616f76ea924eaa7a0fa4a62& 17:59 < bridge> average chillerbot enjoyer 18:16 < bridge> <_wohoo_> Hello all, i have so much input lag with new 18.0.3 version, related to 18.0 release. 18:16 < bridge> <_wohoo_> I was rendering with OpenGL 2.0.0, now I'm forced to use OpenGL 3.0.0 (worst) or 3.3.0 (slight better). 18:16 < bridge> <_wohoo_> Vulkan not available on my laptop card. 18:16 < bridge> <_wohoo_> How can I fallback to OpenGL 2.0.0? Thank you 18:20 < bridge> windows? 18:44 < bridge> <_wohoo_> Yes 18:55 < bridge> what gpu do u have? you said laptop, is there a chance it has two gpus? like intel + nvidia? 19:08 < bridge> <_wohoo_> Correct, Intel + nvidia discrete 19:09 < bridge> <_wohoo_> With 18.0 it rendered with (custom) opengl 2.0.0 19:14 < bridge> and did u try opengl 1? does that lag too? 19:14 < bridge> <_wohoo_> Input lag increased using external display, because nvidia output routed through iGPU. 19:14 < bridge> <_wohoo_> I blame them enough for this design choice. 19:14 < bridge> <_wohoo_> Changing OpenGL version made it barely unplayable 19:15 < bridge> @_wohoo_ can you start the client press f1 and show the purple text to me? 19:15 < bridge> it says smth about gfx 19:17 < bridge> <_wohoo_> GPU Vendor NVIDIA 19:17 < bridge> <_wohoo_> GPU renderer GeForce 820M/PCIe/SSE2 19:17 < bridge> <_wohoo_> GPU version: 3.3.0 Nvidia 389.12 19:17 < bridge> thats the problem then 19:17 < bridge> it uses your nvidia card instead of intel 19:17 < bridge> can you look into your windows settings if you can force intel 19:17 < bridge> https://i0.wp.com/pureinfotech.com/wp-content/uploads/2022/01/windows-10-select-gpu-preference-per-app.webp?resize=827%2C489&quality=78&strip=all&ssl=1 19:18 < bridge> it looks smth like this 19:20 < bridge> or does it _only_ allow nvidia to be used here? 19:21 < bridge> <_wohoo_> Intel gpu go black screen after 1h playing. 19:21 < bridge> <_wohoo_> It started crashing some years ago, not first version of teeworlds or ddnet. 19:21 < bridge> <_wohoo_> 19:21 < bridge> <_wohoo_> Since command input still handled, i have to open another ddnet and blindly tape /swap etc... 19:21 < bridge> <_wohoo_> 🤣 19:21 < bridge> anyway: 19:21 < bridge> 1. i don't see the difference between opengl 2 & 3, but you can do `gfx_gl_major 2; gfx_gl_minor 0` to have opengl 2.0 19:21 < bridge> 2. opengl 1.x should then probably also work? XD 19:22 < bridge> anyway: 19:22 < bridge> 1. i don't see the difference between opengl 2 & 3, but you can do `gfx_gl_major 2; gfx_gl_minor 0` + restart to have opengl 2.0 19:22 < bridge> 2. opengl 1.x should then probably also work? XD 19:22 < bridge> <_wohoo_> I will test again forcing opengl version 19:26 < bridge> i am surprised your nvidia card doesnt support vulkan tho. 19:26 < bridge> 19:26 < bridge> it should 19:26 < bridge> <_wohoo_> Ok nice. 19:26 < bridge> <_wohoo_> It created again Custom opengl 2.0.0 19:26 < bridge> <_wohoo_> I configured nvidia panel control with hardcoded path and missed new download release 19:27 < bridge> <_wohoo_> Testing vulkan on igpu 19:27 < bridge> the igpu might not support vulkan 19:27 < bridge> nvidia cards are usually better at such support 19:27 < bridge> <_wohoo_> Ok, not work. 19:27 < bridge> <_wohoo_> Neither discrete 19:27 < bridge> mh weird 19:27 < bridge> i think chairn has an even older nvidia card 19:28 < bridge> and 2.0.0 does not lag? 19:28 < bridge> <_wohoo_> No that's OK 19:28 < bridge> mh ok 19:29 < bridge> but f1 says still nvidia? 19:29 < bridge> with opengl 2 19:29 < bridge> <_wohoo_> Red screen for vulkan 19:29 < bridge> <_wohoo_> No, now it says intel gpu 19:29 < bridge> better red screen than blue screen 😏 19:29 < bridge> ok 19:30 < bridge> but then it confirms that nvidia card creates the input lag. 19:30 < bridge> 19:30 < bridge> but anyway, you can stay with opengl 2.0 i guess ^^ 19:30 < bridge> <_wohoo_> Ok thank you, so moving to igpu solved. 19:30 < bridge> <_wohoo_> I hope i will not see black screen again. 19:30 < bridge> <_wohoo_> Thank you 19:33 < bridge> <_wohoo_> Another thing is about freezer indicator. 19:33 < bridge> <_wohoo_> Some updates ago it was forced snowflake instead of stars. 19:33 < bridge> <_wohoo_> Nobody noticed flickering with many tees frozen in group? 19:33 < bridge> <_wohoo_> Snowflakes are all synced 19:35 < bridge> <_wohoo_> Probably I have too much photosensitivity, 😅 20:40 < bridge> ```c++ 20:40 < bridge> auto Transform = [AvailableExtensions](const char *ExName){ 20:40 < bridge> auto It = std::ranges::find_if(AvailableExtensions, 20:40 < bridge> [ExName](const char *e){ return !std::strcmp(ExName, e); }, 20:40 < bridge> [](const VkExtensionProperties &Props){ return Props.extensionName; }); 20:40 < bridge> 20:40 < bridge> return std::make_tuple(ExName, It != AvailableExtensions.end()); 20:40 < bridge> }; 20:40 < bridge> 20:40 < bridge> bool ExtensionMissing = false; 20:40 < bridge> std::span ct{GLFWRequiredExtensions, GLFWRequiredExtensionCount}; 20:40 < bridge> for(const auto &e : ct | std::views::transform(Transform) | std::views::filter([](const auto &es){ return !std::get<1>(es); })) { 20:40 < bridge> ExtensionMissing = true; 20:40 < bridge> std::cout << "Missing required extension: " << std::get<0>(e) << std::endl; 20:40 < bridge> } 20:40 < bridge> if(ExtensionMissing) 20:40 < bridge> throw std::runtime_error{"Missing required VK extension"}; 20:40 < bridge> ``` 20:40 < bridge> 20:40 < bridge> Today in my C++20 adventure I explored ranges, not half bad to use ngl. They also seem to suffer from bad optimization under gcc though :/ 20:43 < bridge> and before someone says anything, yes I know it's completely pointless 😄 21:23 < bridge> xd 21:23 < bridge> ill just say 21:23 < bridge> don't 21:23 < bridge> modern c++ looks beatiful no kap 21:25 < bridge> btw, I still haven't rendered anything yet :xDe: 21:25 < bridge> xd 21:25 < bridge> well that's a lie, I did render a quick triangle, but nothing else 21:25 < bridge> why dont u learn vk with c 21:26 < bridge> isnt the original api c 21:26 < bridge> vulkan is already complicated 21:26 < bridge> gamedev is like orthogonal to C 21:27 < bridge> besides, I might aswell get my C++ knowledge up to date while I'm at it, I've been stuck at C++11 for too long 21:53 < bridge> 2024-02-04 21:52:31 I gfx: GPU vendor: NVIDIA 21:53 < bridge> 2024-02-04 21:52:31 I gfx: GPU renderer: GeForce GT 740M 21:53 < bridge> 2024-02-04 21:52:31 I gfx: GPU version: Vulkan 1.1.95 (driver: 425.31.0.0) 21:53 < bridge> looks like my driver is more up to date than his 21:58 < bridge>