00:11 < bridge_> ChillerDragon: we have a working fork of teeworlds, it's called ddnet 00:32 < bridge_> what are vim motions 00:32 < bridge_> doesnt git bash use vim motions or smth 01:38 < bridge_> holy shit 01:39 < bridge_> i just got jumpscared by the quitting screen 01:39 < bridge_> :gigachad: 09:04 < bridge_> xd trol 09:04 < bridge_> (@heinrich5991) 09:05 < bridge_> I mean one that merges the vanilla prs and is only focusing on the vanilla gametype. Trying to get merged back into the official repo. 09:06 < bridge_> It uses emacs by default 09:06 < bridge_> (@_voxeldoesart) 09:08 < bridge_> ChillerDragon better try to convince oy to finally add new contributors 09:09 < bridge_> Maybe it would even be the best if it's no one from ddnet either 09:09 < bridge_> I try to contact oy since forever 09:09 < bridge_> He gone dark 09:09 < bridge_> Robyte was a good try. But... 09:10 < bridge_> ddnet yoinked robsti 09:23 < bridge_> 😄 09:25 < bridge_> skin renderer? 09:25 < bridge_> not open source i believe 10:35 < bridge_> its monday 10:35 < bridge_> time to compile firefox update 10:41 < bridge_> lol 10:41 < bridge_> ``` 10:41 < bridge_> 0:22.13 error: failed to run custom build command for `proc-macro2 v1.0.59` 10:41 < bridge_> 0:22.13 note: To improve backtraces for build dependencies, set the CARGO_PROFILE_RELEASE_BUILD_OVERRIDE_DEBUG=true environment variable to enable debug information generation. 10:41 < bridge_> 0:22.13 Caused by: 10:41 < bridge_> 0:22.13 process didn't exit successfully: `/var/tmp/portage/www-client/firefox-116.0/work/firefox_build/release/build/proc-macro2-de9a4003ce515990/build-script-build` (exit status: 127) 10:41 < bridge_> 0:22.13 --- stderr 10:41 < bridge_> 0:22.13 /var/tmp/portage/www-client/firefox-116.0/work/firefox_build/release/build/proc-macro2-de9a4003ce515990/build-script-build: error while loading shared libraries: /var/tmp/portage/www-client/firefox-116.0/work/firefox_build/release/build/proc-macro2-de9a4003ce515990/build-script-build: DT_RELR without GLIBC_ABI_DT_RELR dependency 10:41 < bridge_> ``` 10:42 < bridge_> firefox failed to compile 10:43 < bridge_> https://blog.rust-lang.org/2023/08/03/cve-2023-38497.html 10:43 < bridge_> https://blog.rust-lang.org/2023/08/03/Rust-1.71.1.html 10:47 < bridge_> apparently firefox fails to compile with clang + mold? 10:47 < bridge_> but im using gcc iirc 10:49 < bridge_> https://bugs.gentoo.org/911591 10:57 < bridge_> https://twitter.com/joshmanders/status/1688255209259102208 11:00 < bridge_> Are those both contracted or both noop? 😄 11:01 < bridge_> I can see NASA not wanting to die in prod 11:01 < bridge_> * ERROR: www-client/firefox-116.0::gentoo failed (setup phase): 11:01 < bridge_> * No LLVM slot <= 16 satisfying the package's dependencies found installed! 11:01 < bridge_> * 11:01 < bridge_> * Call stack: 11:01 < bridge_> * ebuild.sh, line 136: Called pkg_setup 11:01 < bridge_> * firefox-116.0.ebuild, line 523: Called llvm_pkg_setup 11:01 < bridge_> * llvm.eclass, line 246: Called get_llvm_slot '16' 11:01 < bridge_> * llvm.eclass, line 151: Called die 11:01 < bridge_> * The specific snippet of code: 11:01 < bridge_> * die "No LLVM slot${1:+ <= ${1}} satisfying the package's dependencies found installed!" 11:01 < bridge_> firefox is fucking with me today 11:01 < bridge_> ❯ sudo equery l llvm 11:01 < bridge_> * Searching for llvm ... 11:01 < bridge_> [IP-] [ ] sys-devel/llvm-15.0.7-r3:15/15 11:01 < bridge_> [IP-] [ ] sys-devel/llvm-16.0.6:16/16 11:51 < bridge_> I don't want to say anything 11:52 < bridge_> xd 12:11 < bridge_> furry 12:12 < bridge_> Who knows how to use the MousePos() function? 12:18 < bridge_> Exactly like that 12:21 < bridge_> Did you try to resync? I've managed to snipe partial updates before 12:23 < bridge_> resync? 12:23 < bridge_> i do sudo emaint -a sync 13:01 < bridge_> Hey there short question; How long does it take to register a server? I created one (and also hosted it) but it just keeps saying "I register/6/ipv6: registering..." and "I register/6/ipv4: registering..." multiple times for 10minutes now. Is this normal? Didn't find anything for it in web. 13:23 < bridge_> Another one of those yeah 13:47 < bridge_> ChillerDragon: no, teeworlds didn't want to have robyte ^^ 13:48 < bridge_> ☕ 13:48 < bridge_> it should take less than 5 seconds 13:50 < bridge_> btw, do you know nitter.net? https://nitter.net/joshmanders/status/1688255209259102208 13:50 < bridge_> i know, but im lazy 13:51 < bridge_> also i thought it stopped working 13:51 < bridge_> with api change 13:51 < bridge_> the good thing is that you can replies on nitter.net 13:51 < bridge_> without logging in 13:56 < bridge_> https://jasony.me/publication/hotos23-qdlp.pdf 13:57 < bridge_> > FIFO can be Better than LRU: 13:57 < bridge_> > the Power of Lazy Promotion and Quick Demotion 13:57 < bridge_> > LRU has been the basis of cache eviction algorithms for 13:57 < bridge_> > decades, with a plethora of innovations on improving LRU’s 13:57 < bridge_> > miss ratio and throughput. While it is well-known that FIFO- 13:57 < bridge_> > based eviction algorithms provide significantly better through- 13:57 < bridge_> > put and scalability, they lag behind LRU on miss ratio, thus, 13:57 < bridge_> > cache efficiency. 13:57 < bridge_> > We performed a large-scale simulation study using 5307 13:57 < bridge_> > block and web cache workloads collected in the past two 13:57 < bridge_> > decades. We find that contrary to what common wisdom sug- 13:57 < bridge_> > gests, some FIFO-based algorithms, such as FIFO-Reinsertion 13:57 < bridge_> > (or CLOCK), are, in fact, more efficient (have a lower miss 13:57 < bridge_> > ratio) than LRU. Moreover, we find that qick demotion 13:57 < bridge_> > — evicting most new objects very quickly — is critical for 13:57 < bridge_> > cache efficiency. We show that when enhanced by qick 13:57 < bridge_> > demotion, not only can state-of-the-art algorithms be more 13:57 < bridge_> > efficient, a simple FIFO-based algorithm can outperform five 13:57 < bridge_> > complex state-of-the-art in terms of miss ratio. 15:15 < bridge_> @ryozuki 15:15 < bridge_> New Intel processor refresh are rumored to have 6GHz turbo boost 😬 15:15 < bridge_> https://benchlife.info/raptor-lake-s-refresh-intel-core-i9-14900k-max-frequency-will-be-6ghz/ 15:17 < bridge_> Maybe my next PC is not Ryzen xdd 15:22 < bridge_> :NekoDrink: 16:01 < bridge_> @robyt3 just noticed the atlas pic, is this from ur change? 16:01 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1138109670876270736/glyphs.png 16:01 < bridge_> it looks effin cool 16:03 < bridge_> https://wiki.ddnet.org/wiki/File:Glyphs.png 16:03 < bridge_> :gigachad: 16:04 < bridge_> @ryozuki lul did u see that 16:04 < bridge_> https://github.com/Lxtharia/minegrub-theme 16:04 < bridge_> hehe, if I see that correctly, the top-left corner is populated by the letters required when the client starts ("Loading DDNet Client") 16:05 < bridge_> why the red channel 16:05 < bridge_> and not green or blue? 16:05 < bridge_> fun to see how you can make out the history of the client a little by looking at the atlas 16:05 < bridge_> true 16:05 < bridge_> lol epic 16:13 < bridge_> Because because. It has no reason XD. Probably BCS u usually say rgb 16:13 < bridge_> In earlier OpenGL in was the alpha channel 16:13 < bridge_> opengl used red channel for some stuff iirc 16:13 < bridge_> ah 16:13 < bridge_> this 16:13 < bridge_> xd 16:45 < bridge_> @jupeyy_keks this guy doesnt have vulkan 16:46 < bridge_> Ban him 16:47 < bridge_> I bet he is on arch Linux 16:47 < bridge_> Those fucker maintainers that closed my issue BCS lazy 16:47 < bridge_> Still makes me mad 16:48 < bridge_> @heinrich5991 weird, shouldnt cmake fail if its missing libs? 16:48 < bridge_> He's on ubuntu 16:48 < bridge_> Normally u cannot install VK without VK headers there, weird 16:49 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1138121654812278884/image.png 16:49 < bridge_> 😬 16:49 < bridge_> @jupeyy_keks 16:49 < bridge_> 🥀 16:49 < bridge_> road to 1k 16:49 < bridge_> Well sadly not all packages follow the PKG cfg idea 16:49 < bridge_> wasnt it always 500 16:49 < bridge_> So we even include VK_platform.h 16:50 < bridge_> I don't even know why xd 16:50 < bridge_> back in my age we had less than 200 16:50 < bridge_> and we wanted to prevent reaching 400 16:50 < bridge_> xd 16:50 < bridge_> :PES4_Old: 16:51 < bridge_> Most issues come from backward compability 😬😬😬😬😬 16:51 < bridge_> Haters in 321 16:51 < bridge_> 0, -1, -2 16:51 < bridge_> @jupeyy_keks https://github.com/ddnet/ddnet/issues/6788 16:51 < bridge_> can we close this 16:52 < bridge_> anyways yeah when i have enough days off i finally want to start making concept art for ddnet 2.0 16:52 < bridge_> Well what should I answer xd 16:52 < bridge_> yes 16:52 < bridge_> I'll not be able to reproduce 16:53 < bridge_> If u ask me close it. Or ask the guy if it's fixed 16:53 < bridge_> the easy open another issue if it still happens 16:53 < bridge_> 👍 16:54 < bridge_> where do i modify the server tickrate? 16:54 < bridge_> It's no use 16:55 < bridge_> If u do, you change the physics 16:55 < bridge_> If you don't care. U can find it in protocol.h I think 16:55 < bridge_> yeah its in protocol.h but it glitches out when there are 2 players in the server 16:55 < bridge_> and makes them teleport all over the map 16:56 < bridge_> so i was wondering how to change that 16:56 < bridge_> The client needs to change it too 16:57 < bridge_> then where do you change the client tick 16:57 < bridge_> Here, not perfect, but ok 16:57 < bridge_> 16:57 < bridge_> https://github.com/ddnet/ddnet/commit/ba0155c8686e3a3c8ce3c7b5767a0587530d7d85 16:58 < bridge_> The quantization still sucks 17:00 < bridge_> thanks 17:00 < bridge_> whats he trying to do 17:07 < bridge_> @jupeyy_keks look im a active dev 17:07 < bridge_> now ur turn 17:11 < bridge_> After holidays I continue with ddnetpg. And then i rq ddnet and only build Vulkan compute shader software xd 17:12 < bridge_> xd 17:12 < bridge_> https://github.com/ddnet/ddnet/issues/6839 17:12 < bridge_> this seems important 17:13 < bridge_> For Heinrichs quic impl maybe 17:13 < bridge_> I'd favor putting network in a different thread 17:13 < bridge_> I think it'd generally be good for correctness 17:13 < bridge_> the code as it currently is looks like it's attempting something but not doing it 17:14 < bridge_> I'd say having to rely on it isn't really correct 17:14 < bridge_> net on a thread means lot of sync right? 17:14 < bridge_> I don't think alternative implementations are that important to whether this particular issue should be fixed or not 17:14 < bridge_> Better than sys calls 17:15 < bridge_> Why is that an issue tho 17:15 < bridge_> The client should simply have min 50fps 17:15 < bridge_> In minimized u probably don't care 17:16 < bridge_> currently, this thing does: 17:16 < bridge_> ``` 17:16 < bridge_> while(total wait time has not passed) 17:16 < bridge_> { 17:16 < bridge_> sleep_unless_a_network_packet_is_waiting() 17:16 < bridge_> } 17:16 < bridge_> ``` 17:16 < bridge_> this seems incorrect no matter how I look at it 17:16 < bridge_> Yeah bro BCS u wanted it xd 17:16 < bridge_> I said just sleep 17:16 < bridge_> that's not what I wanted 17:16 < bridge_> it should sleep and check if total right? 17:17 < bridge_> it should sleep and wait for a network packet 17:17 < bridge_> if there's a network packet, process it 17:17 < bridge_> It should just sleep. Nothing more 17:17 < bridge_> then continue sleeping 17:17 < bridge_> why do you think that's better than what I propose? 17:17 < bridge_> BCS network should not rely on main thread 17:17 < bridge_> No sys calls 17:18 < bridge_> sleep is a syscall. no network sys calls? 17:18 < bridge_> why not? 17:18 < bridge_> anyway 17:18 < bridge_> this is orthogonal 17:18 < bridge_> the current code is wrong for the current architecture 17:18 < bridge_> hence I opened the issue 17:19 < bridge_> ```cpp 17:19 < bridge_> do 17:19 < bridge_> { 17:19 < bridge_> net_socket_read_wait(m_aNetClient[CONN_MAIN].m_Socket, SleepTimeInNanoSecondsInner); 17:19 < bridge_> auto NowInnerCalc = time_get_nanoseconds(); 17:19 < bridge_> SleepTimeInNanoSecondsInner -= (NowInnerCalc - NowInner); 17:19 < bridge_> NowInner = NowInnerCalc; 17:19 < bridge_> } 17:19 < bridge_> while((SleepTimeInNanoSecondsInner / std::chrono::nanoseconds(1us).count()) > 0ns); 17:19 < bridge_> ``` 17:20 < bridge_> should it be a do while? 17:20 < bridge_> no need to wait if the wait time already passed 17:20 < bridge_> no 17:20 < bridge_> A sleep has almost deterministic runtime tho 17:20 < bridge_> rly? 17:20 < bridge_> i heard it doesnt 17:20 < bridge_> Network syscalls create laggs 17:20 < bridge_> Almost = as good as it can 17:21 < bridge_> even if this is a problem, it seems orthogonal to the issue. unless the architecture is changed, the main thread does the network 17:21 < bridge_> Yes, that's the filled glyph atlas with my change, I added a temporary debug feature to render it for testing 17:21 < bridge_> hence the code is currently wrong. network syscalls will happen on the main thread 17:21 < bridge_> hence the code is currently wrong. network syscalls will happen on the main thread anyway 17:21 < bridge_> But the current network code seems fine, and as described in the issue, there are other problems 17:22 < bridge_> So if u fix it, fix it everywhere 17:22 < bridge_> And if u fix it everywhere fix the architecture 17:22 < bridge_> we can discuss the architecture separately 17:23 < bridge_> Yet another useless discussion opened 17:23 < bridge_> or we can also not discuss it if you want 17:23 < bridge_> Sounds good 17:23 < bridge_> I prefer to just code my style and then rq 17:23 < bridge_> xd 17:23 < bridge_> but if we don't change the architecture, I think that the issue I created describes a bug with the current code 17:23 < bridge_> by architecture u mean the code structure? 17:23 < bridge_> design 17:23 < bridge_> But also fixes one tbf 17:24 < bridge_> One that imo is more drastic 17:24 < bridge_> current architecture = network on main thread 17:24 < bridge_> after fixing the issue I opened, the bug you fixed should still be fixed, no? 17:24 < bridge_> after fixing the issue I opened, the bug you fixed should still be gone, no? 17:24 < bridge_> Yes, go ahead 17:26 < bridge_> i wont do it cuz idk much 17:27 < bridge_> @ryozuki try xd 17:27 < bridge_> basically you'd need to invert the logic, I think 17:27 < bridge_> sleep at the end of the main loop, go through it again if an event happened, but only render if the next render target has passed 17:29 < bridge_> it returns 1 17:29 < bridge_> if it read? 17:29 < bridge_> net_socket_read_wait is undocumented 17:32 < bridge_> yea, looks a bit complicated to get this right ^^ 17:32 < bridge_> yeah i cant wrap my head around this rn xd 17:32 < bridge_> yea, seems to return 1 if a network packet is waiting 17:33 < bridge_> ah net_socket_read_wait doesnt block? 17:34 < bridge_> > go through it again if an event happened, 17:34 < bridge_> what is an event here 17:34 < bridge_> > but only render if the next render target has passed 17:34 < bridge_> what is a render target 17:38 < bridge_> Better only handle network packets of possible. The whole main loop creates extra noise again 17:38 < bridge_> If 19:05 < bridge_> Does 17.2 client support 128 slots? 19:12 < bridge_> LOL 19:12 < bridge_> OH MY GOD 19:13 < bridge_> multiview works perfectly as intended 19:13 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1138157880122556527/screenshot_2023-08-07_19-12-17.png 19:13 < bridge_> No problem slave 19:13 < bridge_> I've good news for you, your keyboard is still functional 19:14 < bridge_> F10 is working xd 19:14 < bridge_> What does the image even show xd 19:14 < bridge_> look at my tee 19:15 < bridge_> wait im gonna record a video 19:20 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1138159770180796417/DDNet_Client_2023-08-07_19-18-13.mp4 19:20 < bridge_> sometimes it makes the game play audio 10 times louder, sometimes it breaks my tee like in the screenshot after rejoining 19:23 < bridge_> Does it happen only when you set `cl_multiview_zoom_smoothness` to a low value? Or any time you change this variable? Or is it unrelated? 19:23 < bridge_> only low values 19:23 < bridge_> LOL 19:26 < bridge_> Nice fps drops 19:27 < bridge_> I think this happens when cursor is at 0,0 19:27 < bridge_> Causing Nan in float calc 19:38 < bridge_> also u cant multiview when theres no one near but u can remove everyone from multiview so it does this weird thing idk lol 19:38 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1138164239467544718/DDNet_Client_2023-08-07_19-35-09.mp4 19:39 < bridge_> idk if its intended 19:43 < bridge_> Xd 19:43 < bridge_> Totally intended 19:43 < bridge_> If u got GH account just add it to the issue 19:53 < bridge_> good afternoon 19:53 < bridge_> no it's still morning technically 19:54 < bridge_> ChillerDragon: cool video btw 19:56 < bridge_> We could do default colors, but we tried to apply custom colors, which is harder than we've guessed 🙂 Sometimes its failing, but there are some good ones too 20:02 < bridge_> is there a way to create a random skin with the console 20:03 < bridge_> ive been trying to find it rn for a bit 20:04 < bridge_> a nvm there is no console command 20:04 < bridge_> makes sense 21:04 < bridge_> > also u cant multiview when theres no one near 21:04 < bridge_> Yes, if there is nothing you are looking at, it wont turn on. Thats intended. 21:04 < bridge_> > but u can remove everyone from multiview so it does this weird thing idk lol 21:04 < bridge_> Yes, if no tee is selected it will spectate the whole team instead 21:05 < bridge_> I will look into it thanks. 22:27 < ChillerDragon> @scrumplex sos 22:27 < ChillerDragon> 9.031 D "error_description" : "AADSTS700016: Application with identifier '549033b2-1532-4d4e-ae77-1bbaa46f9d74' was not found in the directory 'Microsoft Accounts'. This can happen if the application has not been installed by the administrator of the tenant or consented to by any user in the tenant. You may have sent your authentication request to the wrong tenant.\r\nTrace ID: 22:27 < ChillerDragon> 65352137-2fe2-4c76-8c7f-a5be862ae200\r\nCorrelation ID: 0b06044a-0209-4706-a76f-8ecc3e34d641\r\nTimestamp: 2023-08-07 20:27:07Z" 22:27 < ChillerDragon> @heinrich5991 the debugger talk? 22:28 < ChillerDragon> @heinrich5991 yea robsti did not get a warm welcome by Oy 22:29 < bridge_> ChillerDragon: yes, the debugger talk 22:34 < ChillerDragon> @scrumplex a seems to be a selfcompiled issue you probably dont ship the proper creds/config for the ms stuff 23:35 < bridge_> uh we do :O 23:36 < bridge_> our client id is this though: https://prismlauncher.org/.well-known/microsoft-identity-association.json 23:41 < bridge_> another multimc fork? why 23:54 < bridge_> prism is dope 23:54 < bridge_> multimc is a bad author 23:54 < bridge_> what does it have over multimc 23:55 < bridge_> a not angry dev for publishing it on arch aur 23:55 < bridge_> anyway its good 23:55 < bridge_> paired with modrinth 23:56 < bridge_> it also handles curse anti user rights quite well 23:56 < bridge_> he's mad that he can't publish it on extra or what 23:56 < bridge_> he is mad he used a free software license and someone other than him published a aur package 23:56 < bridge_> or smth 23:56 < bridge_> he is stupid 23:56 < bridge_> oh 23:56 < bridge_> lol 23:56 < bridge_> i think it was this, this drama 23:56 < bridge_> i forgot 23:57 < bridge_> hmm 23:58 < bridge_> can u get curse modpacks from within the prism client 23:58 < bridge_> i remember reading that without a per-user key it wouldn't be possible 23:58 < bridge_> i hate overwolf & all the other clients 23:58 < bridge_> i hate overwolf & all the other big $$$ clients 23:59 < bridge_> having to use the twitch app to download modpacks was prob the reason i didn't get into them earlier