00:18 < bridge> 200 ping playable if map have only 1 tele out 😂 04:19 < bridge> hi 04:19 < bridge> if you're reading this, thank you 06:47 < bridge> what u mean 06:47 < bridge> Gorpu 07:16 < bridge> yw 07:16 < bridge> (i have no idea what this is referencing) 07:26 < bridge> moooooooooooooooooooooooooooorning vietnam 07:27 < bridge> https://www.youtube.com/watch?v=khZq6JKVnlk 07:47 < bridge> KAFFIENE COMMITING UPSTREAM 07:47 < bridge> NO WAY 07:48 < bridge> ANDI T GETS DEF INSTA MERGED 10:36 < bridge> are there any alternatives for ddnet to this library https://github.com/ChillerDragon/teeworlds-protocol ? 10:40 < bridge> are there any alternatives for ddnet to this library https://github.com/ChillerDragon/teeworlds_network ? 10:40 < bridge> are there any alternatives for ddnet to this library https://github.com/ChillerDragon/teeworlds_network ? 11:10 < bridge> https://github.com/teeworlds-go/protocol 11:10 < bridge> there are alternatives from Chiller in pretty much any language 11:11 < bridge> assembly for example 👀 11:15 < bridge> ty bro 11:24 < ws-client> wowo jiggsel the go lib is not listed there 11:25 < ws-client> and milkeey lib is also missing 11:26 < bridge> listed where? 12:24 < ws-client> my msg did not get sent to irc omg 12:24 < ws-client> @zelnix there is a list of libraries here https://chillerdragon.github.io/teeworlds-protocol/ 12:24 < ChillerDragon> my webchat is struggeling to stay connected 12:25 < bridge> ty 13:51 < bridge> - GF can't join video session for her swiss university 13:51 < bridge> - IT support finds out that the uni blocks traffic from abroad 13:51 < bridge> - turns on VPN and just joins 13:51 < bridge> - IT support in chambles 14:17 < bridge> @essigautomat: classical remote lite. You don’t have to be in university in person. But can’t be too far either. 14:38 < bridge> I don't understand, why they don't have their own vpn, but that's not my issue 14:38 < bridge> I am not their sys admin 14:46 < bridge> https://tenor.com/view/compass-tabletennis-pingpong-developers-sysadmin-problems-compass-tabletennis-pingpong-compass-developers-gif-19175136 15:06 < bridge> I'm very grateful to you for https://gitlab.com/teeworlds-network/twnet_parser bro 15:17 < bridge> what kind of university doesn't have its own vpn ? 15:26 < ws-client> zelnix confirmed pyson enjoyer 16:15 < bun_bun> boo 16:16 < bun_bun> hi storma 16:17 < bridge> wha 16:17 < bun_bun> if this is connected to #developer thats kinda not good 16:17 < bun_bun> for a global chat xd 16:18 < bun_bun> maybe can ask murpi to add like a channel 16:18 < bun_bun> for global ddnet chat 16:19 < bridge> :justatest: 16:19 < something> :deenstar: 17:27 < bridge> I downloaded Zen Browser and was disgusted by how opinionated it was 17:27 < bridge> Got Floorp and I'm probably never using another Firefox fork 17:28 < bridge> you can't even disable some stuff in Zen but I found a way to make Floorp even more minimal (which is Zen's goal) 17:28 < bridge> if anyone knows a strong independent ~~woman~~ browser that renders CSS3 well, let me know 17:28 < bridge> i was going to try ladybird but idk 17:31 < bridge> gm devs, does Graphics' `ClipEnable` require any special handling..? i seem to get lots of segfaults because of it 17:32 < bridge> (i am clipping in-game objects) 17:37 < bridge> how did you find out it's the clipping that causes the segfault 17:37 < bridge> oh huh, seems more like it crashes because i zoom out too much 17:38 < bridge> modding the codebase, the only thing I added so far is use of `ClipEnable` and `ClipDisable` 17:39 < bridge> sometimes it crashes out of the blue, but i guess that was because i didnt properly close enable with disable 17:40 < bridge> on the side note, when i zoom out too much, it crashes too, so i guess i linked these two to the same cause 17:40 < bridge> even though it crashes when zooming out even without clipping 17:40 < bridge> 🤔 17:40 < bridge> The only thing i could imagine that could make problems is if you use a width or height of <= 0 17:40 < bridge> send a stacktrace 17:41 < bridge> uhm... how so? it gets SIGSEGV and outputs nothing, and im pretty sure i did that `cmake -DCMAKE_BUILD_TYPE=Debug ..` thing 17:42 < bridge> attach gdb 17:42 < bridge> then do bt 17:42 < bridge> brilliant 17:43 < bridge> hold on, let me test on vanilla to be 100% sure its not just my poor code 17:44 < bridge> what does vanilla have to do with that 😄 17:44 < bridge> I mean, unmodded code 17:44 < bridge> who knows what else did i break 😅 17:46 < bridge> ye, its not my client 17:46 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1374775341964595313/image.png?ex=682f46cf&is=682df54f&hm=29542866424159fcf24455805d2aa02e09357fc026761b58f3a26968c81a3260& 17:46 < bridge> repo's checked out at 19.2 i think 17:49 < bridge> can this be an SDL2 problem? cause I got only SDL3 installed along with their compatibility lib 17:55 < bridge> just realized, why is this opengl even 17:55 < bridge> i had vulkan set as default since forever 18:05 < bridge> tja 18:06 < bridge> Is the crash reproducable for you on some map? 18:07 < bridge> yes, on tutorial map? just slam `-` till it zooms out to infinity 😎 18:07 < bridge> it might be resolved for me since this problem does not appear on vulkan renderer 18:08 < bridge> I cannot reproduce it on OpenGL 3.0. How far do you zoom out? 18:08 < bridge> still not good that its there 18:09 < bridge> I am on mesa 25 too 18:09 < bridge> AMDGPU 18:09 < bridge> dunno.. maximum allowed? it starts lagging and stuff, I can almost see the whole map before it dies 18:09 < bridge> Make a screenshot or smth 18:09 < bridge> So i can estimate it 18:10 < bridge> hmm, vulkan-radeon for me 18:10 < bridge> wait what? 18:10 < bridge> `glxinfo | grep string` 18:11 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1374781605041672222/image.png?ex=682f4ca4&is=682dfb24&hm=a8a6975037c10e73694a5b84261d386a9411c6d6cc3a7330089c94b63f77f965& 18:11 < bridge> Oh ok 18:11 < bridge> oh uh, there are just multiple drivers available on arch, the one im using is called vulkan-radeon or something 18:12 < bridge> radeonsi 18:12 < bridge> but that is the only opengl driver afaik 18:12 < bridge> there are multiple vulkan drivers tho 18:13 < bridge> the last frame 18:13 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1374782225899585616/image.png?ex=682f4d38&is=682dfbb8&hm=a4aedd2c213e6dd959509ec6cd86cd946a4220120f1e06428e024a2ed280755a& 18:13 < bridge> before it died 18:13 < bridge> On which OpenGL is that now? 18:13 < bridge> oh wait, it's still alive 18:14 < bridge> 1.4.0, extremely laggy 18:14 < bridge> Well 1.4 you should not remove the zoom limit 18:14 < bridge> wait i think i know the problem 😂 I might have ran out of vram or something 18:14 < bridge> cause it did crash when i had like 3 clients open 18:15 < bridge> This crash was defs not on opengl 1.4 18:15 < bridge> it was either ogl 3.0 or 2.0 18:15 < bridge> This is what i get on ogl 3.0 18:15 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1374782637628985364/Bildschirmaufnahme_20250521_181330.webm?ex=682f4d9b&is=682dfc1b&hm=b2ae2e23932805d80fdbbe52c2bf123e4ebd9827d622d607894bf4c5f6decc79& 18:15 < bridge> yes, 3.0.0, crashed here 18:15 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1374782650644037674/image.png?ex=682f4d9e&is=682dfc1e&hm=5a101e171b47f9d563273dae06cce4a0f12ec0d959d19158621398f01e8b2386& 18:16 < bridge> this time for real 18:18 < bridge> Funny, in self compiled debug mode i also get this crash 18:19 < bridge> :susge: 18:19 < bridge> It crashes in a `__memmove_avx512_unaligned_erms` 18:19 < bridge> inside the driver 18:19 < bridge> I'll check if we do a out of bounds call, otherwise you can report it to mesa 18:20 < bridge> nobody uses opengl these days :nouis::nouis: 18:20 < bridge> apart from some people with extremely outdated setups 18:21 < bridge> very obscure bug i'd say 18:21 < bridge> In OGL 3.3 I also cannot reproduce it. 18:21 < bridge> 18:21 < bridge> The offset of the render call is 4 bytes aligned and seems to be fully in range of the buffer object 18:21 < bridge> So I assume the driver does smth invalid here 18:23 < bridge> but i dont wanna call big guys at mesa... im not even competent enough to draw a triangle :justatest::justatest::justatest: 18:24 < bridge> I mean you can just report it under fake name or smth xd 18:26 < bridge> this is the feature ive been working btw (so ctf gamers stop accusing me of cheating with zoom) 18:26 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1374785465978323014/image.png?ex=682f503d&is=682dfebd&hm=1d97a08d5b0c73d23ebad8b9b9363f5f6231ac4ad782f1c2254ee36de7821abd& 18:27 < bridge> (notice the cut-off tee at point beyond default zoom 10) 18:27 < bridge> lmao 18:27 < bridge> You play 20:9 or what 18:27 < bridge> 16:9 but the monitor is DAMN big and im sitting almost in front of it :greenthing: 18:28 < bridge> XD 18:28 < bridge> 16:9 is the aspect ratio basically everyone in this world uses 18:28 < bridge> oh right there was this funny cheating idea I had, having wm keybinds to change the aspect ratio of the game window 18:28 < bridge> Do you uses non-fullscreen? 18:28 < bridge> its so that i can play with zoom other than 10, 7 in this case 18:28 < bridge> cheating in ctf with 42:9 monitor 18:29 < bridge> without cheating 18:29 < bridge> Ah you _want_ to zoom 18:29 < bridge> off 18:29 < bridge> uff 18:29 < bridge> :banhammer: 18:29 < bridge> yes yes, sorry my eyes are all over when its 10 18:29 < bridge> i just cant see 18:29 < bridge> its that simple :greenthing: 18:29 < bridge> hes cheating to not cheat :banhammer: 18:29 < bridge> use a smaller game window :balls: 18:29 < bridge> U have to make sure to not cheat in the vertical direction too 18:30 < bridge> its basically clipping a rectangle in the middle 18:30 < bridge> just as i would see with zoom 10 18:30 < bridge> its al good 18:30 < bridge> But still I have to say this is a cheat and defs bannable 18:30 < bridge> hmm works correctly no? 18:30 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1374786418861277288/image.png?ex=682f5120&is=682dffa0&hm=7fe8af7d98df88cbf6da7775d5b88425777bed402e0271240fd3b0007a27e96d& 18:30 < bridge> looks like zoom 10 to me 18:30 < bridge> these star particles though.... they dont render on the same layer as other particles or what?? they dont give a damn about my clips 18:30 < bridge> Looks correct, since it automatically zooms in 18:31 < bridge> idk how seeing more of the map is a significant advantage 18:31 < bridge> yes there is no way to zoom with a different aspect ratio 18:31 < bridge> seeing others. 18:31 < bridge> you can react earlier 18:31 < bridge> Doesnt matter if it is significant. you have an advantage 18:31 < bridge> is the edge of the clipping rectangle any different to the edge of your screen in this sense? 18:32 < bridge> what? 18:32 < bridge> any ideas what to do about this (ive tried everything) 18:32 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1374786890552578138/image.png?ex=682f5191&is=682e0011&hm=58d131d071f052c8b742b2a05ed88747b141e3a9bf93225142746fefa0229571& 18:32 < bridge> @ar1gin btw what are you trying to do? there is no way to see more without zooming 18:32 < bridge> except dyncam 18:33 < bridge> there is, you definitely see players earlier with unlocked zoom on fng 18:33 < bridge> he tries to clip the whole ingame objects 18:33 < bridge> unlocked zoom is cheat 18:33 < bridge> I recommend you use windowed mode 18:33 < bridge> otherwise it's a cheat 18:33 < bridge> im trying to uncheat it 😎 18:33 < bridge> what does this mean? 18:33 < bridge> they're hiding tees outside the area that you'd normally see with zoom 10 18:33 < bridge> ?? the default client doesn't have it unlockd 18:33 < bridge> look at the screenshot, the tee is cut off 18:33 < bridge> basically clip everything as if they are at normal zoom 18:33 < bridge> serverside? 18:34 < bridge> client 18:34 < bridge> clientside 18:34 < bridge> screenshot would be nicer if the other tee wasn't super low opacity 18:34 < bridge> bro he posted 20 screenshots 18:34 < bridge> took me a couple seconds to see it 18:34 < bridge> but only cheatclients can unlock the zoom bro 18:34 < bridge> what is he changing then wtf 18:34 < bridge> sorry, i dont wanna join random fng servers for tests 😭 18:34 < bridge> change the map to not tutorial maybe 18:35 < bridge> anyways, i see you guys still think of this as cheating 🤔🤔🤔 18:35 < bridge> @jupeyy_keks can you join vc rq 18:35 < bridge> why are you doing this again? 18:35 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1374787695452360805/image.png?ex=682f5250&is=682e00d0&hm=411d1addeab9f8e67856cb8b680035796ba10a27691aa8ccd4086bd77017adad& 18:35 < bridge> bro i see it 18:35 < bridge> but why 18:35 < bridge> its TOO hard to see anything with zoom 10 on this big monitor (but i dont wanna uglify my gameplay with windowed mode) 18:36 < bridge> I don't think it's a big deal but it's probably for the best you don't try to argue it anyways 18:36 < bridge> solution is big black bars, surely :) 18:36 < bridge> that's actually what some people do in osu! 18:36 < bridge> yea just make the window smaller 18:36 < bridge> the game just straight up has that as a feature 18:37 < bridge> so you can be fullscreen but still have a smaller viewport 18:38 < bridge> think if this was in the official client 🤔🤔 18:38 < bridge> speaking of seeing more, would people have issues with a minimap? 18:38 < bridge> would be nice to have for sure 18:38 < bridge> probably, most likely, especially on competitive servers 18:39 < bridge> you could have the minimap not show others in competitive modes 18:39 < bridge> or take the heavy handed approach and completely disable it in competitive 18:39 < bridge> \> see more 18:39 < bridge> \> see further 18:39 < bridge> \> zero drawbacks 18:39 < bridge> in drag or gores it really shouldn't be an issue 18:40 < bridge> ig in gores it matters because you can't just spec whenever you want 18:40 < bridge> @inv41idu53rn4m3 are you doing a minimap 18:40 < bridge> and zooming in and out while playing is a bit difficult 18:41 < bridge> just thought of the idea the other day but it's not like high on my list of things I'd want to add 18:41 < bridge> I was just thinking it could help with playing drag maps since you'd need to spec less at times 18:42 < bridge> wanting to know whether the bottom of a pit is dangerous is such a common scenario 18:43 < bridge> wanted to try to draw a minimap, turned out my skills at low-level graphics are garbage and i dont know ddnets codebase enough :feelsbadman: 18:43 < bridge> sure is a kickass feature to have though! 18:43 < bridge> esp with /showall 18:44 < bridge> i'd say ddnets codebase simply makes it too hard xd 18:44 < bridge> maybe... its too hard to develop something without encroaching on every system 18:44 < bridge> its a major turn off since i wanna keep up to date and rebase every release 18:45 < bridge> perhaps, isolate it in a component and it'll be fine? 18:45 < bridge> anyway, i'll post clip feature when I finish it and gather some insight and opinions :greenthing: 19:50 < bridge> seems all that really needs to be done for QUIC is testing, timeout protection, and ensuring disconnect and max conn / throttling edgecase stuff works as intended, as well as testing windows + cross-comp 19:50 < bridge> @jupeyy_keks do you think you would be able to work in part of hein's PR to see if you can affect the model over QUIC? 19:51 < bridge> also, probably more QUIC types needed for the auth methods you added (not sure if he had considered Steam during it) 19:53 < bridge> I guess we also need to update the game in preparation, error messages for the network, like account not linked, account key invalid/incorrect, rate limit, etc. 19:53 < bridge> and decide if this should be otherwise transparent in the UI 19:57 < bridge> you mean aggounts? 19:57 < bridge> steam + email are supported 19:58 < bridge> ye okay 19:58 < bridge> i talked with hein a bit about the PR 19:58 < bridge> seems like its just annoying edge case type stuff left 20:00 < bridge> timeout protection is probably the biggest thing 20:01 < bridge> k 20:02 < bridge> I'd have said timeout protection can be done on higher level 20:26 < bridge> Why does timeout protection need to change? I think if you provide the timeout code the server should just let you resume your session. Im can't think of any vulnerabilities with that but idk. 20:26 < bridge> Why does timeout protection need to change? I think if you provide the timeout code the server should just let you resume your session. I can't think of any vulnerabilities with that but idk. 21:03 < bridge> I guess ensuring the account session is maintained, not removed 21:03 < bridge> I'm not sure how timeout really works honestly 21:06 < bridge> Then we'd need random timeout codes. Else other servers can steal account sessions ig. Dunno how current timeout codes work, but probs static? 21:08 < bridge> the timeout code generation right now is pretty good, I think it's seeded with the map name, server address, and a private seed stored on your PC. 21:09 < bridge> but I don't think you'd need to change anything about accounts to make it functionally fully random 21:10 < bridge> k 21:10 < bridge> I guess it just needs to be sent and associated and nothing elsr 21:10 < bridge> I guess it just needs to be sent and associated and nothing else 21:10 < bridge> Is timeout set on map join? 21:10 < bridge> for the server 21:11 < bridge> ideally we'd have something like a cookie storage system, where the server can just send you a timeout code and it would automatically expire and be deleted from your PC eventually 21:11 < bridge> yeah server authoritative would be better because there's one less step to the handshake right 21:12 < bridge> but client would still need to ack I guess? 21:12 < bridge> I think both are secure enough technically 21:12 < bridge> yaya 21:12 < bridge> i don't know how "reliable" QUIC is anyway 21:12 < bridge> or really what that means in networking :P 21:13 < bridge> But when we have accounts why not simply use the acc info? XD 21:13 < bridge> accounts technically are a session 21:13 < bridge> good point 21:13 < bridge> idk I guess you could yeah 21:13 < bridge> you would just be logging back in essentially 21:14 < bridge> possible optimize prediction? 21:14 < bridge> is it too slow for you? 21:14 < bridge> ye 21:14 < bridge> why 21:14 < bridge> idk 21:15 < bridge> the game lags? 21:15 < bridge> hmm doesn't tclient have a fine tuned system already 21:15 < bridge> it's not faster tho 21:15 < bridge> you can tweak it at least 21:15 < bridge> the tweaks make it slower technically 21:15 < bridge> ahh 21:16 < bridge> wouldn't you have to add some kind of prediction based on the mode 21:16 < bridge> like if someone is playing ddrace, likely held inputs with 1 hooked and in freeze will continue 21:16 < bridge> etc 21:16 < bridge> it handles that already 21:17 < bridge> the prediction changes slightly depending on the mode 21:17 < bridge> but for more complex cases it could be extended, though opinionated 21:17 < bridge> is there like "fly" prediction? 21:18 < bridge> fly on high ping still feels bad, it would be nice if a whole hammer cycle ahead could be predicted 21:18 < bridge> with some adjustable buffer that tries to figure out if a fly is happening 21:32 < bridge> thought u were talking about the game mode 22:37 < bridge> wdym, kinda talking about it? 23:32 < bridge> gonna try matrix bridge 23:45 < bridge> looks like it's not a successfull try 23:45 < bridge> nah got busy 23:59 < bridge> done