00:00 < bridge> or does it include some kind of "time until next tick" info 00:00 < bridge> oh new day already 00:00 < bridge> the client does not include that info only the server 00:00 < bridge> but does it send in a fixed rate? 00:00 < bridge> how else can the server guess if the timing was off 00:00 < bridge> the client sends the "intendedtick" along with the input 00:01 < bridge> so the server looks at its own gametick and references the input's intended tick 00:01 < bridge> sure but how does the server know the client was 2ms too late 00:01 < bridge> the server knows it's sub tick timing 00:01 < bridge> so the client sends in a fixed rate? 00:01 < bridge> no not really 00:01 < bridge> it sends whenever it feels like it 00:01 < bridge> but when the client sends 20 inputs for a single tick 00:01 < bridge> all were off or what 00:01 < bridge> the client basically sends: this packet is supposed to ariive at "12:00:00.000" 00:02 < bridge> so a timestamp? 00:02 < bridge> the server sends back the timestamp then? 00:02 < bridge> and then the server says "your packet that was supposed to arive at "12:00:00.000" actually arrived 0.120s late 00:02 < bridge> and then the server says your packet that was supposed to arive at "12:00:00.000" actually arrived 0.120s late 00:02 < bridge> your input was for tick 25, which is in 5ms 00:02 < bridge> this? 00:02 < bridge> it says "input for tick 25 was 5ms early" 00:03 < bridge> the "tick" is the timestamp here 00:03 < bridge> which client and server share 00:03 < bridge> yeah, and it sends it back in the snapshot packet i guess? 00:03 < bridge> I think it's a separate packet 00:03 < bridge> not currently 00:03 < bridge> NETMSG_INPUTTIMING 00:03 < bridge> but sent in the same packet as the snap 00:03 < bridge> it could be in the snap if bandwidth was a concern but it's not that big 00:03 < bridge> oh maybe it is 00:03 < bridge> does it flush? 00:04 < bridge> ok it doesnt 00:04 < bridge> heinrich is correct 00:04 < bridge> I think only connection establishment stuff, NETMSG_INPUT and NETMSG_SNAP flush 00:05 < bridge> i'd still say this information is already in the ping + snapshots packet timing 00:05 < bridge> anyway, I'm off to sleep. good night 🙂 00:05 < bridge> gn8 00:05 < bridge> gn 00:07 < bridge> but manually acking has one huge advantage, it includes the overhead of the network code 00:07 < bridge> 00:07 < bridge> even if it might be small, CPUs ofc also generate a bit of jitter, and this must be handled by a margin for example. 00:07 < bridge> but manually acking has one huge advantage, it includes the overhead of the network code (on server & client) 00:07 < bridge> 00:07 < bridge> even if it might be small, CPUs ofc also generate a bit of jitter, and this must be handled by a margin for example. 00:08 < bridge> (it must be handled by a margin, if it wouldnt be so accurate as with normal ack'ing ^^) 00:08 < bridge> isn't manually acking already what happens? 00:08 < bridge> my old version used ping + snapshot timing only 00:08 < bridge> ah you mean dd pg 00:08 < bridge> yeah 😄 00:15 < bridge> @totar does your issue about the prediction code also help with ping jitter, or was that unrelated? 00:15 < bridge> 00:15 < bridge> 00:15 < bridge> like that it resets the auto prediction margin all the time 00:15 < bridge> idk 00:17 < bridge> you would need to spike to 800+ ms for long enough that the m_PredictedTime smoothtime reaches a high enough value to trigger the prediction reset. which maybe takes 0.2-0.4s of input timing messages from the server. 00:18 < bridge> it can certainly happen but only for super major lag spikes, not really normal jitter 00:19 < bridge> the other issue with csmoothtime being wrong does have a huge impact on ping jitter 00:19 < bridge> csmoothtime should really be re-written because it's not very good at it's job but if you patch that one issue it becomes kinda ok 00:20 < bridge> mh ok 00:20 < bridge> i thought the reset was related to an early packet arriving 00:21 < bridge> vs slow ones 00:21 < bridge> another issue with csmoothtime is that the speed it adjusts your prediction amount is proportional to your ping. so if you have 500ms it literally adjust like 20x faster than if you have 30ms. so people with high latency will experience small jitters much worse 00:22 < bridge> reset can't really happen from a single packet, you need a real high ping connection 00:22 < bridge> well 00:22 < bridge> oh yeah 00:23 < bridge> actually maybe it can happen for a single packet if the server sends you one that says you're like 20000ms off? lol 00:23 < bridge> i swear, a good prediction timer is harder than a vulkan backend 00:23 < bridge> not joking xD 00:23 < bridge> i struggle so much with it 00:23 < bridge> hmm 00:23 < bridge> trml did a lot of good work to make the cpp version much better 00:24 < bridge> how do you react to laggs 00:24 < bridge> how fast should you consider a connection stable again 00:24 < bridge> how far back should your ping/timer data reach back to give a good automatic prediction margin 00:24 < bridge> how to deal with packet loss 00:24 < bridge> 00:24 < bridge> and so on 00:24 < bridge> so many factors 00:24 < bridge> have you found any articles describing how to implement this system? I know that other games also use client prediction in the same way as ddnet but I haven't found any that explain their methods 00:25 < bridge> only that overwatch GDC talk explains it a little bit but they do not talk about the deep specifics 00:26 < bridge> nope, i couldnt even find math/statistics that is directly connected to it. 00:26 < bridge> 00:26 < bridge> i only find things for physics that are similar to a ping jitter from the curves 00:27 < bridge> also this article but it's not a full implementation https://gabrielgambetta.com/client-side-prediction-server-reconciliation.html 00:27 < bridge> i sometimes write test cases with specific start conditions and jitter and see how it evolves xD 00:27 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1275219586714697728/image.png?ex=66c51849&is=66c3c6c9&hm=aecbd661351d79af97d0e614180b7029e089681aa990a8d6f84971634dd67ce8& 00:28 < bridge> for the prediction timer, which is basically csmoothtime in the cpp version i guess 00:30 < bridge> https://tenor.com/view/cat-touch-dont-gif-9258567 00:30 < bridge> cpp version is basically this gif 00:30 < bridge> except the fish is dropping an input message on the server 00:31 < bridge> my version still suffers insanly from when the lag initially happens, once it's stable, it's kinda ok 00:31 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1275220581100421190/lag.mp4?ex=66c51936&is=66c3c7b6&hm=5beaf7c171e17436679714078c746703ea80404742705a89cf89d837b2d251c9& 00:31 < bridge> on the right you see the ping jitter etc. 00:32 < bridge> I didn't still any insane suffering lol 00:32 < bridge> I didn't see any insane suffering lol 00:32 < bridge> ok i show you, don't laugh xd 00:32 < bridge> cpp version also has insane suffering so I don't think it can be much worse 00:33 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1275221233012834304/laggyaf.mp4?ex=66c519d2&is=66c3c852&hm=9674ab4c562ed86fe7750b26578418213ac7e5034474f68994080be74290b788& 00:33 < bridge> if you introduce delay instantly there's physically no way you can smoothly adjust without misprediction 00:34 < bridge> it is 00:34 < bridge> i still struggle to find a way to make the lag not too impactful 00:34 < bridge> but if this would be ddos, rip 00:34 < bridge> have you tested this also on cpp? 00:34 < bridge> it looks similar tbh 00:34 < bridge> it cannot deal with it without your issue being fixed xD 00:34 < bridge> ah 00:35 < bridge> but yeah it adjust to the lag faster, but you cannot really play well with the lag 00:35 < bridge> mine adjust extremly bad, but once it's adjusted it's playable 00:35 < bridge> idk I was waiting on trml to give me his ancient wisdom on what "prediction time reset" does before I delete it 00:36 < bridge> I guess we can just increase MaxLatencyTicks() to 3s without any issues but the predicton time reset can still cause issues 00:36 < bridge> I guess we can just increase MaxLatencyTicks() to 3s to fix high ping but the predicton time reset can still cause issues 00:37 < bridge> ok gotta head out, else i die by sleep loss xd 00:37 < bridge> night 00:37 < bridge> ok gn 00:37 < bridge> gn big boys 00:38 < bridge> I'm not going to sleep yet lol. it's only 5pm for me 02:32 < bridge> PikoTee? Love that guy 02:36 < bridge> Yes 02:36 < bridge> irc moment I realized too late it was an edit xd 02:36 < bridge> .\_. 02:40 < bridge> yea Pikotee @chillerdragon 02:41 < bridge> He is OG 02:43 < bridge> Also jopsti I do my gaming on the beach 02:44 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1275253979504447540/ima_12b463e.png?ex=66c53851&is=66c3e6d1&hm=11250174049132d66526c45df150710bcca4046cfb84c26a5d3bbd0a6856fda0& 03:44 < bridge> qwertz sux 04:14 < bridge> usa moment 😉 05:02 < ws-client> LMAO I just pumped teeworlds music max volume by accident in a quiet co working space. I forgot that vanilla teeworlds client has music on by default xxxxxxD then i tried to panic quit with "q" but i was in the pick a name menu xxxxD 05:02 < ws-client> resetting settings to default is dangerous 07:08 < bridge> So Close to installing teeworlds on here fr 07:08 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1275320518194626661/rn_image_picker_lib_temp_57eb46a7-931f-4f2d-a0f0-08498dd6cf1d.jpg?ex=66c57649&is=66c424c9&hm=172192bcf66af3be0d64ae8c678160c21b0f30e7172c12518782b97c5f039133& 07:09 < bridge> Gm btw 07:23 < bridge> at work i am developing a web application for monitoring and analytic such machines like fanuc / mitsubishi / headman, etc) 07:23 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1275324183185723463/image.png?ex=66c579b3&is=66c42833&hm=9be5ef2a9a6dd8e41a389ee57b3f7c167a75580cef7d4f10d78ecfcca606a276& 07:23 < bridge> Oh wow thats cool! 07:31 < bridge> poor chiller has to play with himself 07:38 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1275327958139142215/image.png?ex=66c57d37&is=66c42bb7&hm=2eb8ff329defe2463bf151b0f310b873e42506a197dc11917ece7d09475e3c39& 07:38 < bridge> Wording Freddie.. WORDING!!! 07:38 < bridge> 07:38 < bridge> By himself = alone 07:38 < bridge> With himself = {redacted} 07:38 < bridge> Tf is this 07:38 < bridge> CNC milling 07:38 < bridge> Could host a ddnet server on a cisco switch at work 07:38 < bridge> They support docker 07:39 < bridge> lmao 07:46 < bridge> clearly i meant the latter 07:47 < bridge> :justatest: 07:48 < bridge> i mean... there are two chillers on screen. definitely what i mean. 07:48 < bridge> 👀 08:23 < bridge> he must love his laptop so much. 08:46 < bridge> hello, please download the latest version of the mobile ddnet at the moment, I am grateful in advance 08:47 < bridge> it says Matodor top right? You use your gamer tag at work? 08:51 < bridge> yes, its account login) 08:51 < bridge> yes, its account name) 08:54 < bridge> gm 08:54 < ws-client> i always try not to leak my gamertag at work xd 08:59 < bridge> because you are Batman. Chiller at night Hans Müller during the day. 09:01 < ws-client> xd 09:01 < ws-client> @robyt3 09:01 < ws-client> rossbit u here? 09:02 < ws-client> idk what to do about 0.7 skin info message stuff 09:14 < ws-client> @jxsl13 so people greet you as jiggsel at work? xd 09:16 < bridge> Ive met people from ddnet in real Life a few Years ago, they Always called Me Melon :kek: 09:16 < ws-client> i would call you melon too 09:16 < bridge> I would call you drache 09:16 < ws-client> xd 09:18 < ws-client> @ryozuki when transfer ddnet points from Ryozuki to edg-l 09:32 < ws-client> @soulyvevo very epic pr btw. You are such smooth githubber heinrich and deen forgot to say something like "thanks for your first contribution" they were blown away by your professionalism and thought you were OG develeoper already. 09:38 < bridge> yes 🥲 09:39 < bridge> Chilleroni, Chiller at night, peperoni during the day :p 09:41 < ws-client> peperoni lol 09:47 < bridge> öäüß 😘 09:47 < bridge> Ñ 09:51 < bridge> chillerdragon: did you include the stencil thing that we never merged in 0.7 09:56 < bridge> ~~i actually don't like that we now intentionally used different default asset from teeworlds, but that's just me and probably doesn't matter that much.~~ 10:09 < bridge> rich experience 10:11 < ws-client> @tsfreddie nah because im not porting it to vulkan 10:11 < bridge> @jupeyy_keks help 10:11 < ws-client> indeed 10:11 < bridge> my thought exactly really 10:11 < ws-client> jopsti biggest 0.7 supported can do it 10:11 < ws-client> @tsfreddie what different default asset? 10:12 < bridge> what different? 10:12 < ws-client> you said that 10:12 < bridge> oh, the souly pr 10:12 < ws-client> ah 10:12 < ws-client> we can also send a pr in teeworlds 10:12 < bridge> i'm not sure if our 0.7 is directly copied from 0.7 to begin with tho 10:12 < ws-client> im sure oy will merge it 10:13 < bridge> helped 10:13 < bridge> chillerdragon: do you know if we actually want the adaptive marking tho 10:13 < ws-client> @soulyvevo send same pr in github.com/teeworlds/teeworlds 10:13 < bridge> i can learn vulkan and do it || probably || 10:14 < ws-client> @tsfreddie i do want it. soli does want it. 10:14 < bridge> ye that's what I mean, does anyone apart from souly wants it lol 10:14 < ws-client> oy? 10:14 < bridge> we have to update all marking assets too 10:14 < bridge> that would introduce a stencil buffer. 10:14 < bridge> i dunno if we are ready for that 10:15 < ws-client> Stiopa 10:15 < ws-client> Sonix 10:15 < bridge> it's pretty trivial in 0.7's sdl 10:15 < bridge> or are you talking about we just don't want to enable stencil buffer for this one particular use case 10:16 < bridge> i would agree enabling an entire buffer for this is a lot 10:16 < bridge> 10:16 < bridge> yeah i guess this 10:16 < bridge> plus overhead per tee 10:17 < bridge> wait. aren't scrollrect already stenciled 10:17 < bridge> clipped 10:17 < bridge> oh right 10:17 < bridge> i remember now 10:17 < bridge> wasnt there one already? 10:17 < bridge> or at least i remember some talk about it 10:17 < ws-client> i also remember some talk 10:18 < ws-client> https://github.com/teeworlds/teeworlds/issues/1398 10:18 < ws-client> this? 10:18 < ws-client> or is this what caused it xd 10:18 < ws-client> idk i care too little about the feet to search for me they all look the same 10:19 < ws-client> lmao fred 10:19 < ws-client> https://github.com/teeworlds/teeworlds/issues/1400#issuecomment-139928311 10:24 < bridge> but it looks cool tho 10:24 < bridge> :kek: 10:24 < bridge> yeah 10:25 < bridge> raytracing also looks cool 😬 10:25 < bridge> call it DDNet2 so the fps drop is expected 10:26 < bridge> in my opinion the 0.7 skin system isnt good enough anyway. 10:26 < bridge> 10:26 < bridge> changing some markings is cool yes. 10:26 < bridge> 10:26 < bridge> But look at our "normal" database for skins: 10:26 < bridge> https://ddnet.org/skins/index.php?search=%24type%3Anormal 10:26 < bridge> for 99% of the skins the 0.7 concept simply aint working 10:26 < bridge> i thought that's why we kept both 10:27 < bridge> if you really want to do a modular skin system you have to plan what you need before hand. 10:27 < bridge> 10:27 < bridge> 10:27 < bridge> like if u want hats, your base skins must never include hats or ears or other stuff that blocks hats 10:28 < bridge> yeah but that is quite some work.. also ppl seem to enjoy having a database behind it. the 0.7 database defs has more overhead 10:28 < bridge> it needs to query per changable item e.g. 10:28 < bridge> yeah but that is quite some work.. also ppl seem to enjoy having a database behind it. the 0.7 database defs has more overhead (would have9 10:28 < bridge> yeah but that is quite some work.. also ppl seem to enjoy having a database behind it. the 0.7 database defs has more overhead (would have, if it existed) 10:28 < bridge> imo 0.7 skins are already planned out too bad. 10:29 < bridge> ears shouldnt be part of the main skin 10:29 < bridge> ears shouldnt be part of the base skin 10:29 < bridge> i don't think we need to have whether 0.7 skins are good idea just to avoid stencil 10:29 < bridge> i don't think we need to have whether 0.7 skins are good idea convo, just to avoid stencil 10:30 < bridge> i'd like to stick to whether we should try to implement the adaptive markings 10:31 < bridge> then you have adaptive markings and then? 10:31 < bridge> 10:31 < bridge> u still cant have hats xd 10:31 < bridge> 10:31 < bridge> it sounds to me like this is just one idea of a new better skin system generally 10:32 < bridge> i have no words. i won't touch it then 10:33 < bridge> i wont prevent you from. but you should make clear that we add a buffer that has to be cleared every frame, and has overhead for every tee rendered 10:33 < bridge> for markings xd 10:33 < bridge> on 0.7 10:33 < bridge> xDD 10:34 < bridge> @tsfreddie the idea is cool tho 10:34 < bridge> i mean i kinda want your help 10:34 < bridge> i don't think i can make vulkan work in a short amount of time 10:34 < bridge> i'd love to see some mmo teeworlds that allows to wear custom clothes etc 😄 10:35 < bridge> if it is not immensely complicated we could benchmark a bit to weight our options before merging. 10:35 < bridge> ahaha thats one way to revive teecity! 10:36 < bridge> it's dead? 10:36 < bridge> I dont See many people playing it - its AFK city 10:37 < bridge> It was really cool Back then tho - No Zoom meaning you actually Had to follow the signs, getting oneshot by npcs etc 10:37 < bridge> i mean it's not like ddnet currently runs terribly. *unless it is, cuz i haven't played it for quite some time now* 10:39 < bridge> yeah but e.g. stencil buffers are not core in ogl 1.x, so we have to make sure it runs on every hardware and stuff like that 10:39 < bridge> it's quite a bit of complexity 10:39 < bridge> if u want to try out, you can also start with opengl 10:40 < bridge> we already did in 0.7 10:40 < bridge> and its not merged, and not realed to public 10:40 < bridge> and its not merged, and not released to public 10:41 < bridge> not merged sure, not sure how released to public relates to anything. 10:42 < bridge> you could say it's not thoroughly tested, that i would agree. 10:42 < bridge> how does it play together with multi sampling? 10:43 < bridge> that's why i think we need your help 10:44 < bridge> xd 10:44 < bridge> iirc in 0.7 i drew the border on top of the markings to hide the edges so i actually never checked how it looks by itself 10:48 < bridge> i guess it's no secret that i'd prefer if we don't depend on 0.7, which is a dead project for me. 10:48 < bridge> if we want a better skin system, then it's not 0.7 ones IMO 10:56 < bridge> 10:56 < bridge> we know that 10:58 < bridge> at least you didnt explode 10:58 < bridge> 11:36 < bridge> maybe having skin system with attachment feature (hats, glass, earing, glowing effect or particles) would be better than having 0.7 customized skin system that even not work on most of DDNet's skins that currently exist in skins database 11:41 < bridge> if chiller can add 0.7 into ddnet, someone could add new skin system too. it's just no one took the liberty of designing the thing, probably due to fear of potentially throwing a huge amount of efforts to waste. 11:41 < bridge> tru 😄 11:41 < bridge> 11:41 < bridge> and you need artists that help you design your concept 11:42 < bridge> i also never thought chiller's 0.7 pr could go anywhere 11:42 < bridge> and i'm apparently wrong 11:42 < bridge> i just hope we get android & accounts this year 11:42 < bridge> and don't put too much focus on other things 11:43 < bridge> I like hope 11:43 < bridge> i wish i could help, but my life is weird rn 11:43 < bridge> 2024 year when ddnet gets accounts 11:43 < bridge> and apparently has been for years 11:43 < bridge> we have 4 months left 11:44 < bridge> 2024 year when ddnet gets accounts_FINAL_123123_TEST.psd 11:44 < bridge> speedrun time 11:44 < bridge> 2024 year when ddnet gets accounts_FINAL_123123_TEST_2025.psd 11:44 < bridge> does chiller work? 11:44 < bridge> no, he chills on the beach 11:45 < bridge> i wonder how he can pump out so many teeworlds projects 11:45 < bridge> since 1 year 11:45 < bridge> 11:45 < bridge> he is paid by the 0.7 consortium 11:46 < bridge> maybe i should stop throwing myself into difficult situations and just accept that I should have a normal job 11:46 < bridge> 11:46 < bridge> he secretely lives inside a tw map, probably with a beach and a hut 11:46 < bridge> i forgot to have lunch again... 11:47 < bridge> it's 5:46pm 11:47 < bridge> don't have lunch at that hour, have dinner XD 12:05 < bridge> maybe focusing on android release and accounts is best purpose for now other thing will get implementation by time, so far DDNet really improved from time that I was played for first time I think that was DDNet 6 or 8 with DDRace tournament of map Kobra 2 I think 12:05 < bridge> maybe focusing on android release and accounts is best purpose for now other thing will get implementation by time, so far DDNet really improved from time that I was played for first time I think that was DDNet 6 or 8 with DDRace tournament of map Kobra 3 I think 12:27 < bridge> @diariesvexar: don’t worry our best man is working non stop on android 12:28 < bridge> @headshotnoby: when port ddnet to android 12:29 < bridge> DDNet RUS1 - Oldschool 12:29 < bridge> Address: ddnet://37.230.162.50:8304 12:29 < bridge> My IGN: паша кру 12:29 < bridge> 37.230.162.50:8304 is an official DDraceNetwork (DDNet, Region: 🇷🇺) server. 12:29 < bridge> Quick join: 12:30 < bridge> I thought Robyt working on android :justatest: 12:30 < bridge> :lol: dont trick the system 12:31 < bridge> Yes he is 12:31 < bridge> He is the „best man“ I was talking about 12:31 < bridge> yeah I agree 12:31 < bridge> I just pinged headbot because he ported ddnet to every console on the planet 12:31 < bridge> he's hard trying on ddnet 12:31 < bridge> Oh yeah I remember now 12:32 < bridge> I don't know that consoles but we have ddnet on all of them 12:32 < bridge> Yeah xd 12:32 < bridge> I think he's trying to port on smart watches 12:33 < bridge> cuz we want to play DDNet in bus or in toilet 12:33 < bridge> I play zCatch on my smart fridge thanks to headbot 12:33 < bridge> thank you Headshot for porting DDNet to all kind of devices 12:34 < bridge> do we have it on smart tv ? 12:34 < bridge> Don’t think so .\_. 12:34 < bridge> Just attach a pc duh 12:34 < bridge> hmm I think it's need android version first 12:34 < bridge> Smart tv = android 12:34 < bridge> Somehow 12:34 < bridge> of course I don't have even any kind of console just pc 😄 12:34 < bridge> yeah I noticed after I asked 12:35 < bridge> @diariesvexar: when become ddnet contributor and help with android version? 12:35 < bridge> is there any wiki or guide to learn ddnet UI edit ? 12:35 < bridge> Let’s go 12:36 < bridge> There is a smol ddnet dev wiki 12:36 < bridge> I can just test it on arm-8a as 64 cpu if robyt need it 😄 12:36 < bridge> And a blog from king Edgar 12:36 < bridge> oh Ryo educated that too hmm 12:37 < bridge> then I should check it 13:33 < bridge> chillerdragon: are you available for hire 13:48 < ws-client> @tsfreddie mostly not. What do you need? 13:48 < bridge> nothing, just asking lol 13:49 < ws-client> ?xd 13:49 < ws-client> why 13:49 < bridge> wondering what do you do normally 13:49 < ws-client> oh thats confidental 13:49 < bridge> german rich spy 13:49 < bridge> chillerdragon is ur pc sand proof 13:50 < bridge> code teeworlds for fun when not spying 13:50 < ws-client> @reitw dont think so. But i usually dont put it IN the sand 13:50 < bridge> do you own that island 13:50 < bridge> yes but wind can put sand on urs, so when u play tw on the beach u risk it 13:50 < ws-client> @tsfreddie i can not comment on that 13:51 < ws-client> @reitw its like smoking. Everything that can kill you is cool. 13:51 < bridge> chillerdragon: have you been interrogated by russian spy and deemed neither an asset nor a threat. 13:52 < ws-client> stop doxing me i wont reveal anything 13:52 < bridge> ok then 13:53 < bridge> @blaiszephyr :issou: 13:53 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1275422431523110933/image.png?ex=66c5d533&is=66c483b3&hm=2c5d2e71e1140334359fdf80a5cabdd96565f598c679e4be8ac78d5ae9a1a401& 13:53 < bridge> lets bring a ddnet server on a Cisco sw 13:55 < bridge> LETSGOOOOOO 13:55 < bridge> what 15:07 < bridge> Chiller is 1000000% an international arms or drugs trader. He has like 500 different emails, refuses to give his phone number to any service 15:07 < bridge> Don’t ask too many questions or he’ll have you terminated 15:08 < bridge> okay 15:17 < bridge> Mafia boss 15:17 < bridge> 102% 15:17 < bridge> Prob why he works in different places, like forest or beach 15:17 < bridge> :KEKW: 15:17 < bridge> He's hiding 15:17 < bridge> Also the fluctuating timezones that are wildly apart 15:18 < bridge> we need to double check all his commit to see if he planted backdoor in ddnet to spy on players 15:18 < ws-client>