03:36 <+bridge_> @robyt3 btw i found that the default value of snd_buffer_size (512) delays the sound a lot in the web version, setting it to 256 fixes the issue. using the minimum of 128 doesnt work in a browser, it throws an error. 03:52 <+bridge_> it may be required 10:01 <+bridge_> I wonder if I can crash clients by sending them to max int in a map 10:02 <+bridge_> because map size itself is unsigned 10:29 <+bridge_> quads break already before x=100.000 10:29 <+bridge_> literally unplaylable 10:30 <+bridge_> the editor doesn't allow me to create maps bigger than 100.000 🙁 10:32 <+bridge_> "DDNet is an open world racing game …" :happy: 10:32 <+bridge_> but game is still stable ✅ 10:32 <+bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1514547822815346698/screenshot_2026-06-11_10-31-30.png?ex=6a2bc413&is=6a2a7293&hm=ccebf1882ca88008d6489979f3678b5bb043b1d1395132d97573e8c8388e73f6& 10:33 <+bridge_> doesn't say infinite, I can imagine making any game infinite - meaning hardware limited 10:33 <+bridge_> since server allow you to proxy to others, you could implement infinity by moving the player to another map every 100.000 chunks :p 10:34 <+bridge_> you can also host multiple maps on the same server with a mod 10:34 <+bridge_> I guess with this system you are limited to the number of ports x number of ipv6 adresses 10:35 <+bridge_> (and you need to convice the whole world to host ddnet servers) 10:35 <+bridge_> you don't need a mod for this, just run multiple server processes in parallel 10:36 <+bridge_> this map has 2 x 100.000, I guess a normal player runs out of vram if we introduce any height 10:37 <+bridge_> this map has 3 x 100.000, I guess a normal player runs out of vram if we introduce any height 10:37 <+bridge_> oh maybe not, I wonder if this is optimized away 10:38 <+bridge_> so normal ram is the issue 10:39 <+bridge_> just put your swap file into a cloud … noob error, srsly! 10:39 <+bridge_> for every chunk explore u gotta pay up $1 10:40 <+bridge_> ram becoming expensive so i just allocate my 60TB HDD as swap 10:40 <+bridge_> Most maps are quite compressible, and especially something that big 10:41 <+bridge_> do you remember, that I made a map generator? I can put a 100.000 x 100.000 labyrinth in it and add just one exit for the next chunk 10:41 <+bridge_> Just store compressed chunks and decompress on demand 10:41 <+bridge_> do you remember the map generator i made? 10:41 <+bridge_> no 😄 10:41 <+bridge_> or make procedural maps, problem solved 10:41 <+bridge_> i have a demo somewhere of me completing the 10,000x10,000 map in 51h 10:41 <+bridge_> well i lost it 10:42 <+bridge_> but i sent it here someday 10:42 <+bridge_> cant find it anymore xd 10:42 <+bridge_> can you somehow filter for file types? 10:42 <+bridge_> discord shit 10:42 <+bridge_> so I expect you to finish a 100.000x100.000 maps in 5100 h 10:42 <+bridge_> I wonder if you die of old age before 10:43 <+bridge_> Maybe we add an "ancestor tag" along with the clan tag, so multiple generations can work on this 10:44 <+bridge_> And you could totally expand beyond 100k with floating origin fun 10:44 <+bridge_> Next I wonder if I can still do bigger maps with twmap 10:44 <+bridge_> because this limit surely is not in the maploader, just in the editor buttons 10:45 <+bridge_> I feel like converting to floating origin would be more reasonable than some sort of arbitrary precision fixed point 10:46 <+bridge_> floating point looses precision, so you'd see other bugs instead 10:46 <+bridge_> for example minecraft has some far lands that are really buggy/laggy 10:46 <+bridge_> Hence floating origin 10:47 <+bridge_> would be way easier to support a chunk tree structure where the origin resets when switching chunks 10:47 <+bridge_> Probably 10:47 <+bridge_> this all is just a throught experiment btw - I am just testing rendering 😄 10:48 <+bridge_> Yeah I don't expect anyone to make a map a million tiles across 10:49 <+bridge_> I expect some mods to do some weird teleport shenanigans 10:49 <+bridge_> bigger issue is, that physics have floating point shenanigans as well, so the same part might work in one place, but doesn't in the other 11:01 <+bridge_> @teero777 do you have RenderDoc running? 11:02 <+bridge_> I am pretty sure we are rendering entities off screen, if we activate /showall 11:03 <+bridge_> and this is honestly not intended, yes the client then knows about the entities and can render them in a demo, but it should not render them ingame if you're not looking at them 11:16 <+bridge_> i could 11:16 <+bridge_> but cant you? 11:16 <+bridge_> would have to set it up again :justatest: 11:16 <+bridge_> sorry for my laziness 11:22 <+bridge_> i meant this one 11:22 <+bridge_> but i cant find the demo 11:24 <+bridge_> really grainy video xD 11:39 <+bridge_> wdym set up? what is there to set up xd 11:53 <+bridge_> download the software, hook up ddnet 12:04 <+bridge_> omg I got trolled by the antiping - how does this work 12:08 <+bridge_> I guess predicted projectiles have a different rendering loop somewhere 12:09 <+bridge_> yes, directly above ... 12:13 <+bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1514573281485918311/image.jpg?ex=6a2bdbc9&is=6a2a8a49&hm=52052000600ca5531a3f0bc7ebf3a24c482976ac29f2c3a232f091b78da53af1& 12:13 <+bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1514573282026717345/image.jpg?ex=6a2bdbc9&is=6a2a8a49&hm=dcbf0338bd881591097904f755393c695d7a23a92b6e0fee69c5d7fb804ed3bb& 13:15 <+bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1514588758035665057/65a9dcc3a3ee2675efa107f0b9ab3913.jpg?ex=6a2bea33&is=6a2a98b3&hm=51925bc71fe43021cce4840bfed981290626490a3fa2910c6136d0d024c592fc& 13:22 <+bridge_> `IsVisisble` .... 13:26 <+bridge_> xd 13:27 <+bridge_> As long as I am maintainer, I'll not get rid of the spellchecker xD 13:49 <+bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1514597522298114088/image.jpeg?ex=6a2bf25d&is=6a2aa0dd&hm=81a9c08a305a553d3ef85a926cee39ff87174b0997cb092d5d31141439f29ab2& 13:49 <+bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1514597522671276166/image.jpeg?ex=6a2bf25d&is=6a2aa0dd&hm=ca6d81959b115b036da5d139b9905c557707fd5b98f6af931b7497bd56ca72a2& 13:58 <+bridge_> https://github.com/jart/cosmopolitan 13:58 <+bridge_> til 14:00 <+bridge_> ofc with a honeybadger 14:06 <+bridge_> nice stars 14:06 <+bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1514601601367543990/image0.png?ex=6a2bf629&is=6a2aa4a9&hm=8edaf81f50ca3c8649a116835b942293144f5a48f3301810d784da87401a8116& 14:12 <+bridge_> what happened that I fetch so many issues about uninitialized data now in unrelated code 14:16 <+bridge_> > I started a project called Cosmopolitan which implements the αcτµαlly pδrταblε εxεcµταblε format. I chose the name because I like the idea of having the freedom to write software without restrictions that transcends traditional boundaries. 14:16 <+bridge_> He should have learned foreign alphabets first :D 14:20 <+bridge_> @fokkonaut at https://github.com/ddnet/ddnet/blame/ba95a4455e7a1ac161081e701c0698fac7bd46b0/src/game/client/prediction/entities/laser.cpp#L216 14:20 <+bridge_> 14:20 <+bridge_> We don't initialize the m_Predict field of CLaserData. I assume this should be set to true here? I don't know much about this or what it does 14:22 <+bridge_> also @nuborn (since you wrote that 3 years ago) 14:27 <+bridge_> this flag looks unused in this context to me ... 14:30 <+bridge_> yes it should be `true`, even tho the value is never used after this 14:35 <+bridge_> omg pyson such a troll lang ngl 14:35 <+bridge_> `pytest .` fails 14:35 <+bridge_> but `pytest tests/foo_test.py` passes 14:35 <+bridge_> smh running two tests breaks because one has side effects on the other wtf 14:36 <+bridge_> somehow default parameters in python have memory xd 14:39 <+bridge_> https://paste.zillyhuhn.com/NR 14:39 <+bridge_> prints 14:39 <+bridge_> 14:39 <+bridge_> ``` 14:39 <+bridge_> t1: [6, 7] 14:39 <+bridge_> t2: [6, 7] 14:39 <+bridge_> ``` 14:39 <+bridge_> because both objects use the same array 14:39 <+bridge_> and smh this even works across unit tests xd 14:40 <+bridge_> from different files 14:55 <+bridge_> This one literally gets everyone 14:58 <+bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1514614786774532178/glasses-c-developer-learning-python-learning-c-python-developer.jpeg.jpg?ex=6a2c0271&is=6a2ab0f1&hm=d6df169245f7fb4f2bbc9c173d794934d70da1e5a13f8a7035a7388117ecdde7& 15:07 <+bridge_> math is so weak 15:07 <+bridge_> so useless 15:08 <+bridge_> this one I remember to avoid nowadays but I still get pranked by lambda captures of loop variables from time to time 15:09 <+bridge_> what's your beef with math 15:11 <+bridge_> every time i do anything i get hit with "blah blah this isn't possible you should use this approximation that converges to 99.9% of the real solution in 10000 steps" 15:12 <+bridge_> it's soooo slowwww 15:12 <+bridge_> math has let me down so many times 15:15 <+bridge_> Stop letting down math and go solve ∫(sin(x)/x) 15:17 <+bridge_> https://en.wikipedia.org/wiki/Risch_algorithm :Oldge: 15:19 <+bridge_> this won't help :happy: 15:21 <+bridge_> the Risch algorithm is AFAIR the way to proof that it cannot be solved :kek: 15:24 <+bridge_> Yes, but he shall solve it! 15:31 <+bridge_> > The complete description of the Risch algorithm takes over 100 pages. 15:31 <+bridge_> -# https://en.wikipedia.org/wiki/Risch_algorithm 15:31 <+bridge_> :> 15:35 < multiplebedbugJ> Hi Guys! It's Madeleine Czura! Just thought I'd leave my number here in case you're lonely ;) . 15:35 < multiplebedbugJ> You can reach me on +44-7599248843 or maddyczura@gmail.com or madeleine.czura@arcadis.com anytime! 15:35 < multiplebedbugJ> Linkedin: uk.linkedin.com/in/maddy-czura 15:35 < multiplebedbugJ> Instagram: instagram.com/maddy_czura 15:35 < multiplebedbugJ> Brothers: Benjamin V Czura(Ben), Timothy P Czura(Tim), Greg Czura 15:35 < multiplebedbugJ> Sisters: Eleanor Lousie Czura, Josie Czura 15:35 < multiplebedbugJ> Mom: Lucy P Czura 15:35 < multiplebedbugJ> Dad: Peter Christopher Czura 15:35 < multiplebedbugJ> Dad Emails: peter.czura@ntlworld.com, peter.czura@corrigenda.co.uk 15:35 < multiplebedbugJ> Business Address: Top Floor, 164 Plashet Road, London, United Kingdom, E13 0QT 15:35 < multiplebedbugJ> Home Address: Comrie, 8 Southampton Road, Fareham, Hampshire, United Kingdom, PO16 7DY 15:36 <+bridge_> @Discord Mod 15:36 <+bridge_> frickin.. 15:39 <+bridge_> thanks :D 16:10 <+bridge_> @learath2 what do you think about asymmetry in humanoid robots? 16:11 <+bridge_> I don't think I'm familiar with that term 16:11 <+bridge_> with what term xd 16:12 <+bridge_> "asymmetry in humanoid robots" 16:12 <+bridge_> Are we talking about actual asymmetry? Like shapewise? 16:12 <+bridge_> yes 16:12 <+bridge_> Oh, idk I don't really have an opinion on it? Am I supposed to? 16:12 <+bridge_> Sounds wrong, humans are quite symmetric mostly 16:14 <+bridge_> yes that's why i was wondering if there are positives about designing a robot asymmetrically 16:16 <+bridge_> Hmmm, idk, I can't really think of any, maybe for very niche applications having different arm lengths could be useful somehow. Otherwise if you expect it to move like a human it's not really possible to introduce much of it 16:18 <+bridge_> every thought i have about how asymmetry could be better than symmetry basically just falls apart if you build the robot modular enough 16:18 <+bridge_> What are some of the ideas you came up with? 16:22 <+bridge_> you could use one of the arms for extra battery storage, so it is heavier but provides extra power for anything powerful on the other arm. 16:22 <+bridge_> 16:22 <+bridge_> another idea would be some special kind of movement, some kind of hook or similar on one arm. that arm has to be more robust than the other one. so one arm for tools and the other for movement. but who says you couldn't have both on both arms xd 16:23 <+bridge_> often when i see concept art of robots they are asymmetrical and it looks really cool, not just copy paste 2 arms 2 legs 16:26 <+bridge_> It's probably not a good idea to place dead mass in moving parts 16:26 <+bridge_> wouldn't it make the robot with kinda bad mass center 16:26 <+bridge_> and encourage people to you microslop os? /s 16:27 <+bridge_> doesn't matter if you account for it 16:27 <+bridge_> wdym dead mass? 16:28 <+bridge_> arms are moving quite often, you always accelerate the batteries in different directions, this costs energy 16:29 <+bridge_> oh, ye 16:29 <+bridge_> depends, ofc. A boxer would appreciate it 16:30 <+bridge_> @milkeeycat the minimum version in the error message is 1.100.0, but in your PR you are checking against 1.65.0 16:30 <+bridge_> ehh see most of the thing like cars planes and living structures are symmetrical and it's basically better for movement 16:30 <+bridge_> in case of those big building machines yes? if robot is big enough and well made it can be good choice for some sort of more practical benefits 16:31 <+bridge_> I don't have `rustc` that's < `1.63.0`, but it'd be the same error message 16:31 <+bridge_> depends on the purposes 16:31 <+bridge_> so the error you are showing in the PR screenshot is not the one that ddnet would produce? 16:32 <+bridge_> then how do you arrive at this number? 16:32 <+bridge_> at what number? minimum rust version? 16:32 <+bridge_> yes 16:33 <+bridge_> https://github.com/search?q=repo%3Addnet%2Fddnet%20%22rust-version%20%3D%20%22&type=code 16:33 <+bridge_> 16:33 <+bridge_> https://github.com/ddnet/ddnet/issues/11995 16:33 <+bridge_> > Our current MSRV is 1.63.0. 16:33 <+bridge_> like it's certainly a sane minimum, but why this specific version 16:35 <+bridge_> thanks 16:39 <+bridge_> I updated the image 16:40 <+bridge_> Yeah small niche modifications of the arm without disturbing the center of mass is probably the limit of what can be asymmetrical without disturbing normal humanoid mechanics. Batteries and stuff can go on the back where they don't disturb left-right symmetry 16:40 <+bridge_> thank you, this is already LGTM but I am not a rust maintainer 😄 16:54 <+bridge_> i wish i could be a rust maintainer 16:55 <+bridge_> > Surface Acoustic Wave (SAW) Bolts: Companies like Sensideon engineer smart fasteners using SAW transponders. The bolt head contains a piezoelectric crystal. When a remote reader fires an RF signal at the bolt, the crystal converts it into an acoustic wave that travels through the stressed metal and bounces back. Because tension alters the crystal's properties, the returned radio frequency instantly changes based on the exact clamping force. 16:55 <+bridge_> 16:55 <+bridge_> woah this is cool 16:56 <+bridge_> you could basically get a robot to feel touch if you mount all the panels with these bolts. idk how sensitive they are tho 16:57 <+bridge_> hydrophones work this way 16:57 <+bridge_> (the ones on submarines) 16:57 <+bridge_> So it works very good :happy: 16:57 <+bridge_> So it works very good™ :happy: 16:58 <+bridge_> for what? to run `rm -fr`? 16:58 <+bridge_> i'd consider rust to be my most proficient language :p 16:59 <+bridge_> i dont feel comfortable being a C++ maintainer tho, too many things i dont know 16:59 <+bridge_> WHAT 16:59 <+bridge_> noooo 16:59 <+bridge_> i did start learning rust when i got myself hyped for ddnet-rs :FrogSip: 16:59 <+bridge_> Mine was `.BAT` back in these days, that's not an argument :p 16:59 <+bridge_> and now i'm sort of stuck with it 16:59 <+bridge_> And exactly this is rust: hype. 17:00 <+bridge_> rewrite the rust part in go 17:00 <+bridge_> and be go maintainer 17:01 <+bridge_> I just use a cheat sheet and program in anything I want 17:01 <+bridge_> true 17:01 <+bridge_> with exceptions like haskell :justatest: 17:02 <+bridge_> im struggling with lisp xd 17:02 <+bridge_> its so many brackets 17:02 <+bridge_> its so many parentheses 17:02 <+bridge_> pro tip, the IDE does these automatically 17:02 <+bridge_> i dont think Rust is a hyped-up language 17:02 <+bridge_> zig on the other hand, idk why people praise it like its a gift from jesus 17:03 <+bridge_> its a cool language but its syntax is making me want to commit a crime 17:03 <+bridge_> and that comes from someone who enjoys rusts weird syntax 17:03 <+bridge_> I guess you need to come from C++ to like it xD 17:05 <+bridge_> whats an ide 17:06 <+bridge_> it's a program where you can install a plugin to make it like vim 17:06 <+bridge_> oh i heard about that, emacs wasnt it? 17:06 <+bridge_> ohh so its like a worse version of a text exitor? 17:06 <+bridge_> yeah 17:07 <+bridge_> @teero777 uses `echo` 17:07 <+bridge_> Or just a cable for shortcircuiting pins on a serial port :p 17:08 <+bridge_> just use a morse-key and write your code in binary atp 17:09 <+bridge_> `echo "#include \nint main(void){printf(\"hiii\\n\");return 0;}"` 17:09 <+bridge_> does echo even work with \n 17:09 <+bridge_> with a switch 17:09 <+bridge_> echo -e 17:09 <+bridge_> ah ye 17:10 <+bridge_> butterfiles 17:10 <+bridge_> butterflies 17:11 <+bridge_> i recently had a stripped debian install 17:11 <+bridge_> it didn't have any editor installed 17:11 <+bridge_> not even ed 17:11 <+bridge_> like wtf 17:21 <+bridge_> no vi? 17:25 <+bridge_> We're talking about editors. vi is a great operating system (… with a lousy editor) :happy: 17:26 <+bridge_> https://en.wikipedia.org/wiki/Vi_(text_editor) :grrr: 17:26 <+bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1514652104956842176/image.png?ex=6a2c2532&is=6a2ad3b2&hm=8850ff271ab09b5af77bf9d1b18573ee61f1029b2be181a2f73d3412e360e5d0& 17:27 <+bridge_> the joke goes the other way around 17:27 <+bridge_> or is this ragebait 17:27 <+bridge_> never heard that with vi, only emacs 17:27 <+bridge_> vi vs. emacs is a classic war 17:28 < bewilderedsquir> Hi Guys! It's Madeleine Czura! Just thought I'd leave my number here in case you're lonely ;) . 17:28 < bewilderedsquir> You can reach me on +44-7599248843 or maddyczura@gmail.com or madeleine.czura@arcadis.com anytime! 17:28 < bewilderedsquir> Linkedin: uk.linkedin.com/in/maddy-czura 17:28 < bewilderedsquir> Instagram: instagram.com/maddy_czura 17:28 < bewilderedsquir> Brothers: Benjamin V Czura(Ben), Timothy P Czura(Tim), Greg Czura 17:28 < bewilderedsquir> Sisters: Eleanor Lousie Czura, Josie Czura 17:28 < bewilderedsquir> Mom: Lucy P Czura 17:28 < bewilderedsquir> Dad: Peter Christopher Czura 17:29 < bewilderedsquir> Dad Emails: peter.czura@ntlworld.com, peter.czura@corrigenda.co.uk 17:29 < bewilderedsquir> Business Address: Top Floor, 164 Plashet Road, London, United Kingdom, E13 0QT 17:29 < bewilderedsquir> Home Address: Comrie, 8 Southampton Road, Fareham, Hampshire, United Kingdom, PO16 7DY 17:29 <+bridge_> @Discord Mod 17:29 <+bridge_> how many times has this happened already 17:29 <+bridge_> nope 17:29 <+bridge_> too many times 17:30 <+bridge_> ChillerDragon: would a PoC help you understand #12271? 17:30 <+bridge_> https://github.com/ddnet/ddnet/issues/12271 17:35 <+bridge_> vterm in vi/vim is terrible ngl 17:36 <+bridge_> i still do use nvim as the shell for alacritty though since there isn't anything better really 17:36 <+bridge_> don't know, to be honest. I'm nerd enough to use arch, but I was never nerd enough for all this terminal stuff. It's okay for me to use geany :happy: 17:36 <+bridge_> bro 17:36 <+bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1514654667701747773/1.jpg?ex=6a2c2795&is=6a2ad615&hm=7f9a597a536e056eea7a8c84413910fbf253428977a2b114bb70695c30dbff65& 17:36 <+bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1514654667991285960/2.jpg?ex=6a2c2795&is=6a2ad615&hm=07a369cebac52cb192b6c0dbbc0ded450a78e56e786a9bc92674a886cce07d6b& 17:36 <+bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1514654668339417188/3.jpg?ex=6a2c2795&is=6a2ad615&hm=d14ef2b4cabf793dfb13c35e031fec26d599ce2e5c26a5a0b124c2fb5220efec& 17:36 <+bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1514654668679020574/4.jpg?ex=6a2c2795&is=6a2ad615&hm=e4f08828f9f122c659560ee82f7fa140291df697a58dd9257f297dd70ce5ec50& 17:37 <+bridge_> smart 17:37 <+bridge_> i handle a ticket for 2 minutes, come back and see someone is giving out free money again 17:37 <+bridge_> rude 17:37 <+bridge_> always late to the party 17:37 <+bridge_> i already claimed it ez 17:38 <+bridge_> i bought 4000 core supercomputer from mrbeast money 17:41 <+bridge_> im actually suprised how many people get hacked 17:41 <+bridge_> scammers making alot of money 17:42 <+bridge_> should look into becoming a scammer 17:42 <+bridge_> forreal 17:42 <+bridge_> and its easy to not get caught 17:44 <+bridge_> Send me 1 bitcoin, I'll send you back 2 bitcoin in 2 weeks 17:44 <+bridge_> Trust 17:44 <+bridge_> i've never seen an active person, or a person in power having his account stolen, except @smashtw 17:44 <+bridge_> never make him forget 17:45 <+bridge_> Sure! where can i get your bitcoin address 17:46 <+bridge_> 1CasHVjA4ZfcDsbSiBDcYiUjtiUT3vgmvU 17:46 <+bridge_> `1CasHVjA4ZfcDsbSiBDcYiUjtiUT3vgmvU` 17:46 <+bridge_> how? did he download a trojan 17:46 <+bridge_> having your wallet contain `CasH` is a flex in itself 17:47 <+bridge_> and I didn't even need to bruteforce it, just randomly got it 😄 17:47 <+bridge_> dont bitcoin addresses start with bc? 17:48 <+bridge_> All standard bitcoin addresses start with `1` 17:48 <+bridge_> mine is `bc1q40t6tt58ru7xs57s44t2pw4hnn2kemdacjejcr` 17:48 <+bridge_> `3` is for p2sh 17:48 <+bridge_> `bc1` are bech32 encoded addresses 17:49 <+bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1514657726024781896/waterfox_1Ihd0jXLLc.png?ex=6a2c2a6e&is=6a2ad8ee&hm=e5dec387890266dbcaae7eb88969be06501f46f2d8ca7c9b0c5a520b2c42d2ef& 17:49 <+bridge_> i got a cooler one `bc1q722l2dz86g2k6rjrdlxlmq02un2xldued7v5kk` 17:49 <+bridge_> quite a few L's in there cuh 17:49 <+bridge_> you have "ru" in yours 17:49 <+bridge_> melon confirmed russian?? 17:49 <+bridge_> privet 17:49 <+bridge_> axaxaaxaa 17:50 <+bridge_> I just have an old wallet my addresses are base58 encoded :pepeW: 17:50 <+bridge_> why do none of your wallets have any activity 17:50 <+bridge_> I need to read a paper, but it's just AI stuff, so I keep getting bored and doing other stuff 17:51 <+bridge_> have AI read it to you 17:51 <+bridge_> https://dl.acm.org/doi/abs/10.1145/3719027.3744828 17:51 <+bridge_> there are websites for that 17:51 <+bridge_> melon read it to him 17:51 <+bridge_> you have a reading voice 17:54 <+bridge_> > High-level takeaway 17:54 <+bridge_> > Disa combines: 17:54 <+bridge_> > 17:54 <+bridge_> > Transformer-based contextual instruction classification 17:54 <+bridge_> > Superset disassembly 17:54 <+bridge_> > Value-set analysis 17:54 <+bridge_> > DWARF-derived training labels 17:54 <+bridge_> > Memory-block-aware downstream binary analysis 17:54 <+bridge_> > Its main claimed advantages are: 17:54 <+bridge_> > 17:54 <+bridge_> > More accurate function entry recovery 17:54 <+bridge_> > Robustness against obfuscation 17:54 <+bridge_> > Better memory block boundary recovery than heuristic methods 17:54 <+bridge_> > Improved indirect call target resolution when integrated with BPA 17:54 <+bridge_> you're welcome, thats 3€ in tokens 17:54 <+bridge_> 3e for that is insane 17:55 <+bridge_> Also I keep seeing my professors name in like every random security/reverse engineering paper 😄 17:56 <+bridge_> The man is too humble 18:31 <+bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1514668506384302202/image.jpg?ex=6a2c3479&is=6a2ae2f9&hm=783d5f65467ff898573b6bd12eb97ad88f0afca5e22293d100c2c04198a50c15& 18:31 <+bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1514668506816188547/image.jpg?ex=6a2c3479&is=6a2ae2f9&hm=dce6201f81a89091c4d7adcdac63f44aae44d39fb88b16a343081894e774b2d0& 18:31 <+bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1514668507453853716/image.jpg?ex=6a2c3479&is=6a2ae2f9&hm=7343f2cc0ba6dfc9cc202780e3498f1d89074918b6fb00799ab224b68732032b& 18:31 <+bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1514668508225470604/image.jpg?ex=6a2c3479&is=6a2ae2f9&hm=4a541d9c26eb28f22aac02e4034792a74bfa0d820a592e612c21acde8504a35d& 18:42 <+ChillerDragon> @melon poc of the api could help but missing the module that uses the api will still be hard 18:43 <+bridge_> i can create a PoC of both, i think 18:44 <+ChillerDragon> i doubt itll help 18:44 <+ChillerDragon> if you cant put it in words because you have to censor everything 18:44 <+ChillerDragon> i assume the code will look the same 18:47 <+bridge_> hmm 18:47 <+bridge_> 18:47 <+bridge_> kinda true? - the PoC would show this: 18:47 <+bridge_> 18:47 <+bridge_> heres the hook being called: 18:47 <+bridge_> which gives me this data: 18:47 <+bridge_> and here is a lib that forwards it: 18:47 <+bridge_> 18:47 <+bridge_> i guess xd 18:47 <+ChillerDragon> yea 18:47 <+ChillerDragon> you can just send the prompt in this chat instead of into your LLM 18:48 <+ChillerDragon> what info does it get in which format 18:48 <+ChillerDragon> and what does it do with the info 18:51 <+bridge_> Wat 18:51 <+bridge_> Thats like 18:52 <+bridge_> > "I can show you how to tie both knots." 18:52 <+bridge_> 18:52 <+bridge_> > "I doubt that'll help. If you can't describe the difference in words, I assume they'll look the same." 18:53 <+bridge_> > you can just send the prompt in this chat instead of into your LLM 18:53 <+bridge_> 18:53 <+bridge_> I dont use AI to write me code, i'm fairly capable of doing that myself, unless i'm not, but then i would've openly said so instead of requesting something that heavy 18:56 <+bridge_> You can't feed the DDNet code in an AI. You'll get a cease and desist letter back :p 18:56 <+bridge_> someone already did it https://deepwiki.com/ddnet/ddnet 18:58 <+bridge_> anyway 18:58 <+bridge_> i've stated which info i would want from every hook-call from every rcon command in the issue: ` client ID, IP, reason, duration, and issuing client` 18:58 <+bridge_> the format is tbd, for ease of use probably JSON. 18:58 <+bridge_> 18:58 <+bridge_> whats done is consuming it to further work with it: 18:59 <+bridge_> > I can't consume it programmatically without external tooling that parses log output, which is fragile and requires either a modified client, external tooling relying on dumps, or something sitting on the server intercepting stdout. 18:59 <+bridge_> 18:59 <+bridge_> example: 18:59 <+bridge_> mod X bans a player for toxicity and goes to sleep 18:59 <+bridge_> X decides he was wrongfully banned, and creates a ban appeal 18:59 <+bridge_> X now has to wait 8 hours for the moderator to wake up so other moderators can decide if he was wrongfully banned, or annoy an admin to provide logs, if the mod forgot to add context before. 18:59 <+bridge_> 18:59 <+bridge_> if i'm able to provide a simple chat history from X (when he's banned, the hook will dump_log and parse for player X, as to not breach GDPR) immediatly, there is no need for moderators to 18:59 <+bridge_> 18:59 <+bridge_> dump_log 18:59 <+bridge_> ban 18:59 <+bridge_> dump_remote_console 18:59 <+bridge_> open explorer 18:59 <+bridge_> %appdata% -> DDnet/dumps/remote_console_dump_20261106.txt 18:59 <+bridge_> ctrl+c 18:59 <+bridge_> open discord 18:59 <+bridge_> open #moderator 18:59 <+bridge_> ctrl+v 18:59 <+bridge_> `hey guys, i banned X for saying this:` 18:59 <+bridge_> Hellou :3 18:59 <+bridge_> instead they can just.. ban and carry on with their day 18:59 <+bridge_> Hellou beautiful :3 19:00 <+bridge_> anyway 19:00 <+bridge_> i've stated which info i would want from every hook-call from every rcon command in the issue: ` client ID, IP, reason, duration, and issuing client` 19:00 <+bridge_> the format is tbd, for ease of use probably JSON. 19:00 <+bridge_> 19:00 <+bridge_> whats done is consuming it to further work with it: 19:00 <+bridge_> > I can't consume it programmatically without external tooling that parses log output, which is fragile and requires either a modified client, external tooling relying on dumps, or something sitting on the server intercepting stdout. 19:00 <+bridge_> 19:00 <+bridge_> example: 19:00 <+bridge_> mod X bans a player for toxicity and goes to sleep 19:00 <+bridge_> X decides he was wrongfully banned, and creates a ban appeal 19:00 <+bridge_> X now has to wait 8 hours for the moderator to wake up so other moderators can decide if he was wrongfully banned, or annoy an admin to provide logs, if the mod forgot to add context before. 19:01 <+bridge_> 19:01 <+bridge_> if i'm able to provide a simple chat history from X (when he's banned, the hook will dump_log and (we can) parse for player X, as to not breach GDPR) immediatly, there is no need for moderators to 19:01 <+bridge_> 19:01 <+bridge_> dump_log 19:01 <+bridge_> ban 19:01 <+bridge_> dump_remote_console 19:01 <+bridge_> open explorer 19:01 <+bridge_> %appdata% -> DDnet/dumps/remote_console_dump_20261106.txt 19:01 <+bridge_> ctrl+c 19:01 <+bridge_> open discord 19:01 <+bridge_> open #moderator 19:01 <+bridge_> ctrl+v 19:01 <+bridge_> `hey guys, i banned X for saying this:` 19:02 <+bridge_> by the way, i'm all ears for other suggestions, its just something i came up with after not being able to figure out for.. a long time :D 19:05 <+bridge_> anyway 19:05 <+bridge_> i've stated which info i would want from every hook-call from every rcon command in the issue: ` client ID, IP, reason, duration, and issuing client` 19:05 <+bridge_> the format is tbd, for ease of use probably JSON or some other structured format. 19:05 <+bridge_> 19:05 <+bridge_> whats done is consuming it to further work with it: 19:05 <+bridge_> > I can't consume it programmatically without external tooling that parses log output, which is fragile and requires either a modified client, external tooling relying on dumps, or something sitting on the server intercepting stdout. 19:05 <+bridge_> 19:05 <+bridge_> example: 19:05 <+bridge_> mod X bans a player for toxicity and goes to sleep 19:05 <+bridge_> X decides he was wrongfully banned, and creates a ban appeal 19:05 <+bridge_> X now has to wait 8 hours for the moderator to wake up so other moderators can decide if he was wrongfully banned, or annoy an admin to provide logs, if the mod forgot to add context before. 19:05 <+bridge_> 19:05 <+bridge_> if i'm able to provide a simple chat history from X (when he's banned, the hook will dump_log and (we can) parse for player X, as to not breach GDPR) immediatly, there is no need for moderators to 19:05 <+bridge_> 19:05 <+bridge_> dump_log 19:05 <+bridge_> ban 19:05 <+bridge_> dump_remote_console 19:05 <+bridge_> open explorer 19:06 <+bridge_> %appdata% -> DDnet/dumps/remote_console_dump_20261106.txt 19:06 <+bridge_> ctrl+c 19:06 <+bridge_> open discord 19:06 <+bridge_> open #moderator 19:06 <+bridge_> ctrl+v 19:06 <+bridge_> `hey guys, i banned X for saying this:` 19:06 <+bridge_> anyway 19:06 <+bridge_> i've stated which info i would want from every hook-call from every rcon command in the issue: ` client ID, IP, reason, duration, and issuing client` 19:06 <+bridge_> the format is tbd, for ease of use probably JSON or some other structured format. 19:06 <+bridge_> 19:06 <+bridge_> whats done is consuming it to further work with it: 19:06 <+bridge_> > I can't consume it programmatically without external tooling that parses log output, which is fragile and requires either a modified client, external tooling relying on dumps, or something sitting on the server intercepting stdout. 19:06 <+bridge_> 19:06 <+bridge_> example: 19:06 <+bridge_> mod X bans a player for toxicity and goes to sleep 19:06 <+bridge_> X decides he was wrongfully banned, and creates a ban appeal 19:06 <+bridge_> X now has to wait 8 hours for the moderator to wake up so other moderators can decide if he was wrongfully banned, or annoy an admin to provide logs, if the mod forgot to add context before. 19:06 <+bridge_> 19:06 <+bridge_> if i'm able to provide a simple chat history from X (when he's banned, the hook will fetch the log and (we can) parse for player X, as to not breach GDPR) immediatly, there is no need for moderators to 19:06 <+bridge_> 19:06 <+bridge_> dump_log 19:06 <+bridge_> ban 19:06 <+bridge_> dump_remote_console 19:06 <+bridge_> open explorer 19:06 <+bridge_> %appdata% -> DDnet/dumps/remote_console_dump_20261106.txt 19:06 <+bridge_> ctrl+c 19:06 <+bridge_> open discord 19:07 <+bridge_> open #moderator 19:07 <+bridge_> ctrl+v 19:07 <+bridge_> `hey guys, i banned X for saying this:` 19:17 <+ChillerDragon> @melon i actually think i get it now 19:17 <+ChillerDragon> finally something tangible 19:18 <+ChillerDragon> but actually i still argue this can be done in antibot 19:18 <+ChillerDragon> go merge https://github.com/ddnet/ddnet/pull/11688 and then add antibot ban command 19:19 <+ChillerDragon> no need to maintain a new api for just that 19:19 <+bridge_> 🐴 19:21 <+ChillerDragon> my alternative suggestion is to remove the moderator rights of the player that bans for toxcity 19:22 <+ChillerDragon> make bans expire faster and get rid of the ban appeal system 19:41 <+bridge_> suggestion declined 19:46 <+bridge_> i think i'd like a general yes/no/maybe from learath/roby since they know the pain-point and are also maintainers, afterwards i'm all ears on the actual details 20:43 <+bridge_> @learath2 if you were working on a generic IR and had to expose target specific functionality, would you do it using target specific intrinsics(what LLVM does) or would you do it using target specific instructions(what cranelift does)? 20:43 <+bridge_> (asking for a friend) 20:47 <+bridge_> Target specific instructions 20:48 <+bridge_> I just feel it's easier to work with, littering intrinsic "calls" all over the place doesn't feel as accurate 20:49 <+bridge_> Ye, I agree 21:09 <+bridge_> Big time hackers, grrr 21:09 <+bridge_> You being Admin is the biggest scam tho 21:11 <+bridge_> did you download a trojan or what happened 21:25 <+bridge_> isnt that intrinsics 21:26 <+bridge_> unless ur IR is literally the target assembly 21:58 <+bridge_> somehow got myself a rat, no clue how that happened 21:58 <+bridge_> somehow got myself a remote access trojaner, no clue how that happened 22:09 <+bridge_> my bad 22:20 <+bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1514725927521615893/jwm0pxw.webp?ex=6a2c69f3&is=6a2b1873&hm=4b96e7221d8137ecff17636438adcece519c386a5b0bab58fb203313e6d8bd47& 22:26 <+bridge_> <12944qwerty> can someone please help mefigure out how to solve this damned issue. 22:26 <+bridge_> <12944qwerty> It only happens when I build ddnet but not with a released one 22:26 <+bridge_> <12944qwerty> https://cdn.discordapp.com/attachments/293493549758939136/1514727581457449111/image.png?ex=6a2c6b7d&is=6a2b19fd&hm=870baedc27df83b7819f784c787a047e441872d785c01df93205ae8b9abfb1c7& 22:27 <+bridge_> Did you try a clean build? 22:27 <+bridge_> <12944qwerty> yeah 22:27 <+bridge_> <12944qwerty> several times 22:27 <+bridge_> <12944qwerty> dvulkan is on by default on linux too so 22:28 <+bridge_> Check the console for errors. And check if shaders have actually been compiled. You should have `prim.frag` and `prim.frag.spv` in `data/shaders/vulkan`. 22:29 <+bridge_> <12944qwerty> prim.frag.spv isn't there 22:30 <+bridge_> I had the same issue on cachy and resolved it by clean building/reconfiguring with cmake 22:30 <+bridge_> <12944qwerty> https://cdn.discordapp.com/attachments/293493549758939136/1514728654171668664/image.png?ex=6a2c6c7d&is=6a2b1afd&hm=29dd1ce4d47774cba4b5420127630e765d0985fe201429cf63836f76e2c95f58& 22:31 <+bridge_> <12944qwerty> im on cachy too 22:31 <+bridge_> Open `CMakeCache.txt`, change the value of `VULKAN_SHADER_FILE_SHA256` to empty or any other value. Then run `cmake` again to configure the build folder. 22:31 <+bridge_> <12944qwerty> i've done 22:31 <+bridge_> <12944qwerty> ```sh 22:31 <+bridge_> <12944qwerty> rm -rf build 22:31 <+bridge_> <12944qwerty> cmake -Bbuild -GNinja -DCMAKE_EXPORT_COMPILE_COMMANDS=ON 22:31 <+bridge_> <12944qwerty> cmake --build build 22:31 <+bridge_> <12944qwerty> ./build/DDNet 22:31 <+bridge_> <12944qwerty> ``` 22:32 <+bridge_> Change the asset mapres or whatever it’s called to a transparent image? Like how it’s done in this video with skins: https://youtu.be/e4l1lZ5uw6M 22:32 <+bridge_> Rude :p 22:32 <+bridge_> Did you install the Vulkan SDK? What does cmake print about Vulkan shaders? Maybe try to build #12270 which should improve the error handling. 22:32 <+bridge_> https://github.com/ddnet/ddnet/pull/12270 22:34 <+bridge_> <12944qwerty> that didn't work either 22:34 <+bridge_> <12944qwerty> https://cdn.discordapp.com/attachments/293493549758939136/1514729568013914224/image.png?ex=6a2c6d57&is=6a2b1bd7&hm=1bc3c9d166b5a6ea4e8e3d8d9f631c495df388f0a5675b7079de1063d9c51e57& 22:34 <+bridge_> <12944qwerty> trying pr 22:34 <+bridge_> CD build try again 22:34 <+bridge_> run from build dir 22:39 <+bridge_> <12944qwerty> gmerwighrwgiwgh ofc that works 22:39 <+bridge_> <12944qwerty> guhgwiyghwe 22:39 <+bridge_> <12944qwerty> that's so annoying 22:39 <+bridge_> <12944qwerty> thanks 22:43 <+bridge_> Did the shaders end up being compiled outside your build folder or why did the configuration step succeed? Do you get any useful error with #12270 when configuring? 22:43 <+bridge_> https://github.com/ddnet/ddnet/pull/12270 22:48 <+bridge_> <12944qwerty> i never did the pr, tried cding into build and it worked 22:48 <+bridge_> <12944qwerty> i never did the pr, just tried cding into build and it worked 22:53 <+bridge_> Would be useful to know if the PR helps. Otherwise we should check it different or document it. 22:55 <+bridge_> chillerdragon rush mistakes be lik 23:02 <+bridge_> You can if you like, but all the backups are already provided in repo. It automatically connects to any DDNet servers with at least 2 players when the map is set to Volleyball 23:03 <+bridge_> there's a couple gotchas in the setup (need argon2 installed, also need some C compilation step for the bot) but i can check here again 23:04 <+bridge_> I can send you the most recent stuff (including auth.json) over XMPP 23:04 <+bridge_> @kebscs bool client id 23:04 <+bridge_> (yes)) 23:04 <+bridge_> (yes) 23:05 <+bridge_> did it break anything in current 128p pr? 23:05 <+bridge_> only visually for laser transparency i think 23:08 <+bridge_> PR is probably not required anymore anyways since most players should be on a 128p client nowadays. Upcoming 128p servers hosted by DDNet should probably disallow older clients from connecting. 23:08 <+bridge_> There are still some bugs fixed in the PR that should be ported tho (https://github.com/ddnet/ddnet/issues/9773) 23:09 <+bridge_> Dropping older clients including 16p clients, and remove the old playermapping aswell 23:09 <+bridge_> (and possibly drop 0.7 support as it doesnt have 128p and is barely used) 23:11 <+bridge_> isnt whole point of this pr to at least have some backwards compat? 23:12 <+bridge_> thats why its taking so long instead of just changing max_clients+teams 23:12 <+bridge_> I think the initial reason was to let everyone update before we host servers without mapping 23:13 <+bridge_> But yea, some mapping is generally good. And this PR solves it in a usable way. We should host a brutal or insane server tho for real testing 23:13 <+bridge_> good that some great mind (me) thought to include client-side support before the actual pr 23:14 <+bridge_> and now that older versions are almost unused, everyone is supported 23:14 <+bridge_> almost everyone, yes 23:14 <+bridge_> real 23:14 <+bridge_> what was the point of multeasy when 90% of ppl just sit on spawn/playground 23:15 <+bridge_> idk 23:15 <+bridge_> at least do 2 days in the back, this map is always full 23:17 <+bridge_> admins!! 23:18 <+bridge_> @kebscs btw i'm not entirely sure about multi-view. Is that a client bug? 23:18 <+bridge_> Random player ping jump to 0, what is that about? 23:20 <+bridge_> yea multi-view is client sided 23:20 <+bridge_> idk if this is the same on 64p, maybe its unrelated to ur pr 23:20 <+bridge_> ping jumping to 0, just look at scoreboard with a lot of players 23:20 <+bridge_> some will have 0 randomly 23:21 <+bridge_> that one doesnt happen on 64p i think 23:34 <+bridge_> windows still uses "NULL" in their win32 documentation though 23:49 <+bridge_> kebs, to me it seems the translation layer works. Clients prior to `VERSION_DDNET_128_PLAYERS` get 64p mapping. Clients after that but prior to `VERSION_DDNET_128_TEAMS` get no translation but a legacy fix for old team handling (`TEAM_SUPER` is still 64). 23:49 <+bridge_> Clients with this PR applied get the correct `TEAM_SUPER` (128) and receive all teams as usual. 23:52 <+bridge_> I have no idea and never seen ping jump to 0 or finish time being applied wrong. Needs more info