00:24 < bridge> <ryozuki> https://reshade.me/
00:24 < bridge> <ryozuki> @jupeyy_keks do u know about this?
00:24 < bridge> <ryozuki> so in ffxiv its common to use reshade
00:24 < bridge> <ryozuki> and im running ffxiv on linux
00:24 < bridge> <ryozuki> i found this https://github.com/DadSchoorse/vkBasalt
00:24 < bridge> <ryozuki> reshade is made to apply post shaders to make a game look nicer for gameplay or pics
00:53 < bridge> <Ewan> cool
00:53 < bridge> <Ewan> is ffxiv on linux natively
01:02 < bridge> <ryozuki> no
01:47 < bridge> <stormaxd> idk should i be proud, its more shamefull, but client deving hit me so hard
01:47 < bridge> <stormaxd> well i was inspired cause of @kekomonter so big thanks
01:48 < bridge> <stormaxd> https://cdn.discordapp.com/attachments/293493549758939136/1177772731102674974/image.png?ex=6573b955&is=65614455&hm=5a8b693657f59956c3f22567460b345e6d11c23b0374d9e7969cccb31be57858&
01:48 < bridge> <stormaxd> https://cdn.discordapp.com/attachments/293493549758939136/1177772731350134847/image.png?ex=6573b955&is=65614455&hm=9ccf3c3d2e6d1bb4ea20b41e4cec30bb4fea77914bd9a95c8b4e1f6cfe60fef6&
01:48 < bridge> <stormaxd> https://cdn.discordapp.com/attachments/293493549758939136/1177772731886997534/image.png?ex=6573b955&is=65614455&hm=52454579bd71c57d396e6a9a73a307392df8d6040b69d8a9383594a1696a8eca&
01:51 < bridge> <kekomonter> I inspired someone to make cool clients :poggers2:
01:51 < bridge> <stormaxd> https://tenor.com/view/niko-happy-omori-oneshot-gif-25864183
02:17 < bridge> <Night_L> Hi, I am looking for some help with the hook mechanism. In this stage (map <bizarre untold> TP93), when dummy fell into the rightmost wall, you can not hook directly by any chance when standing still(checked by hook collision). But you have a slight chance to hook it when you move right. The hook seems to stretch a little bit and become longer. I would like to know why this happens and how it relates to the implementation
02:17 < bridge> <Night_L> https://cdn.discordapp.com/attachments/293493549758939136/1177779971901620344/image.png?ex=6573c013&is=65614b13&hm=e4f6871cd1d500dcc70701f0b3636466e427810f5ae0b7f57debcde3ccb4dd83&
02:18 < bridge> <Night_L> Hi, I am looking for some help with the hook mechanism. In this stage (map <bizarre untold> TP93), when dummy fell into the rightmost wall, you can not hook directly by any chance when standing still(checked by hook collision). But you have a slight chance to hook it when you are walking right. The hook seems to stretch a little bit and become longer. I would like to know why this happens and how it relates to the implementation
02:19 < bridge> <Night_L> https://cdn.discordapp.com/attachments/293493549758939136/1177779971901620344/image.png?ex=6573c013&is=65614b13&hm=e4f6871cd1d500dcc70701f0b3636466e427810f5ae0b7f57debcde3ccb4dd83&
02:19 < bridge> <Night_L> Hi, I am looking for some help with the hook mechanism. In this stage (map <bizarre untold> TP93), when dummy fell into the rightmost wall, you can not hook directly by any chance when standing still(checked by hook collision). But you have a slight chance to hook it when you are walking right. The hook seems to stretch a little bit and become longer. I would like to know why this happens and how it relates to the code
02:19 < bridge> <Night_L> https://cdn.discordapp.com/attachments/293493549758939136/1177779971901620344/image.png?ex=6573c013&is=65614b13&hm=e4f6871cd1d500dcc70701f0b3636466e427810f5ae0b7f57debcde3ccb4dd83&
02:21 < bridge> <Night_L> the collision is far from being able to hook the dummy
02:21 < bridge> <Night_L> https://cdn.discordapp.com/attachments/293493549758939136/1177781125905666058/image.png?ex=6573c126&is=65614c26&hm=af772f1dddce7954fd714de51ab36eff615a72a7b151ba816b6afd85d6e13bed&
02:23 < bridge> <Night_L> Hi, I am looking for some help with the hook mechanism. In this stage (map <bizarre untold> TP93), when dummy fell into the rightmost wall, you can not hook directly by any chance when standing still(checked by hook collision). But you have a slight chance to hook it when you are walking right. The hook seems to become longer and move through normal tiles. I would like to know why this happens and how it relates to the code
02:23 < bridge> <Night_L> https://cdn.discordapp.com/attachments/293493549758939136/1177779971901620344/image.png?ex=6573c013&is=65614b13&hm=e4f6871cd1d500dcc70701f0b3636466e427810f5ae0b7f57debcde3ccb4dd83&
02:33 < bridge> <cyberfighter> because hooklines dont count for momentum and the hook will stop if you are a certain distance from it, so if you are going in the same direction you can hold it for a little bit longer because it will have to travel a bit farther until it gets out of your range
04:12 < bridge> <_voxeldoesart> ‼️‼️‼️‼️
05:01 < bridge> <nytebk201> Am I making it clear that hook moves with tee. So having momentum will cause the endpoint of a single hook to stretch?
05:57 < bridge> <Mr.Gh0s7> good C@/day
06:41 < bridge> <Mr.Gh0s7> btw it reads good S-at-per(der)-day
06:44 < bridge> <lolkildal> the hook has a certain length, but this length is calculated not from the point from where you hooked, but from your tee, so if you move towards the hook with enough speed the hook can fly almost indefinitely.  An example of this is try falling from a great height and hooking downwards.  in general, on the Wiki there is a mention of this in advanced behavior https://wiki.ddnet.org/wiki/Hook
06:45 < bridge> <lolkildal> the hook has a certain length, but this length is calculated not from the point from where you hooked, but from your tee, so if you move towards the hook with enough speed the hook can fly almost indefinitely.  An example of this is try falling from a great height and hooking downwards.  in general, on the Wiki there is a mention of this in advanced behavior https://wiki.ddnet.org/wiki/Hook
07:44 < bridge> <Jupstar ✪> Yes
09:15 < ChillerDragon> dude i thought every person on the planet thought its a genius idea to build a wrapper around chatgpt but none of thes products will ever be used over plain chatgpt
09:16 < ChillerDragon> https://github.com/abi/screenshot-to-code
09:16 < ChillerDragon> somehow this prompting proxy got 19k stars xd
10:39 < bridge> <ryozuki> https://www.tomshardware.com/news/ubuntu-runs-20-faster-than-windows-11-on-amd-threadripper-pro-7995wx
10:39 < bridge> <ryozuki> :justatest:
10:45 < bridge> <cellegenrih> surprise surprise, windows is shit
10:51 < bridge> <Ewan> 😌
10:51 < bridge> <Ewan> should have saved for threadripper SMH
10:52 < bridge> <Ewan> even by now I probably wouldn’t be able to afford it
10:54 < bridge> <murpi> it's a $10k cpu...
10:55 < bridge> <Ewan> 😃
10:55 < bridge> <Jupstar ✪> just wait few years then they cheap on ebay
10:55 < bridge> <Ewan> i want to buy the $10k cpu
10:56 < bridge> <Ewan> https://cdn.discordapp.com/attachments/293493549758939136/1177910579470549042/IMG_7221.jpg?ex=657439b6&is=6561c4b6&hm=d867472bb6456919c0a1ab3394e8b8f940652b9514f2b1cf658f0f5de03df1d3&
10:56 < bridge> <Jupstar ✪> i also want to buy a 10k$ GPU. but then i looked into my bank
10:57 < bridge> <Jupstar ✪> i am something of an millionaire myself
10:57 < bridge> <Ewan> don’t make me think about my bank account
10:57 < bridge> <Ewan> normally I do not share this info but i am feeling funny rn
10:57 < bridge> <Ewan> i have $2.30
10:57 < bridge> <Jupstar ✪> arent u working?
10:58 < bridge> <Ewan> i wish
10:58 < bridge> <murpi> @mpft the cost of the board required to operate the CPU is equivalent to the amount you spent on your current CPU alone xd
10:59 < bridge> <Ewan> yes it’s expensive
12:02 < bridge> <ryozuki> day 2 of writing the blog
12:16 < bridge> <cyberfighter> ho no
12:16 < bridge> <cyberfighter> oh no
12:17 < bridge> <cyberfighter> i honestly think its a good thing
13:31 < bridge> <cyberfighter> huh i got a strange bug, scrolling through layers in the editor was really really fast for a moment but it fixed once i alt tabbed
14:33 < bridge> <stormaxd> https://i.imgur.com/QpC13tn.png
14:33 < bridge> <stormaxd> what a prank
14:34 < bridge> <ryozuki> https://cdn.discordapp.com/attachments/293493549758939136/1177965420687405096/image.png?ex=65746cc9&is=6561f7c9&hm=19936b6b0bc06aed7d7e6ce1d1e1adf9fbe4950f769ffb898cc38f11c689dece&
14:34 < bridge> <ryozuki> arm64
14:34 < bridge> <ryozuki> u got 30 registers?
14:34 < bridge> <ryozuki> also i like these names way more
14:34 < bridge> <ryozuki> easier to memorize
14:34 < bridge> <stormaxd> no pranks anymore
14:34 < bridge> <stormaxd> https://cdn.discordapp.com/attachments/293493549758939136/1177965575285243914/image.png?ex=65746cee&is=6561f7ee&hm=ab5c7a60d0465da60f15cb6be35c13f1df9ef83bf706d8023e121bcf07c9de33&
14:34 < bridge> <stormaxd> fr
14:41 < bridge> <stormaxd> https://i.imgur.com/4jxC2XL.png
14:42 < bridge> <milkeeycat> link to twerking twinbop would be better :feelsbadman:
14:43 < bridge> <stormaxd> ahaha
14:43 < bridge> <stormaxd> so true
14:43 < bridge> <stormaxd> alr gonna be done
14:43 < bridge> <milkeeycat> there's one problem, nobody have drawn twerking twinbop
14:43 < bridge> <stormaxd> i'll use this
14:43 < bridge> <stormaxd> https://cdn.discordapp.com/attachments/293493549758939136/1177967896375328848/insane_clip.mp4?ex=65746f18&is=6561fa18&hm=80798897527f208993d459695e5d22375386f6b3cf52424c390341866aa14782&
14:44 < bridge> <stormaxd> if m_PlayerName == twinbop
15:07 < ChillerDragon> bliet asan fakin useless i thought i can just iterate on arrays how ever i want with random indecies and asan would complain if its bad
15:07 < ChillerDragon> it didnt
15:15 < bridge> <Jupstar ✪> ?
15:15 < bridge> <Jupstar ✪> ofc it does, as soon as it is out of bounds
15:17 < ChillerDragon> thats what i thought
15:18 < ChillerDragon> i went 1 too high out of bounds and it didnt complain it just set some other integer to zero :D
15:18 < ChillerDragon> average c++ moment
15:18 < bridge> <Jupstar ✪> i can just repeat myself
15:18 < bridge> <Jupstar ✪> it must be a skill issue
15:18 < ChillerDragon> yes
15:18 < bridge> <Jupstar ✪> or a serious bug in asan
15:19 < ChillerDragon> which is more likely?
15:19 < bridge> <Jupstar ✪> skill issue xd
15:19 < ChillerDragon> chiler being stoopid or asan being bugged
15:19 < ChillerDragon> who knows
15:19 < bridge> <Jupstar ✪> i mean, u probs on some hipster OS
15:19 < ChillerDragon> i use arch BTW
15:19 < bridge> <Jupstar ✪> but i guess they have some basic tests to detect simple out of bounds like this
15:25 < ChillerDragon> https://paste.zillyhuhn.com/ef
15:25 < ChillerDragon> this is my bugged code ripped out
15:25 < ChillerDragon> ``for(int i = MAX_LINES; i > 0; i--)`` where MAX_LINES is array size
15:26 < ChillerDragon> pls fix asan thank jopsti
15:26 < bridge> <Jupstar ✪> i cant read code without colors
15:26 < bridge> <Jupstar ✪> come matrix and send with highligther
15:28 < ChillerDragon> lerato u here?
15:28 < bridge> <Jupstar ✪> yeah ther eis a out of bounds
15:28 < ChillerDragon> can you make it show in asan?
15:28 < bridge> <Jupstar ✪> now show your cmake command
15:28 < bridge> <Jupstar ✪> does this programm compile?
15:28 < bridge> <Jupstar ✪> ok
15:28 < bridge> <Jupstar ✪> just for you
15:29 < ChillerDragon> ``CC=clang CXX=clang++ CXXFLAGS='-fsanitize=address,undefined -fsanitize-recover=address,undefined -fno-omit-frame-pointer' CFLAGS='-fsanitize=address,undefined -fsanitize-recover=address,undefined -fno-omit-frame-pointer'  cmake .. -DCMAKE_BUILD_TYPE=Debug``
15:29 < ChillerDragon> tbh i only tested within tw 
15:29 < ChillerDragon> not this minimal example
15:29 < ChillerDragon> idk how to asan on a single cpp file have no tooling for that xd
15:30 < bridge> <Jupstar ✪> https://cdn.discordapp.com/attachments/293493549758939136/1177979629810753627/image.png?ex=65747a05&is=65620505&hm=5df0c560b150cc59485c38c0b373f3519ee09b6c39bad78a12f16a2888f7b941&
15:30 < bridge> <Jupstar ✪> u happy now?
15:31 < bridge> <robyt3> `-fsanitize-recover=address` is probably a bad idea, you don't want to continue executing after OOB accesses
15:31 < bridge> <robyt3> You may not get the ASAN report until the program closes otherwise
15:32 < bridge> <Jupstar ✪> there u hear it
15:32 < bridge> <Jupstar ✪> master robyte knows
15:32 < bridge> <robyt3> There are false negatives, i.e. cases were ASAN can miss OOB access, but this is probably not the case. You'll have to read the paper for that https://research.google.com/pubs/archive/37752.pdf
15:33 < ChillerDragon> send command robsti
15:33 < ChillerDragon> why did it not alert in my tw codebase
15:33 < bridge> <robyt3> I assume in your code base there is another member variable directly behind your array, so the OOB access on the array still ends up in valid memory
15:33 < ChillerDragon> a rite
15:33 < ChillerDragon> lets add the var
15:33 < bridge> <robyt3> Not sure if ASAN adds shadow memory around member variables
15:34 < bridge> <robyt3> Maybe only around full objects
15:36 < ChillerDragon> ok watever i cant reproduce the issue in minimal maybe it needs classes
15:36 < ChillerDragon> but thanks robsti ill remove the recover
16:25 < bridge> <learath2> Am here, what need?
16:48 < bridge> <learath2> `CRegister` is such a complicated class for what it does
16:49 < bridge> <learath2> All because we had to add 0.7 support :pepeW:
17:10 < bridge> <Jupstar ✪> @_voxeldoesart https://deepmind.google/discover/blog/transforming-the-future-of-music-creation/
17:14 < bridge> <ryozuki> https://tenor.com/view/shame-on-you-gif-25797108
17:14 < bridge> <ryozuki> its why i want to drop 0.7
17:17 < bridge> <learath2> Oh I also have to replace my `unique_locks` with `CLockScope` :pepeW:
17:36 < bridge> <ryozuki> imagine not having Mutex<T>
17:40 < bridge> <robyt3> If only `std::mutex` had thread-safety annotations, then we wouldn't need the wrappers
17:42 < bridge> <ryozuki> mutex should be a generic wrapping a type
17:42 < bridge> <ryozuki> like in rust
17:42 < bridge> <ryozuki> but since c++ doesnt have sum types like rust it wouldnt be safe anyway
17:43 < bridge> <ryozuki> https://doc.rust-lang.org/std/sync/struct.Mutex.html
17:43 < bridge> <ryozuki> well maybe u can do a safeish lock
17:44 < bridge> <ryozuki> ```rust
17:44 < bridge> <ryozuki> use std::sync::Mutex;
17:44 < bridge> <ryozuki> let mutex = Mutex::new(0);
17:44 < bridge> <ryozuki> 
17:44 < bridge> <ryozuki> let mut guard = mutex.lock().unwrap();
17:44 < bridge> <ryozuki> *guard += 20;
17:44 < bridge> <ryozuki> // out of scope unlocks, or u can manually drop()
17:44 < bridge> <ryozuki> ```
17:44 < bridge> <ryozuki> such a simple idea and so powerfull
17:44 < bridge> <ryozuki> why c++ doesnt have this
17:46 < bridge> <_voxeldoesart> isnt this like that jukebox thing
18:00 < bridge> <_voxeldoesart> also NO WAY THEY'RE TRYING TO MARKET "changing your voice into a different instrument" as something new
18:00 < bridge> <_voxeldoesart> thats been a thing since?? forever
18:00 < bridge> <_voxeldoesart> https://cdn.discordapp.com/attachments/293493549758939136/1178017354689544263/image.png?ex=65749d27&is=65622827&hm=0af6ae831e0d0dc59a463459f31c3d817348670d103b05c7aa657179e88462fa&
18:00 < bridge> <Jupstar ✪> but that well?
18:00 < bridge> <Jupstar ✪> can u give me a link to that tool xd
18:01 < bridge> <_voxeldoesart> https://magenta.tensorflow.org/ddsp-vst
18:07 < bridge> <_voxeldoesart> its kinda fun because you make make shitty synths with it that dont even sound like the original instruiment it's trying to change to
18:07 < bridge> <_voxeldoesart> (it gets loud)
18:07 < bridge> <_voxeldoesart> https://cdn.discordapp.com/attachments/293493549758939136/1178019031702327397/2023-11-25_12-04-09.mp4?ex=65749eb7&is=656229b7&hm=a36d99921bd2172b0cb164fe707d3c448571437a1c7755864cacb1d6166244ed&
18:07 < bridge> <Jupstar ✪> ok but that's also an AI i guess? xd
18:07 < bridge> <_voxeldoesart> i have control over it so its good
18:08 < bridge> <Jupstar ✪> anyway the video preview u sent sounds rather simplistic 😄
18:08 < bridge> <_voxeldoesart> thats like, my one problem with at
18:08 < bridge> <_voxeldoesart> its that i dont want it to do my stuff for me. i still want to be able to manage the things i do
18:08 < bridge> <_voxeldoesart> thats like, my one problem with ai
18:08 < bridge> <_voxeldoesart> its that i dont want it to do my stuff for me. i still want to be able to manage the things i do
18:09 < bridge> <Jupstar ✪> as long as AI is stupid u gotta do that anyway
18:09 < bridge> <ryozuki> https://www.tomshardware.com/news/world-first-linux-gaming-laptop-ryzen-7-7840hs-rx-7600m-xt
18:09 < bridge> <_voxeldoesart> yeah lets keep ai stupid
18:09 < bridge> <Jupstar ✪> i cant tell AI to reimplement ddnet yet 😄
18:09 < bridge> <ryozuki> perfect laptop for ddnettera
18:09 < bridge> <ryozuki> perfect laptop for ddnetters
18:09 < bridge> <Jupstar ✪> WAY too expensive
18:10 < bridge> <Jupstar ✪> who would buy that
18:10 < bridge> <Jupstar ✪> for a middle class laptop
18:10 < bridge> <_voxeldoesart> also you KNOW some executive saw the backlash against ai and went "hey. lets target this TOWARDS artists so that more ppl can use our product"
18:11 < bridge> <_voxeldoesart> "But with any new technology we have to approach it responsibly and that's not lost on us. It's start by recognizing that artificial intelligence is meant to amplify human creativity, not replace it" -actual quote from the video
18:11 < bridge> <_voxeldoesart> like anyone's going to listen to that. i dont even think GOOGLE is going to listen to their own words LOL
18:11 < bridge> <Jupstar ✪> they all talk shit anyway xd
18:11 < bridge> <Jupstar ✪> AI is there to replace humans
18:12 < bridge> <Jupstar ✪> nothing else
18:12 < bridge> <ryozuki> :owo:
18:12 < bridge> <ryozuki> replace rust
18:12 < bridge> <Jupstar ✪> even if they only do it partially, ppl will require more and more capabilities
18:13 < bridge> <Jupstar ✪> and at some point they need the capability of smth over a normal human can do
18:13 < bridge> <Jupstar ✪> and this AI is probably more intelligent than a human, so we also dunno if we can control it
18:14 < bridge> <_voxeldoesart> tho i do admit. hearing ai try to copy already existing music is kinda fun because it fumbles it but you can still tell what medium they're trying to piece together
18:14 < bridge> <_voxeldoesart> https://youtu.be/Ba_uGnIoEVA
18:14 < bridge> <_voxeldoesart> tho i do admit. hearing ai try to copy/extend already existing music is kinda fun because it fumbles it but you can still tell what medium they're trying to piece together
18:14 < bridge> <_voxeldoesart> https://youtu.be/Ba_uGnIoEVA
18:15 < bridge> <_voxeldoesart> i doubt it.
18:15 < bridge> <Jupstar ✪> what exactly?
18:15 < bridge> <Jupstar ✪> that there will be AI more intelligent than huamans?
18:15 < bridge> <Jupstar ✪> that there will be AI more intelligent than humans?
18:16 < bridge> <_voxeldoesart> oh i misread the full context
18:17 < bridge> <_voxeldoesart> anyways sometimes i wish ai wasnt trying to replace us. if people were smart and had ai be sort of like, co-exist with us instead of trying to dominate us, then it'd be better
18:18 < bridge> <_voxeldoesart> but with every new ai model implimented, people just use it as an excuse for them not to do the "hard work"
18:18 < bridge> <_voxeldoesart> the hard work is the best part!
18:18 < bridge> <Jupstar ✪> now you know how animals feel about huamans xdd
18:18 < bridge> <Jupstar ✪> apes in cages
18:19 < bridge> <Jupstar ✪> well humans dominate humans
18:19 < bridge> <Jupstar ✪> so we are a problem in itself already
18:19 < bridge> <Jupstar ✪> ai might be the only solution for this, who knows ^^
18:20 < bridge> <_voxeldoesart> for what? human extiction?
18:20 < bridge> <Jupstar ✪> no, for equal rights between humans
18:40 < bridge> <learath2> It compiles :Celebrate: 
18:40 < bridge> <learath2> It doesn't link :ohNo:
18:47 < bridge> <_voxeldoesart> wtf even is girls frontlines
20:09 < bridge> <chairn> why risc-v had to make such complex vector extension ?
20:09 < bridge> <chairn> https://fprox.substack.com/p/risc-v-vector-programming-in-c-with
21:08 < bridge> <louis> wow that looks dope
21:09 < bridge> <louis> the voice to track
21:10 < bridge> <louis> i am actually waiting for the day when u can make some nice samples w ai
22:35 < bridge> <chillerdragon> I sent jopsi a pstd and he wanted code highlight I think there was a query parameter for that. Was wondering if you remember it. But that is now super outdated jospti already red the plaintext code hehe