00:44 < bridge> Does the Server know when you are spectating someone? 00:45 < bridge> Ye 01:01 < bridge> Does the server know exactly who is specing who? 01:02 < bridge> Yes 01:53 < bridge> otherwise you would not be able to see the tees unless you did /showall 02:07 < bridge> This would work if the client only sent the camera position. But it also sends the id of who you are speccing. I think that was the question 02:07 < bridge> Even tho you could check that even if the client didn't send any info about who they're speccing 02:19 < ws-client> #someone4president 02:19 < ws-client> can we start a vote to make SP | Someone ddnet admin? 03:47 < bridge> i don't think chiller can see the vote 03:51 < bridge> current standings are 5/1 in his favor 05:07 < ws-client> <:poggers2:1008007455936094328> 06:15 < bridge> anyone know why there is a `-1` at the end of these 2 lines? https://github.com/ddnet/ddnet/blob/master/src/game/client/render_map.cpp#L591-L592 06:15 < bridge> 06:15 < bridge> it just seems to make things look weird in my opinion because it cuts off 1 pixel from the edge of the tile 06:15 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1287628270728380437/image.png?ex=66f23cc6&is=66f0eb46&hm=5db6cdaac3470159cf948f11d91f2aa6de934a9657326b546c0767c227d117e8& 06:15 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1287628270967193640/image.png?ex=66f23cc6&is=66f0eb46&hm=8a51d4903742aad8576b30dafb78a4cf0ee83df0f619ec602631d35703996d04& 07:31 < bridge> cuz old mapres were shifted :justatest: afair 07:31 < bridge> any experts of tile render there 07:31 < ws-client> patiga is pro 07:32 < bridge> chiller does my pr affect any existing mod? 07:32 < bridge> because now id predefines tunes for all mods 07:32 < ws-client> i think so 07:32 < bridge> it* 07:32 < bridge> and my pr makes it only for ddrace and ddnet 07:33 < ws-client> im not sure i would have to test it 07:33 < ws-client> so many things to think about 07:34 < bridge> if you have time and will pls 07:34 < ws-client> also why u ignore ma comment xd 07:34 < bridge> because i couldn't figure out how ur issue connects with my pr 07:34 < bridge> what exactly is wrong 07:34 < ws-client> then ignore the issue and only respond to the question 07:34 < ws-client> why is tunes ddrace tiles 07:35 < bridge> because if server wants to predict ddrace tiles its supposedly ddrace based 07:35 < bridge> and needs these tunes to operate 07:35 < ws-client> i disagree 07:35 < ws-client> gun curvature and speedups are two different things 07:36 < bridge> but now it does affect all mods 07:36 < bridge> what do you want, to remove ddrace tiles check? 07:36 < ws-client> why have PredictDDrace and PredictDDraceTiles if tiles does not mean tiles 07:36 < ws-client> @zhn well as proposed in the issue i might even want a PredictDDraceTunes flag :p 07:37 < bridge> i want to fix vanilla demos, not to make another protocol entry xd 07:37 < bridge> i wonder how it works good in game and doesnt in demos 07:37 < bridge> something is wrong 07:37 < ws-client> but why did you add predicttiles whats the use case? 07:38 < bridge> if server sent no predict ddrace but predict ddrace tiles xd 07:38 < bridge> why would u need to predict ddrace tiles and not to have default ddnet tunes 07:38 < ws-client> which server does that? and why does that mean that the tunes are ddrace based? 07:38 < ws-client> @zhn ddnet-insta 07:38 < bridge> also it affects only demos 07:39 < bridge> do u send predict ddrace tiles in ddnet insta? 07:39 < bridge> i would 1 tick freeze by prediction if u did 07:39 < bridge> on health pickup touch 07:39 < ws-client> well yea thats what https://github.com/ddnet/ddnet/issues/8923 is about 07:40 < ws-client> currently vanilla weapons plus ddrace tiles such as speedups and stoppers is a mess to get predicted correctly 07:40 < ws-client> your pr looks like it would make it even worse and i dont see the use case 07:40 < bridge> why would u need speeders and stoppers in vanilla xd 07:40 < ws-client> its cool? 07:40 < bridge> join jsaurus server with ddnet 07:40 < bridge> pew pew with shotgun 07:40 < bridge> and pistol 07:40 < bridge> move to demos 07:41 < bridge> and voila 07:41 < bridge> wrong tunes by default 07:41 < bridge> join any vanilla server 07:41 < bridge> even infclass is affected 07:41 < bridge> its 7 year issue that was introduced by defaulting ddrace tunes 07:41 < ws-client> yes sounds good to fix that but my question stands 07:41 < ws-client> why ddracetiles 07:42 < bridge> tell me what it would break in your current setup 07:42 < ws-client> it just seems wrong 07:42 < ws-client> tiles should mean tiles 07:42 < bridge> what exactly 07:42 < ws-client> your prediction is not about tiles 07:43 < bridge> what about ddrqce tune zones? arent these *tiles*? 07:43 < ws-client> hmm 07:43 < bridge> i wouldn't mind if theyre handled differently 07:43 < ws-client> yea imo this should be PredictDDraceWeapons 07:44 < bridge> it affects only demos btw 07:45 < bridge> if u will encounter anything like that in the future we can hop in making new protocol entry 07:45 < bridge> i just don't want it to be protocol dependent 07:45 < bridge> unless existing ones 07:45 < ws-client> why not prepare something clean now for the future? 08:06 < ws-client> the prediction code is such a rabbit hole. TIL the ddnet client hardcodes a specific hammer tuning for fng https://github.com/ddnet/ddnet/blob/f9df4a85be4ca94ca91057cd447707bcce16fd94/src/game/client/prediction/entities/character.cpp#L334-L346 08:10 < bridge> I guess tuning can't work if you're a vanilla server? 08:11 < bridge> or the server simply doesn't care about correct prediction 08:32 < bridge> everything is fine ingame 08:32 < bridge> demo tunings fuck ups on these hardcoded values 08:33 < bridge> I'm not taking about demos 08:33 < bridge> what then 08:34 < bridge> but vanilla servers have their tunings :justatest: 08:34 < bridge> Why doesn't the fng server simply send hammer strength tunes so the client doesn't need these hard coded values? 08:34 < bridge> for the same reason why ddnet tunes are hardcoded ig 08:34 < bridge> idk which though xd 08:35 < bridge> I think it's not the same but ok 08:35 < bridge> afair jupeyys or whatever fng we have has hardcoded strength instead of tuning 08:35 < bridge> not 100% information 08:36 < bridge> Why do demos need tunes at all? They run physics? 08:36 < bridge> demo is just replaying received snapshots as far as i can tell 08:37 < bridge> so how can tunes mess it up 08:37 < bridge> so its a game runningz actually 08:37 < bridge> @totar 08:38 < bridge> That doesn't answer my question 08:39 < bridge> I just don't understand why tunes should affect the demo 08:40 < bridge> Do you mean the demo playback is wrong or the demo saving is wrong? 08:40 < bridge> https://discord.com/channels/252358080522747904/295908390956433410/1281049686911746069 08:41 < bridge> ah waiy 08:41 < bridge> wrong one 08:41 < bridge> Is it saving or playback tho? 08:41 < bridge> https://media.discordapp.net/attachments/824745028311711814/1281009696853069886/sg_demo_bug.mp4?ex=66f1e3bf&is=66f0923f&hm=b95ac0163dbd126afd638081cf2fd8625f33fdd0bcc15209b6675d9ac73ab7eb& 08:42 < bridge> and it was like this for a long time 08:42 < bridge> :justatest: 08:43 < bridge> Can you just answer the question I don't have my pc to test it 08:43 < bridge> wdym by saving or replaying? 08:43 < bridge> both 08:43 < bridge> rendering is just capturing demos playback 08:44 < bridge> If you save the demo on a patched client and play it back on an unpatched one does it work? 08:44 < bridge> 08:44 < bridge> If you save the demo on an unpatched client and play it back on a patched one does it work? 08:45 < bridge> only playback 08:45 < bridge> That's ambiguous just put yes/no for each 08:45 < bridge> you can watch this demo with patched client and everything is fine 08:46 < bridge> Ok thanks 08:46 < bridge> im typing in -2 by celcius, just wait pls xd 08:52 < bridge> hm interesting what if i hardcode gravity to 0 xd 08:52 < bridge> would demo also try to render it like this 09:01 < bridge> Are gun projectiles not part of snapshot data? 09:01 < bridge> So the client has to predict them during playback? 09:01 < bridge> so weird 09:02 < bridge> I thought for sure they should be in snapshots so idk what's happening 09:20 < bridge> @fantrapreal can you stop please 09:21 < bridge> block him to stop reactions 09:21 < bridge> easy solution 09:59 < bridge> An idea to ponder and consider: 09:59 < bridge> 09:59 < bridge> We already have few states Default, Freeze, Deep Freeze. 09:59 < bridge> Is it possible to somehow make a new state? 09:59 < bridge> let's call it "Mechanized" 09:59 < bridge> 09:59 < bridge> ofc need two blocks like a Freeze / Unfreeze, but for this state. 09:59 < bridge> (for example Mechanize / Unmechanize) 09:59 < bridge> One gives this state and second removes state from Tee. 09:59 < bridge> 09:59 < bridge> Tee with this state "Mechanized" is like a frozen Tee. 09:59 < bridge> Player can not control it by inputs, 09:59 < bridge> but Mechanized Tee can be control by another way. 09:59 < bridge> By touch to the {Mech Command Blocks}. 09:59 < bridge> 09:59 < bridge> for example: 09:59 < bridge> if state of Tee is "Mechanized" and it touches to Mech Command Block "HOLD LEFT", 09:59 < bridge> then it will automatically hold it's virtual key [LEFT] all time... 09:59 < bridge> until it touches another {Mech Command Block} "RELEASE LEFT" 09:59 < bridge> 09:59 < bridge> It can hold or release he's virtual keys [LEFT, RIGHT, JUMP, SHOOT, HOOK] 09:59 < bridge> and change look direction (by touch to Mech Command Blocks). 09:59 < bridge> 09:59 < bridge> i think for it need 13 blocks or more... 10:00 < bridge> 10:00 < bridge> 1) "Mechanize" set state "Mechanized" 10:00 < bridge> 2) "Unmechanize" clear state "Mechanized" 10:00 < bridge> 3) "HOLD LEFT" starts holds it's virtual key [LEFT] 10:00 < bridge> 4) "HOLD RIGHT" starts holds it's virtual key [RIGHT] 10:00 < bridge> 5) "HOLD JUMP" starts holds it's virtual key [JUMP] 10:00 < bridge> 6) "HOLD SHOOT" starts holds it's virtual key [SHOOT] 10:00 < bridge> they are part of it 10:00 < bridge> at least their existence 10:01 < bridge> in the snapshot they only store their start point + direction, not all that much 10:01 < bridge> so the client does need to work out the flight path of the projectile 10:02 < bridge> yes they also run a subset of physics, its called "dead reckoning" in the code 10:04 < bridge> Server should send parabolic coefficients :kek: 10:05 < bridge> in essence, as long as a tee doesnt change inputs and doesn't interact with projectiles/tees, it will be viewed as easily predictable (with dead reckoning, a subset of physics) 10:05 < bridge> so a tee with none of the interactions will not show up as changed in the snap to save space, and the client kknows to predict the movement of that tee 10:05 < bridge> thats also why teecore has a `tick` field. with it you know how many ticks of physics this tee needs to advance 10:05 < bridge> I assumed this was only used during actual gameplay to save bandwidth 10:06 < bridge> I was hoping that too, but then I needed to implement it for my demo renderer 10:09 < bridge> this could get lost in this chat, maybe post your idea in #town-hall or open a github issue :) 10:13 < bridge> Sounds kinda cool but also probably a nightmare to implement. 10:14 < bridge> I think except for hook and shoot you can achieve most of the same effect with freeze and speedups but probably not to the precision you want 10:16 < bridge> thx i do not know where is allowed write ideas that's why i write here thx i will write it in town-hall 10:16 < bridge> github issue <--- i do not know where is it 😄 10:20 < bridge> are you sure your grid is correct? pls send that mapres 10:20 < bridge> that -1 is only in affect if your renderer doesn't support array textures, in which case that range is inclusive (as texture coordinates) and thereby needs -1 10:20 < bridge> note that your renderer peobably supports array textures 10:21 < bridge> :heartw: town-hall is a good place, a github issue is not required 10:23 < bridge> You should probably open an issue for a feature request since it will get lost quickly in this channel 10:24 < bridge> he knows already 10:24 < bridge> I already redirected them to #town-hall , github issues tend to be complicated ^^ 10:25 < bridge> i added that message in town-hall with tag Features it's correct? 10:25 < bridge> looks good 👍 10:35 < bridge> morning peeps 10:35 < bridge> mornin 10:39 < bridge> @learath2 treesitter is soo good 10:39 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1287694875524993048/image.png?ex=66f27ace&is=66f1294e&hm=4c258994173d4bda193f59c9e42542407b82ac26e482ae41c819a426192451f8& 10:40 < bridge> im working on ddlauncher 10:40 < bridge> ddlaunch 10:41 < bridge> transparent launcher? 10:41 < bridge> melons working on that too 10:41 < bridge> in rust 10:41 < bridge> :justatest: 10:48 < bridge> well i mentioned it yesterday i would give it a try 10:48 < bridge> xd 10:50 < bridge> :justatest: 10:51 < bridge> does it have ads 10:51 < bridge> does it have ads? 10:51 < bridge> No in golang :kek: 10:52 < bridge> But I'm not sure if I'll stick with lyne as my way of gui'ing, it's supposed to run on a toaster 10:53 < bridge> But I'm not sure if I'll stick with lyne as my way of gui'ing, it's supposed to run on a toaster and I'm not sure if that may be to heavy 10:53 < bridge> waaaaaat 10:53 < bridge> ikr, I'm crazy! 10:53 < bridge> :kek: 11:01 < bridge> mine will be crazy fast 11:01 < bridge> blazingly xD 11:01 < bridge> why would it have ads? tf 11:01 < bridge> https://github.com/Speykious/cve-rs 11:02 < bridge> Ad revenue to pay for better fishy food 11:02 < bridge> @ryozuki 11:02 < bridge> are you anti rust? 11:02 < bridge> just a joke 11:02 < bridge> that repo was already shared 11:02 < bridge> oki 11:02 < bridge> it misses the point completly anyway of rust 11:02 < bridge> im doing rust right now! 11:02 < bridge> i couldn't be antirust 11:03 < bridge> even if the implementation may (and probs have some CVEs), or even if there is a broken thing in a very obscure code that breaks the borrow checker 11:03 < bridge> rust is still useful 11:03 < bridge> in c++ u just need to pop a empty vector to get a segfault 11:04 < bridge> > Exception safety 11:04 < bridge> > If the container is not empty, the function never throws exceptions (no-throw guarantee). 11:04 < bridge> > Otherwise, it causes undefined behavior. 11:04 < bridge> i mean that would be a skill issue 11:04 < bridge> all is skill issue 11:04 < bridge> yea 11:04 < bridge> just dont code 11:04 < bridge> no segfaults 11:04 < bridge> go start a farm 11:04 < bridge> Can I join 11:04 < bridge> https://tenor.com/view/honest-word-its-honest-work-it-aint-much-it-aint-much-but-its-honest-work-gif-13763573 11:04 < bridge> I take care of fishies 11:05 < bridge> just do some asan/ubsan/tsan and almost everything is easily solveable 11:05 < bridge> copium 11:05 < bridge> we do that in ddnet 11:05 < bridge> i still see segfaults in issues 11:05 < bridge> and else you'd just panic 11:06 < bridge> it's still a crash xd 11:06 < bridge> panics are controlled 11:06 < bridge> segfaults arent 11:06 < bridge> its far from the same 11:06 < bridge> a panic is like a thrown exception in c++ kinda 11:06 < bridge> anyway being panicless is easier than being segfaultless 11:06 < bridge> You wouldn't reach the state of a panic unintentionally like you would a segfault 11:07 < bridge> but its not comparable 11:07 < bridge> panic != segfault 11:07 < bridge> also panics can have user friendly messages 11:07 < bridge> u can pinpoint a issue faster with that msg 11:07 < bridge> a segfault a user faces usually just has an address 11:07 < bridge> sometimes i just don't like ownership stuff in rust. i feel like i could write the same thing in c or c++ in less loc 11:08 < bridge> the crate system and stuff are done really well though 11:08 < bridge> i dont care about locs that much 11:08 < bridge> rust maybe has more locs due to style not code itself 11:08 < bridge> "uwu, my tummy hurts at reading the file, file not found" 11:08 < bridge> in fact c++ has longer names usually too 11:08 < bridge> hi mom i want a std::pmr::monotonic_buffer_resource 11:09 < bridge> im not talking about style xd 11:09 < bridge> i think it makes no sense to relate locs to ownership 11:09 < bridge> ppl dislike ownership because it 90% of time finds their code is faulty 11:09 < bridge> most ppl are dumb 11:09 < bridge> xd 11:10 < bridge> C++ gives me the sheer power to shoot my own leg off and I like it! 11:10 < bridge> you probs have some UB issues in your code 11:10 < bridge> and u wont know it 11:10 < bridge> idk i can do fancy smart stuff in c++ that juggles memory but if i know all the cases nothing goes wrong 11:10 < bridge> "if i know all the cases nothing goes wrong" 11:10 < bridge> said everybody 11:10 < bridge> ubsan is there 11:10 < bridge> yet it segfaults 11:10 < bridge> skill issue 11:10 < bridge> yeah ur skill too probs 11:11 < bridge> ye you can't always do it first try xd 11:11 < bridge> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=87191 11:11 < bridge> https://github.com/llvm/llvm-project/issues/75856 11:11 < bridge> https://github.com/google/sanitizers/issues/940 11:11 < bridge> just a quick google 11:11 < bridge> and already 3 misses 11:12 < bridge> rust doesnt miss, but it tradesoff some valid cases 11:13 < bridge> ubsan is nice, but its just a monkeypatch 11:13 < bridge> like the bare minimum 11:14 < bridge> have u considered zig? 11:14 < bridge> 940 is not even ub, 75856 is nothing you would not notice 11:14 < bridge> not yet 11:17 < bridge> hoping to see another great lang popping off 11:17 < bridge> building on rust greatness 11:17 < bridge> zig is not that tho 11:17 < bridge> but the comptime is interesting 11:38 < bridge> https://github.com/mstange/samply 12:31 < bridge> @heinrich5991 ur pr is missing a li tag 12:37 < bridge> thanks, will fix later 12:37 < bridge> what is li an abbreviation for? 12:37 < bridge> list item 12:38 < bridge> but the html element is just `

` 12:38 < bridge> @robyt3 maybe we can publish ddnet on https://f-droid.org/ once the android port gets out of beta 🙂 12:42 < bridge> with the announcement from @heinrich5991 (1287179183205781575) I wonder if you can use dns poisoning in order to make people download hacked clients. Does the ddnet client check the certificate before updateing? 13:05 < bridge> yes 13:05 < bridge> certificates are checked 13:05 < bridge> yeah. I plan to release a small update to the beta this week. if that doesn't add any news bugs we could think about building a real release for the website and f-droid. 13:07 < bridge> Maybe we should switch default graphics backend to OpenGL for Android though, since Vulkan is unstable for some users and the memory usage is also too high for some users 13:18 < bridge> can you give an example? 14:00 < bridge> @headshotnoby btw, old client versions are unbanned again, all the different ported versions should work aagain 14:04 < bridge> Ellipsis is python are so confusing. I thought its only used for a field that is required but can be set to None, aka code that doesn't exist yet, but apperantly the actual intended use is in the numpy which adds a multidimensional array types 14:05 < bridge> I'm using them as a placeholder in my code anyway :greenthing: 14:05 < bridge> 😄 14:39 < bridge> it's 1024x1024 and each grid tile is 64x64 px, but it can be noticed with basically any mapres with a symmetrical tile, and rotate it 180 degrees it will be 1 pixel off (this is how I originally found this issue), here's a picture with ddnet_walls and it still has this. 14:39 < bridge> 14:39 < bridge> also I think that code is ran regardless of if your renderer supports array textures, I don't see any check that would prevent it. 14:39 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1287755270310854697/debug_wall.png?ex=66f2b30d&is=66f1618d&hm=2e97f768cc82d08775ee9ed732ec92ef36911ebb180b83c1201b7acf5f1268ac& 14:39 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1287755270650462279/image.png?ex=66f2b30d&is=66f1618d&hm=68cc042d70d9ae97f4cf890070d508c367b891b62940daf1cf2eb806a821f300& 14:41 < bridge> this line checks the renderer https://github.com/ddnet/ddnet/blob/f9df4a85be4ca94ca91057cd447707bcce16fd94/src/game/client/render_map.cpp#L603 14:42 < bridge> but tru, your mapres appears correct 14:44 < bridge> have you tried out a different renderer backend? 14:44 < bridge> I have not. will try it now 14:44 < bridge> could you send me the map? 14:45 < bridge> (I'm on my laptop where map stuff is kind of a hassle) 14:45 < bridge> ohh so for some reason my render was opengl 1.5, switched to vulkan and it does not chop off the last pixel 14:46 < bridge> ah nice, yea opengl does some buggy stuff 14:47 < bridge> here's the map if you still want it 14:47 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1287757163489394741/debug_wall.map?ex=66f2b4d1&is=66f16351&hm=f523deb4ab9961753dfce16e3c282c83d6cd0a542251ea9779efbb8139d1920e& 14:49 < bridge> but I don't think it should ever chop off the last pixel even if your renderer doesn't support array textures 14:57 < bridge> yea its definitely not perfect 14:57 < bridge> my guess is that it solves some other problem, maybe texture bleeding, where you'd see the edge of the next tile or something similar 15:04 < bridge> Without the `- 1` you get small gaps between tiles when zooming in with OpenGL 1 15:04 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1287761533757034496/image.png?ex=66f2b8e3&is=66f16763&hm=9bbe68fca3296d4914ed77921d52c9f13ec1ec57ebb2b83dec1b83e1a06a8c7a& 15:05 < bridge> remove ogl1 support 15:08 < bridge> https://en.wikipedia.org/wiki/Parchive 15:08 < bridge> @heinrich5991 @learath2 did u know about this? 15:10 < bridge> no, what's the use case? 15:10 < bridge> Nope, looks interesting 15:10 < bridge> I'm too young to have used Usenet. It was just about to perish when I started internetting 15:11 < bridge> I still see those either way, but I guess with the -1 it's less visible 15:11 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1287763168369705021/image.png?ex=66f2ba68&is=66f168e8&hm=c117c313095199d3780e3ffb9b89ba000aa3e0eaed26bb9146b282d314b82447& 15:11 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1287763168960970877/image.png?ex=66f2ba68&is=66f168e8&hm=73c9c61bfe84ccb2aa55a320302b394797356f7b653f7d5f32d438c9a558434f& 15:22 < bridge> Seem like it's related to the mipmap level. You can replace all `GL_LINEAR_MIPMAP_LINEAR` with `GL_NEAREST_MIPMAP_NEAREST` and `CL_LINEAR` with `CL_NEAREST` and remove the `- 1` to get rid of the gaps, but everything will be more pixelated then. 15:22 < bridge> (replace in `backend_opengl.cpp`) 15:25 < bridge> usenet is big 15:53 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1287773829275582564/image.png?ex=66f2c456&is=66f172d6&hm=f55a8837552550a50ec1f1bc1c905a16e8944ca2b07c6306e574e750c55de3a2& 15:53 < bridge> building a project with llvm deps kek 15:53 < bridge> in rust 15:53 < bridge> well its more a linker thing tbh 15:55 < bridge> yeah they work now 16:01 < bridge> https://rust-lang.github.io/async-book/01_getting_started/01_chapter.html 16:28 < bridge> it has use? i just use it like pass 😹 16:35 < bridge> @patiga Unable to find installation candidates for twmap (0.6.2) on MacOS M2 via Poetry 😄 17:21 < bridge> I haven't looked into cross compilation to macos yet, currently there is no twmap for macos on pypi 17:22 < bridge> Yeah 😦 Noticed too 17:23 < bridge> you can ofc always build the python module locally 17:23 < bridge> if you do start to use it on macos regularly, I'll try to get it into my release process 17:23 < bridge> Well, I tried to integrate the compile with poetry, but I havent managed so far 17:24 < bridge> I would use it on a atleast weekly basis as I have to stay long in the office, so I can work on it our maps-workflow 😄 17:25 < bridge> I'll take a look tomorrow or in a couple of days later. if I forget it and you have that problem again, feel free to ping me :) 17:26 < bridge> Sure, I can try to build too & submit it into the build process if you want to 17:26 < bridge> Maybe gitlab ci? 17:27 < bridge> would be neat :) 17:53 < bridge> i've been using neoqwertz as my keyboard layout for a little over a week now and i'm loving it so far, highly recommend it for people coding alot on a german keyboard layout, much easier to type things like {} () " ' | ~, ( and since it's an executable it's not as annoying to setup like other keyboard layouts) 18:00 < bridge> <01000111g> just dont use german keyboard :kek: 18:01 < bridge> <01000111g> I use US keyboard with EurKey layout (US with alt+a = ä etc) 18:03 < bridge> i very much dislike the US key layout :P 18:05 < bridge> so beautiful :harold: 18:05 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1287806940172390470/image.png?ex=66f2e32c&is=66f191ac&hm=3def196dc40a2dc9850d757e9a726ebecd1bf932995c92a60d3284a4a41da0db& 18:14 < bridge> ah yes 18:14 < bridge> teero when try it out 18:16 < bridge> tf 18:17 < bridge> what us key layout is that? 18:17 < bridge> mine doesnt look like that 18:17 < bridge> nah its the one i currently use :kekw: 18:17 < bridge> it uses modifiers so i dont have to lift of my fingers to type anything, pretty neat if you get used to it 18:17 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1287810167915810877/image.png?ex=66f2e62e&is=66f194ae&hm=ca4fca59dc77c05c395af28a6eb4907f514dedcfec2ce4fd754463df1ce698c8& 18:17 < bridge> wtf 18:18 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1287810298119585883/PXL_20240923_161752871.jpg?ex=66f2e64d&is=66f194cd&hm=af6db9f0097f6edf840d3713982474e7cff38ce57ded77ba482628dd8565029c& 18:18 < bridge> qwerty is ok for typing but ye lots of coding symbols are in annoying places 18:18 < bridge> which mac? 18:18 < bridge> i saw macos is getting a rly nice update 18:18 < bridge> m3 max pro iirc, it has 36gb ram 18:19 < bridge> :pepeW: how much was it 18:19 < bridge> @blaiszephyr i think its over 3.5k€, but i got it for free my company gave it for work 18:19 < bridge> ayo??????????? 18:19 < bridge> its not mine but i can use it for anything 18:20 < bridge> as long as i work here xd 18:20 < bridge> lmao are those math symbols latex or unicode characters 18:20 < bridge> if u want ur devs to be productive, specially in rust, provide them with good hardware kek 18:20 < bridge> but the cost to $$ ratio is insane 18:21 < bridge> only bad companies dont spend on hardware for devs 18:21 < bridge> i think for example in nvidia 1 dev equals millions in profit 18:21 < bridge> or more 18:21 < bridge> hard to estimate tbh xd 18:21 < bridge> but when im at home i use my nasa pc 18:22 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1287811242261741611/image.png?ex=66f2e72e&is=66f195ae&hm=5a58478a99d50ab7044e518be495f60d86af2a15266ec8ba187209b87e10a3d5& 18:22 < bridge> its funny i swapped when i compiled some rust stuff 18:22 < bridge> where 1440p 18:22 < bridge> both 18:22 < bridge> otherwise nasa pc 18:22 < bridge> gonna build cyberdeck with that layout 18:23 < bridge> idk i never got into 1440p 18:28 < bridge> ``` 18:28 < bridge>