00:00 < bridge_> I don't think so 00:15 < bridge_> Thought about it a bit, update: I don't think it's possible to do this without changing the calls and without performance hit 00:17 < bridge_> It is kinda weird that something so simple isn't really possible in C++ without some insane metaprogramming magic that I cant even think about 00:24 < bridge_> Yes ddnet 00:24 < bridge_> (@yekrm) 00:27 < bridge_> nuh uh 00:27 < bridge_> not until you finish your pr 00:29 < bridge_> My pr is already playable 01:01 < bridge_> @robyt3 any plans for working on https://github.com/ddnet/ddnet/issues/5654#issuecomment-1612162585 ? 01:34 < bridge_> ryo if i go back to primarily running Linux on my desktop, do i wanna use gentoo 01:34 < bridge_> I’m not really familiar with the benefits 01:34 < bridge_> I like arch 01:35 < bridge_> yeah i already have used arch for a few years and i like it too 01:35 < bridge_> but i want to try new thing 01:35 < bridge_> in gentoo you spend quite some time compiling stuff AFAIK, but I've never actually tried it, so take my advice with a grain of salt 01:35 < bridge_> ah, then go ahead I guess 01:35 < bridge_> I heard nix or qubes is also interesting 01:35 < bridge_> For a while I was running Windows so I could develop Windows specific apps but I haven’t done it in a while 01:35 < bridge_> yeah i like nix 01:35 < bridge_> interesting packaging 01:36 < bridge_> unfamiliar with qubes 04:19 < bridge_> IMPORTANT QUESTION: 04:19 < bridge_> Would these cursors go inside of `weapons` or `hud`? 04:19 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143368707645636740/weapon_grenade_cursor.png 04:19 < bridge_> I think it'd make more sense if this went into weapons, but at the same time, this is a hud element, or more accurately a gui element. 04:59 < bridge_> bah 04:59 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143378888471359528/image.png 05:15 < bridge_> I forgot how pointers worked. How do I get the value of the name of the assigned pointer for each sprite element? 05:15 < bridge_> For example: `container.sprites.Add(Sprite("part_slice", set_particles, 0,0,1,1))` 05:15 < bridge_> I'm trying to print `particles`, since that's the name of the set it's being pointed to, but I keep getting an error. 05:15 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143382983617806336/image.png 05:17 < bridge_> I forgot how pointers worked. How do I get the value of the name of the assigned pointer for each sprite element? 05:17 < bridge_> For example: ```py 05:17 < bridge_> container.sprites.Add(Sprite("part_slice", set_particles, 0,0,1,1))``` 05:17 < bridge_> 05:17 < bridge_> I'm trying to print `particles`, since that's the name of the set it's being pointed to, but I keep getting an error. 05:17 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143382983617806336/image.png 05:23 < bridge_> nvm i found out how ^^ 05:23 < bridge_> i just forgot to put `.target` before `.set` ^^ 05:31 < bridge_> why python 05:31 < bridge_> Should I keep waiting on this? https://github.com/ddnet/ddnet/pull/6775 05:32 < bridge_> @heinrich5991 05:33 < bridge_> why not 05:33 < bridge_> we are cooking tonight 05:33 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143387449523961876/image.png 05:34 < bridge_> @heinrich5991 good progress on the sorting phase! i think tommorow im going to set up the actual slicing tool! 05:35 < bridge_> What I did was reuse content.py (crediting matricks of course), and edit it a bit so that it includes what location it should be sorted into 05:35 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143387972226519111/image.png 05:37 < bridge_> A 05:39 < bridge_> well here's my plan: i'm using Pillow for Python, and I'm going to be setting it up in a way where you can put all the assets into its own custom folder, and you can run the program and have those assets packaged into its own set 05:40 < bridge_> because if i were to do it with an executable, it's just going to package all the default ones and skip the juicy custom assets (because lets be real i dont think most of the asset modders actually replace the main images. if they do id be surprised) 05:43 < bridge_> why is that what an executable would mean 05:44 < bridge_> because how would it know how to sort the custom assets 05:45 < bridge_> personally i would just do like dilate and have the first arg be a source img and have a second optional arg for output destination for CI or w/e 05:49 < bridge_> like, you drag & drop an image called foo.png and it would slice the images and put them in ./assets/foo/*.png or whatever 05:50 < bridge_> or invoke it from the command like & provide foo.png & an output path 05:50 < bridge_> i mean the thing is, will it just package the single image? 05:51 < bridge_> Wdym 05:51 < bridge_> because i was planning to also resort the entire thing because 05:51 < bridge_> look at how unorganized game.png is 05:51 < bridge_> and why exactly isn't extras a part of particles.png 05:52 < bridge_> ./assets/foo/game/*.png, ./assets/foo/extras/*.png 05:52 < bridge_> ./assets/foo/game/*.png, ./assets/foo/extras/\*.png 05:52 < bridge_> yes but that also begs the question: what happens when you go two game.pngs one after the other? 05:53 < bridge_> will the files overwrite? 05:53 < bridge_> ./assets/bar/*/\*.png 05:53 < bridge_> whats bar? 05:53 < bridge_> oh wait 05:54 < bridge_> ok so that means if i have a custom game.png called Awesome.png, and i put that into the slicer, will it be put into `assets/awesome/game/*.png`? 05:54 < bridge_> Yes 05:54 < bridge_> well 05:55 < bridge_> ./assets/Awesome/game/*.png 05:55 < bridge_> So you're telling me 05:55 < bridge_> I have to rename ALL of these to awesome so that it's completely organized and I can send out the zip file for it to embed properly? 05:55 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143393029462437889/image.png 05:55 < bridge_> So you're telling me 05:55 < bridge_> I have to rename ALL of these to Awesome so that it's completely organized and I can send out the zip file for it to embed properly? 05:55 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143393029462437889/image.png 05:56 < bridge_> because idk how the zip reading thing will work but i can only imagine its going to be like a game pack 05:56 < bridge_> unless each and every sub category is going to be its own zip 05:56 < bridge_> are you reading the zip at runtime 05:57 < bridge_> are we? 05:57 < bridge_> im literally only writing the slicing script 05:57 < bridge_> i lack context 05:57 < bridge_> now that i think about it how WILL the zip files be treated? 05:57 < bridge_> i will now stop coding until i get further context onto how ddnet will use these zip files 06:00 < bridge_> if i were doing it, i would add an option to the slicer to package the input asset as an uncompressed tarball 06:00 < bridge_> and the game could transparently read this tarball or each file individually in a neighboring folder 06:00 < bridge_> so ./assets/whatever.tar would be equivalent to the ./assets/whatever/*/*.png 06:01 < bridge_> so ./assets/whatever.tar would be equivalent to the ./assets/whatever/*/\*.png 06:01 < bridge_> and that’s how it could be packaged 06:01 < bridge_> perhaps the tarball could be the default behavior and the tool would place the tar either in the pwd or neighboring the source file 06:02 < bridge_> i put my question in #6753 06:02 < bridge_> https://github.com/ddnet/ddnet/issues/6753 06:02 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143394746534989954/IMG_3556.png 06:04 < bridge_> xd 06:33 < bridge_> anyway yeah if i was to implement it i would add it as a tool 06:33 < bridge_> call it asset_slice perhaps 06:33 < bridge_> seems to match the grammar in place for the map manip tools 06:33 < bridge_> i would provide a generic way to bind a source image to multiple outputs with defined regions, names, and categories 06:34 < bridge_> so you could have multiple layouts (like for old gameskin layouts? is that a thing?) 06:53 < bridge_> Don't use zip files. It makes content creation more annoying 06:54 < bridge_> Deploy is the only pro thing about it 06:54 < bridge_> But zipping your files every time u change something sucks 07:20 < bridge_> Same as having to use a new game.png imo 07:20 < bridge_> Distributing a game skin comprised of multiple files in a non-archive form would really suck 07:21 < bridge_> A game png would directly be an image 07:21 < bridge_> No need to zip it 07:22 < bridge_> Also zipping PNGs is basically useless 07:22 < bridge_> Or does Windows automatically disable compression 07:22 < bridge_> Default behavior 07:23 < bridge_> Well i find it ok. The downsides of zips are bigger 07:23 < bridge_> No file preview sucks extremely 07:25 < bridge_> If we ever have a skin system like 0.7 . One always has to download the whole zip. Read the whole zip, just to load a part of it 07:40 < bridge_> Maybe import button for deploy xd 07:41 < bridge_> Anyway,i feel like the zip feature is not a clear improvement. Even over what we have now. The pros and cons don't justify it. Maybe we should discuss it more before voxel wates all his time 08:29 < bridge_> What could happen is you can have whole ‘resource packs’ and you can choose their priority within the client 08:29 < bridge_> Like Minecraft 08:29 < bridge_> Some sort of manifest in the top level of the archive describing which assets are provided in the archive 08:29 < bridge_> Cuz ppl already have matching gameskins + entities + whatever else 08:30 < bridge_> So you want to zip a whole data directly basically? 08:30 < bridge_> Directory 08:31 < bridge_> Maybe. I was only thinking images at first but no reason not to have sounds like that too, IDK if ppl have custom sounds yet? 08:31 < bridge_> Also not skins 08:31 < bridge_> Skins should be separate 08:31 < bridge_> Some have custom sounds 08:31 < bridge_> But unless the need is presented then I don’t think it’s worth it 08:32 < bridge_> Going from vaguely categorized images to this elaborate system would be a big jump 08:32 < bridge_> I just think of ppl like broso who have custom entities, gameskin, maybe a map file for entities bg & an autoexec’d cfg file to set that up 08:32 < bridge_> Anyway.. from deployment POV it's cool.. i just wonder if it sucks from creator perspective to have 20 instances of inkscape open for all individual files. Then have to zip everything every time u test ur mod 08:33 < bridge_> And ultimately maybe have higher load times xd 08:33 < bridge_> In my theoretical implementation it would be straight from existing game.png form to archive 08:33 < bridge_> Archive would be an option on by default 08:33 < bridge_> So we support multiple variations 08:34 < bridge_> In other words, bloat xd 08:34 < bridge_> Higher load times only if decompression is needed, which uncompressed tarballs don’t present 08:34 < bridge_> ❓ 08:35 < bridge_> Well this sounds like archive and legacy work both 08:35 < bridge_> What tar balls xd 08:35 < bridge_> U didn’t read what I said earlier? 08:36 < bridge_> Tar balls require 7z on Windows 08:36 < bridge_> Not if the tool does it for them 08:36 < bridge_> But yes i apparently haven't read it 08:39 < bridge_> Tar is incredibly simple 08:40 < bridge_> It is. But it makes previewing impossible on windows 08:40 < bridge_> For normal userd 08:40 < bridge_> S 08:42 < bridge_> Idk why you would want to but fair 08:42 < bridge_> Preview by loading it in the game 08:42 < bridge_> Way more effective than trying to guess how itll look ingame 08:45 < bridge_> Uncompressed zips would also be a good option but unfortunately they are hard to make for most people 08:46 < bridge_> It requires specific cmd line invocation & I don’t remember the option existing in archivers like file roller on Linux, and I don’t think you can do it at all on Windows without the command line version 08:47 < bridge_> It would be less bug prone if we expect users to use the ddnet tool to create these archives. Potentially decompress them too, if that’s what theyre into 08:48 < bridge_> asset_pack/asset_extract? or auto detect task by file extension or lack thereof of first arg 08:48 < bridge_> asset_pack/asset_extract? or one tool which auto detects task by file extension or lack thereof of first arg 08:49 < bridge_> Unrelated but ATX case recommendations? Ideally space for 3-4 2.5” drives and potentially radiator(s) 08:49 < bridge_> And no lights. If lights, no rgb 08:54 < bridge_> dunno, smth like this maybe: 08:54 < bridge_> https://www.amazon.de/NZXT-Flow-Dual-Chamber-ATX-PC-Gaming-Geh%C3%A4use-High-Airflow-perforiertes/dp/B0BJHHPCQL/ 08:55 < bridge_> maybe that's overkill for you 08:55 < bridge_> Dell precision t7400 case xd 08:55 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143438212170256407/Dell_t7400_s_compeve.jpg 08:55 < bridge_> Possibly overkill, I tried looking at NZXT cases specifically but some cases just don’t have pics of the back of the interior (right side from the front) 08:56 < bridge_> Which is where those drives are in my current case. Thermaltake something 08:56 < bridge_> Lol 08:58 < bridge_> I suppose most cases have enough space in the front for a rad + fans if they’re open on the front, right? My current one is not water cooling focused at all but I bet it would fit. I would only do an AIO anyway, no need to shell out all those extra bucks for the water cooling accommodations 08:58 < bridge_> https://www.mindfactory.de/product_info.php/Inter-Tech-CXC2-mit-Sichtfenster-Midi-Tower-ohne-Netzteil-schwarz_1233721.html 08:58 < bridge_> i have this for my home server 08:58 < bridge_> but i turned off the fans, they were bit annoying xD 08:59 < bridge_> Looks pretty good. I kinda like how the drives mount. A bit dated looking though 08:59 < bridge_> I wish I could show you the one I have. It’s a fantastic case. Just small 08:59 < bridge_> i bought it in 2019 apparently 😂 09:01 < bridge_> https://a.co/d/hg8by2U 09:01 < bridge_> This looks decent 09:02 < bridge_> Ah this is what I have 09:02 < bridge_> https://thermaltakeusa.com/products/s100-tempered-glass-micro-chassis-ca-1q9-00s1wn-00 09:04 < bridge_> No photos on the site showing how the drives mount. There are 2 detachable plates on the back which you can screw the drive into . The plate itself has 4 screw holes for the drive, a leveled lowered part so you can insert it into the case and use only 1 fastener (which is a thumbscrew on the other side) 09:06 < bridge_> Also one full size hdd mount thing on the bottom that I could never figure out how to use. It seemed to hold my excess cables just fine. Now that I have a modular power supply it’s got no use 09:08 < bridge_> Looks like its big brother is the S300. Nowhere to get it from, though 09:16 < bridge_> Might grt the NZXT H5 Flow 09:16 < bridge_> I’ll have to sleep on it 09:22 < bridge_> What’s the smallest AIO i can comfortably get away with using for my 7900X, do you think? 09:26 < bridge_> Any 09:26 < bridge_> Would it perform better if I got an air cooler at the same price range 09:26 < bridge_> Would it perform better if I got an air cooler at the same price range as a small AIO 09:27 < bridge_> Probably. If the sink is big 09:27 < bridge_> Yea hmm 09:28 < bridge_> I can get nhd15 for about as much as the avg 240mm aio. maybe a bit less 09:28 < bridge_> It might also be more durable 09:28 < bridge_> True 10:25 < bridge_> ey is it easy to plug in shaders into ddnet? 10:29 < bridge_> fractal design does good cases 10:29 < bridge_> https://www.fractal-design.com/products/cases/ 13:49 < bridge_> everytime i use the ingame updater and the game restarted the resolution is wrong, the edges are a little cut off and the mouse sense feels like with vsync on, aka delayed and not as precise 13:49 < bridge_> since a few updates, cant recall which exactly 14:36 < ChillerDragon> why mfing cargo breaks half of the ci runs 14:37 < ChillerDragon> average rust stability 14:42 < bridge_> it detects if you work on bot content 14:44 < bridge_> also lower load times for uncompressed zips, probably. isn't that great? performance! 14:44 < bridge_> I'm willing to provide a tool that lets creators use the old "everything in one image" way of doing things 14:45 < bridge_> the discussion issue is over here: https://github.com/ddnet/ddnet/issues/6753 14:47 < bridge_> useless, when u want smth u do it anyway. i will not invest too much time into the legacy stuff anymore and just go full ego with my own stuff then rq xd 14:47 < bridge_> since when was this legacy? 14:48 < bridge_> once this goes in this will give people more freedom to add in content 14:48 < bridge_> for me the cpp codebase is legacy 😂 14:48 < bridge_> rusthead 14:48 < bridge_> well i want to see your workflow for new content 14:48 < bridge_> is the speed of a speedup just a unsigned char in the src or is it just limited by the editor? 14:49 < bridge_> first step is replacing most texture atlases with single images 14:49 < bridge_> it is in the source indeed 14:49 < bridge_> answered my own 14:49 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143527378480545792/image.png 14:49 < bridge_> ye 14:49 < bridge_> ;(( 14:49 < bridge_> wanted to make mega boost.... 14:50 < bridge_> yep.. even if i generally like it. I have huge concerns about usuability.. do you open 20 inkscape instances then when modding? 14:50 < bridge_> btw why are the mapsizes ints instead of unsigned ints? 14:50 < bridge_> the thing with lua would be cool for mapping scripts 14:50 < bridge_> or do you draw in a atlas and convert it 14:50 < bridge_> lmao, why does everyone like lua 14:50 < bridge_> no one does 14:51 < bridge_> no, you just open the game.png and then split it afterwards 14:51 < bridge_> but its way simpler to implement than other scripting languages. at least thats what ive heard 14:51 < bridge_> and for the new stuff? 14:51 < bridge_> you edit those seperately 14:52 < bridge_> would that not have a huge performance impact while loading all the images? 14:52 < bridge_> so there are your 20 instances? XD 14:52 < bridge_> or I'm going to provide a script to re-assemble it into some new form of game.png for you 14:52 < bridge_> someone's making a genetic multithread image loader 14:52 < bridge_> just for editing 14:52 < bridge_> someone's making a generic multithread image loader 14:52 < bridge_> no, because they're in a zip file, it's all one file for the file system 14:52 < bridge_> ah okay 14:53 < bridge_> i saw the pr. yea 14:53 < bridge_> ah okay 14:53 < bridge_> well first proove that this is even faster 14:53 < bridge_> without zip, you can more easily opne them with many cores 14:53 < bridge_> you didn't prove that it's slower, either 14:53 < bridge_> just above your message i give a legit reason tho 14:54 < bridge_> other than zips, which uncompress compressed images 14:54 < bridge_> this is exciting. imagine in the future being able to brodcast what theme of weapons you're using through the server. since most of the useless images in a pack arent being loaded itll much more efficient than it used to be before 14:55 < bridge_> how do you think it will help not loading useless parts? 14:55 < bridge_> you always at least have the overhead of the zip header, independent of the content 14:55 < bridge_> they do not, since we'll be using uncompressed zips 14:56 < bridge_> I find it disingenious that you ignore this, not even address why you think it might be false 14:56 < bridge_> bcs your claim is ppl open the zip and swap stuff out 14:56 < bridge_> i'd like to know if windows leaves the zip uncompressed in this case 14:56 < bridge_> probably not 14:57 < bridge_> but I'd guess most people will not do that, and for those, I claim, it'll be faster 14:57 < bridge_> ok, on what machine? 14:57 < bridge_> can i assume 6 cores? pretty standard 14:57 < bridge_> sure 14:58 < bridge_> and ofc a SSD 14:58 < bridge_> yes 14:59 < bridge_> and probably windows ^^ 14:59 < bridge_> so we have to assume windows defender too? 14:59 < bridge_> ah 😦 14:59 < bridge_> why is that even a thing 15:00 < bridge_> lol i just discovered a bug. 15:00 < bridge_> i made a map 10,000,000 big. 15:00 < bridge_> i can only fall upto y 524288 15:00 < bridge_> windows defender probably makes opening/closing many files even slower 15:00 < bridge_> im thinking a tiny bit. should `hit1` and `explo1` stay inside particles or should it be grouped with the weapons, since the gun muzzle flashes are also grouped there 15:00 < bridge_> but it might analyze the whole zip.. that's not better 15:01 < bridge_> not like itd do anything, it all just images 15:02 < bridge_> mhh good question, i also thought about that recently. One could say an explosion is only the cause of a projectile, which is unrelated to the weapon. 15:02 < bridge_> But i wouldn't know a good category for that.. It's kinda physic effects 15:03 < bridge_> are they currently in particles? 15:03 < bridge_> muzzle flashes are in game/weapons. explo and hit are in particles 15:04 < bridge_> and projectile trails? 15:04 < bridge_> pistol bullet trials are in game i think. smoke particles are in particles 15:04 < bridge_> pistol bullet trails are in game i think. smoke particles are in particles 15:05 < bridge_> mh ok.. well gun trails are similary sized to the gun projectile. so maybe that could be the logical argumentation.. But consequently it should be a particle i guess 15:05 < bridge_> i dunno rn how the ddnet source code handles it 15:06 < bridge_> i think when we start moving forward with this we should name the assets things that make more sense 15:06 < bridge_> seems like it's a particle in the source code 15:07 < bridge_> like why is the rifle still named laser? and the pistol simply named Gun? 15:07 < bridge_> hello, I was banned from Multimeyasmap, it was supposed to be opened at this time, but it didn't open, my little brother threw people or something, sorry 15:08 < bridge_> rifle is weird yes 15:08 < bridge_> 15:08 < bridge_> pistol is ok? 15:08 < bridge_> 15:08 < bridge_> gun sounds like the word group for a pistol 15:08 < bridge_> 15:08 < bridge_> e.g. a shotgun is also a gun 15:08 < bridge_> or is that your complain? XD 15:09 < bridge_> the time is in UTC , convert it using the internet 15:09 < bridge_> how do i use the internet 15:10 < bridge_> you already do, so that's not an excuse 15:10 < bridge_> "What is 17 o'clock UTC in Paris" 15:11 < bridge_> wdym why is rifle weird 15:11 < bridge_> I don't know Turkish 15:11 < bridge_> I don't know Turkish 15:11 < bridge_> it's english 15:11 < bridge_> better than saying "lemme get laser" "i am holding a laser" 15:11 < bridge_> shotgun is also laser in ddnet 15:11 < bridge_> so you want to name the laser rifle? 15:12 < bridge_> to me shotgun in ddrace is actually a new weapon: puller or smth 15:12 < bridge_> it simply used shotgun, bcs that's unused 15:14 < bridge_> I never call it rifle ingame, I'm always using "laser" 15:14 < bridge_> also "ninja", not "katana" 15:16 < bridge_> unfreeze_gun 15:16 < bridge_> pull_gun 15:16 < bridge_> useless_gun 15:16 < bridge_> xdd 15:16 < bridge_> i guess in terms of modes it makes a bit more sense 15:16 < bridge_> grenade_gun 15:16 < bridge_> alert_gun 15:17 < bridge_> hammering_gun 15:17 < bridge_> ok fine no renames 15:17 < bridge_> except pistol 15:17 < bridge_> gun_gun 15:18 < bridge_> Voxel what's your current structure tree? 15:18 < bridge_> haha 15:18 < bridge_> that would go so hard tho 15:19 < bridge_> imagine mods with actual different weapon visually 15:19 < bridge_> for how assets are organized? 15:19 < bridge_> yep 15:19 < bridge_> well how they are structured xd 15:20 < bridge_> @jupeyy_keks check at the end. are those just floating point errors? 15:20 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143535064139968665/2023-08-22_15-15-18.mp4 15:20 < bridge_> things are buggy in the distant land of teeworlds 15:20 < bridge_> can you just render what u mean xD if it's at the end 15:20 < bridge_> no no 15:20 < bridge_> need the gui 15:20 < bridge_> i have folders for: 15:20 < bridge_> - weapons 15:20 < bridge_> - particles 15:20 < bridge_> - map_items 15:20 < bridge_> - hud 15:20 < bridge_> - gui 15:20 < bridge_> - emoticons 15:21 < bridge_> i get stuck at 524288 y 15:21 < bridge_> ok, how for example does weapons look like 15:21 < bridge_> im mobile 15:21 < bridge_> does it simply have dirs: 15:21 < bridge_> - gun 15:21 < bridge_> -shogtun etc 15:21 < bridge_> windows defender probs check zip files.more aggressive so its slower on those 15:22 < bridge_> each folder just has all the respecting weapon assets in it. i dont want it to go too deep into folders for the user's sake 15:22 < bridge_> a png is a simple header check i think, zip needs uncompression(?) 15:22 < bridge_> ryo we established it wont be compressed 15:22 < bridge_> for the AV check 15:22 < bridge_> still windows defender has to check it 15:22 < bridge_> hashes or whathever it does 15:23 < bridge_> it has to "unzip" it 15:23 < bridge_> ok.. we could at least think about smth like this 15:23 < bridge_> weapons/default/... 15:23 < bridge_> weapons/my_mod/... 15:23 < bridge_> weapons/another_mod/... 15:23 < bridge_> 15:23 < bridge_> In case we want to plan to allow multiple assets in future... How exactly do you imagine overwriting assets else? 15:23 < bridge_> i mean game.png should™️ be in assets like the game.png mods 15:23 < bridge_> i guess 15:23 < bridge_> another problem is what if you want to change weapon skins but not hud skins 15:24 < bridge_> like ddnet now 15:24 < bridge_> yep 15:24 < bridge_> ~~game.zip, particles.zip~~ 15:25 < bridge_> 😬 15:25 < bridge_> 😬 15:25 < bridge_> Do you know how to make a rainbow skin my brother is crying because of it 15:26 < bridge_> lol??? 15:26 < bridge_> Do you know how to make a rainbow skin my brother is crying because of it 15:26 < bridge_> why is your brother crying over ddnet 15:26 < bridge_> no 15:26 < bridge_> hold on lemme find the thing 15:26 < bridge_> @_voxeldoesartTHANK YOU 15:27 < bridge_> ChillerDragon: @jupeyy_keks i am the first tee in the teeworld to discover the farlands 15:27 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143536954806382672/524288_y.mp4 15:27 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143536955594907659/far_lands.mp4 15:28 < bridge_> Most frequent questions, especially related to binds can be found on the DDNet Wiki: 15:28 < bridge_> :wiki: Wiki page for binds: 15:29 < bridge_> rainbow tee is in the wiki 15:30 < bridge_> @_voxeldoesart haha i still remember once i wrote a script to make 255 cfg files to achieve rainbow xdd 15:30 < bridge_> with binds ofc 15:30 < bridge_> What's your x coordinate? 15:30 < bridge_> Or y 15:30 < bridge_> fascinating 15:30 < bridge_> 524288 y 15:30 < bridge_> 1.5 x 15:31 < bridge_> I c 15:31 < bridge_> Then yes 15:31 < bridge_> Floating point 15:31 < bridge_> floating point already at 500k?? 15:31 < bridge_> weird 15:31 < bridge_> Probably during quantinize 15:31 < bridge_> but why cant i go further?? 15:31 < bridge_> Why already 15:32 < bridge_> Its already inaccurate after a few thousand 15:32 < bridge_> Well the velocity probably quantiszes to 0 15:32 < bridge_> the velocity keeps going until max but i cant get my position further 15:32 < bridge_> ill try with teles 15:33 < bridge_> We conv the float to 1/256 or smth.. and the accuracy is already too bad 15:33 < bridge_> then why are we using floats? xdd 15:33 < bridge_> i remember a map being so long that you basicallu fell down and if you went fast enough you stop midair at the finish line 15:33 < bridge_> I dunno. Legacy reasons 15:33 < bridge_> @louis.place remember we raced 15:34 < bridge_> wtf? default client is limited to map to 10k in each direction xdd 15:34 < bridge_> how tw 15:34 < bridge_> how tf 15:34 < bridge_> you can easily reproduce with falling in blue tps too 15:34 < bridge_> my map is 10 mil 15:34 < bridge_> no the position 15:34 < bridge_> not the position 15:35 < bridge_> i meant this to stopping midair 15:35 < bridge_> hmm okay... 15:36 < bridge_> down60k or smth 15:36 < bridge_> xDDD 15:36 < bridge_> so im not the first one ;( 15:36 < bridge_> no you arent 15:36 < bridge_> i just said that 15:37 < bridge_> waat 15:37 < bridge_> If you reach rly high speed your speed starts getting multiplayed by a ramp value, i visulized it once here https://discord.com/channels/252358080522747904/342454474117218334/1064935260678262845 15:37 < bridge_> what are u saying 15:38 < bridge_> its not cuz of speed 15:38 < bridge_> maybe im talking about sth else 15:39 < bridge_> oh yes didnt see the video 15:40 < bridge_> hmm okay if im a little under 524288 y then the y coordinate just completely freezes 15:40 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143540146084520046/image.png 15:41 < bridge_> and i can do super long hooks ig 15:41 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143540337856495657/image.png 15:41 < bridge_> oh wow 15:41 < bridge_> i could only get my camera to far lands 15:41 < bridge_> using glitches 15:41 < bridge_> the width of the aimline is bugged too 15:41 < bridge_> i remember the tiles being stretched out 15:42 < bridge_> xDD 15:42 < bridge_> and tiles were placed where they shouldnt 15:42 < bridge_> some also werent visible 15:43 < bridge_> Yeah, we need physic changes xd 15:43 < bridge_> 15:43 < bridge_> Will be biggest drama ever xddd 15:43 < bridge_> what is happening with my aimline @jupeyy_keks ? 15:43 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143540963294326874/aimline-wtf.mp4 15:43 < bridge_> lmao 15:43 < bridge_> Maybe with a season 2 it would be acceptable.. 99.99% of the game would probably feel the same anyway 15:44 < bridge_> how tf does my position influence the width of my aimline? 15:44 < bridge_> i dont even have custom width enabled 15:44 < bridge_> Its bcs the hookpos 15:44 < bridge_> The hookline basically simulates a hook 15:44 < bridge_> So the hook is probably also longer 15:44 < bridge_> Or broken 15:45 < bridge_> All calculations that use the pos are basically broken with this 15:45 < bridge_> i dont think it does. or was it in a new update? 15:45 < bridge_> although ive never seen it severe as this 15:45 < bridge_> so here 🏆 15:45 < bridge_> Its always been that way 15:45 < bridge_> it does 15:45 < bridge_> in a map with hook tunes your hookline can become longer or shorter 15:45 < bridge_> then why can i mishook when going fast and the line is green? 15:45 < bridge_> yea 15:46 < bridge_> that just a multiplier 15:46 < bridge_> That's someone different 15:46 < bridge_> It calculated the hook as if you were not moving 15:46 < bridge_> yea 15:46 < bridge_> As soon as u move it's basically invalid what u saw 15:47 < bridge_> ima test some weapons down there 15:48 < bridge_> had to install an old client just to do this 15:48 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143542091570499594/screenshot_2023-08-22_15-47-35.png 15:48 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143542092266733598/screenshot_2023-08-22_15-47-43.png 15:48 < bridge_> textures bugged like that when i take camera far off the map border 15:48 < bridge_> at some point they just dissapear 15:48 < bridge_> and at some other point some tiles just stretch a bit 15:49 < bridge_> werent u trying to show me that a while ago souli 15:49 < bridge_> i did 15:50 < bridge_> even the editor is bugged at 10,000,000 y 15:52 < bridge_> Well everything is bugged with these floats xD 15:52 < bridge_> Minecraft also gets Buggy when doubles getting inaccurate 15:53 < bridge_> U can watch it on YouTube 15:53 < bridge_> damn thats a thick tele right there 15:53 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143543406904225812/image.png 15:53 < bridge_> Minecraft farlands 15:53 < bridge_> if you spawn there do you become ddfat 15:53 < bridge_> JANKY HITBOX 15:53 < bridge_> damn thats a thicc tele right there 15:53 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143543406904225812/image.png 15:57 < bridge_> AHAHAHAH 15:58 < bridge_> ill post a vid later 16:15 < ChillerDragon> @teero777 woah far lands! yikers. I remember years ago i tried creating a 10k x 10k map but my client crashed when i tried to open it xd 16:16 < bridge_> xd 16:16 < bridge_> mine is 10mil x 6 16:16 < ChillerDragon> gaming pc moment 16:17 < bridge_> with 6 i mean 6 not 6mil btw xdd 16:17 < ChillerDragon> a 16:17 < ChillerDragon> xd 16:17 < ChillerDragon> scam 16:17 < ChillerDragon> you flat earther? 16:17 < bridge_> nah 16:17 < bridge_> earth is horizontal 16:17 < ChillerDragon> xd 16:17 < bridge_> i mistyped 16:17 < bridge_> i mean 6 x 10mil ofc 16:42 < bridge_> I can make plans to work on it next 16:50 < bridge_> but will you? 16:54 < bridge_> It's high on my TODO list now 16:55 < bridge_> How long is your TODO list 😬 16:55 < bridge_> @teero777 I also experimented with large maps once. it is quite interesting that you can actually feel the difference between each segment of the map where the float inaccuracy get slightly worse 16:57 < bridge_> you can use the twmap python module to use a scripting language to produce maps :) 16:57 < bridge_> Undefined, my TODOs are mostly stored as a bunch of bookmarks in Firefox and as git branches with an mostly empty "TODO: xyz" commit 16:58 < bridge_> put performance todos on highest 😏 17:00 < bridge_> https://github.com/cppfastio/fast_io 17:00 < bridge_> > RUST IS SLOW. Also Rust does not deal with locale. Think how bad it is. 17:00 < bridge_> :santatrollet: 17:01 < bridge_> xD 17:01 < bridge_> well i tell ya, variadic templates beat everything 17:01 < bridge_> sadly most ppl seam to unaware just how awesome they are xd 17:02 < bridge_> (+ concepts ofc) 17:04 < bridge_> but i guess with procedual macros you can basically achieve the same what i would want 17:04 < bridge_> 17:04 < bridge_> so i dunno why it's slow in their mind.. but maybe standard rust is slow indeed xd 17:04 < bridge_> 17:04 < bridge_> i mean it looses against Lua 17:04 < bridge_> in hello world xd 17:04 < bridge_> shame 17:05 < bridge_> The benchmark where Rust loses was "Print out ten million integers from 0 to 10M to file. Then reopen that file to scan back." 17:05 < bridge_> i see 17:06 < bridge_> sounds like they might have missed buffering 17:06 < bridge_> unfortunately I can't find the source code they used 17:08 < bridge_> xd 17:08 < bridge_> issue opened: RUST AIN'T SLOW BRA 17:08 < bridge_> i kinda would like to see a benchmark game, but only with idiomatic code (per language) 17:09 < bridge_> then c wins everywhere lmao 17:09 < bridge_> there is no idiomatic code in c xd 17:10 < bridge_> https://www.youtube.com/watch?v=QWVPOXt24kM 17:10 < bridge_> kind of really low effort 17:10 < bridge_> but here you go 17:10 < bridge_> @teero777 r u a coder btw? 17:11 < bridge_> yea why? 17:11 < bridge_> i just do it for fun 17:11 < bridge_> well back when i was young and didnt know about such stuff like inaccuracies etc. 17:11 < bridge_> i always found such "bugs" mythical 17:11 < bridge_> 17:11 < bridge_> so i wondered if you feel the same xd 17:12 < bridge_> the only magical thing i have is memory corruption... 17:13 < bridge_> i remember i played a game where you were glitched high up in the air and fell for a long time 17:13 < bridge_> 17:13 < bridge_> it kinda gave me chills 17:13 < bridge_> 17:13 < bridge_> 17:13 < bridge_> same for stuff like walking out of maps (in 3d games) 17:13 < bridge_> 17:13 < bridge_> was always scarry 17:13 < bridge_> nowadays i only see it logically 😭 17:13 < bridge_> ahha whats up with the line spaces xDDD 17:14 < bridge_> lol, I think have that in Geoguessr with what players call "rifts" in the sky 17:14 < bridge_> back in elementary school, my sitting neighbour always put an extra line after every line 17:14 < bridge_> 17:14 < bridge_> and it always looked fancy 17:14 < bridge_> 17:14 < bridge_> i guess that's why i still do it up until today 😂 17:15 < bridge_> I wonder what happens if you remove the quantization 17:15 < bridge_> i once tried to beat zelda ocarina of time on my first pc. i had like 4 months of progress then i fell out of the map and my save file wasnt working after.... 17:15 < bridge_> lol, nice. never saw these things 17:16 < bridge_> You can see examples here: https://geohints.com/Rifts 17:16 < bridge_> that sounds rather depressing xd 17:16 < bridge_> those look cool 17:17 < bridge_> knowing less is sometimes better 17:18 < bridge_> well pretty often tbh xdd 17:18 < bridge_> there is a saying: stupid ppl are more happy 17:21 < bridge_> same, but with tw 17:21 < bridge_> i just wonder and think about the code that does X 17:23 < bridge_> and yes rust doesnt buffer by default 17:24 < bridge_> @jupeyy_keks ich actually do remember 1 magical thing 17:25 < bridge_> Finally Saint Ryo will save us xd 17:25 < bridge_> Ryo when open issue 17:25 < bridge_> blaming them for no skillz 17:25 < bridge_> lmao 17:26 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143566747136442479/image.png 17:26 < bridge_> totally no bias 17:26 < bridge_> I'd start by asking for the rust source code 17:26 < bridge_> tbh in case the benchmark is well done. it's good that it exists, it will make rust better 17:26 < bridge_> 17:26 < bridge_> but there are so many fcking libs(language independent), that just do synthetic benchmarks to look good 17:26 < bridge_> and that's just annoying 17:27 < bridge_> if they dont provide the source code 17:27 < bridge_> then all the readme is bs 17:27 < bridge_> for starters 17:27 < bridge_> it's clickbait xd 17:27 < bridge_> there were always some videos about pokemon ruby/sapphire on how to get X pokemon that is litteraly not avaiable ingame. it was basically just someone with a modded game doing specific steps to get the pokemon. i ofc didnt know it was modded and tried to repeat the steps in the videos for hours xdd 17:27 < bridge_> Robyte baited me clicking it because of that quote xDDD 17:29 < bridge_> well pokemon was such a huge deal back in the days.. 17:29 < bridge_> 17:29 < bridge_> as kids we didnt really know how to google. and back then there werent 2000 pokemon wikis 17:29 < bridge_> 17:29 < bridge_> so stuff like Mew gltich etc. were actually done over mouth propaganda 17:29 < bridge_> 17:29 < bridge_> And it worked so well.. Even tho 50% of these glitches didnt even work xDD 17:29 < bridge_> that was the real fakenews time xDD 17:29 < bridge_> xDDD 17:29 < bridge_> but it was fun 17:29 < bridge_> best time of my life tbh 17:29 < bridge_> @jupeyy_keks i would even dare say he ran rust in debug xd 17:30 < bridge_> quite possible 17:30 < bridge_> 17:30 < bridge_> don't trust the benchmark you didn't fake xdd 17:31 < bridge_> there is a chinese issue 17:31 < bridge_> that mentions all the issues of this repo 17:31 < bridge_> ``` 17:31 < bridge_> fast_io may have incomplete or unfair parts in its comparison with traditional streaming I/O. 17:31 < bridge_> 17:31 < bridge_> In terms of speed: 17:31 < bridge_> 17:31 < bridge_> Intel C++ is not included in the comparison (C++ 11 over C++ 20 may not be right, but performance comes first). 17:31 < bridge_> Failure to include the rest of the compilers that are part of the Windows GCC implementation in the list of comparisons, while picking the most mediocre performer, MinGW-w64. 17:31 < bridge_> Failure to provide (as much as possible) uniform compilation parameters. 17:31 < bridge_> When mentioning the speed of the Rust programming language, not mentioning whether to compile in Debug or Release. 17:31 < bridge_> 17:31 < bridge_> Translated with www.DeepL.com/Translator (free version) 17:31 < bridge_> 17:31 < bridge_> ``` 17:31 < bridge_> https://github.com/cppfastio/fast_io/issues/436 17:32 < bridge_> the collerberator seems chinese too xd 17:33 < bridge_> @robyt3 man how did u find that lib.. did you use google on chinese or what xD 17:33 < bridge_> GitHub recommendations 17:35 < bridge_> @jupeyy_keks this is the mother of compiler optimizations https://en.wikipedia.org/wiki/Static_single-assignment_form 17:35 < bridge_> does that say smth about my character? 17:35 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143569056474091580/image.png 17:35 < bridge_> yes 17:35 < bridge_> it's also my only recommendation xd 17:35 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143569245150654586/image.png 17:36 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143569303594090528/image.png 17:36 < bridge_> xD 17:36 < bridge_> https://github.com/carlini/printf-tac-toe 17:36 < bridge_> that's what i want to see 17:36 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143569485190660179/image.png 17:38 < bridge_> @jupeyy_keks 17:38 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143569899462082720/image.png 17:38 < bridge_> epic xd 17:38 < bridge_> didnt u need a encoder 17:39 < bridge_> i specifically need an audio encoder 17:39 < bridge_> in pure rust 17:39 < bridge_> best is if it would be vorbis 17:39 < bridge_> currently the only thing not rust is `ring` :/ 17:39 < bridge_> what does a audio encoder 17:39 < bridge_> i hope i can remove it some day 17:40 < bridge_> it encodes audio 17:40 < bridge_> xd 17:41 < bridge_> that's probably not removable. hmmm. maybe you could do it with inline assembly 17:41 < bridge_> yeah 17:41 < bridge_> sounds good enough 17:41 < bridge_> if rust compiler supports some asm 17:42 < bridge_> could rust itself standarize some instructions needed? 17:42 < bridge_> or what exactly is blocking it 17:42 < bridge_> rust supports inline assembly 17:42 < bridge_> it does 17:43 < bridge_> asm macro 17:43 < bridge_> https://doc.rust-lang.org/nightly/reference/inline-assembly.html 17:43 < bridge_> the problem is compilers. you can't write constant-time code in any non-assembly languages I know of 17:43 < bridge_> https://doc.rust-lang.org/rust-by-example/unsafe/asm.html 17:44 < bridge_> so between inline asm and ring, which ultimately also uses asm then.. is there any burdon? 17:45 < bridge_> burdon? 17:45 < bridge_> burden 17:45 < bridge_> ah, you mean perf implications? 17:46 < bridge_> i mean anything at all 17:46 < bridge_> is it all about perf or what? 17:47 < bridge_> there are no non-perf differences between rust inlien asm and actual asm AFAIK 17:47 < bridge_> i'd be ok to loose a few ms.. tbh 17:48 < bridge_> are the constant time algorithms used for certification creation? 17:48 < bridge_> are the constant time algorithms used for certificates creation? 17:48 < bridge_> I would guess it is about highly security related code, where you are happy that one implementation seemingly gets it right ^^ 17:49 < bridge_> yey 17:49 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143572545166123058/image.png 17:49 < bridge_> or are side channel attacks also a problem for the encryption and decryption itself 17:49 < bridge_> well crates.io has no binary crates.. so i need a c compiler 17:49 < bridge_> afaik, yes 17:49 < bridge_> that's kinda annoying 17:53 < bridge_> also encryption/decryption. maybe even particularly that 17:53 < ChillerDragon> heinrich sos 17:53 < ChillerDragon> watafak is up w `CNetObj_PlayerInput` 17:53 < ChillerDragon> is that a snap item thats not in the snap? :D 17:53 < ChillerDragon> but instead sent by the client 17:54 < ChillerDragon> which drunk trol decided thats making sense 17:54 < bridge_> `CNetObj_PlayerInput` is weird 😄 17:54 < ChillerDragon> isnt snap supposed to be items the server sends to the client 17:54 < ChillerDragon> watafak 17:54 < bridge_> you described it perfectly 17:54 < bridge_> It's probably a hack we use for demos 17:54 < ChillerDragon> thanks senpaii 17:54 < bridge_> not involved in demos AFAIK 17:54 < bridge_> it's just a "let's use this to generate a struct" AFAICT 17:54 < bridge_> 😄 17:55 < ChillerDragon> im so gonna document the shit out of this 17:55 < bridge_> Ew 17:56 < bridge_> ok but is inline asm different to normal asm? 17:56 < bridge_> 17:56 < bridge_> i still don't get what exactly is the difference between doing it in rust or c 17:57 < bridge_> c/rust makes no difference here 17:57 < bridge_> c is also too high-level for constant-time code 17:57 < bridge_> everything you can do in C is also possible in rust 17:57 < bridge_> ok i mean like 17:57 < bridge_> 17:57 < bridge_> if u rewrite the c part 17:57 < bridge_> what's blocking? 17:57 < bridge_> nothing, I'd guess 17:57 < bridge_> but I think ring also includes assembly? 17:58 < bridge_> well i guess that's not a problem is it? 17:58 < bridge_> you still need an assembler, which is sort of like a C compiler ^^ 17:59 < bridge_> so the rust inline asm, is architecture dependent (like let's say x86 asm)? while the assembler there is using a compiler? 17:59 < bridge_> the rust inline assembly is arch dependent 17:59 < bridge_> the actual assembler on the platform is another binary dependency you'd have to install 17:59 < bridge_> which is annoying, in a similar way as it's annoying to have to install a C compiler 18:00 < bridge_> so in conclusion the problem is: there have to rewrite the asm they already have? 18:01 < bridge_> so in conclusion the problem is: they have to rewrite the asm they already have? 18:01 < bridge_> yes 18:02 < bridge_> hmm 18:02 < bridge_> https://doc.rust-lang.org/core/arch/macro.global_asm.html 18:02 < bridge_> they should create an proc macro that reassmbles the asm to arch dependent asm 18:02 < bridge_> xd 18:02 < bridge_> this means they might not have to do it? 18:02 < bridge_> there is drama rn about sandboxing proc macros with wasm 18:02 < bridge_> and precompiling proc macros 18:03 < bridge_> i have seen that there was a crate that can do that 18:03 < bridge_> I don't understand the drama tbh. also, sandboxing proc macros does very little as long as there are build scripts 18:03 < bridge_> idk either 18:03 < bridge_> also: are compilers even secure enough? 18:03 < bridge_> im not rly into programming during my vacations xd 18:04 < bridge_> wanted a break 18:04 < bridge_> rust knows no break 18:04 < bridge_> but drama was started by dtolnay experiment 18:04 < bridge_> xD 18:06 < bridge_> tbh if you dont trust proc macros.. you also can't trust any crate 18:06 < bridge_> as if you are checking the source of the crate you use 18:06 < bridge_> xd 18:08 < bridge_> some people are doin it 18:08 < bridge_> I'm not, unfortunately 18:08 < bridge_> cargo-vet, I think 18:09 < bridge_> oh, cargo-crev as well 18:09 < bridge_> i have a cargo crev trust repo 18:09 < bridge_> sure but i just don't get where the source is different from the proc macro here 18:10 < bridge_> hm 18:11 < bridge_> well the thing is source doesnt execute code when compiling 18:11 < bridge_> proc macros execute code when compiling 18:11 < bridge_> ok but as soon as your start the exec.. it's over xd 18:13 < bridge_> really depends to which extend sandboxing will help anyway. 18:13 < bridge_> 18:13 < bridge_> they can't sandbox file access.. 18:13 < bridge_> that's not true. build scripts exist 18:13 < bridge_> true 18:13 < bridge_> idk 18:13 < bridge_> build scripts can also depend on arbitrary crates 18:13 < bridge_> i have no opinion on this regard on what they do, either way will be same for me i think 18:14 < bridge_> i just hope they don't destroy proc macros xDD 18:14 < bridge_> i find them cool and abuse them hard 18:14 < bridge_> its backward incompat 18:14 < bridge_> to destroy proc macros 18:14 < bridge_> so they cant xd 18:14 < bridge_> for security they might do 18:15 < bridge_> new rust edition 18:15 < bridge_> this isnt just a thing they thought now btw 18:15 < bridge_> backward compability sucks anyway 18:15 < bridge_> who invented that xd 18:19 < bridge_> He is bringing forbidden knowledge into this world 18:20 < bridge_> :gigachad: 18:21 < bridge_> @jupeyy_keks wasmtime or wasmer btw? 18:22 < bridge_> wasmer 18:22 < bridge_> it claims to be the fastest runtime xdd 18:22 < bridge_> that's a convincing argument for me 18:22 < bridge_> but i use cranelift.. and they both use it afaik 18:22 < bridge_> so they can't be too far off 18:23 < bridge_> except the one does some stuff overcomplicated 18:23 < bridge_> the biggest bottleneck i saw were simply the huge function calls.. like the amount of functions that are called until guest function is reached 18:24 < bridge_> https://00f.net/2023/01/04/webassembly-benchmark-2023/ 18:24 < bridge_> seems to agree, wasmer is a little bit faster than wasmtime 18:24 < bridge_> even wasmer cranelift 18:24 < bridge_> I don't understand this one 18:25 < bridge_> you mean rust functions? 18:25 < bridge_> is wasmer interpreter or JIT? 18:25 < bridge_> lets say u have a function `mod_run` 18:25 < bridge_> calling that function (in the wasmer API) seems to create a rather big call stack 18:25 < bridge_> jit probably, since it's using cranelfit 18:25 < bridge_> ah, I see 18:31 < bridge_> TIL `%` works on `#ifdef`s in vim 18:35 < bridge_> vim is so good 18:35 < bridge_> rip 18:35 < bridge_> there is also a vim gdb integration right 18:35 < bridge_> out of the box 18:35 < bridge_> ah 18:36 < bridge_> its a plug 18:36 < bridge_> :packadd termdebug 18:36 < bridge_> modern vim has a package manager btw 18:36 < bridge_> https://github.com/vim/vim/blob/master/runtime/pack/dist/opt/termdebug/plugin/termdebug.vim 18:37 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143584646727221328/image.png 18:37 < bridge_> https://www.youtube.com/watch?v=wmP3MBjsx20 18:37 < bridge_> they do so much nuclear stuff xD 18:37 < bridge_> xd 18:37 < bridge_> i love it 18:38 < bridge_> its always about massive changes 18:38 < bridge_> nuclear stuff 18:38 < bridge_> biology virus 18:38 < bridge_> space 18:38 < bridge_> let's play geo guesser 18:38 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143584938659156050/image.png 18:38 < bridge_> which countries are meant here xD 18:38 < bridge_> it said 18:38 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143585040459104336/image.png 18:38 < bridge_> us vs ducks 18:38 < bridge_> xDD 18:41 < bridge_> https://media.discordapp.net/attachments/293493549758939136/1139665875121750026/STOP_USING_VIM.png 18:53 < bridge_> use neovim instead 19:31 < bridge_> @mind 20:15 < bridge_> @jupeyy_keks https://rosenzweig.io/blog/first-conformant-m1-gpu-driver.html 20:17 < bridge_> :brownbear: 20:44 < bridge_> "it's not a bug, it's a feature." 20:44 < bridge_> I present you a bug that I found around 2016, which allows you to make a static camera in teeworlds 0.6.* and any DDNet client versions \😎 20:44 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143616798277779507/bandicam_2023-08-22_22-38-40-352.mp4 20:45 < bridge_> test it - `37.230.162.198:8303` 20:45 < bridge_> 37.230.162.198:8303 is not a DDNet or KoG server. 20:51 < bridge_> Matrix? 20:51 < bridge_> hm 🤨 20:53 < bridge_> How does it work ? 21:07 < bridge_> i cant test it because theres a long tutorial lol 21:07 < bridge_> kek 😄 21:08 < bridge_> relog now) 21:13 < bridge_> this will go so hard with volleyball 21:17 < ChillerDragon> 8min dl bra 21:18 < ChillerDragon> fakin flight simulator map matodor 21:18 < bridge_> ill try 21:19 < bridge_> In 2.2 there are camera triggers 21:21 < Huu> Ich hab mal ne frage wenn man ne map finnished wird die ms zeit ab oder auf gerundet oder ist das ein kaufmänisches runden 21:21 < bridge_> imagine ddnet camera triggers 21:21 < bridge_> "bro this gores map is so good I CANT SEE GRAHH" 21:25 < bridge_> Zeit wird noch nicht direkt gemessen sondern in gameticks. Ein gameticks ist 1/50 einer sekunde lang (0,02s). 21:25 < bridge_> also ab gerundet. 21:26 < bridge_> Zeit wird noch nicht direkt gemessen sondern in gameticks. Ein gameticks ist 1/50 einer sekunde lang (0,02s). 21:26 < bridge_> also abgerundet. 21:26 < bridge_> neither rounded down nor rounded up I think 21:26 < bridge_> Not directly but by how the game works 21:27 < bridge_> If you finish at say 1.57 seconds your finish time will be 1.56 seconds 21:27 < bridge_> Since you have not yet reached the next gametick 21:29 < Huu> I am wondering releated to nyan cat, there are 35.44 .46 .48 .50 .52 times https://ddnet.org/maps/Nyan-32-Cat 21:29 < Huu> I can reproduce 48 50 and 52 with an bug at start 21:30 < Huu> I saw also runs with 46 but no difference to the 48 run that why i am asking 21:31 < Huu> So 46 and 44 are for me still a weird that they exist 21:31 < Huu> Or can a potential 48 run can be counted as a 46 run just by tick luck? 21:32 < Huu> Since there is no deterministic 21:32 < bridge_> ddnet is completely deterministic except of teleport positions 21:32 < bridge_> given all information ofc 21:33 < Huu> Hmm then i dont get why my potential 48 run sometimes count as 46 21:33 < Huu> Thats kinda weird 21:33 < bridge_> Given ALL the information even those are deterministic now. We use a PRNG now 21:33 < bridge_> 🦔 21:34 < bridge_> The client just doesnt have the seed and the counter of the PRNG that's why it can't predict yet 21:34 < Huu> Not sure about that i can like show u an example where u get random results actualy on nyan cat 21:34 < bridge_> given all informatio.... 21:35 < Huu> where u get a 0,00 - 0,04 time disadvantage 21:35 < bridge_> is determinism bound to the os of the server? 21:35 < bridge_> the client checkpoint times that get shown aren't actually represented as ticks 21:35 < bridge_> They use actual time calculation 21:36 < bridge_> r u testing with a sv in windows? 21:36 < bridge_> the client checkpoint time diffrences that get shown aren't actually represented as ticks 21:36 < bridge_> curious maybe the time api has quirks xd 21:36 < bridge_> So those might be false 21:36 < bridge_> Never trust checkpoints 21:37 < bridge_> oh checkpoints 21:38 < bridge_> @ryozuki btw quick ddnet hack is to lag the server with many friends and then play in slowmo for better accuracy :troll: 21:38 < bridge_> I'd say it depends on the server triplet in general 21:39 < bridge_> os-arch-abi 21:39 < bridge_> Though ideally OS shouldn't(tm) be a factor 21:39 < bridge_> Actually maybe compiler-arch pair is enough though I can't say for sure we didn't test it 21:40 < bridge_> u sure the api syscalls if they are syscalls dont change a bit 21:40 < bridge_> scheduler or idk 21:40 < bridge_> xd 21:40 < bridge_> @zwelf2 has been reproducing teehistorian files, maybe he knows 21:40 < bridge_> interrupts 21:40 < Huu> I have a quick question, lets say i start Nyan Cat pressing D + Jump i will get a run with 48 but as i soon i hold D + Jump and then press kill i will start with a 0,02 ms delay what is the reason for that 21:40 < bridge_> even memory allocation if is slower 21:40 < bridge_> may slow stuff? 21:40 < bridge_> u never know 21:41 < bridge_> Shouldn't matter(tm) since we don't ever skip ticks 21:41 < bridge_> when you kill you start in the air 21:41 < bridge_> If that's mattera 21:41 < bridge_> If that's matters 21:43 < Huu> Gonna record something 1 sec 21:45 < bridge_> thats just gd 21:46 < bridge_> tele aint random anymore? 21:47 < bridge_> it's basically random 21:47 < bridge_> Pseudo random 21:47 < bridge_> no way for the client to predict without #6775 21:47 < bridge_> https://github.com/ddnet/ddnet/pull/6775 21:47 < bridge_> nya nya nya nya nya nya 21:51 < bridge_> It only appears random now. For the server it's deterministic 21:52 < bridge_> *sad bot noises* 21:52 < bridge_> *sad botter noises* 21:56 < bridge_> ``` 21:56 < bridge_> Total download size: 2906.11 MiB 21:56 < bridge_> Total size of installed packages: 14624.70 MiB 21:56 < bridge_> Update size difference: 1810.77 MiB 21:56 < bridge_> ``` 21:56 < bridge_> 21:56 < bridge_> uff 21:57 < bridge_> starting arch after 2 months xd 21:57 < bridge_> and ofc they give me like 1MB/s xd 21:57 < bridge_> xD 21:57 < bridge_> Why Update anyway if you don't want to install smth new? 21:57 < Huu> Here its visible https://imgur.com/MwsGdej 21:58 < bridge_> yea ik but itd be funny to troll blockers or gores players 21:58 < Huu> Inputs are D + Jump on first run its aligned with ghost and i would get 48 run then i press kill while still holding D + Jump and i will get a 50 run 21:58 < bridge_> or hh players 21:59 < bridge_> why do you assume u clicked the exact same time as ghost 21:59 < bridge_> (jump) 21:59 < bridge_> You will spawn in the air 1 pixel above ground 21:59 < bridge_> jumping changes your speed 22:01 < Huu> Intresting so thats why i lose also my dj intresting 22:01 < bridge_> wow whoever in mozillas team made this first time setup screen is an ass xD 22:01 < bridge_> 22:01 < bridge_> animations with delays 22:01 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143636055405367316/image.png 22:01 < bridge_> pure hate 22:01 < bridge_> :justatest: :justatest: 22:01 < bridge_> I also don't like it 22:02 < bridge_> just had to setup a new windows machine. still hate every step of it. 22:02 < bridge_> yes, i hate setting up windows xD 22:02 < bridge_> 2 million privacy questions 22:03 < bridge_> u click no to all 22:03 < bridge_> 22:03 < bridge_> it still steals anything xd 22:03 < bridge_> waiting 3hrs for basic drivers for the hardware to even work correctly 22:03 < bridge_> waiting 3hrs for basic drivers to install for the hardware to even work correctly 22:04 < bridge_> at least there are some debloating scripts on gh 22:05 < bridge_> It's to protect children obviously 22:07 < bridge_> With nlite u get unattended installation and better performance by removing everything you will never use 22:07 < bridge_> I was today years old when I discovered that there is a windows software to flash the windows iso to a USB stick. The twist is that the USB stick has to have a single ntfs partition for whatever reason 22:07 < bridge_> Else the software doesn't detect it 22:08 < bridge_> bad idea 22:08 < bridge_> Neither does it give any warnings/errory 22:08 < bridge_> Rufus? 22:08 < bridge_> windows to go. i used it at school 22:08 < bridge_> Removing cortana etc 22:09 < bridge_> yea I usually use rufus or dd 22:09 < bridge_> but I had no raw iso and the torrents were slow. 22:10 < bridge_> https://github.com/AveYo/MediaCreationTool.bat 22:10 < bridge_> Those debloater scripts tend to remove important components 22:10 < bridge_> what is the command under pacman to switch packages? 22:10 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143638409714356294/image.png 22:11 < bridge_> [jupeyy@archlinux ~]$ sudo pacman -S mesa 22:11 < bridge_> Idk 22:11 < bridge_> i did this 22:11 < bridge_> It should prompt with a number 22:11 < bridge_> idk im gentoo based rn 22:11 < bridge_> xdd 22:11 < bridge_> wdym? do you have to finish watching the animation in order to continue? 22:11 < bridge_> yes 22:11 < bridge_> we should add that to ddnet 22:11 < bridge_> there is like 2 seconds delay per step 22:11 < bridge_> ok 22:11 < bridge_> wanna break from the ads? if you tap now to watch a short video, you will recieve 30 minutes of ad-free multeasymap, yes, really 22:11 < bridge_> :troll: 22:12 < bridge_> add a ddnet tik tok tab to the client 22:12 < bridge_> -1000 fps 22:12 < bridge_> making a ddnet intro anim that plays while loadinh 22:12 < bridge_> i have mesa-git installed 22:12 < bridge_> making a ddnet intro anim that plays while loading 22:12 < bridge_> but want to downgrade to mesa 22:12 < bridge_> like discord 22:12 < bridge_> If you’re having dependency issues U can do -Rdd mesa-git and -S mesa 22:12 < bridge_> iirc 22:12 < bridge_> i probably need a force command 22:12 < bridge_> trure 22:12 < bridge_> and we gotta add dumbass loading messages too 22:13 < bridge_> Did you know?: The bind for deepfly is- 22:13 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1143639088872824924/IMG_20230822_215850.jpg 22:15 < bridge_> pacman -R --nodeps --nodeps mesa-git 22:15 < bridge_> worked xd 22:16 < bridge_> guess it's the same as this? 😹 22:16 < bridge_> isn’t that -Rdd 22:16 < bridge_> Yeah 22:16 < bridge_> Lol 22:16 < bridge_> GG 22:21 < bridge_> use downgrade commands 22:21 < bridge_> I think it's in the aur 22:22 < bridge_> use downgrade command 22:23 < bridge_> someone should make a video comparing major versions of ddnet 22:23 < Huu> I reproduced a flat nyan cat with straight ground with just the start keep the same, if i jump on that platform i will get a time difference while without jumping i will get a faster time. Can someone explain me why 22:25 < Huu> Also when i dont press Jump my X Speed is constant at 18.59 as soon i jump my X speed jumps between 18.59 to 17.9 (i think) and as soon i land again back to 18.59 22:26 < bridge_> Hello guys just wanna let you know have got some Top quality products Gas🍁 ⛽️ , shrooms🍄, coke, meth &Vapes🥢, Ice capz, Edibles🍪, Chocolate Bars🍫, LSD, Syrups, Skittles, pills , 💊 clone cards etc. Both for smokers are patients, Offer at moderate price, and also some polkadot gummies, edibles,road trip available just wanna know if you're interested in grabbing some? 22:26 < bridge_> https://t.me/lambiaserickLtdstore 22:27 < bridge_> because ramping counts y speed too 22:28 < bridge_> (im pretty sure idk correct me if im wrong lol) 22:36 < Huu> So would mean to improve time on nyan cat to maybe get a 46 or 44 ms i would jump less stay mostly constant in Y i will test that out thanks alot everyone 22:44 < bridge_> test it - `37.230.162.198:8303` [currently offline] 22:44 < bridge_> 37.230.162.198:8303 is not a DDNet or KoG server. 22:45 < bridge_> i found that bug a few months ago too, if i remember right it is possible just by setting the "local" variable in the playerinfo snap to 0 :D 22:45 < bridge_> i found that bug a few months ago too, if i remember right it is possible just by setting the "local" variable in the playerinfo snap to 0 :) 22:48 < bridge_> this one is server side thingie i suppose 22:48 < bridge_> server can do anything with player camera anyway 22:48 < bridge_> yes, the snapshot that the server sends to the client 22:48 < bridge_> oh really? 22:48 < bridge_> was it always like that? 22:49 < bridge_> @matodor разве? 22:49 < bridge_> can't) 22:49 < bridge_> duh 22:49 < bridge_> i just joined the server from matodor, he does some magic i think 22:49 < bridge_> i did some tests back in 2019, was a thing 22:49 < bridge_> but i dont think it actually sets you in spectator mode 22:49 < bridge_> maybe i fucked up something xd 22:50 < bridge_> damn, thats interesting 22:50 < bridge_> would be a cool thing for new mods 22:50 < bridge_> matodor says thats impossible tho 22:50 < bridge_> maybe i was high or smth :P 22:51 < bridge_> anyway i was about to make session competetive mod with classes like bf tee mod 22:51 < bridge_> but drones xd 22:53 < bridge_> :p 23:01 < bridge_> Yes it is possible, I'm using in in infclass mod: https://github.com/infclass/teeworlds-infclassR/blob/production/src/game/server/infclass/infcplayer.cpp#L149-L191 23:01 < bridge_> 1. Snap `CNetObj_DDNetPlayer` with `pDDNetPlayer->m_Flags |= EXPLAYERFLAG_SPEC;` 23:01 < bridge_> 2. Snap `CNetObj_SpectatorInfo` 23:01 < bridge_> Done. 23:01 < bridge_> Yes it is possible, I'm using it in infclass mod: https://github.com/infclass/teeworlds-infclassR/blob/production/src/game/server/infclass/infcplayer.cpp#L149-L191 23:01 < bridge_> 1. Snap `CNetObj_DDNetPlayer` with `pDDNetPlayer->m_Flags |= EXPLAYERFLAG_SPEC;` 23:01 < bridge_> 2. Snap `CNetObj_SpectatorInfo` 23:01 < bridge_> Done. 23:01 < bridge_> Yes it is possible, I'm using it in infclass mod: https://github.com/infclass/teeworlds-infclassR/blob/production/src/game/server/infclass/infcplayer.cpp#L149-L191 23:01 < bridge_> 1. Snap `CNetObj_DDNetPlayer` with `pDDNetPlayer->m_Flags |= EXPLAYERFLAG_SPEC;` 23:01 < bridge_> 2. Snap `CNetObj_SpectatorInfo` 23:01 < bridge_> Done. 23:03 < bridge_> o: 23:03 < bridge_> thanks for sharing that, will play around it sometime 23:03 < bridge_> thanks for sharing that, will play around with it sometime