01:49 < bridge_> can you create a topic on teeworlds.com too? 02:45 < bridge_> are you talking about dynamic libs (compiled alongside DDNet) or shared libs (on the system)? 02:46 < bridge_> this should typically be handled as an install step. whatever environment you’re using with mingw should be smart enough to load the dependencies without any fuss 02:47 < bridge_> for Windows deployment this is tricky because there’s no expectation that the libs will exist on target systems. generally when you “deploy” you should also copy the relevant DLLs over. Copying stuff like that should not be the job of the build system 02:48 < bridge_> I use this script https://gist.github.com/ewancg/de0b47f55b9894b9dc5a7b8609150186 02:48 < bridge_> do you really want make making unneeded copies of all of the system DLLs? 02:50 < bridge_> this should typically be handled as an install step. whatever environment you’re using with mingw should be smart enough to load the dependencies without any fuss as you run it 02:51 < bridge_> for Windows deployment this is tricky because there’s no expectation that the libs will exist on target systems. generally when you “deploy” you should also copy the relevant DLLs over. Copying stuff like that should not be the job of the build system (at compile time) 02:52 < bridge_> @blaiszephyr i forgot if i pinged you with the initial reply... forgive the dupe if so 02:52 < bridge_> as far as the latter is concerned, it should typically be handled as an install step. whatever environment you’re using with mingw should be smart enough to load the dependencies without any fuss as you run it 02:53 < bridge_> this is also a licensing can of worms. you're essentially creating a build with all the contents of a static build, which has its own considerations (some libraries you are restricted from linking statically by license) 03:23 < ws-client> That CI fail is suspicious. https://github.com/ddnet/ddnet/actions/runs/8825297278/job/24229302734?pr=8236 03:23 < ws-client> the pr added an unused method that is never called. And the CI failed in the admin integration test. Are the integration tests getting unstable? 05:37 < bridge_> Melon 05:39 < bridge_> Solely talking about dynamic libs (sqlite, steamapi, avformat, sdl2) - i'll Check If i can modify my cmakelists to do that for me. Thanks ❤️ 05:39 < bridge_> Yes? 05:40 < bridge_> I Saw u typing :) 05:40 < bridge_> :angy: 05:40 < bridge_> I Just woke Up 05:40 < bridge_> :owo: 05:40 < bridge_> Its 5:40 AM 05:40 < bridge_> Help 05:41 < bridge_> Yeeeh me 2, sometimes i awake at midnight hahaha 05:43 < bridge_> Maybe with a Post_build Copy ? Hmm Coffee and 10 hours Work before in can Check :monkaStop: :monkaStop: 06:02 < bridge_> same deal 06:02 < bridge_> environment should handle it and an install/deploy step should be responsible for copying it into the build dir 06:03 < bridge_> if you're using MSYS2 you should be able to start it through you env's terminal because it has /*/lib and whatnot added to the library search paths 06:03 < bridge_> and then use a separate deploy step 06:03 < bridge_> those DLLs are not built alongside ddnet 10:20 < bridge_> I'll Just add a Script to move the necessary Files to my build folder, Just annoying i have to Launch a Script every time i Clean rebuild. I wanna get rid of IDE's, they take to much work off of me and i feel like i know nothing about the actual process when you have a program doing it for you 10:23 < bridge_> im not single anymore :justatest: 10:24 < bridge_> yo since when does the ddnet server kick you when the unused bits of the playerflag inputs are non-zero? xD 10:26 < bridge_> lmao 10:39 < bridge_> https://www.phoronix.com/news/NVK-Implicit-Pipeline-Caching 10:48 < bridge_> @learath2 btw my coworker, aided by our findings found the issue 10:48 < bridge_> the issue was that some structs were misaligned, we use the Layout from rust to know the sizes and aligns of the data we create, but made a wrong assumption, we needed to also pad to align at the end of structs 10:49 < bridge_> tldr, we thought 28 was the correct size for example but llvm assumed 32 10:51 < bridge_> So was it that you were allocating the wrong stack size as I hypothesized? 10:52 < bridge_> yep 10:52 < bridge_> u hit the nail 10:58 < bridge_> :Celebrate: 11:25 < bridge_> LETSGOOOOOOOO 11:28 < bridge_> WAIT WHAT? 11:28 < bridge_> 11:28 < bridge_> that's why our epyc guy i never active on discord lately 11:28 < bridge_> 😏 11:29 < bridge_> xD 11:34 < bridge_> Dont Show your discord Chat History or she s gonna want to learn Rust and never leave her room again 11:36 < ws-client> @pisstine do you still play ctf? I made a new release of ddnet 0.7 that is bug free as far as I know so I would appreciate any bug reports :) 11:36 < ws-client> https://chillerdragon.github.io/ddnet7-client/ 11:37 < bridge_> plot twist. in fact it's Voxel acting like a girl to get ryo away from rust. 11:37 < bridge_> 11:37 < bridge_> His secret master plan all along 11:40 < bridge_> He would've started with me fr 11:53 < bridge_> no its an ai girlfriend 12:05 < ws-client> @fokkonaut i dmed u btw 12:07 < bridge_> chiller 12:07 < bridge_> wtf happened in ur 2nd screenshot xd 12:07 < ws-client> wot 12:08 < bridge_> looks like photoshop 12:08 < ws-client> a 12:08 < bridge_> Yes I still play vanilla thanks to your client, and this version seems to fix the issues that I had, the only minor oddities I notice are that 1.) On vanilla 0.7 servers (dm and ctf at least) no sound is made when I hook another tee now (the sound plays on 0.6 servers in the client) and 2.) The value for the alpha channel on the skin markings goes to 0 every time I reset client on any skin that doesn't have a json file setup (idk if that's true i 12:08 < ws-client> its gimp 12:08 < ws-client> i wanted to show the green checkmark and console :D 12:09 < ws-client> @pisstine you have those bugs with the new version from today? 12:09 < ws-client> those should be fixed now 12:10 < bridge_> I may have accidentally downloaded the wrong 18.1 in that case, I'll try again 12:10 < ws-client> 18.1 should not have those bugs :( 12:15 < ws-client> ah shit i can reproduce nice find thanks @pisstine 12:15 < bridge_> Yeah I am having those issues still in 18.1 12:15 < bridge_> Oh okay nice np ty for client! 12:21 < bridge_> I saw your commit history crumbling. I knew something was up!! 12:52 < bridge_> @chillerdragon i cant get it running on my linux system. `./DDNet: error while loading shared libraries: libavformat.so.58: cannot open shared object file: No such file or directory 12:52 < bridge_> ` 12:53 < bridge_> @chillerdragon i cant get it running on my linux system. `./DDNet: error while loading shared libraries: libavformat.so.58: cannot open shared object file: No such file or directory 12:53 < bridge_> ` any fix for this? 12:53 < bridge_> self compile 12:59 < bridge_> I don't have the source code unfortunately. I compiled the normal ddnet yesterday and everything works perfectly 13:04 < bridge_> You are missing a library, the specific version chiller linked to. You'll need to get that 13:04 < bridge_> It's ffmpeg btw 13:10 < bridge_> I downloaded that from github actions. Was hoping that is portable. 13:12 < bridge_> Maybe I have to do the Linux one manually then. What’s the way to go here? Link statically? What be cmake flag is that? 13:12 < bridge_> Maybe I have to do the Linux one manually then. What’s the way to go here? Link statically? What cmake flag is that? 13:12 < bridge_> u could also disable video rendering 13:13 < bridge_> The source code is at ChillerDragon/ddnet on GitHub in the pr\_07\_client branch 13:14 < bridge_> https://github.com/ChillerDragon/ddnet/tree/pr_07_client 13:14 < bridge_> I am curious on how to publish a release in general 13:15 < bridge_> ddnet-scripts contains all stuff related to publishing 13:15 < bridge_> Yes I know -.- 13:15 < bridge_> but if u want stuff like auto updater.. that is more annoying 13:15 < bridge_> Idk if people even use the Linux binary release 13:16 < bridge_> Is it even good style to statically link libraries in a binary release? 13:16 < bridge_> I guess the intended way is everything is managed by the package manager 13:19 < bridge_> why not xd 13:20 < bridge_> > the pr added an unused method that is never called. And the CI failed in the admin integration test. Are the integration tests getting unstable? 13:20 < bridge_> See #8261, the same test also failed with just a translation change. I assume it's related to the recently changed integration test script/map in #8231. 13:20 < bridge_> https://github.com/ddnet/ddnet/issues/8261 13:20 < bridge_> https://github.com/ddnet/ddnet/pull/8231 13:20 < bridge_> there are static link fans and dyn link fans 13:21 < bridge_> @chillerdragon as far as i can see the ddnet release from website static links ffmpeg too 13:21 < bridge_> ``` 13:21 < bridge_> Dynamic Section: 13:21 < bridge_> NEEDED libfreetype.so.6 13:21 < bridge_> NEEDED libSDL2-2.0.so.0 13:21 < bridge_> NEEDED libdl.so.2 13:21 < bridge_> NEEDED libvulkan.so.1 13:21 < bridge_> NEEDED libGL.so.1 13:21 < bridge_> NEEDED libnotify.so.4 13:21 < bridge_> NEEDED libgdk_pixbuf-2.0.so.0 13:27 < bridge_> Rust biased axaxaxax 13:30 < bridge_> Do you know how to link ffmpeg statically? Or do I have to look it up .\_. 13:33 < bridge_> i dont know by heart 13:33 < bridge_> maybe we have a cmake option for it 13:52 < ws-client> I have a bunch of stale discussions in issues and prs and I am slowly getting impatient :/ 13:53 < ws-client> I am waiting for a yes or no + why in here since last year https://github.com/ddnet/ddnet/issues/7480 13:54 < ws-client> This pr is ready since 2 weeks https://github.com/ddnet/ddnet/pull/8151 13:55 < ws-client> this is somewhat blocking a lot of my projects and nobody responded to my bump in february https://github.com/ddnet/ddnet/issues/7777 14:07 < bridge_> tja 16:00 < bridge_> I don't reply to them because you won't like my answer and you already spent so much time on all of them 16:01 < bridge_> Actually I can reply to one 16:11 < bridge_> i'd love to have 7480 - espacially since blockworlds™ wants to expand ( and having a server redirect on multeasymap might help for once :kek:) 16:11 < bridge_> For the uppercamelcase thing, I absolutely hate that we are adding another long running task to our already long asf CI run 16:11 < bridge_> if its all about readability, refactor everything `CAndEveryClassWouldLookLikeThis` xD 16:13 < bridge_> oh god you have bf now 16:15 < bridge_> redirecting to arbitrary servers is not nice tho 16:19 < bridge_> (at least) 16:20 < bridge_> btw there is another problem. 16:20 < bridge_> 16:20 < bridge_> it might violate GDPR 16:20 < bridge_> Context is what's important, they don't want you tricking the browser into fetching protected resources from another website that can then be extracted using js 16:20 < bridge_> if u redirect to USA server 16:21 < bridge_> A redirect just takes you to a new website, the redirector no longer can extract anything anymore 16:21 < bridge_> as long as redirecting is within the master server itself i dont think you'd violate GDPR 16:21 < bridge_> but the target can 16:21 < bridge_> why so? 16:21 < bridge_> 16:21 < bridge_> if u redirect to google, google can collect your IP without your permission 16:21 < bridge_> if the server is in USA 16:21 < bridge_> that kinda violates GDPR afaik 16:22 < bridge_> considering the user didnt press a button that he wants it ofc 16:22 < bridge_> That's why I said we need to be careful not to send any secrets in the issue. Browsers are very careful to only send secrets belonging to the origin, so after `ddnet.tw` redirects to `malicio.us`, the target can't read cookies for `ddnet.tw` 16:23 < bridge_> just give it a "you will be redirected to: localhost - `127.0.0.1`" 16:23 < bridge_> I think the gdpr issue is the hosters issue not ours. If I redirect you to `healthline.org` you'll need to accept their policies before reading it 16:24 < bridge_> .com not .org 😄 16:24 < bridge_> and yes i know this can be spoofed to a certain extend, but just have prequisite checks before even allowing redirection ( both connected to the master server, (need to have the "isCommunity")) ? 16:24 < bridge_> yes, but i think if u as server actively redirect that's not ok either 16:24 < bridge_> also tw client has no context that needs to be shared between two servers 16:24 < bridge_> at least that's how i understand it 16:24 < bridge_> 16:24 < bridge_> bcs u also cannot host google fonts that are hosted on google CDN 16:24 < bridge_> in USA 16:25 < bridge_> anyway, i am no lawyer. i just know that all this shit is super complicated and i hate it 16:25 < bridge_> I'm unsure whether it's an issue or not. But I'm quite sure that it's not our concern 16:25 < bridge_> if you only care about the redirect itself you could add some message box with "do you agree to redirect and connect to x.x.x.x:xxxx" with yes/no and checkbox with "i trust server x.x.x.x:xxxx" 16:26 < bridge_> The user of the redirect should be the one that makes sure that the redirect they do is lawful 16:26 < bridge_> xddd 16:26 < bridge_> ah yeah that is the case anyway 16:26 < bridge_> but it's maybe not intuitive for everyone 16:26 < bridge_> e.g. ChillerDragon Enterprises, should make sure that the server he is redirecting to respects and processes the data in a gdpr compliant way if he is serving users in europe 16:26 < bridge_> gym bro detected, bf projected 16:27 < bridge_> you are no longer relatable btw 16:27 < bridge_> i also kinda dislike that browsers don't make stuff easier GDPR related. they make it to easy to fuck up for devs 😄 16:27 < bridge_> rip 16:27 < bridge_> yeah @ryozuki how dare you leave us behind - rude 16:27 < bridge_> dude works out for a few years and think he's better than us! 16:27 < bridge_> I'll also confiscate your programmer socks 16:27 < bridge_> says the married man with a kid 16:27 < bridge_> :justatest: 16:28 < bridge_> i dont think ddnet as a community will violate against GDPR by solely giving them the option to redirect user's to another server 16:28 < bridge_> i dont think ddnet as a community will violate against GDPR by solely giving server hoststhe option to redirect user's to another server 16:28 < bridge_> i dont think ddnet as a community will violate against GDPR by solely giving server hosts the option to redirect user's to another server 16:29 < bridge_> I wonder if the redirector is even liable for gdpr violations done by the redirectee. That sounds wrong to me 16:29 < bridge_> i dont think ddnet as a community will violate against GDPR by solely giving server hosts the option to redirect user's to another server - espacially if we add checks that validate the server's identity beforehand 16:29 < bridge_> imagine server ring: server has commands to redirect to next or previous server in the ring 16:30 < bridge_> so you can hide a whole community under 1 public ip 16:30 < bridge_> It wouldn't be the redirector mistreating your data, they aren't forwarding any data, they are forwarding you, and you re provide the data. Though it is involuntary, so idk. We should hire a lawyer at 800 euros an hour to resolve this for us 16:31 < bridge_> well that's sadly not the point. an IP is already personal data. the target server can collect it, and it's the redirectors fault. 16:31 < bridge_> 16:31 < bridge_> That's also what i think EU didn't think well enough about. E.g. they could obfuscate the IP or whatever. But it's super hard to not violate GDPR XD 16:31 < bridge_> esp. bcs USA laws are always company friendly 16:31 < bridge_> send 800 euros to oy :justatest: 16:31 < bridge_> :gigachad: maybe he'll come back and merge some PR's finally 16:31 < bridge_> yeah on clearnet 16:33 < bridge_> while it is true that an IP-Adress is personal data, teeworlds as a whole violates GDPR for years then no?, giving modified server's the "possibility!!" to collect and save said data - so a simple redirect shouldnt hurt XD 16:33 < bridge_> while it is true that an IP-Adress is personal data, teeworlds as a whole violates GDPR for years then no?, giving modified server's the "possibility" to collect and save said data - so a simple redirect shouldnt hurt XD 16:33 < bridge_> and a user currently has to actively connect to a server 16:33 < bridge_> that's like having a consent 16:34 < bridge_> and the master server is in EU 16:34 < bridge_> well then force him to either accept or decline said redirect 16:34 < bridge_> but currently u can only redirect to a server with same IP 16:40 < bridge_> https://cdn.discordapp.com/attachments/293493549758939136/1233064766084354068/image0.gif?ex=662bbc8d&is=662a6b0d&hm=7ed7e2d2ad047e62a09f21712ff907c3a49bc896881970a6da194961d59c6123& 16:51 < bridge_> 18:02 < bridge_> One more time in case I was not clear. You should not need these DLLs in dev builds 18:03 < bridge_> If you are shipping releases then feel free to use the script i sent 18:11 < bridge_> Only if your system provides all the DLLs though, for example `avutil-58.dll` is not available on my system path 18:12 < bridge_> Not if you use the environment (MSYS2) to launch it 18:13 < bridge_> Given you have installed the ffmpeg package for your environment 18:14 < bridge_> interesting, I didn't know it's supposed to work like that. looks like I don't have steam API in there though `DDNet.exe: error while loading shared libraries: steam_api.dll: cannot open shared object file: No such file or directory` 18:16 < bridge_> hmm, the error message appears to be wrong, it can't load `avutil-58.dll` because I renamed it to test, it's unrelated to `steam_api.dll` 18:17 < bridge_> https://github.com/rust-lang/rust/pull/104087 18:17 < bridge_> @jupeyy_keks inline const 18:18 < bridge_> epyc 18:22 < bridge_> its a bit like consteval 21:58 < bridge_> congratulations 22:17 < bridge_> is there a good Google Calendar alternative? 22:18 < bridge_> like if Jitsi had a calendar or something like that 22:50 < bridge_> Nextcloud, maybe?