00:00 < bridge> but why 00:01 < bridge> dependency, compile time, random ci failure, executable size 00:01 < bridge> with no benefit apart from testing itself sometimes more thorougly than the cpp side is 00:01 < bridge> im not sure if all rust tests even have a counterpart 00:07 < bridge> idk if we should be eliminating the potential for rust additions tho 00:08 < bridge> i guess heinrich thought adding the bridge would be enough to get people contributing with rust? i was not around when it was added but i do know it adds very little burden if you have a competent dev env 00:11 < bridge> i guess its for quic 00:12 < bridge> but since its not happening any time soon rn its just bloat 00:13 < bridge> or we can actually try to put it to use 00:15 < bridge> i dont see how rn 00:15 < bridge> we can do pretty much everything in c++ 00:18 < bridge> isn't a time machine just a 4D car? 00:20 < bridge> most time machines are 3 dimensional 00:20 < bridge> the space they move in is 4d 00:20 < bridge> i like the i punched a hole through spacetime to where i wanted to go then drove through it depiction of time machines 00:21 < bridge> theres also the wibbly wobbly timey wimey dimension way 00:21 < bridge> or the magic teleportation 00:21 < bridge> but a 3d object would never be able to propel itself in to the 4th dimention 00:21 < bridge> but a 3d object would never be able to propel itself in to the 4th dimension 00:22 < bridge> just as it is never possible for a 2D object to exist that propels itself in the 3rd dimention 00:22 < bridge> just as it is never possible for a 2D object to exist that propels itself in the 3rd dimension 00:35 < bridge> a 2d object can exist on any slice of a 3d plane 00:35 < bridge> and move across it 00:36 < bridge> i guess the part that rotates you in the 4th dimension has to be 4 dimensional 00:36 < bridge> there could be some weird interaction which allows something 3 dimensional to do timey wimey stuff without being 4th dimensional 00:37 < bridge> like how if you pour water imbetween 2 slides and hold it upright, and let go of the slides, the "2d" water flattens even though it was fully 2d 00:37 < bridge> bad example, the water is 3d and so are all of its particles 00:38 < bridge> if you can imagine a pure 2D object that moves across the 3D space by itself with no other influences then idk. shouldn't be possible i think 00:39 < bridge> an object that you can see only from front/back? 00:39 < bridge> and invisible from sides? 00:39 < bridge> 2D is not an ambiguous term... xd 00:43 < bridge> well it has to be for the time travel to work 00:43 < bridge> thats what makes it fiction 00:43 < bridge> you only need a little wobbly physics to make time travel possible 00:43 < bridge> how the consequences work out is upto whoever divided by 0 00:45 < bridge> How about you 2 travel back in time to your last shower 3 weeks ago? 00:48 < bridge> go back to kog /: 00:48 < bridge> back in time 2 ⁉️ 00:50 < bridge> I cant, perma muted 00:53 < bridge> im going to make sure you get perma muted here too if you use dev channel for ragebait xD 00:53 < bridge> deserved 00:57 < bridge> And what will you gain from that? You try to give this message a soft appearance with a “xD”. But in reality it is just an angry threat. Please look within and consider why you have to direct anger at others like this. 00:59 < bridge> wtf ai? or do you have some autocomplete for " to “ 01:00 < bridge> ?what 01:01 < bridge> `“` is not a character you can type on a normal keyboard, either you are on your phone and your autocorrect corrects `"` to `“` or you can type `“` directly on your keyboard or phone 01:02 < bridge> also c/cpp libraries for QUIC exist, theres ones from msft and facebook that look perfectly fine. 01:02 < bridge> `“` is not a character you can type on a normal keyboard, either you are on your phone and your autocorrect corrects `"` to `“` or you can type `“` directly on your keyboard or phone. 01:02 < bridge> or you made the effort to find out the unicode for this which i don't believe xD 01:03 < bridge> Define a normal keyboard? It takes a next level xenophobe to be prejudice towards different types of keyboards. But to anwser your question, i am using a phone 01:05 < bridge> @totar wanna see my sim 01:05 < bridge> 5 minutes 01:05 < bridge> k 01:09 < bridge> i would spend a chunk of time trying to do it myself once the rust ver is completed 01:09 < bridge> what's the point of doing it again in cpp if the rust version was complete? 01:10 < bridge> you want the rust version as a guide? 01:11 < bridge> yah, the pr looks scary 01:11 < bridge> hmm 01:11 < bridge> i dislike rust because im old 01:12 < bridge> I know this to be false 01:12 < bridge> my dislike of rust does not exist? 01:12 < bridge> or my dislike of rust isnt because im old? 01:12 < bridge> you are not old xd 01:12 < bridge> oh 01:12 < bridge> old-fashioned 01:12 < bridge> ok sure 01:12 < bridge> c enjoyer 01:12 < bridge> im old tho 01:13 < bridge> there are people who look adulty to me and boom theyre 12 like what 01:13 < bridge> thats a child 01:13 < bridge> when did that happen 01:13 < bridge> is the QUIC PR really that scary? Leatro keeps saying that it's easy 01:13 < bridge> is the QUIC PR really that scary? Learto keeps saying that it's easy 01:13 < bridge> consider i dont know rust 01:13 < bridge> and the borrow checker has some bones to pick with me from last time i tried 01:13 < bridge> ok but if you did it in cpp then you don't need rust 01:13 < bridge> yeah which is why i want to do redo it in cpp 01:13 < bridge> 12 stones 1 bird 01:14 < bridge> yeah, I think you should just go for it 01:14 < bridge> 1. no rust woo 01:14 < bridge> 2. i learn rust 01:14 < bridge> 3. i learn whats going on in quic 01:14 < bridge> I don't think you need to wait for the rust ver to be done 01:14 < bridge> ah i see 01:14 < bridge> but look at the checklist?? 01:14 < bridge> why is it just not changing transport protocol 01:15 < bridge> idk lemme glance at it agai 01:16 < bridge> @chairn :( 01:16 < bridge> missclick 01:16 < bridge> I some of those checklist things can be skipped in the initial version? 01:17 < bridge> like idk why we need masterserver registering to go over QUIC 01:17 < bridge> also idk which checklist item you referred to with that message 01:18 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1408228842111369307/image.png?ex=68a8facf&is=68a7a94f&hm=d7d5e59ae874437a1f7a7d2766c27db5d91fb70248a6042998217a12ac8ec4b2& 01:18 < bridge> ah, I think because the way the network system is designed really wants conless UDP, so you need to redesign it 01:19 < bridge> why cant quic be conless? 01:19 < bridge> or pretend to be 01:19 < bridge> that defeats the point 01:19 < bridge> well 01:19 < bridge> maybe you can hide it idk 01:19 < bridge> i thought the point was encryption/compression 01:19 < bridge> we dont have issues with the connection handshake 01:19 < bridge> @learath2 can you explain the idea behind the QUIC impl, you mentioned it being invisible to the game or something how does that work 01:20 < bridge> please spill details 01:20 < bridge> it will just happen and then youll just do the normal (unnecesary) handshake 01:20 < bridge> there's a lot of features in QUIC 01:20 < bridge> those are the ones i care about 01:20 < bridge> also it's definitley not compression 01:20 < bridge> ):< 01:20 < bridge> I dont think QUIC does that for you 01:21 < bridge> might be wrong idk 01:21 < bridge> Well the contents of the packets are completely unchanged, it’s just transported over quic. That’s probably what I was talking about 01:22 < bridge> but network.cpp and similar files still need complete refactor? Is that correct? It's not invisible to them 01:22 < bridge> it looks like theres some registration code 01:23 < bridge> did heinrich keep the UDP functionality working in his PR? He seems to have deleted the entire cpp files for that so I assume he reimplemented those in rust 01:23 < bridge> #6961 is pretty much done if you want to take a look 01:23 < bridge> https://github.com/ddnet/ddnet/pull/6961 01:23 < bridge> I'm reading it 01:23 < bridge> it's large 01:23 < bridge> its extremley scary 01:24 < bridge> nothing is that scary 01:24 < bridge> tw06.rs is where the old style packets are handled 01:24 < bridge> `"protocol must be one of ddnet-18+quic, tw-0.5+udp, tw-0.6+udp or tw-0.7+udp".fmt(f)` 0.5? 01:25 < bridge> can i still connect to 0.5 servers? 01:25 < bridge> yeah I think 01:25 < bridge> maybe? 01:25 < bridge> no mastersrv support tho 01:25 < bridge> download link broken 01:25 < bridge> for 0.5 trunk 01:25 < bridge> 0.5.2 works 01:27 < bridge> idk the network layer is really not *THAT* big in ddnet, I don't think this is an insane thing to take over in cpp. but you'd have to write tests for everything 01:27 < bridge> !remind 5 tater come vc 01:27 < chillerbot> Helo teero777 I will remind you at Fri Aug 22 2025 01:32:53 GMT+0200 (Central European Summer Time) about your matter again. 01:28 < bridge> You want to do quic in C++ instead? 01:28 < bridge> why not? 01:28 < bridge> is there a reason it's harder 01:28 < bridge> doesnt seem to work 01:29 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1408231478252408832/image.png?ex=68a8fd44&is=68a7abc4&hm=772f88591c08761bae9ea433372435dbd3bb05875f980c8fdaa855ab9f6121a8& 01:29 < bridge> i wonder if i can force it 01:29 < bridge> Not really, but I thought you wouldn’t be interested in quic, but maybe just something like dtls instead 01:29 < bridge> Would be easier to bolt on, with minimal changes 01:30 < bridge> I've never heard of that ngl 01:30 < bridge> oh wait 01:30 < bridge> nvm I have 01:30 < bridge> I thought we want QUIC benefits 01:30 < bridge> I didn't think a worse solution would be accepted 01:31 < bridge> I think the quality of the quic libraries in C++ wasn’t all that great. Cloudflares rust implementation does have bindings though iirc 01:31 < bridge> the microsft C library looks pretty good to me idk 01:31 < bridge> https://github.com/microsoft/msquic 01:31 < bridge> just stuck on game info 01:31 < bridge> just stuck on getting game info 01:32 < bridge> Probably wouldn’t be, it would be really nice to get all the quic benefits 01:32 < chillerbot> DO NOT FORGET @teero777 TO: tater come vc 01:33 < bridge> i like that 01:33 < bridge> But maybe if you push it along with an accounts implementation you could get it pushed through 01:33 < bridge> I thought accounts were like integrated into QUIC protocol 01:33 < bridge> They kinda are, in their current form 01:34 < bridge> or I was told something like that 01:34 < bridge> ok 01:34 < bridge> But nothing stops you from completely doing that part differently too. Not like the accounts got deployed anywhere yet 01:34 < bridge> I guess jupsti uses them in ddnet-rs and I doubt he’d switch over, so that would be another nice civil war in 6 years 01:35 < bridge> idk I don't really want to change the account server very much 01:35 < bridge> b-but rust... 😭 01:36 < bridge> also like, why would we integrate accounts into a networking protocol. That doesn't make any sense 01:36 < bridge> You could try to tie the connection to the client certificate like quic does, basically bolting it on top of dtls instead 01:37 < bridge> is that a normal thing to do with quic 01:37 < bridge> do they intend you to use it like that 01:38 < bridge> The accounts are basically just certificates, so currently they are just used as the connections client certificate 01:39 < bridge> Not my area really, you’d have to ask jupstar and heinrich, who are probably tired of people probing around it questioning their design 01:39 < bridge> well they should have finished it after they designed it .-. 01:40 < bridge> As far as I understand what got implemented is pretty much jupstars proposal in the accounts issue 01:41 < bridge> wait but this sounds like you can get the certificates and verify them without QUIC then? 01:41 < bridge> Yes, but I think if the connections shared secret is not derived from the client certificate something something security issue 01:42 < bridge> idk I haven't thought about it 01:43 < bridge> I’m sure it’s all in there somewhere #3411 01:43 < bridge> https://github.com/ddnet/ddnet/issues/3411 01:43 < bridge> I think it's documented in the account repo 01:43 < bridge> https://github.com/ddnet/ddnet-accounts/issues/1 01:43 < bridge> This thing has been in progress for so long that everything is just melted together in my brain now. I don’t remember when what happened where 01:45 < bridge> idk the certificate crypto thing is nice but I feel like it's not what we should have started with 01:45 < bridge> I will not be giving any more opinions on it because it always devolves into some argument 01:46 < bridge> I will lol. I wish it worked like this. I made this a while ago 01:46 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1408235956003143680/image.png?ex=68a9016f&is=68a7afef&hm=1cde35a70541d7404bc1965c04a53b6e46eb252d15f2b6a8d181878d6aba9e52& 01:47 < bridge> I plead the 5th 01:48 < bridge> Something something MITM 01:48 < bridge> but https 01:48 < bridge> no he's right 01:48 < bridge> Not that red arrow 01:48 < bridge> o 01:48 < bridge> but like 01:48 < bridge> we already accept this risk 01:48 < bridge> how 01:49 < bridge> yoink the otc then what 01:49 < bridge> someone can MITM any rcon login 01:49 < bridge> There are ways to do it, you can authenticate servers by certificates and CAs. I designed and implemented that a while back 01:49 < bridge> pretend to be the client, but then there are 2 conns fighting over the thing 01:49 < bridge> A while as in 2017 or something 01:49 < bridge> rcon was insecure anyway 01:50 < bridge> pretend to be the client, but then there are 2 conns fighting over the thing, either can say "im being mitmed please stop" 01:50 < bridge> the red arrow needs some way to encrypt 01:50 < bridge> also importantly, we can upgrade to the certificate fancyness or QUIC or DTLS from this design pretty easily 01:51 < bridge> I do not remember the details, but my proposal for keeping the link plaintext, or just encrypting parts of it wasn’t really popular 01:52 < bridge> remmeber them plez 01:52 < bridge> Something something security issue. Idk been a decade since someone first tried to add accounts 01:52 < bridge> kog is using unentryped tokens linked to a bind for multiple years and nothing bad has happened 01:53 < bridge> If only y’all hadn’t annoyed heinrich off of this channel he would definitely remember 01:53 < bridge> also like, we're not a bank. What is the worst thing someone can do if they compromise your account? That should be so vanishinly rare that it could be handled case by case 01:54 < bridge> we don't even have virtual items 01:55 < bridge> ``` 01:55 < bridge> [68a7b1be][server]: strange message cid=0 msg=0 data_size=55 01:55 < bridge> [68a7b1be][server]: 01 C5 07 09 1A 57 55 30 7F AC 7D 3C 1F AB 3A 06 DF 31 30 2E 34 2E 30 20 62 75 69 6C74 20 6F 6E 01:55 < bridge> ``` 01:55 < bridge> im getting close 01:55 < bridge> x-x 01:56 < bridge> thats the getting ready client info 01:57 < bridge> whats this for? 01:57 < bridge> trying to get ddnet to connect to 0.5 01:57 < bridge> o nice 01:57 < bridge> I assume it doesn't work then 01:57 < bridge> when 0.6down? 01:57 < bridge> why is it sixup 01:57 < bridge> and now seven 01:58 < bridge> and not seven 01:58 < bridge> It’s sleep o’clock. Idk what to do with quic and accounts, maybe jupstar might have an idea how you can hook up his accounts without quic to start with 01:58 < bridge> It’s sleep o’clock. Idk what to do with quic and accounts, maybe jupstar might have an idea on how you can hook up his accounts without quic to start with 01:58 < bridge> idk I could make the entire account server and the client/server implementation for the diagram I posted in like 10% as much time as it would take me to implement QUIC. So it's hard to be motivated to implement QUIC 01:59 < bridge> I think it literally takes like 2 days 02:01 < bridge> I do not think your proposal makes it past code review, or design review really. “we are not a bank” probably won’t fly with many 02:02 < bridge> what if the red line was not red for some reason 02:03 < bridge> Then it has more of a chance making it past the council 02:05 < bridge> But honestly, you could just likely piggyback off of jupstars work. The certificates could be authenticated over any link really, I think it’s just much more convenient (and easy to properly secure) if they are used as client certificates (in quic or tls) 02:07 < bridge> Even more honestly, I think you should spend your time on other things. We couldn’t manage to get an account system agreed on and merged in a decade, it’s quite unlikely to change soon. The closest we got was this attempt, and if it ever gets finished it’ll probably get merged. A completely new proposal will probably just dig up all the old conversations again, for days and days 02:14 < bridge> I mean I didn't really expect my proposal to be accepted any way, since it goes against the work that's already done and I know how this go. But I don't really understand or have interest in the current account system, even if it was finished it doesn't outline how it should be integrated into the website or moderation system in a useful way, and I have no idea who would do that. It's like so much less work to finish a much simpler system 02:14 < bridge> I mean I didn't really expect my proposal to be accepted any way, since it goes against the work that's already done and I know how this goes. But I don't really understand or have interest in the current account system, even if it was finished it doesn't outline how it should be integrated into the website or moderation system in a useful way, and I have no idea who would do that. It's like so much less work to finish a much simpler system 02:15 < bridge> I mean I didn't really expect my proposal to be accepted any way, since it goes against the work that's already done and I know how this goes. But I don't really understand or have interest in the current account system, even if it was finished it doesn't outline how it should be integrated into the website or moderation system in a useful way, and I have no idea who would do that. It's so much less work to finish a much simpler system 02:15 < bridge> the server wont send me snapshots idk why D: 02:19 < bridge> (maybe the onready is a different id in 0.5) 02:20 < bridge> https://github.com/teeworlds/teeworlds/pull/3279 02:20 < bridge> hein is 0.7 maintainer?? 02:22 < bridge> yes, for a while 02:22 < bridge> y'all underestimating how much learath and heinrich have contributed, not only to DDNet, but to teeworlds as a whole 👀 02:22 < bridge> although he doesn't seem to merge stuff so idk 02:22 < bridge> "is it okay to merge this while oy is afk" 02:22 < bridge> 0.7 is as directionless as ddnet 02:22 < bridge> Me neither. Idk what happened along the way. I quite liked the initial proposal, but somehow I just don’t vibe with it anymore. Seeing deen look at the demo and not understand why he had 3 different identifiers may have had something to do with it 02:23 < bridge> step 1: login 02:23 < bridge> step 2: done 02:23 < bridge> accounts r ez 02:23 < bridge> [cat_uwu](https://cdn.discordapp.com/emojis/1230626195649990696.webp?size=48&name=cat_uwu) 02:24 < bridge> sorry if I'm keeping you up past bed, but I'm lacking context. What demo? 3 identifiers? Were these chats in public channel? 02:24 < bridge> goodnight 02:24 < bridge> muscle memory 02:24 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1408245441421643897/image.png?ex=68a90a45&is=68a7b8c5&hm=4c297c4cc5abfd35bc21e69ff43c128a5e0391c952764b5bfa7bbaa2ea6e4a69& 02:24 < bridge> lol 02:25 < bridge> Jupstar demo’d the accounts along with ddnet-rs at the meetup. You basically get a unique id, you get a global username, you get a per network “semi-global” username, you get your display name 02:26 < bridge> I count 4 identifiers 02:27 < bridge> I count 4 identifiers /j 02:27 < bridge> yeah idk, I don't even remember reading about those in the account information description 02:28 < bridge> Honestly, I’m not really the best person sum this up as I don’t really understand it all that well myself, and I was convinced all of the identifiers were useful and necessary for this system to operate with all the design goals. But I have many doubts about UX if deen looked at it like it’s from an alien planet 02:28 < bridge> I mean to be fair, discord also has the same 4 identifiers kinda 02:28 < bridge> snowflake, account name, display name, server nickname 02:28 < bridge> so it's not insane 02:29 < bridge> There is also a possibility that I misunderstood something. But if I am misunderstanding it, joe-layman the professional Linear blocker will have a muuch harder time understanding 02:30 < bridge> It’s not insane, but it might be too complicated. But anyway, I have made peace with it, if this is the account system we can get merged, I’m ready to compromise a bit on the UX 02:31 < bridge> I mean this is opinion, some people prefer having competetive username space for simplicity. I would probably do it how old discord did, where you have 02:31 < bridge> unique id, username+discriminator and that's it 02:32 < bridge> My very first proposal in 2016 (I think), was unique usernames but FFA display names like it is right now 02:33 < bridge> idk that's fine too 02:33 < bridge> Or like steam does it 02:33 < bridge> popular players will probably whine about not getting their prefered unique username tho 02:33 < bridge> and maybe validly 02:33 < bridge> do we really want some random person who isn't Aoe to have the Aoe username 02:34 < bridge> Exactly why we just kinda ended up with a design that explicitly does not address any of these controversial issues 02:34 < bridge> but having 200 Aoe's is maybe not better idk 02:34 < bridge> It says nothing about display names, nothing about points, nothing about moderation 02:37 < bridge> Anyway, I don’t completely understand it and I wouldn’t want to misrepresent other peoples work, so you’ll have to have a chat with the jupman. I really need to sleep, my eyes are half closed 02:39 < bridge> gn 02:54 < bridge> No denying in that :deen_star: 02:55 < bridge> But even big CEOs step down from their roles to keep the company growing 05:41 < bridge> require 3k pts or more to get a name xd 05:41 < bridge> 😓 05:41 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1408295087028310077/image.png?ex=68a93881&is=68a7e701&hm=9bc55fa36a91069c12c0880eb47fdf155c7d517f199f22963f01fb6d81acc24c& 05:45 < bridge> except the points aren't tied to an account current :justatest: 05:45 < bridge> except the points aren't tied to an account currently :justatest: 05:49 < bridge> to get a name not an account 05:49 < bridge> so everyone will play as "account id 1094057561" for the first 3k points? 05:50 < bridge> they need a name somewhere 05:51 < bridge> display name != account name 05:51 < bridge> hmm fair maybe 05:51 < bridge> it could be ~your chosen name 05:51 < bridge> so we dont have to deal with 2 names at once 05:51 < bridge> ~ == guest 05:51 < bridge> `~` == guest 05:52 < bridge> `~` == nameless 05:52 < bridge> `~` == unofficial name 05:52 < bridge> poorly designed map 05:52 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1408297723374862558/image.png?ex=68a93af6&is=68a7e976&hm=35d22a325571d54d76d77d6e0c37d04570825d551b4fd4c45d57b4e7f91eccd0& 05:52 < bridge> if u fail once on left theres no way to get them back 05:52 < bridge> is it legit 05:52 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1408297774960738365/IMG_9963.jpg?ex=68a93b02&is=68a7e982&hm=5d3a4aca908517679912268c745ee9e43c826b34df5e9e6d8a48c5d06511e2bb& 05:52 < bridge> and guessts can be some other symbol 05:53 < bridge> client cooporation we can turn them into symbols 05:53 < bridge> or have them as "auth levels" -1 = unnamed/not using account name -2 = no account 05:53 < bridge> I think a better system would something where you upgrade from a discriminator based name to an exclusive unique name that no one else can then use 05:53 < bridge> 05:53 < bridge> so like multiple people are allowed play as ``bun bun#3241`` or ``bun bun#9956`` but once you reach 3k points you can claim the username ``bun bun`` without the discriminator and no one else can use that username regardless. 05:53 < bridge> I think a better system where you upgrade from a discriminator based name to an exclusive unique name that no one else can then use 05:53 < bridge> 05:53 < bridge> so like multiple people are allowed play as ``bun bun#3241`` or ``bun bun#9956`` but once you reach 3k points you can claim the username ``bun bun`` without the discriminator and no one else can use that username regardless. 05:54 < bridge> people will race to get names but i guess there can always be disputes 05:54 < bridge> well the racing was gonna happen with your method too 05:54 < bridge> yeah 05:54 < bridge> i dont like discriminators 05:54 < bridge> I do! :D 05:54 < bridge> unique names ftw 05:55 < bridge> ur opinion is valid 05:55 < bridge> i want to be able to use a non account name 05:55 < bridge> like for my dummy, my dummy cant also be bun bun 05:55 < bridge> you can still have display names 05:55 < bridge> making everyone [D] would be nice 05:55 < bridge> well i guess i would prefer server telling you which one is dummy 05:56 < bridge> i just want a way to tell whos a dummy of who 05:56 < bridge> unreltaed matter 06:39 < bridge> It is defs bugged. Sometimes severs simply don't respond to a connect request.. No idea why, smth about the dos protection ig. On rus servers it mostly works 06:52 < bridge> Heinrich can only technically merge. But Oy does not want him to actually do it .\_. 08:00 < bridge> gumo ^^ 09:03 < bridge> is there some json or metadata which I can check in order to know which client versions are downloadable? 09:03 < bridge> I am sure crawling ddnet.org/downloads is not the correct way 😄 09:14 < ws-client> **** ``git tag`` 09:14 < ws-client> **** ? xd 09:15 < bridge> As years pass I'm starting to understand more and more why Oy ran things like he used to 09:16 < ws-client> **** ono lerato is doing the oy 09:16 < bridge> Just a nice and simple "wontfix". Wanna argue, feel free to just shout into the void 09:16 < ws-client> **** based 09:16 < ws-client> **** aura maintainer 09:18 < ws-client> **** @learath2 yea i havent found my inner peace to handle that properly yet. I think having some maintainer with strong opinions who can be stubborn is good. And I can do that to some degree. But then also I dont want to make people sad that want to help, have problems or put time into something you dont like :C 09:19 < ws-client> **** i guess communicating up front that any contributions might be rejected without the room for discussion even if multiple users want it could help 09:19 < bridge> Recently I've come to realize the nicer you are the more people just take advantage. The more willing to discuss things you are the more they will just drag things oon and on 09:21 < ws-client> **** I think there is value in discussion even if you finally as a maintainer decide to ignore other peoples opinion xd 09:21 < ws-client> **** i have a little chat with 90% of the cheaters on my fng server before i use rcon 09:22 < ws-client> **** listen to what the others have to say and then based on that decide what to do 09:22 < ws-client> **** even if you decide to fully ignore it 09:23 < ws-client> **** but yes it can take time, but then you just have to shutdown the conversation as soon as it starts to spin circles 09:23 < ws-client> **** @learath2 also pls dont stop to be nice :/ just dont be noob and dont get exploited 09:27 < bridge> You know what we need on the team? A villain, someone okay with just being hated 09:36 < bridge> natural life cycle of open source software 09:53 < bridge> We have a Heinrich xd 09:54 < bridge> But I think he might be too soft 09:55 < bridge> @learath2: where can I help? I am a villain for hire. 09:55 < bridge> Heinrich is an angel, we need an actual demon. "Wontfix", don't like it, cry me a river, making too much noise, banned 09:55 < bridge> I see 09:56 < bridge> I feel like I will be kicked by Heinrich if I do that .\_. 09:56 < bridge> You seem to have the same problem that we do, you said you didn't want to make people that worked on stuff sad 09:56 < bridge> I can try 09:57 < bridge> If you need me to 09:57 < bridge> Lerato I’ll do anything for you uwu 09:58 < bridge> <--- 09:59 < bridge> :kek: 10:08 < bridge> what I have have both, the steam directory and the usual teeworlds data directory in appdata, does the client find scripts in both? 10:13 < bridge> Both steam and website client use the same userdir, inside appdata 10:20 < bridge> I think for my purpose rn i can ignore the steam data directories for this 10:23 < bridge> Any C++ std lib enjoyers? How does C++ infer type to be `std::chrono::time_point` from `std::chrono::time_point`? 10:24 < bridge> Any C++ std lib enjoyers? How does C++ infer type `std::chrono::time_point` from `std::chrono::time_point`? 10:26 < bridge> It's the associated duration type for system clock on your implementation 10:26 < bridge> `std::system_clock::duration` 10:26 < bridge> Ye, I saw that there's such thing, but I don't see how it's used in `std::chrono::time_point` 10:27 < bridge> time_point is a template class, the second parameter has a default, it uses the clocks duration 10:27 < bridge> ```cpp 10:27 < bridge> template 10:27 < bridge> struct time_point; 10:27 < bridge> ``` 10:27 < bridge> nvm 10:28 < bridge> I was looking here -.- 10:28 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1408367097817337927/image.png?ex=68a97b92&is=68a82a12&hm=119387bc67e7fba33e8d5dca9d08e534aa95ef34f9bb18dd6f867cae9e9755ec& 10:32 < bridge> Oic, idk what it's called but stl implementations do do that 10:38 < ws-client> **** milkeey doing C++ wot 10:40 < bridge> I can print hello world in C++ 10:40 < ws-client> **** nice 10:41 < ws-client> **** random social media ah moment https://github.com/ddnet/ddnet/pull/10711 4 different comments have a single reaction from a different person xd 10:41 < ws-client> **** github is the new discord, trust me 10:42 < bridge> 🥛 🐱 🚀 10:53 < bridge> ddnet needs a wm4 10:53 < bridge> <0xdeen> Short term it hurts them, but long term it is good for DDNet and for contributors if they know who they have to talk to about new features and how they should be designed. Less wasted effort, better contributions 10:54 < ws-client> **** woah a deen 10:55 < ws-client> **** @0xdeen how do they know "who" to talk to? 10:56 < ws-client> **** i dont have a matrix available right now to see which msg you replied to. But its proably either the dictator or villain one. Either way if you suggest such strategy for ddnet i think the "who" is not obvious 10:57 < ws-client> **** who makes decisions in this soup of maintainers? 10:58 < bridge> He replied to the one where I said you also don't seem to want to be willing to make people sad 10:58 < ws-client> **** ty 10:58 < ws-client> **** still curious about the "who" part tho 10:59 < bridge> Well the soup of maintainers. Not anyone specific 10:59 < ws-client> **** i see 10:59 < ws-client> **** where a maintainer 10:59 < ws-client> **** i need a maintainer 10:59 < ws-client> **** xd 10:59 < ws-client> **** doesnt sound as funny 11:00 < ws-client> **** @learath2 wat we do if maintainers disagree? 11:01 < bridge> It's not often that that happens, but when it does I guess you need to convince at least 2 11:01 < ws-client> **** wat to do about the epic crypto battle #10634 vs #10636 I do not feel comfortable looking into that ._. 11:01 < chillerbot> https://github.com/ddnet/ddnet/issues/10634 11:01 < chillerbot> https://github.com/ddnet/ddnet/issues/10636 11:01 < bridge> I want 10634. I'll ping heinrich on that one, see if he has any reason to outright reject it 11:01 < ws-client> **