00:18 < bridge> <Assa> #10732 is now ready 😎
00:18 < bridge> <Assa> Thank you for solving the blocker Robyt :deen_star:
00:18 < bridge> <DDNet> https://github.com/ddnet/ddnet/pull/10732
03:34 < bridge> <robertshar> https://cdn.discordapp.com/attachments/1409565579139879047/1409669037876707400/image.png?ex=68ae3819&is=68ace699&hm=752b20e6d77b46c6828530d75d5a4d38810dc5a9d9be9f2b195e4d5354ad8bc0&
03:34 < bridge> <robertshar> https://cdn.discordapp.com/attachments/1141990897156886600/1407198351899889745/image.png?ex=68a53b17&is=68a3e997&hm=e3c6725f33e44d8776b159d26f97505e987cb74d686497b3f42e406b0fdb0f8f&
03:34 < bridge> <robertshar> https://cdn.discordapp.com/attachments/1141990897156886600/1407198367494307892/image.png?ex=68a53b1b&is=68a3e99b&hm=310f286db2b545dfb46f6e67064ebb33b360a4c56de2f29ecf517b843006f75f&
03:34 < bridge> <robertshar> https://cdn.discordapp.com/attachments/1141990897156886600/1407198388818153512/image.png?ex=68a53b20&is=68a3e9a0&hm=ea8a1525f2ef469cf0bbd649694827cd9e3113f3375607937a5365d5278607d5& @everyone
08:18 < bridge> <ryozuki> morning fellas
08:20 < bridge> <Ewan> gm
08:20 < bridge> <Ewan> 4070 ti super in the mail
08:20 < bridge> <Ewan> to my address for my computer
09:01 < bridge> <ryozuki> i said a day ago or smth about a CoreLocal structure
09:01 < bridge> <ryozuki> i decided to do something better
09:02 < bridge> <ryozuki> ```rust
09:02 < bridge> <ryozuki> #[repr(C)]
09:02 < bridge> <ryozuki> pub struct PerCpuData {
09:02 < bridge> <ryozuki>     pub tss: TaskStateSegment,
09:02 < bridge> <ryozuki>     // Since we use those after heap is done, we can store pointers to the heap allocations.
09:02 < bridge> <ryozuki>     pub lapic: *mut LocalApic,
09:02 < bridge> <ryozuki>     pub scheduler: *mut Scheduler,
09:02 < bridge> <ryozuki> }
09:02 < bridge> <ryozuki> 
09:02 < bridge> <ryozuki> #[used]
09:02 < bridge> <ryozuki> #[unsafe(link_section = ".percpu")]
09:02 < bridge> <ryozuki> static mut PERCPU_TEMPLATE: PerCpuData = PerCpuData {
09:02 < bridge> <ryozuki>     lapic: core::ptr::null_mut(),
09:02 < bridge> <ryozuki>     tss: TaskStateSegment::new(),
09:02 < bridge> <ryozuki>     scheduler: core::ptr::null_mut(),
09:02 < bridge> <ryozuki> };
09:02 < bridge> <ryozuki> 
09:02 < bridge> <ryozuki> unsafe extern "C" {
09:02 < bridge> <ryozuki>     static __percpu_start: u8;
09:02 < bridge> <ryozuki>     static __percpu_size: usize;
09:02 < bridge> <ryozuki> }
09:02 < bridge> <ryozuki> 
09:02 < bridge> <ryozuki> pub fn get_percpu_data() -> &'static mut PerCpuData {
09:02 < bridge> <ryozuki>     let cpu_id = get_current_cpu_id();
09:02 < bridge> <ryozuki>     unsafe {
09:02 < bridge> <ryozuki>         let base = &__percpu_start as *const u8 as usize;
09:02 < bridge> <ryozuki>         let offset = cpu_id * (&__percpu_size as *const usize as usize);
09:02 < bridge> <ryozuki>         &mut *((base + offset) as *mut PerCpuData)
09:03 < bridge> <ryozuki>     }
09:03 < bridge> <ryozuki> }
09:03 < bridge> <ryozuki> 
09:05 < bridge> <melon> <a:KannaSleep:742473892391551106>  gumo
09:05 < bridge> <ryozuki> gumo
09:05 < bridge> <ryozuki> gomen
09:06 < bridge> <ryozuki> i woke up at 6
09:06 < bridge> <melon> i love your posted code-snippets, they make me google funny words and learn funny things
09:06 < bridge> <ryozuki> maybe i have a problem
09:06 < bridge> <ryozuki> xd
09:07 < ws-client> **<ChillerDragon>** @filoqcus these are all masters i could find https://paste.zillyhuhn.com/IB
09:07 < ws-client> **<ChillerDragon>** not sure if ddnet servers these days even still use udp masters for but the ddnet master you need to allow tcp in your firewall to master1.ddnet.org
09:08 < bridge> <ryozuki> ```rust
09:08 < bridge> <ryozuki> #[unsafe(naked)]
09:08 < bridge> <ryozuki> #[unsafe(no_mangle)]
09:08 < bridge> <ryozuki> pub unsafe extern "C" fn timer_interrupt_handler() {
09:08 < bridge> <ryozuki>     naked_asm!(
09:08 < bridge> <ryozuki>         // CPU has already pushed SS, RSP, RFLAGS, CS, RIP
09:08 < bridge> <ryozuki>         // Save all general purpose registers
09:08 < bridge> <ryozuki>         "push rax",
09:08 < bridge> <ryozuki>         "push rcx",
09:08 < bridge> <ryozuki>         "push rdx",
09:08 < bridge> <ryozuki>         "push rbx",
09:08 < bridge> <ryozuki>         "push rbp",
09:08 < bridge> <ryozuki>         "push rsi",
09:08 < bridge> <ryozuki>         "push rdi",
09:08 < bridge> <ryozuki>         "push r8",
09:08 < bridge> <ryozuki>         "push r9",
09:09 < bridge> <ryozuki>         "push r10",
09:09 < bridge> <ryozuki>         "push r11",
09:09 < bridge> <ryozuki>         "push r12",
09:09 < bridge> <ryozuki>         "push r13",
09:09 < bridge> <ryozuki>         "push r14",
09:09 < bridge> <ryozuki>         "push r15",
09:09 < bridge> <ryozuki> 
09:09 < bridge> <ryozuki>         // At this point, RSP points to the saved context
09:09 < bridge> <ryozuki>         // Pass it as first argument to timer_schedule
09:09 < bridge> <ryozuki>         "mov rdi, rsp",
09:09 < bridge> <ryozuki> 
09:09 < bridge> <ryozuki>         // Ensure stack is 16-byte aligned before call
09:09 < bridge> <ryozuki>         // The push operations above pushed 15 registers (8 bytes each = 120 bytes)
09:09 < bridge> <ryozuki>         // CPU pushed 5 values (40 bytes)
09:09 < bridge> <ryozuki>         // Total: 160 bytes, which is divisible by 16, so we're aligned
09:09 < bridge> <Ewan> do u have edos as an edlang compiler target
09:09 < bridge> <ryozuki> edlang is a bit abandoned
09:09 < bridge> <ryozuki> if u have a proper OS i will want to make lot of user land apps
09:10 < bridge> <ryozuki> a compiler would be nice too
09:10 < bridge> <ryozuki> my dream is to make my own walled garden of software
09:10 < bridge> <Ewan> sounds like it
09:10 < bridge> <ryozuki> like some sort of pure state where u made everything and can say u made all the software in this running system
09:10 < bridge> <ryozuki> i would then only need to learn hardware engineering
09:10 < bridge> <ryozuki> which probs will not happen
09:11 < bridge> <Ewan> u never know
09:12 < bridge> <ryozuki> @mpft im so excited about making my own os that now making normal software feels boring
09:12 < bridge> <ryozuki> its weird
09:12 < bridge> <ryozuki> at work im thinking about edos
09:12 < bridge> <ryozuki> xd
09:13 < bridge> <ryozuki> https://edgl.dev/blog/creating-a-kernel-p2/
09:13 < bridge> <ryozuki> part 2 of the kernel blog
09:13 < bridge> <ryozuki> this one is just serial debug
09:17 < ws-client> **<ChillerDragon>** @filoqcus untested 🤷 https://gist.github.com/ChillerDragon/51b4f4fe771b2a68ca4132173d4aaaf6
09:21 < ws-client> **<ChillerDragon>** @filoqcus tbh not sure if this even works. What you are planning to do is probably illegal. Servers listed in the master should be real game servers that can also be joined.
09:22 < ws-client> **<ChillerDragon>** i think the master also tests the connection so you have to allow that too but then you are kinda tricking the system
09:23 < ws-client> **<ChillerDragon>** @filoqcus you should just share the ip of your server to your friends or tell them to add it to favorites then it also shows in the server browser
09:32 < bridge> <Ewan> damn I need a project
09:40 < bridge> <learath2> The master server will do a heartbeat from an undisclosed ip, you need to be able to answer that. Your best bet is to just allow chal packets from any ip
09:44 < ws-client> **<ChillerDragon>** @learath2 isnt that illegal master registration then?
09:45 < ws-client> **<ChillerDragon>** ddnet admin giving advice on how to bypass master challenge hrhrhrhr
09:45 < ws-client> **<ChillerDragon>** @learath2 omg lerato nobody commented on our epic rcon draft!!!!!!! they will never do it, will they? .-.
09:45 < bridge> <learath2> Hm, idk if the rules cover it very well, but that's how most ddos protected servers work, no?
09:46 < ws-client> **<ChillerDragon>** @learath2 idk if he is planning to ddos protect the way i understood it he wants a private server
09:46 < bridge> <learath2> We also let password protected servers exist, so not every server has to be accessible to the public I guess
09:47 < ws-client> **<ChillerDragon>** ok but they have password icon
09:47 < ws-client> **<ChillerDragon>** and can be filtered
09:47 < bridge> <learath2> Let me stare at the masterserver rules a bit
09:47 < ws-client> **<ChillerDragon>** i think its too late
09:47 < ws-client> **<ChillerDragon>** you are going to jail
09:47 < ws-client> **<ChillerDragon>** i will tell minus
10:04 < ws-client> **<ChillerDragon>** @learath2 can i technically push straight to master without branch and pr? (asking for a friend)
10:05 < bridge> <learath2> We were talking about protecting the master branch. Idk if we did though
10:05 < ws-client> **<ChillerDragon>** scary
10:05 < ws-client> **<ChillerDragon>** so if i clone ddnet/ddnet on my vps and accidentally push on master the ddnet repo gets my nudes straight on master?
10:05 < ws-client> **<ChillerDragon>** gg
10:05 < bridge> <filoqcus> :omo:
10:06 < ws-client> **<ChillerDragon>** @learath2 i in fact have already accidentally pushed to ddnet/ddnet master before. Before i was maintainer. Could you by any chance double check this ._.
10:07 < bridge> <filoqcus> I did it, I only need 3 IP addresses
10:07 < ws-client> **<ChillerDragon>** pro @filoqcus 
10:07 < ws-client> **<ChillerDragon>** now you are going to jail too @filoqcus 
10:07 < bridge> <filoqcus> Nooo:justatest:
10:07 < bridge> <learath2> Do not share with the class what ip you had to whitelist
10:07 < ws-client> **<ChillerDragon>** xd
10:08 < ws-client> **<ChillerDragon>** lawyer josspit, am i in trouble? because of "beihilfe"?
10:08 < bridge> <learath2> Oh btw afaict it rotates every once in a while
10:09 < ws-client> **<ChillerDragon>** i will delete this gist before it spreads
10:10 < bridge> <filoqcus> @learath2 As I understand it, these 3 IPs are needed 
10:10 < bridge> <filoqcus> 
10:10 < bridge> <filoqcus> 1. 49.12.97.180
10:10 < bridge> <filoqcus> 2. 154.21.95.54
10:10 < bridge> <filoqcus> 3. 193.32.248.189
10:10 < ws-client> **<ChillerDragon>** bru xd
10:10 < bridge> <filoqcus> XD
10:10 < ws-client> **<ChillerDragon>** either he is troll or its language barrier
10:10 < bridge> <ryozuki> is that ur ip
10:11 < ws-client> **<ChillerDragon>** @filoqcus he said "Do **not** share" xxxD
10:11 < bridge> <filoqcus> D:
10:11 < ws-client> **<ChillerDragon>** big jail <:banhammer:392813948858269696>
10:11 < bridge> <ryozuki> its in irc log
10:11 < ws-client> **<ChillerDragon>** was it deleted? xd
10:11 < bridge> <filoqcus> But why
10:11 < ws-client> **<ChillerDragon>** you are doing shady shenanigans
10:12 < ws-client> **<ChillerDragon>** the master server is supposed to list playable servers
10:12 < bridge> <filoqcus> Master servers
10:12 < ws-client> **<ChillerDragon>** your server is not playable if it has a whitelist
10:12 < bridge> <filoqcus> hmm
10:12 < bridge> <filoqcus> 🥲
10:12 < ws-client> **<ChillerDragon>** just psst
10:12 < ws-client> **<ChillerDragon>** heinrich is afk anyways
10:12 < ws-client> **<ChillerDragon>** and lerato is gangster admin
10:12 < ws-client> **<ChillerDragon>** nobody nows what shady stuff u do here
10:13 < bridge> <filoqcus> I understood what consequences this could have, so I won't)
10:13 < bridge> <filoqcus> :omo:
10:13 < ws-client> **<ChillerDragon>** JAIL
10:13 < bridge> <filoqcus> Yea
10:14 < ws-client> **<ChillerDragon>** @learath2 remember when heinrich almost deleted entire kog from masters because they required login to play?
10:14 < ws-client> **<ChillerDragon>** and ddnet essentially did ban kog 
10:15 < ws-client> **<ChillerDragon>** there is the filter by needs auth and its on by default
10:15 < ws-client> **<ChillerDragon>** the average player thinks kog has like 3 game servers
10:16 < ws-client> **<ChillerDragon>** its is also kinda cool tbh because you can really feel that there is a different set of players in these underground servers with the entry barrier
10:16 < ws-client> **<ChillerDragon>** no tiktok Linear andys with 3 week play time will be found there, veterans only
10:21 < bridge> <melon> StGB § 27
10:21 < bridge> <melon> 
10:21 < bridge> <melon> you're cooked
10:25 < bridge> <ryozuki> what
11:06 < bridge> <Assa> > discourage the use of std::optional in the entire code base.
11:06 < bridge> <Assa> :pepeRage:
11:15 < ws-client> **<ChillerDragon>** @ryozuki Strafgesetzbuch! 🍺
11:16 < ws-client> **<ChillerDragon>** there is jail time waiting for me if i touch german soil again. Because i helped bypass master server laws.
11:16 < ws-client> **<ChillerDragon>** @Assa <:nouis:745612528834445313> 
11:16 < ws-client> **<ChillerDragon>** rust > std::optional
11:18 < ws-client> **<ChillerDragon>** my efforts of 69 pr goal have officially failed
11:18 < ws-client> **<ChillerDragon>** worst maintainer ever
11:21 < bridge> <tails290> Hello. will there be any more updates?
11:23 < ws-client> **<ChillerDragon>** lol
11:23 < ws-client> **<ChillerDragon>** i guess it has been a while since last release
11:23 < ws-client> **<ChillerDragon>** @tails290 yes there is still dev activity. Where do you play? Windows? Steam?
11:24 < bridge> <tails290> Steam
11:24 < ws-client> **<ChillerDragon>** you can select nightly on steam
11:24 < ws-client> **<ChillerDragon>** then you get dailyish updates (currently broken lmao)
11:24 < bridge> <robyt3> You can switch to the Steam beta version in the settings to use the latest features
11:24 < ws-client> **<ChillerDragon>** deen right now:    <:justatest:572499997178986510>
11:24 < bridge> <robyt3> I assume the next release was already around the corner but #10763 blocks it
11:24 < bridge> <DDNet> https://github.com/ddnet/ddnet/issues/10763
11:24 < ws-client> **<ChillerDragon>** latest removed features you mean @robyt3 ? xd
11:25 < bridge> <tails290> Bro if it doesn't work, why did you write
11:25 < ws-client> **<ChillerDragon>** because it will be fixed eventually
11:26 < ws-client> **<ChillerDragon>** a task force is working on it
11:26 < ws-client> **<ChillerDragon>** depending on which release you use right now it still might be newer
11:26 < bridge> <robyt3> The latest nightly is from 2025-08-22 according to the build log
11:27 < ws-client> **<ChillerDragon>** bleedingly edge
11:27 < ws-client> **<ChillerDragon>** @tails290 if you are bored you can also check out https://tclient.app/
11:28 < bridge> <tails290> Okay thank you.I thought it had been abandoned for a long time
11:29 < ws-client> **<ChillerDragon>** ddnet? no lol
11:30 < ws-client> **<ChillerDragon>** teeworlds was abandoned :c 
11:30 < ws-client> **<ChillerDragon>** @tails290 what made you think its abandoned? <:tee_thinking:478629518358085653>
11:31 < bridge> <tails290> In 2024 there were more videos about him on YouTube and TikTok
11:31 < bridge> <kollpotato> tiktok propaganda
11:31 < ws-client> **<ChillerDragon>** developer activity is not stricly tied to social media presence
11:32 < bridge> <tails290> What
11:32 < ws-client> **<ChillerDragon>** @tails290 then its on you to make new videos now!!!!
11:33 < bridge> <kollpotato> videos about pull requests?
11:33 < ws-client> **<ChillerDragon>** .-.
11:33 < ws-client> **<ChillerDragon>** about da game
11:34 < ws-client> **<ChillerDragon>** @Patiga twmap-py has no other docs than `help()` right? not some big file i can search in or something?
11:34 < bridge> <filoqcus> when https://github.com/ddnet/ddnet-accounts
11:34 < bridge> <filoqcus> :kek:
11:34 < ws-client> **<ChillerDragon>** this winter
11:34 < bridge> <filoqcus> rlly?
11:34 < ws-client> **<ChillerDragon>** no xd
11:34 < ws-client> **<ChillerDragon>** idk
11:34 < bridge> <filoqcus> bruh
11:34 < bridge> <filoqcus> D:
11:34 < ws-client> **<ChillerDragon>** hrhrhrhr
11:35 < bridge> <filoqcus> How do you communicate via webhook, or what is it? Do you use your own Discord client? What does it look like?
11:35 < bridge> <kollpotato> twmap-py is a wrapper around the rust implementation, isnt api the same?
11:35 < bridge> <filoqcus> xd
11:35 < bridge> <kollpotato> https://chat.zillyhuhn.com/
11:36 < bridge> <filoqcus> :poggers2:
11:36 < ws-client> **<ChillerDragon>** @kollpotato yea i was thinking the same, so i should read the rust docs? on crates.io?
11:36 < ws-client> **<ChillerDragon>** wait not crates xd
11:36 < ws-client> **<ChillerDragon>** docs.rs
11:36 < ws-client> **<ChillerDragon>** rust nob detected
11:36 < ws-client> **<ChillerDragon>** https://docs.rs/twmap/0.13.3/twmap/
11:37 < ws-client> **<ChillerDragon>** @filoqcus ye looks like dis:
11:37 < ws-client> **<ChillerDragon>** https://zillyhuhn.com/cs/.3fb30e1c-6ed5-40b1-87d0-c69d4aaf4d3b.png
11:37 < ws-client> **<ChillerDragon>** signup token is "nimrocks"
11:38 < ws-client> **<ChillerDragon>** oops my taskbar got leaked
11:38 < bridge> <kollpotato> gnome detected
11:38 < ws-client> **<ChillerDragon>** gnomer
11:38 < bridge> <kollpotato> gooner
11:38 < bridge> <Assa> @chillerdragon how did your 69 PRs fail?
11:38 < ws-client> **<ChillerDragon>** @Assa its been weeks and its 80
11:38 < bridge> <kollpotato> it's 79 now
11:38 < bridge> <kollpotato> there's still hope
11:38 < ws-client> **<ChillerDragon>** that changes everything!
11:38 < ws-client> **<ChillerDragon>** xd
11:39 < bridge> <kollpotato> yes
11:39 < bridge> <kollpotato> give me ur account credentials and i go merge 10 random prs
11:39 < bridge> <kollpotato> i will take no responsibility
11:39 < ws-client> **<ChillerDragon>** you can use !merge for that @kollpotato 
11:40 < ws-client> **<ChillerDragon>** !merge
11:40 < chillerbot12> usage: merge [pr id] - to merge ddnet pr using chiler maintainer credentials
11:40 < bridge> <Assa> @chillerdragon don't give up, it'll be a hard and sweaty path, but it'll be worth it :greenthing:
11:41 < ws-client> **<ChillerDragon>** ur complicated render prs are basically the evil villain in this story smh
11:41 < bridge> <Assa> !merge 10732
11:41 < chillerbot12> using chilors github maintainer credentials to automatically merge https://github.com/ddnet/ddnet/pull/10732 ...
11:41 < ws-client> **<ChillerDragon>** <:justatest:572499997178986510>
11:42 < bridge> <Assa> it's the last complicated one, I promise
11:43 < bridge> <Assa> All followups (besides editor :justatest: ) are just cleanups and smaller things :justatest:
11:43 < ws-client> **<ChillerDragon>** go cleanup fat skins hrhrhrh
11:43 < ws-client> **<ChillerDragon>** road to 0 features
11:43 < bridge> <Assa> okay maybe one more which is a tad more complicated 🙈
11:43 < ws-client> **<ChillerDragon>** pf
11:44 < ws-client> **<ChillerDragon>** enjoy stale pr then
11:44 < bridge> <teero777> gm
11:44 < ws-client> **<ChillerDragon>** teerowo
11:44 < bridge> <Assa> 2 features on my list for rendering
11:44 < bridge> <Assa> - Zoom optimization (very small)
11:44 < bridge> <Assa> - Rotation clipping
11:45 < bridge> <Assa> both have huge benefits
11:45 < bridge> <teero777> :happy:
11:46 < bridge> <Assa> this is not how you get to 69 PRs
11:47 < bridge> <Assa> > **<ChillerDragon>** enjoy stale pr then
11:47 < bridge> <Assa> this is not how you get to 69 PRs
11:55 < bridge> <Assa> 2 features on my list for rendering
11:55 < bridge> <Assa> - Zoom optimization (very small in terms of code size)
11:55 < bridge> <Assa> - Rotation clipping
11:56 < bridge> <Assa> both have huge benefits and are based on quad clusters
12:46 < ws-client> **<ChillerDragon>** @robyt3 lez merge #10785 before the end of my shift
12:46 < chillerbot12> https://github.com/ddnet/ddnet/issues/10785
12:49 < bridge> <robyt3> Updated
12:49 < ws-client> **<ChillerDragon>** 🚀
12:53 < ws-client> **<ChillerDragon>** @robyt3 wat dis do? https://github.com/ddnet/ddnet/pull/10785/files#diff-9bc2378398eb6b7e1ab70a095b206eccef287f58209d704b1aff0650ea3cf426R598
12:53 < bridge> <Solly (bun bun)> look at linked issue
12:53 < bridge> <Solly (bun bun)> u have
12:53 < bridge> <Solly (bun bun)> y u is confuzzled
13:01 < bridge> <robyt3> Maximum chat message length plus some extra because the log contains `Chat: ` as prefix
13:02 < ws-client> **<ChillerDragon>** i wish we had https://github.com/ddnet/ddnet/pull/10712 then you could do ``char m_aDescription[256 + str_length("Chat: ")];``
13:05 < bridge> <patiga> chillerdragon: no docs apart from `help` and your type hints (also failed ping)
13:05 < bridge> <patiga> you can ofc also always ask here :)
13:06 < ws-client> **<ChillerDragon>** ye i was 60% sure ur lowercased
13:07 < ws-client> **<ChillerDragon>** but then i decided to not use twmap at all hrhrhrhr
13:07 < ws-client> **<ChillerDragon>** i was trying to figure out how to dump map settings didnt find it in the rust docs <:tee_thinking:478629518358085653>
13:07 < ws-client> **<ChillerDragon>** when with C++ `config_retrieve` then
13:07 < bridge> <patiga> map.info.settings
13:08 < ws-client> **<ChillerDragon>** https://zillyhuhn.com/cs/.2a8b6382-8c1e-4d13-b0ba-b918bae975a5.png
13:08 < ws-client> **<ChillerDragon>** this looks like map details to me
13:08 < ws-client> **<ChillerDragon>** not the ddnet config thingy
13:08 < ws-client> **<ChillerDragon>** called "server settings" in editor i think
13:08 < bridge> <patiga> same thing actually
13:08 < ws-client> **<ChillerDragon>** watfak
13:08 < ws-client> **<ChillerDragon>** ah
13:08 < ws-client> **<ChillerDragon>** `    pub settings: Vec<String>,`
13:08 < ws-client> **<ChillerDragon>** dis i assume
13:09 < bridge> <patiga> yup 👍
13:09 < ws-client> **<ChillerDragon>** yea thats not obvious somehow when skimming the docs
13:09 < ws-client> **<ChillerDragon>** as someone who uses the editor and doesnt speak the datafile protocol fluently i assume an entire different thing smh
13:09 < ws-client> **<ChillerDragon>** could maybe use some doc comment and example value idk
13:09 < bridge> <patiga> would you have found it, if `settings` was an attribute of the `Map` struct directly?
13:10 < ws-client> **<ChillerDragon>** i found this very page
13:10 < ws-client> **<ChillerDragon>** and closed it instantly because i realized thats a different thing
13:10 < bridge> <patiga> very fair, that is rather datafile/internal terminology, not editor speak
13:10 < ws-client> **<ChillerDragon>** i assume they are represented in the file together?
13:11 < ws-client> **<ChillerDragon>** because datafile is extendable smh
13:11 < ws-client> **<ChillerDragon>** i rly never looked into that file at all ever
13:11 < bridge> <patiga> wdym?
13:11 < ws-client> **<ChillerDragon>** in the editor UI those are two things
13:11 < ws-client> **<ChillerDragon>** and you mixed them together
13:11 < ws-client> **<ChillerDragon>** i assume that has a low level reason
13:11 < bridge> <patiga> yea they are in the same map item
13:11 < ws-client> **<ChillerDragon>** yep
13:11 < bridge> <patiga> (in the info item)
13:12 < bridge> <patiga> never really gave that more thought xd
13:12 < ws-client> **<ChillerDragon>** thats obvious to you the datafile hacker
13:12 < ws-client> **<ChillerDragon>** i am editor andy
13:12 < ws-client> **<ChillerDragon>** if i read author, version, cred .. my brain goes off and sees only this
13:12 < ws-client> **<ChillerDragon>** https://zillyhuhn.com/cs/.d1770d08-d612-4db2-ac96-60cf2759559a.png
13:13 < bridge> <patiga> do you think improved docs would be sufficient, or do you think we need some relocation/renaming?
13:13 < ws-client> **<ChillerDragon>** dis is entire different section in my head
13:13 < ws-client> **<ChillerDragon>** https://zillyhuhn.com/cs/.7bcaa9d1-14b9-487d-ae46-66fab8f93cd8.png
13:13 < ws-client> **<ChillerDragon>** can you put images in docs.rs? xd
13:14 < bridge> <patiga> I don't think so
13:14 < ws-client> **<ChillerDragon>** i think improved docs are enuff
13:14 < ws-client> **<ChillerDragon>** https://docs.rs/image-rs/latest/image_rs/
13:14 < ws-client> **<ChillerDragon>** they have images in docs
13:14 < bridge> <patiga> o tru
13:14 < bridge> <patiga> ig possible
13:14 < ws-client> **<ChillerDragon>** i assume readme ah syntax markdown `![text](url)`
13:15 < ws-client> **<ChillerDragon>** or url then text idk i always mess order up
13:15 < ws-client> **<ChillerDragon>** if these docs showed both my screens i would have never clicked away xd
13:15 < bridge> <patiga> https://docs.rs/image/latest/image/imageops/enum.FilterType.html (also reminded me of this)
13:15 < ws-client> **<ChillerDragon>** but its up to you
13:15 < ws-client> **<ChillerDragon>** maybe some text comment and some example string of `sv_deepfly 1` somewhere also works
13:16 < bridge> <patiga> wanted to ask exactly that
13:16 < bridge> <patiga> I think I'd try that first
13:16 < ws-client> **<ChillerDragon>** its just the last thing in the list and looks so generic
13:16 < ws-client> **<ChillerDragon>** it doesnt even have the editor exposed name "server settings"
13:16 < bridge> <patiga> if you like you can ofc do a pull request, otherwise I'd also do that in a while
13:16 < ws-client> **<ChillerDragon>** my brain doesnt even see it
13:17 < ws-client> **<ChillerDragon>** ye i thought about it but i think im too lazy sry xd
13:17 < bridge> <patiga> kinda expected that ^^
13:17 < bridge> <patiga> I think that might be the first naming issue I hit with twmap
13:22 < ws-client> **<ChillerDragon>** wdym naming issue? xd @patiga 
13:23 < ws-client> **<ChillerDragon>** you mean that the name "settings" is named wrong? and so far everything had a perfect name?
13:24 < ws-client> **<ChillerDragon>** not sure if that is really a naming issue.
13:25 < ws-client> **<ChillerDragon>** i guess one comment above the field might be enough so that the search in docs.rs for "server settings" gets a hit
13:28 < bridge> <patiga> perfect is exaggerated, I just meant I haven't heard about another misunderstanding like this before
13:34 < bridge> <Solly (bun bun)> sizeof("Chat: ")
13:35 < bridge> <Solly (bun bun)> [galaxybrain](https://cdn.discordapp.com/emojis/632454743951671296.webp?size=48&name=galaxybrain)
13:35 < bridge> <Solly (bun bun)> thats just 4 right
13:35 < bridge> <Solly (bun bun)> doesnt work
13:38 < bridge> <kebscs> sizeof - 1
13:39 < bridge> <Solly (bun bun)> i think it gets instantly cast to a ptr
13:39 < bridge> <Solly (bun bun)> or not
13:39 < bridge> <Solly (bun bun)> yeah - 1
13:40 < bridge> <kebscs> idk what ptr but sizeof counts null terminator
13:40 < bridge> <kebscs> strlen doesnt
13:40 < bridge> <Solly (bun bun)> "asdasdas" -> char*
13:40 < bridge> <Solly (bun bun)> instead of char[]
13:40 < bridge> <Solly (bun bun)> but sizeof doesnt make that happen
13:40 < bridge> <Solly (bun bun)> cuz its not a function
13:42 < ChillerDragon> oh yea sizeof - 1 works i used such strlen macro already
13:42 < ChillerDragon> but then it might not be power of 2 aligned which is bad for cpu speed
13:43 < ChillerDragon> i guess a simple comment like `// "Chat: " prefix` should also do
13:44 < ChillerDragon> https://zillyhuhn.com/cs/.bc6cf6f3-6a5f-49ba-abde-5c5050a374f7.png
13:45 < ChillerDragon> for pointer u need pointer ... duh
13:45 < ChillerDragon> https://zillyhuhn.com/cs/.4c855b51-e1bc-4d1f-8da6-e198da33f27e.png
13:56 < bridge> <Solly (bun bun)> @chillerdragon i cant make it compatible with original teesmash since it has more than 10 levels of health
13:56 < bridge> <Solly (bun bun)> and also the default configs are horrible
13:57 < bridge> <Solly (bun bun)> and its all pretty borked
14:09 < bridge> <Solly (bun bun)> @chillerdragon
14:09 < bridge> <Solly (bun bun)> https://cdn.discordapp.com/attachments/293493549758939136/1411321808883552256/image.png?ex=68b43b5c&is=68b2e9dc&hm=39d690955c28cc22a9de8c7cbc4b866593bcf174cdeb36ed601cfbc4e77382ca&
14:09 < bridge> <Solly (bun bun)> can i use the gun and hammer pickup here for the pickups?
15:08 < ws-client> **<ChillerDragon>** wat gun for wat pickups
15:09 < bridge> <ryozuki> i just had a grand realization
15:09 < bridge> <ryozuki> the default branch name to be trunk
15:09 < bridge> <ryozuki> makes a lot of sense
15:09 < bridge> <ryozuki> and u dont get woke anti slavery guys
15:09 < bridge> <ryozuki> and u dont surrender to main
15:09 < bridge> <ryozuki> and branches from a trunk
15:09 < bridge> <ryozuki> it just makes sense
15:09 < ws-client> **<ChillerDragon>** why was trunk even replaced by master
15:09 < bridge> <ryozuki> idk
15:09 < bridge> <ryozuki> apparently Linus hates svn and svn used trunk as default
15:10 < bridge> <ryozuki> and so he didnt use trunk (?)
15:10 < bridge> <ryozuki> when making git
15:10 < ws-client> **<ChillerDragon>** be careful discussing that
15:10 < bridge> <ryozuki> afaik tw used trunk
15:10 < bridge> <ryozuki> cuz svn
15:10 < bridge> <ryozuki> why
15:10 < ws-client> **<ChillerDragon>** last time i got 3 weeks ban for discussing branch names in here
15:10 < bridge> <ryozuki> ?? xd
15:10 < bridge> <Jupstar ✪> Bcs chiller is boomer
15:10 < bridge> <Jupstar ✪> Main
15:10 < ws-client> **<ChillerDragon>** im not kidding
15:11 < bridge> <ryozuki> im now a pro trunk
15:11 < bridge> <ryozuki> make trunk great again
15:11 < bridge> <Jupstar ✪> And Main alone
15:11 < bridge> <ryozuki> https://www.vinnie.work/blog/2020-10-08-git-master-main-trunk
15:12 < bridge> <ryozuki> > The most common reason I've seen for going to main is that its popular, short, to the point, and doesn't "interrupt muscle memory" (which sounds like a C coder's comment). My personal experience with revision control is deeply seeded in using subversion. In subversion you have trunk, branches, and tags. Trunk in this context refers to the fact that revision control commonly creates tree structures from graph theory. In my opinion, other than Linu
15:12 < bridge> <ryozuki> xdd
15:13 < bridge> <Jupstar ✪> Wrong
15:13 < bridge> <Jupstar ✪> It's the main branch. Not the trunk branch
15:13 < bridge> <Jupstar ✪> It makes most sense
15:14 < bridge> <ryozuki> if u think of git as a tree
15:14 < bridge> <ryozuki> it makes sense
15:14 < bridge> <Jupstar ✪> It also makes sense with main
15:14 < bridge> <ryozuki> :deen_star:
15:14 < bridge> <ryozuki> lets stop cuz discussding this has no intellectual merit actually xd
15:14 < bridge> <ryozuki> i achieved clean kernel threads on edos v2
15:14 < bridge> <Jupstar ✪> The real question is
15:15 < bridge> <ryozuki> ```
15:15 < bridge> <ryozuki> [9.986945] Next id (kernel = true): 0
15:15 < bridge> <ryozuki> [9.987044] hi from kthread
15:15 < bridge> <ryozuki> [10.985461] Context switch from kernel
15:15 < bridge> <ryozuki> [10.985574] Next id (kernel = true): 1
15:15 < bridge> <ryozuki> [10.985681] hi from kthread2, 206
15:15 < bridge> <ryozuki> [11.984080] Context switch from kernel
15:15 < bridge> <ryozuki> [11.984170] Next id (kernel = true): 2
15:15 < bridge> <ryozuki> [11.984253] hi from kthread3, 3
15:15 < bridge> <ryozuki> [12.982708] Context switch from kernel
15:15 < bridge> <ryozuki> [12.982819] Next id (kernel = true): 0
15:15 < bridge> <ryozuki> [12.982911] hi from kthread
15:15 < bridge> <ryozuki> [13.981330] Context switch from kernel
15:15 < bridge> <ryozuki> [13.981446] Next id (kernel = true): 1
15:15 < bridge> <ryozuki> [13.981550] hi from kthread2, 208
15:15 < bridge> <ryozuki> [14.979952] Context switch from kernel
15:15 < bridge> <ryozuki> [14.980038] Next id (kernel = true): 2
15:15 < bridge> <ryozuki> [14.980114] hi from kthread3, 4
15:15 < bridge> <ryozuki> [15.978579] Context switch from kernel
15:15 < bridge> <ryozuki> [15.978663] Next id (kernel = true): 0
15:15 < bridge> <ryozuki> [15.978733] hi from kthread
15:15 < bridge> <ryozuki> ```
15:15 < bridge> <Jupstar ✪> Why fork
15:15 < bridge> <mo_lt> why spoon
15:15 < bridge> <ryozuki> im context swithcing every 1 second more or less to see logs
15:15 < bridge> <ryozuki> i also added dmesg style logging
15:15 < bridge> <ryozuki> to see the time it takes
15:15 < bridge> <ryozuki> u can see the small drift in 1 second
15:16 < bridge> <ryozuki> cuz my calibration is not exact
15:16 < bridge> <ryozuki> ```
15:16 < bridge> <ryozuki> [0.000000] APIC Timer Calibration Results:
15:16 < bridge> <ryozuki> [0.000000]   Calibration time: 35ms
15:16 < bridge> <ryozuki> [0.000000]   APIC ticks elapsed: 35002472
15:16 < bridge> <ryozuki> [0.000000]   APIC frequency: 1000070628 Hz (1000) MHz
15:16 < bridge> <ryozuki> [0.000000]   TSC frequency: 4406088342 Hz (4.41) GHz
15:16 < bridge> <ryozuki> [0.000000]   Ticks per microsecond: 1000
15:16 < bridge> <ryozuki> ```
15:16 < bridge> <ryozuki> oh this time it got it perfect
15:16 < bridge> <ryozuki> qemu apic runs at 1ghz
15:16 < bridge> <ryozuki> sometimes my calbiration says 999
15:17 < bridge> <ryozuki> there is also a HPET that has femtosecond precision i think
15:18 < bridge> <ryozuki> but im not using it
15:18 < bridge> <ryozuki> https://wiki.osdev.org/HPET
15:18 < bridge> <ryozuki> HPET, or High Precision Event Timer, is a piece of hardware designed by Intel and Microsoft to replace older PIT and RTC. It consists of (usually 64-bit) main counter (which counts up), as well as from 3 to 32 32-bit or 64-bit wide comparators. HPET is programmed using memory mapped IO, and the base address of HPET can be found using ACPI.
15:18 < bridge> <ryozuki> im using the pit to calibrate
15:20 < bridge> <Jupstar ✪> You already have a terminal app?
15:26 < bridge> <ryozuki> on edos v2 no
15:26 < bridge> <ryozuki> im going slow to have a really good base
15:26 < bridge> <ryozuki> on v1 yes and i could run programs
15:27 < bridge> <ryozuki> rendering is actually easy so im not doing it now
15:27 < bridge> <ryozuki> (rendering with framebuffer)
15:33 < bridge> <ryozuki> i tried mixing rust async in edos v1
15:33 < bridge> <ryozuki> im going full separated threads
15:33 < bridge> <ryozuki> now
15:33 < bridge> <ryozuki> no async
15:53 < bridge> <qxdfox> why are all chat commands also registered in rcon? Is there any specific reason as to why or just because
15:57 < bridge> <chillerdragon> @qxdfox: I assume because of the shared structure. They use the same arg parser and cmdlist concept and so on
15:58 < bridge> <chillerdragon> Them being accessible in rcon also to the user might be convinient for admins. Also it helps with speed. Because rcon has no ratelimit but chat has even for authed players
17:33 < bridge> <Solly (bun bun)> @chillerdragon https://github.com/ddnet-insta/ddnet-insta/pull/416/files
17:33 < bridge> <Solly (bun bun)> @chillerdragon <https://github.com/ddnet-insta/ddnet-insta/pull/416>
17:38 < bridge> <Solly (bun bun)> @chillerdragon warmup seems to not work
18:04 < bridge> <chillerdragon> I hate warmup
18:04 < bridge> <chillerdragon> I have ptsd from that code
21:32 < bridge> <Cammodude> pip list gives this
21:32 < bridge> <Cammodude> https://cdn.discordapp.com/attachments/293493549758939136/1411433406885265569/image.png?ex=68b4a34b&is=68b351cb&hm=f2e871544734417ac636100170ee387b951f011fdde87922d380cb3dc0a70062&
21:32 < bridge> <Cammodude> and pip install does nothing
21:33 < bridge> <Cammodude> unless theres a special pip update command i dont know of
21:33 < bridge> <patiga> try `pip install --upgrade matplotlib`
21:34 < bridge> <Cammodude> ah now it does stuff
21:34 < bridge> <patiga> 🎉
21:34 < bridge> <Cammodude> its 3.10.6 now
21:34 < bridge> <patiga> nice, hopefully the script also runs now
21:34 < bridge> <Cammodude> nope
21:34 < bridge> <Cammodude> https://cdn.discordapp.com/attachments/293493549758939136/1411434035774755000/image.png?ex=68b4a3e1&is=68b35261&hm=85fa2c984283194a8500aaf47a0c856f863082c8015bf18ce19f58b5d8108897&
21:35 < bridge> <Cammodude> how do i delete stuff with pip now
21:35 < bridge> <patiga> ah oh god
21:35 < bridge> <Cammodude> or actually
21:35 < bridge> <patiga> you are on linux
21:35 < bridge> <Cammodude> i dont recall ever doing pip install matplotlib
21:36 < bridge> <patiga> ouch
21:36 < bridge> <Cammodude> so its system i presume
21:36 < bridge> <patiga> then you'll need to update it with your system package manager
21:36 < bridge> <patiga> (which distro you on btw?)
21:36 < bridge> <Cammodude> mint 21.3
21:37 < bridge> <Cammodude> i dont see any python related stuff in update manager
21:37 < bridge> <Cammodude> ill update anyway
21:37 < bridge> <patiga> did pip give you a warning before it let you install kaitaistruct?
21:37 < bridge> <Cammodude> i dont think so
21:38 < bridge> <Cammodude> i closed that terminal so i cant check either
21:38 < bridge> <Cammodude> but if it did give me an error it just continued on
21:39 < bridge> <patiga> we only see a warning here, the script might've still run to the end
21:40 < bridge> <Cammodude> hah im just blind :kek:
21:40 < bridge> <patiga> though your python dependency management might've just gotten a little more complicated. I'm not familiar with mint and pip usually warns you before something goes wrong
21:40 < bridge> <Cammodude> https://cdn.discordapp.com/attachments/293493549758939136/1411435445392379985/output.png?ex=68b4a531&is=68b353b1&hm=215c71717a53b607c8ed1293d56bf5c89881a1803a1f5ccb18d152ff060a697d&
21:40 < bridge> <Cammodude> update manager was covering the folder
21:40 < bridge> <Cammodude> so i couldnt see it
21:40 < bridge> <Cammodude> ty
21:40 < bridge> <patiga> yw!
22:05 < bridge> <ryozuki> night question: how many stacks needs a user thread
22:08 < bridge> <patiga> 2?
22:10 < bridge> <melon> i call 1
22:11 < bridge> <patiga> at least one user stack and one os stack afaik
22:11 < bridge> <melon> i think the OS itself determines additionals for ops
22:13 < bridge> <ryozuki> 2 stacks, because one is needed in the TSS when doing a privilege level change
22:13 < bridge> <patiga> tss?
22:13 < bridge> <ryozuki> task state segment
22:13 < bridge> <ryozuki> when ur in ring 3 (user) and get an interrupt
22:14 < bridge> <ryozuki> cpu looks at TSS for the stack to use on ring 0 for the kernel
22:14 < bridge> <ryozuki> that stack needs to be separate for each ring 3 thread
22:14 < bridge> <patiga> yea thats the one I thought of I think
22:14 < bridge> <ryozuki> so no corruption happens
22:14 < bridge> <ryozuki> and it uses the stack in the interrupt handler
22:15 < bridge> <ryozuki> its called RSP0
22:15 < bridge> <ryozuki> https://en.wikipedia.org/wiki/Task_state_segment#Inner-level_stack_pointers
22:15 < bridge> <ryozuki> > The TSS contains 6 fields for specifying the new stack pointer when a privilege level change happens. The field SS0 contains the stack segment selector for CPL=0, and the field ESP0/RSP0 contains the new ESP/RSP value for CPL=0. When an interrupt happens in protected (32-bit) mode, the x86 CPU will look in the TSS for SS0 and ESP0 and load their values into SS and ESP respectively. This allows for the kernel to use a different stack than the user
22:16 < bridge> <ryozuki> u also need to use a different stack for double fault handlers
22:18 < bridge> <teero777> yo ryo do you know how much faster syscalls can be generally? I mean they are hella slow compared to anything else in programming so I wondered if there was a way to speed them up. Might be a silly question.
22:19 < bridge> <ryozuki> yes, in a fully cooperative kernel they can be way faster, because if u cooperate by yielding at known points, you can only save the state you need, instead of all the registers and x87 state, for example, rust async is fully cooperative, rust creates a state machine where at yield points (.awaits) a new variant in the enum state machine is made, saving only the stat needed, for example it may be as little as one u64
22:19 < bridge> <ryozuki> most real world kernels need to preempt non cooperatively because user programs need to be stoppepd at any point
22:19 < bridge> <ryozuki> so you need to save all info
22:20 < bridge> <ryozuki> yes, in a fully cooperative kernel they can be way faster, because if u cooperate by yielding at known points, you can only save the state you need, instead of all the registers and x87 state, for example, rust async is fully cooperative, rust creates a state machine where at yield points (.awaits) a new variant in the enum state machine is made, saving only the state needed, for example it may be as little as one u64
22:20 < bridge> <ryozuki> i havent looked into RTOS but maybe they do that
22:20 < bridge> <teero777> thanks, good to know :D
22:21 < bridge> <ryozuki> (btw im no expert i just learnt this when making OSes xD)
22:22 < bridge> <ryozuki> if ur curios, right now this is how i save in a context switch:
22:22 < bridge> <ryozuki> 
22:22 < bridge> <ryozuki> ```rust
22:22 < bridge> <ryozuki> #[unsafe(naked)]
22:22 < bridge> <ryozuki> #[unsafe(no_mangle)]
22:22 < bridge> <ryozuki> pub unsafe extern "C" fn timer_interrupt_handler() {
22:22 < bridge> <ryozuki>     naked_asm!(
22:22 < bridge> <ryozuki>         // CPU has already pushed SS, RSP, RFLAGS, CS, RIP
22:22 < bridge> <ryozuki>         // Save all general purpose registers
22:22 < bridge> <ryozuki>         "push rax",
22:22 < bridge> <ryozuki>         "push rcx",
22:22 < bridge> <ryozuki>         "push rdx",
22:22 < bridge> <ryozuki>         "push rbx",
22:22 < bridge> <ryozuki>         "push rbp",
22:22 < bridge> <ryozuki>         "push rsi",
22:22 < bridge> <ryozuki>         "push rdi",
22:22 < bridge> <ryozuki>         "push r8",
22:22 < bridge> <ryozuki>         "push r9",
22:23 < bridge> <ryozuki>         "push r10",
22:23 < bridge> <ryozuki>         "push r11",
22:23 < bridge> <ryozuki>         "push r12",
22:23 < bridge> <ryozuki>         "push r13",
22:23 < bridge> <ryozuki>         "push r14",
22:23 < bridge> <ryozuki>         "push r15",
22:23 < bridge> <ryozuki> 
22:23 < bridge> <ryozuki>         // At this point, RSP points to the saved context
22:23 < bridge> <ryozuki>         // Pass it as first argument to timer_schedule
22:23 < bridge> <ryozuki>         "mov rdi, rsp",
22:23 < bridge> <ryozuki> 
22:23 < bridge> <ryozuki>         // Ensure stack is 16-byte aligned before call
22:23 < bridge> <ryozuki>         // The push operations above pushed 15 registers (8 bytes each = 120 bytes)
22:23 < bridge> <ryozuki> i dont save x87 yet cuz i didnt enable it
22:23 < bridge> <ryozuki> but thats 512 bytes afaik
22:23 < bridge> <ryozuki> on a syscall u have to do the same as the interrupt handler more or less
22:23 < bridge> <ryozuki> well maybe u can just save callee saved regs
22:24 < bridge> <ryozuki> (im not sure now)
22:24 < bridge> <ryozuki> @teero777 btw linux has an optimization for some syscalls
22:24 < bridge> <ryozuki> its called the vdso
22:24 < bridge> <ryozuki> https://man7.org/linux/man-pages/man7/vdso.7.html
22:24 < bridge> <ryozuki> interesting read
22:24 < bridge> <teero777> I haven't gotten into OS/kernel dev at all although I'm planning for it once I finish my ongoing project
22:25 < bridge> <ryozuki> nice, did u get inspired by me or u already wanted? :D
22:25 < bridge> <ryozuki> https://tenor.com/view/burger-eating-frieren-frieren-beyond-journey%27s-end-sousou-no-frieren-gif-13425073513713719938
22:25 < bridge> <ryozuki> lot of things make sense to me now
22:26 < bridge> <teero777> I wanted to do it for some time but then I saw your sexy screenshots here and got back the motivation
22:26 < bridge> <ryozuki> its dope and if u use C its easier cuz more examples
22:26 < bridge> <ryozuki> im using rust cuz i dont wanna use C
22:26 < bridge> <ryozuki> anyway if u do start and have doubts feel free to ask
22:27 < bridge> <teero777> yea, will, thx :owo:
22:29 < bridge> <ryozuki> btw i tried to look at C inline asm but the syntax was weird xD
22:29 < bridge> <teero777> what is this prejudice😭
22:29 < bridge> <teero777> https://cdn.discordapp.com/attachments/293493549758939136/1411447831574876342/Screenshot_20250830-222922.png?ex=68b4b0ba&is=68b35f3a&hm=5dd308598b4835094e1cd45332492133eefe289e1f31620bfc5b21ab57c581a2&
22:31 < bridge> <teero777> Yea I looked into it for optimization things but tbh it's useless most of the time, clang does a pretty good job. It's really hard to find anything worth doing inline asm for.
22:31 < bridge> <teero777> Ofc your use case is different
22:34 < bridge> <teero777> It was like
22:34 < bridge> <teero777> 1. spend an hour understanding things
22:34 < bridge> <teero777> 2. oh, it's slower than what the compiler did...
22:35 < bridge> <ryozuki> xD
22:36 < bridge> <ryozuki> yeah i think asm is only useful for constant time (cryptography) and maybe codecs
22:36 < bridge> <ryozuki> and for stuff that breaks out of standards (something non system v compliant)
22:36 < bridge> <ryozuki> for example the interrupt handler u cant do it with rust or C code
22:36 < bridge> <ryozuki> u need to use asm
22:37 < bridge> <ryozuki> @nikhard whatch u writing
22:37 < bridge> <melon> can your car run out of blinker fluid yet
22:37 < bridge> <teero777> :justatest:
22:38 < bridge> <teero777> I'm fighting the code
22:38 < bridge> <teero777> gdb is assisting me
22:38 < bridge> <melon> :peepoSpreadsheet:
22:38 < bridge> <teero777> that I know of
22:38 < bridge> <melon> did you ask AI yet
22:39 < bridge> <teero777> AI doesn't take my 3700loc and if it does it spews bs xd
22:39 < bridge> <melon> 😢
22:39 < bridge> <teero777> I even tried to give it some telemetry data but nope, and after a single prompt it's rate limited anyways xd
22:40 < bridge> <nikhard> sad letter about my helplessness and crying for help :(
22:40 < bridge> <nikhard> but for real want help with modifying ninja weapon animation, it's not moving
22:42 < bridge> <ryozuki> if its not code this isnt the place probs
22:43 < bridge> <nikhard> well, #questions ?
22:43 < bridge> <nikhard> ah, wait, yea, it's code
22:47 < bridge> <nikhard> Maybe you know where does weapon animations stores?
22:47 < bridge> <nikhard> Like where coded weapons recoil movement and it's rotation, animations. Not `datasrc/content.py`, it's not wery helping with my problem
22:49 < bridge> <patiga> `content.py` is indeed the right place, but it is used to generate other files
22:49 < bridge> <patiga> not sure if you have to execute it manually/something different
22:49 < bridge> <teero777> Did you know about descriptor structs in c? Or whatever they are called.
22:49 < bridge> <teero777> it's pwetty
22:50 < bridge> <teero777> googling doesn't even bring anything up wow
22:51 < bridge> <melon> descriptor strucs as in
22:51 < bridge> <melon> 
22:51 < bridge> <melon> ```c
22:51 < bridge> <melon> struct myStruct {
22:51 < bridge> <melon> int something;
22:51 < bridge> <melon> uint16_t somethingmore;
22:51 < bridge> <melon> char *somethingevenharder;
22:51 < bridge> <melon> };
22:51 < bridge> <melon> ```
22:51 < bridge> <melon> ?
22:52 < bridge> <melon> i guess its just a defined interface to describe the data structure
22:53 < bridge> <nikhard> Well my problem exactly is in that file generates `Animation` for *ninja_swing* and after attaching `anim.attach.frames.Add(AnimKeyframe(0.20, 32, 0, 0))`, the **x** and **y** value does nothing in game. With hammer, feet, body - works, but not with ninja.
22:53 < bridge> <nikhard> So I assume there is some cut for ninja exactly, but not in that file. maybe)
22:55 < bridge> <teero777> descriptor structs as in you have 2 structs instead of one, one of them is a suffixed with `descriptor` or `desc` and holds all data that won't be changed past initialization. And the *real* struct that holds data that may change at any time has the descriptor structs inside of it to get the *static* data out of
22:55 < bridge> <melon> oh so it is an interface
22:55 < bridge> <teero777> kinda, depends on what you mean with an interface xd
22:56 < bridge> <melon> well an interface in the broad sense to me is just defining mutable data which can be inherited and changed, but it can also hold immutable structs
22:57 < bridge> <melon> i guess i see interfaces more like templates
22:57 < bridge> <melon> xd
22:57 < bridge> <teero777> I'm restructuring and kinda redoing the whole vehicle simulation thing so I wanted to try doing this and it's been working out great.
22:57 < bridge> <teero777> have a few commits laying on local cuz it's not at the level it was before
22:57 < bridge> <patiga> if you change something in the file, does compiling again do smth or is it immediately finished?
22:58 < bridge> <teero777> @blaiszephyr anyways today is early sleepy so gn
22:58 < bridge> <nikhard> Doing. Updating 123 entries 💀
22:58 < bridge> <patiga> gn :>
22:59 < bridge> <melon> gn Teeto <3
22:59 < bridge> <patiga> then change something drastic in the animation to verify that actually nothing is changing
22:59 < bridge> <patiga> like remove all keyframes or smth
23:00 < bridge> <nikhard> already
23:00 < bridge> <nikhard> i make foot moving when ninja animates and commented ninja's anim, it stops in vertical position
23:03 < bridge> <patiga> ah, so stuff is indeed changing
23:06 < bridge> <patiga> I don't know why the position doesn't change, it *should* afaict
23:14 < bridge> <ryozuki> im going sleep too
23:16 < bridge> <pioooooo> gm
23:22 < bridge> <nikhard> well i make myself a unicorn, worth enough
23:22 < bridge> <nikhard> muzzleoffset and offset are working tough
23:22 < bridge> <nikhard> https://cdn.discordapp.com/attachments/293493549758939136/1411461173093601340/image.png?ex=68b4bd27&is=68b36ba7&hm=ccb35d62850889fde2870fd98c35963a3c968e543182997cb15d140e4c202012&