02:02 < bridge> <teero777> Yea same lol, with std::fill I would need to check cpp reference to know what it actually does
02:24 < bridge> <kebscs> It's used a lot in the ddnet code anyway
02:24 < bridge> <kebscs> Fill just fills the variable, it's not that confusing as other cpp stuff
02:26 < bridge> <gorp_tw> I just found this today
02:26 < bridge> <gorp_tw> I think I've been affected by this bug for a long time and just haven't noticed the exact cause / thought it was random until now
02:26 < bridge> <gorp_tw> or that I was going crazy because normally the tz setting that gets eaten is like off screen and the whole thing shifts and stuff seems to be randomly broken lol
02:27 < bridge> <gorp_tw> so, the fix is to check the full string before replacement
02:27 < bridge> <gorp_tw> (up to the value, obviously if the key were the same it should replace it)
02:27 < bridge> <gorp_tw> (up to the value, obviously if the key for what is being tuned were the same it should replace it)
02:29 < bridge> <gorp_tw> very frequently i will find that i can actually combine some tunes for simplicity, then doing this actually just sucks and i have to end up adding them all instead and deleting the old ones
02:29 < bridge> <gorp_tw> not to mention there is no easy way to choose where a setting appears in the list
02:29 < bridge> <gorp_tw> (at cursor position should be an option)
02:32 < bridge> <gorp_tw> I think overall the fact that it replaces redundant tunes is... a bit weird UX wise, I would prefer a warning but I understand why the check is done at all
02:33 < bridge> <blazulite> ddnet-rs' DDrace hud concept
02:33 < bridge> <blazulite> https://cdn.discordapp.com/attachments/293493549758939136/1414045916658270329/image.png?ex=68be2462&is=68bcd2e2&hm=db26266900661eaa9aa619517c4c3300768b3e975262fb1ba5ac4e7f914a4d7c&
06:54 < bridge> <ryozuki> morning
08:27 < bridge> <Ewan> gm
08:36 < bridge> <filoqcus> This is wonderful:poggers2:
08:41 < bridge> <blazulite> Aw tyyyyyyyyy :3
08:41 < bridge> <blazulite> I love the difference between #showroom and #developer
08:41 < bridge> <blazulite> https://cdn.discordapp.com/attachments/293493549758939136/1414138538227666975/image.png?ex=68be7aa4&is=68bd2924&hm=d6012ef933b5e1151d43dc72e9118a89d06b801586c93704ff8e700ae481a83f&
08:41 < bridge> <filoqcus> but we are still waiting for accounts...
08:42 < bridge> <filoqcus> :feelsbadman:
08:42 < bridge> <chillerdragon> They are done
08:42 < bridge> <chillerdragon> This is ddnet-rs hud and ddnet-rs has accounts
08:43 < bridge> <filoqcus> so we just have to wait for ddnet-rs to be released?
08:43 < bridge> <filoqcus> :nouis:
08:43 < bridge> <chillerdragon> Why wait
08:43 < bridge> <chillerdragon> Play it today
08:43 < bridge> <filoqcus> Well, it's not completely done D:
08:44 < bridge> <chillerdragon> That probably won’t change in the near future
08:44 < bridge> <chillerdragon> But it’s quite complete actually
08:44 < bridge> <blazulite> lies lies lies
08:44 < bridge> <chillerdragon> I would play it already if gameplay wouldn’t feel off
08:44 < bridge> <chillerdragon> But the game sense might be a me issue
08:44 < bridge> <chillerdragon> Maybe it’s ready to play for others
08:45 < bridge> <blazulite> only thing that feels of for me is the emote wheel like oh my goodness gracious why can't i go all the way on the scroll
08:45 < bridge> <chillerdragon> See
08:45 < bridge> <blazulite> and also why is there a specific cursor for it
08:45 < bridge> <chillerdragon> @filoqcus: listen to dis guy he says it’s playable
08:45 < bridge> <filoqcus> By the way, what an idea to create a clan system in DDNetPP, and clean up the db table with accounts a bit, because there are +- 100 values and half of them are not used
08:46 < bridge> <chillerdragon> Yes the db should be cleaned
08:46 < bridge> <chillerdragon> But not sure if I want to do a major db refactor now
08:46 < bridge> <chillerdragon> I rather focus on polishing the bugs first
08:47 < bridge> <chillerdragon> Also im a bit busy with ddnet-insta and ddnet and becoming top1 gores player
08:48 < bridge> <blazulite> ah but what about becoming top 1 employee  /j
08:48 < bridge> <chillerdragon> Employee?
08:48 < bridge> <chillerdragon> Where
08:49 < bridge> <chillerdragon> Im free dancer
08:49 < bridge> <filoqcus> I know 😄
08:49 < bridge> <filoqcus> https://cdn.discordapp.com/attachments/293493549758939136/1414140516932517990/IMG_2935.png?ex=68be7c7c&is=68bd2afc&hm=e66fae44470f9b5b9617cb81343e84eb9638ed38adde9cc5f8d3ea24939696f1&
08:49 < bridge> <filoqcus> Pro Kogger
08:50 < bridge> <chillerdragon> yes the green full impact carry bush is often on kog
08:53 < bridge> <filoqcus> when is Teeworlds in 2025? :omo:
09:05 < bridge> <chillerdragon> I usually upload these new years but not sure if there is one coming this year
09:17 < ws-client> **<ChillerDragon>** can someone pls send some git merge driver pls
09:18 < ws-client> **<ChillerDragon>** how is it that in 2025 i still get a git conflict when two new messages were added to the same array
09:18 < ws-client> **<ChillerDragon>** https://zillyhuhn.com/cs/.d574303f-e3d9-43e1-9c13-6b3839edd637.png
09:18 < ws-client> **<ChillerDragon>** like bro i want both adds. Just take it and be quiet. Watafak i want with a conflict here. So useless.
09:23 < bridge> <Jupstar ✪> Wat
09:23 < bridge> <Jupstar ✪> Just use AI to resolve merge conflicts like every 2025 programmer would do
09:32 < bridge> <Jupstar ✪> Why are harddrives so expensive rn xD
09:33 < bridge> <Jupstar ✪> I want 8TB ssd, but these mf costs like 600+ bucks
09:34 < bridge> <Jupstar ✪> Why do we have nobody here that works for Samsung xd
09:37 < bridge> <Ewan> good question
09:38 < bridge> <Ewan> boot times are way better on my new drive but I moved a lot of stuff around & now NixOS just lives on multiple drives where before it was spread out
09:38 < bridge> <Ewan> I'm happy with my computer upgrade. time to get 64gb more ram
09:42 < bridge> <Jupstar ✪> I swear RAM is never enough.
09:42 < bridge> <Jupstar ✪> 
09:42 < bridge> <Jupstar ✪> Hard to believe just few years ago 4GB was the standard
09:42 < bridge> <Jupstar ✪> I am rarely under 32GB usage nowadays xD
09:43 < bridge> <Jupstar ✪> 2-3 code projects open + browsers = rip
09:49 < bridge> <Ewan> same
09:50 < bridge> <Ewan> if ur a mega multitasker u may ahti have open VMs or games
09:50 < bridge> <Ewan> if ur a mega multitasker u may also have open VMs or games
09:50 < bridge> <Jupstar ✪> Ok I tried codex from openai now on ddnet-rs. Simple task
09:50 < bridge> <Jupstar ✪> 
09:50 < bridge> <Jupstar ✪> Prompt
09:50 < bridge> <Jupstar ✪> ```
09:50 < bridge> <Jupstar ✪> the system.rs in lib/base currently only exposes time.
09:50 < bridge> <Jupstar ✪> 
09:50 < bridge> <Jupstar ✪> I want to rename it to steady_clock.rs
09:50 < bridge> <Jupstar ✪> 
09:50 < bridge> <Jupstar ✪> Additionally the SystemTimeInterface can be removed. time_get -> now
09:50 < bridge> <Jupstar ✪> The whole System struct can be renamed to SteadyClock and merged with SystemTime. It should still be clonable tho -> start_time: Arc<std::time::Instant>.
09:50 < bridge> <Jupstar ✪> 
09:50 < bridge> <Jupstar ✪> Make sure the whole program still compiles, since it's used in lots of places
09:50 < bridge> <Jupstar ✪> ```
09:51 < bridge> <Jupstar ✪> 
09:51 < bridge> <Jupstar ✪> https://github.com/ddnet/ddnet-rs/pull/143/commits/6e5b15f9faa25021cb31ac16cc5b33024c65517e
09:51 < bridge> <Jupstar ✪> 
09:51 < bridge> <Jupstar ✪> First of all, it compiles.. Nice.
09:51 < bridge> <Jupstar ✪> 
09:51 < bridge> <Jupstar ✪> But it throws 1 warning.
09:51 < bridge> <Jupstar ✪> 
09:51 < bridge> <Jupstar ✪> Additionally from a human I'd have expected renames like this:
09:51 < bridge> <Jupstar ✪> ```diff
09:51 < bridge> <Jupstar ✪> -sys: Arc<dyn SystemTimeInterface>,
09:51 < bridge> <Jupstar ✪> +time: SteadyClock,
09:51 < bridge> <Jupstar ✪> ```
09:51 < bridge> <Jupstar ✪> This AI thing simply left them called `sys`.
09:51 < bridge> <Jupstar ✪> 
09:51 < bridge> <Jupstar ✪> I'd rate it a 7/10. I mean I had to do nothing except creating the prompt, but I'll need to refactor stuff. Also it took really long, I think the AI often compiled the project to check for errors (instead of e.g. using clippy).
09:51 < bridge> <Jupstar ✪> My pc fans were on most of the time..
09:51 < bridge> <Jupstar ✪> 
09:51 < bridge> <Jupstar ✪> It took like 15 minutes for that prompt
09:51 < bridge> <Jupstar ✪> I think I could have done it in a similar time
09:51 < bridge> <melon> ohoho, thats not half bad
09:51 < bridge> <melon> turn on prompt, say you're compiling and grab a coffee
09:52 < bridge> <melon> :nouis:
09:54 < bridge> <Ewan> just realized even tho there's not enough space in the case, I have a free 16x slot for my old gpu now but I can mount it vertically...
09:54 < bridge> <Ewan> fill the thing with fans and put it somewhere too far to hear
09:59 < bridge> <Ewan> opinions?
10:00 < bridge> <Jupstar ✪> i dont understand how you want to put your gpu far away xD
10:02 < bridge> <Ewan> no the whole computer
10:02 < bridge> <Ewan> it will be so loud to air cool all that
10:02 < bridge> <Ewan> especially in my case
10:03 < bridge> <Jupstar ✪> Buy new case
10:03 < bridge> <Jupstar ✪> `Arc<SteadyClock>` ok that is one example where AI is retarded again.. SteadyClock itself is already clonable
10:04 < bridge> <Jupstar ✪> -1 point for such an obvious error
10:06 < bridge> <Jupstar ✪> `Arc<SteadyClock>` ok that is one example where AI is retarded again.. SteadyClock itself is already clonable (even mentioned in the prompt)
10:10 < bridge> <Assa> brilliant, I am going to ask chatgpt to write a rewview about a PR
10:12 < bridge> <Jupstar ✪> I revise my call, AI still sucks too much
10:12 < bridge> <Jupstar ✪> Gotta wait another 5 years
10:15 < bridge> <Assa> My test is over, AI failed me, it just states obvious things and wants me to write unit tests
10:29 < ws-client> **<ChillerDragon>** promting skill issue
10:30 < ws-client> **<ChillerDragon>** my 2yo brother already prompts like a gigachad he is just AI native. We boomers just dont get it.
10:31 < bridge> <Jupstar ✪> if i prompt every step in detail then i dont need ai 😬
10:31 < bridge> <Jupstar ✪> but true
10:32 < bridge> <Jupstar ✪> 5 years old props can better deal with AI than us boomers already
10:33 < chillerbot12> Elo @ChillerDragon I just wanted to remind you that: twmap check for missing solo tiles around spawn for https://github.com/ddnet/ddnet/pull/10844
10:34 < ws-client> **<ChillerDragon>** @Jupstar ✪ u still write prompts by hand??? ur so 2024
10:34 < ws-client> **<ChillerDragon>** use AI to write prompt
10:36 < ws-client> **<ChillerDragon>** AI is like mongodb. You turn it on and it scales right up. The only limiting factor is the human using it.
10:36 < bridge> <Jupstar ✪> I attached my mind reader cables now.. how stupid from me
10:36 < ws-client> **<ChillerDragon>** I replaced my self at work with AI 3 months ago. So far nobody noticed.
10:37 < bridge> <Jupstar ✪> You still work?
10:37 < ws-client> **<ChillerDragon>** no anymore
10:37 < bridge> <Jupstar ✪> What even is work?
10:37 < ws-client> **<ChillerDragon>** reviewing ddnet prs is work
10:37 < bridge> <Jupstar ✪> I was born in a world where machines do everything
10:38 < bridge> <Jupstar ✪> As soon as i hit the jackpot I gonna be fulltime ddnet dev too
10:38 < bridge> <Jupstar ✪> Just wait few days
10:38 < ws-client> **<ChillerDragon>** ok enuff trolling back to work i have short shift today
10:38 < bridge> <Jupstar ✪> Ah yeah true, sunday
10:38 < ws-client> **<ChillerDragon>** MERGEDAY
10:39 < ws-client> **<ChillerDragon>** rossbit already started at like 00:01 AM
10:39 < ws-client> **<ChillerDragon>** embracing merge day with early bird mentality
10:39 < bridge> <Jupstar ✪> Robyte at 23:59.
10:39 < bridge> <Jupstar ✪> "Only 1 minute left til I can finnaly start a new day of work"
10:39 < ws-client> **<ChillerDragon>** xd
10:39 < ws-client> **<ChillerDragon>** fr
10:40 < ws-client> **<ChillerDragon>** he is not hooman i swear
10:40 < bridge> <Jupstar ✪> Maybe he is AI.. have u ever seen him IRL?
10:40 < ws-client> **<ChillerDragon>** I have not
10:40 < ws-client> **<ChillerDragon>** have you?
10:40 < bridge> <Jupstar ✪> Yes I have, but have you met me IRL?
10:40 < ws-client> **<ChillerDragon>** I have not
10:40 < bridge> <Jupstar ✪> I could be his companion
10:40 < ws-client> **<ChillerDragon>** have you?
10:40 < bridge> <Jupstar ✪> I do
10:40 < ws-client> **<ChillerDragon>** Hm
10:40 < ws-client> **<ChillerDragon>** I have meet Zwelf irl 
10:41 < ws-client> **<ChillerDragon>** he can confirm you are real
10:41 < bridge> <Jupstar ✪> But did he?
10:41 < ws-client> **<ChillerDragon>** he did not.
10:41 < bridge> <Jupstar ✪> You are real bcs someone else confirmed you
10:41 < ws-client> **<ChillerDragon>** whO????
10:41 < bridge> <Jupstar ✪> You know xd
10:41 < ws-client> **<ChillerDragon>** i know
10:42 < ws-client> **<ChillerDragon>** Meskalin saw my passport
10:42 < bridge> <teero777> What xD
10:42 < ws-client> **<ChillerDragon>** yes
10:42 < ws-client> **<ChillerDragon>** and i saw his left egg
10:42 < bridge> <Jupstar ✪> ?
10:42 < bridge> <teero777> is that enough confirmation though
10:42 < ws-client> **<ChillerDragon>** okay now i rly gotta go back to productiveness
10:42 < bridge> <Jupstar ✪> Chiller u gey?
10:42 < ws-client> **<ChillerDragon>** yes
10:42 < bridge> <Jupstar ✪> ogay
10:43 < bridge> <teero777> meskalin isn't though, how did you manage to see that lmao
10:43 < ws-client> **<ChillerDragon>** it was soo funny ngl
10:43 < ws-client> **<ChillerDragon>** but i dont wanna dox him here
10:43 < ws-client> **<ChillerDragon>** he is bae
10:56 < bridge> <Jupstar ✪> yo chiller
10:58 < bridge> <Jupstar ✪> what hertz your monitor has that you used yesterday ingame?
10:58 < bridge> <Jupstar ✪> and did you also check fps during our gameplay?
10:58 < bridge> <Jupstar ✪> 
10:58 < bridge> <Jupstar ✪> bcs the default fps target is 4x hertz. just for fair comparision you'd need same fps in ddnet
10:58 < bridge> <Jupstar ✪> and if it's fps problem that ofc needs fixed xd
10:59 < bridge> <Jupstar ✪> and the freakin prediction timer needs fix
11:00 < bridge> <Jupstar ✪> harder than vulkan backend, 100% guaranteed
11:00 < ws-client> **<ChillerDragon>** @Jupstar ✪ its 10 year old 60hz 4k monitor
11:01 < ws-client> **<ChillerDragon>** u see fps in the clip right? 700 something
11:01 < bridge> <Jupstar ✪> oke
11:01 < ws-client> **<ChillerDragon>** its not a fps problem i think
11:01 < ws-client> **<ChillerDragon>** i play ddnet with 666 fps
11:02 < bridge> <Jupstar ✪> Well maybe this is not a FPS problem, but FPS is a problem
11:02 < bridge> <Jupstar ✪> When I play 500 fps it sucks hard
11:02 < bridge> <Jupstar ✪> 1000 fps smoothie
11:03 < bridge> <Jupstar ✪> Just want to minimize other factors
11:03 < bridge> <Jupstar ✪> :deen_star:
11:05 < ws-client> **<ChillerDragon>** @learath2 @robyt3 can i only move parts of system.cpp to str.cpp for now? Doing this in smaller prs will be so much more fun and safer. Otherwise I have to try to speedrun it now.
11:05 < ws-client> **<ChillerDragon>** wait
11:05 < ws-client> **<ChillerDragon>** how did i never see this?
11:05 < ws-client> **<ChillerDragon>** do we have numbered admins?????
11:06 < ws-client> **<ChillerDragon>** heinrich5991 learath2 robyt3
11:06 < ws-client> **<ChillerDragon>** 1 2 3
11:06 < ws-client> **<ChillerDragon>** WATATFAKA
11:06 < ws-client> **<ChillerDragon>** ILLUMINATI
11:06 < ws-client> **<ChillerDragon>** @Jupstar ✪ my laptop has more hz i think it was also wonkey there iirc
11:06 < bridge> <learath2> Ofc you can do it in pieces
11:07 < ws-client> **<ChillerDragon>** <:poggers2:1008007455936094328> @learath2 did i ever tell you that you are my favorite maintainer?
11:08 < ws-client> **<ChillerDragon>** dennis felsin9 would set the max of admins to 9 i assume?
11:08 < ws-client> **<ChillerDragon>** interesting
11:08 < ws-client> **<ChillerDragon>** gonna rename to ChillerDrag0
11:08 < ws-client> **<ChillerDragon>** ez admin glitch
11:09 < ws-client> **<ChillerDragon>** im sober i swear
11:10 < ws-client> **<ChillerDragon>** @learath2 how do we order?
11:11 < ws-client> **<ChillerDragon>** first all non utf8 and then all utf8?
11:11 < ws-client> **<ChillerDragon>** or mix them to have str_truncate and str_utf8_truncate together?
11:18 < bridge> <totar> You can ask for ai review on any PR I think and the githuh AI will do it, I think you just request a review from co pilot.
11:20 < bridge> <learath2> Not all that important to me, but first all non utf then all utf sounds a tad better
11:20 < ws-client> **<ChillerDragon>** to me as well
11:20 < bridge> <learath2> You need to ask githubs ai objective stuff. It refuses to form any opinion 😄
11:21 < ws-client> **<ChillerDragon>** u mean subjective?
11:21 < bridge> <learath2> No I mean objective. If you ask it to have a subjective opinion it doesnt answer
11:22 < bridge> <learath2> E.g. "do you think this change is for the better" didn't work last I tried
11:22 < ws-client> **<ChillerDragon>** How can objective stuff have a subjective opinion? xd
11:22 < ws-client> **<ChillerDragon>** isnt objective = clear
11:22 < ws-client> **<ChillerDragon>** and subjective = opinions possible
11:23 < ws-client> **<ChillerDragon>** Like objective is when you have clear evidence
11:24 < ws-client> **<ChillerDragon>** Reminds me of the one time our Ethics teacher said in front of a class of 14 year olds that there are objectively ugly people xxxxxxxxxxxxxxxD
11:24 < bridge> <totar> Ive never tried it tbh I only seen people use it on their own PRs when no one else will review it
11:25 < bridge> <learath2> Humm, maybe I could have worded that better. Idk how to word it. Maybe specificity is more important?
11:25 < bridge> <learath2> Idk, I know I couldn't get it to just give it's general opinion about a pr
11:29 < bridge> <byfox> :justatest:
11:30 < furo> ChillerDragon: You can fix #10854 with `tune explosion_strength 0`
11:30 < chillerbot12> https://github.com/ddnet/ddnet/issues/10854
11:40 < ws-client> **<ChillerDragon>** hmm
11:40 < ws-client> **<ChillerDragon>** is that good style furo?
11:41 < ws-client> **<ChillerDragon>** do vanilla servers even have that ye right?
11:42 < furo> Vanilla does not have it.
11:43 < ws-client> **<ChillerDragon>** Oh i thought ddnet only added like 2 new tunes or sumsin
11:43 < ws-client> **<ChillerDragon>** thats rip then
11:43 < ws-client> **<ChillerDragon>** So it doesnt help with https://github.com/ddnet/ddnet/issues/10854 but with https://github.com/ddnet-insta/ddnet-insta/pull/411 :p 
11:51 < ws-client> **<ChillerDragon>** @byfox let me cook! I am onto something
11:51 < ws-client> **<ChillerDragon>** yikes gotta go in 10mins aaaaaaaaaaa
11:52 < bridge> <byfox> What do you want to cook?
11:52 < ws-client> **<ChillerDragon>** never mind it only made it worse xd
11:53 < ws-client> **<ChillerDragon>** i tried to unset this https://github.com/ddnet/ddnet/blob/bf28bd8201bbc4d6a2cdb6445b070afb7574726d/src/game/server/entities/projectile.cpp#L434
11:53 < ws-client> **<ChillerDragon>** but then the nade still bounces you and goes through walls xd
11:53 < ws-client> **<ChillerDragon>** but there probably is a cleanish way to actually do it
11:53 < ws-client> **<ChillerDragon>** without any downsides
11:57 < ws-client> **<ChillerDragon>** yea watefek non explosive projectiles antiping predicts to go through walls omg
11:58 < ws-client> **<ChillerDragon>** we need proper prediction for the ball bounces i think anything else is cursed
11:58 < ws-client> **<ChillerDragon>** fak
12:02 < bridge> <kebscs> Does the projectile have m_bouncing and still explodes?
12:03 < bridge> <chillerdragon> Ou there is bouncing? Nice
12:04 < bridge> <chillerdragon> I only tried explosive so far
12:04 < bridge> <chillerdragon> Why is there even bouncing? Does ddnet use that somewhere?
12:04 < bridge> <Assa> wasntme 🙁
12:04 < bridge> <teero777> Bouncing bullets
12:04 < bridge> <chillerdragon> Where
12:04 < bridge> <chillerdragon> Since when
12:05 < bridge> <teero777> ??
12:05 < bridge> <Jupstar ✪> I mean generally not related to anything we coded 😄
12:05 < bridge> <chillerdragon> Never seen them
12:05 < bridge> <teero777> Every bullet bounces xd
12:05 < bridge> <chillerdragon> How do I get them
12:05 < bridge> <teero777> ????
12:05 < bridge> <chillerdragon> Off walls?
12:05 < bridge> <teero777> yes
12:05 < bridge> <teero777> you are trolling
12:05 < bridge> <chillerdragon> Your trolling
12:05 < bridge> <chillerdragon> Xd
12:05 < bridge> <teero777> the shotgun bullets
12:05 < bridge> <chillerdragon> Bro whf
12:05 < bridge> <teero777> That freeze you
12:05 < bridge> <chillerdragon> They hit the wall and stop there
12:05 < bridge> <chillerdragon> Oh ddrace
12:05 < bridge> <teero777> And or explode
12:06 < bridge> <teero777> ddrace yes
12:06 < bridge> <chillerdragon> Oh and laser
12:06 < bridge> <chillerdragon> Right
12:06 < bridge> <teero777> laser is not a projectile
12:06 < bridge> <chillerdragon> Right
12:06 < bridge> <chillerdragon> So my troll question made more sense
12:06 < bridge> <chillerdragon> Because real shotgun doesn’t bounce
12:06 < bridge> <chillerdragon> Only laser does
12:06 < bridge> <chillerdragon> And laser is not a projectile ez
12:07 < bridge> <chillerdragon> So ddnet has no bouncing projectiles after all
12:07 < bridge> <teero777> you are working on ddnet code though no?
12:07 < bridge> <chillerdragon> Yes yes shatap xd
12:07 < bridge> <teero777> tf you mean. bullets bounce in ddnet
12:07 < bridge> <chillerdragon> Why would you call a laser a shotgun projectile
12:07 < bridge> <chillerdragon> Oh true
12:07 < bridge> <teero777> ??? It's not
12:07 < bridge> <chillerdragon> Aaaaa
12:07 < bridge> <chillerdragon> Now I know which one you mean
12:07 < bridge> <chillerdragon> The tile not the weapon
12:08 < bridge> <teero777> you dumbass
12:08 < bridge> <chillerdragon> Yes sorry I am new to the game
12:28 < ws-client> **<ChillerDragon>** i only get different variations of bad and ridiculously horrible when playing with bounce and explosive flags
12:57 < bridge> <Jupstar ✪> I think chiller
12:57 < bridge> <Jupstar ✪> we onto smth
12:58 < bridge> <Jupstar ✪> i play the game in 0.25x speed, and it has weird micro stutters when changing directions
12:58 < bridge> <Jupstar ✪> why is this so unsmooth.. why are the physics so unsmooth generally
12:59 < bridge> <Jupstar ✪> and wtf is wrong with the server ghost
12:59 < bridge> <Jupstar ✪> https://cdn.discordapp.com/attachments/293493549758939136/1414203523297447997/ghosty.mp4?ex=68beb72a&is=68bd65aa&hm=11ce2b0dcf39dc88d8d0373eaa90c62666ba6b7d65a46dd5d452de8ce30e605a&
13:00 < bridge> <Jupstar ✪> why does it jump around as if the LAN server is on northpole
13:00 < bridge> <Jupstar ✪> we need physics change
13:54 < bridge> <inv41idu53rn4m3> Isn't that just the relatively harsh quantization?
13:55 < bridge> <inv41idu53rn4m3> imo most noticeable part of it is that all velocities under 1/32 per tick get reduced to 0
13:55 < bridge> <Jupstar ✪> Looks like a rather unsmooth curve tho.
13:55 < bridge> <Jupstar ✪> 
13:55 < bridge> <Jupstar ✪> I'd like to have that plotted
13:55 < bridge> <Jupstar ✪> Interpolated positions during a direction change
13:56 < bridge> <inv41idu53rn4m3> that's 1/1600 tiles per second
13:56 < bridge> <Jupstar ✪> must be smth ugly like dis
13:56 < bridge> <Jupstar ✪> https://cdn.discordapp.com/attachments/293493549758939136/1414217820530802840/image.png?ex=68bec47b&is=68bd72fb&hm=f20cfedbfd7948e111d70f646529c707aa6b5f285c92ea7e82f34447d0e79d77&
13:58 < bridge> <inv41idu53rn4m3> that's 1.5625 tiles per second
13:58 < bridge> <inv41idu53rn4m3> kinda ridiculous actually that 1 tile per second is too slow for the game to handle
15:13 < bridge> <learath2> Hm, is this not exactly how it's meant to be? The fact that the hook applies uneven force causes the changes in slope, can't exactly smooth those out very easily without visually lying a lot
15:13 < bridge> <learath2> Is it also stuttery if you just walk at normal walk speed?
15:13 < bridge> <Jupstar ✪> It is, but it's not very visible
15:14 < bridge> <Jupstar ✪> I mean from tick based games perspective probably
15:14 < bridge> <Jupstar ✪> But real physics would be smooth
15:16 < bridge> <inv41idu53rn4m3> Not only is the slope not smooth, the velocity is limited to a discrete set of values
15:16 < bridge> <learath2> I think it's just the relatively low tickrate, we can never have truly smooth physics but with tinier steps the assumption that the curve is linear is a better and better assumption
15:17 < bridge> <Jupstar ✪> yeah
15:17 < bridge> <Jupstar ✪> trash geme
15:17 < bridge> <inv41idu53rn4m3> We could also use a better integrator
15:17 < bridge> <inv41idu53rn4m3> I assume Teeworlds just uses explicit Euler integration?
15:18 < bridge> <inv41idu53rn4m3> I mean that won't help interpolation but it would help "correctness"
15:18 < bridge> <learath2> We could do a cubic interpolation with one more point
15:18 < bridge> <Jupstar ✪> but this point would probs need come from the past too and that would noticable make the game have even more input lag
15:19 < bridge> <Jupstar ✪> with 100 ticks it would be possible
15:19 < bridge> <learath2> While talking about interpolation I missed my metro stop, now I need to go back
15:19 < bridge> <Jupstar ✪> Tja
15:19 < bridge> <learath2> Yep, but it would be smootherly laggy 😛
15:21 < bridge> <Jupstar ✪> But
15:21 < bridge> <Jupstar ✪> I wonder about demos
15:22 < bridge> <ryozuki> @learath2 do u know what is MSI and MSI-X
15:22 < bridge> <ryozuki> now the MSI brand name makes more sense or maybe its unrelated
15:22 < bridge> <inv41idu53rn4m3> positions being multiples of 1/32 of a tile is fine, but velocities being multiples of 1.5625 tiles per second feels wrong
15:22 < bridge> <ryozuki> https://en.wikipedia.org/wiki/Message_Signaled_Interrupts
15:22 < bridge> <learath2> Only what they are, nothing more
15:23 < bridge> <ryozuki> @learath2 basically allows to have more control over setting up interrupts
15:23 < bridge> <inv41idu53rn4m3> Microsoft Installer
15:23 < bridge> <ryozuki> i think for example in AHCI, u can have like 6 ports, so for example u can setup different interrupts for each port
15:23 < bridge> <learath2> Isn't it just in-band interrupts?
15:23 < bridge> <ryozuki> with traditional interrupts u cant iirc
15:23 < bridge> <ryozuki> u just have 1 ahci interrupt
15:25 < bridge> <ryozuki> > MSI increases the number of interrupts that are possible. While conventional PCI was limited to four interrupts per card (and, because they were shared among all cards, most are using only one), message signalled interrupts allow dozens of interrupts per card, when that is useful
15:25 < bridge> <ryozuki> > There is also a slight performance advantage. In software, a pin-based interrupt could race with a posted write to memory. That is, the PCI device would write data to memory and then send an interrupt to indicate the DMA write was complete. However, a PCI bridge or memory controller might buffer the write in order to not interfere with some other memory use. The interrupt could arrive before the DMA write was complete, and the processor could rea
15:27 < bridge> <ryozuki> MSI (first defined in PCI 2.2) permits a device to allocate 1, 2, 4, 8, 16 or 32 interrupts. 
15:27 < bridge> <ryozuki> MSI-X (first defined in PCI 3.0) permits a device to allocate up to 2048 interrupts.
15:27 < bridge> <ryozuki> > According to a 2009 Intel benchmark using Linux, using MSI reduced the latency of interrupts by a factor of almost three when compared to I/O APIC delivery.
16:07 < bridge> <reitw> hello, is the wiki down ?
16:09 < bridge> <kollpotato> no it was deleted
16:10 < bridge> <kollpotato> some data from 2024 will hopefully be recovered in the next few weeks
16:21 < bridge> <reitw> tf why
16:41 < bridge> <milkeeycat> !wiki
16:41 < chillerbot12> dead, tldr is ryo upgraded debian 13, somehow the mediawiki instance inside docker says db is read only mode, ryo thought backups worked but they didnt for a year cuz a tool renamed from mysqldump to mariadb-dump
16:45 < bridge> <inv41idu53rn4m3> Actually, we have tee velocity, both internal and actual resulting velocity post-quantization, which could be used to refine interpolation
16:47 < bridge> <learath2> Could it really be? What would you use those values for?
16:47 < bridge> <inv41idu53rn4m3> wait, do we or do we not know both v0 and v1?
16:48 < bridge> <learath2> We can know it if it's useful
16:49 < bridge> <learath2> The issue is that I think that information is in reality redundant. The distance between the points already encodes the velocity due to the fixed tickrafe
16:49 < bridge> <learath2> The issue is that I think that information is in reality redundant. The distance between the points already encodes the velocity due to the fixed tickrate
16:49 < bridge> <inv41idu53rn4m3> that means we know the slope at both ends of the interpolation interval
16:49 < bridge> <inv41idu53rn4m3> actually yeah I was thinking actual internal velocity would be useful but now I'm second-guessing myself
16:49 < bridge> <learath2> Ah, I guess that can actually be used, you could have a bezier curve
16:50 < bridge> <inv41idu53rn4m3> actually yeah I was thinking internal velocity would be useful but now I'm second-guessing myself
16:50 < bridge> <inv41idu53rn4m3> now instead of a polyline we get a squiggly line
16:51 < bridge> <learath2> Maybe, I think, could be useful
16:54 < ws-client> **<ChillerDragon>** @Jupstar ✪ that clip is c++ right? not ddnet-rs? Then its the low tick rate i guess.
16:54 < ws-client> **<ChillerDragon>** is it a high bandwidth server?
16:58 < bridge> <learath2> So we are going from t1 to t2, we have x1 and x2, we have v1 and v2. Normally we just do a linterp from x1 to x2. That implies a constant v between t1 and t2, however that creates a discontinuity on the v curve, because we assume a constant v.
16:58 < bridge> <learath2> 
16:58 < bridge> <learath2> So we are looking for a smooth curve v(t) s.t. v(t1) == v1, v(t2) == v2 and \int_{t1}^{t2} v(t) dt == x2 - x1
16:59 < bridge> <Jupstar ✪> I code next level game
17:06 < bridge> <learath2> Don't have pen and paper with me, but you can definitely fit a quadratic there, so maybe we can try one. It would be more expensive to evaluate, but not that bad
17:06 < bridge> <kebscs> maybe it can fix interpolation with tele prediction too
17:07 < bridge> <kebscs> bcs velocity going into tele is usually 0 so it wont be interpolated in the middle between tele and to
17:21 < ChillerDragon> @avolicious sos GER3 little bit laggy omagawd helpo
17:56 < bridge> <aegisub> look up catmull-rom splines
17:56 < bridge> <learath2> Those would work too, aren't they just quadratic too?
17:57 < bridge> <aegisub> actually wait I'm drunk
17:58 < bridge> <aegisub> hermite splines are what your use case is
17:59 < bridge> <learath2> Nope, I misremembered catmull rom splines are cubic, but a quadratic can fit this, so anything higher degree can also be fitted np
18:00 < bridge> <learath2> Also cubic, but yes it might be very easy to fit a hermite spline here
18:00 < bridge> <aegisub> the (cubic) hermite spline would be for x(t) so the velocity would be quadratic, yeah
18:03 < bridge> <learath2> Probably a cubic bezier is what we'd want here. Even easier to control the derivatives at the end and start
18:05 < bridge> <aegisub> in the end they're all equivalent since there's usually a unique cubic satisfying the conditions, but cubic hermite splines are exactly the formulation where you specify the values and first derivatives at every node
18:07 < bridge> <aegisub> as opposed to bezier curves where you specify four control points (which have somewhat direct relations to input/output velocity, but with hermite splines it's a direct one-to-one)
18:07 < bridge> <learath2> Oh, I was misremembering what hermite splines were, yeah, those would be easiest to fit
18:39 < bridge> <ryozuki> ```
18:39 < bridge> <ryozuki> [1.433649] <ahci:k:1> === Testing Read Sectors with GPT Header ===
18:39 < bridge> <ryozuki> Reading GPT header from LBA 1...
18:39 < bridge> <ryozuki> Successfully read GPT header!
18:39 < bridge> <ryozuki> Valid GPT signature found!
18:39 < bridge> <ryozuki> GPT Revision: 0x10000
18:39 < bridge> <ryozuki> Header Size: 92 bytes
18:39 < bridge> <ryozuki> Number of partition entries: 128
18:39 < bridge> <ryozuki> GPT Header (first 64 bytes):
18:39 < bridge> <ryozuki> 0000: 45 46 49 20 50 41 52 54 00 00 01 00 5c 00 00 00 
18:39 < bridge> <ryozuki> 0010: ef ec a9 70 00 00 00 00 01 00 00 00 00 00 00 00 
18:39 < bridge> <ryozuki> 0020: ff ff 1f 00 00 00 00 00 22 00 00 00 00 00 00 00 
18:39 < bridge> <ryozuki> 0030: de ff 1f 00 00 00 00 00 fe 7b 2e 51 dc 37 34 4e 
18:39 < bridge> <ryozuki> 
18:39 < bridge> <ryozuki> 
18:39 < bridge> <ryozuki> Testing multi-sector read (LBA 0-1, 2 sectors)...
18:39 < bridge> <ryozuki> Successfully read 2 sectors!
18:39 < bridge> <ryozuki> MBR signature at end of first sector: 55 aa
18:39 < bridge> <ryozuki> GPT signature in second sector: "EFI PART"
18:40 < bridge> <ryozuki> ```
18:40 < bridge> <ryozuki> im back with ahci, this time way better
18:40 < bridge> <learath2> GPT?????
18:40 < bridge> <learath2> When go back to MBR?
18:40 < bridge> <ryozuki> @learath2 i didnt implement any filesystem
18:40 < bridge> <ryozuki> im just reading the first 2 sectors
18:40 < bridge> <ryozuki> i have no sense of filesystem yet
18:40 < bridge> <ryozuki> just ahci reads
18:40 < bridge> <learath2> GPT sounds too much like ai, we should replace it
18:40 < bridge> <ryozuki> xd
18:41 < bridge> <learath2> What are you planning for fs? Something novel or just fat32?
18:41 < bridge> <ryozuki> i need fat32 for sure, because thats a filesystem i need to support for "installing" the OS itself, and many things like limine use it
18:41 < bridge> <ryozuki> then i plan to make edfs
18:41 < bridge> <learath2> I have had a veeeery long lived goal of making a filesystem suited for storing multimedia
18:42 < bridge> <ryozuki> the boot partition uses fat too
18:42 < bridge> <ryozuki> its a must
18:42 < bridge> <ryozuki> we are bound to M$$ partitions even in computers
18:42 < bridge> <ryozuki> life sucks
18:42 < bridge> <learath2> Btreefs is not a bad point of inspiration
18:42 < bridge> <ryozuki> a fs is a complex btree change my mind
18:42 < bridge> <learath2> Eventhough it's author is an entitled ass
18:43 < bridge> <learath2> Eventhough its author is an entitled ass
18:43 < bridge> <ryozuki> i kinda like fat32 simplicity, its just limited by old time stuff
18:43 < bridge> <ryozuki> like max 10 char filenames
18:43 < bridge> <learath2> Btreefs and btrfs would agree with you
18:43 < bridge> <learath2> And max 4g files
18:43 < bridge> <learath2> exFAT is nice
18:43 < bridge> <ryozuki> (the vfat hack extends filenames but imho its a ugly hack)
18:44 < bridge> <ryozuki> also no notion of user ownership
18:44 < bridge> <ryozuki> edfs v1 will probs be inspired by fat32 but with user and perms and not that limited
18:44 < bridge> <learath2> Another thing I always wanted to do was try implement NTFS. I just find it insane that it takes so many LoC
18:45 < bridge> <ryozuki> why does ntfs take so much?
18:45 < bridge> <ryozuki> @learath2 anyway i want to keep it simple because i also want to get into the world of network driver implementation, and i hear it has a lot of quirks and bugs, will be fun
18:46 < bridge> <learath2> No idea, that's why I wanna try. Every NTFS driver is soooo large 😄
18:46 < bridge> <learath2> For a specific NIC?
18:46 < bridge> <ryozuki> i guess there is no generic nic i think
18:46 < bridge> <learath2> Veery annoying stuff, very annoying to test too, passthrough didn't work great last I tried
18:46 < bridge> <ryozuki> ill search the most common one and the one qem usupports
18:47 < bridge> <ryozuki> (i specifically avoid virtio because i dont want qemu only)
18:47 < bridge> <learath2> I guess you could always do a virtio driver yeah
18:47 < bridge> <learath2> How about a linux driver compatibility layer for edos? Instant support for literally everything
18:48 < bridge> <ryozuki> idk how it would work
18:48 < bridge> <ryozuki> sounds complex
18:48 < bridge> <learath2> Extremely
18:48 < bridge> <ryozuki> also part of fun is making them myself
18:48 < bridge> <ryozuki> im learning a lot
18:48 < bridge> <ryozuki> now i know about fat32 internals
18:48 < bridge> <ryozuki> dir entries, sector chains
18:49 < bridge> <ryozuki> btw bytemuck is a nice crate
18:49 < bridge> <ryozuki> makes working with PODs ez
18:49 < bridge> <learath2> I find old things so soothing. Such elegantly simple solutions
18:49 < bridge> <ryozuki> yeah, optimizing for performance often makes things ugly
18:50 < bridge> <learath2> Sure it wasn't extremely efficient, sure it wasn't without limitations. But it always looked so nice. None of this weird 2050 monkeysass sextuple tree with simd
18:51 < bridge> <ryozuki> also currently i have a terminal and i have a "raw input syscall" but in the future i want to add "device" files that u mmap like linux
18:51 < bridge> <ryozuki> ioctl
19:03 < bridge> <cellegenrih> o7 wiki
19:05 < bridge> <learath2> Are you just mimicing linux for everything? You might find it more fun to experiment, I considered having most of init be part of the kernel in the past
19:05 < bridge> <learath2> Another interesting idea is sockets for everything instead of files for everything
19:10 < bridge> <bingbong669> what's negative stability buffer
19:46 < bridge> <ryozuki> i guess yeah, its just that syscalls for some stuff can be expensive so it makes sense to mmap stuff in user land
19:52 < bridge> <ryozuki> doing it from user is more complex tho
19:53 < bridge> <ryozuki> i think
19:53 < bridge> <ryozuki> i wonder howit would work, what if another program decides to mess up
19:54 < bridge> <ryozuki> also some stuff from kernel threads i can have "atomic" parts bytemporarly disabling interrupts, i wonder how that would work in user
20:02 < bridge> <ryozuki> @learath2 do u have any idea for alternative ways to handle errors in syscalls? by alternative i mean for example an alternative to errno
20:07 < bridge> <learath2> Well idk what you need, but you can technically return whatever you'd like in whatever method you like from there
20:08 < bridge> <learath2> You could even return json if that's what you want
20:08 < bridge> <ryozuki> xd
20:09 < bridge> <ryozuki> returning a single u64 is easy
20:09 < bridge> <ryozuki> if i need to return a json i would need to allocate it for the user, or maybe the user provide a buffer
20:09 < bridge> <ryozuki> but that buffer can fail
20:10 < bridge> <learath2> Don't leave it to the user. You can have a thread local buffer for the error that the kernel manages
20:10 < bridge> <ryozuki> true
20:10 < bridge> <ryozuki> ur rigfht ive been mimicking linux a lot
20:10 < bridge> <ryozuki> even my syscalls have nearly same numbers
20:10 < bridge> <ryozuki> altho i added some syscalls that dont exist
20:11 < bridge> <ryozuki> ```
20:11 < bridge> <ryozuki> const SYS_READ: u64 = 0;
20:11 < bridge> <ryozuki> const SYS_WRITE: u64 = 1;
20:11 < bridge> <ryozuki> const SYS_CLOSE: u64 = 3;
20:11 < bridge> <ryozuki> const SYS_PIPE: u64 = 22;
20:11 < bridge> <ryozuki> const SYS_MMAP: u64 = 9;
20:11 < bridge> <ryozuki> const SYS_MUNMAP: u64 = 11;
20:11 < bridge> <ryozuki> const SYS_EXIT: u64 = 60;
20:11 < bridge> <ryozuki> const SYS_ERRNO: u64 = 0x400;
20:11 < bridge> <ryozuki> const SYS_GETPID: u64 = 39; // get process ID
20:11 < bridge> <ryozuki> const SYS_DRAW_RECT: u64 = 100;
20:11 < bridge> <ryozuki> const SYS_RENDER: u64 = 101;
20:11 < bridge> <ryozuki> const SYS_SCREEN_INFO: u64 = 102;
20:11 < bridge> <ryozuki> const SYS_DRAW: u64 = 103;
20:11 < bridge> <ryozuki> const SYS_RAW_INPUT: u64 = 200;
20:11 < bridge> <ryozuki> ```
20:11 < bridge> <learath2> (Tbh I don't really mind linux, it made some pretty good decisions, but if you are making your own kernel might aswell have fun)
20:11 < bridge> <ryozuki> i have these
20:11 < bridge> <ryozuki> (yes i have syscalls to render)
20:11 < bridge> <ryozuki> xd
20:11 < bridge> <learath2> I would definitely want a kernel level bus in mine e.g.
20:12 < bridge> <learath2> Though I guess people did propose kdbus already so it's not very novel
20:12 < bridge> <aegisub> multimedia 👀
20:12 < bridge> <aegisub> what would such a filesystem entail
20:12 < bridge> <ryozuki> what do u mean by bus?
20:12 < bridge> <ryozuki> like udev?
20:12 < bridge> <learath2> I want the concept of tags to be native to the filesystem
20:12 < bridge> <learath2> Like dbus
20:12 < bridge> <ryozuki> dbus*^
20:12 < bridge> <aegisub> ah
20:13 < bridge> <ryozuki> yeah thats a nice idea
20:13 < bridge> <ryozuki> i kinda have already "mailboxes" for inter thread comms
20:13 < bridge> <ryozuki> but they require known data format
20:13 < bridge> <ryozuki> <T>
20:13 < bridge> <ryozuki> i would need to do one for serializable data
20:15 < bridge> <aegisub> fwiw, it's not directly in the file system, but I was made aware of https://hydrusnetwork.github.io/hydrus/index.html recently and it looks cool
20:15 < bridge> <ryozuki> ```rust
20:15 < bridge> <ryozuki> pub fn sys_keyboard_raw(timeout_milis: u64, scancodes_buffer: *mut u32, size: usize) -> i64 {
20:15 < bridge> <ryozuki>     let thread = sched().current_thread_mut();
20:15 < bridge> <ryozuki>     thread.errno = Errno::Clear;
20:15 < bridge> <ryozuki> 
20:15 < bridge> <ryozuki>     if scancodes_buffer.is_null() {
20:15 < bridge> <ryozuki>         thread.errno = Errno::EINVAL;
20:15 < bridge> <ryozuki>         return -1;
20:15 < bridge> <ryozuki>     }
20:15 < bridge> <ryozuki> 
20:15 < bridge> <ryozuki>     let mut buf = Vec::with_capacity(size);
20:15 < bridge> <ryozuki> 
20:15 < bridge> <ryozuki>     x86_64::instructions::interrupts::enable();
20:15 < bridge> <ryozuki> 
20:15 < bridge> <ryozuki>     let rx = KEYBOARD_BROADCAST.subscribe_or_get();
20:16 < bridge> <ryozuki> 
20:16 < bridge> <ryozuki>     while buf.len() < size
20:16 < bridge> <ryozuki>         && let Some(key) = rx.try_recv()
20:16 < bridge> <ryozuki>     {
20:16 < bridge> <ryozuki>         buf.push(convert_key(key));
20:16 < bridge> <ryozuki>     }
20:16 < bridge> <ryozuki> 
20:16 < bridge> <ryozuki>     if buf.len() < size
20:16 < bridge> <ryozuki>         && let Ok(key) = rx.recv_timeout(Duration::from_millis(timeout_milis))
20:16 < bridge> <ryozuki>     {
20:16 < bridge> <ryozuki>         buf.push(convert_key(key));
20:16 < bridge> <ryozuki>     }
20:16 < bridge> <ryozuki> 
20:16 < bridge> <ryozuki>     unsafe { core::ptr::copy_nonoverlapping(buf.as_ptr(), scancodes_buffer, buf.len()) };
20:16 < bridge> <ryozuki> 
20:16 < bridge> <ryozuki>     buf.len() as i64
20:16 < bridge> <ryozuki> xd
20:16 < bridge> <learath2> Nothing that can't really be implemented on top tbh, but I kinda like the idea of files being accessible as if they are real files but through many paths without having to build all the symlinks by hand
20:17 < bridge> <ryozuki> hm i dont know what tags are
20:17 < bridge> <learath2> Maybe versioning built right into the fs
20:17 < bridge> <ryozuki> http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.60.4187 https://github.com/mayuresh/dhtfs
20:17 < bridge> <learath2> Tags as in "rock", "familyphoto", "nature"
20:18 < bridge> <learath2> Think a heavily indexed flat file store that emulates a folder structure
20:18 < bridge> <learath2> Yeah literally this. Someone stole my idea :pepeW:
20:19 < bridge> <ryozuki> the world is already made
20:19 < bridge> <learath2> 11y ago too, I proabably first had it back then too. I wonder if it is someone that got inspired by me talking about it in many places 😄
20:19 < chillerbot12> Elo @chillerdragon I just wanted to remind you that: fix ddnet++ null GetInt loggedin
21:35 < ws-client> **<ChillerDragon>** imagine if there was an emote for lags 
21:35 < ws-client> **<ChillerDragon>** like a broken signal thingy
21:35 < ws-client> **<ChillerDragon>** everyone would spam during ddos xd
21:42 < bridge> <cellegenrih> As if that would be a normal thing to experience
21:42 < bridge> <cellegenrih> good idea
21:45 < ws-client> **<ChillerDragon>** xd
21:45 < ws-client> **<ChillerDragon>** plenty of laggers and high pingers would love it
21:46 < ws-client> **<ChillerDragon>** trust
21:46 < ws-client> **<ChillerDragon>** u know that `bind x say lag` is a thing right?
22:17 < bridge> <Ewan> yea
22:35 < bridge> <pathos3005> @chillerdragon another big bug, v1,7 every time someone joins
22:35 < bridge> <pathos3005> https://cdn.discordapp.com/attachments/293493549758939136/1414348305768386590/image.png?ex=68bf3e01&is=68bdec81&hm=8295adbb011a590b1989a687810e6c91877c94ab32217b8a8b0019611e1cb64f&
22:52 < bridge> <chillerdragon> What’s the bug? @pathos3005 also holy shit SP someone is still alive?
22:54 < bridge> <chillerdragon> Oh you mean they didn’t send the message? Yea nobody ever had that bug that’s quite suspicious. But sorry I don’t take bug reports for old versions @pathos3005 :p let’s fix the new bugs together instead
22:54 < bridge> <blazulite> le chat
22:54 < bridge> <blazulite> https://cdn.discordapp.com/attachments/293493549758939136/1414353141213691985/image.png?ex=68bf4282&is=68bdf102&hm=060575cf44c4a8b3b28cb393606bce9beae53c2dc94c572d201bba31ad1e8e81&
22:57 < bridge> <chillerdragon> Why is it that you are the only one that has weird bugs nobody else has? @pathos3005 is there anything unusual about your setup?
22:58 < bridge> <Solly (bun bun)> like the current chat is very ... bulky
22:59 < bridge> <Solly (bun bun)> works on wide res
23:07 < bridge> <blazulite> huh?
23:09 < bridge> <blazulite> idrk what you mean with "bulky"
23:10 < bridge> <blazulite> idrk what you mean by "bulky"
23:10 < bridge> <kebscs> i like the new function solly added, i could make the background invisible
23:10 < bridge> <kebscs> https://cdn.discordapp.com/attachments/293493549758939136/1414357245831020714/image.png?ex=68bf4654&is=68bdf4d4&hm=e2045024130140baaddd99c232231f1065dfb8ff3ea2294ddfdb9d6f91ecb20b&
23:11 < bridge> <kebscs> kind of league of legends chat feel
23:11 < bridge> <kebscs> no bloat
23:12 < bridge> <Solly (bun bun)> lmao i use it the opposite way
23:12 < bridge> <Solly (bun bun)> i cant see the text xd
23:13 < bridge> <Solly (bun bun)> everythings big
23:13 < bridge> <Solly (bun bun)> better than the current ddnet-rs chat
23:15 < bridge> <blazulite> it's quite quite smaller than the current ddnet-rs' though
23:15 < bridge> <Solly (bun bun)> yar but thats not the baseline
23:16 < bridge> <blazulite> btw
23:16 < bridge> <blazulite> working on a "classic mode" chat style
23:16 < bridge> <blazulite> cuz I'm pretty sure the so called og players wouldn't like this discordish style
23:16 < bridge> <blazulite> https://tenor.com/view/kitty-cute-yarn-gif-13128288
23:16 < bridge> <blazulite> cuz I'm pretty sure the so called og players wouldn't like this discord-styled chat
23:22 < bridge> <Solly (bun bun)> https://tenor.com/view/done-overit-giveup-fall-fail-gif-7356905
23:23 < bridge> <cellegenrih> he never left and he's still trolling
23:24 < bridge> <cellegenrih> maybe it's too large, imagine this chat rapidly going by with 64 players
23:25 < bridge> <cellegenrih> but regardless, I like this chat look. I hope it's as easy to interact with it while on the menu
23:27 < bridge> <blazulite> https://cdn.discordapp.com/attachments/293493549758939136/1414361442626572400/image.png?ex=68bf4a3d&is=68bdf8bd&hm=f0cf6284b6c8198967ce444a1fa8b305e33ea29e930c4f2971fbff4887429a29&
23:27 < bridge> <blazulite> Timestamps will still appear but only when you hover on the messages with the chat open
23:44 < bridge> <cellegenrih> could you make the player chats' margin lower, while the margin for the command chat the same?
23:45 < bridge> <cellegenrih> same with the playerboard on the bottom right, a bit closer together
23:47 < bridge> <blazulite> Like this?
23:47 < bridge> <blazulite> https://cdn.discordapp.com/attachments/293493549758939136/1414366587309391892/image.png?ex=68bf4f08&is=68bdfd88&hm=544ecc01ab47316a37a126f231d554a7f1f2ebd2422a02121124d7b856882d1d&
23:49 < bridge> <cellegenrih> I think yeah, I kinda see the messages being closer, perhaps a bit more closer. But try to make the command messages further away from the player messages if able
23:49 < bridge> <cellegenrih> I think yeah, I kinda see the messages being closer, perhaps move it a bit more closer. But try to make the command messages further away from the player messages if able
23:49 < bridge> <blazulite> oh sure
23:49 < bridge> <cellegenrih> the point is to make the command messages and the player messages separate for readability, while making more space for them
23:50 < bridge> <cellegenrih> also based message example
23:50 < bridge> <blazulite> > perhaps move it a bit more closer
23:50 < bridge> <blazulite> oh jeez even closer?
23:50 < bridge> <blazulite> DD:
23:50 < bridge> <cellegenrih> I mean, what like 2 pixel - 4 pixel separation?
23:50 < bridge> <blazulite> the 10 pixels gap was precisely to improve readability
23:50 < bridge> <blazulite> I reduced it to 5 in this example
23:51 < bridge> <cellegenrih> imo the sweet spot is inbetween 2 and 4
23:51 < bridge> <blazulite> 2 is way too much DDDDDDDDDDDDDDDDDDD:
23:51 < bridge> <blazulite> https://cdn.discordapp.com/attachments/293493549758939136/1414367429710319656/image.png?ex=68bf4fd0&is=68bdfe50&hm=fe0b704a6a666e86b839343871c00d8c43ccb399923c0e59680e9e2680175a09&
23:51 < bridge> <cellegenrih> well this is what I wanted so XD
23:51 < bridge> <cellegenrih> like, show the full screenshot
23:52 < bridge> <blazulite> 4 is okay
23:52 < bridge> <blazulite> https://cdn.discordapp.com/attachments/293493549758939136/1414367632257187840/image.png?ex=68bf5001&is=68bdfe81&hm=4aa223d6b491f76fa30158e1bcc7d6b9b7544e1ea05807c1b80097318cde319a&
23:52 < bridge> <cellegenrih> kinda, but see how the letters actually get deformed a bit?
23:52 < bridge> <blazulite> how?
23:52 < bridge> <cellegenrih> set it to 8,
23:52 < bridge> <cellegenrih> divisible by 4
23:52 < bridge> <cellegenrih> letters tend to follow this rule
23:53 < bridge> <blazulite> I don't get what you mean by this
23:53 < bridge> <blazulite> nothing got deformed at all
23:53 < bridge> <cellegenrih> being a nerd at image compression algorithms, so I recognize the outlines being off by a pixel
23:53 < bridge> <blazulite> 8 is honestly perfect
23:53 < bridge> <blazulite> https://cdn.discordapp.com/attachments/293493549758939136/1414368022524854313/image.png?ex=68bf505e&is=68bdfede&hm=bd9962977aa8da736554b9f1356918b307e4dfd62b5d4856ccdf587376d69a25&
23:53 < bridge> <cellegenrih> don't ask how I know that
23:53 < bridge> <cellegenrih> yeah, there you go
23:53 < bridge> <cellegenrih> this fine by me
23:54 < bridge> <blazulite> ...ok?
23:54 < bridge> <blazulite> I don't know how is that relevant with the conversation honestly LOL
23:54 < bridge> <blazulite> I don't know how is that relevant to the conversation honestly LOL
23:54 < bridge> <cellegenrih> Anyways
23:55 < bridge> <cellegenrih> about the player ranks on the bottom right, a bit closer?
23:55 < bridge> <blazulite> current padding for it is 20
23:55 < bridge> <cellegenrih> set it to around 15, but if it's too close, then around 18
23:56 < bridge> <blazulite> there we go, 8 for you close gaps enjoyer
23:56 < bridge> <blazulite> https://cdn.discordapp.com/attachments/293493549758939136/1414368666530746431/image.png?ex=68bf50f7&is=68bdff77&hm=8a24ca2108c0368259bfba4b36ae1986d2c5a0114501d15ca7487c558dc48bd5&
23:56 < bridge> <blazulite> alr
23:56 < bridge> <cellegenrih> xd,
23:56 < bridge> <cellegenrih> before vs after
23:56 < bridge> <cellegenrih> https://cdn.discordapp.com/attachments/293493549758939136/1414368843329044611/image.png?ex=68bf5121&is=68bdffa1&hm=c6d27478eb1ea94b592359c48221f8cf1b2d39bfab0aff0ee20bfbc45b0181c6&
23:56 < bridge> <cellegenrih> https://cdn.discordapp.com/attachments/293493549758939136/1414368843794747433/image2.png?ex=68bf5122&is=68bdffa2&hm=34583b0ad7a6e439e6b1c3282a27f998db286b7aeb6bc1645ff008c1c3456573&
23:56 < bridge> <blazulite> jokes aside 8 really really fits
23:57 < bridge> <cellegenrih> imo the after is good
23:57 < bridge> <blazulite> yeah
23:57 < bridge> <cellegenrih> top left panel looks a bit weird, but idk how you wanna deal with it
23:57 < bridge> <blazulite> idk either if I'm honest 💯
23:58 < bridge> <cellegenrih> maybe add like a dropdown which will contain the other UI stuff like doublejump, deep, etc?
23:58 < bridge> <cellegenrih> and only show above the dropdown the weaons you can have?
23:58 < bridge> <blazulite> it looks super strange without the bg
23:58 < bridge> <blazulite> https://cdn.discordapp.com/attachments/293493549758939136/1414369243587412089/image.png?ex=68bf5181&is=68be0001&hm=e4b60c5fb5e472ad36e949c8d1734cac86a1c3763be1883135af28ec65d9df0d&
23:58 < bridge> <cellegenrih> yeah, the bg is necessary
23:59 < bridge> <cellegenrih> try it with a dropdown if you could, no need for animations currently, just only to hide the UI elements below
23:59 < bridge> <soulyvevo> eh idk i think this part would be nice if it was the same as in regular ddnet