11:44 < bridge> <b l u e  b e n c i e> why does Visual Studio ship python 3.9.13 :sadKEK:
11:45 < bridge> <b l u e  b e n c i e> why does Visual Studio ship python 3.9.13 :sadKEK:
11:45 < bridge> <b l u e  b e n c i e> i mean they do state its unsupported, but.. why arent they just shipping a newer version then °-°
11:47 < bridge> <b l u e  b e n c i e> why does Visual Studio ship python 3.9.13 :sadKEK:
11:47 < bridge> <b l u e  b e n c i e> i mean they do state its unsupported, but.. why arent they just shipping a newer version then
12:42 < bridge> <cellegenrih> testing on production :troll:
13:30 < bridge> <milkeeycat> does anyone know if all participants in a pr get an email when someone requests a review from 1 person?
13:44 < bridge> <Jupstar ✪> i think only affected ppl
14:01 < bridge> <Assa> you mean a notification
15:52 < bridge> <Solly (bun bun)> github emails me for everything :D
15:52 < bridge> <Solly (bun bun)> https://cdn.discordapp.com/attachments/293493549758939136/1429104850016407702/image.png?ex=68f4ed1b&is=68f39b9b&hm=e78f9deef7691e47df1a57985b6e3905d37a20f26ef08d46ab1acc90910843de&
15:52 < bridge> <Solly (bun bun)> tclient build is failing ):
15:58 < bridge> <Assa> yes but a lot of people (including me) disable email options, but still keep the notifications
15:58 < bridge> <Assa> because I already get thousands on emails on my trash email account, too many on my google account and a lot on my buisness mail
15:58 < bridge> <Assa> because I already get thousands of emails on my trash email account, too many on my google account and a lot on my buisness mail
15:59 < bridge> <Assa> I need to eat something today <.< I confused you with Milkey cat
15:59 < bridge> <Solly (bun bun)> ... dont eat me!
16:00 < bridge> <Assa> I like bunnies :3
16:00 < bridge> <Solly (bun bun)> nooo
16:00 < bridge> <Solly (bun bun)> we are so small its more effort to eat me than its worth
16:00 < bridge> <Solly (bun bun)> im like all fur and bones
16:00 < bridge> <Solly (bun bun)> plz
16:00 < bridge> <Assa> These are my Ubuntu igpu benchmarks, I find it really interesting, that I don't see the regression and that Abyss is sooo bad here
16:00 < bridge> <Assa> https://cdn.discordapp.com/attachments/293493549758939136/1429106920404287549/DDnet-benchmarks-Ubuntu.png?ex=68f4ef09&is=68f39d89&hm=148231f22bb0d7ed222b77e5100de48806febc8d70b4a4ea416a1d6a78b819c2&
16:02 < bridge> <Assa> I didn't bother to add version 15.x and 16.0 was not collecting properly for some reason
16:39 < bridge> <Jupstar ✪> 270fps? what kind of shit laptop is this
16:39 < bridge> <Jupstar ✪> or is the driver buggy
16:39 < bridge> <Jupstar ✪> does it have vk support?
16:48 < bridge> <Teero> Chiller goated commenter xD
16:48 < bridge> <Teero> https://cdn.discordapp.com/attachments/293493549758939136/1429118893942771872/image0.png?ex=68f4fa30&is=68f3a8b0&hm=f492c08f24c71947b4591d4269d37825ac4746dbcf17655c7200de27e05386f4&
16:48 < ws-client1> **<ChillerDragon>** wot stats is dat @Teero 
16:48 < bridge> <Teero> my channel stats
16:49 < ws-client1> **<ChillerDragon>** o
16:49 < ws-client1> **<ChillerDragon>** i fan #1 u know
16:50 < bridge> <Assa> it has vk support, but it's not a special laptop
16:52 < bridge> <Assa> it even struggle with fullscreen, I just did the benchmarks in a smaller window
16:56 < bridge> <remakepower> I get build failed every morning since I forked a repo that has a scheduled action.
16:59 < bridge> <remakepower> When I push a commit to a branch related to a Teeworlds PR, I receive 4 emails about build failures. Two are for my branch, and two are for the PR.
17:00 < bridge> <remakepower> :nouis:
17:01 < bridge> <Assa> dell laptop,
17:01 < bridge> <Assa> dell laptop,
17:01 < bridge> <Assa> - 11th Gen Intel® Core™ i5-1145G7 × 8
17:02 < bridge> <Assa> - Intel® Iris® Xe Graphics (TGL GT2)
17:02 < bridge> <Assa> 
17:02 < bridge> <Assa> Is this even an igpu?
17:02 < bridge> <remakepower> That
17:02 < bridge> <remakepower> made me suprised.
17:02 < bridge> <inv41idu53rn4m3> That's like a new iGPU as well, I'm a bit surprised
17:03 < bridge> <Jupstar ✪> yeah but honestly, this must be some weird power saving thing
17:03 < bridge> <Jupstar ✪> it's almost twice as slow as my CPU
17:03 < bridge> <Jupstar ✪> and i really mean CPU
17:03 < bridge> <remakepower> As I thought my old laptop was worst laptop that still was used today before.
17:03 < bridge> <Assa> it was on my docking station (so not beeing in battery safe mode or something)
17:03 < bridge> <inv41idu53rn4m3> Then again, no matter which iGPU, you're going to have horrible memory bandwidth compared to anything discrete
17:03 < bridge> <Jupstar ✪> i think even my ivybridge laptop has more fps
17:04 < bridge> <Jupstar ✪> intels drivers are probs a mess of buggy weird decisions to make it looks like intel cpus are power saving or some shit
17:04 < bridge> <Jupstar ✪> 270fps on linear is really nothing
17:04 < bridge> <Jupstar ✪> i can draw it by hand
17:05 < bridge> <Assa> xD
17:05 < bridge> <Jupstar ✪> lavapipe
17:05 < bridge> <Jupstar ✪> https://cdn.discordapp.com/attachments/293493549758939136/1429123158149169172/grafik.png?ex=68f4fe28&is=68f3aca8&hm=8cafa0b92c745e63231bc4d40f0518f619a041f5d2042bfb8e70a4c3fddd59c0&
17:05 < bridge> <Jupstar ✪> so fully on CPU
17:05 < bridge> <Jupstar ✪> no hardware acceleration
17:05 < bridge> <remakepower> Same fps on my old laptop
17:06 < bridge> <Jupstar ✪> only case i know that has low fps is nvidia + intel
17:06 < bridge> <Jupstar ✪> bcs of nvidia optimimus
17:06 < bridge> <remakepower> No
17:06 < bridge> <Jupstar ✪> this might be an ass of a pc, which is arguably a waste of money and resources
17:06 < bridge> <Jupstar ✪> but no way it only has 270 fps
17:07 < bridge> <remakepower> No.
17:07 < bridge> <remakepower> You can set the graphics mode to use only nvidia to render better.
17:08 < bridge> <Jupstar ✪> bcs of nvidia optimus
17:09 < bridge> <remakepower> At least, DDNet could run with 1000 fps on my old laptop.(nvidia 940M)
17:09 < bridge> <inv41idu53rn4m3> This has specifically been an issue for me on X11
17:09 < bridge> <Jupstar ✪> yes
17:09 < bridge> <Jupstar ✪> bcs it's a trash potato game
17:10 < bridge> <Jupstar ✪> if it runs below 600fps, it's 100% some scam from intel or nvidia and not bcs their hardware is too bad
17:10 < bridge> <inv41idu53rn4m3> On my laptop I go from 300fps to 3000fps when I switch to Wayland
17:10 < bridge> <Jupstar ✪> some software scam*
17:10 < bridge> <Jupstar ✪> (which ofc can be bcs monitor display is not connected to either of them)
17:10 < bridge> <Jupstar ✪> sick
17:11 < bridge> <inv41idu53rn4m3> I just haven't bothered to set up my desktop stuff for Wayland so I'm still on X11
17:12 < bridge> <Assa> I error proofed, this are my acutal fps
17:12 < bridge> <Assa> https://cdn.discordapp.com/attachments/293493549758939136/1429125055706697839/screenshot_2025-10-18_17-12-14.png?ex=68f4ffed&is=68f3ae6d&hm=4de79af52ab633e8cb85c93669507961130de96109df456c0b1eb0f619d056e5&
17:13 < bridge> <Assa> idk maybe something is missconfigured, or intel drivers are a mess
17:13 < bridge> <Jupstar ✪> ok it's wqhd?
17:13 < bridge> <Jupstar ✪> what kind of laptop is that lol
17:13 < bridge> <Jupstar ✪> why do they put trash into a high res display
17:14 < bridge> <inv41idu53rn4m3> At normal zoom my phone gets 500 fps on Linear
17:14 < bridge> <Assa> it's connected to my dockingstation which in turn is connected to my monitor, which is wqhd
17:14 < bridge> <Jupstar ✪> but yeah even at wqhd this is a scam
17:14 < bridge> <Assa> I didn't buy it ^^
17:15 < bridge> <Assa> I just use it normally as a glorified terminal
17:17 < bridge> <Assa> maybe it makes sense for you now, why I am so eager to make fps high xD
17:17 < bridge> <inv41idu53rn4m3> Benchmarks on my 1.83GHz Intel Celeron next? :D
17:17 < bridge> <inv41idu53rn4m3> I'd need to get that into a functioning computer first...
17:18 < bridge> <Assa> give me a second, going to make an update to the repo
17:18 < bridge> <inv41idu53rn4m3> (it's a laptop motherboard separate from all the laptop bits)
17:18 < bridge> <Jupstar ✪> switch to wayland
17:18 < bridge> <Jupstar ✪> i want to know if that can really fix smth
17:18 < bridge> <Jupstar ✪> but honestly.. best would be to benchmark offscreen
17:18 < bridge> <Jupstar ✪> i dunno what kind of bad integrations they can have
17:19 < bridge> <Assa> isn't ubuntu 24 on wayland natively? :justatest:
17:19 < bridge> <Assa> (clueless)
17:21 < bridge> <inv41idu53rn4m3> Problem is it's a bit hard to tell if an external tool does the job correctly, and I'm not sure if we want to edit every DDNet version to support off screen rendering...
17:22 < bridge> <inv41idu53rn4m3> no harm in trying tho!
17:23 < bridge> <Jupstar ✪> no i am just interested in if a 100$ phone is really better than intel's midclass laptops
17:23 < bridge> <Jupstar ✪> I just want to know how hard customers can be scammed
17:23 < bridge> <Jupstar ✪> For my future monopolistic company
17:25 < bridge> <Assa> @inv41idu53rn4m3 the repo should be ready now, clone with git, then just run ./create-benchmarks.sh
17:25 < bridge> <Assa> **this may take an hour or two**
17:25 < bridge> <Assa> so you maybe want to configure client_list.json before
17:30 < bridge> <Assa> for some reason my account was set to ubuntu-on-xorg :pepeRage:
17:30 < bridge> <Assa> https://cdn.discordapp.com/attachments/293493549758939136/1429129614139064481/screenshot_2025-10-18_17-28-27.png?ex=68f5042b&is=68f3b2ab&hm=639528f504527055a2e1c0abdd713884491d8a0c85e0faf0c4b27adceb9addff&
17:31 < bridge> <Assa> thanks @jupeyy_keks :deen_star:
17:33 < bridge> <Jupstar ✪> wat
17:33 < bridge> <Jupstar ✪> wayland really fixed it?
17:33 < bridge> <Jupstar ✪> sick
17:33 < bridge> <Assa> that's also a first for me
17:33 < bridge> <Jupstar ✪> rare wayland W
17:36 < bridge> <Assa> suddenly I can play this on fullscreen as well, without the laptop screaming at me
17:36 < bridge> <Assa> like its unbelievable how much this increased performance
17:38 < bridge> <Jupstar ✪> 🧑‍🚀
17:38 < bridge> <manikoo> is your benchmark thing open source?
17:39 < bridge> <manikoo> or you do it yourself ingame
17:39 < bridge> <Assa> <https://github.com/AssassinTee/ddnet-benchmarks>
17:39 < bridge> <Assa> 
17:39 < bridge> <Assa> Fully automated
17:39 < bridge> <manikoo> 👍
17:40 < bridge> <Assa> please read the readme before running ./create-benchmarks.sh, as you may not want to benchmark the last 5 years of ddnet
17:40 < bridge> <manikoo> :greenthing:
17:41 < bridge> <inv41idu53rn4m3> Mine is a $300 phone :)
17:41 < bridge> <inv41idu53rn4m3> ...from 2018
17:41 < bridge> <inv41idu53rn4m3> though it was a "flagship killer" at the time, so it has the best SoC that was available back then
17:43 < bridge> <Jupstar ✪> https://cdn.discordapp.com/attachments/293493549758939136/1429132874816159865/Screenshot_20251018-174323_DDNet-19.5.png?ex=68f50735&is=68f3b5b5&hm=0e4291239af309452f01ed8a6c8de6b91b1849f8c084216f471545a8c6f0af35&
17:44 < bridge> <Jupstar ✪> poco f5
17:44 < bridge> <inv41idu53rn4m3> bro how does my poco F1 outperform that
17:45 < bridge> <inv41idu53rn4m3> oh yours is a higher resolution ig
17:45 < bridge> <Jupstar ✪> which resolution?
17:45 < bridge> <inv41idu53rn4m3> 1080 x 2246
17:45 < bridge> <inv41idu53rn4m3> wacko
17:45 < bridge> <Jupstar ✪> not much worse than mine
17:45 < bridge> <Assa> still more than the laptop
17:45 < bridge> <Assa> https://cdn.discordapp.com/attachments/293493549758939136/1429133388748689658/Screenshot_20251018-174526.png?ex=68f507af&is=68f3b62f&hm=575ffba78229c2c0fcbf476e4b6a6f04bcfa2b379bdf746f6728a7f403981315&
17:45 < bridge> <Jupstar ✪> sick
17:46 < bridge> <Jupstar ✪> rich boi phone
17:46 < bridge> <Assa> was exactly 500€ when I bought it
17:46 < bridge> <Assa> current flagships are 1K +
17:46 < bridge> <Teero> Idk it's too hard to write cl_show_fps 1 into f1 on phone
17:47 < bridge> <Assa> Settings->Graphics()->ShowFPs[x]
17:47 < bridge> <inv41idu53rn4m3> mine was like 400€ after retailer markup and VAT
17:47 < bridge> <Assa> I wrote too much c++ it seems
17:47 < bridge> <Jupstar ✪> https://cdn.discordapp.com/attachments/293493549758939136/1429133758690492487/Screenshot_20251018-174710_Termux.png?ex=68f50808&is=68f3b688&hm=74cf10be1bc5ddd9b6ed01b75934721ffc0c4b5853d26c18e5ead8a30b965fd7&
17:47 < bridge> <inv41idu53rn4m3> oh I should probably update my phone's ddnet version to get a good comparison...
17:48 < bridge> <inv41idu53rn4m3> maybe there's a regression on latest
17:48 < bridge> <Jupstar ✪> show screenshot that u have more fps first
17:48 < bridge> <Assa> mine was on 19.4
17:48 < bridge> <Jupstar ✪> mine is newest
17:48 < bridge> <Jupstar ✪> with vullkan
17:49 < bridge> <Jupstar ✪> with vulkan
17:49 < bridge> <Assa> and mine also doesn't show me an update button
17:49 < bridge> <Jupstar ✪> there is none
17:49 < bridge> <Assa> omg I was on GLES 3.0 on phone
17:49 < bridge> <Teero> I get around 500. But making a screenshot bugs the client xd
17:49 < bridge> <Teero> https://cdn.discordapp.com/attachments/293493549758939136/1429134382475776091/Screenshot_20251018-174922.png?ex=68f5089c&is=68f3b71c&hm=ccfe113a2246f23fd970c535db8805021a612e5d04e0e1b696d91ee2ba26e120&
17:49 < bridge> <Teero> https://cdn.discordapp.com/attachments/293493549758939136/1429134382836350997/Screenshot_20251018-174903.png?ex=68f5089c&is=68f3b71c&hm=941109a4bfdf476163de44461a05c4c68487940214a06210b680ee9db9567171&
17:49 < bridge> <Teero> Same
17:50 < bridge> <Jupstar ✪> sick
17:50 < bridge> <Assa> Vulkan has actually **less** fps
17:50 < bridge> <Jupstar ✪> teero has faster phone than me
17:50 < bridge> <Jupstar ✪> holy moly
17:50 < bridge> <Jupstar ✪> unexpected
17:50 < bridge> <Jupstar ✪> that is how teero must feel all the time
17:50 < bridge> <inv41idu53rn4m3> uhhh
17:50 < bridge> <inv41idu53rn4m3> is my ancient version of mobile discord just gonna refuse to send it??
17:50 < bridge> <Teero> With vulkan I get ~600 fps
17:51 < bridge> <Jupstar ✪> which phne
17:51 < bridge> <Jupstar ✪> teero
17:51 < bridge> <Teero> nothing phone 1
17:51 < bridge> <Jupstar ✪> how expensive was that
17:51 < bridge> <inv41idu53rn4m3> 18.8.2
17:51 < bridge> <inv41idu53rn4m3> https://cdn.discordapp.com/attachments/293493549758939136/1429134844281225347/Screenshot_2025-10-18-18-49-02-548_org.ddnet.client.jpg?ex=68f5090a&is=68f3b78a&hm=9b123e43b650c3c1b038adcef502f187714b96c1a45814247912f300529c0903&
17:51 < bridge> <Jupstar ✪> that is on poco f1?
17:52 < bridge> <Jupstar ✪> i mean the cut out is not small, but other than that sick
17:52 < bridge> <Teero> Idk I got it for 150€ maybe? Got it as a present
17:52 < bridge> <Jupstar ✪> k but on amazon it's more like 450€
17:53 < bridge> <Teero> yea don't look on Amazon lmao
17:53 < bridge> <inv41idu53rn4m3> oh and this is Vulkan
17:53 < bridge> <Jupstar ✪> oke
17:53 < bridge> <Jupstar ✪> than rip me
17:53 < bridge> <Jupstar ✪> got scammed by xiamo
17:53 < bridge> <Jupstar ✪> i
17:54 < bridge> <Teero> https://cdn.discordapp.com/attachments/293493549758939136/1429135437867516024/Screenshot_20251018-175350.png?ex=68f50998&is=68f3b818&hm=6904ca204df193a8bdfe422101c0ffe09b4e8b0fb5c2687bec7da1f6068ae108&
17:54 < bridge> <inv41idu53rn4m3> I thought F5 was like a high performance phone??
17:54 < bridge> <Teero> But I got black
17:54 < bridge> <inv41idu53rn4m3> Or was that only like the F5 pro or something stupid
17:54 < bridge> <Jupstar ✪> i dunno, it's midclass
17:54 < bridge> <Jupstar ✪> but 420€ or smth
17:54 < bridge> <Jupstar ✪> so i epxected it to outperform a 150€ one xD
17:54 < bridge> <Teero> F5 really changed my perspective
17:54 < bridge> <Teero> (minecraft)
17:55 < bridge> <Teero> @jupeyy_keks does your phone have a case?
17:56 < bridge> <Jupstar ✪> yes
17:56 < bridge> <Teero> Idk it feels impossible to even break my phone. It doesn't have any protection and I've dropped it so many times and nothing ever happened
17:56 < bridge> <inv41idu53rn4m3> oh right, my performance would probably drop a lot after ten minutes due to it heating up
17:56 < bridge> <inv41idu53rn4m3> and cases aren't very good for heating either
17:56 < bridge> <inv41idu53rn4m3> and cases aren't very good for cooling either
17:57 < bridge> <Assa> you can put a cooling pad out of the fridge on the back of your phone :kek:
17:57 < bridge> <Teero> XD
17:57 < bridge> <inv41idu53rn4m3> for optimal performance I would put the cooling pad on the screen :kek:
17:57 < bridge> <Assa> or do your benchmarks in a cooling chamber
17:59 < bridge> <manikoo> no man it's too hard for me
17:59 < bridge> <manikoo> as i got i need linux to run it
17:59 < bridge> <Teero> reading is hard
17:59 < bridge> <Assa> you don't need linux to run it
17:59 < bridge> <inv41idu53rn4m3> ayo what
17:59 < bridge> <inv41idu53rn4m3> "there was a problem parsing the package"
17:59 < bridge> <inv41idu53rn4m3> I just got the latest version from ddnet.org/downloads
17:59 < bridge> <inv41idu53rn4m3> and my phone doesn't want to install it
18:00 < bridge> <Assa> hmm maybe I didn't hink of it, I installed git bash and can just run normal bash scripts on windows
18:01 < bridge> <manikoo> i mean yes but i run ./create-benchmarks.sh and it opens in vscode or notebook to edit 😭 
18:01 < bridge> <manikoo> im too stoopid
18:01 < bridge> <manikoo> i mean yes but i run ```./create-benchmarks.sh``` and it opens in vscode or notebook to edit 😭 
18:01 < bridge> <manikoo> im too stoopid
18:01 < bridge> <Teero> do you have bash? xD
18:01 < bridge> <manikoo> no
18:01 < bridge> <Teero> use wsl
18:01 < bridge> <Assa> @manikoo you can do all the steps in the script manually, like create the directories, run the python scripts one by one
18:02 < bridge> <Assa> and the scripts are save to run multiple times
18:02 < bridge> <manikoo> oo ok
18:03 < bridge> <inv41idu53rn4m3> I wonder, are those scripts python 2 compatible?
18:03 < bridge> <Assa> no
18:03 < bridge> <Assa> python 2 is EOL since 5 fucking years
18:03 < bridge> <inv41idu53rn4m3> yeah but many systems STILL want you to write python3
18:03 < bridge> <inv41idu53rn4m3> :greenthing:
18:03 < bridge> <inv41idu53rn4m3> I got hit by this recently
18:04 < bridge> <Assa> there is a package named python-is-python3 or something
18:04 < bridge> <Assa> what bitch ass os did you install?
18:04 < bridge> <Assa> I only know of one instance doing this, like debian 7 or 8 was it
18:04 < bridge> <inv41idu53rn4m3> see I didn't install it is the issue
18:04 < bridge> <inv41idu53rn4m3> maybe "many" is an overstatement
18:05 < bridge> <inv41idu53rn4m3> I'm just living in the past
18:05 < bridge> <inv41idu53rn4m3> the most recent installment of me being annoyed by this was on a school system
18:06 < bridge> <Assa> I mean yeah we could update the shell script <.<
18:06 < bridge> <inv41idu53rn4m3> Ubuntu 24.04 apparently
18:07 < bridge> <Assa> I am pretty sure ubuntu 24 does not do that natively like this
18:08 < bridge> <inv41idu53rn4m3> lol create-benchmarks.sh is so simple that if you just removed `set -e` which isn't even needed anyways you'd be Windows compatible
18:08 < bridge> <Assa> really? xD
18:08 < bridge> <inv41idu53rn4m3> yeah our school has a lot of custom stuff going on
18:09 < bridge> <inv41idu53rn4m3> well Windows does have echo, and the script does nothing but echo and run python
18:09 < bridge> <robyt3> Are you using storage.cfg to use a default config for each version? Do you set the backend manually? Otherwise the results depend on the default backend in each version.
18:09 < bridge> <inv41idu53rn4m3> you might want to have @echo off at the start
18:09 < bridge> <robyt3> Or use a python script overall
18:09 < bridge> <Assa> I checked this, as long as I run any vulkan capable version, it will run vulkan. So as long as I am at >= 16.0 the backend doesn't switch
18:09 < bridge> <Assa> at one point it switches from vulkan 1.0.0 to 1.1.0
18:10 < bridge> <Jupstar ✪> yeah
18:10 < bridge> <Jupstar ✪> obs cannot record vk 1.0 apps
18:10 < bridge> <Jupstar ✪> lol
18:11 < bridge> <Assa> :justatest:  if you take a look at the output of the python scripts, you won't care about a few echos anymore
18:12 < bridge> <inv41idu53rn4m3> `@echo off` makes it so that it doesn't print the commands, only their outputs, but yeah I believe you
18:12 < bridge> <inv41idu53rn4m3> When looking at them yesterday I was a bit worried that it might mess with my main ddnet config
18:12 < bridge> <inv41idu53rn4m3> a storage.cfg could be nice
18:12 < bridge> <manikoo> lol i as look your idea is pretty simple but genius in same time
18:12 < bridge> <manikoo> good tool
18:13 < bridge> <manikoo> lol as ilook your idea is pretty simple but genius in same time
18:13 < bridge> <Assa> it might actually does, if starting an old client removes newer configs
18:13 < bridge> <manikoo> lol as i look your idea is pretty simple but genius in same time
18:13 < bridge> <inv41idu53rn4m3> also having it depend on your personal configuration could mess with the results
18:13 < bridge> <Assa> I assume you're not switching your personal configuration between runs
18:15 < bridge> <Assa> for me this was so far sufficient in order to create benchmarks tbh. yes we could improve on the method
18:15 < bridge> <inv41idu53rn4m3> making your personal tool public was a grave mistake
18:16 < bridge> <Assa> no, it's a proof for me. Hiding it would also be besides the point, as my benchmarks would have no values, if I could have just generated them with chatgpt
18:16 < bridge> <inv41idu53rn4m3> yeah I'm just kidding
18:16 < bridge> <Assa> no, it's a proof for me. Hiding it would also be besides the point, as my benchmarks would have no value, if I could have just generated them with chatgpt
18:18 < bridge> <robyt3> Depends on which client version. Some will keep unknown settings. You should probably use storage.cfg to use the default settings for every version.
18:18 < bridge> <Assa> I am a scientist by profession, we're currently in the discussion part. And using storage.cfg is correcter, but this doesn't immediately discredit my results
18:20 < bridge> <Assa> not using wayland however fully invalidated my ubuntu results, all I am showing is that xorg is shit on my laptop xD
18:27 < bridge> <Assa> I'll add the storage.cfg, this should be really simple and is part of the setup step
18:41 < bridge> <manikoo> 😭 😭  it was the last client(i had 10 cuz didn't know it will be that long) and i got some crazy python error
18:41 < bridge> <Assa> at which step?
18:42 < bridge> <Assa> don't worry, you won't need to redo everything, everything that is already created will be skipped 🙂
18:42 < bridge> <manikoo> opening demo on new client(19.5)
18:43 < bridge> <manikoo> ```[RUN] DDNet.exe "play demos/Abyss-Benchmark.demo; exec benchmark-Abyss.cfg" for client clients\ddnet-19.5-win64\DDNet-19.5-win64\DDNet.exe```
18:43 < bridge> <manikoo> oo gud
18:43 < bridge> <Assa> and what failed?
18:44 < bridge> <manikoo> || ```Traceback (most recent call last):
18:44 < bridge> <manikoo>   File "C:\Users\solarqueen\AppData\Local\Programs\Python\Python313\Lib\threading.py", line 1043, in _bootstrap_inner
18:44 < bridge> <manikoo>     self.run()
18:44 < bridge> <manikoo>     ~~~~~~~~^^
18:44 < bridge> <manikoo>   File "C:\Users\solarqueen\AppData\Local\Programs\Python\Python313\Lib\threading.py", line 994, in run
18:44 < bridge> <manikoo>     self._target(*self._args, **self._kwargs)
18:44 < bridge> <manikoo>     ~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
18:44 < bridge> <manikoo>   File "C:\Users\solarqueen\AppData\Local\Programs\Python\Python313\Lib\subprocess.py", line 1615, in _readerthread
18:44 < bridge> <manikoo>     buffer.append(fh.read())
18:44 < bridge> <manikoo>                   ~~~~~~~^^
18:44 < bridge> <manikoo>   File "C:\Users\solarqueen\AppData\Local\Programs\Python\Python313\Lib\encodings\cp1251.py", line 23, in decode
18:44 < bridge> <manikoo>     return codecs.charmap_decode(input,self.errors,decoding_table)[0]
18:44 < bridge> <manikoo>            ~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
18:44 < bridge> <manikoo> UnicodeDecodeError: 'charmap' codec can't decode byte 0x98 in position 8753: character maps to <undefined>
18:44 < bridge> <manikoo> Traceback (most recent call last):
18:44 < bridge> <manikoo>   File "C:\Downloads\ddnet-benchmarks-main\ddnet-benchmarks-main\scripts\create-benchmarks.py", line 71, in <module>
18:45 < bridge> <manikoo>     main()
18:45 < bridge> <manikoo>     ~~~~^^
18:45 < bridge> <manikoo>   File "C:\Downloads\ddnet-benchmarks-main\ddnet-benchmarks-main\scripts\create-benchmarks.py", line 67, in main
18:45 < bridge> <manikoo>     run_benchmark(Path(client), Path(demo), windows)
18:45 < bridge> <manikoo>     ~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
18:45 < bridge> <manikoo>   File "C:\Downloads\ddnet-benchmarks-main\ddnet-benchmarks-main\scripts\create-benchmarks.py", line 43, in run_benchmark
18:45 < bridge> <manikoo>     for line in result.stdout.split('\n'):
18:45 < bridge> <manikoo>                 ^^^^^^^^^^^^^^^^^^^
18:45 < bridge> <manikoo> AttributeError: 'NoneType' object has no attribute 'split'``` ||
18:45 < bridge> <manikoo> biggo text
18:45 < bridge> <manikoo> some shit probably
18:45 < bridge> <Assa> result is none, maybe the client was closed or crashed <.<
18:45 < bridge> <Assa> does just rerunning the script work?
18:45 < bridge> <manikoo> gimme sec
18:46 < bridge> <manikoo> which step
18:46 < bridge> <Assa> `python3 scripts\create-benchmarks.py`
18:47 < bridge> <Assa> `python3 scripts/create-benchmarks.py`
18:47 < bridge> <manikoo> sure
18:47 < bridge> <manikoo> all skipped
18:47 < bridge> <manikoo> but this one also
18:47 < bridge> <Assa> was also skipped, so it created the benchmarks at least
18:48 < bridge> <manikoo> oh i see this error comes out in every part of benchmark on 19.5
18:49 < bridge> <inv41idu53rn4m3> > UnicodeDecodeError: 'charmap' codec can't decode byte 0x98 in position 8753: character maps to <undefined>
18:49 < bridge> <inv41idu53rn4m3> how does that even happen
18:49 < bridge> <Assa> okay oO good thing is, that it's only doing this after the subprocess ran
18:49 < bridge> <inv41idu53rn4m3> is this Windows terminal being wacky?
18:49 < bridge> <manikoo> maybe
18:50 < bridge> <manikoo> it runs demo but crashes during it playing
18:50 < bridge> <Assa> maybe this happens when you open or write a file in python, and all we need to do is set utf8
18:50 < bridge> <Assa> do you use my demos or create your own?
18:50 < bridge> <manikoo> your
18:50 < bridge> <Assa> hmm oO
18:51 < bridge> <manikoo> it was all gud on previous versions
18:51 < bridge> <Assa> nice you're looking at my trash gameplay xD
18:51 < bridge> <manikoo> ```  "client_list": [
18:51 < bridge> <manikoo>     "16.1",
18:51 < bridge> <manikoo>     "16.2.2",
18:51 < bridge> <manikoo>     "16.6",
18:51 < bridge> <manikoo>     "17.0.3",
18:51 < bridge> <manikoo>     "17.3",
18:51 < bridge> <manikoo>     "18.9.1",
18:51 < bridge> <manikoo>     "19.0",
18:51 < bridge> <manikoo>     "19.1",
18:51 < bridge> <manikoo>     "19.2.1",
18:51 < bridge> <manikoo>     "19.5"```
18:51 < bridge> <manikoo> 9 times
18:52 < bridge> <Assa> what happens if you open the demos on your current client, which should also be 19.5?
18:52 < bridge> <Assa> like without using the script?
18:53 < bridge> <manikoo> one moment
18:54 < bridge> <manikoo> all gud
18:55 < bridge> <inv41idu53rn4m3> Crashes in what way? The script intentionally closes the game after 30 seconds
18:55 < bridge> <Assa> If you have time, we can go into a VC
18:55 < bridge> <Assa> I'd go into the created client directory now, like clients/ddnet-19.5..../ddnet ..../ and directly run DDNet.exe there
18:55 < bridge> <inv41idu53rn4m3> so the game just closing unexpectedly is not unexpected
18:55 < bridge> <inv41idu53rn4m3> so the game just closing "unexpectedly" is not unexpected
18:55 < bridge> <manikoo> i did it exactly
18:55 < bridge> <manikoo> and runned demo there
18:55 < bridge> <manikoo> your one
18:55 < bridge> <Assa> hmm
18:56 < bridge> <Assa> and the script keeps rerunning the "broken clients"?
18:56 < bridge> <manikoo> no, it stops to collect data
18:56 < bridge> <Assa> or does it skip and tries the next demo?
18:56 < bridge> <manikoo> this one
18:56 < bridge> <manikoo> but then also error
18:57 < bridge> <Assa> then the data is created
18:57 < bridge> <inv41idu53rn4m3> Benchmark command broken on Windows???
18:57 < bridge> <Assa> maybe something with python subprocess
18:57 < bridge> <inv41idu53rn4m3> or you mean even manually running benchmark works?
18:58 < bridge> <Assa> you can modify script/create-benchmarks.py and just comment this loop ` for line in result.stdout.split('\n'):` and everything inside it (probably just a print)
18:58 < bridge> <Assa> you can modify script/create-benchmarks.py and just comment this loop ` for line in result.stdout.split('\n'):` and everything inside it (probably just a print) out
19:05 < bridge> <robyt3> You generally need this at the beginning of your Python script if you want to support redirecting its UTF-8 stdout/stderr to a file on Windows (see `scripts/export_settings_commands_table.py`):
19:05 < bridge> <robyt3> 
19:05 < bridge> <robyt3> ```python
19:05 < bridge> <robyt3> sys.stdout = io.TextIOWrapper(sys.stdout.buffer, encoding='utf-8', errors='replace')
19:05 < bridge> <robyt3> sys.stderr = io.TextIOWrapper(sys.stderr.buffer, encoding='utf-8', errors='replace')
19:05 < bridge> <robyt3> ```
19:05 < bridge> <robyt3> 
19:05 < bridge> <robyt3> Or this when you want to launch a process and capture its UTF-8 stdout/stderr (see `scripts/integration_test.py`):
19:05 < bridge> <robyt3> 
19:05 < bridge> <robyt3> ```python
19:05 < bridge> <robyt3> stdout_wrapper = io.TextIOWrapper(self.process.stdout, encoding="utf-8", newline="\n")
19:05 < bridge> <robyt3> stderr_wrapper = io.TextIOWrapper(self.process.stderr, encoding="utf-8", newline="\n")
19:05 < bridge> <robyt3> ```
19:07 < bridge> <Assa> I also never specified a python3 version, there might also be differences 🤷‍♂️ 
19:07 < bridge> <Assa> You are probably right, for me it worked so far alsways, this loop was only printing the client output of the subprocess to the main process, didn't hink of this very long
19:07 < bridge> <Assa> I also never specified a python3 version, there might also be differences 🤷‍♂️ 
19:07 < bridge> <Assa> You are probably right, for me it worked so far always, this loop was only printing the client output of the subprocess to the main process, didn't hink of this very long
19:10 < bridge> <Assa> @manikoo you alright?
19:10 < bridge> <manikoo> sure
19:10 < bridge> <manikoo> im just confused by DDNet_Benchmarking.ipynb
19:11 < bridge> <Assa> yeah you probaly need to set the client_versions again
19:11 < bridge> <Assa> do you use colab or a local server?
19:11 < bridge> <Assa> yeah you probaly need to set the client_versions again (in the notebook, not in the json)
19:12 < bridge> <manikoo> idk :pepeW:  i just run it
19:12 < bridge> <Assa> then a local server 😄
19:13 < bridge> <Assa> you might need to adjust the path, where the results are read from
19:13 < bridge> <manikoo> watafak is this
19:13 < bridge> <manikoo> https://cdn.discordapp.com/attachments/293493549758939136/1429155505099640953/O83TrXG.png?ex=68f51c48&is=68f3cac8&hm=77243d1acb5ff82ac88ad11a959b1db64540c91cba498989cb64f6bcae2d166d&
19:14 < bridge> <Assa> missing requirements, like you are missing python packages
19:14 < bridge> <manikoo> but i have this
19:15 < bridge> <Assa> ah it's complaining, that it's not in the requirements.txt
19:15 < bridge> <Assa> what happens if you run the notebook?
19:15 < bridge> <manikoo> what do you mean?
19:16 < bridge> <manikoo> this file in simple notebook?
19:16 < bridge> <Assa> DDNet_Benchmark.ipynb is an "interactive python note book" -> i py nb
19:16 < bridge> <Assa> DDNet_Benchmark.ipynb is an "interactive python notebook" -> i py nb
19:17 < bridge> <Assa> you should have like cells, that you can execute when you run the notebook
19:17 < bridge> <manikoo> yeah i have
19:17 < bridge> <Assa> then run everything and report errors 😄
19:18 < bridge> <manikoo> hmm
19:18 < bridge> <manikoo> it complains on modules again
19:18 < bridge> <Assa> in the 4th cell you may need to adjust `client_list` and the `name = f"{client}-benchmark-{map}.txt"` bit
19:18 < bridge> <Assa> do you use a venv, or conda environment?
19:19 < bridge> <Assa> maybe you just need to pip install numpy matplotlib pandas
19:19 < bridge> <Assa> maybe you just need to `pip install numpy matplotlib pandas`
19:19 < bridge> <manikoo> i did run ```pip install -r requirements.txt``` before
19:20 < bridge> <Assa> ^ I didn't put them in, because I was never running the notebook on my machine
19:20 < bridge> <Assa> this is the data-analyzing part, I adjust this a lot and play with the data around, that's why I did it like this
19:22 < bridge> <manikoo> maybe i need pc restart or what
19:22 < bridge> <manikoo> i already have this modules but it complains
19:23 < bridge> <Assa> hmm maybe close and reopen the notebook first? 🤔
19:24 < bridge> <manikoo> nah it's crazy
19:25 < bridge> <Assa> you could try to add a cell, add `!pip install numpy matplotlib pandas` and run it
19:25 < bridge> <Assa> if it's still crazy idk, I  am still up for a vc if you want
19:25 < bridge> <Assa> you could try to add a cell, add `!pip install numpy matplotlib pandas` and run it (the `!` is used in order to do bash commands)
19:27 < bridge> <manikoo> holy cow it works
19:28 < bridge> <manikoo> but there is error in client list
19:28 < bridge> <manikoo> https://cdn.discordapp.com/attachments/293493549758939136/1429159142677024808/wyFJHSe.png?ex=68f51fac&is=68f3ce2c&hm=4041611f9939d0548d4cd13c2716e30279d5dd4048b1526bce65ed5d36e479fa&
19:28 < bridge> <Assa> should bie client_list `=`!
19:28 < bridge> <Assa> should bie client_list `=`
19:28 < bridge> <Assa> should be client_list `=`
19:29 < bridge> <manikoo> what do i need to change in second one
19:29 < bridge> <Assa> also your idention is wrong, idk if the notebook cares
19:29 < bridge> <Assa> assuming you run this in the root directory, you might need to add `results/` before
19:29 < bridge> <manikoo> it cared but i solved
19:30 < bridge> <manikoo> i have results folder with data there
19:31 < bridge> <Assa> then you probably need `name = f"results/{client}-benchmark-{map}.txt"`
19:33 < bridge> <manikoo> guess what i forgot
19:33 < bridge> <Assa> running the collection script?
19:33 < bridge> <Assa> running the collection script? idk xD
19:33 < bridge> <manikoo> unlocking my fps
19:33 < bridge> <Assa> XD
19:34 < bridge> <manikoo> https://cdn.discordapp.com/attachments/293493549758939136/1429160703662755860/PvdbQ27.png?ex=68f52120&is=68f3cfa0&hm=e529ed0d0a05814f0849291c6c939be691c8a04ba0d4879d0c206fb947df3850&
19:34 < bridge> <manikoo> but i don't need it as i see
19:35 < bridge> <manikoo> crazy graph
19:35 < bridge> <Assa> for comparison
19:35 < bridge> <manikoo> (igpu)
19:36 < bridge> <Jupstar ✪> that graph looks buggy
19:37 < bridge> <manikoo> yes
19:37 < bridge> <manikoo> maybe something cooked bad cuz i alt-tabed few times
19:37 < bridge> <Jupstar ✪> xd
19:37 < bridge> <manikoo> ok i will cook fresh bench later
19:38 < bridge> <Assa> if you want to redo it without unlocking, you need to delete the results inside the client directories, e.g. `rm clients/ddnet-*/*/benchmark-*.txt`
19:38 < bridge> <manikoo> i will just delete everything it results/
19:38 < bridge> <Assa> and don't forget to remove your fps cap before benchmarking <.<
19:39 < bridge> <Assa> this will not be sufficient unfortunately
19:39 < bridge> <Assa> like everything will be skippt except the collect-results script
19:39 < bridge> <Assa> like everything will be skipt except the collect-results script
19:39 < bridge> <Assa> like everything will be skiped except the collect-results script
19:39 < bridge> <manikoo> ok i will delete everything and download again
19:40 < bridge> <manikoo> https://tenor.com/view/sakujo-sakujo-sakujo-sakujo-mikami-teru-teru-mikami-death-note-gif-8622680172965216876
19:40 < bridge> <Assa> just don't delete the notebook 🙈
19:40 < bridge> <manikoo> ok
19:40 < bridge> <filoqcus> :poggers2:
19:41 < bridge> <Assa> I agree, but having an fps cap and this is the worst 0.01, really scuffs this results
19:42 < bridge> <manikoo> https://cdn.discordapp.com/attachments/293493549758939136/1429162701149700116/DpfzF7v.png?ex=68f522fc&is=68f3d17c&hm=96b0ef78024f17947868a44a80205af8a6427386c2b661e1b32b7622a97d6138&
19:44 < bridge> <Assa> this really looks bad <.<
19:51 < bridge> <Assa> 19.5 doesn't make sense here, maybe the data was actually not collected properly, as you reported it closing somewhat unexpectedly
20:27 < bridge> <manikoo> pheew i did it again
20:27 < bridge> <manikoo> https://cdn.discordapp.com/attachments/293493549758939136/1429174046939152425/SPOILER_EBtXPtr.png?ex=68f52d8d&is=68f3dc0d&hm=6a375ce0cca962312aa43019356d321243ff1689b2e39a2992e3176c9d60fb07&
20:27 < bridge> <manikoo> thx Assa for the help
20:29 < bridge> <manikoo> r5 8600g(radeon 760m), if someone needs
20:39 < bridge> <Assa> did you not get 19.5 to work? :/
20:39 < bridge> <manikoo> same error :/
20:39 < bridge> <manikoo> oh
20:39 < bridge> <manikoo> man i forgot
20:39 < bridge> <manikoo> to change
20:40 < bridge> <manikoo> i forgot to comment that line
20:40 < bridge> <Assa> so it just worked this time?
20:40 < bridge> <manikoo> sure
20:40 < bridge> <Assa> there was a big leap from 19.2 to 19.4, id expect it to be seen here 😄
20:41 < bridge> <manikoo> oh
20:42 < bridge> <Assa> interesting that abyss is also the worst map here (like on my ubuntu), while it's not on windows 🤔
20:44 < bridge> <manikoo> 0:
20:44 < bridge> <manikoo> just tried 19.4 and it's ALSO have this error
20:44 < bridge> <manikoo> i will comment that line really quick
20:44 < bridge> <Teero> i hate abyss it literally stole the mapname from the kog abyss map and now it says i have 30h on the ddnet shit abyss map on ddstats.org.
20:45 < bridge> <Assa> you can try to do that step manually, by starting the client, open the benchmark and run `exec benchmark-Abyss.txt` in f1, stop the demo first, press esc (so the demo overlay is closed as it also affects fps)
20:46 < bridge> <Assa> from a design pov I really like abyss, like I love if mappers actually put effort into their maps and not just use entities basically
20:46 < bridge> <Teero> :(
20:46 < bridge> <Teero> https://cdn.discordapp.com/attachments/293493549758939136/1429178776956436542/image.png?ex=68f531f5&is=68f3e075&hm=fcb027432705778a685336db9386a500a91009ab306e23027ff823f1a7aedd11&
20:46 < bridge> <Assa> I have most hours on KingsLeap, as it's my own map, I spend a hundret of hours mapping and I also hunted it xD
20:46 < bridge> <Teero> yea i just don't like the ddrace gamemode but now ddstats tells me i spent 22h on ddnet Abyss even though kog Abyss is one of my favourite maps
20:47 < bridge> <Teero> xddd
20:47 < bridge> <Teero> https://cdn.discordapp.com/attachments/293493549758939136/1429179063997698250/image.png?ex=68f53239&is=68f3e0b9&hm=0f15bc0cf171387e3c4b30d00c5872c0f393639a6a69b4ca69e43e9e577e34c1&
20:47 < bridge> <Assa> then there also is Abyssal by Nixus :p
20:47 < bridge> <Teero> i ddstats could count local server time i'd have >500h on aip gores XD
20:49 < bridge> <Assa> ddstats.org doesn't load for me 🤔
20:49 < bridge> <Teero> ddstats.tw
20:49 < bridge> <Teero> mb
20:50 < bridge> <Teero> you have 0h on Gores?? XD
20:50 < bridge> <Teero> https://cdn.discordapp.com/attachments/293493549758939136/1429179740538798090/image.png?ex=68f532db&is=68f3e15b&hm=3b172eb62705155c50b2c0bd9a826d6895ba21be07768040450add3b1e54f578&
20:50 < bridge> <Teero> how
20:50 < bridge> <Assa> by _not playing gores_
20:50 < bridge> <Teero> nub
20:50 < bridge> <Teero> not playing the superior gamemode
20:50 < bridge> <Assa> wdym race is right there
20:51 < bridge> <Teero> what is Race?
20:51 < bridge> <Ewan> what's going on #developer 😨
20:52 < bridge> <Ewan> how do you not know what race is
20:52 < bridge> <Assa> he does he is just trolling
20:53 < bridge> <Assa> says he doesn't know race, also posts the ideal line on an unrealeased map in the mapchannel of unique
20:54 < bridge> <Ewan> 😌
20:56 < bridge> <pioooooo> rate 1-10
20:56 < bridge> <pioooooo> https://cdn.discordapp.com/attachments/293493549758939136/1429181289491337276/1e0548f211946f95.png?ex=68f5344c&is=68f3e2cc&hm=a55827957f6b0c13246b48cfd0d87bb6e90adf7dbefc399ecc8389712303391a&
20:57 < bridge> <Assa> not enough race hunting/10
20:58 < bridge> <pioooooo> we don't have unique in china
20:58 < bridge> <pioooooo> D:
20:58 < bridge> <Assa> D: oh I'm sorry
20:58 < bridge> <pioooooo> I finished all but one race on ddnet
20:59 < bridge> <Assa> antibuguse?
20:59 < bridge> <pioooooo> yes
20:59 < bridge> <pioooooo> didn't yet try grind it
21:15 < bridge> <big eyed bencie> good morning (4 pm)
21:15 < bridge> <Assa> my ubuntu benchmarks just crashed the laptop as I ran out of memory oO
21:15 < bridge> <Assa> good evening 21:15
21:18 < bridge> <Assa> I wonder how that was possible, ddnet does not even use 2GB of ram there 🤔
21:19 < bridge> <Assa> lol this is highly version dependend, 16.3 just took 6 GB
21:20 < bridge> <Assa> ah it's depending on the map that is loaded, makes sense
21:26 < bridge> <Assa> ofc the map in question is Abyss ...
21:40 < bridge> <Assa> I get logs like "snap invalidate problem"
21:41 < bridge> <Assa> hmm might come from 16.1 running 19.4ish demos
21:47 < bridge> <robyt3> New demos contain new objects/messages that old clients don't understand. How do you record the demos anyway? Seems pretty inefficient if you have to do it manually. You could probably take one demo and replace the map data and meta data.
21:48 < bridge> <Assa> I just recorded a demo once and automatically generate benchmarks from it
21:48 < bridge> <Assa> I just took a look at the logs and noticed it
21:49 < bridge> <robyt3> The script could also do it automatically from only one demo I suppose
21:50 < bridge> <Assa> what do you mean?
21:51 < bridge> <Assa> my current state: I have a list of demos and a list of clients, all I do is benchmark the matrix from it
21:51 < bridge> <robyt3> You don't need to commit a demo for each map into the repository. You could also specify a list of maps when launching the benchmark, then generate demos from those maps automatically.
21:52 < bridge> <robyt3> I guess it would only really work as spectator though because of physics
21:53 < bridge> <Assa> I wanted to use demos of actual gameplay, I chose it like this deliberately
21:53 < bridge> <Assa> idk how I'd be able to generate demos like this. All I can imagine in recording a bunch of bots on each maps
21:53 < bridge> <Assa> idk how I'd be able to generate demos like this. All I can imagine in recording a bunch of bots on each map
21:54 < bridge> <robyt3> I assumed it was only about benchmarking the map rendering, not other components
21:55 < bridge> <robyt3> To only evaluate rendering on any map, I would record a demo as spectator, then replace the map data in the demo with another map
21:56 < bridge> <Assa> you should also notice benefits on normal gameplay. This way I can reuse the same data for other benchmarks
21:57 < bridge> <Assa> except zoom, because this is not reflected in demos
22:09 < bridge> <Assa> apperently wayland makes my desktop freeze after a while oO
22:09 < bridge> <Jupstar ✪> tja
22:09 < bridge> <Jupstar ✪> wayland L
22:33 < bridge> <Assa> Apparently the crashed versions still create the files 🤔  this is much more inline with windows, you can also see the regression between 17.2 and 19.2. Idk why 19.3 to 19.4 doesn't show any benefits.
22:33 < bridge> <Assa> https://cdn.discordapp.com/attachments/293493549758939136/1429205858667069522/DDNetBenchmarksUbuntu.png?ex=68f54b2e&is=68f3f9ae&hm=5b0d5a601a90f8a9a6bd6b6ac52fe5e55e02dd3622764a81493270947f76a029&