01:15 <+bridge> <Voxel> oops
05:58 <+bridge> <Headshot> :justatest:
05:58 <+bridge> <Headshot> https://cdn.discordapp.com/attachments/293493549758939136/1060784353233350716/Screenshot_20230106_010155_Chrome.png
07:10 <+bridge> <uwu> Every spanish text reminds me of tropico game series
07:10 <+bridge> <uwu> https://tenor.com/view/michael-nesmith-elephant-parts-rio-juice-hat-gif-23073584
07:28 <+bridge> <Simon'> What about hotkeys for tile layers that u can choose urself in the editor?
07:33 <+bridge> <Deëivid!> oops, lil typo
07:36 <+bridge> <Deëivid!> need new glasses
08:55 <+bridge> <cagent> a
10:41 <+bridge> <Ryozuki> b
10:42 <+bridge> <Chairn> feel free to fix 😉
11:03 <+bridge> <Ryozuki> https://cdn.discordapp.com/attachments/293493549758939136/1060861109575356436/FluHdGnWQAE-St-.jpg
11:05 <+bridge> <Jupstar ✪> Epic compiler moment
11:06 <+bridge> <Jupstar ✪> Soon it might be worth using Gentoo if Firefox only requires 5min to compile instead of 20min xd
11:16 <+bridge> <Ryozuki> xd
11:16 <+bridge> <Ryozuki> i  leave my pc compiling when i go to gym
11:46 <+bridge> <ReiTW> is it even useful lol
11:46 <+bridge> <ReiTW> even the worst games on my Ryzen 5 5600 only consume  60%  of my CPU usage
11:52 <+bridge> <Jupstar ✪> That's not how u should look at it. In 5 years games can assume the worst pc to support has a Ryzen 7k and do more stuff
11:59 <+bridge> <Ryozuki> 5.7ghz for compiling must be p fast
12:00 <+bridge> <Jupstar ✪> If it can hold it on 16 cores at once then gg
12:18 <+bridge> <Learath2> I just compile in the background. I have some io issue I need to diagnose but normally it doesnt even stutter
12:26 <+bridge> <Ryozuki> probs not
12:27 <+bridge> <Ryozuki> usually there are a group of cores that run at that
12:27 <+bridge> <Ryozuki> and the other at like 5.6 5.5 idk
12:27 <+bridge> <Ryozuki> i read it yesterday
12:27 <+bridge> <Ryozuki> forgot the link
12:27 <+bridge> <Ryozuki> oh
12:28 <+bridge> <Ryozuki> here
12:28 <+bridge> <Ryozuki> https://chipsandcheese.com/2023/01/05/amds-zen-4-part-3-system-level-stuff-and-igpu/
12:28 <+bridge> <Ryozuki> >  Like prior 16 core AMD desktop parts, not all cores will boost to the CPU’s maximum clock. On our sample of the 7950X, we saw the highest clock frequencies from cores 0, 3, and 4. All three of these cores were located on the first CCD.
12:28 <+bridge> <Ryozuki> @Jupeyy_Keks
12:28 <+bridge> <Ryozuki> https://cdn.discordapp.com/attachments/293493549758939136/1060882595434475611/zen4_boostclk.webp
12:29 <+bridge> <Ryozuki> not bad if u ask me
12:29 <+bridge> <Ryozuki> https://cdn.discordapp.com/attachments/293493549758939136/1060882867175039006/image.png
12:48 <+bridge> <Jupstar ✪> Yeah still better than any core i have;~;
12:59 <+bridge> <deen> that will be more extreme with the 7950x3d though. It seems liike one CCD has the cache on top, so it won't be able to clock so high to keep heat production down since it dissipates worse through the cache layer
13:01 <+bridge> <deen> Will be interesting for scheduling, one numa region with more local cache, the other with higher frequency. I guess benchmarkers will try running some applications on one region vs the other to see where it runs better
15:22 <+bridge> <Jupstar ✪> can someone add time support for gores?
15:22 <+bridge> <Jupstar ✪> https://cdn.discordapp.com/attachments/293493549758939136/1060926413600981062/image.png
15:32 <+bridge> <Ryozuki> iirc its done through the http master?
15:32 <+bridge> <Ryozuki> so tell the closed source kog devs to do it
15:32 <+bridge> <Ryozuki> smh
15:33 <+bridge> <Jupstar ✪> oh thought its bcs client assumes its not a race mod
15:33 <+bridge> <Ryozuki> probs in the past
15:34 <+bridge> <Ryozuki> but i think u can control it now
15:34 <+bridge> <Ryozuki> in the way u register in master
15:34 <+bridge> <Jupstar ✪> ok
15:34 <+bridge> <Jupstar ✪> @Avolicious can u fix this?
15:36 <+bridge> <Avolicious> Yeah, after changing the gamemode to "Gores", time support was disabled
15:36 <+bridge> <Avolicious> AFAIK heinrich fixed it by adding a new field to the JSON master register
15:39 <+bridge> <Ryozuki> @Avolicious u can add any field u want on the master register
15:39 <+bridge> <Ryozuki> what u mean is he probs changed it so the client interprets a specific field of the json entry to know if its a time or a score
15:40 <+bridge> <Avolicious> Sure, you can send all kinds of data in the json, but you have to use the correct field to set it from score to time 🙂
15:40 <+bridge> <Ryozuki> yeah that
15:40 <+bridge> <Ryozuki> i forgot which one it is tho
15:40 <+bridge> <Ryozuki> xd
15:41 <+bridge> <Ryozuki> cool to see the skins get reported, we can know display the tees on website
15:41 <+bridge> <Ryozuki> https://cdn.discordapp.com/attachments/293493549758939136/1060931007961907230/image.png
15:41 <+bridge> <Ryozuki> @Jupeyy_Keks maybe u can render the skins too
15:41 <+bridge> <Ryozuki> ur the pro in that
15:42 <+bridge> <Jupstar ✪> yeah i thought about it, but space is really tight, maybe the player flag needs to be smaller
15:43 <+bridge> <Jupstar ✪> https://cdn.discordapp.com/attachments/293493549758939136/1060931517762777088/image.png
15:43 <+bridge> <Jupstar ✪> our scrollbars are also so insanely fat xD
15:45 <+bridge> <Jupstar ✪> @Robyt3 can we remove this space or does it look bad then.. e.g. in discord there is no such space
15:45 <+bridge> <Jupstar ✪> https://cdn.discordapp.com/attachments/293493549758939136/1060932144681205881/image.png
15:45 <+bridge> <Ryozuki> xd
15:52 <+bridge> <Ryozuki> cc @deen
15:53 <+bridge> <Ryozuki> cant test cuz scripts but probs should work xd
15:56 <+bridge> <zyzhiyou> My game can't turn pages up and down
15:56 <+bridge> <zyzhiyou> https://cdn.discordapp.com/attachments/293493549758939136/1060934898992545882/image.png
15:58 <+bridge> <Jupstar ✪> those dont work
15:58 <+bridge> <Jupstar ✪> do u have a "fn" key?
15:59 <+bridge> <zyzhiyou> yes
15:59 <+bridge> <Robyt3> I already removed some of the margin on the scoreboard in #6240
15:59 <+bridge> <Robyt3> https://cdn.discordapp.com/attachments/293493549758939136/1060935605535645706/211005074-5d113fd3-e541-45e6-95d1-4d7f633b4d12.png
15:59 <+bridge> <DDNet> https://github.com/ddnet/ddnet/pull/6240
15:59 <+bridge> <Jupstar ✪> try to find another page up key
16:00 <+bridge> <Jupstar ✪> nice every pixel counts for server player list xd
16:00 <+bridge> <zyzhiyou> 'fn' also cannot be used
16:01 <+bridge> <zyzhiyou> I can only use it in 'm3 m4'
16:01 <+bridge> <zyzhiyou> https://cdn.discordapp.com/attachments/293493549758939136/1060936160010059866/image.png
16:01 <+bridge> <Jupstar ✪> fn is just a key to enable fn keys
16:03 <+bridge> <Jupstar ✪> i dunno how the software works but i guess we handle raw input
16:03 <+bridge> <Jupstar ✪> maybe u can send it as if the keyboard has sent it somehow
16:03 <+bridge> <zyzhiyou> The mechanical keyboard can turn pages up and down
16:04 <+bridge> <Jupstar ✪> can u photo your whole keyboard?
16:04 <+bridge> <Jupstar ✪> in better quality too
16:04 <+bridge> <zyzhiyou> Notebook keyboard cannot be used
16:05 <+bridge> <Jupstar ✪> problem is that page up on numpad are evaluated as numpad keyys
16:05 <+bridge> <Jupstar ✪> problem is that page up on numpad are evaluated as numpad keys
16:05 <+bridge> <Jupstar ✪> inp_translated_keys 1
16:05 <+bridge> <Jupstar ✪> 
16:05 <+bridge> <Jupstar ✪> might work tho
16:05 <+bridge> <Jupstar ✪> but not for me
16:06 <+bridge> <zyzhiyou> https://cdn.discordapp.com/attachments/293493549758939136/1060937403159810158/IMG_20230106_230517.jpg
16:06 <+bridge> <Jupstar ✪> send the whole keyboard xD is there no additional page up?
16:07 <+bridge> <Jupstar ✪> maybe the arrow keys work with `fn` key
16:07 <+bridge> <Jupstar ✪> else simply rebind whatever u want to do with page up to the ones on the numpad, if u dont need it
16:07 <+bridge> <zyzhiyou> https://cdn.discordapp.com/attachments/293493549758939136/1060937716440768562/IMG_20230106_230711.jpg
16:09 <+bridge> <zyzhiyou> Arrow keys cannot be used either
16:09 <+bridge> <Jupstar ✪> anyway we use SDL's scancodes so we dont have conflicts with keyboard layouts etc.
16:09 <+bridge> <Jupstar ✪> 
16:09 <+bridge> <Jupstar ✪> guess this is one simple disadvantage, so rebinding might be easier
16:13 <+bridge> <TsFreddie> my colemak layout is still weird in ddnet
16:14 <+bridge> <TsFreddie> (but I did use registry to rebind my keys so......)
16:14 <+bridge> <TsFreddie> :troll:
16:14 <+bridge> <TsFreddie> btw long time no see
16:14 <+bridge> <Jupstar ✪> wb good to see you back
16:14 <+bridge> <TsFreddie> wdym back
16:15 <+bridge> <Jupstar ✪> good to see you here for the next 20 seconds
16:15 <+bridge> <TsFreddie> :kek:  true
16:15 <+bridge> <zyzhiyou> 😭
16:16 <+bridge> <zyzhiyou> ts
16:16 <+bridge> <TsFreddie> i'm just here to contain our psych patient @zyzhiyou since i've been hearing him from a distance.
16:16 <+bridge> <zyzhiyou> ？？？
16:16 <+bridge> <TsFreddie> lmao
16:16 <+bridge> <zyzhiyou> 😭😭😭😭
16:17 <+bridge> <zyzhiyou> l'm noob
16:20 <+bridge> <Ryozuki> :monkaS:
16:45 <+bridge> <Ryozuki> https://scholarlykitchen.sspnet.org/2023/01/05/github-is-sued-and-we-may-learn-something-about-creative-commons-licensing/
16:45 <+bridge> <Ryozuki> https://unicourt.com/case/pc-db5-doe-1-et-al-v-github-inc-et-al-1330346
16:47 <+bridge> <Ryozuki> > On November 3, 2022, Doe 1 and Doe 2 (collectively, “Plaintiffs”), on behalf of themselves and all others similarly situated, represented by Joseph R. Saveri, Cadio Zirpoli, and Travis Manfredi of Joseph Saveri Law Firm, LLP, filed a class-action lawsuit against GitHub, Inc., Microsoft Corporation, OpenAI, Inc., OpenAI L.P., OpenAI GP, L.L.C., OpenAI Startup Fund GP I, L.L.C., OpenAI Startup Fund I, L.P., and OpenAI Startup Fund Management, L
16:47 <+bridge> <Ryozuki> :BASEDDEPT:
18:13 <+bridge> <Jupstar ✪> ^ @Ryozuki
18:14 <+bridge> <Ryozuki> poggers
18:16 <+bridge> <Ryozuki> @Jupeyy_Keks i feel like it needs some padding
18:16 <+bridge> <Ryozuki> specially when the time is at hours
18:16 <+bridge> <Ryozuki> or maybe its some trick to the eye
18:16 <+bridge> <Ryozuki> xd
18:16 <+bridge> <Jupstar ✪> yeah its weird, bcs there is no padding left
18:18 <+bridge> <Jupstar ✪> ah i know ^^
18:19 <+bridge> <Jupstar ✪> https://cdn.discordapp.com/attachments/293493549758939136/1060970955947655198/image.png
18:23 <+bridge> <Jupstar ✪> so many servers that diditn restart in weeks xD
18:23 <+bridge> <Jupstar ✪> they all dont provide the information
18:29 <+bridge> <Ryozuki> xd
18:47 <+bridge> <Jupstar ✪> lmao i posted the screen in clan chat, they all didnt notice the difference xDDD
19:20 <+bridge> <deen> not official ddnet?
19:20 <+bridge> <Jupstar ✪> ddnet turkey and others
19:21 <+bridge> <Jupstar ✪> 37.230.210.231:8327
19:21 <+bridge> <Jupstar ✪> 37.230.210.231:8316
19:21 <+bridge> <Jupstar ✪> 78.135.85.66:8330
19:21 <+bridge> <Jupstar ✪> and many more
19:22 <+bridge> <deen> On Turkey I see all were restarted on January 4 or later
19:23 <+bridge> <deen> On RUS all on January 5
19:23 <+bridge> <deen> Why do you think they were not restarted for weeks?
19:24 <+bridge> <Jupstar ✪> well nnot latest ddnet
19:24 <+bridge> <Jupstar ✪> but maybe thats different
19:28 <+bridge> <deen> Then all official servers should be on old version
19:28 <+bridge> <deen> I'll build new one now
19:29 <+bridge> <Jupstar ✪> but most servers had skins in server browser
19:29 <+bridge> <Jupstar ✪> just not all
19:29 <+bridge> <Jupstar ✪> well i dunno xd
20:23 <+bridge> <heinrich5991> not implemented yet, the problem is on our side: https://github.com/ddnet/ddnet/pull/5960
20:28 <+bridge> <murpi> Anyone interested in picking up https://github.com/ddnet/ddnet/issues/5440 ? It'd complement the new ddnet url protocol and help people to report others more easily
20:39 <+bridge> <Sylex> @Pathos
20:45 <+bridge> <Robyt3> What should the server info looks like? Like in the closed PR but without the moderator?
20:45 <+bridge> <Robyt3> What should the copied server info looks like? Like in the closed PR but without the moderator?
20:56 <+bridge> <murpi> Good question 😄 My initial proposal was just to copy the server address, but maybe someone can think of something else?
21:01 <+bridge> <Jupstar ✪> wasnt there a pr for that already
21:01 <+bridge> <Jupstar ✪> #5487
21:01 <+bridge> <DDNet> https://github.com/ddnet/ddnet/pull/5487
21:02 <+bridge> <Jupstar ✪> but he rq xd
21:02 <+bridge> <Jupstar ✪> we were to ungracefully
21:03 <+bridge> <ChRIstMas`iPro> Is there any way to compile DDNet-Server to run on ARM architecture on a raspberry pi? The android mehtods won't work. Any guidance would be appreciated
21:03 <+bridge> <ChRIstMas`iPro> Is there any way to compile DDNet-Server to run on ARM architecture on a raspberry pi? The android methods won't work. Any guidance would be appreciated
21:05 <+bridge> <Jupstar ✪> just compile
21:05 <+bridge> <Jupstar ✪> and enable fsigned-char
21:05 <+bridge> <Jupstar ✪> my client crashes very often when i spec someone on kog servers. anyone know why?
21:06 <+bridge> <Sylex> debug
21:06 <+bridge> <Jupstar ✪> ok it crashes all 5 seconds now, but steam version
21:06 <+bridge> <Jupstar ✪> wtf going on
21:08 <+bridge> <Jupstar ✪> normally core dumps are generated, for some reason not this time :/
21:11 <+bridge> <Jupstar ✪> opengl 3.0 crashing, who broke it xd
21:11 <+bridge> <Jupstar ✪> maybe mesa drivers broken ill try software rendering
21:13 <+bridge> <Jupstar ✪> software runs fine, rip 
21:13 <+bridge> <Jupstar ✪> now i try zink to confirm
21:25 <+bridge> <Jupstar ✪> ok zink tries to use amdvlk for whatever reason and fails, nvm too lazy i just blame mesa
21:26 <+bridge> <Learath2> Mh, if it reliably crashes, can't you just run it in gdb?
21:27 <+bridge> <Jupstar ✪> yes it crashes somewhere in the driver
21:28 <+bridge> <Learath2> Maybe you can try rolling back mesa?
21:28 <+bridge> <Jupstar ✪> i can just try the flatpak version
21:28 <+bridge> <Learath2> I guess software working fine pretty much confirms it's a driver issue, but still
21:29 <+bridge> <Learath2> How to spec is the software renderer anyway?
21:29 <+bridge> <Jupstar ✪> flatpak also runs fine
21:29 <+bridge> <Jupstar ✪> what my english is too bad to understand this sentence
21:30 <+bridge> <Learath2> As in how closely does the software renderer follow the opengl spec? Could it be that the software renderer is more relaxed than the actual gl3 spec?
21:31 <+bridge> <Jupstar ✪> i dunno, but it works with a different mesa version anyway
21:31 <+bridge> <Jupstar ✪> so i assume it was just that
21:31 <+bridge> <Jupstar ✪> im on debian sid, so its quite possible to sometimes get unstable stuff
21:31 <+bridge> <Learath2> Yeah, different mesa working is I guess good enough to rule out a mistake on our part
21:32 <+bridge> <heinrich5991> might still be UB what we do?
21:32 <+bridge> <Jupstar ✪> might also simply be the driver
21:32 <+bridge> <Jupstar ✪> i am not starting to debug some random driver issue xD
21:32 <+bridge> <Jupstar ✪> opengl 3.0 is default and works for pretty much 2k players
21:32 <+bridge> <Jupstar ✪> so i can live with it being a mesa driver bug
21:33 <+bridge> <Learath2> I honestly don't know how one tracks down things lime this in gl. There isn't some verification layer of sorts iirc
21:33 <+bridge> <Learath2> s\/lime/like/
21:34 <+bridge> <Jupstar ✪> opengl spec allows u to do almost all shit anyway, it basically only crashes when u upload invalid buffers
21:34 <+bridge> <Jupstar ✪> source: opengl spec xd
21:34 <+bridge> <heinrich5991> really. I thought it was a huge state machine that mustn't be poked the wrong way
21:34 <+bridge> <Jupstar ✪> opengl is bloat just remove it
21:35 <+bridge> <Jupstar ✪> but what do u want to do to trigger this "wrong way"?
21:35 <+bridge> <Jupstar ✪> it generates invalid_operation if smth is wrong
21:35 <+bridge> <Jupstar ✪> it usually doesnt crash
21:36 <+bridge> <Learath2> But it is possible that there is a missing `return invalid_operation` in the driver. That wouldn't make our mistake any less of a mistake. The driver and us might be wrong at the same time
21:37 <+bridge> <Jupstar ✪> once upon a time, i checked all opengl returns
21:37 <+bridge> <Jupstar ✪> #2553
21:37 <+bridge> <DDNet> https://github.com/ddnet/ddnet/pull/2553
21:39 <+bridge> <Learath2> We probably should do something like that enabled by a debug flag so we dont get a billion branches
21:39 <+bridge> <Jupstar ✪> btw i also ran all backends with zink with vulkan debug layers at max
21:40 <+bridge> <Jupstar ✪> in theory™️ they should have caught UB
21:40 <+bridge> <Learath2> Tho, it might even be negligible with modern branch prediction being so good
21:41 <+bridge> <Learath2> Mh, doesn't that entail some translation to vulkan which might be masking an issue we might be triggering in the gl state machine that would actually be running when we directly render with gl?
21:41 <+bridge> <Jupstar ✪> but zink implements opengl
21:41 <+bridge> <Jupstar ✪> how is that different to a driver implementing opengl
21:42 <+bridge> <Learath2> Well not very different. I'm just saying it has the exact same issue of not being sure whether the issue is in the driver or our code
21:42 <+bridge> <Jupstar ✪> i see
21:43 <+bridge> <Learath2> Zink e.g. might be ignoring an illegal state change and just emiting some vulkan
21:43 <+bridge> <Jupstar ✪> but if 10 drivers work and 1 one not
21:43 <+bridge> <Jupstar ✪> then it might be more realistic that the 1 implemented it wrong
21:43 <+bridge> <heinrich5991> then it might be the fault of our code
21:43 <+bridge> <heinrich5991> that often happens with ocmpilers and libcs etc.
21:43 <+bridge> <heinrich5991> dunno about graphics stuff though
21:43 <+bridge> <Jupstar ✪> then its the fault of the spec not being clear
21:43 <+bridge> <Learath2> Yeah it is a good indication. I just am interested in how one would rigorously debug something like this
21:43 <+bridge> <heinrich5991> then we shouldn't use C/C++ ^^
21:43 <+bridge> <Jupstar ✪> u cant
21:44 <+bridge> <Jupstar ✪> then change that
21:44 <+bridge> <Learath2> E.g. when I debugged the sdl issue on retina screens I ended up in AppKit assembly. Where I 100% verified that the issue is indeed in SDL
21:44 <+bridge> <Jupstar ✪> question is:
21:44 <+bridge> <Jupstar ✪> does it crash in the driver?
21:44 <+bridge> <Jupstar ✪> or does it crash on the GPU?
21:45 <+bridge> <Jupstar ✪> in this case its clearly the driver, so it should be debuggable
21:45 <+bridge> <Jupstar ✪> (or easier debuggable)
21:45 <+bridge> <Learath2> Honestly, not sure if that'll help with this, the rust wrapper around the gl state machine might also suffer from a similar issue, at some point unsafe rust code is needed and it's implementor is also human
21:45 <+bridge> <heinrich5991> this wasn't about graphics, it wouldn't help there
21:46 <+bridge> <heinrich5991> was referring to that and my statement that this does indeed happen with compilers
21:46 <+bridge> <heinrich5991> C++ spec is not very nice to code against
21:46 <+bridge> <Learath2> How does one go about that? Do vendors give some symbols like microsoft does for the ntkernel?
21:46 <+bridge> <Jupstar ✪> lets face it the problem is:
21:46 <+bridge> <Jupstar ✪> opengl is a fat as spec nobody actually cares about anymore, bcs no games use it in the real world
21:46 <+bridge> <Jupstar ✪> and u can add as many testing as u want, some weird threading issue can kill everything
21:47 <+bridge> <Learath2> The thing that saved me in appkit was the fact that it's in objective c. Which has surprisingly pleasent to read assembly with all the indirect calling 😄
21:47 <+bridge> <Jupstar ✪> mesa is open source so you have debug symbols
21:47 <+bridge> <Jupstar ✪> but yeah nvidia gives symbols afaik
21:47 <+bridge> <Jupstar ✪> with their nvidia gpu debugging tool i used once that didnt help at all xD
21:50 <+bridge> <Jupstar ✪> https://www.khronos.org/opengl/wiki/Unexpected_Results
21:50 <+bridge> <Jupstar ✪> "If you use OpenGL in a way that is not covered by the specification, there is no guarantee for any specific result. This can usually be avoided by checking for errors. Note that some implementations are less restrictive than the specification about what is accepted (notably NVIDIA implementations)."
21:50 <+bridge> <Jupstar ✪> 
21:50 <+bridge> <Jupstar ✪> i did that, checking for all errors, so this is probably the best effort i can do
21:50 <+bridge> <Jupstar ✪> if i dont want to waste my life
21:53 <+bridge> <Jupstar ✪> but honestly crashing a driver is different to having UB for opengl
21:53 <+bridge> <Jupstar ✪> 
21:53 <+bridge> <Jupstar ✪> just try out random shit with opengl it might not render smth but u will probably not be able to crash it
21:53 <+bridge> <Jupstar ✪> completely different to vk
21:56 <+bridge> <Jupstar ✪> : CommandLine Error: Option 'help-list' registered more than once!
21:56 <+bridge> <Jupstar ✪> LLVM ERROR: inconsistency in registered CommandLine options
21:56 <+bridge> <Jupstar ✪> zsh: IOT instruction (core dumped)  ./DDNet
21:56 <+bridge> <Jupstar ✪> 
21:56 <+bridge> <Jupstar ✪> 
21:56 <+bridge> <Jupstar ✪> this for example is a llvm bug xddd
21:56 <+bridge> <Jupstar ✪> i remember why i switched to amdvlk
21:57 <+bridge> <Learath2> How common is vk nowadays? Did we switch to it as default yet?
21:57 <+bridge> <Jupstar ✪> we are trying next release
21:57 <+bridge> <Jupstar ✪> but honestly now that AMD fixed the bug i reported its not so important anymore
21:57 <+bridge> <Jupstar ✪> yes they fixed it, so it was probably not a bug in ddnet
21:57 <+bridge> <Jupstar ✪> funny isnt it
21:58 <+bridge> <Learath2> Do we fallback if no vk available?
21:58 <+bridge> <Jupstar ✪> yes
21:58 <+bridge> <Jupstar ✪> the good thing is, except one vulkan problem we havent had a single report yet that wasnt fixable somehow
21:58 <+bridge> <Jupstar ✪> and the reported one was fixed by updating driver
21:59 <+bridge> <Jupstar ✪> so i guess the vulkan drivers are more mature from the start
21:59 <+bridge> <Learath2> It'd be nice if we had some telemetry on this, should we maybe append a gl3 or vk onto the version string so we can track adoption?
22:00 <+bridge> <Jupstar ✪> mhh generally telemetry about config would be nice, then we'd know what we can drop xD
22:00 <+bridge> <Jupstar ✪> 
22:00 <+bridge> <Jupstar ✪> as long as we dont send strings its porbably pretty anon
22:01 <+bridge> <Jupstar ✪> but the word is in a very bad spotlight
22:01 <+bridge> <Jupstar ✪> just saying it will trigger someone
22:01 <+bridge> <heinrich5991> we could add opt-in telemetry
22:02 <+bridge> <Jupstar ✪> tahts useless lets be honest xD
22:02 <+bridge> <heinrich5991> debian has it and it's not useless
22:02 <+bridge> <Jupstar ✪> the 5 users that activate it i can ping in this chat
22:02 <+bridge> <Jupstar ✪> whenever an app asked me, i opt'd out xD
22:02 <+bridge> <heinrich5991> see, then don't work against the user
22:03 <+bridge> <heinrich5991> if the user doesn't want to be tracked, don't track them
22:03 <+bridge> <Jupstar ✪> see but thats the point uploading a crash report is completely different to configs
22:03 <+bridge> <Jupstar ✪> whatever mozilla uploads probably more sensitive than ddnet data can be
22:03 <+bridge> <heinrich5991> "people are doing worse things" is not an excuse in my book
22:03 <+bridge> <Jupstar ✪> in fact the unencrypted ddnet ingame chat is the easiest thing to track ppl
22:04 <+bridge> <heinrich5991> yea, let's fix that
22:04 <+bridge> <Jupstar ✪> well i called it, so no point in discussing
22:04 <+bridge> <Jupstar ✪> i can also not call it telemetry, but config cloud
22:04 <+bridge> <Jupstar ✪> then everyone is happy
22:05 <+bridge> <heinrich5991> if it's opt-in, sure
22:05 <+bridge> <Jupstar ✪> is our discord plugin opt in?
22:05 <+bridge> <Jupstar ✪> no bcs nobody would care then
22:05 <+bridge> <Jupstar ✪> just accept it heinrich
22:06 <+bridge> <Jupstar ✪> opt-in doesnt work on large scale
22:06 <+bridge> <heinrich5991> discord's activity tracking is opt-in IIRC
22:06 <+bridge> <Jupstar ✪> if u opt-in by giving everyone 10€ then maybe yes
22:06 <+bridge> <heinrich5991> on discord's side
22:06 <+bridge> <Jupstar ✪> discord's secret data selling is not opt-in
22:06 <+bridge> <heinrich5991> .
22:06 <+bridge> <Jupstar ✪> accepting reality is an excuse in my book
22:07 <+bridge> <heinrich5991> there's concentration camps in china
22:07 <+bridge> <heinrich5991> I won't accept that in my country
22:07 <+bridge> <Jupstar ✪> well maybe i was not specific enough
22:07 <+bridge> <Jupstar ✪> accepting that these camps exists is realism
22:08 <+bridge> <Jupstar ✪> that doesnt mean u have to like it
22:08 <+bridge> <heinrich5991> I realize that discord tracks users very much
22:08 <+bridge> <heinrich5991> that "realizing" doesn't have anything to do with "tracking ddnet users more" IMO
22:08 <+bridge> <Jupstar ✪> so nobody would activate their useless tracking system xD
22:08 <+bridge> <Jupstar ✪> they know that
22:09 <+bridge> <Jupstar ✪> anyway
22:09 <+bridge> <Jupstar ✪> to end this useless discussion, before we add opt-in telemetry, lets not add anything
22:09 <+bridge> <Jupstar ✪> bcs opt-in telemetry for configs is useless
22:09 <+bridge> <Jupstar ✪> realism
22:10 <+bridge> <heinrich5991> opt-in telemetry for debian wasn't useless
22:10 <+bridge> <Jupstar ✪> what do they use it for?
22:10 <+bridge> <heinrich5991> perhaps it's just because it has a much larger user base
22:11 <+bridge> <heinrich5991> https://popcon.debian.org/
22:11 <+bridge> <Jupstar ✪> e.g. if i want to drop a config based on 100 telemetry samples, then it will defs trigger 100 ppl that didnt opt-in xd
22:12 <+bridge> <Learath2> Well I dont propose tracking users. Would gfx backend being part of the version string really be a privacy concern even?
22:12 <+bridge> <Jupstar ✪> it allows more "tracking" than any other config we have(that isnt a string)
22:13 <+bridge> <Jupstar ✪> realism
22:46 <+bridge> <heinrich5991> yea, seems to be okay to me
23:01 <+bridge> <Jupstar ✪> troll