00:15 <+bridge> <sans._.> @super_ray11
00:15 <+bridge> <sans._.> i dont think ive used opengl ever since vulkan was added
00:15 <+bridge> <sans._.> so basically never
00:17 <+bridge> <sans._.> ill take a server hostage tomorrow and test it
00:17 <+bridge> <sans._.> btw could this be related to https://discord.com/channels/252358080522747904/757720336274948198/1506489024377000047 ?
05:46 <+bridge> <12944qwerty> what's the average memory usage for a server (1-16 players for now)
05:46 <+bridge> <12944qwerty> what's the average memory usage for a server (lets say 1-16 players)
07:25 <+bridge> <super_ray11> i can't recreate the issue, but maybe the drivers are outdated ?
07:26 <+bridge> <sans._.> system is up to date
07:26 <+bridge> <sans._.> https://cdn.discordapp.com/attachments/293493549758939136/1507615628352290846/image.png?ex=6a128bf7&is=6a113a77&hm=27b39c7bfaf604f5fcf123a99de5b192e400c41b0d1c1ab872455ee2db1b57dc&
07:26 <+bridge> <sans._.> idk
07:27 <+bridge> <super_ray11> can you recreate the issue ?
07:28 <+bridge> <sans._.> that's what im trying rn
07:28 <+bridge> <super_ray11> alright gl
07:32 <+bridge> <chillerdragon> githubbing
08:32 <+bridge> <sans._.> I tried on multiple maps
08:32 <+bridge> <sans._.> even the ones I remember crashing while joining
08:33 <+bridge> <sans._.> Maybe it's not a map issue?
08:33 <+bridge> <sans._.> given this, I would guess it's got something to do with the client trying to render players, or process some data about them
08:33 <+bridge> <sans._.> idk
08:57 <+bridge> <milkeeycat> Just learned about https://github.com/matplotlib/matplotlib/pull/31132 today. Clanker wrote a blog post about how it's unfair that his PR didn't get merged, just because he's not a human. Also there were people who were on the clanker's side, it's sad to think they're breathing the same air as the rest of the people.
09:27 <+bridge> <sans._.> damn clankers
09:27 <+bridge> <sans._.> smh
09:31 <+ChillerDragon> how to make sense of fakin windows compile errors
09:32 <+ChillerDragon> they are much harder to read than linux ones, change my mind
09:34 <+bridge> <teero777> it's true
09:34 <+bridge> <teero777> chiller its so nice outside
09:34 <+bridge> <teero777> you should go outside
09:35 <+bridge> <teero777> what error did you get?
09:37 <+ChillerDragon> i messed with namespaces
09:37 <+ChillerDragon> `C:\Program Files\Microsoft Visual Studio\2022\Enterprise\VC\Tools\MSVC\14.44.35207\include\cinttypes(22,25): error C2039: 'imaxdiv_t': is not a member of '`global namespace'' [D:\a\ddnet_base\ddnet_base\debug\ddnet_base.vcxproj]`
09:37 <+ChillerDragon> `  (compiling source file '../src/ddnet_base/base/fs.cpp')`
09:37 <+ChillerDragon> i am just not sure where to look
09:37 <+ChillerDragon> it lists a bunch of filenames and i struggle to find the problematic one given the error
09:38 <+ChillerDragon> i am kinda outside
09:38 <+bridge> <teero777> wdym kinda xddd
09:38 <+ChillerDragon> cant tell xd
09:38 <+bridge> <teero777> are u coding in a greenhouse??
09:38 <+bridge> <sans._.> he's sipping coffee with his macbook on his lap while on his balcony
09:38 <+ChillerDragon> dont want to leak my current location
09:38 <+bridge> <sans._.> germany
09:39 <+bridge> <sans._.> (i think)
09:39 <+ChillerDragon> otherwise the stream snipers start showing up
09:39 <+bridge> <teero777> yea right balcony is specific enough to get your location xdd
09:39 <+bridge> <teero777> oh no
09:39 <+bridge> <teero777> im on a chair outside rn
09:39 <+ChillerDragon> based
09:39 <+bridge> <teero777> pls don't snipe me
09:39 <+bridge> <sans._.> chiller i think you meant team of snipers
09:40 <+bridge> <teero777> it's so sad i would like to tell you how warm it is but I can't cuz I'd dox myself
09:41 <+bridge> <sans._.> germany
09:41 <+bridge> <sans._.> (i think)
09:41 <+bridge> <sans._.> (again)
09:41 <+bridge> <sans._.> if not then my next best guess is planet earth
09:42 <+bridge> <teero777> okay actually i wont doz mysefl
09:42 <+bridge> <teero777> its 22°C it is very nice outside
09:42 <+bridge> <teero777> okay actually i wont dox myself
09:43 <+bridge> <teero777> it seems its similar everywhere in germany rn
09:43 <+bridge> <teero777> it seems it's similar everywhere in germany rn
09:43 <+bridge> <ryozuki> chiller swapping places everyday
09:44 <+bridge> <ryozuki> i think chiller is back to germany
09:44 <+bridge> <ryozuki> cuz he used to talk in asia hours
09:44 <+bridge> <teero777> germany will fall soon
09:44 <+bridge> <teero777> best to get out
09:44 <+bridge> <teero777> but no where to go to
09:44 <+bridge> <ryozuki> im in (s)pain
09:44 <+bridge> <ryozuki> komm
09:44 <+bridge> <teero777> ok
09:47 <+bridge> <sans._.> can I get 4 digits of precision on that pls
09:47 <+bridge> <teero777> my thermometer doesn't show me that sry
09:48 <+bridge> <sans._.> I can get by with 3 digits but I'll be off by a few kilometers
09:49 <+bridge> <teero777> My thermometer isn't even showing the same values as the internet xddd
09:53 <+ChillerDragon> i already triangulated teeros pos anyways
09:53 <+ChillerDragon> using ping to kog servers GER3 and GER4
10:23 <+ChillerDragon> oh boi github is having a bad pipeline day again
10:23 <+ChillerDragon> plus the huuge hack :D 
10:23 <+ChillerDragon> man this site is kinda fkd
10:24 <+bridge> <melon> time to selfhost :owo:
10:28 <+ChillerDragon> i think codeberg is fine
10:28 <+ChillerDragon> @robyt3 wat is an android package name?
10:28 <+ChillerDragon> `str_copy(aTag, ANDROID_PACKAGE_NAME "/");`
10:28 <+ChillerDragon> i am tempted to just set a garbage value here xd
10:29 <+ChillerDragon> https://github.com/ddnet/ddnet/blob/c6eb77167db75bbd91d9cb890236611cd3b96fba/docs/BUILDING-android.md#how-to-build-the-ddnet-client-for-android
10:29 <+ChillerDragon> > The third parameter denotes the package name of the APK.
10:29 <+ChillerDragon> hm
10:29 <+ChillerDragon> the package name is the package name xd
10:30 <+ChillerDragon> i think i will just put my repo name there
10:30 <+bridge> <melon> i like codeberg, used it when Scrumplex told me he uses it
10:30 <+ChillerDragon> `      message(FATAL_ERROR "ANDROID_PACKAGE_NAME must define the package name when compiling for Android (e.g. org.example.app)")`
10:31 <+ChillerDragon> ah there is an example
10:31 <+ChillerDragon> a domain hm
10:32 <+bridge> <robyt3> Android package names are reverse domain notation. The official client is `org.ddnet.client`
10:32 <+ChillerDragon> i wonder what to use in my lib
10:32 <+bridge> <robyt3> Is it going to support starting a server on Android? :justatest:
10:32 <+ChillerDragon> no
10:32 <+ChillerDragon> but it still errored
10:33 <+ChillerDragon> so i need to set it
10:33 <+bridge> <robyt3> Then you don't really need the package name
10:33 <+ChillerDragon> i turned src/base into a lib so i can have standalone projects use str_copy and shit
10:33 <+ChillerDragon> https://github.com/ddnet-community/ddnet_base
10:33 <+ChillerDragon> https://github.com/chillerbot/chillerbot-ux/actions/runs/26327976116/job/77509020296
10:33 <+ChillerDragon> it fails here
10:33 <+ChillerDragon> `/home/runner/work/chillerbot-ux/chillerbot-ux/src/external/chillerbot_reply/src/external/ddnet_base/src/ddnet_base/base/log.cpp:168:19: error: use of undeclared identifier 'ANDROID_PACKAGE_NAME'`
10:33 <+ChillerDragon> so i will just set it
10:33 <+bridge> <robyt3> Ah, true, we added it to logging
10:34 <+bridge> <robyt3> If you are building a library it should be set by users of the library somehow
10:35 <+ChillerDragon> sounds annoying tho right?
10:35 <+ChillerDragon> i want the thing to just work
10:35 <+bridge> <robyt3> Is it supposed to work on Android at all?
10:35 <+ChillerDragon> yea
10:36 <+ChillerDragon> chillerbot-ux also runs on android
10:36 <+ChillerDragon> it might log
10:36 <+ChillerDragon> will it break if it does not match the package name?
10:36 <+ChillerDragon> ah wait hm
10:36 <+bridge> <robyt3> Then you need some way to define the package name for logging, or you can't parse logs because they all go into one global channel
10:36 <+ChillerDragon> it probably wont log eh idk
10:36 <+ChillerDragon> i see
10:36 <+ChillerDragon> who parses these logs?
10:37 <+bridge> <robyt3> Anyone debugging a product based on your library
10:37 <+ChillerDragon> hm
10:37 <+ChillerDragon> that might never happen xd
10:37 <+ChillerDragon> but good to know
10:38 <+ChillerDragon> i will hardcode a value for now and add a note in the cmake
10:38 <+bridge> <robyt3> If you don't need logging you could remove it. Or define a default for `ANDROID_PACKAGE_NAME` if its undefined and only warn users on configuring.
10:38 <+ChillerDragon> i rather not maintain a patch that removes it
10:38 <+ChillerDragon> settting is seems cheaper
10:38 <+ChillerDragon> ok amazing thanks for the info i think i got it now
10:50 <+bridge> <robyt3> And the bug is not happening anymore with any map? Did you also switch back to the graphics backend that isn't working (OpenGl 3.0). I think https://discord.com/channels/252358080522747904/757720336274948198/1506489024377000047 is a different issue and also happened on a different version. The crash in https://discord.com/channels/252358080522747904/293493549758939136/1507440416918732941 seems more like a bug in mesa/libgallium. It crashes when `glD
10:52 <+bridge> <sans._.> I'm down to run whatever setup you want me to, but out of what I tried, nothing crashed, and I tried all opengl versions in the list
10:53 <+bridge> <robyt3> If it's map-related it should crash on the same map as before, `Dark Descent`. If that also works with OpenGL 3.0 now, then it might have been a driver issue and fixed by restarting.
10:55 <+bridge> <robyt3> And search for `GPU version` in F1 on launch. Does it still say `4.6 (Compatibility Profile) Mesa 25.2.3-1ubuntu1`?
10:56 <+bridge> <sans._.> https://cdn.discordapp.com/attachments/293493549758939136/1507668582727946350/image.png?ex=6a12bd49&is=6a116bc9&hm=87eae3601662fda3df793336f440987dc3d227581af3de047bef6e364b06f573&
10:56 <+bridge> <sans._.> this is running via gdb, same as yesterday
10:57 <+bridge> <sans._.> im using opengl 3.0 now
10:58 <+bridge> <sans._.> it happened
10:58 <+bridge> <sans._.> tell me
10:58 <+bridge> <sans._.> what to type
10:58 <+bridge> <sans._.> in gdb
10:58 <+bridge> <sans._.> same map btw
10:58 <+bridge> <robyt3> `bt all`
10:58 <+bridge> <robyt3> and press enter or so to continue when it asks to page
10:58 <+bridge> <sans._.> no symbol "all" in current context
10:58 <+bridge> <melon> `bt full` ?
10:58 <+bridge> <melon> ah that may be too verbose
10:59 <+bridge> <robyt3> If you are already logging to a file then `bt full` should be fine
11:00 <+bridge> <sans._.> https://cdn.discordapp.com/attachments/293493549758939136/1507669640938262528/gdb_thingy?ex=6a12be45&is=6a116cc5&hm=741bd9d63ff12948ac97b10cee9c72681580519f9d629babb3b7fcb68d5449a8&
11:01 <+bridge> <sans._.> https://cdn.discordapp.com/attachments/293493549758939136/1507669788887879741/gdb_thingy.txt?ex=6a12be68&is=6a116ce8&hm=eebb5b767e2c2e4f21e26f10fefe5792fc69d02cab93f25a0a49e4d2b6376d2b&
11:01 <+bridge> <sans._.> @robyt3 ^^
11:03 <+bridge> <robyt3> Thanks, I'll try to investigate later whether those draw counts and offsets make sense
11:04 <+bridge> <sans._.> @robyt3 Should I try to get you a core dump file?
11:04 <+bridge> <sans._.> I'm not gonna lie I'm ai'ing the crap out of this
11:04 <+bridge> <sans._.> but it sounds useful
11:05 <+bridge> <robyt3> Not sure how much it would help, I rarely use debuggers interactively. And it could contain logins etc. in the dump so you shouldn't share it publicly.
11:08 <+bridge> <sans._.> ok
11:08 <+bridge> <marmare_314> Are interfaces like `IMapImages` supposed to be purely virtual or is it fine to include some non-virtual functions?
11:14 <+bridge> <sans._.> @robyt3
11:14 <+bridge> <sans._.> https://cdn.discordapp.com/attachments/293493549758939136/1507673154443677797/gdb_thingy.txt?ex=6a12c18b&is=6a11700b&hm=a42aaeeacca57c4bbd401e2401f622702c4c1ee8892126fe6d835b7437563971&
11:15 <+bridge> <sans._.> According to AI this is supposed to help
11:29 <+bridge> <bunny_55116> ку
11:34 <+bridge> <sans._.> Этот канал предназначен для активного обсуждения разработки клиента DDNet. Язык обсуждения — английский. Для отправки сообщений сюда (или в любой другой канал, кроме #off-topic) я рекомендую использовать этот веб-сервис:
11:34 <+bridge> <sans._.> $deepl
11:34 <+bridge> <DDNet> Hi! Most of us communicate in English. If you’re having trouble with English, use <https://www.deepl.com/en/translator> to help you out.
11:34 <+bridge> <sans._.> Thank u <a:nod:604500081114087434>
12:33 <+bridge> <melon> anything concrete should go into the impl class, not the interface, a non-virtual default is rarely needed.
12:33 <+bridge> <melon> any usecase why you'd consider it?
12:43 <+bridge> <Assa> I'd say as long as it is supported by every `IMapImage` and a very general function it's fine - if it's implementation specific obviously not
12:44 <+bridge> <Assa> we also have an IGraphics interface, while graphics.cpp exists
12:45 <+bridge> <Assa> The functions in graphics.cpp should work for every backend, these are operations on the interface itself
12:51 <+bridge> <marmare_314> I was just thinking about how #10753 could be done. Some functions in `CMapImages` would then also be useful for the editor, so they could be moved to `IMapImages`.
12:51 <+bridge> <DDNet> https://github.com/ddnet/ddnet/issues/10753
12:52 <+bridge> <marmare_314> There is just one implementation of it as far as I can tell 😄
12:53 <+bridge> <Assa> there are people forking ddnet and this just provides a clean interface - just imagine there were two 😄
12:53 <+bridge> <Assa> maybe in 100 years we'll add svg support for example
12:56 <+bridge> <marmare_314> Makes sense, specifically I am talking about these. They don't really depend on how the textures are loaded
12:56 <+bridge> <marmare_314> ```
12:56 <+bridge> <marmare_314>     virtual IGraphics::CTextureHandle GetOverlayBottom() = 0;
12:56 <+bridge> <marmare_314>     virtual IGraphics::CTextureHandle GetOverlayTop() = 0;
12:56 <+bridge> <marmare_314>     virtual IGraphics::CTextureHandle GetOverlayCenter() = 0;
12:56 <+bridge> <marmare_314> ```
12:59 <+bridge> <marmare_314> But moving these to `IMapImages` is not easy since `TextRender()` and `Graphics()` are used for these. But adding them makes inheritance of `CComponent` not work since they also add them.
13:12 <+bridge> <marmare_314> For 10753 I was thinking we could add a new classes/interfaces `IRenderMap`, `IRenderGroup`, `IRenderLayer` that are quite similar to the editor mapitems. The editor mapitems could inherit them, and ingame they would be implemented using `CLayers` and `IMap` (much like it is done currently in `CRenderLayer`). This would provide a common interface without too much rewriting. @essigautomat do you think this would work? (If I even added enough det
14:04 <+bridge> <robyt3> Did you already try updating Mesa? Version 26.1.1 should be available https://archlinux.org/packages/extra/x86_64/mesa/
14:05 <+bridge> <robyt3> According to the logs you are using 25.2.3
14:06 <+bridge> <sans._.> ?
14:06 <+bridge> <sans._.> https://cdn.discordapp.com/attachments/293493549758939136/1507716229908140112/image.png?ex=6a12e9a9&is=6a119829&hm=460fb10aa9fafe73c99924a304740318a5fd57fe4505c9b1648817a1fe7cf00d&
14:06 <+bridge> <robyt3> The crash happens in `/lib/x86_64-linux-gnu/libgallium-25.2.3-1ubuntu1.so`, which is part of Mesa :thonk:
14:08 <+bridge> <robyt3> Ah never mind, you are not on arch, that was another reporter
14:08 <+bridge> <robyt3> Check if `mesa-libgallium` is installed and update it
14:25 <+bridge> <sans._.> latest
14:25 <+bridge> <sans._.> https://cdn.discordapp.com/attachments/293493549758939136/1507721217686241362/image.png?ex=6a12ee4e&is=6a119cce&hm=1447c79e7f8e04a246ca8d895848f4f974ae823fc7f44a34675dace1ab45a940&
14:40 <+bridge> <robyt3> Thanks for testing. Seems like there's nothing you can do except not use OpenGL 3.0 in particular. It's a bug in Mesa according to Jupeyy. See #11033 -- I added your information as it seems like exactly the same crash in `glDrawArrays`.
14:40 <+bridge> <DDNet> https://github.com/ddnet/ddnet/issues/11033
18:37 <+bridge> <alw5> can #3239 be reopened i dont want to create an issue just for it to be closed as duplicate
18:37 <+bridge> <DDNet> https://github.com/ddnet/ddnet/issues/3239
18:39 <+bridge> <alw5> maybe also a bonus the discord bot should be able to take an argument for a key so its not default to c
18:47 <+bridge> <zaramka> # check my bio 😁
19:56 <+ChillerDragon> @alw5 why reopen
19:58 <+ChillerDragon> @alw5 go comment on the issue why you want it open again and im sure some maintainer will reopen it for you
20:03 <+bridge> <alw5> done
22:35 <+bridge> <robyt3> @essigautomat It looks like you didn't add the change to the commit in #11711
22:35 <+bridge> <DDNet> https://github.com/ddnet/ddnet/pull/11711
22:45 <+bridge> <Assa> fixed, thank you, I checked my git history and "git add ." doesn't work if I am in the build directory - I am really tired from yesterday and failed to notice, sorry :/
22:57 <+bridge> <kebscs> https://x.com/hd_nvim/status/2057843081441903062
22:58 <+bridge> <kebscs> rust rewrite takes yet another victim
23:05 <+bridge> <aegisub> well it was already being vibed before the rust rewrite
23:12 <+bridge> <melon> yeah that take is weird, bun is literally an AI-driven repo
23:18 <+bridge> <totar> It makes more sense when you realize they just need a functional Javascript runtime so literally any inconvenience is more than enough justification to throw it out. Idk why they even support multiple, just a gimmick
23:21 <+bridge> <12944qwerty> Finally xdd
23:41 <+bridge> <peter0x44> they should just pick duktake or quickjs