07:52 < bridge_> <ryozuki> morning
07:52 < bridge_> <ryozuki> back to work :pepeW:
09:14 < bridge_> <Assa> morning, same :pepeW:
09:15 < bridge_> <Assa> @fokkonaut one info for tunings: tunings are applied post tick, so You can enter a tune zone with old physics and skip through them if fast enough. This is the reason you can skip bouncy tunings for example
09:36 < bridge_> <learath2> https://cdn.discordapp.com/attachments/293493549758939136/1364142722130382950/1bNR5tqtadbrlrY-MTjjFOg.png?ex=68089869&is=680746e9&hm=0eca73a26650b8c201d9ee3acd1b8203bb9073060c91ae50eb45fc0516b34d84&
09:46 < bridge_> <Assa> I have a question: Is MapIO supposed to save invalid/unused data or not? So if I map invalid tiles, save, then load the map, should the invalid tiles still be there?
09:49 < bridge_> <Assa> Well you check for valid ids but don't reset the tiles but send the ID to 0 🤔 time for a PR
09:50 < bridge_> <Assa> Well you check for valid ids but don't reset the tiles but set the ID to 0 🤔 time for a PR
09:50 < bridge_> <robyt3> Depends on whether you allow using unused tiles in the map
09:55 < bridge_> <zerolacqua> I remember kebs worked on this in #9786
09:55 < bridge_> <DDNet> https://github.com/ddnet/ddnet/pull/9786
09:56 < bridge_> <matodor> why everyone skip this screenshot? has anyone seen something like this before?
09:58 < bridge_> <syrinok> Skins should permanently stay in your skins folder when added to favorites / be removed when unfavorited. Seems odd that your favorite skins can just disappear from your skins @learath2 fix :greenthing:
09:59 < bridge_> <zerolacqua> It's weird. but I haven't seen this, so I have no idea. :oop:
09:59 < bridge_> <Assa> thank you, time to make this obsolete when saving 🙈
09:59 < bridge_> <Assa> hmm should actually ask in a popup
10:01 < bridge_> <matodor> https://cdn.discordapp.com/attachments/293493549758939136/1364149112819286077/hook_demo.demo?ex=68089e5d&is=68074cdd&hm=c4b687df91ee724899671c8fe3e93cdaf18f2f82ee042b79ba4851164be1581e&
10:02 < bridge_> <syrinok> it's well known, the graphic of the hook detaches beyond a certain distance
10:03 < bridge_> <matodor> max distance?
10:03 < bridge_> <syrinok> I don't know the details, nor why
10:04 < bridge_> <kebscs> This pr is ready
10:04 < bridge_> <learath2> Send 300€ in bitcoin and I fix
10:04 < bridge_> <syrinok> If I had to guess, performance reasons, or in the original code no one expected anyone to hook someone that far
10:05 < bridge_> <syrinok> sent
10:05 < bridge_> <gorp_tw> is this related to why hook teleport looks awful?
10:05 < bridge_> <Assa> OH this is not merged
10:05 < bridge_> <Assa> And I was like "WHERE ARE KEBS CHANGES"
10:08 < bridge_> <syrinok> Nope, the only map that has had this issue is that one fun map where you dodge things going up
10:09 < bridge_> <syrinok> You usually don't hook far enough for it to be a problem in maps normally, you'd usually only see it if the map allows you to get outside of it
10:09 < bridge_> <Assa> no it's not :pepeW:
10:20 < bridge_> <Jupstar ✪> @syrinok how do they disappear? Is that new?
10:21 < bridge_> <syrinok> Always been like that afaik, if you don't have the skin downloaded, and simply favorited, they disappear since they're just temporarily stored from the DB
10:21 < bridge_> <Jupstar ✪> Not really
10:21 < bridge_> <Jupstar ✪> All skins from db are downloaded
10:21 < bridge_> <Jupstar ✪> There is no concept such as temporary
10:21 < bridge_> <syrinok> They disappear from the skins tab all the time though?
10:22 < bridge_> <Jupstar ✪> No
10:22 < bridge_> <Jupstar ✪> Not for me
10:22 < bridge_> <syrinok> I thought the whole point of the skin DB and auto downloading was to avoid having a bunch of skins downloaded, so the game didn't take 10 years to launch
10:22 < bridge_> <Jupstar ✪> That is only about lazy loading, which is the default in the next release anyway
10:23 < bridge_> <Jupstar ✪> Favorite skins are loaded tho
10:23 < bridge_> <Jupstar ✪> Else someone broke it
10:23 < bridge_> <syrinok> I see, well favorites disappear / skins disappear from the skin list when fresh launch
10:23 < bridge_> <syrinok> And it's always been like this from what I've seen
10:23 < bridge_> <Jupstar ✪> @robyt3 Was this changed
10:23 < bridge_> <Jupstar ✪> I'm 99.99% sure I made it so that favorite skins are shown in list
10:24 < bridge_> <syrinok> I have lolhair and other skins favorited
10:24 < bridge_> <syrinok> https://cdn.discordapp.com/attachments/293493549758939136/1364154813662761022/image.png?ex=6808a3ac&is=6807522c&hm=793f35fa5cff5730468cd8539ca9f4515974cac94b9059f833883d68c54245b7&
10:24 < bridge_> <syrinok> only this one shows up, cause I have it downloaded
10:24 < bridge_> <Jupstar ✪> Mh ok
10:24 < bridge_> <syrinok> and in my skins folder
10:24 < bridge_> <syrinok> connecting my dummy that uses lolhair
10:24 < bridge_> <syrinok> https://cdn.discordapp.com/attachments/293493549758939136/1364154999067770971/image.png?ex=6808a3d8&is=68075258&hm=b63380f3cb7fa7627667f0cc7aa34e51f88daf4e8915bca6c1f673dbe79c49bb&
10:25 < bridge_> <syrinok> It stays until I close the game
10:25 < bridge_> <robyt3> Favourites should always be shown in the list, with the next release the main and dummy skins will also always be loaded
10:25 < bridge_> <Jupstar ✪> Maybe it got buggy then?
10:25 < bridge_> <Jupstar ✪> @syrinok do you use nightly?
10:25 < bridge_> <syrinok> I'm on nightly, yes
10:26 < bridge_> <syrinok> This has been my experience with the skin system for awhile now though, I'm quite positive
10:26 < bridge_> <Jupstar ✪> Yeah alr, then it broke some day most likely
10:29 < bridge_> <totar> how does a skin undownload at all? regardless of favorite status
10:29 < bridge_> <totar> I thought if it shows up in your list it's on your disk
10:29 < bridge_> <totar> I thought if it shows up in the list it's on your disk
10:30 < bridge_> <robyt3> Downloaded skins are not shown in the list currently unless they are used or favorited
10:30 < bridge_> <totar> ah ok
10:30 < bridge_> <totar> sounds easy to fix then
10:31 < bridge_> <robyt3> But once we have a skin index for individual databases we can also show all skins of the database in the client immediately
10:31 < bridge_> <totar> like index for community skin urls?
10:33 < bridge_> <Jupstar ✪> If we are ok with switching to a micro service instead of using nginx, then I'd gladly make this happen.
10:33 < bridge_> <Jupstar ✪> 
10:33 < bridge_> <Jupstar ✪> I wanted a better preview for the skin discord bot anyway some day. The new discord bot required could then do that too
10:36 < bridge_> <totar> I think you can have both services at the same time without any issues, the index can pull all the skins from the skindb?
10:36 < bridge_> <robyt3> Like a file of all skin names and properties so the client can show that list, otherwise the client doesn't know which skins are available in each database
10:37 < bridge_> <totar> ok I'm noticing the "download all" button on the web page gets them individually 💀 so I guess it's not trivial
10:37 < bridge_> <totar> https://cdn.discordapp.com/attachments/293493549758939136/1364158059697668147/image.png?ex=6808a6b2&is=68075532&hm=c6ff914a14d80dbe57d5c02d47a8274e2b56cc4158093c83725bc5a24111e2b5&
10:37 < bridge_> <Jupstar ✪> Ah it's more about changing the PHP upload script, which I don't want to touch xD
10:39 < bridge_> <totar> why do you need to change the upload script?
10:40 < bridge_> <Jupstar ✪> To generate these asset indices
10:40 < bridge_> <totar> what about all the skins already uploaded?
10:40 < bridge_> <Jupstar ✪> They currently use one big index
10:40 < bridge_> <Jupstar ✪> But besides that, I also want to add better meta data like mulitple authors etc.
10:41 < bridge_> <totar> oh man, the skins page does not lazy load the skin images. You just get 25mb of skins everytime you visit the page
10:41 < bridge_> <Jupstar ✪> It should lazy load them
10:42 < bridge_> <Jupstar ✪> Well maybe it broke lol
10:43 < bridge_> <Jupstar ✪> That would explain some crashes with the debugger xd
10:45 < bridge_> <totar> what does this mean? where does the data even go?
10:45 < bridge_> <totar> is there a SQL server?
10:45 < bridge_> <Jupstar ✪> A json file
10:48 < bridge_> <totar> hmm ok
10:49 < bridge_> <Jupstar ✪> https://skins.ddnet.org/skin/skins.json
10:49 < bridge_> <totar> so what does this get replaced with?
10:49 < bridge_> <totar> you want to get skin data indivudally?
10:50 < bridge_> <Jupstar ✪> No, but one index per skin type (community, normal, hd, non-hd)
10:50 < bridge_> <totar> oh ok
10:50 < bridge_> <totar> ok but that doesn't sound that hard
10:51 < bridge_> <Jupstar ✪> Depends on what you call hard. It's not technically hard. But it requires to have a different upload process
10:51 < bridge_> <totar> but you already have all the data you need to generate those indexes? no?
10:52 < bridge_> <Jupstar ✪> Depends again on the goal xd
10:52 < bridge_> <Jupstar ✪> 
10:52 < bridge_> <Jupstar ✪> I'd add multiple authors support, maybe tags & description at the same moment
10:52 < bridge_> <Jupstar ✪> Something more future proof
10:52 < bridge_> <totar> I suppose
10:53 < bridge_> <totar> but you can just take the current json and rearrange it how you want as an ad-hoc step seemingly pretty easily
10:53 < bridge_> <Jupstar ✪> Yes, probably need a script for that anyway to not reupload all currently existing skins
10:59 < bridge_> <murpi> @robyt3 looks like 58a6b722958698ebe2af4372c41718ab5bd2b550 is causing the issue mentioned in #10095
11:00 < bridge_> <DDNet> https://github.com/ddnet/ddnet/issues/10095
11:00 < bridge_> <totar> this file is pretty small btw, I'd be very confident a good LLM can handle changes.
11:02 < bridge_> <totar> this file is pretty small and self contained btw, I'd be very confident a good LLM can handle changes.
11:04 < bridge_> <Jupstar ✪> Yes, but there is yet another reason I want a micro service.
11:04 < bridge_> <Jupstar ✪> I want to host a hashed version of the skin (adding the hash to the name)
11:04 < bridge_> <Jupstar ✪> and non hashed for compat and easier finding/typing.
11:04 < bridge_> <Jupstar ✪> 
11:04 < bridge_> <Jupstar ✪> That's also what ddnet-rs uses, so it's already coded anyway
11:05 < bridge_> <Jupstar ✪> Using an index could completely skip the download request we currently do using the hash information
11:06 < bridge_> <totar> ah good point
11:47 < bridge_> <Assa> Starve and do a rust PR or go shopping?
11:47 < bridge_> <learath2> go shopping
11:47 < bridge_> <Assa> but but but 😦
11:49 < bridge_> <kebscs> no rust
12:12 < bridge_> <patiga> ddnet with outlines :D
12:12 < bridge_> <patiga> https://cdn.discordapp.com/attachments/293493549758939136/1364182130711724144/image.png?ex=6808bd1d&is=68076b9d&hm=befd42f72f1d4768f1efb354310af566f208da3821ca1363582c7503b5a14b66&
12:13 < bridge_> <patiga> (probs a driver bug with my web renderer, very noticable when you squish the canvas into very small space)
12:40 < bridge_> <patiga> @jupeyy_keks I made some input/startup improvements to https://mapview.patiga.eu/ (though I didn't improve the startup time, only opened an issue ^^ https://gitlab.com/Patiga/twgpu/-/issues/7)
12:40 < bridge_> <patiga> it *should* now *maybe* work a little better on mobile
12:40 < bridge_> <patiga> though I found out that mobile firefox is a little buggy with the touch input
12:42 < bridge_> <patiga> and could you please take a look at the mipmapping currently in twgpu? I feel like it selects a too high-res level
12:42 < bridge_> <patiga> (as in: please take a look at the rendered output, not any code)
12:44 < bridge_> <oxyzo.> hi developers can we have a thing to be able to change where the screenshot go
12:49 < bridge_> <patiga> not user friendly, but you could create a sym-link from your desired screenshot folder to the config-dir screenshot folder. this way the screenshots will be automatically redirected to your desired location by windows. you'd probably need to use a terminal for that tho
12:53 < bridge_> <Assa> now I have the Pizza and RedBull to do Rust even more efficient 💪
12:54 < bridge_> <Assa> @patiga you've seen my (typo) PR?
12:57 < bridge_> <patiga> yes. I'm not the hugest fan of typo PR's, but I honestly don't know why.
12:57 < bridge_> <patiga> could you just adjust the pr title to smth like `tools/check-ddnet: Fix typo`?
12:57 < bridge_> <patiga> (then I'll just merge)
12:58 < bridge_> <Assa> How else am I supposed to fix typos? It's a typo visible to everyone running into this issue and it annoys me, because bad english
12:58 < bridge_> <Assa> my english not yellow from egg
12:59 < bridge_> <patiga> like I said, I don't know why I'm not the biggest fan of single typo PR's. I'll merge :)
13:00 < bridge_> <Assa> I would totally understand, if it was just a typo in the codebase with no effect what-so-ever
13:01 < bridge_> <patiga> merged ⚡
13:02 < bridge_> <Assa> Merci ❤️
13:05 < bridge_> <Assa> Should I add `faulty speedtiles` to the bot/twmap-ddnet-check or would you prefer generalizing it with `faulty_tele` to `faulty_tiles` ?
13:06 < bridge_> <Assa> Dobby needs guidance from master :owo:
13:09 < bridge_> <Assa> Oh I think putting them togheter doesn't make sense, one of these are invalid, the others are logical mapping errors
13:12 < bridge_> <Jupstar ✪> looks kinda cool, but the image is really low res xd
13:13 < bridge_> <Jupstar ✪> Since it was mostly the first start I could still imagine it's about downloading the wasm blob (?)
13:13 < bridge_> <Jupstar ✪> Or why else is it cached in later runs
13:14 < bridge_> <Jupstar ✪> For me it was buggy on both browsers with input
13:14 < bridge_> <Jupstar ✪> I could only zoom out by using a physical mouse
13:14 < bridge_> <learath2> @zwelf2 if you can generate an hour or two it'd be nice if you can form an opinion on the two teehistorian PRs I have open
13:15 < bridge_> <zwelf2> ok, will try to do in evening
13:18 < bridge_> <patiga> yea, another option is fine. At some point I'll try to merge some of those (e.g. the faulty hookthrough/tiles with invalid rotation are also candidates)
13:19 < bridge_> <patiga> (gonna piggyback here and also note that my pr might also be interesting ^^)
13:20 < bridge_> <patiga> buggier than before?
13:21 < bridge_> <patiga> with touch inputs, could you try out if placing one non-moving finger, and then doing the pinch motion only with the second finger works better?
13:21 < bridge_> <Jupstar ✪> Was, so didn't test new yet, didn't know you changed smth
13:23 < bridge_> <patiga> honestly this is a mystery to me. the wasm blob is 'only' 3 MB, it shouldn't take long
13:24 < bridge_> <Jupstar ✪> Maybe your VPS was dosed by all sudden requests xd
13:24 < bridge_> <learath2> @sans._. when judge me for my js
13:24 < bridge_> <Jupstar ✪> But yeah loading a map on my phone rn and it takes ages again
13:25 < bridge_> <Some Moderator> yeah I'm getting to that, rn I'm packing but first I'm deep cleaning this hunk of metal
13:25 < bridge_> <Some Moderator> https://cdn.discordapp.com/attachments/293493549758939136/1364200460986220564/rn_image_picker_lib_temp_8fb674f2-7798-4257-bf36-039c127a92a8.jpg?ex=6808ce2f&is=68077caf&hm=44a538db00d99b7fc8a0e3d1e401bee015fb06c0e89fa023fa43619d65b838dd&
13:25 < bridge_> <Some Moderator> I need to ve ready with my stuff by tomorrow noon
13:25 < bridge_> <Some Moderator> :heh:
13:25 < bridge_> <Some Moderator> I need to be ready with my stuff by tomorrow noon
13:26 < bridge_> <Assa> BeQuiet
13:26 < bridge_> <learath2> no u
13:26 < bridge_> <Assa> :justatest:
13:27 < bridge_> <patiga> but thats cloudflares job :c
13:28 < bridge_> <Some Moderator> What, it's on the tier list
13:28 < bridge_> <Some Moderator> I at least _tried_ :(
13:28 < bridge_> <Assa> you can see it on the picture, I am just stating it's presence 😄
13:28 < bridge_> <Some Moderator> lol
13:28 < bridge_> <Assa> you can see it in the picture, I am just stating it's presence 😄
13:29 < bridge_> <Some Moderator> reddit PSU shaming has given me ptsd
13:30 < bridge_> <Assa> let's not part shame here, you can apparently play ddnet on the windows 1986 communist edition
13:30 < bridge_> <Some Moderator> tf
13:30 < bridge_> <Some Moderator> dk that's a thing
13:30 < bridge_> <Assa> ^ 🤣
13:31 < bridge_> <Some Moderator> ddnet for xbox when
13:32 < bridge_> <Assa> :justatest: shouldn't be too hard, actually
13:33 < bridge_> <Some Moderator> I expect a prototype by may 5th then :heh: I'm getting a minifridge
13:33 < bridge_> <Jupstar ✪> @totar Did you notice the `NETMSG_INPUTTIMING` msg from the server is not flushed?
13:33 < bridge_> <Jupstar ✪> 
13:33 < bridge_> <Jupstar ✪> Doesn't this affect prediction quite a bit? I see often 20ms wrong timings because of this (at least I assume so)
13:33 < bridge_> <Jupstar ✪> Probably bcs it only gets send when another package causes a flush then (mostly snapshots)
13:33 < bridge_> <Jupstar ✪> 
13:33 < bridge_> <Jupstar ✪> Or does that not matter for the client?
13:33 < bridge_> <Jupstar ✪> @patiga still not loading, after 5 minutes of waiting xD
13:33 < bridge_> <Jupstar ✪> On both browsers
13:34 < bridge_> <Jupstar ✪> Was kingsleap removed from testing?
13:34 < bridge_> <Jupstar ✪> Oh it loaded, apparently my browser history removed the case sensitivity
13:35 < bridge_> <Jupstar ✪> @patiga now the rendering is broken for me
13:35 < bridge_> <Jupstar ✪> In old versions it worked fine
13:35 < bridge_> <patiga> wat
13:35 < bridge_> <Jupstar ✪> In the android internal browser it works
13:35 < bridge_> <Jupstar ✪> https://cdn.discordapp.com/attachments/293493549758939136/1364203073295089735/Screenshot_20250422-133549_Fennec.png?ex=6808d09e&is=68077f1e&hm=04ee7cbdb576a0059059f8047edfcfb657838755430a63b9b0e2dc475ac8ebeb&
13:36 < bridge_> <patiga> ah true, I updated the wgpu version, something could've broken by that
13:36 < bridge_> <Jupstar ✪> I dunno
13:36 < bridge_> <Jupstar ✪> It only breaks when zooming
13:37 < bridge_> <patiga> so it looks alright at first, but when you zoom stuff starts disappearing?
13:37 < bridge_> <Assa> I know this map :owo:
13:37 < bridge_> <Jupstar ✪> https://cdn.discordapp.com/attachments/293493549758939136/1364203478611791882/screen-20250422-133721.mp4?ex=6808d0fe&is=68077f7e&hm=ac2f9c6062475b6ff1455cc9592b7f97557cc211a1d7e6d63183a8cfabab1332&
13:37 < bridge_> <Jupstar ✪> @patiga
13:38 < bridge_> <patiga> heh, that was also that way for me on the older version
13:38 < bridge_> <patiga> iirc
13:38 < bridge_> <patiga> but on desktop it still works, right?
13:38 < bridge_> <patiga> (also in browser)
13:39 < bridge_> <Jupstar ✪> I mean it also works in the other mobile browser
13:39 < bridge_> <patiga> yea that is so weird, probs something that would need to be fixed in wgpu
13:40 < bridge_> <patiga> though it says in the webgl support that off-canvas rendering isn't really supported or smth along those lines
13:40 < bridge_> <patiga> and I do that for the mipmaps
13:40 < bridge_> <patiga> I mean this looks like all the mipmaps are just missing
13:40 < bridge_> <Jupstar ✪> On desktop it works everywhere
13:40 < bridge_> <patiga> darn
13:41 < bridge_> <patiga> is the input at least better?
13:41 < bridge_> <patiga> on mobile
13:41 < bridge_> <Jupstar ✪> Yes
13:41 < bridge_> <patiga> hm
13:41 < bridge_> <Jupstar ✪> @patiga here
13:41 < bridge_> <patiga> W(?)
13:41 < bridge_> <Jupstar ✪> This was old image from fennec
13:41 < bridge_> <Jupstar ✪> So it worked in old versions
13:41 < bridge_> <patiga> yea crap
13:42 < bridge_> <patiga> when I did usb-debugging for this issue I didn't see anything in the console
13:43 < bridge_> <patiga> I'll try that again
13:46 < bridge_> <patiga> honestly no clue how to debug that
13:46 < bridge_> <patiga> It kinda looks like its only using the top-most mip level here, from the pattern of the white bricks
13:47 < bridge_> <Jupstar ✪> I assume you cannot compile to native android yet?
13:47 < bridge_> <Jupstar ✪> Else I could debug with renderdoc probs
13:47 < bridge_> <patiga> never tried it
13:48 < bridge_> <patiga> hm though I could try out running it with proper webgpu in the browser
13:48 < bridge_> <patiga> does mobile chrome support that nowadays already?
13:49 < bridge_> <patiga> oh it does
13:49 < bridge_> <patiga> hmm, thats worth a shot :)
13:50 < bridge_> <Jupstar ✪> Oh wow, we broke compat to vanilla 0.6, sick progress
13:50 < bridge_> <Jupstar ✪> Oh wow, we broke compat to vanilla 0.6 clients, sick progress
13:50 < bridge_> <patiga> o.o with what?
13:52 < bridge_> <Jupstar ✪> I dunno, I just noticed using the libtw2 downloader test thing that old servers send Character info and new servers only send DdnetCharacter info
13:54 < bridge_> <Jupstar ✪> Well maybe it's also just a libtw2 bug
13:54 < bridge_> <Jupstar ✪> it often says `item decode error IntOutOfRange`
13:54 < bridge_> <Jupstar ✪> And in some snapshots the Character info is in
13:55 < bridge_> <patiga> isn't there some sort of client feature detection?
13:57 < bridge_> <Jupstar ✪> Probably, but apparently something broke compat with at least libtw2 xD
13:57 < bridge_> <Jupstar ✪> Either itself or the server
13:59 < bridge_> <teero777> yippee
14:02 < bridge_> <Jupstar ✪> @patiga apparently new servers set the ammo_count to -1
14:02 < bridge_> <Jupstar ✪> and libtw2 refuses that xD
14:02 < bridge_> <Jupstar ✪> Dunno if vanilla is also so strict about that tho
14:04 < bridge_> <Jupstar ✪> Where even is heino
14:04 < bridge_> <Jupstar ✪> He is perma afk
14:12 < bridge_> <learath2> I can't believe that every time I use Rust I find something to annoy me
14:12 < bridge_> <learath2> It's like the entire language is purposefully designed to lack the things I want from a language
14:13 < bridge_> <Jupstar ✪> At least tell us what annoys you
14:20 < bridge_> <Jupstar ✪> @patiga probably a webgpu thing, that is out of your control, but in chromium I actually get:
14:20 < bridge_> <Jupstar ✪> https://cdn.discordapp.com/attachments/293493549758939136/1364214197344534528/image.png?ex=6808dafa&is=6807897a&hm=86b29711a87071a9f6d0f05952d5a94e4b9c58bb09d3be0b23e864ba739d26e4&
14:20 < bridge_> <Jupstar ✪> So it doesn't work in chromium rn
14:21 < bridge_> <milkeeycat> Can you give a list of things you want from a language(asking for a friend)?
14:23 < bridge_> <Some Moderator> @learath2 I'm just now taking a first look at the tool, may I ask wtf?
14:24 < bridge_> <Some Moderator> It's so slow XD
14:24 < bridge_> <Some Moderator> Also I may have bent a few (3) pins on my mobo and had to bend em back <a:9889pain:1190395129958240376>
14:24 < bridge_> <Some Moderator> it looks like im not cooked but we'll see
14:25 < bridge_> <learath2> ```c++
14:25 < bridge_> <learath2> void read_int(std::integral auto &i) {
14:25 < bridge_> <learath2>     constexpr int width = std::numeric_limits<std::remove_reference_t<decltype(i)>>::digits;
14:25 < bridge_> <learath2> }
14:25 < bridge_> <learath2> ```
14:25 < bridge_> <learath2> This snippet that took me 20 seconds to think up in C++ is surprisingly annoying to reproduce with rust generics
14:25 < bridge_> <learath2> Rust generics somehow make me miss SFINAE, and I hate SFINAE
14:25 < bridge_> <Jupstar ✪> Wtf does it even do xD
14:26 < bridge_> <learath2> It lets you pass in any integer type, and I can figure out and use it's width
14:26 < bridge_> <Jupstar ✪> Well you need a trait that has what you need on your objects
14:27 < bridge_> <Jupstar ✪> std::integral is some kind of trait too?
14:27 < bridge_> <patiga> will push a fix in a sec
14:27 < bridge_> <Jupstar ✪> (Or constraint as they call it in fancy cpp)
14:27 < bridge_> <learath2> Yes, which doesn't seem to be in std, or in the num-traits crate. So I now need implement it myself, or at that point I might aswell just copy and paste it 4 times for `read_i8` `read_i16` `read_i32` `read_i64`
14:28 < bridge_> <learath2> Concept 😛
14:28 < bridge_> <Jupstar ✪> So the std sucks
14:28 < bridge_> <Jupstar ✪> Accepted, I also think it lacks some QOL stuff
14:32 < bridge_> <learath2> I guess I can use `num-traits::PrimInt` and determine the bitwidth by doing 8 * `std::mem::size_of`
14:33 < bridge_> <Assa> @patiga 
14:33 < bridge_> <Assa> ```rust
14:33 < bridge_> <Assa> .filter(|(_, speedup)| (speedup.type != 0 && speedup.force == 0))
14:33 < bridge_> <Assa> ```
14:33 < bridge_> <Assa> `435 |         .filter(|(_, speedup)| (speedup.type != 0 && speedup.force == 0))
14:33 < bridge_> <Assa>     |                                         ^^^^ expected identifier, found keyword`
14:33 < bridge_> <learath2> Cursed markdown
14:33 < bridge_> <Assa> type is the property name, but also a keyword
14:34 < bridge_> <Assa> @patiga 
14:34 < bridge_> <Assa> ```rust
14:34 < bridge_> <Assa> .filter(|(_, speedup)| (speedup.type != 0 && speedup.force == 0))
14:34 < bridge_> <Assa> ```
14:34 < bridge_> <Assa> ```435 |         .filter(|(_, speedup)| (speedup.type != 0 && speedup.force == 0))
14:34 < bridge_> <Assa>     |                                         ^^^^ expected identifier, found keyword
14:34 < bridge_> <Assa> ```
14:34 < bridge_> <Assa> is it renamed and how am I supposed to know
14:35 < bridge_> <learath2> `speedup.r#type` time
14:36 < bridge_> <Assa> it's renamed to id, thanks to lerath the error now contains the info what types are available 🙂
14:36 < bridge_> <learath2> 😄
14:36 < bridge_> <Assa> it's renamed to id, thanks to learath the error now contains the info what types are available 🙂
14:38 < bridge_> <patiga> if that is the macro, you could also try writing it without the macro
14:39 < bridge_> <Assa> ```
14:39 < bridge_> <Assa>     let invalid_speedup_count = speedup_tiles
14:39 < bridge_> <Assa>         .indexed_iter()
14:39 < bridge_> <Assa>         .filter(|(_, speedup)| (speedup.id != 0 && speedup.force == 0))
14:39 < bridge_> <Assa>         .filter(|(yx, _speedup)| {
14:39 < bridge_> <Assa>             maybe_print_path(path);
14:39 < bridge_> <Assa>             println!("Invalid (maybe invisible) speedup at x: {}, y: {} {}", yx.1, yx.0, "Force 0 is forbidden!");
14:39 < bridge_> <Assa>             true
14:39 < bridge_> <Assa>         })
14:39 < bridge_> <Assa>         .count();
14:39 < bridge_> <Assa> ```
14:39 < bridge_> <Assa> 
14:39 < bridge_> <Assa> Please explain for rust beginners :justatest:
14:39 < bridge_> <Assa> ```rust
14:39 < bridge_> <Assa>     let invalid_speedup_count = speedup_tiles
14:39 < bridge_> <Assa>         .indexed_iter()
14:39 < bridge_> <Assa>         .filter(|(_, speedup)| (speedup.id != 0 && speedup.force == 0))
14:39 < bridge_> <Assa>         .filter(|(yx, _speedup)| {
14:39 < bridge_> <Assa>             maybe_print_path(path);
14:39 < bridge_> <Assa>             println!("Invalid (maybe invisible) speedup at x: {}, y: {} {}", yx.1, yx.0, "Force 0 is forbidden!");
14:39 < bridge_> <Assa>             true
14:39 < bridge_> <Assa>         })
14:39 < bridge_> <Assa>         .count();
14:39 < bridge_> <Assa> ```
14:39 < bridge_> <Assa> 
14:39 < bridge_> <Assa> Please explain for rust beginners :justatest:
14:50 < bridge_> <Assa> does `cargo build` not overwrite existing executables?
14:50 < bridge_> <Assa> I see, that a .d file has an up-to-date timestamp, but the executable was not updated
14:52 < bridge_> <patiga> cargo build overwrites executables
14:52 < bridge_> <Assa> I had to delete it and rebuild it, otherwise clap was not adding the new option
14:52 < bridge_> <patiga> did you maybe switch up debug/release?
14:53 < bridge_> <Assa> no, I don't have a single debug build
14:53 < bridge_> <patiga> you can also use `cargo run --bin twmap-check-ddnet -- <args>` to directly run the newly built binary
14:55 < bridge_> <Assa> ```
14:55 < bridge_> <Assa> $ ./twmap-check-ddnet.exe --faulty-speedup ~/AppData/Roaming/Teeworlds/maps/KingsLeap.map 
14:55 < bridge_> <Assa> 
14:55 < bridge_> <Assa> $ ./twmap-check-ddnet.exe --faulty-speedup ~/AppData/Roaming/Teeworlds/maps/Cansada_invalid.map 
14:55 < bridge_> <Assa> Invalid (maybe invisible) speedup at x: 394, y: 66 Force 0 is forbidden!
14:55 < bridge_> <Assa> Invalid (maybe invisible) speedup at x: 518, y: 68 Force 0 is forbidden!
14:55 < bridge_> <Assa> Invalid (maybe invisible) speedup at x: 531, y: 137 Force 0 is forbidden!
14:55 < bridge_> <Assa> Invalid (maybe invisible) speedup at x: 486, y: 429 Force 0 is forbidden!
14:55 < bridge_> <Assa> Invalid (maybe invisible) speedup at x: 487, y: 429 Force 0 is forbidden!
14:55 < bridge_> <Assa> Invalid (maybe invisible) speedup at x: 679, y: 483 Force 0 is forbidden!
14:55 < bridge_> <Assa> ```
14:55 < bridge_> <Assa> 
14:55 < bridge_> <Assa> It's working 🥳
14:55 < bridge_> <Assa> should call it speedtile instead
14:55 < bridge_> <patiga> I think speedtile is the common terminology
14:55 < bridge_> <patiga> *speedup
14:56 < bridge_> <Assa> speedup-tile? 🤓
14:56 < bridge_> <patiga> I think speedup is fine
14:56 < bridge_> <Assa> well okay I think it's clear enough like this
15:02 < bridge_> <Assa> PR is ready ❤️
15:05 < bridge_> <Assa> it's written like that in order to make the change that allows force 0 for the new speedtiles easy and readable
15:06 < bridge_> <Assa> it's written like that in order to make the change that allows force 0 for the new speedtiles easy and readable, if we ever come around to do this ofc 😄
15:20 < bridge_> <Assa> yet another backend incoming?
15:20 < bridge_> <Jupstar ✪> That is no backend
15:22 < bridge_> <Assa> it shows as a graphics backend, what is glew then?
15:22 < bridge_> <patiga> can you try again? you can also now use the url parameter `backend=webgpu`/`backend=webgl` to force a backend
15:22 < bridge_> <Jupstar ✪> It's an OpenGL extensions loader
15:23 < bridge_> <Jupstar ✪> That's cool
15:23 < bridge_> <Jupstar ✪> Give it a few minutes
15:23 < bridge_> <Jupstar ✪> :lol:
15:23 < bridge_> <patiga> on my linux chromium, the webgpu backend was hella slow
15:23 < bridge_> <Jupstar ✪> It works for you? Did you have to manually enable it?
15:24 < bridge_> <patiga> though on linux/chromium, webgpu support is still experimental
15:24 < bridge_> <patiga> I enabled it via the unstable cli parameter `--enable-unsafe-webgpu`
15:25 < bridge_> <patiga> NVM
15:25 < bridge_> <patiga> I broke it I think
15:25 < bridge_> <patiga> lemme fix the site
15:25 < bridge_> <patiga> hrmpf
15:27 < bridge_> <patiga> it works locally :c
15:30 < bridge_> <Jupstar ✪> https://cdn.discordapp.com/attachments/293493549758939136/1364231891607093288/image.png?ex=6808eb75&is=680799f5&hm=7cd577710861f37763efeec885686cf333035dde31e99c3bf7d1c8090da701e7&
15:30 < bridge_> <Jupstar ✪> @patiga
15:30 < bridge_> <patiga> yea I failed to fix that one
15:30 < bridge_> <patiga> reload works tho
15:31 < bridge_> <patiga> I thought I only get that locally with too fast loading xd
15:31 < bridge_> <patiga> but for you the website sorta works? it won't do anything for me rn
15:31 < bridge_> <patiga> `Uncaught LinkError: import object field '__wbg_requestAdapter_5e45466d2792e15f' is not a Function`
15:35 < bridge_> <Jupstar ✪> Yes, but the above error persists.
15:43 < bridge_> <kebscs> :D
15:43 < bridge_> <kebscs> https://cdn.discordapp.com/attachments/293493549758939136/1364235135368826956/image.png?ex=6808ee7a&is=68079cfa&hm=2a4e9df635a7148ad3708ded00664234ea580961fa14f78569e8cc9101664217&
15:47 < bridge_> <patiga> and you didn't have that error before right?
15:47 < bridge_> <patiga> why the hell is everything falling apart now xd
15:47 < bridge_> <patiga> I can't even load the site, and you get that weird error
15:47 < bridge_> <patiga> should I just let run spin until the window says its width and height are non-zero?
15:48 < bridge_> <Jupstar ✪> I think it worked few weeks ago
15:48 < bridge_> <Jupstar ✪> I dunno xd
15:48 < bridge_> <Jupstar ✪> It works now btw
15:48 < bridge_> <Jupstar ✪> Have you changed smth?
15:48 < bridge_> <patiga> NO
15:49 < bridge_> <patiga> and its still just as broken for me xd
15:49 < bridge_> <patiga> wth
15:49 < bridge_> <Jupstar ✪> Ok it only works sometimes
15:49 < bridge_> <patiga> ok I'll restart my pc for a quick sanity-check
15:49 < bridge_> <learath2> New Rust deficiency landed, no generic bidirectional iterator trait
15:50 < bridge_> <learath2> Enough winning for one day, I'll go do something else for a bit
15:50 < bridge_> <Jupstar ✪> DoubleEndedIterator or smth?
15:50 < bridge_> <Jupstar ✪> https://doc.rust-lang.org/std/iter/trait.DoubleEndedIterator.html
15:50 < bridge_> <learath2> That's an iterator that can iterate from both ends simultaneously
15:50 < bridge_> <Jupstar ✪> What do you want? Linked list?
15:51 < bridge_> <Jupstar ✪> Back and forth?
15:51 < bridge_> <learath2> Back and forth aye
15:51 < bridge_> <Jupstar ✪> That will probably never exist in Rust
15:51 < bridge_> <Jupstar ✪> At least not for mutable stuff
15:51 < bridge_> <learath2> It probably can't because you can probably create multiple mutable refs with it
15:51 < bridge_> <learath2> Thanks Rust, very cool
15:51 < bridge_> <Jupstar ✪> xdd
15:52 < bridge_> <patiga> @learath2 pls open https://mapview.patiga.eu/ for a second and tell us what error you get in the console :d
15:52 < bridge_> <learath2> You want the "This isn't actually an error" error?
15:52 < bridge_> <patiga> not that one
15:52 < bridge_> <patiga> the site just works for you`?
15:53 < bridge_> <learath2> I don't get any other error, it works fine
15:53 < bridge_> <Jupstar ✪> He uses a browser written in Rust, bcs he likes Rust so much
15:53 < bridge_> <patiga> I'll blame cloudflare caches maybe?
15:53 < bridge_> <Jupstar ✪> I got this in firefox exactly once btw. After reload it then magically worked
15:54 < bridge_> <learath2> Hm, I do get a `WebGPU supported: false` at the start, maybe I'm getting a different codepath?
15:54 < bridge_> <learath2> idk what you are debugging
15:54 < bridge_> <patiga> I can't open the site cause of some uncaught link error. jupii gets the surface error with width/height = 0, Learath just doesn't bother and gets happy path
15:54 < bridge_> <patiga> site doesn't work at all for me, it works perfectly for you
15:54 < bridge_> <patiga> I get `Uncaught LinkError: import object field '__wbg_requestAdapter_5e45466d2792e15f' is not a Function`
15:55 < bridge_> <patiga> @jupeyy_keks okay please clarify: you got the linkerror once, then after reload the surface error?
15:56 < bridge_> <Jupstar ✪> No, after the reload it worked, but firefox only
15:56 < bridge_> <Jupstar ✪> chromium is now completely broken
15:56 < bridge_> <patiga> with the link error
15:56 < bridge_> <Jupstar ✪> Actually just reloaded and worked xD
15:56 < bridge_> <Jupstar ✪> But at least _mostly_ broken
15:56 < bridge_> <learath2> Chrome works fine for me too
15:56 < bridge_> <patiga> was it broken with the surface or link error on chromium @jupeyy_keks
15:56 < bridge_> <Jupstar ✪> With surface stuff, but rn it magically works, dunno
15:57 < bridge_> <Jupstar ✪> Yeah only surface problem
15:57 < bridge_> <patiga> wooo now it also works for me again
15:57 < bridge_> <patiga> and on chrome you should actually get the webgpu backend now
15:57 < bridge_> <Jupstar ✪> Dunno rn it works
15:57 < bridge_> <patiga> the linkerror might've actually just been some cloudflare caching I suppose
15:57 < bridge_> <Jupstar ✪> "map_inspect_web.js:2665 WebGPU supported: false"
15:58 < bridge_> <patiga> on chrome?
15:58 < bridge_> <patiga> chromium?
15:58 < bridge_> <learath2> ```
15:58 < bridge_> <learath2> map_inspect_web.js:1736 Failed to create WebGPU Context Provider
15:58 < bridge_> <learath2> imports.wbg.__wbg_requestAdapter_5e45466d2792e15f
15:58 < bridge_> <learath2> ```
15:58 < bridge_> <learath2> I do get this on chrome
15:58 < bridge_> <Jupstar ✪> But i can try with this
15:58 < bridge_> <learath2> but then it says `WebGPU supported: false` and seems to bail to a different codepath that works
15:58 < bridge_> <patiga> chrome doesn't officially support webgpu for linux
15:58 < bridge_> <Jupstar ✪> @patiga I defs still get the Surface error
15:58 < bridge_> <Jupstar ✪> It's completely random
15:59 < bridge_> <patiga> on mobile chrome it should be supported
15:59 < bridge_> <patiga> your pc is too fast or smth
15:59 < bridge_> <patiga> buy slower pc
15:59 < bridge_> <Jupstar ✪> Believe me, I already did that :lol:
16:00 < bridge_> <learath2> enabling it by hand and it still works, so I guess on my specific configuration it always just works 😄
16:00 < bridge_> <patiga> does it say that it uses webgpu in the console?
16:00 < bridge_> <learath2> Yes
16:01 < bridge_> <learath2> ```
16:01 < bridge_> <learath2> AdapterInfo { name: "", vendor: 0, device: 0, device_type: Other, driver: "", driver_info: "", backend: BrowserWebGpu }
16:01 < bridge_> <learath2> ```
16:01 < bridge_> <patiga> is it horribly slow? :D
16:01 < bridge_> <patiga> on my laptop it tries to boil my cpu or smth
16:01 < bridge_> <patiga> might work better with an actual graphics card or smth
16:01 < bridge_> <learath2> Eeeh, it is a bit slow, and it makes my fans go turbo mode 😄
16:01 < bridge_> <patiga> yeaaa
16:01 < bridge_> <learath2> I get 8 entire fps
16:02 < bridge_> <patiga> https://mapview.patiga.eu/?backend=webgl
16:02 < bridge_> <patiga> ^ and with webgl you prob get full fps with no turbo mode?
16:02 < bridge_> <learath2> yep
16:02 < bridge_> <patiga> lovely
16:02 < bridge_> <learath2> gl on top
16:02 < bridge_> <learath2> 💪
16:05 < bridge_> <Jupstar ✪> It completely doesn't work for me again
16:05 < bridge_> <Jupstar ✪> without error
16:05 < bridge_> <learath2> Have you considered getting better hardware?
16:06 < bridge_> <Jupstar ✪> @patiga see it's defs a network thing
16:06 < bridge_> <Jupstar ✪> 
16:06 < bridge_> <Jupstar ✪> Why should it take like over 30s to download them xD
16:06 < bridge_> <Jupstar ✪> https://cdn.discordapp.com/attachments/293493549758939136/1364240949177483335/image.png?ex=6808f3e4&is=6807a264&hm=6fb55be37be4219124a6d93c5d2e23c72870a7f859fe30379496d1de42a81c89&
16:07 < bridge_> <Jupstar ✪> https://cdn.discordapp.com/attachments/293493549758939136/1364241172402405448/image.png?ex=6808f419&is=6807a299&hm=05da3cd61cc9a49aad62581faf62d6429c7cead1685e34e111f0efd11d16fd73&
16:07 < bridge_> <Jupstar ✪> Still downloading xD
16:08 < bridge_> <patiga> did you anger cloudflare?
16:08 < bridge_> <learath2> So blame cloudflare?
16:08 < bridge_> <Jupstar ✪> I blame Patiga.eu
16:08 < bridge_> <patiga> patiga.eu is github pages behind cloudflare
16:08 < bridge_> <Jupstar ✪> More like gitlab
16:08 < bridge_> <patiga> ah right
16:08 < bridge_> <Jupstar ✪> Host it on github xD
16:09 < bridge_> <Jupstar ✪> It's probs a USA server
16:09 < bridge_> <Jupstar ✪> Bcs I have really bad ping to it
16:09 < bridge_> <Jupstar ✪> Oh it loaded
16:09 < bridge_> <Jupstar ✪> Finally
16:09 < bridge_> <patiga> are you browsing via tor or smth?
16:09 < bridge_> <learath2> Add serverside rendering for the first frame
16:09 < bridge_> <Jupstar ✪> https://cdn.discordapp.com/attachments/293493549758939136/1364241763052818515/image.png?ex=6808f4a6&is=6807a326&hm=01f4c1349122383d969b745ef56bac5f3605cc1bf9a7ad516e88cceac7010f55&
16:09 < bridge_> <Jupstar ✪> Only took 2 minutes
16:09 < bridge_> <Jupstar ✪> lmao
16:09 < bridge_> <patiga> ?????
16:09 < bridge_> <patiga> @learath2 you had normal loading times, no?
16:09 < bridge_> <Jupstar ✪> `map_inspect_web.js:2665 WebGPU supported: true`
16:10 < bridge_> <Jupstar ✪> epyc vulkan for web moment
16:10 < bridge_> <learath2> RFC1149 moment
16:10 < bridge_> <Assa> I guess if you want to protect your ip from ddos ...
16:10 < bridge_> <Jupstar ✪> Oh no I had disable cache on
16:10 < bridge_> <learath2> Yeah, pretty much instant
16:10 < bridge_> <Jupstar ✪> Here we go again
16:11 < bridge_> <patiga> I could also try to blame jupii's isp
16:11 < bridge_> <patiga> but no matter where the server is, the throughput should still be okay afaik
16:18 < bridge_> <Jupstar ✪> I blame gitlab
16:21 < bridge_> <Assa> maybe there is a tariff on your bandwidth (:
16:40 < bridge_> <Jupstar ✪> Yes indeed, bcs I just restarted router and changed ip, still sucks
16:41 < bridge_> <Jupstar ✪> And in tor browser it's fast xd
16:41 < bridge_> <Jupstar ✪> Guess I've to get one of those losers that pay for VPN 😬
16:43 < bridge_> <meloƞ> :catsob:
16:45 < bridge_> <patiga> but this is the first website you've had problems with? :o
16:49 < bridge_> <patiga> ( @jupeyy_keks )
16:50 < bridge_> <Jupstar ✪> I don't think I've seen lots of websites loading for 2minutes xd
16:50 < bridge_> <Jupstar ✪> 
16:50 < bridge_> <Jupstar ✪> Defs one of the only ones
17:12 < bridge_> <Assa> I once made a leaflet world-map where I placed ~2500 objects. So I send the client 2500 images to render them 🙂 this was beyond 2 minutes xD optimized now
17:31 < ws-client4> <ChillerDragon> > And in tor browser it's fast xd
17:31 < ws-client4> <ChillerDragon> nobody ever
17:50 < bridge_> <Jupstar ✪> Truest story
18:15 < bridge_> <Assa> @patiga can you look over the PR again? I am a bit nervous, that any map gets released with this bug with the new speedtiles before it's detected and I don't want to break it
18:16 < bridge_> <meloƞ> these maps can always be fixed even after the 2 week period if needed if its game/mapbreaking - i wouldnt worry about it to much
18:17 < bridge_> <Assa> do we fix already released maps? I am currently monitoring staging maps a bit manually
18:17 < bridge_> <Assa> There is at least one map in review currently which contains this bug, where the mapper simply didn't answer yet
18:19 < bridge_> <kebscs> how do i `QuadsSetSubsetFree` on quad thats been loaded with `QuadContainerAddSprite`, or any other way to flip it?
18:25 < bridge_> <meloƞ> we fix already released maps with game/mapbreaking bugs
18:25 < bridge_> <meloƞ> we fix already released maps with game/mapbreaking bugs and if they dont alter already made ranks
18:25 < bridge_> <meloƞ> we fix already released maps with game/mapbreaking bugs and if they dont alter already made ranks (outside of the 2 week fixing period)
18:26 < bridge_> <Assa> how do you know, that this bugfixes don't affect any ranks tho?
18:26 < bridge_> <Assa> I wish I could just validate all existing runs and proof
18:26 < bridge_> <Assa> I wish I could just validate all existing runs and proof ( I know we aren't there yet)
18:28 < bridge_> <meloƞ> teehistorian is an open format, good luck! :kek:
18:28 < bridge_> <meloƞ> 
18:28 < bridge_> <meloƞ> of course you cant check every rank, but checking the top X ranks to see if they use/abuse a gamebreaking/mapbreaking bug is usually enough
18:28 < bridge_> <meloƞ> and of course, the willingness of the mapper to fix it
18:35 < bridge_> <kebscs> in CItems::RenderPickup
18:45 < bridge_> <Assa> I can fix it for them, for free
18:45 < bridge_> <Assa> In fact, I did already
18:55 < bridge_> <patiga> I'm home in a few mins and can merge then
19:10 < bridge_> <patiga> damn I often forget to push my own changes to twmap
19:12 < bridge_> <patiga> I did an illegal force push, but only with the most recent two commits
19:12 < bridge_> <patiga> (sry @essigautomat, but the history was really messed up with the twmap-tools releases)
19:16 < bridge_> <patiga> @essigautomat rlsed twmap-tools 0.4.6
19:18 < bridge_> <Assa> huh?
19:18 < bridge_> <Assa> I assume you managed to fix your issue and we have a beatiful new version of twmap-tools? 🥳
19:20 < bridge_> <Assa> @murpi can you update the bot please, so I can sleep tonight? :owo:
19:21 < bridge_> <murpi> Uhm, yes
19:21 < bridge_> <Assa> loveley, thank you!
19:36 < bridge_> <murpi> twmap-tools should be on 0.4.6 now
19:45 < bridge_> <Jupstar ✪> where is hex viewer chilllerdragon when u need him most
19:47 < bridge_> <kebscs> broke ddnet repo github actions
19:47 < bridge_> <kebscs> https://cdn.discordapp.com/attachments/293493549758939136/1364296661580382293/image.png?ex=680927c7&is=6807d647&hm=56d0bba681e5dfc50611e5072d9dea3a185539e756b8aa3d45a90dbcbf1b1816&
19:47 < bridge_> <kebscs> @patiga
19:48 < bridge_> <robyt3> So much for preventing the CI from breaking by pinning the version :owo:
19:49 < bridge_> <patiga> I thought we pinned with the commit
19:50 < bridge_> <robyt3> But did your force push invalidate the previous commit ID?
19:50 < bridge_> <Jupstar ✪> GitHub has the same GitLab download issues as me
19:50 < bridge_> <Jupstar ✪> :deen_star:
19:50 < bridge_> <patiga> god damn it, I thought I fixed the commit order
19:50 < bridge_> <patiga> I should've double-checked
19:51 < bridge_> <patiga> aaah I get it
19:52 < bridge_> <patiga> I made a local change into a release
19:52 < bridge_> <patiga> -> but didn't push it
19:52 < bridge_> <patiga> then two commits from Assa with more validation
19:52 < bridge_> <patiga> -> no release, but pinned in ddnet
19:52 < bridge_> <patiga> -> then I rebase those commits and order them after the release
19:53 < bridge_> <patiga> eh I could rebase again and restore the original ordering
19:53 < bridge_> <patiga> then the history looks really odd and pretty wrong, but the original commit hashes are correct again
19:54 < bridge_> <patiga> I'll do that I suppose. lesson learned: don't force push, even only a few commits to fix the release history :c
19:54 < bridge_> <patiga> hm, or I quickly create a pr
19:55 < bridge_> <patiga> @robyt3 which option do you prefer?
19:55 < bridge_> <robyt3> I'm fine with updating the hash in the CI as well
19:55 < bridge_> <patiga> I'm fine with either
19:56 < bridge_> <robyt3> Maybe using a tag for a release version instead of hash would be better as well
19:58 < bridge_> <patiga> I prefer hashes. I don't use tags as of yet, so that would be the only usage rn
19:58 < bridge_> <patiga> what would be the advantage of tags?
20:01 < bridge_> <Assa> easily checking out old versions, checking out "stable" versions
20:05 < bridge_> <robyt3> You could have changed which hash the tag points to without having to make a PR in this case
20:09 < bridge_> <patiga> I honestly don't see why my rebase broke that hash, it is more in the past
20:10 < bridge_> <patiga> I just confirmed how I got this result
20:10 < bridge_> <patiga> I only pushed my two local commits right below the commits I just merged
20:10 < bridge_> <patiga> which I thought would be fine
20:11 < bridge_> <patiga> but apparently this broke the chain of merges somehow, and modified the hashes of previous commits
20:13 < bridge_> <patiga> this is the start
20:13 < bridge_> <patiga> https://cdn.discordapp.com/attachments/293493549758939136/1364303055457226923/image.png?ex=68092dbb&is=6807dc3b&hm=6090a5f206656f54c70b17c568976b3373e0d8b26ff39c629d62ad8cf4f204ae&
20:13 < bridge_> <patiga> and I pushed down the top two commits right below 'Fix typo'
20:14 < bridge_> <patiga> and the pinned commit in ddnet is the first Merge of branch feature/new-speedtile
20:19 < bridge_> <patiga> I fixed the commit history, the pipeline should work again @robyt3
21:58 < bridge_> <totar> I don't think it matters, the client calculates the timing based on the data in the packet, not the actual time the packet arrives.
21:58 < bridge_> <totar> I don't think it matters, the client calculates the timing based on the data in the NETMSG_INPUTTIMING , not the actual time the packet arrives.
21:58 < bridge_> <totar> I don't think it matters, the client calculates the timing based on the data in the NETMSG_INPUTTIMING, not the actual time the packet arrives.
21:59 < bridge_> <totar> I suppose it would be slightly more accurate if you flushed it every tick, but I don't think it's nessecary idk
22:16 < bridge_> <totar> > I see often 20ms wrong timings because of this (at least I assume so)
22:16 < bridge_> <totar> Shouldn't these show up pretty clearly in the network debug graph? all NETMSG_INPUTTIMING get charted on it
22:35 < bridge_> <Jupstar ✪> Yeah I mean like ping pong between input and timing ack.
22:35 < bridge_> <Jupstar ✪> 
22:35 < bridge_> <Jupstar ✪> But I assume the client simply doesn't care.
22:35 < bridge_> <Jupstar ✪> If the timing would not be delayed then the client could simply adjust client timings a bit earlier
22:38 < bridge_> <totar> the current system is basically only intended to handle stable latency situations. It makes no attempt to calculate your actual jitter and adjust for it
22:38 < bridge_> <totar> https://cdn.discordapp.com/attachments/293493549758939136/1364339490830946324/409812906-7baab461-5e75-4037-adc5-54bba410556d.png?ex=68094faa&is=6807fe2a&hm=1e13fbd8f61dbfe4dcbb946bc8c7da77858d49f292c38023cc5919893fe187f9&
22:38 < bridge_> <totar> it's like a slightly janky moving average
22:39 < bridge_> <totar> so telling the client that the latency has changed 20ms sooner won't really help it
22:47 < bridge_> <Assa> @jupeyy_keks can you review my render force 0 PR again? I want to be able to tell people to download nightly in order to fix the bug the ddnet bot is telling them to fix 🙈
23:06 < bridge_> <Assa> Thank you ❤️
23:41 < bridge_> <kebscs> #10098
23:41 < bridge_> <DDNet> https://github.com/ddnet/ddnet/pull/10098
23:41 < bridge_> <kebscs> ready for review :)