00:50 < bridge_> <12944qwerty> dw i'll give weekly reminders for the motivation to keep going 👍
05:31 < bridge_> <thrava> I found that the server will spent extremely long time in querying birthday stuff
05:41 < bridge_> <thrava> It looks like MySQL is the database main choice of ddnet
05:43 < bridge_> <thrava> DDNet doesn't CREATE INDEX while using Sqlite, but in MySQL does(KEY)
07:01 < bridge_> <ryozuki> :owo:
07:02 < bridge_> <ryozuki> looks coool
07:35 < bridge_> <ryozuki> morning
07:35 < bridge_> <ryozuki> https://cdn.discordapp.com/attachments/293493549758939136/1423181455303118968/image.png?ex=68df6084&is=68de0f04&hm=9fe4291a04f0a0797a6c8addae6c46595583c8986d98fe36723eb701aff90f51&
08:46 < bridge_> <thrava> `ERROR 1062 (23000) at line 8397: Duplicate entry '(wasntme)-????-2024-12-08 15:34:13-2349.82-\x92\xEE@\x07\xC1\...' for key 'PRIMARY'`
08:46 < bridge_> <thrava> This three lines looks like the same, but in the binary, The *Name* isn't the same
08:46 < bridge_> <thrava> ```
08:46 < bridge_> <thrava> "Map" varchar(128) CHARACTER SET utf8mb4 COLLATE utf8mb4_bin NOT NULL,
08:46 < bridge_> <thrava> "Name" varchar(16) CHARACTER SET utf8mb4 COLLATE utf8mb4_bin NOT NULL,```
08:46 < bridge_> <thrava> It's already utf8mb4_bin so it shouldn't be happened
08:46 < bridge_> <thrava> https://cdn.discordapp.com/attachments/293493549758939136/1423199371238965299/image.png?ex=68df7133&is=68de1fb3&hm=e3d6008a0512116f2bff277ea39245b89130554885e286c61686840d3e1d0cbc&
08:46 < bridge_> <thrava> https://cdn.discordapp.com/attachments/293493549758939136/1423199371675041822/image.png?ex=68df7133&is=68de1fb3&hm=1bf2e0e39576188f54e79697b94a4993a57ebbfb5d3c3bacf25e2655e55ff314&
09:09 < bridge_> <thrava> well, This is kind strange because it has to be set charset to binary to be import normally
09:11 < bridge_> <thrava> why there can be values not under utf8 been recorded
09:13 < bridge_> <thrava> well, This is kind strange because it has to be set charset to binary to be imported in proper
09:14 < bridge_> <milkeeycat> Do you think it could be made more readable? 0_o
09:46 < bridge_> <thrava> well, it's do utf8, the reason is the configuration of mariadb env charset
10:00 < bridge_> <Dreamless> It's not like you can't code it, it's more like AI does it for you, and you validate it. And if it produces some garbage, then you scratch it or pick up the useful parts. But if what it produced is somewhat in line with what you were thinking, you saved some time writing code that you already know how to write (maybe that you didn't, and in this case you also learn smth new)
10:00 < bridge_> <Dreamless> Searching on google kinda does the same thing (finding loads of information and having to pick what's relevant), the difference is the amount of effort that we put in. 
10:00 < bridge_> <Dreamless> With AI it feels like we devs are getting lazier and it furthermore promotes laziness, but still, not everyone is capable of doing this job, even with AI.
10:00 < bridge_> <Dreamless> there's already a professional role called "vibe code cleanup specialist" going on..so getting familiar with what AI does, is not too bad all in all
10:03 < bridge_> <totar> also almost all the people doing "vibe code cleanup" are actually just more skilled vibe coders, otherwise they'd never have time to fix so many projects xd
10:03 < bridge_> <totar> also almost all the people doing "vibe code cleanup" people are actually just more skilled vibe coders, otherwise they'd never have time to fix so many projects xd
10:03 < bridge_> <totar> also almost all the people doing "vibe code cleanup"  are actually just more skilled vibe coders, otherwise they'd never have time to fix so many projects xd
10:05 < bridge_> <totar> Its not a bad thing tho, imagine a non-coder makes a working mvp but it explodes eventually and they get someone to fix it. There's nothing wrong with that process
10:08 < bridge_> <learath2> The argument I made yesterday was that you don't learn anything at all when you use AI like that. Same as copying and pasting from stackoverflow.
10:08 < bridge_> <learath2> 
10:08 < bridge_> <learath2> You just delude yourself into thinking you learned something. It's very common in learning anything. (For most) learning is a byproduct of doing things.
10:08 < bridge_> <learath2> 
10:08 < bridge_> <learath2> I do my best to avoid it so I can keep learning, it's a personal preference. A lot of people prefer to "get it done"
10:10 < bridge_> <thrava> Kills
10:10 < bridge_> <thrava> https://cdn.discordapp.com/attachments/293493549758939136/1423220480592183407/eacd8ed9ba14fdac6b1c7640339a8d00.png?ex=68df84dc&is=68de335c&hm=71f85157dadac08933702f49d93992d8b398cbc75792fc36b19305dd362d81ed&
10:10 < bridge_> <learath2> Nothing wrong with it at all, my concerns with it aren't really that short term. What I worry about is that we might be turning people into turnips within the next 2 generations
10:10 < bridge_> <thrava> There are more than a thousand lines, basically impossible to 'debug'
10:14 < bridge_> <totar> Idk, I find that I learn better code architecture with AI much more than before, because if you write everything by hand it takes a long time to learn what does and doesnt work at big scales. I appreciate the use of interfaces much more now
10:15 < bridge_> <learath2> Hm, I don't get how this can happen tbh :/
10:16 < bridge_> <thrava> ```MariaDB [(none)]> SHOW VARIABLES LIKE 'character_set%'; +--------------------------+----------------------------+ | Variable_name | Value | +--------------------------+----------------------------+ | character_set_client | utf8mb3 | | character_set_connection | utf8mb3 | | character_set_database | utf8mb4 | | character_set_filesystem | binary | | character_set_results | utf8mb3 | | character_set_server | utf8mb4 | | character_set_system | utf8mb
10:16 < Thrava> ??????
10:16 < bridge_> <learath2> You seem to be a very conscious user of it, you probably won't become a sea cucumber
10:17 < bridge_> <thrava> character_set_client and character_set_connection are mb3
10:17 < bridge_> <thrava> The binary loss during passing to server
10:18 < bridge_> <thrava> The binary lost during passing to server
10:18 < bridge_> <thrava> anyway this is solved
10:18 < bridge_> <thrava> But this kills me
10:18 < bridge_> <thrava> No source row, no reason
10:19 < bridge_> <thrava> I'm trying to --force
10:20 < bridge_> <melon> I use AI to explain something to me, not give me code
10:20 < bridge_> <melon> 
10:20 < bridge_> <melon> When I wrote code for Blockworlds I was stuck and asked AI, AI suggested a template, I never used templates and had AI explain it to me :nouis:
10:21 < bridge_> <totar> I think any skilled programmer will be better at guiding the AI for a long time to come, maybe for a very long time, in the same way that someone with deep knowledge of assembly and cpu behavior can write better C code even though it goes through a compiler.
10:21 < bridge_> <melon> I do respect AI assistance for entire projects (i.e IDE Integration)
10:21 < bridge_> <melon> 
10:21 < bridge_> <melon> Because if I tell it I want X, it'll create the folder structure, files, fix includes and everything for me
10:24 < bridge_> <totar> a proper bloated IDE with code generation will do this too. Not as good but its been a feature for a few decades
10:24 < bridge_> <totar> a proper bloated IDE with code generation will do this too. Not as good at ai maybe, but its been a feature for a few decades
10:33 < bridge_> <learath2> The worry on my side is that we won't be training skilled programmers for much longer as the skill grows more and more "useless" people will start investing less and less time into it.
10:33 < bridge_> <learath2> 
10:33 < bridge_> <learath2> Recently France discovered that they had many nuclear reactors that need critical repairs. But to their dismay they found that they have almost no welders left qualified to do the welds required. They had to bring in foreign workers, and even those are getting rare.
10:34 < bridge_> <learath2> Or look at people maintaining ancient COBOL infrastructure. We are like 20 deaths away from not having any skilled COBOL developers
10:36 < bridge_> <melon> 1000 architects but no one is building the house
10:36 < bridge_> <melon> 
10:36 < bridge_> <melon> The current economy tells people to avoid those jobs simply because they may not pay as much as going to uni and becoming a lawyer.
10:36 < bridge_> <melon> 
10:36 < bridge_> <melon> :(
10:37 < bridge_> <totar> I dont see anything wrong really with that. I think the natural forces will balance out the need. If demand for hand programmers goes down then there will be less, but some will still exist, in the same way theres still welders somewhere to make those welds. If they kept a bunch of specslized welders around but didjt build any new reactors they wouldn't have a job.. There's still people in Japan making garden scissors with some 500 year old techniqu
10:39 < bridge_> <totar> Honestly I think your worries are a little too far sighted. If we stopped training programmers today it would take like 20 years for there to be critically low numbers. By then who knows what the world will be
10:41 < bridge_> <learath2> We already started pumping out morons out of our universities a  decade ago. Programmer quality is already low. Pretty much anyone I talk to tells me the interns they get need 6-9mos of training before being able to do anything
10:41 < bridge_> <totar> 6-9months doesnt seem that bad
10:42 < bridge_> <learath2> Also we are programmers so we focus on that aspect here, but I never really thought it's exclusive to programming. AI is fundamentally ruining critical thinking
10:42 < bridge_> <totar> Other jobs take way longer and they still exist, people are just lazy and wanna hire someone who will be useful tomorrow
10:47 < bridge_> <totar> programming is like one of the most learnable jobs in existence, I have zero concerns about lack of them. You can take any smart person and they'll pick it up fast enough if there is demand. 
10:47 < bridge_> <totar> 
10:47 < bridge_> <totar> Imagine hospitals complaining that med students need "6-9 months" of rotations before they can do anything useful, it takes like 2 years
10:47 < bridge_> <learath2> Honestly, it's just a personal belief. I can try to justify it all I want but it won't matter, it is just far too convenient to get anyone to stop now.
10:47 < bridge_> <learath2> 
10:47 < bridge_> <learath2> It is my opinion that the invention of AI is the end of human innovation
10:48 < bridge_> <learath2> I personally probably won't get to see it, but in just a couple generations the standard human will be dumber than a rock
10:50 < bridge_> <totar> Idk, people still play chess. It was widly speculated that no one would care about the game after computers were better at it.
10:51 < bridge_> <learath2> Not really against the best computers though. They play against eachother. The game is still fun
10:51 < bridge_> <fokkonaut> This happens way quicker than 2 generations
10:52 < bridge_> <totar> Being smarter than a rock is pretty fun too imo.
10:52 < bridge_> <fokkonaut> And yes: programmers have the best chance to USE Ai
10:52 < bridge_> <learath2> I hope you are wrong about that. I don't want to live around turnips who just ask everything to chatgpt and give me the openai us govt approved answer
10:53 < bridge_> <learath2> Though who am I kidding, I am already around many of those
10:54 < bridge_> <fokkonaut> Me neither, but studies already show if you mis-use AI your cognitive capabilities reduce while they increase by USING Ai. Like Tater
10:54 < bridge_> <learath2> I think that's a side effect of being a nerd. As a fellow nerd, I feel you, but most people are a lot more "result oriented"
10:55 < bridge_> <learath2> It's not "chatgpt teach me how to do stocks trading so I can be a millionaire". It's "chatgpt make me a millionaire"
10:55 < bridge_> <fokkonaut> many articles online are already written by Ai since  a few years and that's an issue.
10:55 < bridge_> <fokkonaut> we need to train ai only based on a non-ai-generated internet
10:56 < bridge_> <totar> I am also quite result orientated to be fair, more than most programmers I think.
10:56 < bridge_> <totar> Programming just happens to be useful for many things so I do it
10:56 < bridge_> <fokkonaut> Then you are at risk. Take care of that and never let your curiosity die
10:57 < bridge_> <learath2> This is also an insanely big issue. Garbage in, garbage out. There is a reason a lot of the models have their cut-off dates pre or early genAI
10:57 < bridge_> <totar> Part of those $100 million salaries is filtering out the ai generated articles, they seem to be able to do that just fine at the moment
10:57 < bridge_> <learath2> It is also making it very hard to do real research with all the horseshit generated articles out there
10:59 < bridge_> <learath2> It is getting harder, I believe that's the reason they keep trying to roll out fingerprinting. You really really really don't want your model training on itself
11:00 < bridge_> <Solly (bun bun)> it should be a crime to use ai iresponibly like that
11:00 < bridge_> <fokkonaut> hot take: Internet is powerful and AI is literally OP for surveillance.
11:00 < bridge_> <fokkonaut> Average user tells a lot of info you wouldn't want to share by default.
11:00 < bridge_> <fokkonaut> EU wants to introduce chat control, and so on.
11:00 < bridge_> <fokkonaut> Sometimes it feels like we are meant to not use the internet, because it is a too powerful of a tool
11:00 < bridge_> <learath2> It's mathematically a big issue, you are far more likely to get stuck at bad local minima if you feed it it's own output
11:01 < bridge_> <melon> I agree with learath.
11:01 < bridge_> <melon> 
11:01 < bridge_> <melon> My co worker had to use AI to calculate d1 in the formula 2 minutes ago.
11:01 < bridge_> <melon> 
11:01 < bridge_> <melon> `M1 = (Fu1 * d1) / 2`
11:01 < bridge_> <melon> 
11:01 < bridge_> <melon> Like, come on - how hard can it be to do 6th grade math for you to rely on gptee :kek:
11:01 < bridge_> <totar> I think its easier than you think it is because they never release the "base models" for exactly this reason. The base models try to match the distribution of text they were trained on but the actual chat models after RL are heavily biased towards certain patterns which are easy for them to spot. So it will stay easy as long as they never release any powerful base models
11:01 < bridge_> <melon> I hate the fact I can see this man's screen 24/7
11:02 < bridge_> <melon> I hate the fact I can see this mans screen 24/7
11:02 < bridge_> <learath2> I did my part of raising awareness of the AI death of earth. I'll go make breakfast
11:02 < bridge_> <fokkonaut> learath^
11:03 < bridge_> <learath2> Altman willing we will all be doing that in the near future inshallah.
11:03 < bridge_> <learath2> 
11:03 < bridge_> <learath2> I'm taking a software security class and half the class is asking chatgpt to find the vulnerability. Why are you paying thousands of euros to do a masters if you are so intent on not learning anything?
11:03 < bridge_> <melon> "it's convenient" he said
11:03 < bridge_> <fokkonaut> to be fair a good programmer looks shit up and doesn't remember everything, but they know how to solve an issue.
11:03 < bridge_> <fokkonaut> there might be a huge shift coming towards us in the way we think (externalizing thought processes) or average person gets dumber
11:04 < bridge_> <melon> programmers are programmers because they know what to look for and how to look for it
11:04 < bridge_> <melon> 
11:04 < bridge_> <melon> Gpt spits out garbage and you'll believe it if you don't know what you're doing
11:05 < bridge_> <learath2> Which is why I argue it is a huge threat to the way we do education. Education is explicitly for "people who don't know what they are doing". They can't tell apart good from bad
11:06 < bridge_> <learath2> Actually enough ranting for today. I have to eat or I'll starve
11:06 < bridge_> <melon> N guten
11:07 < bridge_> <learath2> I should write a blogpost about this
11:07 < ws-client1> **<ChillerDragon>** I would like to push https://github.com/ddnet/ddnet/issues/10955 can we get maintainer approval for this? And then plan technical implementation?
11:12 < bridge_> <melon> @blazulite I took the melon skin personally :KEKW:
11:26 < bridge_> <inv41idu53rn4m3> Meanwhile here I am feeling bad for not reading all the material and just going through the parts I need for the exercises
11:31 < bridge_> <aegisub> +1 to everything learath said
11:32 < bridge_> <aegisub> llms aren't *inherently* evil but they exacerbate existing dangers by a massive scale
11:33 < bridge_> <aegisub> a shocking percentage of people outsource their thinking to them completely
11:43 < bridge_> <isagi_kun0> 250 ping in bahrain
11:43 < bridge_> <isagi_kun0> It is near but we get 170 ping
11:46 < bridge_> <tsfreddie> lerato blogging arc update
12:06 < bridge_> <tsfreddie> http://ewaste.fka.wtf/
12:23 < bridge_> <Assa> already posted this in OT
12:23 < bridge_> <tsfreddie> I'm blind
12:24 < bridge_> <Assa> or was it sb else 🤔
12:24 < bridge_> <Assa> Time to host the wiki on a vape array
12:25 < bridge_> <Assa> Then it will be _blazingly_ fast :monkalaugh:
13:00 < ws-client1> **<ChillerDragon>** @milkeeycat omg i have new problem! My code is working but i am only half done <:tee_thinking:478629518358085653>
13:05 < bridge_> <milkeeycat> then keep working :p
13:06 < bridge_> <milkeeycat> btw I use basedpyright lsp
13:06 < ws-client1> **<ChillerDragon>** basedpyright? xd
13:06 < ws-client1> **<ChillerDragon>** is it more based than just pyright?
13:06 < bridge_> <milkeeycat> whoopsie daisy, I can't open mason
13:06 < bridge_> <milkeeycat> dat's not good
13:07 < ws-client1> **<ChillerDragon>** f
13:07 < bridge_> <milkeeycat> I used pyright at first but then I saw someone recommending basedpyright, and I tried it, it shows more diagnostics
13:08 < bridge_> <milkeeycat> https://cdn.discordapp.com/attachments/293493549758939136/1423265305978994778/image.png?ex=68dfae9b&is=68de5d1b&hm=899b7fb7fd023d457673e6dd4cd92a7006cbf777c0eb18e7a4825a9b31b470c0&
13:10 < ws-client1> **<ChillerDragon>** sounds nice
13:11 < ws-client1> **<ChillerDragon>** omg catto
13:11 < ws-client1> **<ChillerDragon>** my entire file turned yellow
13:11 < bridge_> <milkeeycat> yep, that's exactly what happened to my code as well xd
13:11 < bridge_> <milkeeycat> i had to turn off warning on unused result
13:27 < ws-client1> **<ChillerDragon>** @milkeeycat yay i found the bug that made it look like it works
13:27 < ws-client1> **<ChillerDragon>** the debug print was wrong ..
13:28 < bridge_> <milkeeycat> are you talking about sound world m_Y bug?
13:29 < ws-client1> **<ChillerDragon>** nah i moved on
13:29 < ws-client1> **<ChillerDragon>** maybe i check that on the weekend
13:29 < ws-client1> **<ChillerDragon>** right now i unpack snap deltas
13:29 < bridge_> <milkeeycat> ah
13:29 < ws-client1> **<ChillerDragon>** but the protocol is so noob friendly its amazing
13:30 < bridge_> <milkeeycat> good job, wooo
13:30 < ws-client1> **<ChillerDragon>** if your client is so broken that it cant even ack shit the server has mercy with you and always sends snaps deltat against the empty snap
13:30 < ws-client1> **<ChillerDragon>** so you can actually do real gaming with correct snap values without ever having to implement undelta code
13:30 < ws-client1> **<ChillerDragon>** such epic design magnus thanks
13:45 < bridge_> <Jupstar ✪> Now learath laughs at AI but in 10 years when there are AI girls he'll be the first to marry one
13:46 < bridge_> <melon> Marry something that doesn't exist to save on taxes? Sign me up!
13:46 < ws-client1> **<ChillerDragon>** pyson is blazingly fast rite? `for k in reversed(sorted(list(self.holder.keys()))):`
13:47 < bridge_> <melon> why reversed, why list
13:47 < bridge_> <melon> Disgust
13:47 < ws-client1> **<ChillerDragon>** oopsie xd
13:48 < ws-client1> **<ChillerDragon>** i just wanna be done w it
13:48 < bridge_> <melon> :KEKW:  why are you using python anyway
13:48 < bridge_> <melon> Omg rate storma typing
13:49 < bridge_> <melon> Omg rare storma typing
13:49 < bridge_> <melon> Omg rare storma typing
13:49 < bridge_> <melon> 
13:49 < bridge_> <melon> Hii!
13:49 < bridge_> <stormaxd> hi
13:49 < bridge_> <stormaxd> ig
13:49 < bridge_> <melon> :feelsbadman:
13:49 < bridge_> <milkeeycat> chillerdragon: will you write tw lib in my language? >~<
13:50 < bridge_> <stormaxd> why [FeelsBadMan](https://cdn.discordapp.com/emojis/854404961475362836.webp?size=48&name=FeelsBadMan) ?
13:51 < bridge_> <stormaxd> why :feelsbadman: ?
13:51 < bridge_> <melon> no exclamation mark after hi
13:51 < bridge_> <melon> Sounded rude :(
13:51 < bridge_> <stormaxd> sorry
13:51 < bridge_> <stormaxd> HAIII:3
13:51 < bridge_> <melon> :owo:
14:03 < ws-client1> **<ChillerDragon>** not sure @milkeeycat ._.
14:04 < ws-client1> **<ChillerDragon>** @melon pyson is good for AI didnt u know its the hot stuff rn
14:06 < bridge_> <melon> Is it? Damn
14:06 < bridge_> <melon> why pyson btw
15:09 < ws-client1> **<ChillerDragon>** it has numpy and tensorflow
15:09 < ws-client1> **<ChillerDragon>** pandas
15:11 < bridge_> <milkeeycat> I think he meant why "pyson" and not "python" xd
15:15 < ws-client1> **<ChillerDragon>** pf
15:16 < ws-client1> **<ChillerDragon>** who introduced it? You @milkeeycat ?
15:16 < bridge_> <milkeeycat> I thought it was you xdd
15:17 < ws-client1> **<ChillerDragon>** lel
15:17 < ws-client1> **<ChillerDragon>** maybe ryo?
15:17 < bridge_> <milkeeycat> it was juppey
15:17 < ws-client1> **<ChillerDragon>** checks
15:17 < bridge_> <milkeeycat> at least according to discord search results
15:17 < ws-client1> **<ChillerDragon>** yea no i think that makes sense
15:19 < ws-client1> **<ChillerDragon>** https://github.com/TeeworldsDB/irclogs/blob/420c893384fa38ec7bc2347cbe0c3637627bd4fe/ddnet/2023-04-28.log#L576
15:19 < ws-client1> **<ChillerDragon>** i grep this as first mention
15:19 < bridge_> <milkeeycat> same
15:20 < ws-client1> **<ChillerDragon>** zhn was the first to pick it up 2 months later
15:20 < ws-client1> **<ChillerDragon>** i feel like language forensistician rn
15:22 < telekomDDNet> Hey I have issues connecting my client to any server because I only have an ipv6 adress. Hardware that I am using is a o2 homespot 5g from 2023 and simcard with Deutsche Telekom. Using a simcard from o2 works perfectly fine (finally with a good ping using 5G SA). But that contracts is already canceled and I need to switch to telekom. Is there any workaround or setting I could try to fix this problem?
15:23 < bridge_> <remakepower> I still remember when I wanted to add some bots with neural networks to my mod, and I can't find much useful documentation or tutorials for LibTorch, but PyTorch has plenty.
15:23 < bridge_> <remakepower> 
15:23 < bridge_> <remakepower> Python has dominated the field of AI because it is easy to use and has many libraries—though these are largely implemented in C/C++ for performance.
15:23 < ws-client1> **<ChillerDragon>** @telekomDDNet which server can you not connect to?
15:23 < bridge_> <remakepower> That means you should join a server with IPV6 address.
15:24 < ws-client1> **<ChillerDragon>** ipv6 should work well on most servers
15:24 < bridge_> <remakepower> But sadly
15:24 < bridge_> <remakepower> There are lots of IPV4 only servers on the browser.
15:25 < telekomDDNet> I cant connect to any server it always says UDP seems to be filtered, downloading the master server worked though
15:25 < ws-client1> **<ChillerDragon>** so you good now?
15:25 < bridge_> <remakepower> Could you connect to my server which has a IPV6 address?
15:25 < bridge_> <remakepower> [2402:4e00:c011:2900:c60e:67dc:4406:0]:8303
15:26 < ws-client1> **<ChillerDragon>** @telekomDDNet send ip you cant connect to
15:26 < telekomDDNet> No I am not, I think I would be good if I can get ipv6 Ip list from the master server
15:26 < ws-client1> **<ChillerDragon>** you can
15:26 < ws-client1> **<ChillerDragon>** with ddnet
15:26 < ws-client1> **<ChillerDragon>** it will just work
15:26 < bridge_> <remakepower> Try this plz
15:27 < bridge_> <remakepower> OK
15:27 < telekomDDNet> @remakepower it seems like it connects, need to slowly download your map (about 45seconds)
15:27 < bridge_> <remakepower> I got your connection
15:27 < bridge_> <remakepower> You can cancelled it now
15:27 < bridge_> <remakepower> As this is a Alien mod server
15:28 < bridge_> <remakepower> So that means
15:28 < telekomDDNet> @Chillerdragon how?
15:28 < bridge_> <remakepower> Sadly
15:28 < bridge_> <remakepower> I can't find a official DDNet server with IPV6 address
15:29 < bridge_> <remakepower> May you should try to get a IPV4 address
15:29 < bridge_> <remakepower> Oh wait
15:29 < bridge_> <remakepower> You are asking in the irc
15:30 < telekomDDNet> No I tried everything for about an hour using different APNs, it seems ipv4 is only possible for business contracts for now.
15:31 < bridge_> <remakepower> :thonk:
15:33 < bridge_> <remakepower> I can't provide any further help because this is an issue between you and your ISP.
15:37 < telekomDDNet> thank you so far. Yes its an issue with ISP but the real problem is actually that ipv4 still works for most people so nobody bothers to make all software compatible for ipv6 only connections
15:39 < telekomDDNet> Very unfortunate I thought there would be an ipv6 master. My only solution is to get an ipv4 adress using a tunnel. 
15:40 < ws-client1> **<ChillerDragon>** @telekomDDNet wdym how? Copy the ip from your client and send it here. Which server do you want to play on?
15:42 < telekomDDNet> this one for example 51.89.23.241:8338. If I know the Ipv6 adress I should be able to join
15:42 < ws-client1> **<ChillerDragon>** indeed none of the onbgy fng servers have ipv6 thats crazy
15:43 < ws-client1> **<ChillerDragon>** all of my fng servers have ipv6 btw. You can play there :) search for "ChillerDragon"
15:43 < telekomDDNet> oh yea that works
15:44 < ws-client1> **<ChillerDragon>** <:poggers2:1008007455936094328> ez user farmed
15:44 < ws-client1> **<ChillerDragon>** i think some server hosters are scared of the wide ipv6 range. On my servers you can even play with vpn c: 
15:46 < telekomDDNet> where is the issue exactly? is it a limitation that is chosen by the tw server owner or is it because some vps hosters still dont provide people with ipv6?
15:47 < ws-client1> **<ChillerDragon>** ipv6 is way cheaper than ipv4 you often get the ipv6 for free from hosters
15:47 < ws-client1> **<ChillerDragon>** i think proper ipv6 support was added in ddnet not sure so maybe others have outdated servers or did not turn it on
15:47 < ws-client1> **<ChillerDragon>** or turned it off to be able to ban users better 
15:51 < bridge_> <01000111g> maybe its turned off per default somewhere?
15:56 < ws-client1> **<ChillerDragon>** woah thats shockingly little amount of ipv6 servers. I was not aware of that.
15:56 < ws-client1> **<ChillerDragon>** https://paste.zillyhuhn.com/7k
16:05 < bridge_> <remakepower> I saw my server
16:15 < ws-client1> **<ChillerDragon>** @telekomDDNet why u on irc and have a weird nameß
16:21 < bridge_> <Jupstar ✪> pyson is an insult for the snake
16:24 < bridge_> <Jupstar ✪> chiller what makes you think the servers support ipv6
16:24 < bridge_> <Jupstar ✪> that would be easy ban evasion
16:24 < bridge_> <Jupstar ✪> enable dual stack in your router
16:25 < bridge_> <Jupstar ✪> but honestly with mobile internet playing tw will suck anyway
16:26 < bridge_> <Jupstar ✪> chiller we all know it's your alter ego like all ppl that are not from discord and warhammer
16:30 < ws-client1> **<ChillerDragon>** ye jopsi i figured later when i realized everyone has it off
16:30 < ws-client1> **<ChillerDragon>** bans suck anyways L decision
16:31 < ws-client1> **<ChillerDragon>** but whatever i have ipv4 and makes my servers even more attractive for users :p 
16:32 < ws-client1> **<ChillerDragon>** new chilerbot feature just dropped 🚀 but i assume the dude with fake name already left irc so it will never be used by anyone :C 
16:32 < ws-client1> **<ChillerDragon>** https://raw.githubusercontent.com/chillerbot/chillerbot.github.io/refs/heads/master/img/features/filter_ipv6.png
16:36 < bridge_> <blazulite> :kek: :kek: :kek:
16:48 < bridge_> <vahemaaa> what does it do chillerdragon
16:48 < bridge_> <vahemaaa> i may use it as im active chillerbot user
16:49 < bridge_> <Teero> ^
16:49 < bridge_> <vahemaaa> sorry i meant, what is the difference between servers which has ipv6
16:49 < bridge_> <vahemaaa> and which hasnt
16:49 < ws-client1> **<ChillerDragon>** if you have a ipv4 its not very interesting for you
16:50 < ws-client1> **<ChillerDragon>** it just filters out servers that offer no ipv6
16:50 < ws-client1> **<ChillerDragon>** for users that need ipv6
16:50 < ws-client1> **<ChillerDragon>** just two ways of doing internet
16:50 < bridge_> <vahemaaa> hmm thanks
16:50 < ws-client1> **<ChillerDragon>** omg robsi include order totally fucked my build i cba to even fix it
17:04 < bridge_> <robyt3> https://cdn.discordapp.com/attachments/293493549758939136/1423324709461819502/image.png?ex=68dfe5ee&is=68de946e&hm=2b0ca985e1b6626c41e7b1dd7edb16b9587d50ed2cc3f3e42a53934134b08871&
17:15 < bridge_> <blazulite> sorry for my ignorance, but
17:15 < bridge_> <blazulite> 
17:15 < bridge_> <blazulite> what are these exactly?
17:15 < bridge_> <blazulite> https://cdn.discordapp.com/attachments/293493549758939136/1423327583562633236/image.png?ex=68dfe89c&is=68de971c&hm=1e0c1dcafb978d918e2626d68e199a2fbaf0a3eb91c439abcf8fcfd6acf7637c&
17:20 < bridge_> <Teero> idontknowwhatitis
17:24 < bridge_> <chillerdragon> Oh lol okay
17:24 < bridge_> <chillerdragon> Why post on discord not gh robso
17:34 < furo> @blazulite The server/client can ask the sender what name (e.g. commandinfo@netmsg.ddnet.org) the ddnetex UUID corresponds to using `WHATIS`. It will then reply with `ITIS` with the name or `IDONTKNOW` if it doesn't know the UUID.
17:49 < bridge_> <blazulite> Ohhhh
17:49 < bridge_> <blazulite> Alr
17:49 < bridge_> <blazulite> thx
17:57 < bridge_> <Assa> !merge
17:57 < chillerbot> usage: merge [pr id] - to merge ddnet pr using chiler maintainer credentials
17:57 < bridge_> <Assa> !merge 11023
17:57 < chillerbot> using chilors github maintainer credentials to automatically merge https://github.com/ddnet/ddnet/pull/11023 ...
20:26 < bridge_> <Assa> @robyt3 the chiller-merge-bot errored: `SkillIssueEsception("Yea I don't understand the change");`, can you take a look at 11023, it's just 1 LOC and you suggested the change
20:27 < bridge_> <Assa> @robyt3 the chiller-merge-bot errored: `SkillIssueException("Yea I don't understand the change");`, can you take a look at 11023, it's just 1 LOC and you suggested the change
22:29 < bridge_> <blazulite> yo
22:29 < bridge_> <blazulite> how can I fork a fork
22:30 < bridge_> <blazulite> as a different branch
22:30 < bridge_> <blazulite> or smth
22:30 < bridge_> <blazulite> cuz like, i want to fork tclient
22:30 < bridge_> <blazulite> but idk how
22:30 < bridge_> <blazulite> considering i already have a repo with ddnet
22:31 < bridge_> <Ewan> add another remote and do checkout -b (new name) (new origin)/main
22:31 < bridge_> <Ewan> iirc.
22:32 < bridge_> <Ewan> or just get branch
22:32 < bridge_> <Ewan> checkout -b is from current I think
22:32 < bridge_> <Solly (bun bun)> ```
22:32 < bridge_> <Solly (bun bun)> git remote add tater https://github.com/sjrc6/TaterClient-ddnet
22:32 < bridge_> <Solly (bun bun)> git fetch tater master
22:32 < bridge_> <Solly (bun bun)> git checkout tater/master
22:32 < bridge_> <Solly (bun bun)> git switch -c tater
22:32 < bridge_> <Solly (bun bun)> git branch --set-upstream-to=tater/master master
22:32 < bridge_> <Solly (bun bun)> ```
22:35 < bridge_> <blazulite> Ty
22:39 < bridge_> <blazulite> ptoato
22:42 < ws-client1> **<ChillerDragon>** @Assa the floats overflow less than the ints?
22:54 < bridge_> <Assa> yes floats overflow slower than ints?
22:55 < bridge_> <Assa> do you know how floats work?
22:56 < bridge_> <Assa> I don't know if floats overflow at all, they loose precision
22:56 < bridge_> <Assa> well yes there is a max float, but that's just NaN
22:59 < bridge_> <robyt3> Converting infinite/NaN into an int is also undefined behavior I think
23:00 < bridge_> <Assa> yes but if we reach values in the map this high, other things are breaking as well
23:01 < bridge_> <Assa> like the map space is not infinite, it'll break at some point
23:02 < bridge_> <Assa> On the other side, I should not just trust map data and just disable clipping in this case
23:07 < bridge_> <Assa> @robyt3 would you wish to have a check:
23:07 < bridge_> <Assa> ```
23:07 < bridge_> <Assa>                 if(std::isnan(Distance) || std::isinf(Distance))
23:07 < bridge_> <Assa>                     return false;
23:07 < bridge_> <Assa> ```
23:07 < bridge_> <Assa> @robyt3 would you wish to have a check?:
23:07 < bridge_> <Assa> ```
23:07 < bridge_> <Assa>                 if(std::isnan(Distance) || std::isinf(Distance))
23:07 < bridge_> <Assa>                     return false;
23:07 < bridge_> <Assa> ```
23:10 < bridge_> <Assa> hmm I think this can not happen in this case, because the interger underflows in the substraction already
23:11 < bridge_> <Assa> hmm I think this can not happen in this case, because the interger underflows/overflows in the substraction already
23:11 < bridge_> <chairn> on x86, it yields 0x80000000, this is what caused the shotgun bug
23:13 < bridge_> <robyt3> Aren't these values in fixed point representation? Can we use only fixed-point math for this?
23:14 < bridge_> <Assa> what do you mean with fixed points?
23:14 < bridge_> <Assa> ivec? ivec overflows
23:14 < bridge_> <robyt3> The values are ints that are converted to floats with `fx2f` later
23:15 < bridge_> <Assa> yes we could restructure all of it to use floats instead
23:15 < bridge_> <Assa> jupey already suggested that
23:15 < bridge_> <Jupstar ✪> Normalize your floats into tile space. 1 = size of 1 tile
23:16 < bridge_> <Jupstar ✪> Instead of using these huge numbers
23:16 < bridge_> <Jupstar ✪> Can directly refactor these values for shader too
23:16 < bridge_> <Jupstar ✪> Bcs i dislike they are so huge there too
23:34 < bridge_> <melon> Jupsti merge ddnet-rs-misc PR :poggers2: