00:36 < bridge> wrote a small utility for uploading demo files (or any kind of files that are being actively written) to a discord channel, maybe someone else has also a use for it: https://github.com/jxsl13/demo-uploader 01:07 < bridge> That's so cool :o 04:09 < bridge> gm 07:13 < bridge> morning peeps 07:17 < bridge> https://adventofcode.com/ 07:20 < bridge> A little early innit 07:41 < bridge> I know the language im doing advent of code in :santatrollet: 07:42 < bridge> Imagine 07:42 < bridge> morning 🌄🌅 08:14 < bridge> there a lot of generated names, like 5 symbols random strings 08:24 < bridge> there a lot of generated names, like 5 symbols random strings 08:24 < bridge> 08:24 < bridge> p.s ok you were talking about the ddnet base 08:32 < bridge> I guess it stores namechanges during the Game too? 08:33 < bridge> Yeah. I guess if he played for at least 5 seconds or something he is in the list already ^^ 08:35 < bridge> In pm furo told me even with a Finnish that were like 1.9m. 08:35 < bridge> 08:35 < bridge> With at least 100 points it were around 100k 08:36 < bridge> In pm furo told me even with a finish that were like 1.9m. 08:36 < bridge> 08:36 < bridge> With at least 100 points it were around 100k 08:41 < bridge> what is the order based on? 08:50 < bridge> The order in the video? 08:50 < bridge> yeah 08:50 < bridge> I shuffled it, so random 08:50 < bridge> ah 08:53 < bridge> In pm furo told me even with a finish that were like 1.3m. 08:53 < bridge> 08:53 < bridge> With at least 100 points it were around 100k 09:01 < bridge> pppArgv is insane naming :kek: 09:01 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1293105993848328225/image.png?ex=67062a4d&is=6704d8cd&hm=5cfef15a618f8df31d2df341364637c1a88928fd51818cbfa46c0f1a9527fce7& 09:11 < bridge> triple pointer is insane variable 09:49 < bridge> @learath2 syntactic macros look so cool 09:49 < bridge> https://interpreterbook.com/lost/ 5.1 5th paragraph 11:34 < bridge> ПРИВЕТ 11:34 < bridge> КАКДЕЛА 11:34 < bridge> Я ВАХУЙ 11:34 < bridge> ЕТО ПЕЗДЕЦ 11:34 < bridge> Я ИГРАЮ И ТУТ МЕНЯ ЕБУТ 11:34 < bridge> АЩАЗЗААЗАЗАЗАААЗВХАЗФЫВПЩФЫВАП 11:34 < bridge> ХВ-ХВХВВХХВХВХВХВХВХВХХВХВ 12:03 < bridge> Sir this is a Wendy's ? 12:06 < bridge> <__fredy__> guys how to host public server? please help 12:06 < bridge> $lan 12:06 < bridge> This command was not found. 12:06 < bridge> $wiki lan 12:07 < bridge> There we go -> setting up your lan server for online play 12:07 < bridge> <__fredy__> ty 12:15 < bridge> <__fredy__> i used sv_register 1 but its not making my server public 12:20 < bridge> Did you port forward ? 12:21 < bridge> <__fredy__> oh i thought there is 2 ways to make it public, mb 12:22 < bridge> @jupeyy_keks https://rust-gpu.github.io/ 12:24 < bridge> <__fredy__> can i know what should i type in "internal ip adress"? 12:25 < bridge> Yeah it's based on https://github.com/EmbarkStudios/rust-gpu 12:25 < bridge> 12:25 < bridge> Actually found the project really interesting, but it's also rather slow moving 12:28 < bridge> I thought client sends like 1-2 messages and you're ready to play but holy molly it does so much more xd 12:28 < bridge> ``` 12:28 < bridge> CLIENT SERVER 12:28 < bridge> 12:28 < bridge> CClient::Connect 12:28 < bridge> NET_CTRLMSG_CONNECT -----------------------> CNetServer::OnTokenCtrlMsg 12:29 < bridge> CNetConnection::Feed: 445 <------------------ NET_CTRLMSG_CONNECTACCEPT 12:29 < bridge> (sets state to online & sets security token recieved from da packet) 12:29 < bridge> NET_CTRLMSG_ACCEPT ------------------------> CNetServer::TryAcceptClient 12:29 < bridge> 12:29 < bridge> CClient::PumpNetwork: 2492 12:29 < bridge> CClient::SendInfo 12:29 < bridge> NETMSG_CLIENTVER --------------------------> CServer::ProcessClientPacket 12:29 < bridge> NETMSG_INFO -------------------------------> CServer::ProcessClientPacket 12:29 < bridge> CServer::SendRconType 12:29 < bridge> CClient::ProcessServerPacket: 2141 <--------- NETMSG_RCONTYPE 12:29 < bridge> SendCapabilities 12:29 < bridge> CClient::ProcessServerPacket: 1477 <--------- NETMSG_CAPABILITIES 12:29 < bridge> SendMap 12:29 < bridge> CClient::ProcessServerPacket: 1453 <--------- NETMSG_MAP_DETAILS 12:29 < bridge> CClient::ProcessServerPacket: 1493 <--------- NETMSG_MAP_CHANGE 12:29 < bridge> NETMSG_READY ------------------------------> CServer::ProcessClientPacket: 1584 12:29 < bridge> CClient::ProcessServerPacket: 1665 <--------- NETMSG_CON_READY 12:29 < bridge> CGameClient::OnConnected 12:29 < bridge> NETMSGTYPE_CL_STARTINFO -------------------> CGameContext::OnStartInfoNetMessage 12:29 < bridge> CGameClient::OnMessage: 978 <---------------- NETMSGTYPE_SV_READYTOENTER 12:29 < bridge> CClient::EnterGame 12:29 < bridge> NETMSG_ENTERGAME --------------------------> CServer::ProcessClientPacket: 1607 12:29 < bridge> ``` 12:33 < bridge> how did you generate this? 12:33 < bridge> i wrote it myself xd 12:34 < bridge> what are the numbers? 12:34 < bridge> line numbers 12:34 < bridge> ah 12:35 < bridge> so much effort 12:35 < bridge> Uff, when I try to have the debug console of chromium open while I scroll in our skin database i get 12:35 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1293159849512800326/image.png?ex=67065c75&is=67050af5&hm=9903affe56391d3916241be8a0ea91a7d9d02f68d61fc6e14b214b971fe03d4d& 12:35 < bridge> SIGILL 12:35 < bridge> chrome 0day? :p 12:35 < bridge> Apparently chrome changed smth about max xmlhttp requests 12:36 < bridge> so i wanted to dbg it 12:36 < bridge> But no chance xD 12:36 < bridge> Dunno 😄 12:36 < bridge> 12:36 < bridge> But only happens if console is open, so maybe a bug in the debugging tools 14:36 < bridge> <__fredy__> can anyone help me please, how to fix this problem? I just want to make public server 14:36 < bridge> <__fredy__> https://cdn.discordapp.com/attachments/293493549758939136/1293190248049934386/image.png?ex=670678c5&is=67052745&hm=5bc2994a20717f2389443926a1378b51479864e30be41dfaf33f14cbc30a84bc& 14:41 < bridge> Stop posting in all channels 14:41 < bridge> sv_register ipv4 14:43 < bridge> <__fredy__> thanks 14:56 < bridge> <__fredy__> https://cdn.discordapp.com/attachments/293493549758939136/1293195291364692018/image.png?ex=67067d77&is=67052bf7&hm=8d3d61be3f804dc10259a125f63360f50337f6f235238acf6414048e50adbda9& 14:56 < bridge> if you don't understand this msg, use google 14:57 < bridge> <__fredy__> i used already, i did a lot of things and nothing 14:57 < bridge> i wont help you to bypass the protection of yout NAT tho 14:57 < bridge> it doesn't look like you host the server on a VPS 14:57 < bridge> <__fredy__> what is vps 14:58 < bridge> https://google.com 14:59 < bridge> <__fredy__> so how can i host the server on a vps? 14:59 < bridge> buy a vps 14:59 < bridge> read tutorials 14:59 < bridge> <__fredy__> cant i host without vps? 14:59 < bridge> you can, but i wont help you with that. ask others 14:59 < bridge> <__fredy__> eh 😦 14:59 < bridge> <__fredy__> alr thanks 15:00 < bridge> Maybe https://steamcommunity.com/sharedfiles/filedetails/?id=2474013400&searchtext=server helps 15:03 < bridge> But yes, you need to open port 8303 (or to whatever sv_port is set to in your config) in order to register the server in the internet tab. 15:03 < bridge> and to do that, you need access to your router interface, if you don't know how, then either read your routers manual or google the routers brand 15:17 < bridge> <__fredy__> i did everything there and still no results... 15:18 < bridge> Even if you did everything right, there is a chance that your internet provider simply doesn't allow it 15:20 < bridge> Since you've did everything, can you screenshot your routers port forwarding interface which clearly displays you've opened port 8303? 15:22 < bridge> <__fredy__> https://cdn.discordapp.com/attachments/293493549758939136/1293201825448591471/image.png?ex=6706838d&is=6705320d&hm=b4d7f4c888b1fb6410480b02c7e56f5ccb685e770b52e87988613c16c2c7c04a& 15:23 < bridge> and does the internal IP address match the local IP of the device you're using to run the server? 15:23 < bridge> (screenshot that too) ^^ 15:25 < bridge> <__fredy__> yes its 100% right, i have copied that ipv4 address from cmd 15:27 < bridge> <__fredy__> yes its 100% right, i have copied that ipv4 address from cmd and typed it there 15:27 < bridge> <__fredy__> yes its 100% right, i copied that ipv4 address from cmd and typed it there 15:27 < bridge> Can you screenshot that? Just making sure 15:30 < bridge> <__fredy__> ill dm you, ok? 15:30 < bridge> Sure 16:31 < bridge> @learath2 @jupeyy_keks @heinrich5991 any weird language u find interesting to learn? for the upcoming https://adventofcode.com/ 16:31 < bridge> maybe ill learn ocaml 16:31 < bridge> go-lang 16:31 < bridge> this lang is also rising https://odin-lang.org/ 16:31 < bridge> i already know golang 16:31 < bridge> I was planning on trying it in zig 16:32 < bridge> do at least one puzzle in the GPU 16:32 < bridge> with risv-v 16:32 < bridge> if i dont find a interesting alternative ill probs try zig too 16:33 < bridge> https://en.wikipedia.org/wiki/Whitespace_(programming_language) 16:34 < bridge> > As a consequence of its syntax, Whitespace source code can be contained within the whitespace of code written in a language that ignores whitespace – making the text a polyglot. 16:36 < bridge> https://crystal-lang.org/ 17:09 < bridge> Todays useless word no one has heard since 1856: `salubrious`, means `healthy` 17:21 < bridge> gonna use that in an upcoming exam 17:21 < bridge> thx 18:10 < bridge> is header less c++ a thing? 18:11 < bridge> didn't know why i replied that 18:11 < bridge> is header less c++ a thing :justatest: 18:11 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1293244475899908126/screenshot_2024-10-08_19-08-45.png?ex=6706ab46&is=670559c6&hm=33fdcaef75ff689736ff575156598fd8540e95f5669c3820b65a8f05c98c0833& 18:11 < bridge> No words, only emotions 18:13 < bridge> https://cdn.discordapp.com/attachments/293493549758939136/1293244952163389492/screenshot_2024-10-08_19-13-04.png?ex=6706abb7&is=67055a37&hm=49227f7f326b124e27387a7a523183b9d14a4a3e0fdbb4c5b1a5ada2566edf89& 18:14 < bridge> I think you wanted to post that in #showroom ? XD 18:14 < bridge> cool #developer content 18:14 < bridge> But thanks for sharing 18:14 < bridge> I feel your sorrow 18:14 < bridge> Not a sorrow, rather anger 18:17 < bridge> nah, developers should finally fix physics to run 100x perf 18:18 < bridge> Ok i have a secret for you: 18:18 < bridge> 18:18 < bridge> ||You are a developer too|| 18:18 < bridge> When RTX in DDNet...? 18:19 < bridge> As soon as you created the pr for it 18:20 < bridge> noone looked at my pr in 3 weeks :feelsbadman: 18:21 < bridge> I create the PR to let you do the job? 👍 18:21 < bridge> that would be called an issue 18:22 < bridge> I look at it sometimes. Then close it again 18:22 < bridge> D: 18:22 < bridge> k wtf 18:22 < bridge> 0.7 pr 18:23 < bridge> no way 18:23 < bridge> where 18:23 < bridge> closed again 18:24 < bridge> mhh..my issues are orders... 18:24 < bridge> dictatorships are discussed in #off-topic 18:34 < bridge> I'm not sure if this is intentional, but if you register a chat command, like /logout, you won’t be able to use it because the same command name exists for both rcon and econ 18:35 < bridge> says access denied 19:24 < bridge> https://github.com/rust-lang/rust/issues/131374 19:26 < bridge> Wow that is quite minimal code to reproduce such a issue xD 19:32 < bridge> https://nullprogram.com/blog/2024/10/07/ some windows madness 19:37 < bridge> кто няшка? 19:38 < bridge> https://www.theverge.com/policy/2024/10/7/24243316/epic-google-permanent-injunction-ruling-third-party-stores 19:38 < bridge> I wonder how US courts can rule that google is more of a monopoly wrt. app store than apple 19:38 < bridge> but here we are 20:14 < bridge> Crazy news. 20:15 < bridge> btw, thanks heinrich, I have setup caddy and supplied the necessary urls to `NETMSG_MAP_DETAILS` and it actually works. 20:16 < bridge> (I added some baseurl config var and append the map name in `CServer::SendMap()`) 20:17 < bridge> crazy :0 20:57 < bridge> @murpi is the skin upload bot still not in auto start? xd 20:57 < bridge> it's down again 20:57 < bridge> I don't know :$ 21:29 < bridge> https://www.researchgate.net/publication/344037263_A_Direct_Proof_of_the_Riemann_Hypothesis I found this on youtube 😄 21:35 < bridge> what did youtube say? 21:35 < bridge> about that paper 21:38 < bridge> And where is their mistake? 21:39 < bridge> https://www.youtube.com/watch?v=B7POSJlOLgs 21:40 < bridge> I don't know enough math to say that, but I'm fairly certain nothing "obvious" was missed with this problem given how long it stood 😄 21:41 < bridge> I mean where is the flaw in the proof xd 21:41 < bridge> 21:41 < bridge> Was it all so easy or what? 21:42 < bridge> As said, I'm not the calibre of mathematician required to critique this guys proof. All I'm saying is that it's unlikely someone found a simple direct proof that fits in 7 pages and can be demonstrated on geogebra 😄 21:43 < bridge> and it's even more unlikely that someone that solved a millenium problem refuses to post in an established journal 21:43 < bridge> Ah yeah, looks to good to be true, him just casually using some paint skillz :lol: 21:44 < bridge> but hey, maybe I'm wrong and the riemann hypothesis is actually proven true, either way win win 😄 21:45 < bridge> I have to assume they just want some attention. My brain doesn't allow it 21:46 < bridge> I think this dude has been talking about his "direct proof" of the remann hypoethesis for a long time now 21:47 < bridge> His last update video starts off by saying the original paper had a mistake in it that could be shown with simple algebra, so he fixed it to this current version by reducing the proof to a single lemma 21:48 < bridge> idk you would think if his approach had any merit then it would have been confirmed by now 21:48 < bridge> also lol 21:48 < bridge> https://www.youtube.com/watch?v=kPSFiL-q_1s 21:49 < bridge> if he was close people would flock to fix and complete his proof 21:49 < bridge> exactly 21:50 < bridge> this is 2020. if this had any merit, you would have heard of it already 21:50 < bridge> what everyone else already said 21:50 < bridge> nvm 21:50 < bridge> Last update is Sept 24 2024, he fixed the "trivial issue" reduced it to a single lemma 😄 21:51 < bridge> Fake it 'til you make it 😬 21:51 < bridge> directly disproven: Grigori Jakowlewitsch Perelman 21:52 < bridge> iirc arxiv.org 21:52 < bridge> He was an eccentric person, whose proof was confirmed mostly quickly and furthermore it wasn't 5 pages 21:52 < bridge> 😄 21:52 < bridge> there are a lot of schizo marhematicians 21:52 < bridge> vixra is a fun place to find them btw 21:53 < bridge> Actually this paper is quite short, I wonder if you could type it up in something like Coq and have the computer check it 21:54 < bridge> kog mentioned 21:54 < bridge> Then you could write a paper on how this supposedly trivial proof of riemann hypothesis was disproven using computer assisted proof tech 21:54 < bridge> or Lean 3 or sth 21:54 < bridge> nah, that doesn't work 21:55 < bridge> only the reverse works 21:55 < bridge> you can't say a proof doesn't work just because you weren't able to formalize it in lean 21:56 < bridge> Hm, that's interesting. Can't you somehow show that you did formalize it properly? As in show the equivalence between the papers claims and your formalization? 21:56 < bridge> Hm, that's interesting. Can't you somehow show that you did formalize it properly? As in show the equivalence between the papers claims and your formalization 21:56 < bridge> Hm, that's interesting. Can't you somehow show that you did formulate it properly? As in show the equivalence between the papers claims and your formalization 21:56 < bridge> two problems with that 21:57 < bridge> My professor proved his statements by saying "trivial". 21:57 < bridge> That's my spirit. 21:57 < bridge> The proof is trivial, alas too large for this margin, it's left as an exercise for the reader 21:57 < bridge> a) no, you can't. that's one of the fundamental problems of formalization. you can't show that the model you have corresponds to what you want to have 21:57 < bridge> b) even if it has the correct model, that doesn't mean the absence of a proof means anything — it could just be you being too stupid to fill in the gaps the author left 22:00 < bridge> Interesting, I would have thought proofs like this one would be rather simple to put into computer language, then a computer can check whether the claims hold at each step 22:00 < bridge> a paper proof is almost never detailed enough for a computer to check 22:00 < bridge> IOW whether atleast the papers claims are valid under the assumption that all it's assertions hold 22:01 < bridge> you always have to fill in the gaps 22:01 < bridge> aint they working on formalizing the riemann hypothesis in lean or something 22:02 < bridge> I see, interesting. I've never seen a computer assisted proof done except for examples so complex that I can't follow 22:02 < bridge> try proving something really trivial from your first semester 22:02 < bridge> the four color theorem is probably the only one that comes to my mind 22:03 < bridge> I actually like working with proof assistants 🙂 22:03 < bridge> kinda like a small puzzle game 22:03 < bridge> Which one would you suggest playing around with? 22:03 < bridge> lean is the hip shit, I heard 22:03 < bridge> I mostly worked in coq so far, but would try lean nowadays 22:04 < bridge> one of the first things I did after going through various coq tutorials was proving that a group is commutative if each element is self-inverse 22:04 < bridge> this was slightly nontrivial but still interesting 22:04 < bridge> Say I filled in "all" the gaps/hidden assumptions in this mans proof, would it still not be enough to show that this paper doesn't hold? 22:04 < bridge> if you filled in all the gaps, then you'd have the proof 22:04 < bridge> so there must be some gap missing 22:05 < bridge> this is probably where a false statement appears 22:05 < bridge> hm, maybe I misunderstood the word "gap" there 22:05 < bridge> you can find that counterexample if it's simple enough 22:05 < bridge> but if you do, the answer is going to be that this doesn't matter for the correctness of the proof 22:05 < bridge> and that you just need slightly different assumptions 22:06 < bridge> (which is also something that happens for actual mathematics btw, that statements have slightly wrong assumptions, but the whole proof still works if you correct them) 22:06 < bridge> Is it not possible to show that a gap is "unfillable"? 22:06 < bridge> I'm guessing not in general 22:07 < bridge> you can sometimes prove that statements are false, yes. but since this is a well-known hypothesis, both a proof and a proof of the inverse are very hard 22:07 < bridge> so if you construct your fake proof well enough, the person trying to show that your proof doesn't work has to almost prove the riemann hypothesis right or wrong 22:07 < bridge> to conclusively prove that you're wrong 22:08 < bridge> I see, that makes some sense yeah 22:09 < bridge> btw my favourite mathematical problem ever is the Collatz conjecture. I really hope one day we have a beautiful proof for it 22:09 < bridge> not one of those billion page computer generated proofs, nor a 2 sentence counterexample 😄 22:11 < bridge> some of the very hard long standing problems are in a little gap between the problem having sufficient complexity to encode a turing machine, the problem being solvable. So you end up almost running into the halting problem but not quite 22:11 < bridge> It does feel very closely related to something like the halting problem though, so idk maybe it can be shown to actually be impossible to prove 22:11 < bridge> some of the very hard long standing problems are in a little gap between the problem having sufficient complexity to encode a turing machine, and the problem being solvable. So you end up almost running into the halting problem but not quite 22:12 < bridge> concrecte examples are probably not impossible to prove 22:13 < bridge> *for concrete, non-constructed, examples, it's probably hard to prove that it's impossible to prove 22:13 < bridge> but turing showed really early that the generalized collatz problem is equivalent to the halting problem 22:14 < bridge> Impossible is just the lack of entering the world of infinity 22:14 < bridge> What sort of generalization are we talking about here? 22:15 < bridge> instead of 2 cases, choose n cases and give a polynomial step formula for each 22:15 < bridge> IIRC 22:17 < bridge> I read about busy beaver numbers a couple months back, it's such an interesting way to "quantify" complexity 22:18 < bridge> Proving the Collatz Conjecture is as hard as computing BB(15) apparently 22:18 < bridge> @learath2 What do you think, how many percent of today's known math do you know? 22:18 < bridge> Hi 22:18 < bridge> Probably less than 10%, say 10 with all the curiousities I skimmed over my life 22:18 < bridge> wow really? bb(15) is absurd 22:18 < bridge> can anyone recommend gameskin? 22:19 < bridge> Try #general or #questions 22:19 < bridge> default is legendary 22:19 < bridge> Opps sorry 22:19 < bridge> try that 22:19 < bridge> > There is a 15 state Turing machine that halts if and only if a conjecture by Paul Erdős (closely related to the Collatz conjecture) is false. Hence if BB(15) was known, and this machine did not stop in that number of steps, it would be known to run forever and hence no counterexamples exist (which proves the conjecture true). This is a completely impractical way to settle the conjecture; instead it is used to suggest that BB(15) will be very h 22:19 < bridge> 10% sounds like too much IMO. even for some math researcher 22:20 < bridge> I guess it's hard for me to estimate that percentage given I don't know what I don't know 22:20 < bridge> So that 10% is more like, I know some things about 10% of the math I've heard about ever even in passing 22:20 < bridge> the longer text states the opposite: computing BB(15) is at least as hard as proving the collatz conjecture 22:21 < bridge> It's still interesting to philosophy about such questions. 22:21 < bridge> 22:21 < bridge> We probably cannot even comprehend how much math there is xd 22:21 < bridge> 22:21 < bridge> What do you think, how many percent do you know? :DD 22:21 < bridge> If you consider the ability to understand the cutting edge math of any subject in under a week I think Tao knows more than 10% 22:21 < bridge> just because he hasn't read every paper ever seems unfair to say that he only understands <10% of mathematics 22:23 < bridge> Hm, perhaps I misremembered, it's been months since I took a gander at bb 22:24 < bridge> terence tao is a genius 22:24 < bridge> good read https://gwern.net/doc/iq/high/smpy/1984-clements.pdf 22:24 < bridge> Possible. 22:24 < bridge> 22:24 < bridge> What I often notice in software development is, how many libs we don't know but depend on. 22:24 < bridge> So much _random_ code that is actually used by many ppl, but I never heard of. 22:25 < bridge> 22:25 < bridge> Ofc some code solves the same problem as others. But generally there are always new things. 22:25 < bridge> 22:25 < bridge> So if you say ok, but if you know most stuff about CPU arch and GPU arches and how memory works and what do I know, how much do you really know then? 22:25 < bridge> Possible. 22:25 < bridge> 22:25 < bridge> What I often notice in software development is, how many libs we don't know but depend on. 22:25 < bridge> So much _random_ code that is actually used by many ppl, but I never heard of. 22:25 < bridge> 22:25 < bridge> Ofc some code solves the same problem as others. But generally there are always new things. 22:25 < bridge> 22:25 < bridge> So if you say ok, but if you know most stuff about CPU arch and GPU arches and how memory works and what do I know, how much do you really know about software dev then? 22:26 < bridge> To be fair, isn't it fact that there is no consistent system of axioms that can prove all truths of natural numbers? So some of the true things will never be provable 22:26 < bridge> Idk it's hard to find a parallel to Terence Tao is the computing world 22:26 < bridge> So there is "math" if you can even call it "math" that we would never be able to ever explore 😄 22:26 < bridge> It's also kinda two sided. A good kernel dev might understand the problems of a "normal" program. But does that make him a good developer in this field? 😄 22:27 < bridge> Terence Tao competes against great mathmaticians, and he is alive, probs the best math guy in the world rn 22:27 < bridge> software is mostly about labor than it is deep insights. 22:28 < bridge> I think there's probably many software engineers who given enough years could replicate the linux kernal, but no one has time for that. 22:28 < bridge> Well is that so? 😄 22:28 < bridge> 22:28 < bridge> Do you learn by doing this labor? If you can learn from it, what is there to learn if it's just about doing stuff 😄 22:28 < bridge> gg 22:28 < bridge> well there are many great software guys 22:28 < bridge> alan turing 22:28 < bridge> i forgot the other big name 22:28 < bridge> @patiga gg 22:28 < bridge> omg 22:28 < bridge> I don't mean computer science, I mean software as a practical industry 22:28 < bridge> the part that has a $10 trillion dollar value 22:29 < bridge> Well there is still theoretical computer science going on, with many deep insights required 22:29 < bridge> Software engineering is just putting pieces together, so yeah, that is a little less "deep" I guess 22:29 < bridge> software practical industry is driven by computer science imho 22:30 < bridge> yes but those deep insights alone do not get you to useful software 22:30 < bridge> but they kind of do 22:30 < bridge> the field of computer science is hugely influenced by industry moreso than pure mathematics 22:30 < bridge> Yeah I can defs understand your point. It seems like development doesn't require huge skills, since so many ppl can create useful stuff. 22:30 < bridge> a language like rust is based on computer science, the borrow checker is literally a paper 22:30 < bridge> there is no deep insight you can have that lets you reproduce the linux kernal in a week or a month 22:30 < bridge> Probably bcs our computers are so fast 😄 22:30 < bridge> with lot of theory 22:30 < bridge> and rust is useful software 22:30 < bridge> I'm talking about the ratio of theory:work 22:31 < bridge> I just had a very deep insight that will allow me to do that 22:31 < bridge> `git clone https://github.com/torvalds/linux.git` 🙃 22:31 < bridge> And do you feel the theory is needed to invent smth like Rust? 22:31 < bridge> 22:31 < bridge> Or could you just do it and do the theory later? 😄 22:31 < bridge> https://plv.mpi-sws.org/rustbelt/stacked-borrows/paper.pdf 22:31 < bridge> rust lies strong on the theory to prove its safe 22:31 < bridge> in its definition of safe 22:32 < bridge> engineering is just an application of science 22:32 < bridge> there are ofc good engineers 22:32 < bridge> imho its 50/50 22:32 < bridge> LLVM the best compiler in the world, along gcc is powered by its R&D 22:33 < bridge> After looking into making my own, I can confirm that just going at it is not possible unless you are like a savant for whom it's immediately obvious 22:33 < bridge> If you are not careful and don't have a strong theoretical background you immediately run into the halting problem trying to implement a borrow checker 😄 22:34 < bridge> just because you can use theory to make software doesn't mean it's necessary. strictly speaking you could write every program in C with enough time, and the end user would not notice 22:34 < bridge> its true tho that a lot of science papers are crazy and ignore completly how a computer works in the real world 22:34 < bridge> and dont do optimized algorithms 22:34 < bridge> whathever 22:34 < bridge> what i mean is sometimes they are far fetched fro mreality yes 22:34 < bridge> but there is some practical stuff 22:34 < bridge> (for example, of useless things: galatic algorithms) 22:35 < bridge> https://en.wikipedia.org/wiki/Galactic_algorithm 22:35 < bridge> yes. gödel's incompleteness theorem 22:35 < bridge> There are more theoretical papers and there are more concrete practical papers, I've seen both kinds in the wild 22:35 < bridge> like if you make a program will only run if you input the lean proof of the Poincaré conjecture and then say that "see you need theory in order to make software" 22:35 < bridge> like you make a program will only run if you input the lean proof of the Poincaré conjecture and then say that "see you need theory in order to make software" 22:36 < bridge> > An example of a galactic algorithm is the fastest known way to multiply two numbers,[3] which is based on a 1729-dimensional Fourier transform.[4] It needs O ( n log ⁡ n ) {\displaystyle O(n\log n)} bit operations, but as the constants hidden by the big O notation are large, it is never used in practice. However, it also shows why galactic algorithms may still be useful. The authors state: "we are hopeful that with further refinements, the algo 22:36 < bridge> :pepeW: , 22:36 < bridge> :pepeW: 22:36 < bridge> thought he's mostly the most visible one of the very good mathematicians 22:36 < bridge> from the ppl i know he is held as the best math alive 22:36 < bridge> mind you, im NOT implementing it in Tee Galaxy 22:37 < bridge> i recommend u read the biography paper i sent 22:37 < bridge> its rly interesting 22:37 < bridge> he was identified as a genius and was raised as such, not ignored 22:37 < bridge> lucky to have rich parents tho 22:37 < bridge> Tao is very impressive because of his breadth, it's hard to say if there is an equally smart mathematician who only focused on one subject their entire career. 22:38 < bridge> His research includes topics in harmonic analysis, partial differential equations, algebraic combinatorics, arithmetic combinatorics, geometric combinatorics, probability theory, compressed sensing and analytic number theory.[4] 22:38 < bridge> Well not all theory fits to computation. You can't really "just make" an AI without understanding any theory. AI is about as close to "just theory" as you can get 22:38 < bridge> his dedication must've been insane 22:38 < bridge> There are probs tons of legendary mathematicans alive today, that we don't know. 22:38 < bridge> 22:38 < bridge> The world also doesn't know us, the Legends. 22:38 < bridge> > Childhood 22:38 < bridge> > 22:38 < bridge> > A child prodigy,[20] Terence Tao skipped 6 grades.[21][22] Tao exhibited extraordinary mathematical abilities from an early age, attending university-level mathematics courses at the age of 9. He is one of only three children in the history of the Johns Hopkins Study of Exceptional Talent program to have achieved a score of 700 or greater on the SAT math section while just eight years old; Tao scored a 760.[23] Julian Stanley, Director of the Stu 22:38 < bridge> > 22:39 < bridge> > Tao was the youngest participant to date in the International Mathematical Olympiad, first competing at the age of ten; in 1986, 1987, and 1988, he won a bronze, silver, and gold medal, respectively. Tao remains the youngest winner of each of the three medals in the Olympiad's history, having won the gold medal at the age of 13 in 1988.[25] 22:39 < bridge> god damn, at the age of 9 22:39 < bridge> young neilson would shit itself 22:39 < bridge> hmmm. rust is a very practical language I think, very distant from academic languages. it tries to solve real-world problems, such as interactions with C, something that academics are largely unconcerned about 22:40 < bridge> yeah, i dont disagree (?) 22:40 < bridge> it still relies on a lot of CS, just like other languages 22:40 < bridge> i wish they made a movie about tao 22:41 < bridge> i seen the ramanujan one 22:41 < bridge> i like seeing movies about smart ppl 22:41 < bridge> xd 22:42 < bridge> I think all theory does fit in computation, by definition. Even cutting edge AI research doesn't not require math knowledge as deep as physics or other areas, most people can learn all the math for transformers in a couple years it's not insane stuff. 22:42 < bridge> What does it mean to be the "best" mathematician anyway? He is exceptional in his field but is he say better than everyone in every subbranch of mathematics? Is he the best to ever live? Is he "better" than say Euler e.g.? or Ramanujan? 22:42 < bridge> @totar tell tao to invent better ultraforce if you see him around 22:42 < bridge> lol 22:42 < bridge> idk 22:42 < bridge> the nice thing is he is alive 22:42 < bridge>