00:06:18 <esolangs> [[A returns a]] https://esolangs.org/w/index.php?diff=84360&oldid=84300 * Jetison333 * (+0) /* Code examples */
00:19:50 -!- leah2 has quit (Read error: Connection reset by peer).
00:34:01 -!- Thelie has quit (Remote host closed the connection).
01:16:11 <oerjan> <b_jonas> Corbin: https://esolangs.org/wiki/Combinatory_logic says "Blackbird" <-- after a little wikiwalk inspired by wikipedia's featured picture, turns out those are not actually corvids
01:18:43 <oerjan> (neither the european nor american version)
01:20:11 <oerjan> *old world nor new world
01:23:34 <Corbin> oerjan: Zoology is wack, huh? And bluejays are corvids despite looking like passerines. TIL, thanks.
01:27:46 <keegan> corvids are a subset of passerines, aren't they?
01:28:02 <keegan> https://en.wikipedia.org/wiki/Corvidae "Order: Passeriformes"
01:29:20 <keegan> we don't have blue jays here but we have scrub jays and Steller's jays
01:29:23 <keegan> https://en.wikipedia.org/wiki/Steller's_jay
01:29:39 <keegan> very striking appearance
01:29:58 <b_jonas> oh heck, so you need to know both ornithology and bird sociology to tell whether there's a combinatorial bird that's also a corvid
01:30:32 <b_jonas> or more likely, solve some kind of puzzle where the forest's keeper tells you about rules that somehow involve corvids and bird application, Smullyan style
01:31:59 <b_jonas> "After observing this forest for twenty years, I have discovered three interesting rules. (1) For every corvid A, there is a bird B such that BA is not a corvid. [...]
01:33:22 <b_jonas> `I have pondered your observations all night, and have discovered a striking fact. None of the birds in this forest are corvids.' [...] `Curious. I've been observing this forest for 20 years and never realized that.'"
01:33:24 <HackEso> I? No such file or directory
01:33:51 <b_jonas> (The actual punchline in the book that none of the birds ever sing in the forest is much better than the one with corvids, I admit.)
01:37:54 <Corbin> keegan: Oh, interesting, TIL again. I thought "passerine" was just how zoologists say "songbird". The most common corvids here are crows, and they don't exactly ♩ sing ♪
01:38:02 <b_jonas> or perhaps it would be something like "From these three rules, detective Craig could derive the following properities. 1. All bluebirds are corvids. 2. None of the kestrels in this forest are corvids. Can you also prove those properties?"
01:39:34 <b_jonas> I think the most common songbirds here are common blackbirds
01:40:01 <b_jonas> but maybe they're just the most obviously recognizable, with their yellow beak but otherwise black body, while some of the other corvids are all black and hard to distinguish from each other
01:40:31 <b_jonas> crows might be more common
01:43:26 <keegan> crows are very common here
01:43:32 <keegan> almost as common as pigeons
01:56:29 <b_jonas> it's sparrows that are common here, the little greedy guys
01:56:32 <oerjan> crows are more common than pigeons here
01:57:08 <oerjan> although they compete with seagulls, depending on time of year
01:57:20 <b_jonas> I think pigeons are more common than crows, though there also seem to be less pigeons than there used to be. maybe the generation of old city folks who fed pigeons have aged out?
01:57:56 <oerjan> there are probably more pigeons in the city center (or at least used to be)
01:59:13 <b_jonas> perhaps. there are pigeons right on our roof, and so are there on the roof over my parent's home, but yes, probably more common in the city center
02:00:55 -!- ircseeker has joined.
02:05:20 <keegan> i think that deliberately given food is probably not a huge fraction of their diet but that a lot of it is trash that will be more widely available in the city center
02:05:43 <keegan> we also have seagulls (being by the sea and all)
02:05:46 <keegan> and they love trash too
02:05:57 <keegan> they like to hang out at the parking lot of a burger place i go to
02:18:11 -!- user3456 has quit (Quit: ZNC - https://znc.in).
02:20:18 -!- user3456 has joined.
02:30:45 -!- oerjan has quit (Quit: Nite).
03:30:21 -!- Lord_of_Life has quit (Ping timeout: 252 seconds).
03:30:25 -!- Lord_of_Life_ has joined.
03:31:38 -!- Lord_of_Life_ has changed nick to Lord_of_Life.
04:47:09 -!- ircseeker has quit (Ping timeout: 264 seconds).
05:09:20 -!- tromp has joined.
05:19:48 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
05:20:36 -!- sprocklem has changed nick to sprock.
05:46:09 -!- tromp has joined.
05:59:48 <shachaf> int-e: So I was thinking about how to actually to meet-in-the-middle attacks with the lambda algorithm and it's not clear to me how to do it.
06:07:40 -!- t3hyoshi has joined.
06:07:41 -!- t3hyoshi has quit (K-Lined).
06:07:50 <int-e> shachaf: I don't know. It /says/ it's a meet in the middle attack but it actually computes a collision between two distinct hash-like functions, somehow relying on the second half of the process to be invertible. I imagine you have a double DES, E_k1(E_k2(X)) = Y for 128 bit known plaintext X and ciphertext Y, and you rewrite that as D_k1(Y) = E_k2(X) and then treat k1 and k2 as parameters
06:08:13 <int-e> and then... do something with that
06:08:37 <int-e> to get an ~2^64 operations attack. Or 2^56 with some extra care.
06:09:07 <shachaf> There are lots of collisions if you just look for any pair of keys that collide in the middle for a single plaintetxt-ciphertext pair.
06:09:51 <shachaf> So you probably want two pairs, and then maybe to encrypt/decrypt both texts, so f,g : 2^56 -> 2^128 or something.
06:10:16 <int-e> shachaf: I wrote "128 bits" for a reason
06:10:41 <shachaf> Right, I'm not disagreeing with you, just writing things out.
06:10:54 <int-e> you can do some hashing on the 128 bit values to get f,g : 2^128 -> 2^128
06:11:13 <int-e> but it feels weird... I'm not sure whether it works
06:11:31 <int-e> I don't know what I want.
06:11:36 <shachaf> It seems odd because if you construct h out of f and g, you're not just exploring a huge space with lots of collisions.
06:11:58 <shachaf> You want to explore approximately the entire domain of h, or something?
06:12:14 <int-e> I have not really worked out that meet-in-the-middle attack part of the paper. I pointed it out because it actually *does* is find a collision between two distinct functions.
06:12:53 <int-e> And for that purpose, it does work.
06:13:33 <int-e> So I mostly ignored the meet-in-the-middle angle. I can't say whether it works, but it does strike me as a bit odd.
06:16:47 <int-e> For example, the problem with "you can do some hashing on the 128 bit values" is that the effective codomain is still just of size 2^56
06:17:56 <int-e> shachaf: I've been answering you wrong, I think. What I should have written is that I think that's a good question, that I haven't worked it out and that I also have some doubts about whether this works, but I have not read the text carefully enough to figure it out for sure.
06:18:18 <int-e> There was no need to get defensive or aggressive about it.
06:18:23 <shachaf> Sure, I was just vaguely thinking about it out loud or something at this point.
06:18:56 <shachaf> I certainly didn't meet to accuse you of saying something wrong, I thought it was interesting when you said it and I didn't really think about it until today.
06:22:39 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
06:29:02 -!- SiFuh3 has joined.
06:29:02 <SiFuh3> ⧸!⧹ ΤΗIЅ ⲤНΑⲚNEL ᎻAS MОVEᎠ ΤO IRC.ᏞIᗷΕRA.CᎻΑΤ #HΑМᎡΑᗪІО ᜵ⵑ⧵
06:29:03 <SiFuh3> ᜵!＼ ΤHE ЈΕWᏚ HΑVΕ ΤAⲔЕN ⲞVEᎡ ᖴREЕΝODΕ‚ СዘATᏚ HAVЕ MOⅤED TΟ IRⅭ․LІBЕᎡΑ.CHАΤ /﹗＼
06:29:04 -!- SiFuh3 has quit (K-Lined).
06:30:21 <myname> i don't get why those bots even spam in libera about switching to libera
06:33:26 <int-e> shachaf: looking more closely, it searches for collisions for a single block, then validates them on a second block, and the main benefits are better tradeoff of memory for performance and better parallelization, both thanks to the distinguished points method.
06:37:20 <int-e> shachaf: it's worse noting that the basic work factor in (9) is 2^84 = 2^28 * 2^56, scaling down by sqrt(w) for a table of size w, so it becomes 2^56 when we get a table entry per key
06:38:20 <int-e> in any case, there's quite a lot going on there on top of the collision search
06:40:13 <int-e> a big part is randomizing the auxiliary functions that you have collide so that you get /different/ collisions to work with, each with a chance of being the key you're looking for
06:55:31 <int-e> Oh what fun... https://www.theregister.com/2021/06/04/google_chip_flaws/
07:15:29 -!- river has joined.
07:28:13 <zzo38> And, why are some of the letters mixed up on thise spam messages?
07:28:56 <shachaf> Yes, I looked at it and it looked tricky.
07:29:16 <int-e> zzo38: to thwart spam filter, obviously
07:29:43 <int-e> worse noting, I really make th wortht typos
07:30:32 <zzo38> Yes, I suppose so, but a spam filter could be made to look for such mixed up letters instead. (A long time ago I had seen spam messages with too many spaces at the end of the subject line, so I used that as a filter criteria; sometimes it is base64 encoded or whatever, so I used that asa filter criteria; etc)
07:33:15 -!- warlock has quit (Quit: Lost terminal).
07:41:11 -!- Sgeo has quit (Read error: Connection reset by peer).
08:11:11 -!- tech_exorcist has joined.
08:14:49 <fizzie> Re why spam about it here, I don't know how plausible it is, but it *could* be an attempt to get bridge bots get killed on the other side.
08:17:55 <int-e> seems too elaborate
08:18:14 -!- tromp has joined.
08:38:12 <nakilon> 09:30:21 <myname> i don't get why those bots even spam in libera about switching to libera
08:38:41 <nakilon> because the whole point was just to call people jews, koreans, etc., not switching the network for any profit
08:38:52 <nakilon> people will start realising it maybe in few years
08:39:03 <nakilon> some things are too hard to understand for majority
08:39:59 <nakilon> the only "reason" switching from freenode was these same spambots who were telling you to do it and you just followed their orders, lol
08:45:55 <nakilon> I imagine the TPUs can have miscalculations happen every second and no one will spot it behind the "NN magick"
08:46:57 <nakilon> the whole point of "computers being reliable" has devaluated thanks to NN hype marketing
08:48:10 <nakilon> today people with no math/engineering skills design, approve and even implement most of the software
08:49:19 <myname> i switched way before the spambots started
08:52:29 <myname> i suspect those bots to be somehow paid by freenode to fud on libera
08:53:52 <nakilon> spambots started before, they just looked different, were not automated
08:54:14 <int-e> it can easily be just a third party who revels in drama... the spam resembles the supernets spam so much
08:54:43 <int-e> And there's really no need to accuse Andrew+crew of things that can't be proven
08:55:16 -!- craigo has quit (Read error: Connection reset by peer).
08:55:42 <nakilon> people didn't need anything to be proven to switch from freenode though
08:56:38 <nakilon> I was asking like two times what are the evidence that something bad is happening and all you had was a wikipedia link to some website where guy is called a prince and it's somehow bad or something
08:57:01 <myname> it was a hostile takeover, that's reason enough
08:57:22 <nakilon> yeah, people say these two words "hostile takeover" with no sense put in it
08:57:31 <myname> it doesn't even matter if there were legal threats or not, it doesn't matter what the intentions from whoever are
08:58:13 <myname> okay, even if it was not a hostile takeover, i still follow the staff, not the name of the network
08:58:31 <int-e> Andrew's first blog post by itself would've convinced me to leave.
08:58:35 <myname> if the staff saw a reason to leave, so do i
08:59:17 <nakilon> so stuff are in fact your masters? maybe that's how the "takeover" looks like? they move you from one place to another as they wish
08:59:46 <int-e> "masters"? They're good hosts.
08:59:55 <myname> what is your reason to choose a network if not the people who operate it?
08:59:58 <nakilon> how isn't Andrew a good host?
09:00:47 <myname> the operators are what define the network, not the name of it
09:00:50 <int-e> Going with the restaurant analogy, he's been serving burned food and yelling at the customers from day one.
09:01:15 <nakilon> I have no idea who operates networks and this neither has or have to have any nifluence on my choice
09:01:34 <nakilon> my socket connection and ability to talk and run bots works fine on any network
09:02:30 <nakilon> how is it bad to be independent and not a pawn being moved elsewhere because of someon else's personal dramas lol
09:03:15 <int-e> nakilon: so you're largely unaffected by the move and complaining about it because you think that just because you don't care about Andrew's antics, nobody else should care either?
09:03:19 <nakilon> int-e I didn't have any burned food
09:03:29 <myname> i'm not a pawn being moved. i know staffers personally and i trust them. i have yet to find any positive fact about andrew
09:04:43 <nakilon> neither of you did have a burned food
09:04:50 <nakilon> that was just their personal drama
09:05:05 <nakilon> some internal job communication problems
09:05:07 <int-e> What? I saw #haskell being taken over.
09:06:35 <nakilon> what do you mean by "#haskell taking over"? something like with gentoo? the bug in a script with the following apologizing from Andrew's side and purposeful hiding this by gentoo's admins to manipulate their pawns who are happy to be like that?
09:08:52 <FireFly> I think it's less "taken over by a prince and that's bad" and more "taken over by a businessman running it as some sort of startup, apparently with several reports of harassment against them, opering up known abusers", at least from my POV
09:08:56 <myname> so either he straight up lies or has no clue on how to run a network in a robust way. great start so far. i also like how you always point out that others believe everything and at the same time are totally fine with him claiming it was an accident that should never ever happen outside of test servers or channels
09:09:06 <nakilon> or another picture I see here is like starting shitting on the walls. drawing words "ACAB" with a shit, until cops come to stop you and you say "see?! they are arresting me -- that's why I was right shitting on the walls and no one will check the chronology, pawns"
09:09:25 <FireFly> but either way, it doesn't really matter particularly, people are presumably free to choose where they want their communities *shrug*
09:10:48 <nakilon> "lies or has no clue on how to run a network in a robust way" -- what would you expect when half of staff have left? maybe it was even one of their sabotage scripts, saying "hey dude, run this"
09:11:32 <myname> sure, i also always run unchecked code from people that just left me
09:12:12 <myname> if "half" of staff left, i expect him to either not run untested scripts or just test them locally
09:12:24 <nakilon> maybe that sctipr author left on the next day
09:12:43 <myname> i cannot trust him maintaining a reliable network either way
09:12:52 <nakilon> all you know is that it was ran and there were apologizings but you don't mentions the latter because that's not the goal of drama
09:13:42 <int-e> https://nitter.fdn.fr/kurtseifried/status/1398712468394811395#m
09:13:51 <myname> no, the apology doesn't matter. if he doesn't know what it does, he should never have run it in the first place. if he did know it, the apology is useless
09:13:52 <FireFly> I don't think the problem is the script or apologise per se, moreso than the sudden changes in what has previously been perfectly fine for at least two decades on freenode..
09:14:19 <FireFly> the whole thing like "can't mention another network in topics because that's advertisement" is... pretty foreign
09:14:33 <nakilon> yep, people self inflicted the network damage
09:15:02 <myname> what, don't you love censorship in your free network?
09:15:16 <nakilon> "he should never have run it in the first place" -- he was the one who tried to fix things unlike others
09:15:30 <myname> what things? what was wrong?
09:16:27 <nakilon> how spamming the libera network name in topic better than spamming hamradio channel name?
09:17:22 <FireFly> I think "this community has moved to X" is uh, not exactly spamming
09:17:35 <nakilon> so you call stopping you from shitting the walls a censorship?
09:17:46 <FireFly> just, like, a friendly PSA to the community in question, if they've decided to move.. be it to libera, or discord, or matrix or whatever
09:18:10 <myname> okay, at first you didn't know what bots were and now you don't know what spamming is. if channel owners want to include something in their topic, it's their choice. unless it violates law, a network must not forbid me to put whatever i want in there
09:19:01 <myname> if i spray something on the inside of my wall and get arrested for it, i indeed do call it censorship
09:19:34 <myname> i can do whatever i want in my house
09:19:43 <nakilon> the network mentions in topic was not a choice -- it was a result of a spam and person-based hate flame effect on channels owners
09:20:26 <myname> ah, all ops of the hundreds of channels were bad because they decided to inform their respective community about a change
09:20:56 <nakilon> it wasn't only on your inside walls, the whole apartment was smelling with this
09:21:09 <nakilon> people were smelling shit in all the channels
09:21:28 <myname> yeah, because each house owner decided to paint something on their walls
09:21:56 <myname> that is nothing that should ever be changed by a network owner
09:23:33 <nakilon> that's exactly what landlords should do -- stop the hype of shitting the walls no matter if they are inside or outside because the whole building started smelling
09:23:53 <myname> you can probably argue whether or not the choice to switch networks was reasonable. but _if_ i decide to move a community as an op, i should be allowed to put that information in a topic.
09:24:16 <myname> nakilon: i highly doubt landlords can do anything to you if you spray the inside of your house
09:24:18 <nakilon> btw, it's not even the correct analogy -- channels are not personal flats but a common areas like cafes or cinemas
09:24:25 <myname> at least where i live, they cannot do shit about it
09:24:54 <myname> it's a fundamental right that i can do whatever i want in my house
09:25:03 <nakilon> so it was a shopping center and cafe owners started serving shit
09:25:41 <esolangs> <LKoen> well you can't farm drugs in your house
09:25:44 <myname> if other people don't get affected by it, sure thing
09:26:06 <myname> yeah, i did mention earlier that you cannot break the law
09:26:44 <myname> the whole "stinking" argument is nonsens. if i spray my wall, you will not be able to smell that on the outside after like a few days max
09:26:49 <nakilon> everyone was affected; people hated this shit so much that agreed to switch just in hope that they'll stop smell what cafe owners did
09:27:44 <nakilon> channel topics are the outer side of your door anyway even if ti's a personal flat
09:27:53 <nakilon> people are passing by every day
09:28:13 <nakilon> and there are rules on what you can do with your door for sure -- you can't shit it
09:28:20 <myname> how are channel topics the outer side?
09:29:45 <nakilon> they are in /list and /chanserv info and on the websites that track the channel topics
09:29:56 <esolangs> [[FolderCode]] M https://esolangs.org/w/index.php?diff=84361&oldid=84322 * Dominicentek * (-41)
09:30:13 <nakilon> also it's the same as your balcony and there are rules too about what you can place on your balcony
09:30:19 <myname> you know why that is? so people can see relevant information. like changes in the community and stuff
09:30:39 <nakilon> yeah nazism was revelant change in 30s
09:30:55 <myname> sure, you can go that route, but that is a reason against freenode. i want to be in a network where i can put whatever legal stuff i want to in topics
09:31:01 <nakilon> so Italy could learn it too
09:31:06 <int-e> https://en.wikipedia.org/wiki/Godwin's_law
09:31:40 <myname> that comparison is bullshit
09:31:53 <myname> in most places, it's highly illegal to do that
09:32:05 <nakilon> ahah, you won't be able to put the other network name all over the place if that would be caused by that some guy from these team is a prince of some Evil Axis country or something
09:32:11 <myname> it is however not illegal to name a website that is totally legal
09:33:00 <nakilon> int-e there should be some law ahout probability that mentioning the logic, ANd and OR would be too
09:33:18 <nakilon> that's called a learning on mistakes
09:33:32 -!- leah2 has joined.
09:33:57 <nakilon> the thing is that you again ignore the chronology
09:34:21 <myname> also, you are comparing a legit network by trusted people with some grey area, which is still bullshit
09:34:35 <nakilon> the freenode drama started with Andrew's nationality in the first place
09:34:50 <myname> what? i don't even know andrew's nationality
09:34:58 <int-e> nakilon: You really don't get to question other people's personal decisions. We have reasons, you think they're invalid reasons, we get it. Move on, let it rest.
09:35:30 <myname> how should i know? it was never mentioned by any of the staffers letters afair
09:35:50 <nakilon> his surname is such a secret
09:36:00 <nakilon> this pretending is too blatant
09:36:27 <myname> okay, now i pretend because i say i don't care about the nationality of somebody?
09:36:57 <myname> maybe i read his surname, not sure, i never bothered to remember. i wouldn't even know his first name if you didn't mention it
09:38:03 <myname> okay, so he is called lee. so what?
09:48:35 -!- imode has quit (Ping timeout: 272 seconds).
10:16:42 -!- moon has quit (Quit: leaving).
10:17:14 -!- moon has joined.
10:23:11 <esolangs> [[Column]] N https://esolangs.org/w/index.php?oldid=84362 * Dominicentek * (+2637) Created page with "Column is an esoretic programming language developed by Dominicentek. It's code is executed column by column in a text file, rather than line by line like in a typical program..."
10:33:31 <esolangs> [[User:Dominicentek]] https://esolangs.org/w/index.php?diff=84363&oldid=84210 * Dominicentek * (+13)
10:34:34 <esolangs> [[Language list]] https://esolangs.org/w/index.php?diff=84364&oldid=84343 * Dominicentek * (+13) /* C */
12:00:30 <esolangs> [[Special:Log/newusers]] create * .yazic * New user account
12:35:45 -!- andydude has joined.
12:51:00 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
12:56:50 <esolangs> [[Language list]] https://esolangs.org/w/index.php?diff=84365&oldid=84364 * ResU * (+56) added NOTE and Truth Machine
13:07:01 -!- tech_exorcist has quit (Ping timeout: 272 seconds).
13:08:50 -!- tromp has joined.
13:09:10 -!- anonymous has joined.
13:09:48 -!- andydude_ has joined.
13:16:42 -!- arseniiv has joined.
13:18:14 -!- andydude has quit (*.net *.split).
13:18:14 -!- leah2 has quit (*.net *.split).
13:18:14 -!- river has quit (*.net *.split).
13:18:14 -!- andydude_ has changed nick to andydude.
13:20:44 <esolangs> [[Column]] M https://esolangs.org/w/index.php?diff=84366&oldid=84362 * PythonshellDebugwindow * (+24) /* Interpreter */ Category
13:25:32 -!- leah2 has joined.
13:27:21 -!- andydude has quit (Quit: andydude).
13:38:35 -!- anonymous has changed nick to river.
13:44:23 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
13:45:36 <b_jonas> "<int-e> worse noting, I really make th wortht typos" => you managed to type "thwart" correctly
13:46:48 -!- arseniiv has quit (Quit: gone too far).
13:54:08 -!- tromp has joined.
13:55:09 <b_jonas> ok, let's do it right this time, rather than just return from a family break to see the drama played out
13:56:31 <b_jonas> dear esolangs community, if this channel or the network collapses somehow, even twelve years from now, we will meet on the wiki and on the oftc IRC server channel #esolangs , using whichever works as a lifeboat to decide where to go.
14:19:09 -!- iovoid has quit (Read error: Connection reset by peer).
14:19:51 -!- Bowserinator has quit (Ping timeout: 272 seconds).
14:19:55 -!- user3456 has quit (Ping timeout: 252 seconds).
14:20:42 -!- user3456 has joined.
14:20:48 -!- Bowserinator has joined.
14:21:01 -!- iovoid has joined.
14:59:12 <esolangs> [[NOTE]] https://esolangs.org/w/index.php?diff=84367&oldid=84359 * ResU * (+121) added self-interpreter
15:27:40 <Corbin> That one scene from "Shawshank Redemption" where he's trying to remember how to spell "Zihuatanejo" so he can meet up with his friend~
15:28:52 <esolangs> [[Special:Log/newusers]] create * Plasmath * New user account
15:35:43 <nakilon> https://i.imgur.com/8PFnh0P.png -- how do I decide that the 1st of 3 peaks is what I need?
15:35:59 <nakilon> the middle peak can be absent
15:36:53 <nakilon> the question of image binarizing
15:42:10 <river> have you thought of using a neural network to detect the (rotated) letters?
15:43:04 <nakilon> no, I'm doing better than NN can do
15:44:36 <nakilon> my captcha solver will have no NN cancer
15:47:36 -!- dutch has quit (Quit: WeeChat 3.1).
15:48:22 <river> why do you say these things?
15:49:00 <river> how are letters recognized without?
15:49:01 <myname> yeah, the most advanced technology we have to date do deal with image recognition is totally useless for image recognition
15:49:22 <nakilon> years of observations, math knowledge and marketing understanding
15:49:34 <myname> undistorted letters should be pretty easy, though
15:50:14 <nakilon> "how are letters recognized without?" -- with machine learning
15:50:45 <river> do you have a link that explains how to do this machine learning?
15:50:47 <myname> "NN is useless, just use machine learning"===
15:51:07 <myname> machine learning is in most instances done through NN
15:51:09 <Corbin> myname: Careful not to imagine that society is cumulative and always makes forward progress. Sometimes useless technologies are invented.
15:51:09 <nakilon> river what do you mean by "this machine learning"?
15:51:13 <river> i'v always been curious about making a program that can read text from an image, but never managed it
15:51:25 <nakilon> myname https://en.wikipedia.org/wiki/Machine_learning
15:51:26 <myname> Corbin: i do know that
15:52:18 <myname> nakilon: i am well aware. how does that disproof my statement?
15:52:40 <nakilon> your statement isn't proved in the first place
15:53:05 <Corbin> myname: Neural nets are generally machine-learning, yes, but not all machine-learning is done through neural nets.
15:53:20 <myname> Corbin: i never stated that
15:53:24 -!- dutch has joined.
15:53:35 <Corbin> I implemented some example code from https://arxiv.org/abs/1603.01446 the other day, and the learning is done through black-box optimization, not using any gradients or neurons.
15:53:36 <myname> it's just convinient to use for most cases
15:53:39 <river> nakilon, how to read text from an image
15:54:02 <nakilon> river I already have 10 lines long prototype that recognizes text better than a tool made by 50 people using several multimillion costing NN technologies
15:54:05 <Corbin> Maybe it's convenient with a toolkit, but NNs to me are expensive, slow, and temperamental.
15:54:27 <river> im not the one arguing with you
15:54:30 <nakilon> and the prototype isn't even aware of language dictionaries
15:54:31 <river> the chat is very busy
15:54:34 <myname> NN is so big that you are probably easier off just using whatever NN framework you have available in your language of choice than writing a machine learning algorithm from scratch
15:54:36 <nakilon> so once I add them it will do even better
15:54:40 <river> hopefully you are not mixing my messags up with anyone elses
15:55:11 <nakilon> river I understand, it's hard to talk meaningful when every 2nd line is a purposeful flame even if not yours
15:55:21 <river> yes it is quite frustrating
15:58:02 <myname> having a prototype that recognizes text better than a tool made by 50 people is a bold claim. unless the other tool isn't specifically designed to recognize text, i highly doubt you can give a reasonable metric for that
16:01:43 <nakilon> their tool was made specifically for text recognition, my tool is just a wrapped on another thing of mine that isn't even about text
16:01:57 <nakilon> it was just a library usage example
16:03:34 <myname> in that case, better publish a paper or even better sell that product
16:04:41 <Corbin> myname: Aha! Okay, this is an understandable perspective. I'm usually working in a language I built, so I have to pay the cost of every abstraction. NNs would take a lot of work.
16:04:48 -!- vyv has joined.
16:05:53 <myname> that's fine. but if you just want to recognize something and are using python (and aren't doing it for research purposes), there is not much benefit writing a specialized solution over taking one of the larger existing frameworks
16:10:51 <nakilon> I make this captcha solver for just a personal usage script and maybe will reuse somewhere else in future
16:11:30 <nakilon> but I checked how much it costs to solve your captcha with a service where a thousand of people are sitting and and solving them
16:12:10 <nakilon> it's ~$1.5 per 1000 captchas but they are doing it in 30 seconds in average while my one is gonna work within 5 sec
16:13:14 <myname> "yeah, i have this technology written in like 10 lines of codes that could generate me easy passive income, but i will only use it for personal usage, because who needs free money anyways"
16:46:26 -!- craigo has joined.
16:46:37 -!- river has quit (Quit: Leaving).
16:49:20 <b_jonas> not totally useless, alas, but it's overmarketed and somewhat overused
16:52:03 <b_jonas> a lot of products have some kind of useless modern machine learning part that is technically in it but not really needed or used, just so they can market it as a product that uses machine learning, because that's fashionable. (blockchain is similar, but even worse; in most products advertised with it it's absolutely useless, as opposed to just mostly useless)
16:52:37 <b_jonas> it used to be JIT and multithreaded, then GPU programming, but the marketing fashions are changing
16:54:38 <myname> b_jonas: did you know about that first idea about an electronic vaccination passport in germany?
16:57:01 <b_jonas> *sigh* no, I haven't been following that political nonsense and I'm not too interested in it
17:00:35 <nakilon> I guess I'll just take the peak within the section that ends when it's the first fall to 20% of the peak
17:08:53 -!- arseniiv has joined.
17:11:44 -!- Sgeo has joined.
17:16:25 -!- tech_exorcist has joined.
17:23:51 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
17:26:48 -!- lukalot has joined.
17:34:41 -!- lukalot has quit (Quit: Leaving).
17:35:01 -!- lukalot has joined.
17:35:19 -!- lukalot_ has joined.
17:35:20 -!- lukalot_ has quit (Client Quit).
17:38:33 <zzo38> Is it possible to define a escape sequence that cancels any further responses from the terminal emulator for previously accepted escape sequences (but does not cancel further ones)?
17:41:28 <b_jonas> but where does Zach propose to get all the energy for the electrolysis? (re "https://www.smbc-comics.com/comic/titan")
17:42:02 <pikhq> well it's a nuclear sub, so presumably nuclear power
17:42:15 <pikhq> wasteful? yes. possible? also yes
17:44:59 <b_jonas> zzo38: wouldn't work too well, because there's a race condition: you wouldn't know when the terminal actually reacted to the escape sequence, and in fact the terminal will likely respond to its input in sequence, so it will react to previous escape sequences before it even reads your recent one. what you can do is to send some escape sequence to which the terminal does react by a response that you can
17:45:05 <b_jonas> recognize (you can make that from cursor reads unless the terminal is like 1x1 character in size), then read from the terminal and skip everything until you see that reply
17:47:04 <b_jonas> zzo38: there's also an ioctl or some such system call api to flush the input queue of the terminal, but that only really works for a local linux console on /dev/ttyN , for anything else there's a race condition because the terminal logic that gives you responses is in an asyncroniously running process or even on another host
17:48:02 <zzo38> I know that it won't work for remote terminals; it is for local terminals that I ask about
17:48:11 <b_jonas> for the local terminal, this is useful and getty or its equivalents do us it, so that the new login doesn't get input left there by a previous login, possibly from a malicious user. for a remote terminal like ssh, you usually start a new terminal connection instead; for serial it's a bit more complicated
17:48:16 <zzo38> (specifically, local pseudoterminals)
17:48:29 -!- Koen has joined.
17:48:40 <b_jonas> zzo38: for local pseudoterminals, the terminal emulator process is still asynchroniously scheduled
17:49:12 <b_jonas> the latency is probably less than if you're talking through an internet connection on ssh, but qualitatively it's the same
17:50:02 -!- Koen has quit (Client Quit).
17:50:30 <b_jonas> on /dev/tty this can only work because the kernel can and does actually access the read queue of the terminal and flush it and does this in a way that synchronizes with its other terminal processing
17:51:11 <b_jonas> you can still flush a pty's queue the same way, but it doesn't really solve the race condition, you still need to wait for the terminal's response to some command you send after that and skip previous responses
17:51:28 <zzo38> For example, if you put such a escape sequence in the shell prompt string then any responses to other escape requests (e.g. cursor position, window title, etc) will not be seen by the shell. This can also include keyboard mode; if previous escape sequences set some mode, any keys will be transmitted using the default keyboard mode instead of the new one.
17:51:36 <b_jonas> unless you can just start a new virtual terminal, which you can actually often do and may be easier (but slower) than resetting an existing one
17:52:00 <zzo38> (It could reset other modes too)
17:53:23 <b_jonas> zzo38: that's two different problems. there is a sequence "\ec" that resets almost every setting (in some cases it keeps the palette, window size, fonts, window title and icon title, etc), but you still have to skip the input
17:53:50 <b_jonas> for which I recommend sending a sequence of absolute cursor moves and cursor position reads and wait for the matching response, but perhaps there's a better way
17:54:29 <b_jonas> you may need something more complicated than just "\ec", I'm not sure, but "\ec" is what does most of the work
17:55:24 <b_jonas> heck, it even resets the weird "\e12m" mode on linux console that, for some reason, systemd's built-in getty doesn't fix on this debian system.
17:55:51 <b_jonas> why doesn't systemd properly reset the terminal? I have no effing clue
17:58:01 <esolangs> [[Special:Log/newusers]] create * GreenThePear * New user account
17:58:06 <zzo38> There are many problems with systemd that I read about, and many people don't like it, you can use something else instead
17:59:07 -!- imode has joined.
18:00:03 <b_jonas> zzo38: on this computer there are bigger problems with the local console logins than that: the VESA bios is broken and sometimes doesn't properly set the video mode when switching from X11 to the mode that linux graphics mode console uses
18:00:10 <b_jonas> so I don't particularly care about that bug for now
18:00:48 <b_jonas> I don't know how they messed that up, the bios and the built-in video card are on the same motherboard sold together
18:01:14 <b_jonas> it's not a separate extension video card (yet)
18:05:05 <esolangs> [[User:ResU]] https://esolangs.org/w/index.php?diff=84368&oldid=84225 * ResU * (+30)
18:07:55 <esolangs> [[Esolang:Introduce yourself]] https://esolangs.org/w/index.php?diff=84369&oldid=84336 * GreenThePear * (+307)
18:10:14 -!- tromp has joined.
18:22:36 -!- river has joined.
18:44:09 <esolangs> [[NOTE]] https://esolangs.org/w/index.php?diff=84370&oldid=84367 * ResU * (+544) Added prime calculator (my own algorithm)
18:48:06 <esolangs> [[NOTE]] https://esolangs.org/w/index.php?diff=84371&oldid=84370 * ResU * (+6)
18:49:43 <esolangs> [[NOTE]] https://esolangs.org/w/index.php?diff=84372&oldid=84371 * ResU * (+6) /* Prime calculator */
18:58:33 -!- Guest32 has joined.
18:59:16 -!- Guest32 has quit (Client Quit).
19:00:01 -!- vyv has quit (Quit: Konversation terminated!).
19:03:47 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
19:19:05 <esolangs> [[Special:Log/newusers]] create * TinyGuy32 * New user account
19:30:28 -!- oerjan has joined.
19:32:35 -!- tromp has joined.
19:51:33 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
20:00:21 -!- arseniiv has quit (Ping timeout: 252 seconds).
20:01:09 -!- arseniiv has joined.
20:04:26 -!- tromp has joined.
20:20:10 <esolangs> [[What's the dog doin?]] N https://esolangs.org/w/index.php?oldid=84373 * GreenThePear * (+3893) Created
20:21:22 <esolangs> [[What's the dog doin?]] https://esolangs.org/w/index.php?diff=84374&oldid=84373 * GreenThePear * (+55)
20:22:04 <esolangs> [[What's the dog doin?]] https://esolangs.org/w/index.php?diff=84375&oldid=84374 * GreenThePear * (+26)
20:22:11 <esolangs> [[What's the dog doin?]] M https://esolangs.org/w/index.php?diff=84376&oldid=84375 * GreenThePear * (+0)
20:22:20 <esolangs> [[What's the dog doin?]] M https://esolangs.org/w/index.php?diff=84377&oldid=84376 * GreenThePear * (+2)
20:25:14 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
20:28:42 <esolangs> [[What's the dog doin?]] https://esolangs.org/w/index.php?diff=84378&oldid=84377 * GreenThePear * (-3)
20:28:54 -!- river has quit (Read error: Connection reset by peer).
20:30:26 <esolangs> [[NOTE]] https://esolangs.org/w/index.php?diff=84379&oldid=84372 * ResU * (+234)
20:35:34 <esolangs> [[NOTE]] M https://esolangs.org/w/index.php?diff=84380&oldid=84379 * ResU * (+10) slightly modified the quine so it is a valid program
20:36:18 <esolangs> [[What's the dog doin?]] https://esolangs.org/w/index.php?diff=84381&oldid=84378 * GreenThePear * (-26)
20:37:39 <esolangs> [[Esolang:Introduce yourself]] https://esolangs.org/w/index.php?diff=84382&oldid=84369 * TinyGuy32 * (+44)
20:39:37 -!- Trieste_ has joined.
20:41:06 -!- Trieste- has joined.
20:41:06 -!- Trieste has quit (Ping timeout: 264 seconds).
20:44:06 -!- Trieste_ has quit (Ping timeout: 245 seconds).
20:47:43 <esolangs> [[What's the dog doin?]] https://esolangs.org/w/index.php?diff=84383&oldid=84381 * GreenThePear * (+46) Decided to use 's for the raw value instead
20:48:16 -!- Trieste- has quit (Remote host closed the connection).
20:48:34 <esolangs> [[Language list]] https://esolangs.org/w/index.php?diff=84384&oldid=84365 * GreenThePear * (+27) /* W */
20:48:55 <esolangs> [[Language list]] https://esolangs.org/w/index.php?diff=84385&oldid=84384 * GreenThePear * (+0) /* W */
20:49:11 -!- Trieste has joined.
20:49:53 <esolangs> [[What's the dog doin?]] M https://esolangs.org/w/index.php?diff=84386&oldid=84383 * GreenThePear * (+0)
20:52:36 <esolangs> [[What's the dog doin?]] https://esolangs.org/w/index.php?diff=84387&oldid=84386 * GreenThePear * (+3) /* Expressions */
20:52:53 <esolangs> [[Esolang:Introduce yourself]] M https://esolangs.org/w/index.php?diff=84388&oldid=84382 * Plasmath * (+240) Added my signature.
20:53:07 <esolangs> [[What's the dog doin?]] https://esolangs.org/w/index.php?diff=84389&oldid=84387 * GreenThePear * (+7) /* Expressions */
20:54:39 <esolangs> [[Special:Log/upload]] upload * Plasmath * uploaded "[[File:Godencode Memory Samples.png]]"
20:58:23 <esolangs> [[User:TinyGuy32]] N https://esolangs.org/w/index.php?oldid=84391 * TinyGuy32 * (+829) Created page with "==NumberPankackes== ==Commands== 1X: pushes X to the stack (you must have a colon at the end of the command) 2 coppys the top of the stack and pushes it..."
21:00:36 <esolangs> [[User:TinyGuy32]] https://esolangs.org/w/index.php?diff=84392&oldid=84391 * TinyGuy32 * (-829) Blanked the page
21:00:50 -!- river has joined.
21:03:53 <esolangs> [[User:Plasmath]] N https://esolangs.org/w/index.php?oldid=84393 * Plasmath * (+128) Created page with "Hello! I've made some languages, here they are: (I haven't made pages for any of my languages yet, I'll put them here when I do)"
21:17:26 <esolangs> [[User:TinyGuy32]] https://esolangs.org/w/index.php?diff=84394&oldid=84392 * TinyGuy32 * (+829)
21:43:56 -!- river has changed nick to riv.
21:58:23 -!- imode has changed nick to FuckAndrewLee.
22:02:02 -!- FuckAndrewLee has changed nick to imode.
22:07:15 -!- tech_exorcist has quit (Ping timeout: 272 seconds).
22:24:10 -!- tromp has joined.
22:35:42 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
22:40:03 -!- tech_exorcist has joined.
22:44:29 -!- tromp has joined.
22:56:34 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
23:01:09 <esolangs> [[Godencode]] N https://esolangs.org/w/index.php?oldid=84395 * Plasmath * (+26343) Created the page.
23:06:02 <esolangs> [[Talk:Godencode]] N https://esolangs.org/w/index.php?oldid=84396 * Plasmath * (+1282) please help finish interpreter
23:06:23 <esolangs> [[Talk:Godencode]] https://esolangs.org/w/index.php?diff=84397&oldid=84396 * Plasmath * (-46)
23:15:37 <esolangs> [[Primes]] M https://esolangs.org/w/index.php?diff=84398&oldid=84355 * Meloons * (+0) lowercase
23:17:05 -!- craigo has quit (Read error: Connection reset by peer).
23:23:59 -!- tech_exorcist has quit (Quit: tech_exorcist).
23:37:42 -!- lukalot_ has joined.
23:40:05 -!- andydude has joined.
23:40:38 -!- andydude has quit (Client Quit).
23:40:55 -!- lukalot has quit (Ping timeout: 272 seconds).
23:43:39 -!- arseniiv has quit (Ping timeout: 252 seconds).
23:57:34 <esolangs> [[NOTE]] M https://esolangs.org/w/index.php?diff=84399&oldid=84380 * PythonshellDebugwindow * (+0) /* Commands */ Modify wording