00:06:57 -!- ghosthell has joined. 00:07:41 Non Terrestrial Or Terrestrial Beings which can help me with Trans Universal Transportation (Please PM Me)9 00:07:58 wat 00:10:06 cd: it's just the usual "psychic needed, you know where to apply" call. if you don't have alien technology to time travel and build pyramids, don't worry about it. 00:10:41 [[RCEM]] https://esolangs.org/w/index.php?diff=83194&oldid=83191 * Kaveh Yousefi * (-460) Changed the Common Lisp interpreter, basing it upon a hash table instead of a vector, such that a potentially infinite tally of cells as well as a negative indices become possible. 00:10:43 i found some funny looking shapes in my back yard, i bet aliens did it 00:12:58 Not bridging joins and parts makes those things come a bit out of the blue, but bridging them would be just awful. 00:17:33 -!- TheLie has quit (Remote host closed the connection). 00:27:52 `` cat $(which mkx) 00:27:53 key=$(mk "$@") && echo "$key" && chmod +x "$key" 00:28:17 `` cat $(which mk) 00:28:18 [[ "$1" == ?*//* ]] || { echo usage: "mk[x]" file//contents >&2; exit 1; }; key="${1%%//*}"; value="${1#*//}"; echo "$value" > "$(echo-p "$key")" && echo "$key" 00:50:24 Needs some actual integration work, for now I'm just running a parallel instance of the exact same frontend except with a different root path. 00:59:38 fizzie: yeah, you told us that you're using an IRC library called IIRC "multibot" that you find hard to adapt to what you want 01:00:40 multibot is hackeso's lib. Now that i've looked through it, it's... interesting 01:00:50 very weird tbh 01:05:19 Yeah, multibot's for HackEso, and it is pretty quirky. For esolangs (logging, wiki recent changes, bridging) I'm using my own custom C++ thing, so any deficiencies in it I can just blame on myself. 01:06:04 tbh multibot is so simplistic it probably isn't too hard to write a replacement 01:06:28 it's just over 600 lines of C 01:06:30 not bad 01:07:15 Yeah, it's like the bare minimum of C to handle parsing the IRC protocol, and executing external executables (or scripts) in response to events. 01:07:29 It doesn't even do the sockets, it relies on socat for that. 01:09:24 The other bot's basically https://github.com/fis/bracket/ which is my homegrown single-threaded C++ event loop and other utilities of inconsistent quality. 01:10:43 I sort of have a long-term plan to write my own IRC client thingy. I have some existing IRC code that is terrible but contains useful notes about the IRC protocol and freenode that I'll be able to use. 01:19:46 -!- ghosthell has quit (Ping timeout: 260 seconds). 01:25:55 -!- dbohdan has quit (Read error: Connection reset by peer). 01:41:29 -!- tromp has joined. 01:45:47 -!- tromp has quit (Ping timeout: 246 seconds). 01:56:48 Ada (or a specific compiler actually) has magic variable names: 01:56:49 "Or give the variable a name that contains any of the strings discard dummy ignore junk unused (case insensitive) 01:56:49 " 01:56:57 (To suppress unused variable warnings) 02:13:49 [[Numberlang]] N https://esolangs.org/w/index.php?oldid=83195 * Batata * (+4915) Created page with "{{infobox proglang |name=Numberlang |author=[[User:Batata|Batata]] |year=[[:Category:2021|2021]] |class=[[Turing Complete|Turing Complete]] |files=.numl }} '''Num..." 02:15:07 -!- tromp has joined. 02:16:15 [[Numberlang]] https://esolangs.org/w/index.php?diff=83196&oldid=83195 * Batata * (+4) /* Reverse Cat */ 02:17:43 [[User:Batata]] https://esolangs.org/w/index.php?diff=83197&oldid=83163 * Batata * (-479) 02:19:06 Reminds me of GCC's magic fall-thru comment, with a configurable sensitivity. 02:19:21 https://gcc.gnu.org/onlinedocs/gcc/Warning-Options.html -Wimplicit-fallthrough=n 02:19:26 -!- tromp has quit (Ping timeout: 240 seconds). 02:20:31 Level 2 matches /.*falls?[ \t-]*thr(ough|u).*/ case-insensitively, but it's the list of level 3 that's really the amusing one. 02:21:25 With gems like: [ \t.!]*([Ee]lse,? |[Ii]ntentional(ly)? )? fall(s | |-)?thr(ough|u)[ \t.!]*(-[^\n\r]*)? 02:21:56 Though I guess comments for silencing warnings or linters are a little more common. 02:22:24 There was definitely something (else) where you could do an `unused_` prefix on variables though. 02:23:16 I think that's pylint. 02:23:31 the very first major mod to hackbot i made was allowing it to use /me 02:23:40 because yes 02:30:27 I found a NSA document; it looks like it was printed out from a MediaWiki-based web page and then OCR to convert it to text. 02:31:07 Presumably not our wiki, though? 02:35:29 huh 02:42:04 [[MarioLANG]] M https://esolangs.org/w/index.php?diff=83198&oldid=83182 * HaleyHalcyon * (-13) 02:45:09 [[Numberlang]] https://esolangs.org/w/index.php?diff=83199&oldid=83196 * Batata * (+18) 02:46:00 [[Numberlang]] https://esolangs.org/w/index.php?diff=83200&oldid=83199 * Batata * (+16) /* Cat */ 02:47:47 [[Numberlang]] https://esolangs.org/w/index.php?diff=83201&oldid=83200 * Batata * (+18) /* Truth-Machine */ 02:58:04 -!- tromp has joined. 03:02:47 -!- tromp has quit (Ping timeout: 246 seconds). 03:18:40 -!- Lord_of_Life_ has joined. 03:19:02 [[Special:Log/newusers]] create * 8Epic8 * New user account 03:19:56 -!- Lord_of_Life has quit (Ping timeout: 260 seconds). 03:20:03 -!- Lord_of_Life_ has changed nick to Lord_of_Life. 03:26:23 [[Esolang:Introduce yourself]] https://esolangs.org/w/index.php?diff=83202&oldid=83175 * 8Epic8 * (+167) 03:57:35 Sgeo: some languages use variables starting with an underscore to suppress unused variable warnings: in perl and rust it really just suppresses the warnings (in perl you rarely need it because the warning only applies to globals, not lexical locals), but I think in some languages (was it Prolog? Haskell? I can't remember) it is more magical and makes them unique even if the name is repeated 03:58:01 fizzie: if it's from our wiki, my guess is the BANCStar one 03:59:06 -!- j4cbo has left. 04:00:36 fizzie: as for the fallthrough, there's now also a new-style attribute [[fallthrough]] for that (both in C and C++) 04:30:42 -!- delta23 has quit (Quit: Leaving). 04:46:39 -!- tromp has joined. 04:50:56 -!- tromp has quit (Ping timeout: 246 seconds). 05:04:44 [[Special:Log/move]] move * Ilaylevy * moved [[Yogurt]] to [[Dead]] 05:05:31 [[Dead]] https://esolangs.org/w/index.php?diff=83205&oldid=83203 * Ilaylevy * (+7) 05:06:19 -!- tromp has joined. 05:10:32 -!- tromp has quit (Ping timeout: 246 seconds). 05:20:59 -!- tromp has joined. 05:25:35 -!- tromp has quit (Ping timeout: 246 seconds). 05:40:55 TIL from https://www.youtube.com/watch?v=tyixMpuGEL8 why the transparency or alpha channel of images is sometimes called the "matte" channel 05:49:22 -!- tromp has joined. 05:54:09 -!- tromp has quit (Ping timeout: 265 seconds). 06:25:52 [[Stun Step]] https://esolangs.org/w/index.php?diff=83206&oldid=57054 * Bangyen * (+108) 06:44:19 -!- sftp has quit (Excess Flood). 06:46:38 -!- sftp has joined. 06:46:49 Write about ZZT, Free Hero Mesh, tarot cards, and the gemini protocol in the wisdom file, next. 06:47:47 `? tarot 06:47:48 tarot? ¯\(°_o)/¯ 06:53:23 I'm really thinking about drawing a line in an IRC bot between handlers that are hardcoded in it and those that are deployed as a different app serverless 06:55:13 the hot reload capability would stop relying on bot implementation/language ability to do it 06:57:44 -!- tromp has joined. 06:58:18 essentially since Ruby is very dynamic it's trivial and you use the "load" method that reloads the file where you put your "constants", methods and methods stored in constants (constants are identifiers of Modules) but there is still a room for coder's error about the runtime state 06:59:19 while writing the serverless app you are supposed to think stateless and can apply measures to be sure they are called only once 06:59:54 another approach would be just to call a subprocess of course 07:01:01 but that would still need you to update the source code on bot machine that is another risk to kill the bot 07:40:32 Does anyone know an easy-to-compute metric on permutations of letters such that "carthorse" is close to "horsecart" but far from "orchestra"? 07:40:56 -!- stux|RC has quit (Quit: Aloha!). 07:46:14 today's gif and these comments remind me how I (always unsuccessfully) try to teach people IRC https://i.imgur.com/ZH7715s.png 07:47:57 -!- LKoen has joined. 07:49:39 Taneb: you could measure the distances the letters moved. 07:50:11 find all positions of the letters in the first word in the second word, compute their differences, add the differences up. 07:50:25 Taneb if it exists it should be somewhere here https://en.wikipedia.org/wiki/String_metric 07:50:29 either that or something like that via hamming distance. 07:50:44 imode: that would make carthorse far from horsecart, wouldn't it? 07:51:03 Taneb: yeah, valid. you could use some form of common edit distance algorithm. 07:51:58 there's probably something efficient for the permutations case.. hm. 07:52:41 maybe adding and then taking the average? 07:52:44 ACTION shrugs 07:52:58 Hmm, what about something with adjacencies? 07:54:10 [[I-ok]] N https://esolangs.org/w/index.php?oldid=83207 * Kwes * (+784) Created page with "= i-ok = A small 1 digit esolang, for simple 1 letter usage. {| class="wikitable" |+ |- ! chars !! desc |- | i+ or i++ || adds |- | i- or i-- || subtracts |- | i* || multiple..." 07:54:13 actually if you let the differences be negative, it might cancel out. 07:54:29 [[I-ok]] https://esolangs.org/w/index.php?diff=83208&oldid=83207 * Kwes * (-9) 07:55:50 also there is nothing matching here https://en.wikipedia.org/wiki/Edit_distance 07:56:00 so I assume the problem just isn't solved yet or is unsolvable 07:57:27 if you can assume that the string can be split only in two you may try to build something on top of length(str) splits 07:58:19 -!- stux|RC has joined. 07:59:55 or 08:02:02 str.chars.each_cons(2).sort.map(&:join) 08:02:12 results in 08:02:18 ["ar", "ca", "ho", "or", "rs", "rt", "se", "th"] 08:02:25 ["ar", "ca", "ec", "ho", "or", "rs", "rt", "se"] 08:02:26 ["ch", "es", "he", "or", "ra", "rc", "st", "tr"] 08:03:09 and now you implement Levenshtein between these "two-byte-char" string 08:05:22 you see here it's distance 2 between carthorse and horsecart 08:07:36 [[Language list]] M https://esolangs.org/w/index.php?diff=83209&oldid=83094 * Kwes * (+11) added I-ok 08:09:43 [[I-ok]] https://esolangs.org/w/index.php?diff=83210&oldid=83208 * Kwes * (+23) 08:09:57 -!- hendursaga has quit (Ping timeout: 240 seconds). 08:11:13 [[I-ok]] M https://esolangs.org/w/index.php?diff=83211&oldid=83210 * Kwes * (+0) 08:11:33 [[I-ok]] https://esolangs.org/w/index.php?diff=83212&oldid=83211 * Kwes * (+1) 08:12:00 [[I-ok]] M https://esolangs.org/w/index.php?diff=83213&oldid=83212 * Kwes * (+1) 08:12:36 you could maybe do some substring matching like the bioinformatics people do all the time to find the string "horse" in both occurences, remove it and repeat 08:12:51 [[I-ok]] M https://esolangs.org/w/index.php?diff=83214&oldid=83213 * Kwes * (+0) 08:14:05 -!- hendursaga has joined. 08:14:51 O(n)=python 08:18:12 [[User:Kwes]] https://esolangs.org/w/index.php?diff=83215&oldid=83171 * Kwes * (+23) 08:19:30 if you make some reasonable threshold, it should perform pretty well 08:21:56 [[I-ok]] M https://esolangs.org/w/index.php?diff=83216&oldid=83214 * Kwes * (+30) 08:22:57 [[I-ok]] M https://esolangs.org/w/index.php?diff=83217&oldid=83216 * Kwes * (-6) 08:22:59 if you don't know in how many substrings it should be splitted it's N^N 08:23:17 you try h, ho, hor, hors, horse on each level 08:23:22 if I understood you correctly 08:24:34 i wouldn't try it that way. there should be algorithms to find longest substrings in O(n^2). you go through that as long as the substring you find is larger than something like O(sqrt(n)) to allow for reasonable shrinking and after that you just do levensthein or something 08:35:00 oh btw, in my approach you don't need the O(n^2) levenshtein, because lists are sorted already 08:35:51 so it's O(n) to split, O(n log n) to sort and then O(n) again to calculate distance 08:36:08 n+m, whatever 08:37:15 -!- tromp has quit (Remote host closed the connection). 08:37:19 how would you do that in your example? "horse" cannot be put together by two-byte-strings 08:38:52 I don't put the horse together, I operate is as a ["ho", "or", "rs", "se"] sorted array 08:40:25 you can see this subarray included in two arrays in my copypasta above but not in the third one because there was no horse in it 08:40:49 sure it depends on the length of strings, it they are 100 chars long there might be fake horses 08:41:39 *if 08:48:22 -!- tromp has joined. 09:05:32 \help 09:05:44 is prefix \ bot-free currently? 09:15:28 -!- TheLie has joined. 09:28:16 -!- Sgeo has quit (Read error: Connection reset by peer). 09:38:18 -!- kspalaiologos has joined. 09:44:37 -!- TheLie has quit (Remote host closed the connection). 09:47:11 -!- LKoen has quit (Remote host closed the connection). 10:02:48 -!- tromp has quit (Remote host closed the connection). 10:05:17 -!- tromp has joined. 10:10:49 -!- LKoen has joined. 11:38:24 -!- Thelie has joined. 11:48:06 -!- LKoen has quit (Remote host closed the connection). 11:53:59 [[Special:Log/newusers]] create * Nooder Coob * New user account 12:06:25 [[Esolang:Introduce yourself]] M https://esolangs.org/w/index.php?diff=83218&oldid=83202 * Nooder Coob * (+355) /* Introductions */ 12:06:36 "and can apply measures to be sure they are called only once" -- forget this, since the commands I want to deploy are made in a supported language you can run Cloud Functions instead if Cloud Run so it's already solved on their side 12:19:14 [[DotSnap]] N https://esolangs.org/w/index.php?oldid=83219 * Nooder Coob * (+1094) Created page with "== DotSnap Language == ==== What is the DotSnap Language ==== The DotSnap Language is a language dedicated to my stupidity, craziness, and vast imagination yet bounded with my..." 12:25:24 what a description 12:26:24 b_jonas: Go doesn't use an underscore prefix, but it does make the plain _ a special "blank identifier", which you can use for unused parameters, or in a multi-value assignment for values you don't care about, or a few other similar cases. And you can freely repeat it. 12:28:25 in ruby _ is a valid identifier and also you are allowed to use the same identifier as parameters or multi-vaue assignment, so people use _ in the same way 12:30:42 Speaking of bot architectures, I did almost complete a refactoring once of our esowiki/esolangs bot where I left just the "critical" bits (IRC connection management and channel logging) in the executable, and all other logic would have been in separate processes that connect to it over an RPC thing and register for callbacks for events they care about. Mostly to avoid having to restart it so often when 12:30:48 messing about, and also to avoid having the logging go down if some less important bit breaks. Not quite as cloudy, but a bit in that direction. 12:37:31 zzo38: Oh, by the way -- you might have said something once about ignoring the wiki updates? I made the same bot (esowiki) do the bridge from the experimental Libera channel, so if you're ignoring it completely you might have been missing out on discussions happening over there. Sorry about that. 12:37:32 -!- tromp has quit (Remote host closed the connection). 12:45:28 Taneb: does it only have to work on words without repetitions? 12:48:09 The letter bigram bag distance thing sounded reasonable to me. 12:48:42 fizzie: "the plain _" => yes, the plain underscore is a special keyword in rust too, when you use it in a pattern it can behave differently from using a variable that you only mention once 12:49:20 fizzie: a refactor of esowiki? nice 12:50:01 "and also to avoid having the logging go down if some less important bit breaks" => what happens when the web server breaks like it sometimes does from lurk mode? 12:51:28 Fortunately that already doesn't affect logging, just serving, since it's already a separate process. In the post-refactoring model, it'd probably be much the same, except it'd use the standard RPC mechanism instead of the custom-built protocol it now does to implement stalker mode. 12:52:09 If you can call "connect to a socket and just read protos from it" a protocol. Well, maybe: both contain the substring "proto". But I digress. 12:53:43 Incidentally, stalker mode doesn't yet work for this channel's logs. 12:54:00 an RPC mechanism? in my vague plans for the vaporware irc client, the process that keeps the IRC connection alive and does only a few things so you rarely need to replace it just logs (almost) all traffic from the IRC connections into an append-only file, plus notifies subscribers after the file changes, that's the whole RPC mechanism for reading. of course that just pushes the harder part to other 12:54:07 components. writing would be a bit more complicated. 12:55:19 -!- tromp has joined. 12:55:31 That's a reasonable model too. I was going to use the "brpc" system, because, well, because it's there. 12:55:31 [[User:Batata]] https://esolangs.org/w/index.php?diff=83220&oldid=83197 * Batata * (+1) 12:55:42 admittedly reading also might be complicated, because I might want to exclude some sensitive information like passwords from the logfile yet temporarily relay them to a reader process 12:57:21 though I could put them to a separate short-term logfile 12:57:34 I dunno, ideally I never want to handle passwords read from IRC again 12:57:55 [[I-ok]] M https://esolangs.org/w/index.php?diff=83221&oldid=83217 * PythonshellDebugwindow * (+55) Clarify 12:58:10 [[I-ok]] M https://esolangs.org/w/index.php?diff=83222&oldid=83221 * PythonshellDebugwindow * (+0) fix 12:58:44 I got the impression one of Libera's (well, Solanum's) vendor-specific IRCv3 mechanisms was this thing where you get an account tag for every message, so that clients can easily know which identified user account a message came from. 12:58:51 fizzie: Is there a regular expression to distinguish the messages? 12:59:01 fizzie: freenode already has a mechanism for that 12:59:24 I even have a bad implementation for it that works correctly and so I can use it as reference for a good implementation 12:59:46 the protocol details are a bit complicated because it consists of three different parts: 13:00:02 zzo38: All the bridged messages start with " ..." and all the wiki updates start with "[[Page title]]", though the latter has color codes in it. 13:00:42 So I guess ^< should work for detecting the bridged traffic. You can see how they look like in the logs, of course. 13:01:06 (1) you use extended WHO to read the account names when you join a channel, (2) you can opt in to the server sending extended JOIN messages that contain the account name of anyone joining later, and (3) you can opt in to ACCOUNT messages that the server sends whenever someone you see on a channel changes their account (by logging in to nickserv or logging out) 13:01:23 I didn't know there was a new mechanism 13:01:45 do they also have a tag to identify connections uniquely even if they're not joined to a channel? 13:01:46 I didn't look at it so closely. Actually, looks like there's a standard IRCv3 tag: https://ircv3.net/specs/extensions/account-tag 13:01:54 as in, when they're private message me 13:02:09 That one ^ should do, since it's in every message. 13:02:12 fizzie: well ok, but does libera actually implement that? 13:02:35 fizzie: well that's more useful, though not quite what I asked about 13:02:36 I didn't really find the docs of the Libera/Solanum tag, so I only had a very vague impression it was something similar. 13:03:16 From a Libera server: Capabilities supported: account-notify away-notify chghost extended-join multi-prefix sasl tls account-tag cap-notify echo-message solanum.chat/identify-msg solanum.chat/realhost 13:03:35 So "account-tag" *is* mentioned there, but the one I was confused about was "solanum.chat/identify-msg". 13:03:46 I also want a uniqe tag to identify the connection to the server, even for unidentified nicks, so that if a nick privmsgs me twice and they're not joined a channel, I can tell if they're the same connection, and more usefully, to send a private message that the server won't deliver to an unrelated person with the same nick who connected or nicked to that nick shortly before I send them the message 13:04:09 I know you can send a message to a nick!user@host as an approximation, but that's just not the same 13:04:36 and I know the servers do have such unique tags, they use them when the servers talk to each other 13:05:12 "I didn't really find the docs" => thanks, I can just ask on #libera after the drama gets lower, the freenode admins were usually very helpful about that sort of thing 13:05:37 https://github.com/solanum-ircd/solanum/pull/4 implements it, and the comments said they're working on making those URLs link to the documentation, but they didn't yet. 13:06:08 Looking at the code, it's probably something else. 13:08:31 there are also some other small but useful features of freenode's old server and services that libera still supports and that I'm glad for, such as that it resends the 005 replies when you ask for VERSION -- not every IRC server does that, and it's a nice QOL improvement that lets you more safely make it possible that you can emergency hot-restart even the process that is connected to IRC while keeping 13:08:37 the connection alive even if it has a memory corruption and so its knowledge about its connections is mangled 13:08:43 and yes, I know that's a stupid thing and I'd never do that in production, but still 13:10:47 those 005 reply lines helpfully contain my nickname and the nick/channel case folding rule, all of which are really essential state to follow an IRC connection, and contain the network name, useful when you inherit multiple sockets and have no idea which one is supposed to be what 13:14:34 I wish the IRC protocol was at least built such that when the server sends you a NICK/JOIN/PART/KICK message, it distinguishes whether it applies to your connection or someone else, so you don't have to know your own nick, which can change, to decide, but that seems too late to fix 13:15:21 but I haven't seen an extension that does that 13:15:50 for PRIVMSG/NOTICE it's easy, you only get a PRIVMSG targeting a nick if it's you 13:25:25 I guess that's like the sort of things that compiler authors face when there's an existing programming language that used to be well-defined but then optimization technology marches on and suddenly they have to figure out exact semantics of what the optimizer is allowed to do, in smaller scale 13:27:11 and then people who write interesting C or C++ or Rust programs also want to know the exact semantics to tell whether current or future optimizers are allowed to break their code 13:28:17 and then you have to go to freenode/#freenode to ask about the guarantees of how the servers are in a forest, there can't be cycles, and what that implies about the ordering of messages originating from other servers, or to oftc/#llvm to ask about the optimization guarantees for floating point vector addition 13:28:43 even though in production code you wouldn't want really to rely on those guarantees anyway 13:28:54 it's not just me that happens to, right? I mean this is #esoteric 13:31:57 and then you wonder about how that limits future optimizers, including how you'll have to modify TCP when we colonize Mars and then make an IRC-like thing that allows cycles and breaks those guarantees when we colonize a third place with pairwise significant light-speed limit information delay among the three 13:37:24 I'm worried about the TCP thing by the way. with how slow updating basic infrastructure like that is, I really hope someone is working on suitable protocols in advance. there are a lot of theoretical problems, like you can't continue https anywhere if you want to cache webpages on every planet for faster reaction time without everyone who makes webpages having to pay for a server on every planet and 13:37:30 moon. there might be a rebound from https anywhere to allowing caching HTTP proxies for that. at least I hope so, because the alternative where only big companies like facebook and google can host webpages would suck. 13:39:08 -!- tromp has quit (Remote host closed the connection). 13:40:37 -!- tromp has joined. 13:41:01 Hmm, well. I think Cloudflare's CDN services are pretty affordable (they've got that whole Cloudflare for Individuals free tier and all), but it would be kind of a shame if you'd be *forced* to rely on services like that. 13:41:40 huh, why can't I log in to the internet bank? did they break the webpage? 13:41:53 They moved it to Mars. 13:42:23 fizzie: they might be affordable *right now*, perhaps still if we only have presence on two or three planets (like how now many providers have one server in Europe, one in America, and one in Australia) 13:42:34 but that doesn't scale well to a Galactic Empire 13:43:12 fizzie: they could, but they'd notify me of changes that can potentially cause difficulties in accessing the internetbank interface 13:43:16 good, now the webpage reacted 13:44:35 Cloudflare's CDN presumably has a *lot* of Points of Presence on Earth, so from a simple numbers perspective, if you lower the density to one / celestial body, I think you can cover a lot of ground^W space. Though the way interstellar commerce works might turn out to be *quite* different. 13:46:28 We have a bank we no longer have any products from, and they didn't say anything about what happens to the online banking credentials. So I tried it out (once right after, and once a week later), and both times it gave all the signs of passing through a login page, but then when it should have shown the landing dashboard page, the first time it gave an error message, and the second time it just gave a 13:46:34 blank page with an URL that had an error code in it. 13:46:57 I don't know if they've just not considered the situation where someone has their online banking account but no active products. 13:47:37 I'd kind of preferred to have closed the online thing "properly". But I'm not sure I want to start calling their customer support about it, especially now that I can't do it via a message sent through the online system. 13:49:54 maybe we invent practical homomorphic encryption by then (plus obviously bigger and faster hardware available for cheap, which I sort of assume with space colonization), that would make this easier, the user would just download a cached copy of a large snapshot of the state of the bank encrypted homomorphically and run it on their own machine or on the ISP's beefy box 13:50:25 and that state would verify your login details and send you the SMS for two-factor authentication 13:50:39 I mean it presumably wouldn't still be SMS, just some analog 13:53:07 "I don't know if they've just not considered the situation where someone has their online banking account but no active products." => there's a bank where I have some rights to a relative's account, but no own account with them. there are silly limitations that they do to me for this: I can't buy the service where I can authenticate myself on telephone so they can give me help on phone about their 13:53:13 services that I can access on the relative's account, even though the account owner could get such a service for free with their account. 13:55:01 "I'd kind of preferred to have closed the online thing "properly"." => they can't do that, there might still be things you want to do with a closed account on internet bank, like inform them about changes in your address or review transactions that are resolved after you closed your account where someone else claimed you payed for it from your account earlier but there was a delay, or transactions where 13:55:07 you potentially fraudulently tried to pay or write a cheque after your account was closed, or repeating online payments where the other side tries to incur a fee because they don't know you have closed your account 13:55:36 I think if you have a bank account, even if you close everything, the bank has to keep some information about it "forever", or at least for a lot of years 13:56:48 Well, fair, but they could certainly invalidate my login credentials completely, instead of having them still "work" for logging into an error page. 13:56:55 this sort of applies to a lot of businesses, but with much shorter terms, because everyone else has shorter limitations for how long they are allowed to try to say you owe them money after you stopped doing business with them than for how long you can claim that your bank owes you money 13:57:19 yes, logging into an error page is probably not the correct behavior 13:58:03 I'm also interested in what they'll do about the last annual statement, which they've so far sent only electronically, which I would normally download from the website. 13:58:16 Hopefully they'll just mail it this one time. 13:59:21 (Not that I really need one, except to complete the set.) 13:59:46 I don't think they will snail mail you the statement after you asked them to send only electronically, because some clients do that because they worried that unauthorized people steal snail mail from their postbox and the snail mail has sensitive personal information 14:00:07 you can probably get the statement if you walk in in person and ask them, or maybe ask them to mail it to you 14:10:03 does brpc stand for befunge rpc? 14:13:09 "ACCOUNT messages that the server sends whenever someone you see on a channel changes their account" -- does you get these messages when someone does nickserv identify? 14:14:14 16:03:35 So "account-tag" *is* mentioned there, but the one I was confused about was "solanum.chat/identify-msg". -- oh this is it mentioned I guess 14:14:28 nakilon: yes, but only if you opt into it with CAP or CAPAB or some such message 14:14:52 and if you and them share a joined channel at the time 14:15:05 s/such message/such magic/ 14:15:18 and this is for freenode, I haven't tested anything on libera 14:16:15 -!- b_jonas has left. 14:16:36