00:00:21 -!- oerjan has joined. 00:00:43 does "removing the syntax from a language" mean projection it into its computational class? 00:01:12 because only the semantics are left 00:01:40 projection into its paradigm i think 00:01:41 -!- Oj742 has joined. 00:01:52 like lambda calc->functional 00:02:01 projection onto some shit i just made up 00:05:44 Koen_, what do you... no, no it doesn't 00:07:14 poppycock syntax consists of two elements: poppies and cocks 00:07:31 ah, so it's like daisyworld but with sexual reproduction 00:11:29 Phantom_Hoover: well, I've always thought a "programming language" is just a set of "programs", and a program is just a couple (text, meaning) where 'text' is what you would write in the program file and 'meaning' is what it does (ie a function input -> output). now if you remove the syntax, that means you remove the 'text' component of programs; so the C program int main() {write(1, "Hello World!"", 12); return(0);} and the brainfuck p 00:11:29 ++++++++++[>+++++++>++++++++++>+++>+<<<<-]>++.>+.+++++++..+++.>++.<<+++++++++++++++.>.+++.------.--------.>+.>. become equivalent 00:12:24 Koen_, well the thing is that equivalence of programs is at least as hard as the halting problem 00:12:28 where's the return 0 in bf 00:12:35 feeling ripped off here 00:13:11 also what Bike said (IO is a complete nightmare to formalise) 00:13:32 Bike: yeah I was thinking the same thing when I wrote that :( 00:14:06 not even that. you could say that the "output" of the function described by BF is like, the value in the zeroth cell, but then you just declared a whole shitload of programs equivalent 00:14:10 like ais' proof that wolfram's turing machine is TC ran into some controversy because of the particulars of the IO 00:14:21 or you could combine all the cells in whatever way, but then what about other ways 00:16:07 Bike: yeah that's something I'm always frustrated about when people write a proof that a language is TC and then someone else uses reduction from that language to prove a second language tc 00:16:18 what? 00:16:24 "did the reduction and the initial proof used the same convention for IO?" 00:16:32 use 00:16:45 why would that matter 00:20:18 -!- Sgeo has joined. 00:24:03 Phantom_Hoover: youre horrible 00:24:47 augur_, well the actual parser wouldn't be that hard, i think! 00:25:27 perhaps i should say, it should be a one-dimensional language :P 00:25:44 all these crazy requirements! 00:26:02 what does one-dimensional mean? 00:26:12 is scheme one dimensional 00:26:13 Phantom_Hoover: like a normal language :P 00:26:33 yes! 00:26:40 why 00:26:47 i mean eodermdrome's graph representation is a 'one-dimensional' string if that's what you're complaining about 00:26:53 because newlines dont matter 00:27:05 -!- Sprocklem has quit (Ping timeout: 245 seconds). 00:27:07 so python is multidimensional. 00:27:17 yes, python is crucially >1d 00:27:18 how many dimensions does python have 00:27:23 who knows! 00:27:31 -!- Bike has left. 00:27:38 off to find out, i suppose 00:28:25 i don't... why are newlines so important 00:28:30 they're just another token 00:28:44 Phantom_Hoover: yes 00:29:01 but aligntment isnt 00:29:03 you could replace them with semicolons or whatever in the eodermdrome spec and it'd fit your apparent requirements 00:29:27 alignment doesn't matter in eodermdrome 00:29:56 -!- heroux has quit (Ping timeout: 246 seconds). 00:30:49 -!- heroux has joined. 00:33:42 -!- oerjan has set topic: The channel of the chimæric hellos | Magnus! | Koirammekokaan ei lennä? :( | https://dl.dropboxusercontent.com/u/2023808/wisdom.pdf | logs: http://codu.org/logs/_esoteric/ or http://tunes.org/~nef/logs/esoteric/. 00:33:52 shachaf: istr you wanted some norwegian in the topic. 00:34:14 magnus? 00:34:23 magnus! 00:34:40 oerjan: Is Tåsen some kind of a place? (It was in one of those videos.) 00:34:52 (first match today) 00:35:10 fizzie: i think it's possibly a suburb of Oslo? 00:35:43 Okay. But not a particularly special one? 00:35:55 more like neighborhood, it seems. 00:36:28 fizzie: well i cannot think of anything particularly special about it :P 00:36:47 but then i'm not too well acquainted with oslo. 00:38:26 the wikipedia article looks boring. 00:38:55 even the norwegian one. 00:39:02 * oerjan checks the turkish version 00:39:57 -!- Oj742 has quit (Quit: irc2go). 00:39:59 i'm sure it contains shocking news if i only could read it. oh wait, google.. 00:44:06 curiously, the turkish version mentions some WWII stuff the others don't. 00:44:23 (not _extremely_ shocking) 00:45:22 oerjan: help 00:45:29 shachaf: problem? 00:45:37 The nearest Norwegian speaker departed seconds ago. 00:45:51 I AM STILL HERE YOU INSENSITIVE CLOD 00:46:02 wait, did you kill him 00:46:12 no, he went home 00:46:16 good, good 00:47:31 google translate seems to be somewhat buggy with https. or maybe it's the turkish wikipedia. 00:48:28 i like bugs 00:49:53 -!- Phantom_Hoover has quit (Ping timeout: 265 seconds). 00:50:19 hm seems you do, i just got to the point in the logs. 00:51:44 * oerjan overloads his cheese quota on that youtube link. 01:02:27 -!- Koen_ has quit (Quit: Koen_). 01:06:05 -!- Phantom_Hoover has joined. 01:11:01 -!- conehead has quit (Quit: Computer has gone to sleep.). 01:30:09 "After all, the great advantage of BosonSampling is that, unlike with (say) factoring or quantum simulation, we don’t actually care which problem we’re solving!" 01:30:25 (new shtetl-optimized post) 01:33:07 -!- Bike has joined. 01:35:11 -!- ter2 has changed nick to tertu. 01:35:13 -!- Nisstyre-laptop has joined. 01:35:46 -!- Nisstyre-laptop has changed nick to nisstyre. 01:37:47 If there some MUD game that has a flag you can enable which makes it so that any other players can attack you as if you are a NPC, and kill you, earn all of your stuff and experience points and whatever (it is only one-way; to kill other players they have to set a flag too)? 01:39:02 you accidentally the question. 01:39:15 s/If/Is/ 01:41:39 oerjan: What does that mean? 01:42:12 what does what mean 01:42:53 "After all, the great advantage of BosonSampling is that... 01:45:28 -!- shikhin has joined. 01:45:43 zzo38: that they're just trying to make a physical experiment that has a proof that it's (probably, assuming generally believed computational complexity conjectures) hard to simulate with a classical computer, and don't really care what if anything it computes. 01:46:07 oerjan: Ah, OK. 01:50:15 -!- Phantom_Hoover has quit (Ping timeout: 244 seconds). 02:05:52 -!- nooodl has quit (Ping timeout: 264 seconds). 02:07:30 BosonSampling sounds like a hi-fi audio product 02:07:50 -!- yorick has quit (Remote host closed the connection). 02:09:23 -!- nooodl has joined. 02:12:12 -!- Phantom_Hoover has joined. 02:13:01 Jafet: different boses 02:13:06 "Boses belong to the Kayastha clan, a sub-caste of Kshatriyas (warriors/rulers) that originated from Kannauj, the capital of India during much of the classical period, and emerged in eastern India during the 11th century AD." 02:21:37 you know you regret reading a comment field when the commentors fail to distinguish cold and hot fushion. 02:22:00 -!- lexande has changed nick to lx. 02:23:20 -!- lx has changed nick to lexande. 02:45:15 how amazing that the android ssh server app keeps the server running after I tell it to shut down, and also is shit and doesn't work 02:45:27 who could possibly have predicted this outcome oh wait it was me 02:45:29 :( 02:48:43 -!- Phantom_Hoover has quit (Quit: Leaving). 02:57:38 :̀( 02:57:48 COMBINING SINGLE TEAR BELOW 02:58:04 unicode needs modular smileys 03:02:39 i distinctly see that as above 03:03:22 also in the logs, although there it's above a different character. 03:04:35 "It is well known that no model of computation characterizes exactly the set of total functions" can someone give this dummy (me) a reference better than "it's well known" 03:05:28 the halting problem? 03:05:56 oh i guess that works oops 03:07:06 well, no. rice's theorem means there's no procedure for picking out, say, primitive recursive functions, either, but that's still a model 03:18:17 it might depend on what is meant by "model of computation" 03:19:12 something i can fuck around with, preferably 03:19:34 i meant in the quote, thank you 03:19:44 well it has no context. 03:19:49 i mean obviously you can describe the set of total functions, but you can describe the primitive recursive functions, uh, "better" 03:21:06 oh, actually in context it gives primitive recursive functions and arithmetic functions as "models of computation" 03:22:23 well if we assume that you have (1) a string representation of "computations" (2) a total computable function for determining whether a string represents a computation (3) an interpreter for computation strings (on a TM or in whatever normal TC language) 03:22:41 then you might call that a model. 03:23:16 i guess (1) is a trivial point 03:23:22 is that an "and" or an "or" there 03:23:32 "and" 03:24:18 so you need a computable grammar, is what you're saying 03:24:55 yes. otherwise you could just define the grammar to be all the strings that happen to be total as functions. :D 03:25:15 well that makes some sense, thank 03:25:41 i'm still not sure that your quote is actually true for this definition though 03:26:12 what would a definition it's well-known-to-be-true for be, then 03:26:42 the missing point is a way to map total functions in some other representation into the model. 03:27:38 hm actually even then it might not be obvious, because it could do anything on non-total functions 03:28:24 no wait. that works, you could solve the halting problem if you could do that, i think. 03:28:52 er or wait 03:28:58 If you had a computable grammar for the total functions plus a map from turing machines to strings satisfying the grammar? 03:30:11 yes. although it doesn't need to do anything sensible if the turing machine doesn't represent a total function. 03:31:31 how would you solve the halting problem with that? i mean there are perfectly good functions that are only defined on total functions, like evaluation 03:31:43 i said "er or wait" 03:32:06 oh. i'm not enough of a mathematician to do things in my head and then just imply them. 03:32:16 :p 03:33:08 the thing here, is that this definition implies nothing about being able to compose these computations in any way (except by passing back via a TM) 03:33:31 yeah, that would be nice. 03:33:39 http://farm8.staticflickr.com/7051/6799351990_a716c10ae8_z.jpg 03:33:44 maybe throwing "also this is closed under composition" in there is too ugly 03:33:58 kmc: :3 03:34:06 well i don't mean just (.) composition, since that's still total. 03:34:28 there are other kinds of composition? 03:34:45 looping and stuff 03:35:04 general structured programming 03:35:52 however... what this does give you is a map from TMs to total functions which preserves the semantics of total functions. 03:35:52 hm i vaguely remember that being able to take variable powers of functions did... something 03:36:03 <^v> kmc, programming on macs is impossible 03:36:05 maybe just threw you up the gregzowieskbhch hierarchy 03:36:18 <^v> especialy with someone with add like me ;_; 03:36:21 ^v: what's the joke 03:36:32 Bike: well primitive recursion is something like that 03:37:24 oh. right. 03:38:56 oh wait, found a proof. 03:39:03 ! 03:40:34 to solve the halting problem: given a TM computation, make a new computation which calculates how many steps it takes. translate this. run the translation (which is total.) run the original TM for the number of steps returned. 03:41:14 How many steps it takes to halt? 03:41:18 yes. 03:41:56 if it doesn't halt, the translation will return some nonsense, but in either case the final check will reveal this. 03:42:44 clever 03:43:07 thanks 03:43:37 does that actually solve the halting problem, though? i mean there's no guarantee the "make a new computation" step is effectively computable is there 03:44:10 yes there is, these are TM computations so just "normal" program embedding. 03:44:59 like, the new computation runs the TM for one step, reports back if it halts, runs another step, etc.? 03:45:00 it's really just "add a step counter to the program" 03:45:30 that's not total is it 03:45:32 yes, and it reports the number of steps taken in total. 03:45:38 no, that's not total. 03:45:44 so how do you translate it 03:45:51 but the translation turns everything into something total. 03:46:14 because the other computation model has only total functions. 03:46:41 it just gives unrelated nonsense if the original wasn't total. 03:47:11 that's the assumption of how the translation works. 03:48:21 OK, wait, so let me lay this out. We have (a) the model of general computation, which is turing machines or whatever. (b) our total model, which is a language with a total "is the input in this language" function, and a total "run the input" function. (c) a partial computable function from (a) to (b) defined for all total (a). Right? Did I miss something? 03:48:34 from (a) to sentences of (b), or whatever 03:49:27 yeah. (c) should always halt itself, even if given something non-total, though. 03:50:20 OK, I think I get how your proof works, thanks for letting me indulge myself in description i guess 03:52:01 hm, can you use this to construct a rice violator for different types of b 03:52:31 argh 03:52:36 ? 03:52:45 brain overload :D 03:53:30 * oerjan is briefly pondering if (c) _really_ needs to always halt, given that it's functions it's translating, not single computations 03:54:18 i think requiring (c) to be total lets you violate rice's theorem. you do like you did except say "given a TM computation, make a new partial function which returns true when the TM fulfills [some ricey condition]" 03:54:19 because the functions have infinite domain, so (c) can't simply test all inputs before returning. 03:54:33 or... something 03:54:50 -!- nisstyre has quit (Remote host closed the connection). 03:55:00 -!- Slereah has quit (Ping timeout: 245 seconds). 03:55:10 Bike: well i thought most of that stuff is trivial once you can solve that halting problem 03:55:17 *the 03:55:18 probably 03:55:19 maybe for "models" it would be more natural to reverse c, though... 03:55:51 -!- Slereah has joined. 03:57:09 like there's obviously a total, computable map from sentences describing partial recursive functions or [total functional language du jour] to turing machines and you don't care about the other direction /that/ much 03:57:21 or maybe you do. i'm tired 03:58:12 well if you had no map the other way your language could be as total as you wanted but you'd never be able to translate any _given_ total function to it... 03:59:10 that kind of seems okay to me? when i define a program i can use some non-TM-at-all model, like arithmetic or something 03:59:11 you would know each string computes a total function but you might have no idea of which one 04:01:08 are you still trying to figure out the original quote 04:01:16 iunno 04:01:26 right now i'm wondering how i would prove the program i work on at work is total 04:01:48 translate it to agda duh 04:01:55 obviously 04:01:59 it's probably not hth 04:02:44 i think it is actually 04:03:22 well the halting problem has been solved for Bike already 04:03:26 it's called brakes 04:04:26 think you only need "loop [easy term] number of times" and "return first element in finite list that equals [easy term]" where easy terms are just parameters and floating point arithmetic on them 04:05:25 floating point arithmetic /!\ 04:05:49 i'm not exactly working on a compiler aight 04:09:30 loop NAN number of times 04:09:57 throw modular arithmetic in there if you must >: 04:10:17 loop NAN `mod` 0 number of times 04:10:24 yes 04:10:40 :t fmod 04:10:41 Not in scope: `fmod' 04:10:41 Perhaps you meant `mod' (imported from GHC.Real) 04:10:51 fmodular arithmetic 04:10:56 i'm sure it's somehwere 04:11:07 how about modular arithmetic only and we just do floating point on top of that, jerk 04:11:43 who's the real jerk here 04:11:56 me probably 04:12:22 > [0.. 0/0] 04:12:23 [] 04:12:27 > 0/0 04:12:29 NaN 04:12:52 > [0/0 .. 1/0] 04:12:53 [] 04:13:02 so uncooperative 04:13:03 > [-1/0 .. 1/0] 04:13:04 [-Infinity,-Infinity,-Infinity,-Infinity,-Infinity,-Infinity,-Infinity,-Inf... 04:13:18 > [-1/0 .. 0/0] 04:13:20 [] 04:13:21 :( 04:13:30 > [-1/0, -1/0 .. 0/0] 04:13:32 [] 04:13:50 > [-1/0, 0/0 .. 1/0] 04:13:51 [] 04:14:33 > [1,0 .. 0/0] 04:14:35 [] 04:15:02 > 0/0 < 1/0 04:15:04 False 04:15:08 > 0/0 > 1/0 04:15:10 False 04:15:56 > 1/0 - 0/0 04:15:57 NaN 04:16:08 enlightening 04:17:01 you know that writing style where you have lots of text and you bold lots of sentences in it 04:17:20 never heard of it 04:17:30 it's the worst :'( 04:17:52 like bad comics? 04:19:17 no, that's a separate thing 04:19:30 i really don't get that 04:19:40 what does all that bolding in those comics even mean 04:20:06 -!- conehead has joined. 04:21:18 -!- ^v has quit (Quit: http://i.imgur.com/MHuW96t.gif). 04:22:38 ¯\_(ツ)_/¯ 04:23:03 -!- nisstyre has joined. 04:26:52 -!- ^v has joined. 04:34:14 -!- nooodl has quit (Quit: Ik ga weg). 05:04:19 -!- Froox has joined. 05:04:52 -!- ^v has quit (Quit: http://i.imgur.com/MHuW96t.gif). 05:06:03 I managed to send a message to a channel in ifMUD while not connected. As far as I know, this wasn't intended to be possible. 05:09:23 how did you do it? 05:09:55 -!- variable- has joined. 05:11:27 -!- Frooxius has quit (Read error: Connection reset by peer). 05:11:28 -!- ggherdov has quit (Ping timeout: 248 seconds). 05:11:28 -!- variable has quit (Ping timeout: 248 seconds). 05:11:28 -!- fungot has quit (Ping timeout: 248 seconds). 05:17:23 -!- shikhin has quit (Ping timeout: 272 seconds). 05:18:02 -!- variable- has quit (Changing host). 05:18:02 -!- variable- has joined. 05:20:09 -!- ggherdov has joined. 05:20:59 -!- contrapumpkin has joined. 05:21:26 -!- copumpkin has quit (Ping timeout: 246 seconds). 05:29:59 -!- shikhin has joined. 05:31:00 kmc: If you know ifMUD, try to figure it out... (I made up a channel specifically for testing this so as not to annoy people) 05:33:14 zzo38: I don't know ifMUD. 05:33:47 shachaf: I didn't expect you to. 05:35:02 -!- shikhin has quit (Ping timeout: 264 seconds). 05:39:10 The plug for my clock keeps falling off! 05:39:34 is the time leaking out 05:39:47 it's time for a new clock 05:40:07 No, it isn't the clock that is defective; it is the position of the bed. 05:41:02 or the socket 05:41:07 -!- variable- has changed nick to variable. 05:41:19 Yes, it is also the position of the socket. 05:42:42 -!- shikhin has joined. 05:42:42 Oh, wait, "clock" doesn't mean "watch". 05:42:48 I was sure zzo38 meant a watch. 05:43:22 In this case it is not a watch; it is a digital alarm clock. 05:55:46 -!- conehead has quit (Quit: Computer has gone to sleep.). 05:57:35 the bed was the clue 05:57:57 correct 05:58:11 it' sbeen years since i confused clock and watch wow 05:58:17 c razy man 05:58:32 help 05:58:55 what is a gender-neutral phrase that has similar connotations to "whoa, dude" 06:09:41 -!- shikhin_ has joined. 06:11:55 -!- shikhin__ has joined. 06:12:29 -!- shikhin has quit (Ping timeout: 272 seconds). 06:14:38 -!- shikhin_ has quit (Ping timeout: 264 seconds). 06:15:31 'holy fucking shit' 06:17:34 -!- shikhin_ has joined. 06:19:57 -!- shikhin__ has quit (Ping timeout: 244 seconds). 06:21:31 -!- nisstyre has quit (Quit: Leaving). 06:24:28 -!- shikhin__ has joined. 06:24:31 -!- shikhin_ has quit (Ping timeout: 272 seconds). 06:28:24 -!- shikhin_ has joined. 06:31:33 -!- shikhin__ has quit (Ping timeout: 248 seconds). 06:32:20 -!- shikhin__ has joined. 06:35:38 -!- shikhin_ has quit (Ping timeout: 264 seconds). 06:36:23 shachaf: after more than a year those people finally changed the description of their users from "craftsmen" to "software engineers" 06:36:40 -!- shikhin has joined. 06:37:36 I assume this was a long painful process that involved buying a suit and burning a pile of thinkgeek shirts in a big bonfire 06:37:54 which people 06:38:02 http://javascriptmvc.com/ 06:38:08 oh 06:38:23 it's not a celebrating software engineership thing 06:38:35 almost 06:39:28 -!- shikhin__ has quit (Ping timeout: 264 seconds). 06:39:42 -!- shikhin_ has joined. 06:40:57 I made up a "Critical Vulnerability" flaw for Dungeons&Dragons game: Any damage you take from critical hits that is multiplied due to the critical hit is doubled again (so x5 becomes x10 and so on). If someone try to hit you with a damaging attack and roll a natural 1, you take 1 point of damage anyways. 06:41:57 zzo38: Are those two separate clauses of the flaw? 06:42:55 -!- shikhin has quit (Ping timeout: 245 seconds). 06:42:59 shachaf: They are two separate sentences; I don't know if you would consider them two separate "clauses" or not, or if that is important. 06:45:27 It's not. 06:45:40 The first one is bad and the second one is good? 06:47:02 -!- shikhin_ has quit (Ping timeout: 264 seconds). 06:48:44 -!- shikhin_ has joined. 06:50:56 i think they're both bad, if someone hits a 1 they usually miss entirely or worse? 06:52:32 oerjan: Yes that is correct; they usually miss entirely (or sometimes hit themself). 06:54:09 -!- shikhin_ has quit (Ping timeout: 252 seconds). 06:54:29 -!- shikhin__ has joined. 06:59:05 -!- carado has joined. 07:01:10 http://www.theonion.com/articles/giant-burrito-to-solve-all-of-area-mans-problems-f,34479/ 07:03:51 confirm 07:05:08 kmc: what about beer/bonghits 07:05:25 I think that's qdoba 07:05:42 oh it says that in the first sentence :( 07:05:54 my burrito detective skills are not needed here :( 07:07:20 -!- shikhin_ has joined. 07:08:11 have you been to costena.com yet 07:10:26 -!- shikhin__ has quit (Ping timeout: 264 seconds). 07:15:22 -!- shikhin__ has joined. 07:17:16 -!- shikhin_ has quit (Ping timeout: 264 seconds). 07:18:02 no! 07:18:07 will they sell me the World's Largest Burrito 07:18:32 probably not 07:18:49 i think someone else got the record now anyway 07:19:15 are the normal size burritos good 07:19:16 but they have good burritos 07:19:39 that's kind of far from everything 07:20:49 well, it's kind of near google 07:20:57 true 07:21:02 that's not a real place for me 07:21:03 and i think it's near evan's house? 07:21:07 oh 07:21:11 i don't actually know where evan lives 07:21:19 didn't we go there once 07:21:24 or am i mixing people up 07:21:24 yes but in a car 07:21:26 true 07:21:33 at night 07:21:37 so it might as well have been a teleporter 07:21:49 I actually walked from his house to Caltrain once but it was before I had any kind of map of the area 07:21:51 since -ito is a diminutive suffix, the world's largest burrito should be a burro. 07:21:57 true~ 07:22:15 oh, they will sell "Burrito Gigante: The 4-foot long burrito" 07:22:16 -!- shikhin__ has quit (Ping timeout: 245 seconds). 07:22:22 "Big enough for 12-16 people for only $60.00!" 07:22:36 oerjan: there's a taquería chain around here named El Faro, and another named El Farolito 07:22:38 a bargain 07:22:45 the founder of the latter is the son of the founder of the former, or something 07:23:45 Far & Sønn 07:23:57 also it's moving according to the website 07:24:00 i wonder where they'll move 07:24:15 google won't translate "taquería" into icelandic :/ 07:28:06 well icelandic wikipedia doesn't even have a taco article 07:28:09 http://is.wikipedia.org/wiki/Tortilla 07:30:06 there are also burritos at google but they do not compare 07:31:01 i heard that in arizona they call them "burros" can you confirm 07:32:12 no but I can go on a fact finding mission 07:32:16 i'll need an advance and a per diem 07:32:35 how much 07:32:47 tortilla þýðir „lítil kaka“ á spænsku 07:33:16 today i wondered how hard it is to make tortillas at home 07:35:28 shachaf: also the print version of The Onion is ceasing publication :( 07:35:48 what? 07:35:48 i heard :'( 07:35:49 dammit. 07:35:58 `rwlcome tertu 07:36:03 /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: rwlcome: not found 07:36:15 `welcome tertu 07:36:18 tertu: Welcome to the international hub for esoteric programming language design and deployment! For more information, check out our wiki: . (For the other kind of esoterica, try #esoteric on irc.dal.net.) 07:37:18 you can't do that 07:37:29 `run echo 'welcome "$@" | sed 's/aeiou//'" > bin/wlcm; chmod +x bin/wlcm 07:37:31 bash: -c: line 0: unexpected EOF while looking for matching `'' \ bash: -c: line 1: syntax error: unexpected end of file 07:38:05 `run echo 'welcome "$@" | sed '\'s/aeiou//\'' > bin/wlcm; chmod +x bin/wlcm 07:38:07 bash: -c: line 0: unexpected EOF while looking for matching `'' \ bash: -c: line 1: syntax error: unexpected end of file 07:38:14 `run echo 'welcome "$@" | sed '\'s/aeiou//\' > bin/wlcm; chmod +x bin/wlcm 07:38:19 No output. 07:38:19 `wlcm shachaf 07:38:22 shachaf: Welcome to the international hub for esoteric programming language design and deployment! For more information, check out our wiki: . (For the other kind of esoterica, try #esoteric on irc.dal.net.) 07:38:42 wow i'm so bad at this :'( 07:38:44 `revert 07:38:46 Done. 07:39:22 what is it missing other than g 07:40:05 oh, wait 07:40:16 i switched from tr to sed and did it completely badly 07:40:32 `run echo 'welcome "$@" | sed '\'s/[aeiou]//ig\' > bin/wlcm; chmod +x bin/wlcm 07:40:34 No output. 07:40:35 `wlcm shachaf 07:40:37 shchf: Wlcm t th ntrntnl hb fr strc prgrmmng lngg dsgn nd dplymnt! Fr mr nfrmtn, chck t r wk: . (Fr th thr knd f strc, try #strc n rc.dl.nt.) 07:41:33 ths s gttng t f hnd 07:42:25 `eoe oerjan 07:42:50 `echo hi 07:42:51 hi 07:43:04 `unicode > ` 07:43:06 Unknown character. 07:43:17 `ord > ` 07:43:18 62 32 96 07:43:29 `cat bin/eoe 07:43:31 cat: bin/eoe: No such file or directory 07:43:35 `which eoe 07:43:37 No output. 07:43:47 my method was so low tech oerjan doesn't even suspect 07:43:48 terrifying 07:46:04 `eoe like this? 07:46:32 except it seems almost impossible to verify if my guess is correct. 07:46:43 "tr -d would've been simpler" hth 07:47:02 oerjan: guess? just look at the logs 07:47:04 (yes) 07:47:18 fizzie: yes but i remembered something about tr being bad at unicode or something 07:47:31 shachaf: um my browser has a disturbing tendency to ignore control characters. 07:47:46 oerjan: curl http://codu.org/logs/_esoteric/2013-11-09-raw.txt | xxd | less 07:47:53 That would've been an issue only if you wanted to delete a non-ASCII character. 07:48:08 fizzie: what about when we convert everything to utf-16 07:48:31 Oh no, are we converting everything to UTF-16? :( 07:48:43 No. 07:49:02 Scary thought. 07:50:23 well i guessed correctly. 08:04:59 Do you like "Where is my keys" soup? 08:05:35 What's that? 08:06:07 It is OK if tr is bad at Unicode; that is why UTF-8 is design the way it is, that you can delete non-ASCII characters in a search/replace even if the program is otherwise ASCII only. 08:06:11 shachaf: I don't know. 08:06:59 zzo38: Why did you bring it up? 08:07:41 One of the levels in a computer game I play today is titled: "Where is my keys" soup 08:07:43 zzo38: you can't really delete specific UTF-8 characters with tr, can you? 08:08:05 kmc: Probably not with tr, but with other programs such as sed you probably can, even if it is the ASCII version of sed. 08:08:24 The ASCII version of sed operates on seven-bit characters. 08:10:06 shachaf: Yes, but just one extended to 8-bits is good enough then; you don't need Unicode or anything; it works on any 8-bit character set which is compatible with ASCII, and UTF-8 has the properties needed for it to work (Shift-JIS doesn't). 08:10:25 Therefore the same program can be used for UTF-8, CP437, and other character encodings. 08:19:46 -!- prooftechnique has quit (Ping timeout: 245 seconds). 08:21:12 For some values of "works"; you can replace single characters with the s/// command, but e.g. '.' won't necessarily match a single character, and the y/// command is not going to work any better. 08:21:47 all we need is character classes for "UTF-8 start byte" and "UTF-8 continuation byte" 08:22:01 quick somebody write a UTF-8 to UTF-16 decoder in sed 08:22:04 using Lies And Trickery 08:22:31 re-coder? whatever 08:22:47 kmc: Best is if you are allowed to design your own character classes; it would be useful for other things too 08:40:08 kmc: transcoder is the term? 08:45:57 -!- oerjan has quit (Quit: leaving). 08:47:18 -!- AnotherTest has joined. 08:49:14 Well, I did recently write one in C, that does a lot of other things too, not only UTF-8 and UTF-16, but also CESU-8, VLQ, LEB128, Modified UTF-8, linebreak conversion, byteswap, and some other things. Most of these are just an emergent property of the features of the program, though. 08:49:39 (It doesn't actually have support for CESU-8 built-in but using a pipe you can easily convert to/from CESU-8) 08:53:56 CESU-8, VLQ, LEB128, UVB-76, MDZhB 08:53:58 -!- AnotherTest has quit (Quit: ZNC - http://znc.in). 08:54:10 What's Modified UTF-8? 08:54:19 obviosuly it should be abbreviated ΜTF-8 08:54:24 shachaf: Using overlong encoding for zero 08:54:33 Did you invent that? 08:54:45 (Mainly for compatibility with C programs, and no I did not invent that.) 08:55:18