00:00:39 !c printf("%ld\n", -180447893567515161L / 918168945243822051L); 00:00:45 0 00:09:09 -!- ZOMGMODULES has joined. 00:10:15 -!- sftp has quit (Ping timeout: 258 seconds). 00:12:22 zomg it's the modules again 00:12:32 -!- sftp has joined. 00:12:38 -!- augur has joined. 00:12:43 -!- augur has quit (Remote host closed the connection). 00:15:48 -!- variable has joined. 00:17:03 zomg 00:19:12 -!- sftp_ has joined. 00:19:23 -!- sftp has quit (Read error: Connection reset by peer). 00:20:31 ZOMG OM NOM NOM NOM NOM 00:21:23 import Zomg.Cookies(ChocolateChip,Almond) 00:22:41 main = mapM_ eat $ cycle [ChocolateChip, Almond] 00:22:45 hmm 00:22:53 > -180447893567515161 / 918168945243822051 00:22:53 -0.19653016419499658 00:22:56 > -180447893567515161 `div` 918168945243822051 00:22:57 0 00:23:00 yay! 00:23:11 > let (/) = div in -180447893567515161 / 918168945243822051 00:23:12 0 00:23:14 yay! 00:23:36 elliott: i believe unary minus has lower precedence than `div` 00:23:42 > let (/) = div in (-180447893567515161) / 918168945243822051 00:23:43 -1 00:23:47 fucking 00:24:00 gah 00:24:04 why does nothing follow C semantics here? 00:24:19 > (-180447893567515161) `quot` 918168945243822051 00:24:20 0 00:24:33 because haskell has both versions 00:24:47 because you might need both someday 00:25:03 div and mod are more mathematical 00:25:21 oerjan: yeah, but x86 does quot :) 00:25:29 while quot and rem are frequently ... yes 00:25:39 erm, does rem ignore the sign? 00:25:43 > 2 `rem` (-3) 00:25:44 2 00:25:46 argh! 00:25:58 oerjan: that's fucked up, C has quot and mod 00:26:04 heh 00:26:18 Gregor: ^ Are you sure we want to make Fythe follow such fucked-up semantics :P 00:26:27 > -2 `rem` 3 00:26:28 -2 00:26:37 These seem to be C's semantics. 00:26:43 Uhhhhhhhhhhhhhhhhhhhh 00:26:46 > -5 `rem` 3 00:26:47 -2 00:26:47 That's sort of not what you said in #plof 00:26:49 elliott: both of haskell's versions fulfil the (a%b)+(a/b)*b = a property 00:26:54 I said I was wrong in #plof :P 00:26:59 For C? 00:27:00 divisor < 0 00:27:00 That's it? 00:27:00 which is important algebraically 00:27:11 Gregor: Hay wanna fix fytheBignumMod in bignum.c? 8D 00:27:12 *== 00:27:18 It's actually fairly easy to follow for the bignum functions :P 00:27:20 x % y = sign(x) * (abs(x) % abs(y)) 00:27:25 ^^^ This is C semantics 00:27:43 Yes yes, go fix it, I'm taking a short break from bignum.c to test the bastards :P 00:28:36 me, I would just throw an exception if either argument was negative, or if the rhs was zero 00:29:03 but I'm a bastard that way 00:29:04 Gregor: i repeat, haskell's unary minus has lower precedence that `div` etc. 00:29:17 oerjan: Argh :P 00:29:22 > (-5) `rem` 3 00:29:23 -2 00:29:28 OK, still right :P 00:30:20 ERROR: (335272555365323022 * -725962439160346310) % (-640621279300857530 * -981358903806926303) = -243395282076532154171070906535748820 00:30:20 Fythe thinks: 0 00:30:23 Well that's not good. 00:30:49 i'm not sure using a negative divisor with this stuff is recommended under any circumstances 00:30:54 -!- augur has joined. 00:32:15 ERROR: (335272555365323022 * -725962439160346310) % (-640621279300857530 * -981358903806926303) = -243395282076532154171070906535748820 00:32:15 Fythe thinks: 0 00:32:17 Well that's not good. 00:32:19 got throttled there 00:32:21 dunno if that got through 00:32:32 it did 00:33:18 -!- augur has quit (Remote host closed the connection). 00:33:20 Wrong channel :P 00:33:48 Gregor: There's overspill :P 00:34:00 Gregor: I'm gonna have to fix fytheIModBignum myself, aren't I :( 00:34:53 -!- augur has joined. 00:38:34 -!- augur has quit (Remote host closed the connection). 00:38:37 i'm not sure using a negative divisor with this stuff is recommended under any circumstances 00:38:42 oh but you might need it someday 00:39:04 who is the library designer to second-guess what the library's users may need someday? 00:40:06 oerjan: literals default to integer in haskell right? 00:40:18 > :t 4 00:40:19 : parse error on input `:' 00:40:21 :t 4 00:40:22 forall t. (Num t) => t 00:40:41 yes i know :) 00:40:42 looks like no 00:40:55 "ooh! numberlike!" 00:42:15 ZOMGMODULES: yes, but in ghc -e 00:42:19 defaulting rules 00:44:03 me, I agree with lambdabot. in general, literals shouldn't prescribe their type 00:44:19 4.0 is an integer 00:44:52 ew 00:44:52 i mean 00:44:53 yeah it is 00:44:54 but ew :) 00:46:10 if you think it isn't you've been spending too much time with computers 00:46:49 ZOMGMODULES: as in pretty much all my life, yes 00:46:50 I mean 00:46:51 yes 00:46:53 it obviously is an integer 00:47:01 but nobody writes "4.0" unless the context is non-integers 00:47:10 so it's a fairly good hint to a computer that you don't want a pure integral type imo 00:47:19 elliott: default defaulting is to Integer, then Double; it has nothing to do with _literals_ per se 00:47:27 YES YES YES 00:47:51 :t 4.0 00:47:52 forall t. (Fractional t) => t 00:49:01 > 4.0 :: Integer 00:49:02 No instance for (GHC.Real.Fractional GHC.Integer.Type.Integer) 00:49:02 arising f... 00:49:21 :t let x -> 10 in x / 2; goto 10 00:49:23 parse error on input `->' 00:49:48 that's wrong in so many ways 00:50:05 unlike say fractional integers 00:50:31 :t 6/3 00:50:33 although someone _did_ hack together a partial BASIC DSL in Haskell 00:50:33 forall t. (Fractional t) => t 00:51:08 forall t. (Line Number t) => t 00:51:49 oerjan: it was lennart right? 00:51:52 LLVM-powered :D 00:52:07 i cannot recall, i thought it was pure haskell... 00:53:06 oh maybe it was a monad for compiling into LLVM 00:53:34 yeah 00:56:33 -!- augur has joined. 00:58:14 I am here for five minutes. Let's see what people have said. 00:58:30 hello 00:58:39 NO ONE TELL ANYTHING 00:59:29 tswett, there's a new person 00:59:42 Hi, new person. 00:59:45 but he's just a sham! 01:01:07 he will totally convince tswettbot to eat itself though OM NOM NOM NOM 01:01:23 that may be. 01:01:35 * tswett nods. 01:01:44 Who's the new person? 01:03:43 i'm an old person, actually 01:04:03 * tswett nods further. 01:04:06 Well, good night, everyone. 01:04:25 but definitely a new sham 01:05:21 :t 5 / "hello" 01:05:22 No instance for (Fractional [Char]) 01:05:23 arising from a use of `/' at :1:0-10 01:05:23 Possible fix: add an instance declaration for (Fractional [Char]) 01:05:46 ngood suggestion! 01:05:49 *good 01:06:01 newsham: you might want to file for a name change just to avoid the puns 01:07:42 i'm not one to shy from puns 01:08:09 like two peanuts in central park, one was assaulted. 01:08:19 peanut 01:08:55 newsham: erm. be very careful 01:08:57 oerjan is in charge of puns around here. 01:09:02 you don't want to get on his bad side 01:10:06 i'm just annoyed no one got my pun without spoonfeeding 01:10:27 oerjan: um you realise people get your puns then ignore them 01:10:33 NO WAY 01:10:50 that cannot be legal 01:11:09 I need an emoticon which succinctly signifies "ACKNOWLEDGE PUN" and nothing else 01:11:23 :X: perhaps 01:11:44 the official response is "*groan*" 01:11:45 a home groan emoticon 01:11:50 at least thats what I get 01:11:56 :X: 01:12:14 newsham: here we swat instead. 01:12:19 but that's usually a fuss to type. 01:12:26 ?slap oerjan 01:12:26 go slap oerjan yourself 01:12:51 * oerjan swats newsham -----### 01:13:01 ?slap ZOMGMODULES 01:13:02 * lambdabot smashes a lamp on ZOMGMODULES's head 01:13:10 ?slap ?slap 01:13:10 I'd rather not; ?slap looks rather dangerous. 01:13:15 ?slap lambdabot 01:13:15 * lambdabot slaps lambdabot 01:13:25 ?shapr shapr 01:13:25 I'd rather not; shapr looks rather dangerous. 01:13:27 ?slap auh 01:13:27 * lambdabot will count to five... 01:13:39 ?asdfasdf asdfasdf 01:13:39 Unknown command, try @list 01:13:44 ?shapr Sgeo 01:13:44 *SMACK*, *SLAM*, take that Sgeo! 01:13:57 o.O at shapr 01:14:06 ?slap shapr 01:14:07 I'd rather not; shapr looks rather dangerous. 01:14:36 ?shapr Sgeo 01:14:37 * lambdabot moulds Sgeo into a delicous cookie, and places it in her oven 01:14:50 ?vixen kinky! 01:14:50 i can be very kinky at times 01:15:09 if we abuse lambdabot too much it'll get taken away again! :p 01:15:40 s/it/she/ 01:15:49 sheeeit 01:17:26 * Sgeo is having some difficulty finding an OpenGenera emulator 01:17:44 brad parker has VLM which runs in x64 linux 01:18:03 * ZOMGMODULES is having no difficulty at finding an OpenGenera emulator. Of course, he's not actually trying. 01:18:28 http://weblog.mrbill.net/archives/2008/05/18/finally-got-open-genera-running/ 01:18:37 http://www.unlambda.com/download/genera/ 01:18:44 docs here http://www.cliki.net/VLM_on_Linux 01:19:07 I saw unlambda, just didn't realize it also had OpenGenera 01:21:01 newsham: oh, is that newer than the one going around on teh torrentz? 01:21:16 I have the OpenGenera distribution, but the snap4 emulator is buggy as all hell 01:21:18 i dont know. 01:21:26 hmm 01:21:39 talk to brad, he was talking about rolling a new dist but i dont think he ever did 01:21:59 btw, he's got a huge refrigerator sized lispm that he's giving away free (as of earlier this week) if anyone wants one.. arlington ma. :) 01:22:44 hehe 01:22:46 i was gonna buy a lispm 01:22:49 but no way i'd be able to afford the shipping 01:22:56 of one of the real ones, I mean, not any of this macivory crap :) 01:23:16 ooh 01:23:18 Vorpal: http://www.cliki.net/Linux%20VLM%20workarounds 01:23:22 Vorpal: that's a workaround for having to start it as root 01:26:21 I take it no one has ever made a smalltalk machine? 01:26:53 * ZOMGMODULES feels space and time start bending around him 01:28:31 Sgeo, you should make one. 01:28:48 * Sgeo doesn't have the attention span 01:31:24 * ZOMGMODULES is insulted 01:31:30 oh, not at Sgeo 01:31:48 at Bitbucket, for not including BASIC in their "What language is this project in" dropdown 01:32:28 They do have R though. God bless their little hearts! 01:33:04 They have like a half-dozen languages I've never heard of, and frankly, that I doubt exist. 01:33:30 "Duby"? SRSLY 01:34:13 "SuperCollider"... I so hope that one's as interesting as it sounds 01:34:24 THINGS BETTER BE FRICKEN COLLIDING IN THAT 01:35:33 -!- pizearke has joined. 01:36:23 ZOMGMODULES: supercollider is awesome 01:36:27 ZOMGMODULES: it's an audio processing language thing 01:36:33 yaxu uses it 01:36:41 http://upload.wikimedia.org/wikipedia/en/4/4e/SuperCollider_screenshot3.jpg 01:36:48 well, yaxu's software uses it :) 01:37:06 ZOMGMODULES: duby is a boring "ZOMG RUBY WITH TYPES" on the jvm. 01:37:10 utterly disregardable. 01:37:21 also, hello pizearke 01:37:28 Oh, hey. 01:38:09 Funny question about fractran... 01:38:24 i think oerjan is the one who knows fractran :D 01:38:33 Oh, cool. 01:38:34 slightly 01:38:46 Do you know how you would go about representing a list in fractran? 01:38:58 -!- augur has quit (Remote host closed the connection). 01:38:59 Because there's no way of introducing new primes. 01:39:38 right, it would have to be encoded into the exponent of one prime (or possibly more, but one is enough) 01:39:56 So there is a way. 01:40:21 if the list elements are bits, then you can for example encode using binary 01:40:42 True. 01:40:51 0,1,1,0 -> 10110_2 = 22 01:41:26 But there's no way to use an infinite amount of primes, right? 01:41:30 no 01:42:01 a fractran program can only meaningfully use primes that divide one of the listed numerators or denominators 01:42:17 i.e. a finite number 01:43:37 That's what I though. 01:43:38 *t 01:44:12 basically a fractran program essentially ends up using the exponents on its primes as registers, and the rest of the trick of programming it is to treat that as a minsky/counter machine 01:45:12 well not exactly but essentially 01:45:46 I get you. 01:49:07 whoa, fractran, sounds weird. 01:49:13 -!- augur has joined. 01:51:25 the not exactly is because you also need some primes for the states of the minsky machine 01:52:00 newsham: that conway is a freaky guy 01:52:09 he should get a life 01:52:32 i think it's a bit late for that, how old is he? 01:52:43 :X: 01:52:55 and oerjan missed it?? 01:53:02 74 01:53:02 [15:07] < elliott> oerjan is in charge of puns around here. 01:53:07 ZOMGMODULES: you should really fix the part where you eyes 01:53:09 ZOMGMODULES: DAMN 01:53:13 and also an X for a mouth 01:53:30 elliott: IT'S SUPPOSED TO BE A BASKEY 01:53:33 *BASKET 01:53:36 er 73 01:54:13 "Conway resides in Princeton, New Jersey. He has seven children and three grandchildren, and has been married with his wife Diana since 2001." 01:54:16 he's not even been alive for a full generation! 01:54:17 DO YOU GET IT 01:54:22 "GENERATIONS" IN THE GAME OF LIFE HAHAHAHAHAHAHAIRUHSJFGOUJRWHTREIWJQOFDGHUIEWQ9EWNG 01:54:47 elliott: i just got caught up in the literal meaning, sheesh 01:54:52 also, carrots 01:55:30 DO 01:55:31 YOU 01:55:32 GET 01:55:36 also i wasn't complaining at oerjan 01:55:46 no i'm just waffling around 01:56:18 ironic 01:56:19 after some amount of reddit browsing, that is an automatic reaction to seeing HAHAHAHA 01:56:26 where are his other forty-six grandchildren, if he has seven children, he should have forty-nine grandchildren 02:00:13 -!- augur has quit (Remote host closed the connection). 02:00:22 ZOMGMODULES: there were so many that most died from overcrowding 02:00:24 hmm.. seems like fractran could make a tiny interpretter 02:01:04 newsham: and probably has 02:01:15 oerjan: OF COURSE 02:02:00 would be amusing if someone twisted a fractran program inside a larger program as some sort of integrity check as part of a copy protection scheme or something 02:06:17 newsham: the J interpreter is tiny :) 02:06:21 for fractran, that is 02:06:58 dont know J (know of it) 02:07:08 newsham: it's like APL but better! 02:07:46 -!- Zuu has quit (Read error: Connection reset by peer). 02:08:44 The problem with unoptimized fractran interpreters is that writing anything remotely practical is impossible. 02:09:07 wait, how can *fewer* Greek letters be an improvement? 02:09:28 Unless it's in a programming language with infinitely large integers... 02:10:22 pizearke: Like all decent programming languages you mean? :) 02:10:42 What do you mean by that? 02:11:18 Well, most high-level languages have bignums. 02:12:06 -!- Zuu has joined. 02:12:28 -!- augur has joined. 02:12:38 !c printf("%lld\n", (long long) 2 << 63) 02:12:44 0 02:12:47 ... 02:12:49 ... 02:12:55 !c printf("%lld\n", 2LL << 63LL) 02:12:56 -!- augur has quit (Remote host closed the connection). 02:12:57 0 02:13:01 I disagree. 02:13:03 !c printf("%lld\n", (long long) 1 << 63) 02:13:04 -9223372036854775808 02:13:13 Gregor: 2 << 63 == 2^64 isn't it 02:13:14 And so 0 02:13:19 Or... 02:13:20 Erm, 2 lol :P 02:13:20 Something like that. 02:13:23 2 * 2^63 02:13:24 = 2^64 02:13:30 !c printf("%lld\n", 1LL << 63LL) 02:13:32 -9223372036854775808 02:13:33 !c printf("%lld\n", 1LL << 62LL) 02:13:35 4611686018427387904 02:13:45 !c printf("%lld\n", 1LL << 63LL - 1) 02:13:48 4611686018427387904 02:13:57 Uhhhwtf 02:13:58 !c printf("%lld\n", (long long) 1 << ((long long) 1 << 63)) 02:14:00 1 02:14:09 !c printf("%lld\n", 1LL << 63LL - 1LL) 02:14:12 4611686018427387904 02:14:12 ORLY 02:14:33 ... shouldn't the absolute values of those numbers be off by one? 02:14:39 !c printf("%lld\n", 1LL << (1LL << 63)) 02:14:42 1 02:14:48 !c printf("%lld\n", 4611686018427387904LL) 02:14:51 4611686018427387904 02:14:52 !c printf("%lld\n", 4611686018427387904LL + 1) 02:14:55 4611686018427387905 02:15:00 DAMN IT C 02:15:19 X-D 02:15:55 omg this guy on rationalwiki is like Sgeo! 02:15:57 !c printf("%lld\n", 1LL << (1LL << (1LL << 63))) 02:15:59 2 02:16:05 elliott, oh? 02:16:25 Sgeo: yes 02:16:31 Linky 02:16:51 i have this vision of a single 1 bit getting v. dizzy 02:17:06 Sgeo: no 02:17:19 :( 02:17:20 * Sgeo sads 02:17:46 Sgeo: you don't want a link :) 02:17:54 Why not? 02:17:55 > let ft = unfoldr . (find ((==1).denominator) .) . mapM (*) in numerator <$> ft [17/91, 78/85, 19/51, 23/38, 29/33, 77/29, 95/23, 77/19, 1/17, 11/13, 13/11, 15/14, 15/2, 55/1] 1 02:17:57 Couldn't match expected type `(a, b)' 02:17:57 against inferred type `GHC.Re... 02:18:13 grmbl 02:19:29 oh duh 02:20:52 * ZOMGMODULES waits patiently for fix making awesome fun time 02:24:28 > let ft = unfoldr . ((((id&&&id)<$>).find ((==1).denominator)) .) . mapM (*) in numerator <$> ft [17/91, 78/85, 19/51, 23/38, 29/33, 77/29, 95/23, 77/19, 1/17, 11/13, 13/11, 15/14, 15/2, 55/1] 2 02:24:30 [15,825,725,1925,2275,425,390,330,290,770,910,170,156,132,116,308,364,68,4,... 02:24:41 -!- azaq23 has joined. 02:24:43 fix is a little bit ugly :( 02:24:57 it would make excellent graffitti 02:27:00 yes 02:32:26 > let ft = unfoldr . (((join (,) <$>).find ((==1).denominator)) .) . mapM (*) in numerator <$> ft [17/91, 78/85, 19/51, 23/38, 29/33, 77/29, 95/23, 77/19, 1/17, 11/13, 13/11, 15/14, 15/2, 55/1] 2 02:32:28 [15,825,725,1925,2275,425,390,330,290,770,910,170,156,132,116,308,364,68,4,... 02:32:35 that looks better 02:34:28 newsham: ^ 02:56:52 Sgeo: http://webcache.googleusercontent.com/search?q=cache:cOXGIY9Gl1kJ:desk.org:8080/ToytownStLab/Smalltalk%2520hardware+smalltalk+machine&cd=4&hl=en&ct=clnk&gl=us&client=ubuntu&source=www.google.com <-- stuff about Smalltalk machines 02:58:46 In other news, I just got word from cpressey that he's migrated his crap onto Bitbucket. The publication model of his website will stay the same for now (modulo internal workflow differences) but he might start moving a few of his projects to be public Bitbucket projects, as a sort of pilot test. 02:59:55 https://bitbucket.org/catseye/ 03:00:06 ZOMGMODULES: but that's so... 03:00:08 conventional... 03:00:30 I accuse that Pressey fellow of reading The Cathedral and the Bazaar. 03:01:14 free private version control hosting is free private version control hosting, man 03:01:21 I did say "might" 03:01:55 free beats paying svnrepository.com $5 a month or whoever that is and whatever it is they're charging him 03:02:03 ZOMGMODULES: Define "private" :P 03:02:16 I mean 03:02:22 "but he might start moving a few of his projects to be public Bitbucket projects, as a sort of pilot test." 03:02:33 but you...sorry, Chris was all going on about *private* hosting :P 03:02:42 I suppose when people say they're going to move a few things, they don't mean all things. 03:02:44 what was i going to say. 03:03:05 ZOMGMODULES: heck, bitbucket at $100/mo would be better than svn hosting that pays you because svn 03:04:29 i smell an ig nobel price http://news.sciencemag.org/sciencenow/2011/04/scienceshot-safe-sex-duck-style.html?rss=1 03:04:39 *prize 03:04:42 svn was awesomesauce when the alternative was cvs 03:05:04 http://news.sciencemag.org/sciencenow/assets/2010/07/30/sn-ducks.jpg pictured: duck penis 03:05:17 ZOMGMODULES: at least cvs was fast :) 03:05:43 i dunno though, that's basically a 2-year limit on svn being acceptable for projects that aren't too big for darcs 03:05:49 svn - 2000; darcs - 2002 03:06:03 so I'm only 9 years out of date I MEAN HE 03:08:17 you can search Bitbucket with "haskell" to find all 140 haskell projects hosted there; but "haskell game" will not return any of the games written in haskell. 03:08:59 ZOMGMODULES: i smell trolling! 03:10:25 OMG AND THEY SUCK TOO 03:10:30 er 03:12:32 OH OH OH https://bitbucket.org/zoibot/bfhs/overview 03:13:07 https://bitbucket.org/zoibot/bfhs/overview 03:13:13 >:( 03:13:19 mem = array (1,20000) [(i,0) | i <- [1..20000]] 03:13:38 lol 03:13:39 I am stunned 03:13:43 why 03:13:53 oh wow that is the worst program 03:13:56 i just looked at it 03:13:57 Very high-quality code you find in these projects 03:14:11 lol, bos rules the haskell results 03:14:21 ZOMGMODULES: github proably has higher-quality stuff ;P 03:14:23 *;) 03:14:26 ;P is Vorpal smiley. 03:16:54 elliott: probably probably 03:19:02 much more active development in python projects... lua too, it seems 03:19:49 yes, all 3 people who use lua because they don't like python also use hg because they don't like git 03:20:15 * ZOMGMODULES nods vigorously 03:20:56 i like how searching for "ruby" and "erlang" have in their first few hits a project like "foo_python is a version of foo_{ruby,erlang} for python" 03:22:42 -!- ZOMGMODULES has quit (Quit: evaporated). 03:30:00 -!- augur has joined. 03:30:44 -!- azaq231 has joined. 03:31:03 "In newLISP, all variables are dynamically scoped by default." 03:31:19 * Sgeo ... slightly mindboggles, but maybe it makes more sense if I look at it more 03:32:05 sounds like oldlisp 03:32:17 -!- azaq23 has quit (Ping timeout: 252 seconds). 03:32:17