00:00:00 `2014 00:00:01 Hello, world! 00:00:01 `2014 00:00:02 `2014 00:00:02 No output. 00:00:02 No output. 00:00:03 `2014 00:00:04 No output. 00:00:04 `2014 00:00:05 No output. 00:00:21 You did it. 00:00:29 Woo 00:01:09 Good bye, 2014. 00:04:01 Should `2014 be kept? 00:04:17 yes 00:04:19 otherwise it defeats the point 00:06:14 2014 wasn't 2014 for just 2014, but for all time. That's true in the abstract, but it helps drive the point home having a running implementation. 00:13:34 -!- adu has quit (Quit: adu). 00:29:27 -!- fizzie` has joined. 00:29:28 -!- fizzie` has quit (Changing host). 00:29:28 -!- fizzie` has joined. 00:32:07 -!- shikhin_ has joined. 00:34:53 -!- shikhin has quit (Ping timeout: 240 seconds). 00:36:00 -!- fungot has quit (*.net *.split). 00:36:00 -!- fizzie has quit (*.net *.split). 00:36:00 -!- zemhill_ has quit (*.net *.split). 00:40:31 -!- zzo38 has quit (Remote host closed the connection). 00:50:15 -!- zemhill_ has joined. 00:50:27 -!- weissschloss has quit (Remote host closed the connection). 00:54:10 -!- pikhq has joined. 00:54:11 -!- weissschloss has joined. 00:59:19 -!- big3_brother has joined. 01:02:32 -!- adu has joined. 01:05:32 -!- Solace has joined. 01:23:11 okay, i made my first ever golfing attempt in haskell for Kimariji and needed 139 bytes 01:24:24 132 with less whitespace 01:25:23 What is white space? 01:25:47 -!- boily has joined. 01:26:40 is there something like µ recursion in haskell? 01:27:13 What is μ recursion? 01:27:25 Excuse me, µ recursion. 01:27:28 like, giving a predicate, return the largest possible value that satisfies it 01:27:50 Largest? 01:28:09 i am doing last(0:[x|x<-[1..99],pred x]) now 01:28:33 0.. might solve that leading 0 problem 01:28:56 indeed it does 01:29:20 so last [x|x<-[0..99],pred x] 01:30:22 120 now 01:30:36 i don't think i will shorten it much more with my attempt 01:33:46 oh, 121 on the page 01:33:47 oh well 01:35:19 myname: check for final newlines 01:35:58 vim does not show a second line, so i guess there is none 01:36:30 vim adds a final newline by default 01:36:35 :O 01:36:39 (at least for me) 01:36:52 how do you remove them? 01:36:57 I usually use the form for submission 01:37:16 i have 3 times the word "take" :( 01:37:31 i may reduce it to 2 01:37:33 and I don't remember how to omit the final newline in vim (it was discussed here before, it's possible somehow) 01:37:41 myname: you could use some sed trickery, I think. 01:37:57 not worth it 01:38:56 okay, can i do something like x<-"bar" to get "b", "ba", "bar"? 01:39:01 inits i guess 01:39:58 oh dear, that makes stuff worse 01:40:03 inits, but you need to import Data.List for that, so it's generally not worth it. scanl(flip(:))[][1,2,3] comes close, but generally even that is too long 01:40:34 i would need to replace one usage of take with one of length 01:40:40 doesn't sound that smart 01:41:56 ghci won't let me run that line of yours, what the hell 01:42:13 > scanl(flip(:))[][1,2,3] 01:42:15 [[],[1],[2,1],[3,2,1]] 01:42:34 it should work just like that in ghci, too. 01:42:45 i don't think that would help with this problem 01:42:54 > map reverse.tail.scanl(flip(:))$[][1,2,3] 01:42:54 it doesn't. 01:42:55 Couldn't match type ‘[a0] -> [[a0]]’ with ‘[[a]]’ 01:42:55 Expected type: [a0] -> [[a]] 01:42:55 Actual type: [a0] -> [a0] -> [[a0]]Couldn't match expected type ‘[t1] -> [... 01:43:02 meh. 01:43:14 > map reverse.tail.scanl(flip(:))[]$[1,2,3] 01:43:16 [[1],[1,2],[1,2,3]] 01:43:21 scanl/scanr do come in useful from time to time, so it's good to keep in mind that they're in the prelude. 01:44:33 myname: one can learn a lot of tricks by looking at previous entries, but usually the main trick is to come up with a good algorithm. 01:44:47 yeah 01:45:43 what i do now is building a list of numbers where the prefix is NOT identifying and adding one 01:45:47 (from previous entries you can learn tricks like defining custom infix operators (often better than alphanumeric names for function definitions), guards, pattern guards...) 01:45:57 but checking if it is identifying takes quite some space 01:46:23 -!- Phantom_Hoover has joined. 01:46:55 why the negation, why don't you take the minimum of those lengths that *are* identifying? 01:47:17 (also my algorithm is completely different, but that's all I intend to say about it) 01:47:21 i could do that 01:47:39 but i don't think it'll be much shorter 01:47:54 i waste 42 bytes to check if a prefix is identifying 01:48:59 I'd expect this to be good for a couple of characters, 3 to 5 perhaps. 01:49:00 it is 3 bytes shorter with that 01:49:43 (mostly because it's saving a +1) 01:49:48 yeah 01:50:05 but i need a == instead of a < somewhere else now 01:50:18 time to hoogle 01:51:44 when golfing haskell, are you allowed to use the fancy unicode operators? if so, do they count as single chars? 01:52:00 heh. my big Haskell golfing source file (I keep all the programs in a single file) has the output of :browse Prelude at the end 01:52:05 what fancy unicode operators? 01:52:13 boily: anagol counts bytes 01:52:39 besides {-# LANGUAGE UnicodeSyntax #-} is just too long 01:52:50 (even after omitting the two superfluous spaces) 01:53:45 int-e: noooooooooo :( 01:54:05 myname: λ, ＊ and similar. 01:54:17 what the hell do they do 01:54:21 (not sure about ＊ though, but it'd be typographically feng shui.) 01:54:48 myname: the first is lambda, the other is something that I don't think I have ever seen... 01:55:13 (too lazy to try to type the triple-bar-equal-sign. that one I know is actually used.) 01:55:16 boily: well, does it behave different than \? 01:55:17 :k Int 01:55:18 * 01:55:39 myname: absolutely not. it is purely graphical. 01:55:50 lol 01:56:18 boily: this may come as a surprise to you, but generally speaking, golf code is NOT PRETTY 01:56:18 in fact, there's some vim madness that you can use to change the way operators are displayed in your source files, but keep the underlying ascii intact. 01:56:39 int-e: OKAY. 01:56:58 * boily wants to prettify his golf 01:57:12 Actually, some people prefer to not use λ because it is a Greek letter, not a symbol. 01:59:04 -!- hjulle has joined. 01:59:27 i wish there were some kind of any2 02:00:03 any2? 02:00:43 check if there are at least 2 values that satisfy a property 02:01:57 -!- big3_brother has quit (Quit: lsd rules!). 02:01:57 -!- drdanmaku has quit (Quit: Connection closed for inactivity). 02:02:30 boily: have you seen the "beautiful" indentation here? http://golf.shinh.org/reveal.rb?Prime+Substrings/henkma%28oerjan%29_1415625927&hs 02:03:34 is there any reason not to interact with lines/unlines? 02:03:52 `2014 test 02:03:52 No output. 02:04:04 i can hardly believe getLine and putStrLn are that much shorter 02:04:45 myname: it depends, there is no general rule for that 02:04:51 okay 02:05:08 i am however curious about that brainfuck solution 02:05:14 since here, the whole input has to be read before anything can be processed, I believe 'interact' will be superior. 02:05:35 int-e: ... eghbhghhhghh... 02:05:47 boily: what's the problem? 02:05:53 looks perfectly fine for me 02:05:54 it's... disquieting. 02:05:56 m@main deserves an award 02:06:15 i do think it's quite normal to do this 02:07:04 That indentation is a ghc extension, though 02:08:48 myname: I expect the brainfuck is a variation on this idea: http://golf.shinh.org/reveal.rb?Text+Compression/nuko%28cheat%29_1201178750&bf ... the interpreter puts the program code after the data area, so it's possible to skip over the data and use verbatim strings from the program itself 02:08:53 too bad i have to access the whole input in my check, i have a seperate function just to give it a name 02:09:20 boily: ghc is the only haskell there is on anagol ;-) 02:10:16 oh 02:10:30 oh I got the direction wrong. 02:10:59 i am doing something like "let f i=do something with i in interact f" now and i am a bit unhappy about it 02:11:04 (since it's obviosuly moving left) 02:11:10 but i can't see how to avoid it at the moment 02:11:22 myname: define a global id 02:11:33 ? 02:11:46 main=interact f;f i=do something with i 02:11:51 ah 02:11:56 (or do it on two lkilines) 02:11:59 ouch 02:12:01 on two lines 02:12:42 that's indeed 6 bytes shorter 02:13:35 'let' and 'where' are hardly ever worthwhile 02:13:44 but there are exceptions. 02:13:54 if i could somewhere get rid of naming it directly, it'd be pretty nice 02:14:29 main=interact(\i->do something with i) is worth a try sometimes. 02:15:03 also, point-free code, but there's no S combinator (or anything comparable that can duplicate an argument) in the Prelude 02:15:23 too bad 02:15:49 (\i-> works but has exactly the same length 02:16:01 oh, no 02:16:07 1 byte shorter 02:16:07 > ap (+) (2*) 1 2 -- this "ap-tly" named one is in Control.Monad 02:16:08 Could not deduce (GHC.Num.Num (a0 -> t)) 02:16:08 arising from the ambiguity check for ‘e_1212’ 02:16:08 from the context (GHC.Num.Num (a -> t), GHC.Num.Num a) 02:16:18 > ap (+) (2*) 1 02:16:20 3 02:16:22 stupid me. 02:16:38 myname: are you using 'lines' twice? 02:16:53 no 02:17:54 don't see why i should do that 02:18:56 just interesting because I originally started out with main=interact$unlines.f.lines 02:19:07 i did, too 02:19:21 well, i do still 02:19:32 but why should i line there twice? 02:20:53 never mind. I'm just surprised that the \i-> helped you rather than being neutral 02:21:34 i'm down to 102, it's getting hard 02:22:38 "on" is not in prelude, is it? 02:22:45 nope 02:22:49 too bad 02:23:13 i really dislike my check 02:23:42 > let f a b c|t<-take a=t b==t c in f 3 "abcd" "abcef" 02:23:43 True 02:23:56 wait wait wait 02:24:08 what? 02:24:12 behold: pattern guard abuse 02:24:25 i don't even 02:24:36 what? 02:24:46 I'm with myname here. what the fungot is going on here. 02:24:54 FireFly: FUNGOOOOOOOOOOOOOOOOT! 02:25:00 s/FireFly/fizzie/ 02:25:31 > let f a | Just _ <- guard (a==5) = 4 | 1 > 0 = a in (f 3, f 5) 02:25:33 (3,4) 02:25:57 what the hell does take do with 3 parameters? 02:25:58 just like that, but with a more boring pattern (`t`) 02:26:00 ah 02:26:01 aaaaah 02:26:18 @quote view.pattern 02:26:18 benmachine says: let (\x -> case x of (\x -> x -> x) -> x -> x) = 0 in x / best view pattern evar 02:26:19 woah 02:26:25 that is crazy stuff 02:26:45 so you rename "take a" to t 02:26:46 But view patterns are not in Haskell 2010. Pattern guards are. 02:28:11 is it possible to do something like this in a list comprehension? 02:28:54 well, x<-[foo] and let x=foo happen to be the same length 02:29:13 -!- shikhin_ has quit (Ping timeout: 244 seconds). 02:29:17 err. wait, what am I saying 02:29:34 the former saves one character. 02:30:16 > let x|x<-x=x|x=x in x 02:30:19 mueval-core: Time limit exceeded 02:31:40 well, it actually makes my code longer 02:31:54 oh no 02:32:12 99 02:32:24 that should be enough for now 02:32:58 dream up other algorithms 02:33:31 i am quite happy with it as is for my first ever golfing attempt in haskel 02:33:52 i discussed the problem aside from golfing and people suggested a trie 02:33:55 i like that idea 02:34:08 but it may be incredibly long to implement 02:37:31 right. tries or suffix arrays should give good performance 02:40:11 is there something like length$filter...? i am using sum[1| right now 02:40:49 that's good 02:41:39 (I recall discovering that trick; it felt good.) 02:42:50 sum[1|x] is also shorter than fromEnum x (conversions from tests to 0 or 1 is one of the most annoying problems in Haskell golfing, I've found; I've spent a lot of time on trying to rewrite code so that this won't be necessary) 02:43:18 -!- SopaXorzTaker has joined. 02:43:19 -!- SopaXorzTaker has quit (Changing host). 02:43:19 -!- SopaXorzTaker has joined. 02:45:01 > ([1|False],[1|True]) 02:45:03 ([],[1]) 02:45:40 > [()|] 02:45:42 :1:5: parse error on input ‘]’ 02:46:13 Jafet: I would use that in production code without blushing (except with spaces, [1 | test]) 02:47:04 Booleans in production code? Shocking 02:47:28 I'm surprised that desugaring comprehensions gives that result 02:47:37 Jafet: I would not use the sum[1 | ...] version though; that should clearly be fromEnum ... 02:47:57 why? 02:48:03 > guard (1==1) :: [()] 02:48:05 [()] 02:48:10 > guard (1==2) :: [()] 02:48:11 too bad this isn't vba, where adding 1 to z gives you aa 02:48:11 [] 02:48:20 i could brute force everything 02:49:30 Jafet: Hmm. I wonder how the Haskell report defines this, whether it's in terms of guard or in terms of filter ... but the former makes more sense to me. 02:50:37 It uses a recursive definition with [x|True] = [x] as the base case 02:52:38 adding 1. to z. gives you aa. AAAAAAAAAAAAAAAAAAAAH! 02:52:45 https://www.haskell.org/onlinereport/haskell2010/haskellch3.html#x8-420003.11 --- oh it's spelled out, but it's essentially the 'guard' version. The trick in translating [x|p] is to use [x|p] = [x|p,True] = if p then [x|True] else [] = if p then [x] else []. 02:53:06 AAAAAAAAAAAAAAAAAAAAG + 1 02:53:25 Jafet: so adding an artificial ,True to avoid the case that Q could be empty in the translation of the 'p' guard. 02:54:17 fun 02:54:35 Hey, we're having *fun* without fungot! 02:54:58 We haven't got fungot fun. 02:55:23 while we are at the haskell topic: would somebody recommend what of all these arrays i need? i want to have a text with O(1) access with lines that don't have the same length 02:55:25 fizzie`: WHARGHARBL! IEUAAAAH! FLBLBLBLBLBL! [INARTICULATE SUPRASEGMENTAL YELLING]! 02:55:34 so it should be an array of arrays of chars 02:55:38 but... what array? 02:55:46 * boily disappears to watch the TV specials... 02:55:56 -!- boily has quit (Quit: APHONIC CHICKEN). 02:56:22 :t array 02:56:23 Ix i => (i, i) -> [(i, e)] -> Array i e 02:56:40 aren't there like 6 different arrays? 02:57:13 yes, there's a matrix of them: unboxed/boxed and pure, for IO, and for ST 02:57:43 ST being what? 02:58:04 Since the lines will be some boxed type (ByteString or Text?), I think you'll be using a pure, boxed array, that's Array 02:58:33 myname: software transactional memory 02:58:33 Control.Monad.ST - the "state thread" monad which allows pure computations that work with (think: temporary) arrays. 02:58:43 oh wait, ST, not STM 02:58:44 my bad 02:58:52 well, since i want to access each character i most likely want something like Array Int (Array Int Char) 02:59:20 okay 02:59:46 should still be linear time to build it up 02:59:51 that's fine 02:59:54 myname: or Array Int (UArray Int Char) -- but seriously I'd store the lines as either ByteStrings or as Text. 03:00:12 int-e: i need O(1) access to every single char 03:00:18 (UArray is the unboxed, pure array; it's much more compact than the boxed one) 03:00:26 ByteString will give you that 03:00:31 compact being what? 03:00:56 oh, i should read about bytestring then 03:01:17 but it makes sense 03:01:35 i guess ByteString is similar to what C thinks is a string 03:02:31 Array Int Char will store a pointer per character, which points to a Char value. UArray Int Char will be a flat array of - hmm - wide characters? 03:03:03 so UArray Int Char should basically a ByteString? 03:03:12 but a bit uglier 03:03:46 An array of Word32s 03:04:32 it is surprisingly simple 03:05:05 i think i overestimated them 03:07:03 myname: nah, the UArray Int Char stores (32 bit, I checked) characters, ideally unicode code points; ByteString stores bytes, so it's closer to the C's char*. 03:07:47 okay 03:08:25 there are no ready to use function to make [[Char]] into Array Int ByteString i guess 03:08:43 should be pretty easy, though 03:10:23 Actually, since Char is boxed I think it is 16 bytes on amd64 03:10:37 -!- Phantom_Hoover has quit (Remote host closed the connection). 03:11:56 (And String uses 40 bytes per character.) 03:13:00 what 03:13:44 (The haskellwiki mentions a Char pool; presumably ASCII Chars only use 8 bytes on average.) 03:14:33 Idris functions are anti-parametric unless exported as abstract 03:16:44 how come i need (i, i) to build an array? 03:17:18 isn't it usually directly dependent of the list? 03:17:48 Scanning the list to find its size is a bad idea. 03:17:56 okay 03:18:50 (Your computer may not have enough memory to force a list spine, even if its contents would fit into a UArray.) 03:26:34 Jafet: the Char pool (is that a misspelled car pool?) covers latin1, hmm. (it goes up to 255.) 03:27:07 or simply, chars-as-bytes. 03:37:44 ah I may have reproduced oerjan's 90 character version. 03:39:01 -!- callforjudgement has joined. 03:39:54 [wiki] [[Special:Log/move]] move * Notjohnconway * moved [[CA-1]] to [[CA (programming language)]]: set of languages 03:40:26 is it interesting? 03:40:37 -!- ais523 has quit (Ping timeout: 240 seconds). 03:42:01 myname: If I'm right then this snippet offers food for thought: map take[1..] 03:42:28 oh 03:42:40 that-s neat 03:44:14 :t take 03:44:15 Int -> [a] -> [a] 04:01:41 (that's good for 88 characters in my attempt) 04:07:53 -!- GeekDude has joined. 04:07:54 95 now 04:11:29 cheers 04:20:12 -!- Solace has quit (Quit: Connection closed for inactivity). 04:29:58 93 04:46:59 -!- nys has quit (Quit: quit). 05:00:54 Happy new years! 05:01:28 happy new year! 05:02:07 And it's still the year of the horse. 05:04:16 Happy New Year, anyone in my timezone! 05:07:03 Yay, 1/1460 of the year is already past. 05:13:43 int-e: lol 05:30:22 -!- coppro has quit (Ping timeout: 240 seconds). 05:35:42 -!- GeekDude has quit (Quit: {{{}}{{{}}{{}}}{{}}} (www.adiirc.com)). 05:36:39 -!- GeekDude has joined. 05:42:22 -!- coppro has joined. 05:49:09 -!- oerjan has joined. 05:50:00 * oerjan changes a 4 to a 5 in the log url 05:50:12 happy new year! 05:53:24 Do the two kinds of vararg systems have names? The system where the caller gives a number of arguments that the function has to accept (e.g. most Lisps) vs where the function decides, based on arguments or other things how many arguments to take (e.g. C and Idris)? 06:03:09 -!- zzo38 has joined. 06:04:05 -!- dts|pokeball has joined. 06:05:15 Why doesn't DOSBox have a "UNMOUNT" command? 06:05:56 this is stupid. i am trying to find the english original for http://heltnormalt.no/img/tommytigern/2014/12/31.jpg but neither tin-eye nor google manage it. you'd think it was a simple thing to achieve... 06:06:33 (tin-eye only give the same page, while google only gives everything else on the same page.) 06:06:53 (which are of course, entirely different comics.) 06:09:54 -!- MoALTz__ has joined. 06:12:43 -!- MoALTz_ has quit (Ping timeout: 255 seconds). 06:14:19 -!- GeekDude has quit (Quit: {{{}}{{{}}{{}}}{{}}} (www.adiirc.com)). 06:14:28 zzo38: why not? it also has a mount command... 06:14:44 zzo38: sorry 06:15:01 zzo38: missed the negation. I guess the answer is that you can just restart dosbox. 06:15:45 text searches like "calvin and hobbes playing cards" don't find it either :( 06:17:58 applying extra quotes doesn't help 06:19:04 http://www.gocomics.com/calvinandhobbes/1994/01/01 - that was easy. 06:19:28 oerjan: note the year (between first and second panel) and the date (next to the signature in the last panel) 06:19:35 yay! but HOW? 06:19:46 * oerjan already tried searching on that site... 06:19:51 "read between the lines" 06:20:13 ah. 06:23:27 but yeah, a calvin&hobbes text search would be nice. too bad that (probably illegal) transmogrifier site is (long) dead. 06:36:06 int-e: I don't know what that transmogrifier site is, but I once made a local copy of a site that did C&H text search. 06:37:54 http://www.transmogrifier.org/no_more_ch.html?ch/comics/list.cgi <-- this one 06:38:11 It has a .js file with a transcription of every strip, I think. 06:38:21 Nope, different site. 06:40:03 whoa, I forgot I had this. 06:43:15 oerjan: should have tole me yesterday 06:43:24 i gave my father the complete c&h collection for christmas 06:44:54 I wonder how copyright should apply to an index. 06:45:46 keywords may be fine; transcripts are clearly subject to copyright law. 06:46:00 I bet Google knows about that. 06:50:22 hmm, transcripts may be a transformative use (under US law)... interesting question. 06:50:51 Note: I'm speculating. 07:10:06 argh. who but GG makes paper doll cliffhangers 07:10:32 Yeah, I was going to complain about that too. 07:11:13 "ne'er do wells who write this electro-rag" - indeed. 07:11:38 err. I can't copy correctly. 07:11:41 "ne'er do well hacks who write this electro-rag" - indeed. 07:12:11 oh end of a volume? so it's actually the entirely volume that ends on a cliffhanger. 07:12:24 *-ly 07:12:44 grammar, where dideth you went 07:13:12 I hate being 30 seconds off-phase wrt. to oerjan. 07:13:27 fancy 07:13:33 (well, at least if I'm 30 seconds behind) 07:13:49 being 30 seconds ahead feels kinda nice. 07:13:50 technically i cheated by not reading linearly 07:20:08 meh, it's a new year and I still haven't figured out how henkma did 'identity matrix' (and some other open-ended problems where I haven't submitted solutions yet because they'd score less than 10k) 07:21:09 (otoh, henkma hasn't caught up on http://golf.shinh.org/p.rb?kM4_ ) 07:22:04 (almost everybody is totally cheating on that one.) 07:23:20 that seems like one it would be silly _not_ to cheat on 07:25:43 int-e: I know you can just restart DOSBox, but I hope to have UNMOUNT command so that you can reassign a drive letter without having to restart DOSBox!! 07:27:50 zzo38: actually, mount -u does just what you want 07:28:51 (dosbox 0.74) 07:29:39 O, OK, but it doesn't seems to have that in the built-in documentation 07:30:15 same here. dosbox(1) documents it though. 07:33:35 Is it possible for the mount command to specify whether or not programs on a mounted drive are allowed to access special DOSBox functions such as mounting a drive? 07:36:14 config -securemode may be of interest, but that's a one-way street. 07:37:34 I don't see how dosbox would distinguish between programs loaded from one drive or another. 07:38:09 It displays the name of the program at the top, so probably it can know. 07:42:44 int-e: it seems to me like embedding the output won't get anywhere near 39 or 40, you need something like main=do n<-readLn;print$...random cleverly searched for math formula here... 07:43:01 I also believe DOSBox ought to include its own versions of DEBUG and MODE commands. Other than that, the existing commands are probably fine. 07:43:38 and henkma just hasn't searched as cleverly as you 07:44:56 oerjan: yeah I think that's a fair summary 07:45:24 oerjan: I suspect that henkma does everything by hand. 07:48:29 oerjan: oh http://golf.shinh.org/p.rb?Dancing+Kids is another annoying one. 07:48:50 (I have a 101 character solution...) 07:49:14 actually, two variations of that. 07:51:33 -!- rade has quit (Quit: Leaving). 07:52:44 -!- shikhin has joined. 08:01:05 never mind. 08:01:56 -!- DrMcHood has joined. 08:03:47 Have studied Ebonics for many decades, have dived into programming the past 2 years. Any thoughts on a convergence? 08:03:52 (Now it's just a can't-believe-I-missed-that-the-last-time-I-worked-on-this nuisance.) 08:04:57 DrMcHood: Sorry, I don't know. (That doesn't imply everyone else also doesn't know.) 08:05:20 -!- zzo38 has quit (Remote host closed the connection). 08:08:55 keep it(100):[tree.fitty.4°[ice cold*2^chains]] 08:14:09 -!- DrMcHood has quit (Quit: Page closed). 08:16:32 how do I set an environment variable for both X and login sessions? I thought that was what .profile was for but it does not seem to be working 08:19:33 streaming at http://twitch.tv/coppro 08:31:40 `2014 this is going to become the new o 08:31:40 No output. 08:34:03 coppro: I think that has become impossible to answer; it depends on your X11 login and session managers. 08:34:28 :( 08:34:48 `cat bin/2014 08:34:49 #!/bin/sh \ if [ $(date +%Y) = "2014" ] \ then echo "Hello, world!" \ fi 08:34:52 hahaha 08:39:32 coppro: some possible suggestions can be found here ... https://superuser.com/questions/19044/how-do-i-set-an-environmental-variable-for-a-gnome-session 09:11:36 * oerjan notes that the last `2014 hello world appeared on the january 1 log page. 09:12:12 `2014 09:12:12 No output. 09:12:14 -!- adu has quit (Quit: adu). 09:14:33 int-e: thanks 09:40:27 -!- fizzie` has changed nick to fizzie. 09:41:58 -!- fungot has joined. 09:43:13

and I don't remember how to omit the final newline in vim (it was discussed here before, it's possible somehow) <-- something like :set binary noeol 09:43:32 @time fizzie 09:43:33 Local time for fizzie is Thu Jan 1 11:43:32 2015 09:48:03

(also my algorithm is completely different, but that's all I intend to say about it) <-- I COULD GUESS THAT 09:48:25 really! 09:48:34 you're SO smart 09:48:44 YEP 09:51:32 i suspect myname has converged towards the algorithm i'm using, but is missing a trick or too 09:51:39 but that doesn't help us beat you 09:51:43 *two 09:51:57 did I guess that snippet correctly? 09:52:07 wat 09:52:17 i haven't got to that in the log if it's there 09:52:30 "snippet" should be in the log? 09:53:26 i have no map take 09:53:47 interesting :) 09:54:55 oh well. 2 down, 3 to go ... identity matrix, I'll get you yet... 09:56:02 GOOD, GOOD 09:56:34