00:00:00 what's the definition of "helloworldless"? no characters from "helloworld"? 00:00:04 int-e: please, like telling you it's definitely completely possiblew ouldn't have motivated you even more 00:00:08 *impossible wouldn't 00:00:09 ais523\unfoog: right 00:00:15 that's how I got to the solution in the first place, after all 00:00:21 ais523\unfoog: "Hello, world!" actually 00:00:36 elliott: thus no spaces either, which I assume is one of the harder parts 00:00:37 -!- shikhin has quit (Ping timeout: 244 seconds). 00:00:41 although this is Haskell, you have $ to substitute 00:00:42 nah 00:00:49 not really hard, you can use tabs usually 00:00:53 or newlines 00:00:55 or whatever whitespace 00:00:57 depending on the language 00:01:29 oh right 00:01:30 my first solution didn't have tabs. it extracted a space from the SUDO_COMMAND environment variable... 00:01:55 I felt silly after seeing other solutions that used tabs 00:02:41 ais523\unfoog: it turns out that everything is easy to do in haskell _except_ finding an actual output function not containing those chars 00:02:56 there just isn't one 00:06:21 except for that, it has both numerical string escapes and map(succ) which would work 00:07:14 > succ 'd' 00:07:15 'e' 00:07:49 but succ.succ or pred.pred work. 00:07:58 err 00:08:00 no, you cannot use pred 00:08:10 succ.succ works 00:08:43 > map((-1+).(-1+))"Jgnnq.\"yqtnf#" 00:08:45 No instance for (GHC.Num.Num GHC.Types.Char) 00:08:45 arising from a use of ‘GHC.Num.+’ 00:08:48 ah 00:08:51 sniff. 00:08:55 heh 00:09:43 > [pred,pred.pred] <*> ["Hello, world!"] 00:09:45 No instance for (GHC.Enum.Enum [GHC.Types.Char]) 00:09:45 arising from a use of ‘GHC.Enum.pred’ 00:09:50 oops 00:09:54 > [pred,pred.pred] <*> "Hello, world!" 00:09:56 "Gdkkn+\USvnqkc Fcjjm*\RSumpjb\US" 00:10:39 "\R" 00:10:45 > "\R" 00:10:46 :1:3: 00:10:47 lexical error in string/character literal at character 'R' 00:10:50 -!- Bicyclidine has joined. 00:10:52 wat 00:10:57 > "\RS" 00:10:58 "\RS" 00:11:02 > succ "\RS" 00:11:04 oh 00:11:04 No instance for (GHC.Enum.Enum [GHC.Types.Char]) 00:11:04 arising from a use of ‘GHC.Enum.succ’ 00:11:09 > succ '\RS' 00:11:10 '\US' 00:11:28 > succ '\US' 00:11:30 ' ' 00:11:41 oh well that's no shorter than \31 and \30 00:11:57 which are presumably easier to remember 00:11:58 space, the final frontier, right next after the United States... 00:12:05 ooh 00:12:07 but it's so nicely obscure. 00:12:28 > ("\US0123","\310123") 00:12:30 ("\US0123","\310123") 00:12:53 > length"\310123" 00:12:55 1 00:13:52 > '\310':"123" 00:13:54 "\310\&123" 00:14:07 > map length["map(succ)\"G\\99kkn+\\USvnqkc\"","map(succ.succ)\"Fcjjm*\\RSumpjb\\US\""] 00:14:09 [27,33] 00:14:26 looks like two succs isn't worth it 00:14:38 fair enough 00:15:20 succ less, quote more 00:15:46 oh hm wait 00:16:00 i forgot the final ' ' of the first 00:16:15 > '\SP' 00:16:15 > map length["map(succ)\"G\\99kkn+\\USvnqkc\32\"","map(succ.succ)\"Fcjjm*\\RSumpjb\\US\""] 00:16:17 ' ' 00:16:17 can't find file: L.hs 00:16:24 > map length["map(succ)\"G\\99kkn+\\USvnqkc\32\"","map(succ.succ)\"Fcjjm*\\RSumpjb\\US\""] 00:16:25 [28,33] 00:16:34 oh wait 00:16:41 > map length["map(succ)\"G\\99kkn+\\USvnqkc\\32\"","map(succ.succ)\"Fcjjm*\\RSumpjb\\US\""] 00:16:42 [30,33] 00:16:49 still not worth it 00:18:07 > 111111*123 00:18:09 13666653 00:20:07 what's that about 00:20:46 -!- vanila has joined. 00:21:54 nothing useful. 00:22:34 -!- bananagram has joined. 00:23:17 `relcome bananagram 00:23:18 bananagram: 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.) 00:23:57 hi 00:26:53 `welcome vanila 00:26:54 vanila: 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.) 00:27:43 int-e: haha, I like that SUDO_COMMAND thing 00:34:01 Hello 00:35:10 hi! 00:35:29 http://lpaste.net/raw/113500 00:35:34 I wrote this 00:47:23 -!- MDude has joined. 00:49:53 how do you iterate over dictionary keys in JavaScript? 00:50:12 `fetch http://lpaste.net/raw/113500 00:50:14 2014-10-31 00:50:07 URL:http://lpaste.net/raw/113500 [2316] -> "113500" [1] 00:50:36 so it converts brainfuck to C, neat. 00:50:59 oh it needs input? 00:51:13 yes. 00:51:38 ais523\unfoog: Object.keys? 00:51:38 `find esobf8 00:51:39 find: `esobf8': No such file or directory 00:51:46 lifthrasiir: that isn't a loop 00:51:46 *sigh* 00:51:52 ais523\unfoog: for in and checking hasOwnProperty? Or do you mean an actual Map? 00:52:02 Sgeo: it's the syntax for "for in" I'm failing at 00:52:13 https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/for...in 00:52:17 for (variable in object) {... } 00:52:22 for (var key in {a:1, b:2}) console.log(key); // a, b 00:52:25 `url ibin/bf8 00:52:26 http://codu.org/projects/hackbot/fshg/index.cgi/file/tip/ibin/bf8 00:52:31 but as Sgeo said, hasOwnProperty is meh 00:52:56 lifthrasiir: I get "missing ; before statement" 00:53:03 `run echo ",[.,]" | interps/egobf/src/egobfi8 113500 00:53:03 #include \ #include \ uint8_t a[1024]={0};uint8_t*p=a;main(){setbuf(stdout,NULL);*p=getchar();while(*p){putchar(*p);*p=getchar();}} 00:53:21 ais523\unfoog: uh, really? 00:53:34 ais523\unfoog: pastie sample? 00:53:35 what JS impl do you use? 00:53:42 it can self host! 00:53:44 I've tried it in Chrome 00:53:56 for (var y in x) print y; 00:53:58 SyntaxError: missing ; before statement 00:54:01 lifthrasiir: I'm using the Firefox console 00:54:11 Had to delete the b in the comment, because Chrome was trying to autocomplete it when I pressed enter 00:54:15 its cool how the bot downloaded and ran it 00:54:16 print is not a statement 00:54:25 Use console.log() 00:54:27 ais523\unfoog: console.log(y); should work 00:54:49 Except in IE where you should check for the existence of console first 00:54:51 Because IE 00:54:56 vanila: HackEgo runs nearly arbitrary linux commands 00:55:10 lifthrasiir: err, right, I'm being silly 00:55:11 thanks 00:55:25 oerjan: well, it runs arbitrary commands, just those commands don't always do what you'd expect 00:55:30 `` uname -a 00:55:31 Linux umlbox 3.13.0-umlbox #1 Wed Jan 29 12:56:45 UTC 2014 x86_64 GNU/Linux 00:56:58 It would be nice to write a brainfuck to assembly compiler, but it's harder because you have to match brackets 00:57:42 the hard part is the forward jumps when a [ condition doesn't match 00:58:27 If I ever get around to writing a Braintrust impl. in Clozure CL like I intend at some point, that serves for allowing brainfuck compilers to be written in brainfuck that compile to machine code 00:58:37 Without any knowledge of actual machine code 00:58:47 Although I wonder if using a lisp like that is 'cheating' 00:59:22 Braintrust? 00:59:30 ^wiki Braintrust 00:59:31 http://esolangs.org/wiki/Braintrust 01:00:19 one of the more interesting brainfuck derivatives 01:00:35 :) 01:01:19 -!- Phantom_Hoover has quit (Ping timeout: 255 seconds). 01:01:35 Phantom_Hoover couldn't stand for that, naturally 01:01:41 would that turn a self interpreter into a compiler? 01:01:58 I don't really understand the (!) thing 01:02:44 ! takes the stored up commands and compiles them. So, if the target is Haskell, and ! gets executed, Haskell code will get output, corresponding to the program in the code block 01:03:16 The crucial thing is that the language that ! compiles in the code block is the -current- language, not necessarily Braintrust 01:04:27 so that puts 99% of the effort onto the implementation? 01:04:31 Yes 01:04:42 Including any knowledge of what the target is 01:04:55 but how would you go from a brainfuck compiler in braintrust, to a brainfuck compiler in brainfuck? Or is that not possible 01:05:16 So, the same Braintrust program can be a Brainfuck derivative 346 to machine code compiler, and a Brainfuck derivative 346 to Haskell compiler 01:06:44 Vanilla brainfuck doesn't have instructions for 'compile this code' 01:07:38 so you need to implement braintrust interpreter in brainfuck - then you can compile the self applicable braintrust compiler to get a brainfuck compiler in brainfuck? 01:07:43 once you start stacking implementations, which one of those is the regarded as the "current" one? 01:08:37 The one used to compile the program that is currently executing 01:08:47 The Haskell impl stores a stack of each implementation 01:12:32 that's really cool though 01:13:02 So, suppose the compiler stack is BT, L0, L1, L2. A program P is in the code block. ! is basically BT(L0(L1(L2(P)))) 01:13:08 Where BT is the primitive compiler 01:13:14 vanila: thank you 01:14:53 why can't ! pick any old language/ 01:14:54 * 01:14:55 *? 01:14:58 It's all inspired by Reflections on Trusting Trust 01:15:00 rather than being restricted to the implementation language 01:15:23 elliott: hmm... a second code block for the language to use? 01:15:28 what I mean is: if I write a haskell braintrust interpreter, and compile it to C, is it no longer a braintrust interperter? 01:15:35 because ! should be outputting C, but it's outputting haskell 01:16:00 possibly I don't understand braintrust at all. 01:16:29 -!- hjulle has quit (Ping timeout: 265 seconds). 01:17:00 elliott: if... I understand properly, this is just a definitions issue. I think. But the primitive compiler doesn't need to be written in the target language, I think. 01:17:30 (If it does, then that's a spec bug) 01:19:56 What I would think interesting is somehow decoupling ! in such a way that ! requires its input to be primitives, but somehow extra commands that make it simple to do the quineiness needed for something that behaves as this ! 01:40:28 https://eval.in/108854 i love these 01:55:54 -!- boily has quit (Quit: SQUASH CHICKEN). 02:17:19 -!- shikhout has quit (Ping timeout: 245 seconds). 02:22:51 > logBase 2 (10**9 * 26**7) 02:22:53 62.80043088097391 02:23:28 -!- bananagram has left ("WeeChat 0.4.3"). 02:23:56 that should fit nicely 02:24:13 with a bit to spare 02:25:54 i just invented an esolang 02:26:16 its called edwardk-oriented-programming 02:27:45 > let f(m,n)=logBase (10^m * 26^n) in maximumBy f [(m,n)| m<-[0..64],n<-[0..64], f(m,n)<=64] 02:27:47 Couldn't match type ‘(b, b1)’ with ‘GHC.Types.Ordering’ 02:27:47 Expected type: (b, b1) -> (b, b1) -> GHC.Types.Ordering 02:27:47 Actual type: (b, b1) -> (b, b1) -> (b, b1) 02:28:02 > let f(m,n)=logBase (10^m * 26^n) in maximumBy (comparing f) [(m,n)| m<-[0..64],n<-[0..64], f(m,n)<=64] 02:28:03 Could not deduce (GHC.Classes.Ord (a0 -> a0)) 02:28:04 arising from a use of ‘Data.Ord.comparing’ 02:28:04 from the context (GHC.Real.Integral b1, GHC.Real.Integral b) 02:28:04 bound by the inferred type of 02:28:04 it :: (GHC.Real.Integral b1, GHC.Real.Integral b) => (b, b1) 02:28:12 oops 02:28:39 > let f(m,n)=logBase 2 (10^m * 26^n) in maximumBy (comparing f) [(m,n)| m<-[0..64],n<-[0..64], f(m,n)<=64] 02:28:41 (15,3) 02:29:08 > logBase (10^15*26^3) 02:29:10 Double> 02:29:14 > logBase 2 (10^15*26^3) 02:29:16 63.93024057773372 02:30:22 vanila: http://cokmett.github.io/cokmett/ 02:30:52 XD 02:31:27 Ugh. 02:31:28 -!- shachaf has left. 02:33:16 shachaf got a haskell overload 02:34:05 vanila: have you seen Taneb's proof that lens functions are turing complete 02:34:09 :O 02:34:26 i havent1 02:34:33 actually i'm not sure _i_ have, did he finish it 02:34:54 hm looks idle 02:36:01 also i am still wondering if that cokmett thing is statistically guaranteed to stop, i couldn't find the source code from which the javascript was generated 02:38:23 hm the google hits give logs for this channel 02:39:23 -!- shachaf has joined. 02:41:32 http://codu.org/logs/log/_esoteric/2014-02-17#092119Taneb 02:41:49 -!- Bicyclidine has quit (Ping timeout: 258 seconds). 02:41:59 it seems google _is_ currently crawling the channel logs, both tunes and codu 02:42:06 that didn't use to be the case 02:42:34 * Sgeo always googles the logs 02:44:36 i guess this means i'll have to stop complaining they're no longer in HackEgo 02:44:52 well, or scale it down somewhat. 02:45:53 -!- shachaf has left. 02:56:58 > [n|a<-[0..4],b<-[0..9],c<-[0..9],n<-[a*100+b*10+c],n==0||n`mod`(a+b+c)`elem`[0,a,a+b]] 02:57:00 [0,1,2,3,4,5,6,7,8,9,10,11,12,13,17,18,20,21,22,24,26,27,30,31,33,35,36,39,4... 02:57:39 far too long, of course, but faster than henkma's 02:57:55 hm i thought there would be such a formula 02:58:30 I have a fairly neat show- and read-free version, but it's 76 characters long. 02:58:45 the speed seems about right though 03:02:23 -!- nys has quit (Quit: quit). 03:02:47 btw it's better to use the mod prefix there 03:05:07 fromEnum is so stupidly long :( 03:05:09 and it's worthwhile to factor out the [0..9] 03:05:35 but might still be better than that 03:06:36 I wouldn't paste a serious contender, not intentionally anyway 03:06:42 heh 03:07:46 @type fix.shows 03:07:48 Show a => a -> String 03:08:07 you don't have fix, mind you 03:08:10 I wonder why fix is not in the Prelude. 03:08:14 I know! 03:08:23 import Data.Function ... 03:08:27 go wild ;-) 03:09:27 i haven't yet seen any case where importing was best when it was _possible_ without 03:09:50 although i had a single char difference once 03:10:15 > let f a | c <- b = c where b = a in f "arg" 03:10:17 "arg" 03:10:39 oerjan: Data.List.sort, when required, is worth it. 03:10:46 i suppose. 03:30:31 meh perhaps I should submit head -c790 /dev/random and see if I get lucky... 03:31:36 -!- MDude has changed nick to MDream. 03:32:02 :t insert 03:32:03 Ambiguous occurrence ‘insert’ 03:32:03 It could refer to either ‘Data.List.insert’, 03:32:04 imported from ‘Data.List’ at /home/lambda/.lambdabot/State/L.hs:91:1-16 03:32:09 :t Prelude.insert 03:32:10 Not in scope: ‘Prelude.insert’ 03:32:11 Perhaps you meant one of these: 03:32:11 ‘Prelude.sqrt’ (imported from Prelude), 03:32:13 huh. 03:32:52 > length "{-#OPTION -F -pgmFsh #-}\nhead /dev/random>$2" 03:32:54 Ambiguous occurrence ‘length’ 03:32:54 It could refer to either ‘GHC.List.length’, 03:32:54 imported from ‘Data.List’ at L.hs:91:1-16 03:32:54 (and originally defined in ‘base:GHC.List’) 03:32:54 or ‘Data.List.NonEmpty.length’, 03:32:59 @undef 03:33:00 Undefined. 03:33:01 > length "{-#OPTION -F -pgmFsh #-}\nhead /dev/random>$2" 03:33:03 44 03:33:23 > length "{-#OPTION -F -pgmFsh#-}\nhead /dev/random>$2" 03:33:24 43 03:33:28 int-e: coming soon to anagol: control the entropy pool 03:33:39 how much information can you stuff into $$ 03:33:47 maybe you can shave off a few bytes with that 03:33:49 15ish bits? 03:33:53 I guess -- yeah 03:33:56 probably not enough 03:36:45 > length "{-#OPTION -F -pgmFsh#-}\nhead /dev/r*om>$2" -- /dev/ra*m works, too 03:36:46 41 03:36:58 > length "{-#OPTION -F -pgmFsh#-}\nhead /*/r*om>$2" -- /*/ra*m works, too 03:36:59 39 03:40:03 int-e: does OPTION without the S work? 03:40:39 no. 03:41:00 int-e: I bet you could inject some state into the system by, like, contacting some daemon or using some setuid program that keeps some kind of cache. 03:41:06 and then fish a cheat out of there. 03:53:13 neat. you called forbidden system call (socketcall). 03:54:07 ok, enough curiosity 03:59:53 oh duh, it's up to 500, not including 04:00:08 not that it makes a difference for lenght 04:00:09 *th 04:02:00 so nicely fast, but 84 chars 04:04:47 elliott: you can store some information in the kernel message buffer if you like. that's the only hole I've found so far. there seem to be no writable directories, no network access, and forked processes seem to get cleaned up after each job is done. 04:05:14 int-e: sysv ipc was a hole that used to work 04:05:29 ipc to where? 04:05:36 how can you get arbitrary stuff into the kernel message buffer? 04:05:48 http://john.freml.in/codegolf-cheating 04:07:13 although this isn't short i just love my use of scanr main=mapM print$0:[n|n<-[1..499],t:s<-[scanr(+)0[read[c]|c<-show n]],mod(-n)t`elem`s] 04:07:43 it's not a function frequently used 04:08:35 what are you implementing 04:08:48 vanila: http://golf.shinh.org/p.rb?Belgian+Numbers 04:11:23 oerjan: oh, I have an extra pair of () in my version of that: main=mapM_ print$0:[n|n<-[1..499],(x:r)<-[scanr(+)0[read[c]|c<-show n]],mod(-n)x`elem`r] 04:11:46 see topic 04:12:19 also an extra _, i see 04:23:46 right. the last thing I did with it was count the number of bytes of the expected output. runhaskell prints the result of mapM, too. 04:24:13 ah 04:26:55 problem 8 here is pretty good to golf: https://www.student.cs.uwaterloo.ca/~cs241/a6/ 04:27:08 my record is 53 characters of perl, but I think that's with a trailing \n 04:27:33 elliott: I was hoping for core dumps, but the best way is probably to use the "socket" limitation and try to connect to a couple of IP addresses; those can then be found in the kernel log. 04:28:02 XD 04:28:09 int-e: christ 04:28:27 @tell Gregor Have you challenged your students go golf assignment 6 problem 8 yet? 04:28:28 Consider it noted. 04:36:49 elliott: /proc/sys/kernel/msgmax reads 0, so I guess those message queues have been plugged. 04:54:13 are there any non-hg-based persistence mechanisms in HackEgo? 05:00:34 To make "Helloworldless Hello world" with SQL, is this it? SELECT+X'48656C6C6F2C20576F726C6421'; Is there shorter way? 05:03:55 it's unlikely that there'd be a shorter way 05:04:21 I believe you 05:04:22 is the hellowordlessness requirement case-sensitive? 05:04:33 if it is, you could do it in INTERCAL pretty easily 05:04:41 if it isn't, you can't, because you can't write any valid statement identifier 05:04:48 The check is case-sensitive 05:05:15 (Look at examples to see that) 05:05:32 Therefore, "SELECT" should be acceptable. But, commas and spaces are prohibited. 05:13:54 I am sure it is completely impossible in Unlambda, though. 05:20:37 elliott: I found a better hole: http://golf.shinh.org/reveal.rb?Helloworldless+Hello+world/int-e+%28cheat+horribly%29_1414732809&sh 05:21:14 int-e: is that shell? 05:21:19 yes. 05:21:48 unquoted: crontab -l | cut -c 2- 05:22:09 but o, r, l and space are not allowed. 05:22:17 you stored information in the crontab? 05:22:22 ais523\unfoog: yes! 05:22:27 ahahahaha 05:22:44 that seems rather worse-sandboxed than HackEgo 05:23:04 now I'm wondering if shinh.org is busy hello-worlding every minute 05:23:30 I made it a comment. 05:24:22 ("Hello, world!" has the wrong format for a crontab anyway) 05:24:42 can you abuse the crontab to get a zero-length entry?> 05:25:06 Perhaps you can report it, so that they can fix that bug 05:25:17 unlikely, mostly because the program files are not writable by the golf user 05:25:27 I doubt you can get a zero-length entry 05:25:46 it seems like crontab abuse would allow for a universal submission, in a less hilariously-unlikely-to-work way than outputting a random number of random charactres 05:25:48 *characters 05:32:55 ais523\unfoog: and yes, it's a worse sandbox; it looks like just sudo, perhaps in connection with a chroot. 05:33:18 (make that probably) 05:33:25 I'd been informed that anagolf's sandbox was really incredibly complex 05:33:58 just spin up a new VM for each entry 05:35:06 ais523\unfoog: the complexity seems to be in the syscall auditing, http://shinhoge.blogspot.jp/2014/10/sandbox-of-my-golf-server.html 05:36:29 I like the bit where shinh redefined what getpriority and setpriority do to allow communication with the kernel, rather than adding a procfile 05:36:57 it...works! 05:39:26 I was looking for some unused pipe special file that the golf user can write to, but I didn't find any. 05:42:38 *counts negations* "I don't think my server is not uncheatable" 05:42:43 *chuckles* 05:46:53 `` crontab -l 05:46:53 crontab: your UID isn't in the passwd file. \ bailing out. 05:47:29 `` cat /etc/passwd 05:47:31 cat: /etc/passwd: No such file or directory 05:47:37 `` ls -l /etc 05:47:38 total 8 \ drwxr-xr-x 2 0 0 4096 Jan 29 2014 alternatives \ drwxr-xr-x 5 0 0 4096 Jan 29 2014 java-6-openjdk 05:47:55 root-owned and we can't write to it 05:48:03 `` ls /dev 05:48:03 agpgart \ audio \ audio1 \ audio2 \ audio3 \ audioctl \ console \ core \ dsp \ dsp1 \ dsp2 \ dsp3 \ fd \ full \ kmem \ loop0 \ loop1 \ loop2 \ loop3 \ loop4 \ loop5 \ loop6 \ loop7 \ mem \ midi0 \ midi00 \ midi01 \ midi02 \ midi03 \ midi1 \ midi2 \ midi3 \ mixer \ mixer1 \ mixer2 \ mixer3 \ mpu401data \ mpu401stat \ null \ port \ ptmx \ pts \ ram \ 05:48:08 `` ls /dev | wc 05:48:09 89 89 538 05:48:26 `` ls -lad /dev/shm/ 05:48:26 ls: cannot access /dev/shm/: No such file or directory 05:48:42 `` mount | grep tmpfs 05:48:43 tmpfs on /tmp type tmpfs (rw,relatime) 05:48:55 `` ls /tmp 05:48:56 No output. 05:49:04 `` touch /tmp/abc 05:49:04 No output. 05:49:06 `` ls /tmp 05:49:07 No output. 05:52:48 -!- b_jonas has quit (Ping timeout: 250 seconds). 05:56:21 -!- kakaouette has joined. 05:57:38 -!- kakaouette has quit (Remote host closed the connection). 06:05:23 -!- b_jonas has joined. 06:07:36 [wiki] [[Talk:FakeASM]] http://esolangs.org/w/index.php?diff=40732&oldid=40725 * Rdebath * (+829) 06:09:41 [wiki] [[Talk:FakeASM]] http://esolangs.org/w/index.php?diff=40733&oldid=40732 * Rdebath * (+65) /* Um Esoteric Language ? */ 06:21:54 -!- vanila has quit (Quit: Leaving). 06:28:10 [wiki] [[Talk:FakeASM]] http://esolangs.org/w/index.php?diff=40734&oldid=40733 * CosmoConsole * (+475) 06:35:34 https://www.reddit.com/search?q=aurora+borealis&sort=relevance&t=all 06:35:57 I ... still think a solar eclipse is higher up on my bucket list... but aurora is so much more achievable 06:36:11 So far my bucket list consists of two items anyway 06:37:12 I remember a partial solar eclipse visible in the UK 06:37:19 we watched it using improvised pinhole cameras 06:37:34 s//total/g 06:39:03 I stared directly at a partial solar eclipse when I was a kid 06:39:05 Maybe for one second 06:54:37 There was a total solar eclipse in Finland in 1990, I vaguely remember being up at some unlikely hour watching it at our summer place. 06:56:57 (It was in June, and this was... well, not quite northern Finland but a third of a way up there, so sunrise was at 3:44am; I think the eclipse was around 5am.) 06:57:01 Sorry, July. 07:01:28 fizzie: right, I was going to say "solar eclipses only happen in daytime by definition", but forgot how weird daytime can be at extreme latitutdes 07:02:16 I wouldn't go as far as to call it "extreme", it isn't even within the arctic circle. 07:02:52 well, it gets weirder as it gets extremer 07:03:01 Sunrise at 3:44am and sunset at 10:27pm, and proper (civil) dusk for a bit less than two hours, apparently. 07:03:04 I think if I see it once, I might be disinclined to try again 07:03:07 (For the day of that eclipse.) 07:03:19 It's why I'm not so enthusiastic about lunar eclipses, I think. Saw one as a kid 07:03:31 Or maybe those aren't as pretty 07:04:09 Next total solar eclipse visible from Finland is apparently from 2126 (and an annular one at 2039). 07:04:19 s/from 2/in 2/ 07:05:20 s/at 2/in 2/ probably too. 07:05:27 [wiki] [[Talk:FakeASM]] http://esolangs.org/w/index.php?diff=40735&oldid=40734 * CosmoConsole * (+497) 07:05:28 It's like my preposition engine is totally random. 07:07:56