00:03:08 -!- AndoDaan has joined. 00:05:33 -!- Bicyclidine has quit (Ping timeout: 260 seconds). 00:21:20 `run gcc -Wall hello.c 00:21:22 hello.c:1:12: warning: ‘main’ is usually a function [-Wmain] 00:38:29 -!- Phantom_Hoover has quit (Remote host closed the connection). 00:38:37 -!- LordCreepity has quit (Ping timeout: 245 seconds). 00:44:54 why does english not even have a singular of clothes 00:45:22 -!- GeekDude has quit (Quit: {{{}}{{{}}{{}}}{{}}} (www.adiirc.com)). 00:51:13 -!- LordCreepity has joined. 00:51:13 -!- LordCreepity has quit (Client Quit). 00:51:28 -!- LordCreepity has joined. 01:08:54 -!- mcfrish has joined. 01:09:29 -!- mahem1 has joined. 01:16:53 Hey all, speaking about befunge here. Is there any fast way to do advanced stack manipulations. For instance 'right rotating' a LIFO stack (the first value becomes the third, third, become second, second becomes first) 01:17:24 93? 98? 01:17:42 -!- mcfrish has left. 01:18:57 98. Although, not entirely sure of the difference between them. 01:19:48 in 93 you have almost no way to directly or indirectly access the stack element past second-to-the-top 01:21:34 Excluding sets and puts? Or does 93 not have those? 01:21:43 yeah, excluding sets and puts 01:22:09 ok, and in 98 what commands let you do that? 01:22:10 of course, you are obviously limited in 80*25 bytes then 01:22:41 afaik there is no direct command, but you can combine stack stack ops to get the similar effect, I think 01:25:06 I've tried to work out how to use the command seen here 9http://en.wikipedia.org/wiki/Befunge) paste the second, but I can't seem to . 01:25:11 Example? 01:26:12 -!- zzo38 has joined. 01:28:27 -!- idris-bot has quit (Ping timeout: 246 seconds). 01:30:24 mahem1: the Wikipedia article exclusively discusses Befunge-93 (which is far smaller language than Funge-98), you should look at the "official Funge-98 specification" in the external links. 01:30:45 of course, unless you want to constrain yourself to Befunge-93 01:35:32 lifthrasiir, Dang, buh, I've been stuck in a small little world... Thanks for setting me free from the small world of befunge93! 01:36:12 the constraint is often a source of creativity though ;) 01:39:26 True, it's quite fun and intriguing actually. But I feel like I am cheating when I just use gets and puts as variables with mostly an empty stack... 01:39:41 we all do it. 01:39:49 so much room for variables. 01:40:06 just lying there, under the ip's feet. 01:40:43 Hmmmm, no goto in Funge98 01:41:01 Yeah, we all do, but then it kills the challenge. 01:41:21 http://users.tkk.fi/~mniemenm/befunge/ccbi.html best '98 interpreter, has most of the fingerprints (modules) 01:41:44 is cfunge still in development? 01:41:57 It's fun trying to figure out when to duplicate the top value for use way later in your program. 01:43:25 I doubt it, lifthrasiir. but idk. 01:43:30 -!- yiyus has quit (Ping timeout: 260 seconds). 01:43:46 -!- yiyus has joined. 01:44:32 kinda lost my appetite for befunge when i started golfing. all that creativity doesn't help with competetiveness. 01:51:39 -!- drdanmaku has quit (Quit: Connection closed for inactivity). 02:03:50 I don't think CCBI is very in-development, either. 02:06:15 funge's best days are behind him i'm afraid. 02:14:25 -!- zzo38 has quit (Ping timeout: 258 seconds). 02:23:02 -!- shikhin has joined. 02:26:18 -!- shikhout has quit (Ping timeout: 272 seconds). 02:33:55 -!- Bicyclidine has joined. 02:52:54 Hmmm, how do I get this 3rd dimension working... 02:53:18 are you using the windows' binary? 02:53:47 in ccbi i mean. 03:00:18 I dunno I got my interpretter from here. https://github.com/catseye/Befunge-93/blob/master/src/bef.c 03:00:22 On, linux. 03:02:36 I don't think '98 is implemented in that one. 03:02:55 and the 3D stuff is '98. 03:04:01 is there any particular reason for a finite number of dimensions 03:04:09 -!- J_Arcane has quit (Remote host closed the connection). 03:04:43 idk hilbert space just seems.... tooo big. 03:05:13 -!- J_Arcane has joined. 03:05:24 i'm just saying, as long as you're adding a whole dimension, might as well go for broke. 03:06:52 yeah, but after 3... maybe 4, it's just useless trying to visualize. and for me, personally, i'm useless without visualization. 03:06:57 looks like funge 98 is generic anyway? you could have a unary "turn in this direction" operator in addition to the usual 0ary ones, so you wouldn't even need infinite opcodes 03:08:00 have you ever played nethack or something? they do three dee just by successive slices of 2D, and it works fine since the topology is discrete. shouldn't be too hard in n dimensions 03:09:28 hmm, i've been looking a nethack recently. and i couldn't make out if it was a topdown view, or side on. 03:09:34 topdown 03:09:37 it takes practice though yeah 03:10:00 Whoops that is what I am using now. But I meant in funge98. 03:10:17 basically i just mean each floor is a flat surface, and the dungeon as a whole is just a bunch of those taped together 03:10:18 Bicyclidine: it already has instructions to set delta 03:10:24 hell yes 03:10:26 so you can do arbitrary dimensions, the only thing you need to specify is file encoding really 03:10:39 very forward thinking 03:10:43 it'd need tweaks for infinite dimensions though. 03:10:50 Like what code is excuted in the 3rd dimension. How do you write it? . Or do I need to let it write it's own code in 3d? 03:10:50 how are three dee spaces encoded 03:11:04 h and l i think 03:11:07 ^L increases z and resets x and y, basically 03:11:12 ^L as in the control code 03:11:26 don't we have a befunge '98 bot here? 03:11:28 i mean how would you describe the 3d space of a program 03:11:32 fungot: help 03:11:32 Bicyclidine: we were at this arcade. do you remember when i gave you five hundred bucks, and told you to upgrade my computer for doom 3? 03:11:40 I'm not sure what you mean 03:11:44 I'm talking about the file format 03:11:49 you use ^L to separate 3D slices, basically 03:11:56 OK, that makes sense then 03:12:00 i thought you meant those were instructions 03:12:22 catseye mentions other topologies so that's uh, something 03:12:53 continuity sounds like it would be annoying. 03:14:48 with a certain number of dimensions having instructions might be redundant. you only need like... i think seven continuous dimensions to run a UTC 03:15:58 * mahem1 is still confused on how to use l and h... 03:16:10 -!- Sprocklem has joined. 03:16:42 the befunge 98 spec link is broken :( 03:17:26 easy to find on github though i guess, if that's what this is 03:17:38 http://quadium.net/funge/spec98.html 03:19:14 the delta function uses absolute space! how am i supposed to generalize this to a kerr-newman metric 03:19:39 All the spec really seems to say is. 03:19:41 The h "Go High" instruction causes the IP to travel up (delta <- (0,0,1)); the l "Go Low" instruction causes the IP to travel down (delta <- (0,0,-1)). These instructions are not available in Unefunge or Befunge. 03:20:20 So you're wondering if that's a one time move or a more permanent change like <> etc? 03:20:44 Permanant change I would assumen, cause it's changing the delta. 03:20:54 so what don't you get 03:21:00 But I am askign what code does it go and excute. 03:21:06 Cause you only write code in 2d 03:21:12 well, the code above, presumably 03:21:18 elliott_ said you specify that with some ^L thing 03:21:31 otherwise i guess it's whatever's in fungeland by default, which is probably nops? 03:21:36 the quadium.net spec is out of date 03:21:51 I don't have a 3d text editor. :P 03:21:52 use https://github.com/catseye/Funge-98/blob/master/doc/funge98.markdown 03:22:04 What's an example of using ^L? 03:22:04 your text editor is weak 03:22:05 anyway, I don't get the confusion 03:22:08 the IP starts at (0,0,0) 03:22:11 if you write 03:22:12 abc 03:22:12 def 03:22:15 ^L 03:22:18 ghi 03:22:20 jkl 03:22:23 then the IP starts at a 03:22:25 e is at (1,1,0) 03:22:28 g is at (0,0,1) 03:22:31 the internal links are broken on the github, fuckin a 03:22:31 l is at (2,2,1) 03:22:37 (iirc) 03:22:46 "In Trefunge-98, the Form Feed (12) character increments the z coordinate and resets the x and y coordinates to zero." 03:22:51 I guess you might need to omit the newlines around the ^L there. 03:23:07 all this is just about interpreting source files; the ^Ls don't exist in fungespace or anything. 03:23:15 just like the newlines don't, for 2d funge 03:23:43 i have no idea what this lahey space thing is talking about 03:23:54 I see. Pretty crazy stuff. Lemme try.. 03:24:05 Bicyclidine: it's just talking about how wrapping works. 03:24:21 Bicyclidine: since the space is infinite, but if you have e.g. "foobar" then it runs f, o, o, ..., a, r, f, o, o, ... 03:24:31 rather than running f, o, o, b, a, r and then running spaces (which just get skipped) forever 03:25:13 you can think of it as just being a finite wrapping space that expands as you place stuff outside the boundaries 03:25:18 but that doesn't quite specify all the behaviour. 03:25:32 lahey space is just the overly-fancy "mathematical" model of the space being used. 03:25:39 of course. 03:25:47 i don't understand how this is distinct from a torus i admit 03:26:11 oh, maybe for weird deltas 03:26:12 or something 03:26:17 because toruses are finite 03:26:20 fungespace is infinite. 03:26:39 it's not a fixed, wrapping size, it's both infinitely large and wrapping. 03:26:54 when you wrap you go over an /infinite/ amount of space, conceptually 03:26:55 so a resizeable torus, practically speaking 03:27:10 yeah, but that's not an actual specification of the space. :p 03:27:15 and there are edge-cases in terms of how you do that resizing. 03:27:17 nerds... 03:27:23 but yeah that makes sense i gues 03:27:31 note that this is relevant in cases like foo"bar 03:27:45 i've been reading about strange attractors for hours so everything looks like topological garbage 03:27:49 that runs f, o, o, and then pushes I think "bar foo" to the stack 03:27:54 manifolds... everywhere... 03:27:56 (by which I mean, it pushes those ascii values) 03:27:57 oh that's pretty cute 03:28:08 and then executes b? 03:28:10 in stringmode, consecutive spaces get collapsed, like in HTML 03:28:16 it's just collapsing aleph_null of them :p 03:28:20 Bicyclidine: *nod* 03:28:23 i bet you could write some really horrid programs that way 03:28:36 though i guess befunge is probably good for that in general 03:29:56 maybe i should like... actually write a program. 03:32:25 does esolang have any that are properly lorentz invariant 03:32:46 or whatever the words are 03:35:33 do you think running a bunch of semiconductors counts as a programming languages 03:44:59 -!- tromp_ has joined. 03:54:29 You only need two registers for TC, right? what's the instruction set like again 04:02:46 -!- zzo38 has joined. 04:11:46 I tried to figure out the decks of Pokemon Card GB2 cards only, assuming all random stuff to your benefit and your opponent is cooperating but nobody can concede, how to win with the largest number of cards in your draw pile at the end. 04:12:38 and? 04:12:46 I came up with: Your cards = 1x GAMBLER, 1x any basic pokemon card. Opponents cards = 1x MEOWTH [Lv.14], 3x PORYGON [Lv.12], 2x RATTATA [Lv.9], 1x RECYCLE ENERGY, 4x IMAKUNI?, 1x ITEMFINDER. 04:13:38 -!- Left_Hand_6969 has joined. 04:13:44 Bicyclidine: minsky machine 04:14:06 Bicyclidine: http://esolangs.org/wiki/Minsky_machine 04:14:47 "alternate state transition" is basically a jump right 04:14:52 it's been a while since i've played... the gambler does it? 04:14:58 yeah 04:15:09 god it must be a pain in the ass to write anything 04:15:24 AndoDaan: GAMBLER card means: Shuffle all cards from your hand into your draw pile. Toss a coin. If heads, draw eight cards; if tails, draw one card. 04:16:16 how many cards do i have in my hand on average? 04:16:54 You start with seven cards, and then draw one card on the beginning of each turn (if you cannot, you lose). 04:17:27 oh wait... 8-> and maybe eight back. 04:17:48 i'm imagining how to compile from brainfuck and, seriously, pain in the ass 04:21:21 Someone told me that "while(setjmp(buf));" is like resumable exceptions. Does resumable exceptions have anything to do with the RESUME command in BASIC? I can see some similarity. 04:22:57 anyway i guess i'm pretty sure that trefunge with only <>^vlh is tc. 04:23:32 I doubt it. 04:23:37 how could you ever branch 04:24:06 you use your position in x as reg1, position in y as reg2, and position in z as the ip. the program is fixed so it should be okay 04:24:13 branching would be going to wherever in z 04:24:30 how do you branch on any of those registers 04:25:08 you position an h in the appropriate place on the zero of the appropriate axis 04:25:12 or an l or whatever 04:25:47 you can only branch to constant instructions, so you know all the branches when you're setting up fungespace, and there's a finite number, so you can space things out appropriately 04:26:29 well... ok. write it up :p 04:26:35 I'm too tired to analyse. 04:26:43 it sounds interesting though. 04:27:35 i think i'll just believe minsky that two registers is TC and that'll be that 04:28:12 -!- shikhin has quit (Ping timeout: 260 seconds). 04:28:30 anyway there must be some number of dimensions it works for, i don't think discretizing the space ought to make physics totally impossible and obviously with a given number of continuous dimensions you only need to move cos phase fuckin space 04:30:21 i guess the problem is that actually finding a program written in a minsky machine is like... why would you do that 04:32:14 -!- tromp_ has quit (Read error: Connection reset by peer). 04:32:21 -!- tromp_ has joined. 04:34:45 Bicyclidine: there's one linked to from the Etre page 04:35:39 so there is! let me see if i can even understand this syntax 04:35:43 Bicyclidine: also, wouldn't it be easier to simulate Collatz functions in trefunge with those commands? 04:36:14 maybe but i'm not thinking about that 04:36:39 oh i thought you just wanted to prove that subset TC 04:37:16 well, i'm thinking of this in the context of strange attractors. more proofs the merrier i guess 04:38:58 e.g. i think reducing it to two dimensions would /not/ be TC, but i can't prove that off the top of my head 04:39:24 and if i said "by the poincare-bendixson theorem" people would just be confused 04:46:30 i guess the initial condition of fungespace has to be infinitely large, though repetitive. oh well 04:47:49 should i put it on the funge talk page or what 04:48:48 -!- Left_Hand_6969 has quit. 05:02:44 -!- Sprocklem has quit (Ping timeout: 272 seconds). 05:05:47 oh durr the paper i was thinking of is only three dimensional :x 05:33:14 -!- Bicyclidine has quit (Ping timeout: 250 seconds). 05:34:36 If a recruiter is blatantly targetting me personally, but suggests they want to talk today or tomorrow, is it ok to hold off a bit? I mean, I'm the one in demand 05:38:22 Just ask yourself "would a stupid person hold off a bit?" And if the answr is "no" then yes. 05:41:55 * Sgeo mutters something about reversed stupidity, and also not being smart enough to model a stupid person 05:43:09 -!- kcm1700 has quit (Remote host closed the connection). 05:43:17 -!- kcm1700 has joined. 05:43:38 -!- Bicyclidine has joined. 05:51:59 [wiki] [[Talk:Befunge]] http://esolangs.org/w/index.php?diff=40585&oldid=40244 * 69.166.47.145 * (+4051) hot singles in fungot's area 05:52:50 -!- kcm1700 has quit (Remote host closed the connection). 05:52:57 wow, that thing's quicker than i thought 05:57:52 -!- kcm1700 has joined. 06:00:16 What, the recent changes feed? It's very pushy. 06:01:07 i know "somebody" "hypothetically" "named" "monotone" with a similar irc widget that is much slower, is all 06:04:52 It's built on the MediaWiki UDP recent changes feed, plus a socat-based shell oneliner to mangle the message into raw IRC and write it to the Unix domain socket that HackEgo accepts extra stuff on. 06:05:18 the other one is, uh, dokuwiki i think. maybe dokuwiki sucks 06:05:40 Maybe it's based on polling, then. 06:05:49 maybe 06:13:37 UDP recent changes feed? wtf, who invents such a thing 06:14:08 people who need up to the millisecond updates on my half baked phase constructions at midnight 06:24:38 -!- MoALTz has quit (Quit: Leaving). 06:26:33 [wiki] [[User:Rdebath]] http://esolangs.org/w/index.php?diff=40586&oldid=40409 * Rdebath * (+7212) Add tests and interpreters (plus upgrades) 06:27:09 http://www.mediawiki.org/wiki/Manual:$wgRCFeeds 06:28:15 It can do UDP and Redis Pub/Sub, and the message itself can be the above (with IRC color codes and all) or a JSON object. 06:29:21 -!- kcm1700 has quit (Remote host closed the connection). 06:29:29 -!- kcm1700 has joined. 06:38:07 -!- AndoDaan has quit (Ping timeout: 276 seconds). 06:40:14 * atehwa acknowledges self-Finnishness @ mroman_ 06:59:37 -!- AndoDaan has joined. 07:06:10 -!- AndoDaan_ has joined. 07:06:24 -!- AndoDaan has quit (Ping timeout: 260 seconds). 07:06:34 -!- AndoDaan_ has changed nick to AndoDaan. 07:10:52 -!- AndoDaan has quit (Ping timeout: 240 seconds). 07:11:21 -!- lambdabot has quit (Write error: Broken pipe). 07:14:36 -!- AndoDaan has joined. 07:14:59 -!- AndoDaan has quit (Client Quit). 07:15:21 -!- AndoDaan has joined. 07:17:02 -!- lambdabot has joined. 07:24:45 moin 07:24:54 @messages-loud 07:24:55 oerjan said 9h 57m 35s ago: The hackage-story with "corrupt acid-state logs" made me really sceptical of using acid-state in future projects <-- afaiu it wasn't _really_ corrupted, the recovery just didn't know how to ignore the junk at the end of the log from the aborted transaction 07:25:28 @ask oerjan Can you do backups of acid-state databases? 07:25:29 Consider it noted. 07:28:21 -!- Patashu has joined. 07:28:37 -!- shikhin has joined. 07:29:07 -!- shikhin has changed nick to Guest72000. 07:30:10 -!- Guest72000 has changed nick to shikhout. 07:30:40 -!- shikhout has changed nick to Guest40212. 07:31:31 -!- Guest40212 has changed nick to shikherr. 07:31:36 -!- shikherr has quit (Changing host). 07:31:36 -!- shikherr has joined. 07:33:01 -!- shikherr has changed nick to shikhin. 07:48:54 -!- Bicyclidine has quit (Ping timeout: 246 seconds). 07:56:19 -!- AndoDaan_ has joined. 07:56:49 -!- AndoDaan_ has left. 07:56:54 -!- AndoDaan_ has joined. 07:59:02 -!- AndoDaan has quit (Ping timeout: 258 seconds). 08:00:30 -!- AndoDaan_ has changed nick to AndoDaan. 08:14:29 -!- brandonson has joined. 08:17:37 -!- S1 has joined. 08:22:12 -!- Phantom_Hoover has joined. 08:22:51 -!- shikhout has joined. 08:23:32 [wiki] [[Talk:EsoInterpreters]] http://esolangs.org/w/index.php?diff=40587&oldid=40572 * Keymaker * (+177) Reply. 08:26:02 -!- shikhin has quit (Ping timeout: 272 seconds). 08:28:03 -!- sebbu2 has joined. 08:28:40 -!- sebbu2 has quit (Changing host). 08:28:41 -!- sebbu2 has joined. 08:29:50 -!- sebbu has quit (Ping timeout: 272 seconds). 08:31:22 -!- zzo38 has quit (Ping timeout: 240 seconds). 08:31:43 -!- sebbu2 has changed nick to sebbu. 08:36:02 Bring back The Esotric Awards! 08:36:40 I mean, The... Essies. 08:38:49 the what? 08:39:13 http://esolangs.org/wiki/Esoteric_Awards 08:39:28 I bet it was star studde 08:39:30 d 08:41:34 hm 08:41:56 your burlesque would be a sho'win. 08:42:07 i don't know how to spell expressions. 08:43:37 I remember talk about the Essies, but nothing about the competitions themselves. 08:44:05 "It is not known if any entries were submitted. It is likewise not known if a winner was picked." 08:44:18 Very appropriate. 08:44:18 seems like that would be the most exciting bit. 08:46:40 well... 08:46:46 if the award ceremony is in finnland 08:46:50 I might participate. 09:03:29 Grumble frumble sox FAIL formats: can't open input file `data/adults/test/man/gp/15z8973a.wav': sph: unsupported coding `pcm,embedded-shorten-v2.00' 09:03:51 I don't quite see why all these corpora go for the NIST SPHERE + Shorten format. 09:04:14 -!- brandonson has quit (Quit: KVIrc 4.2.0 Equilibrium http://www.kvirc.net/). 09:08:48 -!- oerjan has joined. 09:09:57 holy shit a nobel to trondheim 09:10:47 @messages- 09:10:47 boily said 10h 22m 9s ago: thanks for the irradiation! 09:10:47 mroman_ asked 1h 45m 18s ago: Can you do backups of acid-state databases? 09:12:25 -!- brandonson has joined. 09:13:09 @tell mroman_ They did nightly backups. It was never a matter of losing more than that. But if they had understood in time that it was just a single partially written transaction, they could have restored everything up to that as well. 09:13:09 Consider it noted. 09:14:42 @tell mroman_ *up to the last moment, as the system _did_ shut down for safety when it happened. 09:14:42 Consider it noted. 09:21:14 but how do you do backups of that? 09:21:21 @messages- 09:21:22 oerjan said 8m 12s ago: They did nightly backups. It was never a matter of losing more than that. But if they had understood in time that it was just a single partially written transaction, they could have restored everything up to that as well. 09:21:22 oerjan said 6m 39s ago: *up to the last moment, as the system _did_ shut down for safety when it happened. 09:21:50 you'd have to stop the application 09:21:54 then copy the acid-state folder? 09:22:01 it's possible they do? i don't know. 09:22:03 then start the application 09:22:17 I mean... you don't know when acid-state will write to those files 09:22:26 so copying the files mid-write is probably not a good idea 09:26:41 i don't know how this works. ask someone who does. 09:27:28 i want the neighbors' dog to die so much. 09:27:50 -!- shikhout has changed nick to shikhin. 09:28:19 it has this completely eerie ability to start barking the moment i'm trying to concentrate. 09:28:55 and then to keep silent enough of the rest of the time that i don't feel i can complain too much. 09:29:10 (although today is starting to push it.) 09:34:57 mroman_: look at this discussion http://www.reddit.com/r/haskell/comments/xs0t2/acidstate_with_yesod/c5p8af7 09:40:30 `ls 09:40:31 :-( \ a.out \ bdsmreclist \ bin \ binpipes \ canary \ cat \ complaints \ crunchfuck \ :-D \ dc \ dog \ etc \ factor \ gcccomp \ head \ hej \ hello \ hello.c \ ibin \ interps \ lib \ moop.txt \ paste \ pref \ prefs \ quines \ quotes \ share \ src \ test.c \ Wierd \ wisdom \ wisdom.pdf 09:40:40 `rm gcccomp 09:40:41 No output. 09:42:25 `cat moop.txt 09:42:25 Moop! 09:42:28 Useful. 09:42:52 `url binpipes 09:42:52 http://codu.org/projects/hackbot/fshg/index.cgi/file/tip/binpipes 09:43:03 `unidecode ⋖ 09:43:04 [U+22D6 LESS-THAN WITH DOT] 09:43:16 -!- heroux_ has joined. 09:43:33 -!- shachaf_ has joined. 09:43:40 -!- FreeFull_ has joined. 09:43:45 -!- heroux___ has joined. 09:44:47 -!- heroux__ has quit (Excess Flood). 09:44:48 -!- heroux has quit (Write error: Broken pipe). 09:44:52 -!- heroux_ has changed nick to heroux. 09:44:52 -!- heroux has changed nick to 16WAAR95O. 09:44:52 -!- heroux___ has changed nick to 7JTAA4WJ5. 09:45:02 -!- 7JTAA4WJ5 has changed nick to heroux. 09:45:28 i think that was put there by someone who doesn't understand unixy directories. 09:46:06 -!- variable has quit (Ping timeout: 467 seconds). 09:46:45 `` rm binpipes moop.txt 09:46:47 No output. 09:46:52 " Or do you expect very expensive haskell developers spend the rest of their lives creating hundreds of customized reports ?" 09:46:53 It's also not terribly useful over IRC. 09:46:59 haskell developers are expensive? 09:47:06 i didn't think so either 09:47:31 mroman_: i think that vagif guy may be _ever_ so slightly trolly on occasion. 09:47:37 (maybe always.) 09:47:39 -!- shachaf has quit (Remote host closed the connection). 09:47:40 -!- FreeFull has quit (Remote host closed the connection). 09:48:09 `url crunchfuck 09:48:09 http://codu.org/projects/hackbot/fshg/index.cgi/file/tip/crunchfuck 09:48:31 oerjan: He's not completely wrong though 09:48:40 oerjan: I think you cheated on hello hello world, 84 looks impossible ;-) 09:48:49 Interfacing with an acid-state database pretty much relies on you writing some export-my-data-service 09:48:54 `rm crunchfuck 09:48:56 No output. 09:49:09 -!- variable has joined. 09:49:31 int-e: nope, it's completely non-cheating 09:50:33 oerjan: have you seen my terrible prime solution? using or[] and a prime test that works for numbers up to 340 (as the other two 61 character solutions do) is comparatively sane. 09:50:57 oh it's finished? hadn't checked. 09:51:11 > (mod(2^341-2)341,mod 341 11) 09:51:13 (0,0) 09:51:56 the other drawback is you can't use acid-state with large databases on low-memory devices 09:52:19 int-e: so i may still have the shortest _non_-cheating solution on that? :P 09:52:33 even if it's the longest in the list. 09:52:44 oerjan: I think so, yes. 09:53:37 I uploaded my long Forth hello-hello-world because nobody else seemed to be doing Forth, seems that's still true. 09:54:18 hmm, this hello world task again, which task is this? 09:54:23 can you give a link? 09:54:42 http://golf.shinh.org/p.rb?hello+hello+world 09:54:56 thanks 09:56:00 but wasn't that about ternary? how does prime testing come into this/ 09:56:01 ? 09:56:13 fizzie: the trick to get a huge amount of points on anagol is to submit an incredibly large program in a language nobody has yet submitted something just short before deadline 09:56:14 oerjan: however: main=mapM print[z+1|z<-[0..252],p(z+2)||p z];p n=all((>0).mod n)[2..n-1] 09:56:16 b_jonas: no that's http://golf.shinh.org/p.rb?A045718 09:56:22 then you submit your real solution just right after the deadline 09:57:01 your solution is then compared to the best solution before the deadline (i.e. your incredibly large program) 09:57:08 mroman_: hehe 09:57:10 and receives points depending on how much shorter it is 09:57:18 which means you can get more than 10k points 09:57:25 oerjan: and the or[] trick is applicable as well: main=mapM print[z+1|z<-[0..252],or[all((>0).mod n)[2..n-1]|n<-[z+2,z]]] 09:57:39 I should try that. I can generate long programs in strange languages for some tasks. 09:58:59 > length "main=mapM print[z+1|z<-[0..252],p(z+2)||p z];p n=all((>0).mod n)[2..n-1]" 09:59:01 72 09:59:05 oerjan: and another character: main=mapM print[z|z<-[1..252],or[all((>0).mod(n+1))[2..n]|n<-[z,z-2]]] 09:59:19 > length "main=mapM print[z|z<-[1..252],or[all((>0).mod(n+1))[2..n]|n<-[z,z-2]]]" 09:59:20 70 10:00:55 i also had a version using or instead of any 10:01:03 was the same length, though. 10:01:46 actually it was using _both_ or and and. 10:02:12 l=[2..255];main=mapM print[z|z<-1:l,or[and[x*y-z/=s|x<-l,y<-l]|s<-[-1,1]]] 10:02:30 (I submitted the 70 characters one) 10:03:41 -!- S1 has quit (Quit: S1). 10:03:43 -!- AndoDaan_ has joined. 10:04:12 -!- AndoDaan has quit (Ping timeout: 272 seconds). 10:04:43 I have a 20-character Burlesque A006520 that runs in 1.1 seconds on my own computer, but is too slow for anagolf. :/ 10:06:11 -!- conehead has quit (Quit: Computer has gone to sleep). 10:06:55 i guess this mod(2^y-2)y<1 trick is something well-known for golfing. 10:07:49 I discovered it myself, but it appears in several publicly visible entries by henkma 10:07:54 (and others) 10:07:55 so yeah 10:08:02 fizzie: did you compile with -O3? 10:08:33 (well, "discover" is perhaps too strong, since I have read a bit about pseudoprimes before) 10:09:06 also what's ghc's default for -O? 10:09:43 oerjan: what trick? 10:09:46 according to github shinh didn't compile it with -O3 10:10:19 myname: it's a primality test that works up to 340 10:11:33 -!- AndoDaan has joined. 10:12:05 hm... 10:12:43 > let p=nubBy(((<1).).mod)[2..] in p 10:12:45 [2,3,5,7,11,13,17,19,23,29,31,37,41,43,47,53,59,61,67,71,73,79,83,89,97,101,... 10:13:27 -!- AndoDaan_ has quit (Ping timeout: 272 seconds). 10:13:40 `factor 341 10:13:40 341: 11 31 10:13:45 1024 1024?*ro)fCL[ runs in 1.7s on my machine 10:13:48 !blsq 101ro{fcL[2==}f[ 10:13:48 {2 3 5 7 11 13 17 19 23 29 31 37 41 43 47 53 59 61 67 71 73 79 83 89 97 101} 10:13:54 on anagol it runs at least 3.5s but then runs out of stack space 10:16:51 with -O0 on my machine it takes 3.5s to run 10:17:00 that's huge 10:17:08 > let p=nubBy(((<1).).mod)[2..]; c=concat$zipWith(\m n->[m+1..n-1])p(tail p) in c 10:17:10 [4,6,8,9,10,12,14,15,16,18,20,21,22,24,25,26,27,28,30,32,33,34,35,36,38,39,4... 10:17:45 mroman_: I did what your README.md suggested, so probably. 10:17:53 > let p=nubBy(((<1).).mod)[2..]; c=concat$zipWith(\m n->[m+1..n-1])p(tail p); k=2 in find(\y -> mod(k^y-k)y<1)c 10:17:55 Just 341 10:18:01 > let p=nubBy(((<1).).mod)[2..]; c=concat$zipWith(\m n->[m+1..n-1])p(tail p); k=3 in find(\y -> mod(k^y-k)y<1)c 10:18:03 Just 6 10:18:26 hmph 10:18:29 oerjan: it's hard to outperform the 2 10:18:33 ic 10:18:44 > let p=nubBy(((<1).).mod)[2..]; c=concat$zipWith(\m n->[m+1..n-1])p(tail p); k=10103 in find(\y -> mod(k^y-k)y<1)c 10:18:46 Just 561 10:19:02 i guess it's the shortest for its performance, then 10:19:09 1024 1024?*ro{fC++J?*}m[++ is 30s with -O0 and 19s with -O3 :) 10:19:26 oerjan: 561 is a https://en.wikipedia.org/wiki/Carmichael_number, so that's the best possible. 10:19:37 -!- boily has joined. 10:20:19 (for this Fermat style primality test) 10:20:29 int-e: yeah, that's a nice trick, though I think in some languages a slow trial division test might be shorter. 10:20:34 ah 10:21:13 like, there's that crazy regex backreference test that works in perl and some other languages 10:21:49 that's like eww but for such small numbers it should work, and you might be able to embed the plus minus one into the regex 10:21:53 hmm 10:22:05 see http://www.perlmonks.com/?node_id=510925 10:22:44 ('1' x shift) !~ /^(11+)\1+$/ 10:22:49 what's the upper bound for that test? 10:24:59 none, it's perfect 10:25:13 it'll be slow for large numbers though 10:26:32 is that a unicode operator? 10:27:40 huh? 10:27:48 the x? 10:27:52 or is that a regular x 10:28:02 the letter 10:28:05 also... does that use base1 numbers? 10:28:08 or latter, if you like ;-) 10:28:33 I have no idea how that code works 10:28:39 (also I don't know ANY perl) 10:28:42 (11+) <-- match 2 or more 1s. \1+ <- match what the () matched, one or more times 10:29:05 wait 10:29:07 so you match (1+n)(1+m) 1s, where n,m>1 10:29:15 a string of 1 of composite length 10:29:37 but '1' x shift converts to base1? 10:29:42 oh and note that the test fails for 0 and 1. 10:29:44 yes 10:30:15 (that is, 0 and 1 are primes according to that test) 10:30:54 Makes sense 10:31:01 0 is kinda a super-prime 10:31:10 you can divide primes by itself and 1 10:31:21 you can divide 0 only by 1 but not by itself 10:31:36 so it's even primier then every other prime 10:31:39 you can divide 0 by any natural number except 0 10:32:01 of course 10:32:25 you can divide anything by anything 10:32:37 elephant / tofu 10:32:38 see 10:32:55 is that ABNF? 10:32:59 0 is the maximum of the divisibility relation; 1 is the minimum, and primes are the successors of 1. 10:33:02 !blsq 10 1B! 10:33:02 Ain't nobody got time fo' dat! 10:33:09 hmm 10:33:16 yeah. that doesn't work 10:33:22 !blsq "1"10.* 10:33:22 {"1" "1" "1" "1" "1" "1" "1" "1" "1" "1"} 10:33:28 !blsq '110.* 10:33:28 "1111111111" 10:33:33 what does it try and calculate? 10:33:38