00:05:12 <fizzie> http://zem.fi/esostats/people_summary.html PLOOOOTS (also beta)
00:10:36 <quintopia> i'm pretty sure my activity level spikes iff interest in bfjoust spikes
00:14:06 <ais523> quintopia: mine probably falls, as I spend more time thinking BF Joust and less talking on IRC
00:15:01 <quintopia> ais523: your baseline activity level is higher than mine i think
00:16:23 <quintopia> bfjoust spikes are the one time i am more activious than you
00:21:30 <Bike> oh i just realized "ploooots" was supposed to be the plural of "plot"
00:21:37 <Bike> and not a bunch of ploots (what's a ploot)
00:22:32 <elliott> fizzie: I'm still not merged with zuff!!
00:22:57 <elliott> I like how it doesn't say what the colours are
00:23:46 <fizzie> It's still not recalculated. And the colours are obviously obvious.
00:24:02 <elliott> i don't know what the colours mean !!!
00:24:20 <fizzie> Blue is the overall activity, and red is the selected nick.
00:24:55 <elliott> why does it show me as not being active in 2007 at all
00:24:58 <monqy> wow elliott has ungodly huge activity
00:25:32 <ais523> elliott: did you use the nick "elliott" ever then?
00:25:37 <elliott> monqy: iirc i have more than twice the lines said in #esoteric than the second person......
00:25:41 <ais523> it's likely a failure in nick merging
00:25:44 <elliott> ais523: no but it's meant to be merged and stuff
00:25:56 <fizzie> Phantom_Hoover: Which one of those three would be the "canonical" name?
00:25:57 <ais523> is that when you were using "alise"?
00:26:21 <elliott> it was ehird and then maybe a `
00:26:30 <fizzie> ais523: 'elliott': ['elliott', 'ehird', 'alise', 'tusho', 'aliseiphone', 'ehirdiphone', 'zuff'],
00:26:43 <elliott> fizzie: well it's still in that menu at least
00:26:47 <elliott> anyway you need ehird` in there SORRY
00:26:47 <ais523> fizzie: 'ehird`' is the nick you're missing
00:26:54 <fizzie> elliott: Yes, it's old datta.
00:27:01 <Phantom_Hoover> <fizzie> Phantom_Hoover: Which one of those three would be the "canonical" name?
00:27:07 <ais523> you didn't use estoppel very much
00:27:11 <ais523> fizzie: what list do you have for me?
00:27:15 <elliott> imo leave fax separated so you don't have to deal with the mess of who is fax or not
00:27:38 <fizzie> ais523: It was supposed to remove all trailing non-alphanumerics, but maybe it didn't.
00:27:53 <ais523> but ais523 and ais523_ are different computers!
00:28:07 <ais523> and as a result, may talk about different topics
00:28:09 <monqy> how about alphanumerics in the middle
00:28:18 <ais523> ais523_ is substantially more likely than me to bitch about Java, for instance
00:28:22 <shachaf> You're also missing "shachaf".
00:28:28 <monqy> cf phantom____________________________________hoover
00:28:29 <shachaf> "elliott's gr8st pseudonym"
00:28:48 <fizzie> ais523: For you I just have "ais523" and "ais523\unfoog". I made these by getting a top-100 list of overall activity.
00:28:48 <shachaf> monqy: uve been shachaffed!!
00:28:53 <elliott> monqy: i thought it was a "joek"
00:29:06 <fizzie> monqy: I have _, __ and ___ manually listed for that.
00:29:06 <ais523> fizzie: "scarf", "callforjudgement"
00:29:22 <monqy> fizzie: but think about how phantom____hoover feels
00:29:29 <fizzie> ais523: I see "scarf" ended up on the list, too.
00:29:32 <ais523> also "this", though I don't have it registered
00:29:39 <ais523> and only use it for the purpose of trolling myself
00:30:41 <shachaf> ais523 is a silly thing to do.
00:31:10 * ais523 sort of thing is more legible when it starts "this" not "ais523"
00:34:01 -!- Phantom_Hoover has quit (Remote host closed the connection).
00:54:09 <shachaf> kmc: Are you the "turn in crypto challenges in asm" person?
01:41:53 -!- ais523 has quit.
02:29:32 -!- Nisstyre has joined.
02:32:05 -!- Nisstyre_ has quit (Ping timeout: 252 seconds).
02:35:35 -!- Nisstyre has quit (Ping timeout: 255 seconds).
02:36:47 -!- Nisstyre has joined.
02:45:15 <tswett> Wait, I just realized something.
02:45:20 <tswett> Mirrors are Turing-complete, aren't they?
02:45:40 <tswett> Like, you can store an arbitrary amount of accessible information in the position of a beam of light.
02:47:43 <tswett> In particular, if you consider a beam of light to have x and y coordinates (and it's traveling in the z direction), you can use mirrors to perform scaling and translation on x and y, as well as conditionals.
02:48:08 <zzo38> Yes I can understand now how you mean.
02:48:14 <Bike> is this a billiard ball computer?
02:48:40 <tswett> Nope. There's just a single "ball"; it doesn't interact with other moving objects.
02:48:47 -!- copumpkin has quit (Ping timeout: 245 seconds).
02:49:18 -!- copumpkin has joined.
02:49:36 -!- Nisstyre_ has joined.
02:49:45 <tswett> The ball just moves in a constant direction until it hits a mirror, at which point it bounces off and continues moving in the appropriate new direction.
02:50:05 <tswett> If your space is three-dimensional, this configuration is Turing-complete.
02:50:25 -!- evincar has quit (Quit: leaving).
02:51:25 -!- Nisstyre has quit (Ping timeout: 252 seconds).
02:51:28 <tswett> Of course, this wouldn't work in the real world, for many reasons, one of which is that real light diffracts.
02:52:05 <tswett> What if you use an electron instead? I guess electrons also diffract.
02:52:46 <oerjan> all objects diffract, in principle
02:52:54 <Bike> i feel like i should say something about heisenberg but then i'd be pretending to know quantum physics, which i do not
02:53:08 <elliott> Bike: whoa hw man thats so QUNAutm
02:53:26 <tswett> Yeah, I think Heisenberg's principle applies here. A photon does not have one distinct position.
02:53:36 <shachaf> Quantum? I don't even know 'em!
02:54:44 -!- zzo38 has quit (Remote host closed the connection).
02:56:14 -!- Bike_ has joined.
02:57:10 <Sgeo> elliott, why is *: defined when *~ is the same thing?
02:57:28 <Sgeo> (at least monadic *: )
02:58:08 <elliott> does *~ have the inverse that *: does
02:58:20 <elliott> the real answer is that i don't know tho
02:59:11 -!- Bike has quit (Ping timeout: 245 seconds).
03:02:28 <monqy> *~ is the 'reflexive' of * (~ being the 'reflex' adverb); i can't recall how that interacts with obverses
03:03:46 <monqy> ahh, the obverse of *~ is specially specified to be %:
03:03:57 <monqy> i.e. sqrt i.e. same as the obverse of *:
03:04:19 <kmc> tswett: you've reminded me of the optical computer for solving traveling salesman problem
03:04:45 <kmc> you have a bunch of beamsplitters and you connect them with optical fibers whose lengths correspond to the distances between cities
03:04:53 <kmc> then you use interferometry (somehow) to find the shortest path
03:07:30 <Bike_> hm, what's a good academic who does weird-ass computation stuff
03:07:43 <Bike_> i checked for adamatzky doing things like this but he's too busy getting stoned on BZ
03:08:36 <kmc> getting stoned on BZ is a bad idea
03:09:04 <monqy> the one it's a good idea to get stoned on
03:09:05 <kmc> so not the chemical weapon
03:09:25 <Bike_> no, although i've heard apparently people have tried to use that recreationally...
03:09:30 <Bike_> i mean the clock reaction
03:09:43 <kmc> i don't know what that is
03:10:03 <Bike_> http://en.wikipedia.org/wiki/Belousov-Zhabotinsky_reaction
03:10:15 <Bike_> observe: pretty colors!
03:10:16 <kmc> oh those are neat
03:10:31 <Bike_> http://arxiv.org/abs/1212.1344 so, adamatzky does things like this with 'em
03:10:42 <kmc> http://upload.wikimedia.org/wikipedia/commons/d/d9/The_Belousov-Zhabotinsky_Reaction.gif
03:11:41 <Bike_> also a clock reaction is just a reaction that does something periodic, so you could use it as a clock if you wanted your clock to be made of acids.
03:12:00 <Bike_> anyway so not tswett lightmachines
03:12:13 <kmc> oh i heard about this slime mould computer too
03:12:59 -!- DH____ has quit (Read error: Connection reset by peer).
03:13:05 -!- DHeadshot has joined.
03:13:12 <Bike_> oh, there's a journal or something called "unconventional computing", that's more like it!
03:14:09 <kmc> "The denizens of Carlisle, meanwhile, may wonder what objection slime moulds have towards their fine city."
03:14:12 <Bike_> 'Abstract Geometrical Computation 6: A Reversible, Conservative and Rational Based Model for Black Hole Computation' i think i'm going to be distracted
03:16:12 -!- oerjan has quit (Quit: leaving).
03:16:15 -!- Arc_Koen has quit (Quit: Arc_Koen).
03:18:39 <kmc> if you drop a supermassive computer into a rotating black hole while traveling at the speed of light then you meet god and can just ask him to solve your dumb problem
03:22:04 <elliott> `addquote <kmc> if you drop a supermassive computer into a rotating black hole while traveling at the speed of light then you meet god and can just ask him to solve your dumb problem
03:24:26 <Bike_> kmc: "O father who art in heaven, I really need to win this CTF"
03:27:36 <Bike_> 'Can a Computer be “Pushed” to Perform Faster-Than-Light?' i love how far out these questions are
03:27:55 <Bike_> "Let us assume that Alice, Bob, and Charlie, the three classical people of cryptography are not limited anymore to perform a finite number of computations on real computers, but are limited to α computations and to α bits of memory, where α is a fixed infinite cardinal."
03:27:59 -!- Bike_ has changed nick to Bike.
03:28:57 -!- WeThePeople has quit (Quit: Leaving).
03:29:26 <Bike> good assumption i think
03:29:53 <Bike> Oh, hey, I think this one is actually vaguely sort of near what tswett was saying.
03:29:56 <Bike> «We propose to “boost” the speed of communication and computation by immersing the computing environment into a medium whose index of refraction is smaller than one, thereby trespassing the speed-of-light barrier.»
03:32:07 <kmc> you can have fairly mundane materials whose index of refraction is less than one
03:32:21 <kmc> because IoR is defined according to the phase velocity
03:32:31 <kmc> but this doesn't imply superluminal transfer of information
03:32:37 <Bike> want the paper?
03:35:32 <kmc> summertime and the monoids are easy
03:36:09 <kmc> i downgraded my old laptop from a SSD to a spinny disk
03:36:21 <kmc> have to remember not to throw that one at sofas anymore
03:36:28 <kmc> so that i can give the SSD to a friend
03:41:11 -!- Nisstyre_ has quit (Ping timeout: 255 seconds).
03:42:30 -!- Nisstyre has joined.
03:46:58 -!- glogbackup has quit (Remote host closed the connection).
03:51:00 -!- glogbackup has joined.
03:55:35 -!- glogbackup has quit (Ping timeout: 260 seconds).
04:03:21 -!- glogbackup has quit (Remote host closed the connection).
04:14:18 <shachaf> kmc: Did you read _Cat's Cradle_?
04:15:22 -!- glogbackup has quit (Ping timeout: 276 seconds).
04:17:26 -!- WeThePeople has joined.
04:24:45 -!- Nisstyre has quit (Ping timeout: 260 seconds).
04:26:48 -!- Nisstyre has joined.
04:40:34 <shachaf> City on fire! / Rats in the grass / And the lunatics yelling in the streets! / It's the end of the world! Yes!
04:41:01 -!- Nisstyre_ has joined.
04:41:19 <kmc> aren't there always rats
04:41:53 -!- Nisstyre has quit (Disconnected by services).
04:42:01 -!- Nisstyre_ has changed nick to Nisstyre.
04:44:51 <Sgeo> Hmm. J seems like it might be a reasonable language to implement Befunge-98 in
04:46:33 -!- Nisstyre has quit (Ping timeout: 252 seconds).
04:46:36 -!- Nisstyre-laptop has joined.
05:02:02 -!- WeThePeople has quit (Quit: Leaving).
05:14:31 -!- sirdancealot7 has quit (Ping timeout: 276 seconds).
05:15:03 -!- zzo38 has joined.
05:17:19 -!- TeruFSX has quit (Quit: Leaving).
05:27:29 -!- sirdancealot7 has joined.
05:37:44 <shachaf> kmc: By the way, you know ho CoYoneda IORef gives you a "read-only IORef" that can't do any other effects?
05:38:15 <shachaf> That can either be a good thing or a bad thing.
05:38:28 <Sgeo> Is APL more readable than J?
05:38:42 <shachaf> And how do they both compare to Clojure?
05:39:12 <Bike> "Is APL more readable than J" should be the PLT version of "How many angels can dance on the head of a pin".
05:40:12 <shachaf> are monoids more lovable than categories
05:40:46 <monqy> how long until you get tired of this beaky thing
05:41:13 <shachaf> monqy: how long until you get tired of the "monqy style"
05:45:01 <kmc> monoids mo problems
05:52:45 -!- quintopia has quit (Ping timeout: 240 seconds).
05:59:37 -!- quintopia has joined.
06:21:59 <kmc> the noid ruins pizzas
06:24:10 <shachaf> I crashed someone's Haskell bot with a GHC type system bug.
06:24:17 <shachaf> By crashed I mean unsafePerformIO-then-segfault
06:24:32 <kmc> you should own their computer
06:24:46 <shachaf> Instead of that I told them about -dcore-lint.
06:25:55 <Jafet> chmod +s /usr/bin/ghc
06:25:56 <kmc> 'On January 30, 1989, Kenneth Lamar Noid, a mentally ill customer who thought the ads were a personal attack on him, held two employees of an Atlanta, Georgia, Domino's restaurant hostage for over five hours.'
06:26:17 <kmc> -dcore-lint is like a condom for your haskell bot
06:28:08 <kmc> it's pretty easy to go from unsafePerformIO to an arbitrary memory read/write yeah?
06:28:14 <kmc> unsafeCoerce
06:31:20 <shachaf> This time by defining a custom Typeable instance.
06:31:25 <shachaf> He should probably turn on SafeHaskell
06:31:27 <Bike> Is this with -dcore-lint?
06:31:46 <kmc> read :: Word64 -> Word64; word addr = fromJust (unsafeCoerce (addr - 8) :: Maybe Int)
06:31:49 <kmc> or something?
06:32:14 <shachaf> Well, you can run arbitrary IO easily.
06:32:21 <kmc> how's that?
06:32:21 <shachaf> But I guess all the Foreign functions might not be in scope.
06:32:47 <shachaf> IO a = State# RealWord -> (# State# RealWorld, a #)
06:32:50 <kmc> you run IO by unsafeCoercing to something operationally isomorphic to... yeah
06:32:59 <shachaf> So you can cast to () -> () or something.
06:33:06 <kmc> that's tricky though because State# t has a weird STG-type
06:33:15 <kmc> but it probably works out
06:33:20 <shachaf> I managed to run IO both times so far, but I segfaulted it too.
06:33:29 <shachaf> Or actually illegal-instructioned it.
06:33:37 <shachaf> I guess it tried to evaluate the () and jumped to some invalid code.
06:33:54 <kmc> so how would you do a memory write, assuming Foreign stuff isn't in scope?
06:34:20 <shachaf> That seems tricky, actually.
06:34:37 <shachaf> Well, you can probably jump to arbitrary code?
06:34:59 <shachaf> That doesn't necessarily help you that much.
06:35:46 <kmc> does GHC's allocator set page permissions? you might be able to just create a ByteString and then jump into it
06:36:24 <shachaf> I think pages should be nonexecutable by default?
06:36:31 <shachaf> mauke has a "hello world" that works that way.
06:37:44 <shachaf> GHC only sets page permissions explicitly for foreign import wrapper, as far as I know.
06:38:41 <kmc> hm I didn't have to do anything with page permissions for http://mainisusuallyafunction.blogspot.com/2010/09/executing-bytestring.html
06:39:07 <shachaf> Well, it just allocates bytestrings with malloc.
06:40:54 <shachaf> Maybe it has to do with the FFI?
06:41:35 <kmc> did you get a PoC executing bytestring using only unsafeCoerce?
06:41:46 <Sgeo> Hmm. Apparently any single-line pointful verb definition can be converted into pointfree
06:42:00 <Sgeo> Are there expressions that @pl cannot convert in Haskell>
06:42:06 <kmc> can we have a poll to decide what the next Sgeolang is?
06:42:24 <shachaf> It might not be as easy as I said.
06:42:25 <Sgeo> Does J really count as a current Sgeolang?
06:43:25 <kmc> shachaf: I guess you would use a second ByteString to build a fake heap object, whose first word points to the code you want to execute
06:44:04 <kmc> then you get the address of that bytestring into a constructor field and force it
06:44:24 <shachaf> I bet the ByteString is allocated statically?
06:47:42 -!- Nisstyre has joined.
06:49:21 <shachaf> 7ffff6e00000-7ffff6f00000 rwxp 00000000 00:00 0
06:49:26 <shachaf> I guess it does allocate w+x memory
06:49:52 <kmc> good times
06:52:11 <kmc> the p is for profit
06:53:19 <shachaf> Now they turned on SafeHaskell but also left GeneralizedNewtypeDeriving on.
06:53:24 <shachaf> So I can still write unsafeCoerce. :-(
06:53:48 <Jafet> You should just root their system and fix it properly.
06:57:49 <shachaf> I don't see it allocating any w+x memory in strace
06:58:13 <kmc> for real? haha
06:58:34 <Bike> It's beaky in my shachaf fanfiction.
06:58:46 <shachaf> Please don't tell me about shachaf fanfiction.
06:58:56 <shachaf> You can write whatever you want, just don't tell me about it.
06:58:58 <Bike> It's titled "Coercion Ain't Easy"
06:59:19 <Bike> You like BDSM, right?
07:05:39 <Bike> ...is that even legal nowadays?
07:06:04 <shachaf> I like my SDSM blessed greased
07:06:16 <shachaf> I'm sure Jafet can empathize.
07:06:30 <Bike> tch, I bet you use scrolls of genocide too.
07:09:30 <Jafet> shachaf's gaze penetrated through the slits of the balustrade, as the excited crowd of algebra neophytes struggled to grab the professor by any wrinkled limb or lineament they could find. Here, he knew, was a master of the group action.
07:12:49 -!- FreeFull has quit.
07:13:53 <Bike> I think that went well.
07:19:28 <shachaf> kmc: 7fff5a4ea000-7fff5a50b000 rwxp 00000000 00:00 0 [stack]
07:19:41 <shachaf> I don't quite see what's going on. :-(
07:20:21 <shachaf> Maybe it's the ELF file somehow?
07:21:27 <shachaf> I doubt it's remapping the stack.
07:21:48 <kmc> yeah the file may set initial permissions for the stack
07:22:06 <shachaf> It's not just the stack, though.
07:22:27 <shachaf> 7eff280e9000-7eff28269000 r-xp 00000000 08:03 1448930 /lib/x86_64-linux-gnu/libc-2.13.so
07:22:49 <shachaf> 7eff28f90000-7eff28f91000 rwxp 00020000 08:03 1452451 /lib/x86_64-linux-gnu/ld-2.13.so
07:22:56 <shachaf> 7eff28d68000-7eff28d70000 rwxp 00067000 08:03 658337 /usr/lib/x86_64-linux-gnu/libgmp.so.10.0.5
07:24:44 <shachaf> How do I figure this out from the ELF file?
07:24:48 <shachaf> GNU_STACK 0x0000000000000000 0x0000000000000000 0x0000000000000000
07:24:48 <shachaf> 0x0000000000000000 0x0000000000000000 RWE 8
07:26:05 <olsner> it should be documented somewhere, maybe the linker option to set the stack non/executable says how it does it
07:26:52 <shachaf> I'm looking at the flags GHC linked with.
07:28:35 <shachaf> Apparently GHC has some code to mark the stack non-executable.
08:01:22 -!- Nisstyre_ has joined.
08:02:07 -!- Nisstyre has quit (Disconnected by services).
08:02:16 -!- Nisstyre_ has changed nick to Nisstyre.
08:02:44 -!- Nisstyre-laptop has quit (Ping timeout: 240 seconds).
08:06:53 <Sgeo> <Sgeo> ) 0 = 1 o. o. 2 * i. 5
08:06:53 <Sgeo> <jconn> Sgeo: 1 0 0 0 0
08:10:35 <Sgeo> * for signum actually obeys tolerance rules, hmm
08:10:45 <Sgeo> <Sgeo> ) * 1 o. o. 2 * i. 5
08:10:46 <Sgeo> <jconn> Sgeo: 0 0 0 0 0
08:15:52 -!- Nisstyre has quit (Ping timeout: 276 seconds).
08:24:37 -!- Taneb has joined.
08:29:51 -!- Nisstyre has joined.
08:33:55 -!- Taneb has quit (Quit: Leaving).
09:01:34 -!- zzo38 has quit (Remote host closed the connection).
09:01:38 -!- carado has joined.
09:24:32 -!- carado has quit (Ping timeout: 246 seconds).
09:31:45 -!- epicmonkey has joined.
09:35:29 -!- ais523 has joined.
09:37:25 -!- copumpkin has quit (Ping timeout: 260 seconds).
09:38:50 -!- copumpkin has joined.
09:51:42 -!- ais523 has quit (Ping timeout: 276 seconds).
09:51:54 -!- Bike has quit (Quit: bye).
10:05:02 -!- glogbackup has quit (Ping timeout: 256 seconds).
10:05:09 <Sgeo> atriq Taneb I forget Fiora someone someonelese
10:12:23 -!- glogbackup has quit (Remote host closed the connection).
10:14:09 <fizzie> You should put that thing in fungot, that's clearly the most stable bot available.
10:14:09 <fungot> fizzie: i mean... what if it gets a 87% grade when i found a solution to the problem
10:15:44 <fizzie> ^def list ul (Taneb atriq Ngevd Fiora nortti Sgeo alot)S
10:15:52 <fungot> Taneb atriq Ngevd Fiora nortti Sgeo alot
10:21:49 -!- glogbackup has quit (Remote host closed the connection).
10:43:55 -!- DHeadshot has quit (Ping timeout: 260 seconds).
10:55:47 -!- azaq23 has joined.
10:55:56 -!- azaq23 has quit (Max SendQ exceeded).
10:56:27 -!- azaq23 has joined.
11:11:44 -!- monqy has quit (Quit: hello).
11:15:20 -!- glogbackup has joined.
11:20:05 -!- glogbackup has quit (Ping timeout: 260 seconds).
11:23:55 -!- copumpkin has quit (Ping timeout: 244 seconds).
11:24:25 -!- copumpkin has joined.
11:49:37 -!- epicmonkey has quit (Ping timeout: 245 seconds).
11:53:35 -!- Frooxius has quit (Ping timeout: 255 seconds).
12:08:08 -!- DHeadshot has joined.
12:09:11 -!- Phantom_Hoover has joined.
12:12:37 -!- oerjan has joined.
12:15:46 -!- Frooxius has joined.
12:18:16 <oerjan> good afternoon shachaf
12:18:43 <lambdabot> Local time for shachaf is Mon Jan 28 04:18:40 2013
12:21:03 <oerjan> ha ha you people keep faking that timezone thing
12:22:14 -!- DHeadshot has quit (Ping timeout: 252 seconds).
12:24:30 <fizzie> "^echo ^echo" is the best quine in the fungot-^echo-language.
12:24:48 <fungot> shachaf: then one could write ( loop-safety on) or ( x ( get-value) block) will expand to the same event at the same time
12:25:51 <shachaf> (DECLAIM '(OPTIMIZE (LOOP-SAFETY ON)))
12:25:55 <oerjan> sadly it's also the _only_ quine, i think.
12:26:11 <fungot> fizzie: there are a bazillion definitions per item with no discernable purpose and reside in a different order
12:26:30 <oerjan> x ++ x == "^echo " ++ x doesn't have a lot of solutions
12:26:32 <lambdabot> Local time for fungot is ITS NOT A FAKE
12:27:08 <oerjan> * x ++ " " ++ x == "^echo " ++ x doesn't have a lot of solutions
12:27:44 <shachaf> oerjan: Do empty programs count?
12:27:59 <shachaf> That's not an empty program.
12:28:17 <oerjan> shachaf: he said "in the fungot-^echo-language"
12:28:18 <fungot> oerjan: ais523 suggested it x-p you win!!! eheheheeheh trying to find a way to see where one can look at the number of
12:28:46 <shachaf> Do all programs in that language start with ^echo?
12:28:51 <shachaf> In that case I guess it's the only quine.
12:29:08 <fizzie> They all do start with "^echo ".
12:30:14 <shachaf> That's not a ^echo-program?
12:32:50 <fizzie> Hrm, for some reason I confused ^echo with one of the commands where the space is mandatory.
12:33:39 <oerjan> i think the space is handled by fungot's pre-bf parsing
12:33:39 <fungot> oerjan: remembered. i'll tell riastradh when he/ she
12:34:03 <oerjan> the one on input, that is
12:34:27 <fizzie> Yes, it's only the fixed commands that sometimes check also that there is a space.
12:34:32 <fungot> shachaf: like calories. a
12:34:36 <oerjan> so ^command<return> is equivalent to ^command<space><return>
12:34:36 <fungot> shachaf: actually i'll probably change and to and, although it would eventually segfault.
12:35:00 <oerjan> only fungot can make a segfault by changing and to and
12:35:00 <fungot> oerjan: whichever is closer to most module systems, is it
12:36:59 <fizzie> Everyone loves fungot like calories.
12:36:59 <fungot> fizzie: i have to, as the compiler is correct? maybe the implementations of " reverse"
12:37:22 <shachaf> i love fungot like calories. a
12:37:22 <fungot> shachaf: thanks. your new nick? hehe ( sometimes i make too long sentenced to avoid being eaten by a grue.
13:00:13 -!- epicmonkey has joined.
13:07:25 -!- sebbu has quit (Ping timeout: 260 seconds).
13:22:50 -!- Arc_Koen has joined.
13:33:39 <shachaf> Did elliott take my quote out of the topic?
13:35:22 <oerjan> unless you mean the lambda cube, in which case yes
13:35:35 <Jafet> Six simultaneous rotations of the lambda tesseract
13:40:40 <fizzie> Food for thought: NaNoWriMo requires 50000 words in order for something to qualify as a novel; W|A says average English word length is 5.1 characters; allowing for some whitespace and punctuation, 6.5*50000 = 325000 characters. Compare this with http://zem.fi/esostats/fig/activity_chars_20year.png -- since mid-2008, the channel seems to average around 1-2.5 novels a week.
13:41:40 <oerjan> the plot is horrendous, though
13:42:17 <fizzie> And the characters are, frankly, stupid and unbelievable.
13:42:29 <oerjan> except for fungot and zzo38, they'll obviously go on to start spinoff series
13:42:29 <fungot> oerjan: wiliki sure looks a lot like the inverse of f(n) a(n,n)? would it run the other schemes will do something with
13:43:35 <fizzie> Also, total characters over all time: 127,714,454 i.e. 393 novels, which again reminds me that I'd like to see the #esoteric bookshelf as a physical thing.
13:44:11 -!- copumpkin has quit (Ping timeout: 252 seconds).
13:44:42 -!- copumpkin has joined.
13:45:31 -!- augur has quit (Remote host closed the connection).
13:56:07 -!- Jafet has quit (Quit: Leaving.).
13:56:34 -!- Jafet has joined.
14:19:04 <elliott> fizzie: is #esoteric bookshelf just printed logs of #esoteric
14:20:17 <fizzie> Yes. Well, not just "printed", but bound to something for which the word "tome" is appropriate.
14:20:43 <fizzie> Possibly also "copied by a cloisterful of monks" as opposed to "printed".
14:22:43 <oerjan> i guess illumination is mandatory as well.
14:23:30 <elliott> fizzie: well here I was considering giving Lulu a bunch of money to make that dream a reality
14:23:40 <fizzie> http://www.flickr.com/photos/darrellg/6822994478/ I mean something like this maybe?
14:24:49 <elliott> I nominate fizzie to be the monk.
14:25:49 <shachaf> May all your monoids be easy!
14:31:10 -!- boily has joined.
14:45:13 <Jafet> Is the monq monqy?
14:46:27 -!- WeThePeople has joined.
14:47:00 -!- augur has joined.
14:54:02 -!- oerjan has quit (Quit: leaving).
15:18:57 -!- Lumpio- has quit (Ping timeout: 248 seconds).
15:24:35 -!- Taneb has joined.
15:44:15 -!- FreeFull has joined.
16:00:20 -!- glogbackup has joined.
16:02:45 -!- ais523_ has joined.
16:17:11 -!- glogbackup has quit (Remote host closed the connection).
16:44:48 <Phantom_Hoover> "In 2007, Niven, in conjunction with a group of science fiction writers known as SIGMA, led by Pournelle, began advising the U.S. Department of Homeland Security as to future trends affecting terror policy and other topics.[8] One of his suggestions as a member of SIGMA was that hospitals stem financial losses by spreading rumors in Spanish within the Latino community that emergency rooms are killing patients in order to
16:44:48 <Phantom_Hoover> harvest their organs for transplants, in order to reduce illegal immigrants' use of emergency rooms.[9]"
16:44:53 <Phantom_Hoover> larry niven.....................................................
16:46:55 -!- desty has joined.
16:47:00 <elliott> According to author Michael Moorcock, in 1967 Niven was among those Science Fiction Writers of America members who voiced opposition to the Vietnam War.[5] However, in 1968 Niven's name appeared in a pro-war ad in Galaxy Science Fiction magazine.[6]
16:47:04 <elliott> Niven was an adviser to Ronald Reagan on the creation of the Strategic Defense Initiative anti missile policy, as covered in the BBC documentary Pandora's Box by Adam Curtis.[7]
16:47:11 <elliott> hmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmm
16:48:16 <Phantom_Hoover> maybe he got sick of the government asking him to join in with their idiotic initiatives and decided to take the piss
17:06:17 <pikhq> That sounds like him.
17:10:14 -!- copumpkin has quit (Ping timeout: 240 seconds).
17:10:44 -!- copumpkin has joined.
17:27:03 <Taneb> Wow, I just rage-parted a channel
17:28:02 <fizzie> Rage against the channel.
17:29:27 -!- sebbu has joined.
17:30:44 <Taneb> There were two people who differed on opinions regarding GNU very loudly
17:32:46 -!- Taneb has changed nick to Taneb|Away.
17:37:40 <fizzie> http://zem.fi/esostats/charfreq_brainfuck.html it's probably partially BFJoust days?
17:41:49 <quintopia> fizzie: then do a search for "bfjoust" and overlay it
17:45:00 <fizzie> There are so many possible THINGS. But I suppose I'll add a straight-forward keyword-frequency thing similar to those charfreq things. Though I don't know what else than "bfjoust" to include.
17:45:46 <quintopia> case insensitive "bfjoust" should catch just about every day that it was discussed
17:46:13 <fizzie> Yes, but, I mean, maybe there are other things in the world in addition to bfjoust too.
17:46:27 <fizzie> I suppose I can run some kind of a TF-IDF keyword extraction on my dataset, though.
17:51:56 <fizzie> I'd rerun the people-statistics bit, but my computar is so noisy when it's doing things. :/
17:52:04 <augur> who was talking about other me?
17:52:24 <fizzie> The other you ended up in my list of "important" people for some plots plots plots.
17:53:14 <quintopia> is the list of important people the set of all people who posted more than x times in this channel?
17:53:49 <fizzie> Where 'x' was somewhere between two and three thousand.
17:53:56 <augur> Phantom_Hoover: where did fizzie mention my other username? :|
17:54:36 <fizzie> http://zem.fi/esostats/ "individual activity charts"
17:56:01 <fizzie> They will be merged after the next recomputamation.
17:57:23 <augur> ive been pretty inactive in here lately
18:04:38 <boily> meh. I'm not important yet.
18:05:52 -!- sirdancealot7 has quit (Ping timeout: 245 seconds).
18:08:28 <fizzie> boily: You've been made important for the next run. (But you should indeed chat more; your linecount in my logs is just 505.)
18:09:02 -!- sirdancealot7 has joined.
18:09:36 <boily> I need to chat more indeed.
18:09:56 -!- ais523_ has quit (Quit: Page closed).
18:10:25 <boily> quintopia: nope. still canadaing, even if I think its existence hasn't been yet formalized, let alone proved.
18:10:43 <quintopia> lets all go to canada and chat boily
18:10:43 <boily> I wonder who's the nearest member from me in this channel.
18:11:38 <boily> definitely nearer.
18:11:51 <boily> BC is like 5000 km away from me.
18:12:21 <boily> say, does anyone here have experience with OpenGL ES 2.0?
18:12:57 <quintopia> no. i like esolangs, but that one is just too esoteric
18:13:54 -!- surma has joined.
18:16:21 <surma> Hey guys, I need a language where every combination of <whatever> is a valid and executable program. Does anything come to mind? I was thinking of SUBLEQ (or any OISC for that matter), but I was hoping you guys would know something more abstract
18:18:06 <desty> it's like that in many genetic programming implementations, where everything must be valid and defined (even if meaningless or NOP in most cases) since the code is generated by random process
18:18:40 <desty> so, for example, the PushGP language probably allows any input?
18:19:39 <elliott> boily: your use of the fake letter "é" proves you must be french
18:24:09 -!- Bike has joined.
18:26:32 <fizzie> surma: http://esolangs.org/wiki/Jot is one example. (Every string in [01]* is a valid Jot program.)
18:27:23 <boily> elliott: sincé whén is é a faké léttér?
18:30:19 <surma> desty: Ha, the genetic programming application slipped my mind. Makes a lot of sense to look there.
18:33:12 <fizzie> TF-IDF top-16 terms for 2005-03-05: "brainfuck he bf really org than languages possible name tutorial programming hehe ah how bfasm p" ... and for 2005-03-06: "esoapi pesoix easel h will dialect cpressey api up has they level bos o check switch"; well look at that, it actually does sort of hint as to what was the topic of discussion that day. (Disclaimer: IDF computed over a total of three ...
18:33:18 <fizzie> ... days; larger thing now in progress.)
18:34:39 <desty> "brainfuck he bf really org" <- sure it isn't from some S&M chatroom? :P
18:34:46 <boily> serendipitous reddit link: http://www.primaryobjects.com/CMS/Article149.aspx
18:35:34 <elliott> Advantages of Brainf-ck as an AI Programming Language
18:35:40 <elliott> this is certainly something
18:36:20 -!- Taneb|Away has changed nick to Taneb.
18:36:23 <Gregor> http://blog.bitbucket.org/2013/01/28/signup-to-bitbucket-via-google-or-github/ You can now sign in to BitBucket with your GitHub account.
18:36:57 <Bike> why are people so obsessed with self-modifying code -> AI...
18:37:40 <fizzie> Aw dagnabbit, I forgot to filter out initial nick-attributions, this whole output is now just mostly list of people who've been talked to -- http://sprunge.us/iTii -- will have to redo.
18:37:45 <elliott> I see they don't even give any examples of evolving programs to compute a function... just constant output.
18:37:50 <boily> from experience, genetic algos are slow, computation intensive, and they don't work.
18:38:01 <desty> boily: yes, very slow and inefficient
18:38:07 <elliott> fizzie: Might want to ignore lines starting with ! and ` too.
18:38:13 <elliott> Except I guess !bfjoust would be Relevant(tm).
18:38:49 <desty> I wrote a genetic programming engine to try to evolve strings of input to vim (for vimgolf)... it was awful :)
18:39:06 <fizzie> Maybe the command names can stay. But people are boring, I'll get rid of them.
18:39:27 <elliott> "ehird compile cube shit" "sexy ehird bitch" Um, okay.
18:40:01 <fizzie> Hey, that's just what the machine tells me. :p
18:40:25 <kmc> surma: brainfuck and the uncountably many brainfuck derivatives have that property
18:40:47 <kmc> most bytes (or is it characters?) are no-ops
18:40:48 <elliott> kmc: ] is an invalid bf program
18:40:58 <elliott> < too, under most interpretations
18:41:00 <kmc> ok fair enough
18:41:48 <elliott> fizzie: Can you pad every Jot program out to be byte-sized?
18:41:58 <elliott> Or are there hypothetically functions which can only be expressed as Jot programs of a non-integral number of bytes?
18:44:27 <fizzie> I'm not qualified to tell, but I'd hazard a guess you can always at least pad to some multiple of 8.
18:48:32 <fizzie> http://sprunge.us/aBZV <- nick-attribution-filtered tfidf output. Apparently there's still quite a few nicks, possibly since people in fact mention other people also inside their comments.
18:48:45 <fizzie> 2012-12-26's topic: "rape switzerland"
18:55:28 <Taneb> fizzie, does the people one combine Taneb, Ngevd, and atriq?
18:55:45 <coppro> fizzie: how do you feel to have been in this channel for over a decade
18:58:35 <fizzie> coppro: I don't think I really "get" it.
18:59:05 <coppro> get what? this channel?
18:59:10 -!- epicmonkey has quit (Ping timeout: 260 seconds).
18:59:59 -!- desty has left ("Leaving").
19:00:56 <kmc> 'Shipment of 18 human heads found at Chicago's O'Hare airport'
19:01:03 <coppro> fizzie: http://codu.org/logs/_esoteric/2003-01-19.txt
19:01:07 <coppro> that was more than 10 years ago
19:01:26 <Bike> kmc: destined for what?
19:01:49 <elliott> coppro: fizzie has been here since before 2003-01-19.
19:02:11 <coppro> but that's how far back the logs go
19:02:31 <coppro> so I dispute your claim and demand evidence otherwise
19:02:33 <elliott> Actually they go back to http://codu.org/logs/log/_esoteric/2003-01-18. But it's interesting that fizzie's private logs that are in the rsync thing aren't in the web interface.
19:03:33 <kmc> http://www.borderagencyscotland.com/
19:06:41 -!- AnotherTest has joined.
19:09:32 <Taneb> Esolang the movie staring Elijah Wood as Elliott Hird and Viggo Mortensen as rjan Johansen.
19:10:34 <Taneb> coppro, who would play you?
19:10:40 <coppro> Taneb: Andy Serkis obv
19:10:56 <Taneb> I didn't mean to restrict it to LOTR actors
19:11:08 <coppro> you can be played by Anne Hathaway
19:11:11 <Taneb> I think I'd be played by Zachary Quinto
19:12:00 <Gregor> fizzie and oerjan are the olde hats.
19:12:24 <elliott> Taneb: sorry but we have to get Dijkstra to play oerjan.
19:12:46 <Taneb> ...isn't he kind of dead?
19:13:03 <Gregor> Not only is he kind of dead, he's entirely dead.
19:13:03 <elliott> Use necromancy if you have to!
19:13:19 <boily> I like necromancy. all those shiny zombies :D
19:14:09 <Bike> no need for necromancy, you could just prop up his rotten corpse like a puppet
19:14:25 <Bike> cost effective
19:15:08 <Taneb> AnotherTest, should I...
19:15:17 <Taneb> should I do a song and dance routine or something?
19:15:21 <Taneb> Is this a talent contest?
19:15:31 <Bike> only if you're prepared to have it uploaded to youtube
19:26:00 <FreeFull> Is there any number argument to genericTake, where an input of an infinite list will produce an infinite list
19:28:38 <kmc> if you define inductive natural numbers then you can define ω which is an infinite number
19:29:06 <kmc> as elliott said, ω = Succ ω
19:29:08 <kmc> elliott: ;P
19:30:11 <kmc> you could also just define a "numeric" type which is always infinite
19:30:26 <elliott> well then your fromInteger has to lie
19:30:31 <kmc> that's fine
19:31:07 <kmc> fuck tha police
19:31:11 <Bike> do you need a fromInteger in a Numeric type?
19:31:15 <lambdabot> class (Eq a, Show a) => Num a where
19:31:20 <elliott> Bike: behold the worst class in the universe
19:31:27 <elliott> Bike: it gets better though
19:31:29 <lambdabot> class (Num a, Ord a) => Real a where
19:31:34 <Bike> that's not better
19:31:38 <Bike> that's not better at all D:
19:31:59 <Bike> christ is there any good way to make numbers in programming remotely like they are in math
19:32:16 <elliott> abstract algebra is kinda unwieldy to do via typeclasses
19:32:28 <Bike> where's class Group huh
19:32:31 <Bike> where's class Monoid
19:32:39 <Bike> oh, there it is
19:32:40 <elliott> they are so easy, even typeclasses can do them
19:32:56 <Bike> still think it's weird that the operation is called "append"
19:33:04 <Bike> also is mconcat really just there for efficiency
19:33:22 <lambdabot> (>>=) :: forall a b. m a -> (a -> m b) -> m b
19:33:22 <lambdabot> (>>) :: forall a b. m a -> m b -> m b
19:33:28 <kmc> mconcat has a default definition, yeah
19:33:42 <Bike> oh, type classes don't have inheritance or nuthin, do they
19:34:14 <Bike> monads aren't monoids!
19:34:18 <Bike> where are the endofunctors
19:34:20 <Bike> is my life a lie?
19:34:28 <kmc> it's all obfuscated
19:34:29 <FreeFull> I think the default definition for mconcat is something like foldr mappend mempty
19:34:36 <Bike> so i would assume
19:35:14 <lambdabot> Source not found. It can only be attributed to human error.
19:35:26 <elliott> Bike: class Category cat => Monoid cat unit product mon where mempty :: unit `cat` mon; mappend :: product mon mon `cat` mon
19:35:33 <elliott> or, with infix syntax that is sadly no longer allowed:
19:35:49 <elliott> class Category (~>) => Monoid (~>) unit (**) mon where mempty :: unit ~> mon; mappend :: (mon ** mon) ~> mon
19:35:55 <kmc> THEY GOT RID OF THAT SYNTAX?!?
19:35:57 <elliott> now if you let (~>) = (->), unit = (), (**) = (,), you get
19:36:03 <elliott> mempty :: () -> mon; mappend :: (mon, mon) -> mon
19:36:07 <kmc> whyyyyyyyyyyyy
19:36:14 <elliott> if you let f ~> g = (forall a. f a -> g a) (<-- natural transformation)
19:36:24 <elliott> (**) = functor composition i.e. (f ** g) a = f (g a)
19:36:31 <elliott> mempty :: Identity a -> mon a
19:36:36 <elliott> mappend :: mon (mon a) -> mon a
19:36:45 <elliott> which is a Monad (return and join respectively), if mon is a Functor!
19:36:55 <elliott> now you actually need to wrap those latter definitions up into data types and stuff to get it work
19:37:03 <elliott> but with the PolyKinds extension you can actually define a generic Monoid in this way
19:37:10 <elliott> kmc: that syntax is now used for actual infix types, rather than type variables
19:37:17 <elliott> kmc: i.e. you can define data a * b = Product a b or whatever
19:37:20 <FreeFull> You know, the full name for football is associative football
19:37:26 <FreeFull> Does that mean football can be a monoid?
19:37:29 <kmc> so close yet so far
19:37:43 <elliott> kmc: not being able to use infix type variables is annoying, but having to prefix infix types with : is arguably worse
19:38:34 <FreeFull> Nevermind, it's actually association football
19:39:52 <Bike> imo football is a magma at best
19:41:21 <kmc> now i'm picturing a football variant with obstacles in the form of holes filled with molten rock
19:44:03 <fizzie> http://zem.fi/esostats/charfreq_brainfuck.html Aw, the word "bfjoust" doesn't correlate all that well with brainfuck characters. (Though that peak in 2009 followed by a longish bump for the rest of the year probably qualifies.) ((Also there are some other terms plotted in the index.))
19:53:13 <Bike> http://docs.python.org/3/c-api/init.html#threads kmc, is this part of why you were complaining about pythong threads?
19:54:46 <kmc> yes the Global Interpreter Lock is one of the problems
19:56:28 <Sgeo> I thought GIL just prevents good parallelism but not concurrency?
19:56:52 <kmc> did anyone say anything to the contrary?
19:57:07 <kmc> that said I don't think there are provisions for good concurrent I/O either
19:57:12 -!- augur has quit (Read error: Connection reset by peer).
19:57:22 <kmc> despite not supporting parallelism, CPython threads are OS threads and will block in individual system calls
19:57:31 <kmc> rather than having something like the GHC IO manager
19:57:36 -!- augur has joined.
19:57:44 * Sgeo personally cares about concurrency more than parallelism
19:57:59 <Sgeo> Although I guess being OS threads means annoying heavyweight threads
19:58:39 <kmc> the heaviness of OS threads is also sometimes overstated
19:58:49 -!- Taneb has changed nick to Taneb|Away.
19:59:01 <kmc> Linux can definitely support a whole lot of threads
19:59:07 <kmc> it depends on things like how much stack you allocate for each
19:59:31 <kmc> oh also some of it is carryover from 32-bit days
19:59:52 <kmc> because even if the actual stack memory is allocated on demand, you need to reserve address space for each stack
20:00:09 <kmc> which can be a problem when you only have 3GB of address space
20:00:16 -!- ogrom has joined.
20:00:36 * Sgeo is using a 32-bit Linux OS although the machine is 64-bit
20:00:41 <Sgeo> 64-bit but only has 2GB memory
20:00:47 <kmc> why 32-bit OS?
20:02:00 <Sgeo> Not sure, installed it 2010 I think. I guess I figured since I don't have enough memory to really make 64-bit relevant and since I had the impression that 64-bit OS would be incompatible with stuff
20:02:43 <kmc> ok well you don't need lots of physical memory to make a 64-bit address space useful
20:03:06 <AnotherTest> Sgeo: you can make most 32bit programs run on 64bit operating systems
20:03:16 <kmc> also the AMD64 architecture has lots of improvements over i386 that are unrelated to the word size change
20:03:43 <kmc> yeah the compatibility story is pretty good these days
20:03:51 <kmc> you can always install a 32-bit userland in a chroot, but it's rarely necessary
20:04:30 <elliott> well also if you only have 2 gigs of ram running a 64 bit OS will reduce your effective RAM
20:04:36 -!- augur has quit (Ping timeout: 245 seconds).
20:04:53 <kmc> because pointers are twice as big?
20:04:54 <kmc> that's true
20:05:17 <kmc> you could run a 64-bit kernel and a mostly-32-bit userland, but i think that's not generally that worthwhile on x86
20:05:35 <kmc> maybe once x32 is mainstream you would run a 64-bit kernel and a mostly-x32 userland
20:05:45 <kmc> elliott: well int is still 32 bits
20:06:43 <Sgeo> help I need to get my brain fully functional in less than an hour I didn't get enough sleep since I was reading about J and EVE all night
20:06:48 <kmc> i think having twice as many registers and convenient position-independent code beats the word size increase for many tasks
20:08:36 -!- epicmonkey has joined.
20:11:14 <boily> functional position-independent drugs?
20:13:06 <kmc> 32 bit architectures may have killed the idea of memory mapped files :(
20:13:22 <kmc> like, how many apps have some ad hoc caching layer where they read stuff from disk and keep it in memory for a while?
20:13:28 <kmc> wouldn't it be great to make the OS do that for you?
20:14:10 <kmc> i'm not sure the Kids These Days starting as programmers really appreciate what virtual memory can do
20:17:12 * kmc wonders what the mincore() syscall is good for
20:17:32 <AnotherTest> mroman: Nice brainfuck interpreter you got there.
20:20:10 <AnotherTest> I sort of like the language because it allows you to write small programs that do quite a lot
20:20:41 <AnotherTest> In that way, it reminds me a little bit of zetaplex (and its variants)
20:26:40 -!- oerjan has joined.
20:28:16 -!- augur has joined.
20:30:44 <tromp> where's the brainfuck interpreter?
20:31:05 <oerjan> <boily> quintopia: nope. still canadaing, even if I think its existence hasn't been yet formalized, let alone proved. <-- someone formalize it, please
20:31:07 <AnotherTest> ".""X"r~"-""\/^^{vvvv}c!!!-.256.%{vvvv}c!sa\/"r~"+""\/^^{vvvv}c!!!+.256.%{vvvv}c! sa\/"r~"[""{"r~"]""}{\/^^{vvvv}c!!!}w!"r~">""+."r~"<""-."r~"X""\/^^{vvvv}c!!!L[+] \/+]\/+]^^3\/.+1RAp^\/+]\/0RA1RA^^-]\/0RA\/"r~"\'\'1 128r@{vv0}m[0"\/.+pse!vvvv<-sh
20:31:46 <tromp> in what language is that?
20:33:08 <Sgeo> (Note: Does not actually look like J)
20:34:02 <boily> not enought punctuation marks to look like J.
20:34:53 <Sgeo> Why is J considered a productive language and Burlesque isn't?
20:35:00 <quintopia> what would constitute a rigorous proof of the existence of canada?
20:35:57 <boily> I'd have to submit a solid sample of a chunk of canada to a comittee.
20:36:17 <boily> but then, there's always the possibility that I'd fake that sample.
20:37:01 <boily> therefore, one of you will have to sacrifice him/her/itself to go to canada, and witness its existence.
20:37:19 <tromp> you have to express cancada in Zermelo Fraenkel Set theory
20:37:49 <boily> darn. I feared that I'd have to do that. can't I just handwave it?
20:38:20 <tromp> not sure if you need the axion of choice...
20:38:24 -!- Phantom_Hoover has quit (Ping timeout: 252 seconds).
20:39:08 <tromp> ZFC would do it; Zermelo Fraenkel with the axion of Canada
20:39:23 <quintopia> perhaps there is a consistent formal theory in which canada does not exist!
20:39:44 <quintopia> in which case the axiom of canada is necessary for a consistent theory that includes canada
20:40:05 -!- Phantom_Hoover has joined.
20:40:20 <quintopia> let's assume it doesn't exist and see what that implies
20:40:27 -!- Taneb|Away has changed nick to Taneb.
20:41:33 <boily> does the non-existence of canada implies provinces and territories don't exist too?
20:42:12 <quintopia> not necessarily. but they will probably be member states of the U.S.
20:42:44 <Gregor> The Yukon: Lone star state of the far northwest.
20:42:46 -!- DHeadshot has joined.
20:42:53 <ogrom> or the united states would be member states of canada, but canada would have some other name
20:43:03 <Gregor> Such as "The United States of America"
20:43:47 <Bike> the axiomatic independence of canada.
20:43:55 <boily> it's a long-standing open problem.
20:45:42 <oerjan> <elliott> fizzie: Can you pad every Jot program out to be byte-sized? <-- yes, any prefix representing I can be removed, and the article's conversion turns SKS into the 23-bit 11111110001110011111000 (i tried SKK first but that was even length)
20:46:24 <quintopia> why would one want jot to be byte sized? compression?
20:47:20 <olsner> hmm, surely someone must have already made a Brainfuck backend for LLVM?
20:47:22 <quintopia> oerjan: so that means you'd have to add 7 bytes to it that do nothing just to make the encoding work sometimes. so wasteful!
20:48:25 <quintopia> better to just have 1 extra byte that says X if only first X bits of the last byte are part of the jot program
20:48:51 <Sgeo> I thought what I'd do was, I'd pretend I was one of those deaf-mutes
20:49:09 <Sgeo> Hmm, got the quote wrong
20:49:24 <Sgeo> ...actually, I'm not entirely sure what the exact quote is
20:52:22 <Sgeo> Maybe I shouldn't have looked up the quote
20:52:30 <Sgeo> Spoilers for GitS really
20:52:49 -!- carado has joined.
20:52:59 -!- AnotherTest has quit (Quit: Leaving.).
20:53:10 <Sgeo> I am so nervous
20:53:12 <boily> it kinda has become a 'it was its sled', seeing that every other cosplayer has had the spinning face once on their head.
20:53:14 <Sgeo> And I didn't eat yet :(
20:53:36 <tromp> oerjan, how can you remove a prefix representing I without changing the meaning?
20:53:41 <Sgeo> boily, well, the logo itself isn't spoilery, but seeing a name associated with it presumably is
20:54:08 <Sgeo> ...even saying "a name" is spoilery, sorry everyone
20:54:20 <Sgeo> (At least I think it's spoilery)
20:54:33 * Sgeo is going to go briefly insane
20:54:50 <boily> what's in a name? a deaf-mute by any other name would spoiler as much :p
20:55:28 <tromp> you can remove 1 followed by a rep of I
20:55:40 <Sgeo> Although I don't know if it's really an "interview"
20:56:43 <boily> don't worry. the worst that can happen is... oh, you mentioned interview. anything is permitted.
20:56:57 <oerjan> tromp: no, you can remove the rep of I itself because the semantics of FG depends on the string G but only on the semantics of F, so if F has the same semantics as the empty string (i.e. I) then [FG] = [G]
20:57:06 -!- augur has quit (Remote host closed the connection).
20:57:41 -!- augur has joined.
20:57:59 -!- augur has quit (Read error: Connection reset by peer).
20:58:21 -!- augur has joined.
20:59:47 <tromp> the jot page says that [AB] = 1[A][B]
21:00:07 <oerjan> yes, so what you said is also true
21:01:43 <tromp> so code [A] is equivalent to code 1[I][A]
21:02:56 <tromp> and [I] could be [SKK]=1[SK][K] = 11[S][K][K]
21:03:17 <oerjan> tromp: that's not the _definition_ of Jot though
21:04:03 -!- DHeadshot has quit (Ping timeout: 252 seconds).
21:04:04 <tromp> no, it's the implied coding rules from CL to jot
21:04:24 <oerjan> argh http://idiom.ucsd.edu/~barker/Iota/ is a 404 _and_ has been a 404 for as long as wayback has archived it D:
21:06:48 <oerjan> that page was redirected
21:07:17 <tromp> yes, i see you're right too
21:08:57 <tromp> although it seems inconsistent
21:11:06 <tromp> it makes one wonder what is the shortest non empty jot program for identity
21:19:59 <tromp> no wait, it's simply 1
21:22:29 <tromp> isn't 1 ^xy. I (xy) ?
21:22:37 <oerjan> the representations on the article page are restricted to starting with 1, btw, because he wanted to have things give direct binary numbers by concatenation
21:23:12 <Sgeo> If an interviewer keeps saying "interesting" is that a good thing?
21:23:34 <kmc> it depends, that is sometimes just a meaningless filler word
21:23:38 <tromp> it's not good. but it is interesting
21:24:27 <Sgeo> help interviewer wants to see code for one of the projects I mentioned on resume. The code is horribly ugly
21:24:45 <kmc> do you have time to clean it up?
21:24:47 <oerjan> tromp: hm i guess S(KI) = I except that it requires two arguments
21:24:57 <Sgeo> kmc, I didn't think to ask for time to clean it up
21:24:59 <kmc> also what stage is this? phone interview?
21:25:07 <Sgeo> Well, phone conversation
21:25:12 <Sgeo> Not sure if it counts as an interview
21:25:47 <Bike> are you ircing while phone conversing
21:26:21 <Sgeo> No, I'm off phone
21:26:39 <Sgeo> I have IRCed while video chat sexing once
21:27:06 <Bike> were you ircing here? did you spread cooties in the channel?
21:27:32 <tromp> ok, so 1 = I only with eta equivalence, but 010 = I regardless, right?
21:28:15 <Sgeo> Another channel
21:28:42 * oerjan wonders how that [AB] = 1[A][B] rule can possibly be true
21:28:48 <boily> Sgeo: oh, you're still alive! that's interesting!
21:29:01 <tromp> i was wondering that myself...
21:29:09 <tromp> i just took it for granted
21:29:34 <Sgeo> ....I used bad language in the source code
21:33:03 <tromp> only in the comments?
21:33:04 <boily> public static void longLiveVisualBasicSix() ?
21:33:56 <tromp> fsck(that,sh*t); ?
21:34:04 <Sgeo> tromp, as a variable name for a variable only used in debugging
21:34:29 -!- zzo38 has joined.
21:35:14 <tromp> must be in the later stages of debugging, after frustration mounted
21:37:59 <Bike> wait, is tromp a curse
21:40:00 <boily> Bike: google translate isn't much of a help on that matter, sadly.
21:40:08 <kmc> wow another rails vulnerability: https://groups.google.com/forum/?fromgroups=#!topic/rubyonrails-security/1h2DR63ViGo
21:40:23 <Bike> I thought sgeo said ««tromp», as a variable name»
21:40:26 <kmc> 'The JSON Parsing code in Rails 2.3 and 3.0 support multiple parsing backends. One of the backends involves transforming the JSON into YAML"
21:40:45 <Bike> that sounds like a pretty weird way to parse, uh, anything
21:43:16 <Sgeo> sent the email
21:45:29 <Sgeo> I also mentioned some Haskell code I've written, but not sure if it's appropriate to send it along
21:45:33 <Sgeo> (It's about speed dating)
21:49:52 <olsner> Sgeo: your job interview was about speed dating?
21:50:22 <Sgeo> The actually not horrifically ugly code I've written last year was about speed dating
21:50:36 <oerjan> tromp: ah mystery solved. the CL -> Jot conversion satisfies a rather stronger property which makes it work, namely [F[A]] = [F][A]
21:51:38 <oerjan> so it's not enough to plug in an arbitrary Jot program in the 1[A][B] rule, it has to also satisfy that
21:52:26 <tromp> yes, i realized that [AB] = 1[A][B] failes for even simple cases like A=B=0
21:53:11 <tromp> and it doesn't help that [] is used in two ways:(
21:53:30 <oerjan> indeed. maybe i should disambiguate them
21:55:01 <tromp> use {} for encoding function
21:56:22 <tromp> not in [F[A]] = [F][A]
21:58:00 <tromp> i prefer self-delimiting languages :)
22:01:22 <Sgeo> You know what would be cool? A readable J
22:01:29 <tromp> did anyone write a brainfuck interpreter in Unlambda?
22:02:21 <oerjan> well i didn't, although i've pondered it a bit.
22:02:48 <oerjan> it has the usual "needs a complete character table to convert I/O" annoyance
22:03:19 <oerjan> which showed up in my unlambda in unlambda
22:03:47 <boily> elliott: readable, yes. its understandability still remains a completely different matter.
22:04:24 <oerjan> http://oerjan.nvg.org/esoteric/interpreter.unl
22:05:00 * Sgeo just made his first commit in years to PSOX
22:05:21 <Sgeo> http://trac.assembla.com/psox/changeset/98
22:06:06 -!- WeThePeople has quit (Quit: Leaving).
22:06:08 <tromp> to make J more readable, just expand the single letter names according to http://www.jsoftware.com/help/dictionary/vocabul.htm
22:06:45 <tromp> it has 3 operators devoted to taylor series alone:(
22:07:36 <Bike> why is that a :(?
22:07:59 <tromp> actually it has 1 and 2-char operators
22:08:01 <Bike> oh dang they have hypergeometric functions in the base, huh
22:08:02 <boily> tromp: 3 operators for taylor series? that's completely insane! and lovable!
22:08:31 <Bike> u T. n is the n-term Taylor approximation to the function u . " i love it
22:09:17 <Bike> wow, automatic differentiation too
22:09:51 <Sgeo> Wait, it has automatic differentiation?
22:09:57 <Sgeo> I thought it only had numeric and symbolic
22:10:05 <tromp> the only thing missing is the Kitchen Sink operator
22:10:13 <Sgeo> tromp, that's !:
22:10:27 <Bike> I'm just looking at http://www.jsoftware.com/help/dictionary/ddcapdot.htm
22:10:37 <Bike> only for polynomials and their inverses by default
22:10:38 <Sgeo> (Well, ok, that's Foreign, not Kitchen Sink)
22:10:47 <Sgeo> But it's for all the other stuff like file access
22:10:49 <tromp> i want a domestic kitchen sink
22:11:09 <Bike> er, polynomials plus exp, i guess
22:11:41 <tromp> Evoke Gerund sounds like a powerful spell
22:11:49 <Bike> jesus it has exponential generating functions
22:13:15 <tromp> the big question is: why do many alphabet letter go unused?
22:13:27 <Bike> imo needs more coptic
22:16:34 <oerjan> tromp: hm {SK} = 10 satisfies the conversion property
22:22:01 -!- ogrom has quit (Quit: Left).
22:25:07 <oerjan> ah of course, you can read things like a stack language from the right. 0 pushes S and K, 1 pops two elements combining them into one. when you have one element on the stack, the program you used to get there is a conversion to that element.
22:26:32 -!- SirCmpwn has quit (Excess Flood).
22:26:52 <oerjan> 00 -- S K S K, 11100 -- SKSK = K
22:27:26 <oerjan> 000 -- S K S K S K, 11111000 -- SKSKSK = KSK = S
22:28:11 <oerjan> looked at that way, it's not that mysterious how to find representations.
22:30:39 -!- SirCmpwn has joined.
22:30:39 <oerjan> tromp: oh and it means that {11010} = SK(SK) = I
22:32:04 <oerjan> as a conversion which can be used in the conversion rule
22:34:06 <oerjan> this is probably a context-free language of some sort, similar to parenthesis matching. actually translating 1 = ( and 0 = )) and then adding one more ( to the left turns it into an actual parenthesis matching problem
22:38:34 <oerjan> ok it must be CF since this can be trivially seen to be recognizable with a pushdown automaton
22:46:52 <lambdabot> class (Real a, Enum a) => Integral a where
22:46:52 <lambdabot> quot, rem, div, mod :: a -> a -> a
22:46:52 <lambdabot> quotRem, divMod :: a -> a -> (a,a)
22:47:12 <oerjan> i think Integral is not meant to have members not subsets of Integer
22:47:54 -!- Taneb has quit (Quit: Leaving).
22:48:54 <oerjan> <Bike> monads aren't monoids! <-- alas, kind mismatch
22:50:00 <oerjan> you _could_ make all monad value types monoids, but then you wouldn't be able to have any other monoids of form t a without overlappinginstances
22:50:34 <oerjan> oh and of course it would conflict with the instance for [a]
22:52:25 <Sgeo> I really would like this Transcriptic job (if it's telecommute)
22:53:28 <Sgeo> Going to watch some TV to take my mind off of having sent an employer the worst code I have ever written
22:53:55 <Sgeo> And by TV I mean Ghost in the Shell on Hulu
22:55:38 <tromp> oerjan, what is the conversion property expressed with {} ?
22:56:53 <tromp> so F is a string, and A a combinator?
23:01:04 -!- Lumpio- has joined.
23:03:19 -!- WeThePeople has joined.
23:03:28 <Sgeo> GitS is getting on my nerves
23:03:50 <Sgeo> "We suspect XYZ" "Do you have any evidence for that?" ... the topic of any reasoning for suspecting XYZ is dropped
23:03:55 <Sgeo> And no one complains
23:04:23 <Lumpio-> They can't waste time for things with no evidence!
23:05:08 <Sgeo> They don't drop XYZ, just the topic of why XYZ is suspected
23:05:49 <boily> "We suspect Canada" "Do you have any evidence for that?" "Not yet."
23:05:53 <tromp> oerjan, now all makes sense:)
23:09:24 -!- ogrom has joined.
23:10:11 -!- carado has quit (Ping timeout: 246 seconds).
23:14:23 <kmc> that sounds like TV news
23:14:50 <kmc> "Are teenagers inhaling bees for a cheap high?" "Well, are they?" "Beats us, but we got you to sit through the last commercial break"
23:27:31 <oerjan> it's the beginning of the great edinburgh zombie cataclysm.
23:28:30 <oerjan> that's not the zombies, that's their victims hth
23:28:56 <oerjan> or wait are you implying you're not in edinburgh any longer
23:34:15 <oerjan> oh well i guess it's not zombies then. maybe hooligans?