←2013-03-02 2013-03-03 2013-03-04→ ↑2013 ↑all
00:00:01 <oerjan> oh no
00:00:58 <oerjan> is there an equivalent to raw_input() that takes an entire file?
00:01:30 <Jafet> `run python -c 'print open("bin/welcome").readlines()'
00:01:32 <HackEgo> ​['#!/usr/bin/perl -w\n', 'if (defined($_=shift)) { s/ *$//; s/ +/ @ /g; exec "bin/@", $_ . " ? welcome"; } else { exec "bin/?", "welcome"; }\n']
00:01:46 <oerjan> ...i meant entire stdin
00:01:52 <Bike> oerjan: when I showed your colorize script to a python developer he screamed at me and rewrote it in horrror. Wanna see?
00:02:00 <oerjan> Bike: OKAY
00:02:13 <Bike> Now that's enthusiasm I want to see!
00:02:13 <oerjan> it's not quite mine, fizzie wrote the original
00:02:28 <oerjan> (share the blame)
00:02:46 <Jafet> `run welcome | python -c 'print open("/dev/stdin").read()'
00:02:51 <HackEgo> Welcome to the international hub for esoteric programming language design and deployment! For more information, check out our wiki: http://esolangs.org/wiki/Main_Page. (For the other kind of esoterica, try #esoteric on irc.dal.net.)
00:03:12 <oerjan> um and read takes all?
00:03:36 <Jafet> Sometimes
00:03:48 <oerjan> `run cat bin/welcome | python -c 'print open("/dev/stdin").read()'
00:03:52 <HackEgo> ​#!/usr/bin/perl -w \ if (defined($_=shift)) { s/ *$//; s/ +/ @ /g; exec "bin/@", $_ . " ? welcome"; } else { exec "bin/?", "welcome"; }
00:04:00 -!- Sanky has quit (*.net *.split).
00:04:01 -!- atehwa has quit (*.net *.split).
00:04:01 -!- Gregor has quit (*.net *.split).
00:04:01 -!- EgoBot has quit (*.net *.split).
00:04:01 -!- hogeyui has quit (*.net *.split).
00:04:02 <oerjan> i guess that works
00:04:06 <oerjan> oops
00:04:19 <hagb4rd> `run cat bin/welcome | whatever?
00:04:20 <Bike> http://pastie.org/private/k564m0ahv1xxuqaumcafg
00:04:23 -!- EgoBot has joined.
00:04:24 -!- hogeyui has joined.
00:04:27 <Bike> uses sys.stdin
00:04:31 -!- Sanky has joined.
00:04:33 -!- Sanky has quit (Excess Flood).
00:04:41 <Jafet> I like how the new ircd goes to great lengths to not tell you which server went down.
00:05:19 -!- DHeadshot has joined.
00:05:30 <oerjan> Bike: that's not equivalent at all
00:06:01 -!- Sanky has joined.
00:06:02 <Jafet> > comparing length "open('/dev/stdin')" "import sys;sys.stdin"
00:06:03 <lambdabot> LT
00:06:05 <Bike> oh no!
00:06:19 <Jafet> linux wins!!!
00:06:46 <Jafet> `lsb_release -a
00:06:48 <HackEgo> No LSB modules are available. \ Distributor ID:Debian \ Description:Debian GNU/Linux \ Release:n/a \ Codename:n/a
00:07:33 -!- AgonyLang has quit (Ping timeout: 245 seconds).
00:08:22 <oerjan> `run (echo '#!/usr/bin/env python'; echo 'print (lambda s: "".join(['(s[i].upper() if i%2==0 else s[i].lower()) for i in range(len(s)) ]))(open("/dev/stdin").read())') >bin/CaT; chmod a+x CaT
00:08:24 <HackEgo> bash: -c: line 0: syntax error near unexpected token `(' \ bash: -c: line 0: `(echo '#!/usr/bin/env python'; echo 'print (lambda s: "".join(['(s[i].upper() if i%2==0 else s[i].lower()) for i in range(len(s)) ]))(open("/dev/stdin").read())') >bin/CaT; chmod a+x CaT'
00:08:55 <oerjan> `run (echo '#!/usr/bin/env python'; echo 'print (lambda s: "".join([(s[i].upper() if i%2==0 else s[i].lower()) for i in range(len(s)) ]))(open("/dev/stdin").read())') >bin/CaT; chmod a+x CaT
00:08:59 <HackEgo> chmod: cannot access `CaT': No such file or directory
00:09:10 <oerjan> `chmod a+x bin/CaT
00:09:11 <HackEgo> chmod: missing operand after `a+x bin/CaT' \ Try `chmod --help' for more information.
00:09:14 <oerjan> `run chmod a+x bin/CaT
00:09:19 <HackEgo> No output.
00:09:28 <oerjan> `run welcome | CaT
00:09:29 -!- atehwa has joined.
00:09:30 <HackEgo> WeLcOmE To tHe iNtErNaTiOnAl hUb fOr eSoTeRiC PrOgRaMmInG LaNgUaGe dEsIgN AnD DePlOyMeNt! FoR MoRe iNfOrMaTiOn, ChEcK OuT OuR WiKi: HtTp://eSoLaNgS.OrG/WiKi/mAiN_PaGe. (fOr tHe oThEr kInD Of eSoTeRiCa, TrY #eSoTeRiC On iRc.dAl.nEt.)
00:09:41 <oerjan> `cat bin/WeLcOmE
00:09:43 <HackEgo> ​#!/bin/sh \ welcome $@ | python -c "print (lambda s: ''.join([ (s[i].upper() if i%2==0 else s[i].lower()) for i in range(len(s)) ]))(raw_input())"
00:10:20 <oerjan> `run sed -i 's/pyt.*/CaT/' bin/WeLcOmE
00:10:23 <HackEgo> No output.
00:10:34 <oerjan> `WeLcOmE atehwa
00:10:36 <HackEgo> AtEhWa: WeLcOmE To tHe iNtErNaTiOnAl hUb fOr eSoTeRiC PrOgRaMmInG LaNgUaGe dEsIgN AnD DePlOyMeNt! FoR MoRe iNfOrMaTiOn, ChEcK OuT OuR WiKi: HtTp://eSoLaNgS.OrG/WiKi/mAiN_PaGe. (fOr tHe oThEr kInD Of eSoTeRiCa, TrY #eSoTeRiC On iRc.dAl.nEt.)
00:11:14 <oerjan> `run echo "cantcode: ok what did you need help with?" | CaT
00:11:16 <HackEgo> CaNtCoDe: Ok wHaT DiD YoU NeEd hElP WiTh?
00:11:26 <hagb4rd> great
00:11:50 <hagb4rd> can you make a version for zalgo? :)
00:12:18 <oerjan> oh hm
00:12:28 <oerjan> i can't even _see_ zalgo.
00:12:42 <nooodl> what was wrong with old WeLcOmE
00:12:52 <hagb4rd> it's ok
00:12:53 <Bike> not modular enough
00:12:56 <cantcode> i was just trying if it would parse it
00:12:57 <Bike> haven't you heard of unix philosophy?
00:13:08 <oerjan> `run sed -i 's/$@/"$@"/' bin/WeLcOmE
00:13:13 <HackEgo> No output.
00:13:21 <Bike> `run relcome | CaT
00:13:22 <oerjan> `cat bin/WeLcOmE
00:13:24 <HackEgo> ​#!/bin/sh \ welcome "$@" | CaT
00:13:25 <HackEgo> wElCoMe tO ThE iNtErNaTiONaL hUb fOr eSoTeRiC PrOgRaMmInG LANgUAgE dEsIgn AnD DePlOyMeNt! fOr moRe iNfOrMatIoN, ChEcK OuT OUR wIkI: hTtP://EsolaNGS.OrG/WIkI/MaIn_PaGe. (foR tHe oTHEr kInD OF eSoTeRiCa, TrY #eSoTeRic oN iRC.DaL.nET.)
00:13:34 <Bike> See? Beautiful.
00:14:20 <cantcode> it looks like a gas leak in a puddle
00:15:18 -!- copumpkin has quit (Ping timeout: 240 seconds).
00:15:22 <nooga> who's that?
00:15:46 <oerjan> cantcode: HackEgo only takes commands starting with ` . some of the other bots respond to their name (hi fungot!)
00:15:46 <fungot> oerjan: he was now a city, and of the dots perhaps a half inch more. the augmented party now began to advance directly toward the invisible building, and had no idea what the curious image could be.
00:15:49 -!- copumpkin has joined.
00:16:32 <hagb4rd> `run ls bin
00:16:36 <HackEgo> ​? \ @ \ WELCOME \ addquote \ allquotes \ anonlog \ aseen \ botsnack \ bseen \ calc \ CaT \ colorize \ define \ delquote \ elist \ emmental \ emoclew \ emptylist \ etymology \ forget \ fortune \ frink \ fueue \ gaseen \ google \ h \ ?h \ h! \ hatesgeo \ ?hh \ hyfinate \ hyphenate.fi \ interp \ joustreport \ jousturl \ js \ json \ ka
00:16:39 <cantcode> .habla espanol?
00:17:00 <oerjan> not much spanish here, sorry
00:17:08 <cantcode> `oh
00:17:10 <HackEgo> ​/home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: oh: not found
00:17:34 <Bike> `run botsnack | colorize | CaT
00:17:37 <HackEgo> :-D
00:17:45 <hagb4rd> that's it
00:17:56 <oerjan> oonbotti: you are the only bot i remember who actually parses english, no?
00:17:56 <oonbotti> oerjan: Why do you think I am the only bot you remember who actually parses english, no?
00:17:57 <Bike> Indeed, it is that.
00:18:27 <oerjan> oonbotti: habla espanol?
00:18:27 <oonbotti> oerjan: Perhaps the answer lies within yourself?
00:18:38 <nooga> ¿shouldn't you use ¿ for spanish questions?
00:19:00 <nooga> anyway, do spaniards really use ¿ on the internet?
00:19:01 <oerjan> oonbotti: ¿habla español?
00:19:01 <oonbotti> oerjan: Why do you ask that?
00:19:06 <oerjan> nooga: MAYBE
00:19:08 <nooga> ¿
00:19:24 <nooga> ¿, huh, add this to one of our bots
00:19:31 <nooga> ¿ looks nice
00:19:40 <Bike> not usually, in my experience (do they use it)
00:19:59 <Bike> not that i know any actual spaniards, just spanish-speakers from the other half of the world
00:20:01 <zzo38> What's that? ELIZA?
00:20:04 <nooga> ¿que es una gasoliniera cerca de aqui?
00:20:07 <Bike> sure looks like eliza.
00:20:17 <oerjan> zzo38: oonbotti? i think so.
00:20:18 <Bike> nooga: irc.dal.net
00:20:30 <nooga> buen
00:21:02 <nooga> HAR HAR HAR
00:23:37 <hagb4rd> ¿donde esta la zapateria?
00:24:01 <Bike> We need a Parry in here.
00:24:25 <nooga> redio reloj
00:29:00 <Jafet> oonbotti: fungot
00:29:00 <oonbotti> Jafet: I see.
00:29:00 <fungot> Jafet: when i fnord hinted to others about my find, and this was their tragic homecoming. they had, and between each pair of anklets stretched a golden chain that held its fnord to a significance beyond the emotion which it excites and is. for the man who
00:30:36 * hagb4rd nervously jingles his chains
00:31:15 <Sgeo> wtf wat
00:31:18 <Sgeo> "Yes, the ratio of a perfect circle to its radius should be rational. The fact that it is irrational proves that either a. the circle is not perfect, or that b. the space plane the circle occupies is not perfectly flat and consistent."
00:31:56 * Bike imagines riding a space plane, scratching a circle into the window
00:32:12 <Phantom_Hoover> sounds like hagb4rd all right
00:32:42 <Bike> Sgeo: I hear Khinchin's constant is rational.
00:33:07 <hagb4rd> 100% cotton
00:33:17 * hagb4rd unleashed
00:34:17 <Phantom_Hoover> i can't believe it's still not known whether the euler-mascheroni constant is rational
00:34:58 <oerjan> i can't believe it's not maccheroni
00:35:25 <Bike> If you pick a random real number it's probably pretty hard to prove its irrationality.
00:38:09 <Taneb> Don't you need the axiom of choice to pick a random real number
00:39:29 <Bike> psh sets
00:40:59 <Phantom_Hoover> yes but let's face it, what can you prove about a random real number
00:41:20 <Bike> "it exists, maybe"
00:41:29 <Bike> depending on how constructivist you are??
00:42:02 <monqy> it's real
00:42:09 <Bike> Maybe it would be better to restrict it to definable numbers...
00:42:47 <oerjan> the definable numbers are countable and so don't have a uniform probability distribution
00:43:08 <oerjan> well, neither do reals for that matter
00:43:26 <oerjan> conclusion: random real numbers don't exist!
00:43:39 <Bike> oh nooooo
00:44:20 <Bike> I meant like, in [0,1] though.
00:47:11 <oerjan> OKAY
00:47:29 <oerjan> then the definable numbers still don't have a uniform one, hth
00:47:37 <hagb4rd> in fact, one of gods biggest problems might have been designing some unpredictability. otherways it eternatity would be so boring.
00:47:53 <Bike> :(
00:48:13 <madbr> it's it possible to do a turing complete language with only one bignum variable
00:48:22 <Phantom_Hoover> yes
00:48:27 <Phantom_Hoover> obviously
00:48:33 <Bike> psh who needs variables
00:48:52 <Phantom_Hoover> also what bike said, cf. the lambda calculus
00:48:58 <Taneb> madbr, you can encode lists of integers as an integer if you think in terms of product of primes
00:49:04 <Bike> lambda calculus has variables!
00:49:08 * Bike was actually thinking combinators
00:49:17 <Bike> Schönfinkel in da house
00:49:17 <oerjan> madbr: fractran
00:49:17 <Phantom_Hoover> Taneb, just treat it as a bitstream
00:49:19 <madbr> taneb: yeah but that's ok if you have like 2 or 3 variables
00:49:41 <Taneb> [1, 2, 3] -> 2^1 * 3^2 * 5^3 = 2250
00:50:26 <oerjan> which essentially does what Taneb says
00:50:29 <Taneb> madbr, you can do that for infinite variables because there are infinite primes
00:50:33 <Phantom_Hoover> madbr, we proved sumamoito tc essentially using only one variable
00:50:59 <Bike> You could also use a more reasonable code but lol math.
00:51:52 <Phantom_Hoover> (there were a few others because sumamoito has next to no facilities for actually manipulating that variable, but they wouldn't be needed if you had arithmetic builtins)
00:52:52 <Phantom_Hoover> augh
00:53:14 <Phantom_Hoover> the back of my chair went loose and now my neck hurts like hell
00:53:31 <Phantom_Hoover> why didn't i listen when they warned me about good posture
00:58:11 -!- Nisstyre-laptop has quit (Quit: Leaving).
01:05:41 -!- mekeor has joined.
01:06:04 -!- augur has joined.
01:06:44 -!- augur has quit (Read error: Connection reset by peer).
01:06:58 -!- augur has joined.
01:12:52 -!- nooga has quit (Ping timeout: 272 seconds).
01:18:41 -!- nooga has joined.
01:28:12 -!- Taneb has quit (Quit: Leaving).
01:31:46 -!- carado has quit (Ping timeout: 246 seconds).
01:42:30 -!- Gregor has joined.
01:45:57 <Jafet> A random real number isn't even random
01:47:46 -!- oerjan has quit (Quit: Ribbit).
01:49:38 -!- nooga has quit (Ping timeout: 252 seconds).
01:50:22 -!- oerjan has joined.
01:53:40 -!- nooodl has quit (Ping timeout: 248 seconds).
01:55:25 -!- cantcode has changed nick to banana_pee.
02:00:34 -!- nooodl has joined.
02:01:40 <hagb4rd> http://math.stackexchange.com/questions/155156/is-it-generally-accepted-that-if-you-throw-a-dart-at-a-number-line-you-will-neve
02:03:26 <madbr> yeah it's like countably infinite (rationals) vs uncountably infinite (irrationals)
02:03:29 <madbr> or something like that
02:04:04 <Fiora> if you throw a countably infinite number of darts, what's the probability at least one will hit a rational?
02:04:31 <madbr> heh that's a hard question
02:05:11 <Gregor> Sounds like the kind of problem that's fun to complicate unnecessarily.
02:05:16 <Fiora> it does XD
02:05:22 <Gregor> I like complicating the birthday "paradox" by mentioning that birthdays are not evenly distributed.
02:05:33 <pikhq> I am going to arbitrarily say the probability is epsilon.
02:05:35 <Gregor> "How strong is your throwing arm?"
02:05:42 <Gregor> "How far is the number line from the thrower?
02:05:42 <Phantom_Hoover> i like how, if you look at the wikipedia article on 'wedge', the actual article on real-life wedges is near the bottom
02:05:48 <Gregor> "How big are the darts?"
02:05:57 <Phantom_Hoover> <madbr> yeah it's like countably infinite (rationals) vs uncountably infinite (irrationals)
02:05:58 <Phantom_Hoover> <madbr> or something like that
02:06:05 <Fiora> I guess my intuition would say that on average one would hit?
02:06:07 <Phantom_Hoover> i think it's mostly measure
02:06:07 <Phantom_Hoover> s
02:06:13 <Fiora> but this seems like a case where intuition will be totally wrong
02:06:37 <Phantom_Hoover> but idk if you can have a measure space with both uncountable and countable sets with nonzero measure
02:06:38 <pikhq> Intuition hates cardinality.
02:06:47 <oerjan> <Fiora> if you throw a countably infinite number of darts, what's the probability at least one will hit a rational? <-- 0 hth
02:06:50 <nooodl> i'd say 0
02:06:58 <Phantom_Hoover> guys i'm not sure this is cardinality...
02:07:16 <Phantom_Hoover> oerjan, help
02:07:17 <nooodl> ugh why am i reawake
02:07:18 <madbr> essentially rationals and non-transcendental irrationals are countable so they have a surface area of 0
02:07:36 <Phantom_Hoover> Fiora, something something something countable products
02:07:44 <oerjan> Phantom_Hoover: sure you can, you can sum a discrete and a continuous measure easily
02:07:55 <Phantom_Hoover> right
02:09:03 <Bike> Man, I just want to know how hard proving something rational is.
02:09:13 <madbr> countably infinite number of darts, I don't think you'd hit anything
02:09:42 <madbr> you've got countable infinity number of points that can be hit
02:09:56 <oerjan> Phantom_Hoover: you can also split it up into a continuous and a discrete one
02:09:58 <madbr> and countably infinite number of darts that will be thrown
02:09:59 <hagb4rd> a few weeks ago i've heard another interesting analogy including darts, chances of hit, and all that improbable (but true) cicumstances leading straight to the possibility of existance. it's that there are two way of the perfect 'hit': one is aiming and hitting by chance.. the other is like throwing the dart somewhere and drowing the mark around the place it hits :>
02:10:08 <madbr> countable infinity * countable infinity = still countable infinity
02:10:10 <Phantom_Hoover> Bike, 7 hards
02:10:38 <Fiora> Bike: is it proving rational or transcendental that's hard?
02:10:42 <hagb4rd> *drawing
02:10:45 <Fiora> I thought rational wasjust proving it could be expressed as p/q
02:10:47 <Jafet> I know a guy on the street who sells perfect hits
02:11:04 <hagb4rd> waiting for my man26$
02:11:08 <hagb4rd> in my hand
02:11:19 <Gregor> hagb4rd: That sounds like the kind of analogy that a painfully stupid person might use as proof of God or divine design.
02:11:33 <hagb4rd> not really
02:11:37 <Bike> Fiora: well, we've proved less numbers transcendental, and all rationals are algebraic...
02:11:46 <hagb4rd> just some humoristic thoughts
02:11:51 <Fiora> Bike: yeah, that's what I thought
02:11:53 <Fiora> transcendental is the hard one
02:11:57 <oerjan> btw it's simply basic countable additivity of measures, an axiom. mu(A_1 \/ A_2 \/ ...) <= sum mu(A_i) = 0
02:12:14 <Bike> But we still don't know if Euler-Mascheroni or Khinchin are rational, let alone algebraic.
02:12:24 -!- nooodl has quit (Ping timeout: 272 seconds).
02:12:25 <oerjan> (ok a little more work to show it's subadditive when the sets are not disjoint.)
02:12:36 <Fiora> Bike: wow @_@
02:12:47 <Bike> Yeah, it's weird.
02:12:49 <Phantom_Hoover> Fiora, the difficulty is simply that proving /anything/ on an arbitrary real number is really hard
02:12:55 <Bike> Khinchin is a really weird constant anyway.
02:13:07 <Bike> And relevant! Since almost all real numbers conform to it.
02:13:33 <Fiora> Phantom_Hoover: well, it's not arbitrary, it's computable, right?
02:13:36 <Fiora> and computable reals are countable
02:13:58 <Phantom_Hoover> uncomputably countable, but yes
02:14:17 <Phantom_Hoover> otoh proving anything on an arbitrary computable number is harder in even messier ways
02:14:19 <Fiora> I meant like, countable as in countably infinite right
02:14:26 <Bike> The hell does that mean? Obviously it's not recursive but what's that have to do with cardinality
02:14:35 <Phantom_Hoover> nothing?
02:14:46 <Phantom_Hoover> i just thought it was really weird when i first realised
02:14:55 <Bike> well it's still r.e.
02:15:05 <Bike> I... think
02:15:54 <oerjan> <Fiora> Bike: is it proving rational or transcendental that's hard? <-- basically if a number isn't obviously one of those by construction, then discerning them is usually hard.
02:16:08 <Phantom_Hoover> yeah, you can just enumerate the set of programs in some languag
02:16:09 <Phantom_Hoover> e
02:16:40 <Phantom_Hoover> but discerning the ones that correspond to computable reals is isomorphic to the halting problem, of course
02:17:04 <Fiora> yeah, I guess it's like, proving things /not/ rational is hard <.<
02:17:46 <oerjan> and proving things rational is hard if you cannot find the particular rational they're equal to
02:17:59 <shachaf> That's why people like me are still roaming the streets.
02:18:12 <Fiora> searching. searching for the rational numbers hiding in the dark alleys
02:18:32 <Fiora> passing out illegal primes
02:18:32 <shachaf> Hmm, I was thinking along the lines of "no one can prove I'm not rational".
02:18:36 <shachaf> But that works too.
02:18:42 <Jafet> Does anyone really care whether khinchin's constant is irrational
02:19:06 <Phantom_Hoover> think of the implications, Jafet
02:19:12 <Phantom_Hoover> think of the implications
02:19:18 * Jafet thinks.
02:19:49 <shachaf> Not all is lost, though! At least the Euler-Macaroni constant is computable.
02:19:51 <Jafet> ( save the implications! )
02:20:28 <oerjan> also the euler-macarena constant is real
02:20:37 <zzo38> Do you mean Euler-Mascheroni?
02:20:38 <Jafet> Does a program that computes the euler-macaroni constant consist of spaghetti code?
02:21:02 <zzo38> Wikipedia doesn't have Euler-Macaroni.
02:21:30 <shachaf> zzo38: If I want to call it Euler-Macaroni, then by FSM, I will!
02:21:32 <shachaf> No one can stop me!
02:22:05 <oerjan> shachaf: the euler-mascara constant isn't any random person's to name
02:22:25 <zzo38> shachaf: I am not trying to stop you.
02:22:25 <Phantom_Hoover> istr my analysis notes saying we only know the euler-mechanic constant to 20 digits
02:22:38 <Phantom_Hoover> which seems to be very, very wrong
02:22:40 <Bike> Jafet: "it would be really weird"
02:22:56 <Fiora> that was true in... 1809 or so?
02:22:57 <hagb4rd> your analysis notes?
02:23:02 <shachaf> zzo38: I see right through your Euler-Masquerade!
02:23:04 <oerjan> Phantom_Hoover: the wikipedia introduction has 50...
02:23:10 <Bike> hagb4rd: notes from a class on analysis
02:23:16 <hagb4rd> sieged by ignorant misanthropes spitting words
02:23:18 <Fiora> it says we know 28,844,489,545 as of 2009? XD
02:23:22 <hagb4rd> writing songs that voices never share
02:23:38 <Phantom_Hoover> you have ruined that song for me forever
02:23:39 <Bike> Fiora: eh, what's nine orders of magnitude between friends
02:23:44 <Jafet> Phantom_Hoover was probably thinking about bruno's constant
02:24:18 <Fiora> bruno's constant?
02:24:24 <zzo38> shachaf: Do you like to see things that aren't necessarily there?
02:25:05 <shachaf> zzo38: Like a hippopotamus?
02:25:14 <shachaf> zzo38: Oh, are you talking about modal logic here?
02:25:44 <oerjan> in that case, have you ever seen a thing that was necessarily there?
02:25:44 <shachaf> ¬□there, or something?
02:26:05 <shachaf> I.e. ◇¬there?
02:26:33 <zzo38> No, I mean you.
02:27:01 <shachaf> You mean me, and I mean modal logic.
02:27:07 <shachaf> And surely meaning is transitive?
02:27:46 -!- monqy has quit (Quit: hello).
02:28:22 <shachaf> hello to you too, monqy
02:28:24 <oerjan> itym transcendental
02:29:29 <shachaf> zzo38: You should invent a crazy new kind of modal logic that uses crazy new symbols.
02:29:38 <shachaf> Maybe things like 25C9 FISHEYE [◉]
02:29:38 <shachaf> 25CA LOZENGE [◊]
02:31:17 <zzo38> Fisheye logic?
02:31:24 <Bike> Maybe logic for optics.
02:31:48 <shachaf> ◯ ○
02:31:53 <shachaf> See the difference between those two?
02:32:03 <shachaf> One of them is LARGE
02:32:06 <hagb4rd> size matters
02:32:26 -!- Jon1 has joined.
02:32:49 <shachaf> hagb4rd: Have you ever said anything useful?
02:33:24 <Phantom_Hoover> it's yet to be observed
02:33:27 <Phantom_Hoover> `welcome Jon1
02:33:36 <HackEgo> Jon1: Welcome to the international hub for esoteric programming language design and deployment! For more information, check out our wiki: http://esolangs.org/wiki/Main_Page. (For the other kind of esoterica, try #esoteric on irc.dal.net.)
02:34:50 <hagb4rd> shachaf: have ever made someone happy?
02:34:50 <Jafet> `relcome Phantom_Hoover
02:34:55 <HackEgo> Phantom_Hoover: Welcome to the international hub for esoteric programming language design and deployment! For more information, check out our wiki: http://esolangs.org/wiki/Main_Page. (For the other kind of esoterica, try #esoteric on irc.dal.net.)
02:35:23 <Phantom_Hoover> hagb4rd, on this channel, at leasy, he has a considerable lead on you.
02:35:27 <Phantom_Hoover> *at least
02:35:38 -!- Jon1 has left.
02:35:51 <hagb4rd> yes. i follow.
02:35:58 <hagb4rd> you lead
02:36:11 <hagb4rd> hope you have an idea of the next goal to achieve
02:36:28 <Phantom_Hoover> tell me, hagb4rd, do you actually think all the shit that comes out of your mouth is deep, or are you just posturing
02:36:37 <hagb4rd> cmon phanti
02:37:00 <hagb4rd> have a cake!
02:37:32 <hagb4rd> shit that comes out of my mouth
02:37:37 <hagb4rd> my dear
02:37:38 <madbr> how good are compilers at optimizing loop test conditions?
02:38:10 <shachaf> 438 points good.
02:38:58 <Fiora> um, like, what specifically about them?
02:40:16 <Fiora> like, trying to count down to avoid cmp at the end and stuff?
02:41:10 <shachaf> whoa,dude, look at all these TLDs: http://www.iana.org/domains/special
02:41:35 <madbr> no more like guessing that your routine is going to loop like 8 times in one go
02:42:01 <Fiora> you mean, unrolling?
02:42:28 <shachaf> The easy way to figure out how good a compiler is at optimizing a specific piece of code is to compile that piece of code and see what it does.
02:43:40 <madbr> mhm
02:44:02 <shachaf> http://דוגמה.טעסט/
02:44:19 <madbr> wondering about this because I'm working on a, uh, fairly interesting CPU arch
02:44:28 <Fiora> which one :o
02:44:58 <zzo38> Do you have any better documentation of Yamaha OPM?
02:45:09 <Bike> ooh, devanagari?
02:45:28 <madbr> the one where iteration 1 of a loop is done on core 1, iteration 2 is on core 2, iteration 3 is on core 3, iteration 4 is on core 4, iteration 5 is on core 5, iteration 6 is on core 6, iteration 7 is on core 7, iteration 8 is on core 8, iteration 9 is on core 1, iteration 10 is on core 2...
02:45:34 <Fiora> wait, this is a real cpu arch?
02:45:52 <madbr> not yet
02:45:53 <Jafet> "Powered by MEDIAWIKI"
02:45:55 <Fiora> .... oh
02:46:00 <madbr> but it's definitely doable
02:46:02 <Fiora> um. I don't think you can meaningfully ask
02:46:09 <Fiora> what a compiler does on an arch that doesn't exist <.<
02:46:12 <Fiora> >.>
02:46:27 <Jafet> Well, you can, unless the compiler also does not exist
02:46:35 <Jafet> Then it is less meaningful
02:47:14 <shachaf> I suppose it was a more general "what sorts of loopy behaviors are compilers good at recognizing?" sort of question.
02:47:17 <madbr> zzo : hmm... I remember being told that if you try to change synth params on the fly, OPM (or was it OPN2?) can crash
02:47:18 <Bike> well guessing at loop invariants is an optimization thing dating from the 60s
02:47:21 <Jafet> If your CPU has eight cores and fully coherent shared cache, you are doomed to fail in any case.
02:47:27 <Bike> or 50s
02:48:15 <shachaf> 50s = 60s
02:48:16 <madbr> jafet : it's more like 8-way SSE / 8-issue superscalar
02:48:26 <madbr> might cut it down to 4-way
02:48:30 <Fiora> 8-issue superscalar already exists, I think
02:48:33 <Fiora> there's some TI DSPs that do that
02:48:58 <madbr> the nice thing is that the design will be a lot less complicated than some of those
02:49:19 <Fiora> um, are you sure? some ofthose DSPs use less than a watt and are miniscule
02:49:33 <Fiora> I mean, like, they're not complicated
02:49:56 <madbr> yeah and you have to write for them in ASM?
02:50:06 <Fiora> I don't think so...
02:50:22 <Fiora> http://www.ti.com/lsds/ti/dsp/c6000_dsp/c64x/products.page
02:50:53 <Fiora> http://www.ti.com/lit/ds/sprs371f/sprs371f.pdf
02:52:16 <zzo38> madbr: Yes, I think I read something about such things doing so too, but still, I have not found any good documentation for OPM.
02:52:41 <madbr> fiora: hmm, that looks fairly close to what the crusoe did
02:53:11 <madbr> same kind of 8 unit pre-issued super-superscalar vliw design
02:54:00 -!- Arc_Koen has quit (Quit: Arc_Koen).
02:54:55 <madbr> fiora : that's a DSP
02:55:04 <madbr> essentially 2 cores bundled together
02:55:19 <madbr> can you really write efficient code for that in C++? :D
02:55:51 <Fiora> I don't think it's two cores?
02:56:05 <Fiora> http://en.wikipedia.org/wiki/TMS320C64x doesn't say anything about multicore at least
02:56:18 <Fiora> oh, I see. the two sets of functional units.
02:56:50 <madbr> yeah so it's not multicore but it's two units
02:57:52 <Fiora> http://stackoverflow.com/questions/1453718/ti-dsp-programming-is-c-fast-enough-or-do-i-need-an-assembler I found this I guess
02:59:17 <madbr> yeah so you probably have to use intrinsics
02:59:33 <Fiora> I think the intrinsics are just for the built-in SIMD stuff
02:59:38 <Fiora> like the 2x16-bit and 4x8-bit?
02:59:48 <madbr> oh
02:59:54 <Fiora> since, like, you can't really write that in C
03:00:04 <madbr> well anyways
03:00:17 <madbr> the design I have beats that
03:00:22 <madbr> it can vectorize any loop
03:00:48 <madbr> as long as you delay the write operations to the next cycle
03:01:50 <madbr> which is why it has to have special registers for data feedback from iteration to iteration
03:02:08 <madbr> and some opcodes to "lock" memory addresses until the real writeback
03:02:23 <madbr> but other than that it's just a normal RISC
03:05:45 <madbr> like, you only even just need a decoder for the first core
03:06:09 <madbr> and just use delayed versions of the same instructions next cycle on the next core
03:07:52 <madbr> the point is to get performance similar to crazy superscalar RISCs
03:08:06 <madbr> but for much less complexity
03:08:23 <madbr> and without the severe restrictions on datapaths etc you see in DSPs
03:15:14 -!- Phantom_Hoover has quit (Remote host closed the connection).
03:16:45 <hagb4rd> this sounds interesting. would you mind to share a look on your solution/code/paper? (as long as it's not finished and sold/open source)
03:18:47 -!- oerjan has quit (Ping timeout: 240 seconds).
03:18:54 -!- oerjan has joined.
03:23:21 -!- ais523 has joined.
03:23:55 -!- tswett_ has joined.
03:29:00 -!- GeoffreyVanSchau has joined.
03:30:41 -!- tswett has quit (*.net *.split).
03:30:41 -!- myndzi has quit (*.net *.split).
03:38:03 -!- tswett_ has changed nick to tswett.
03:43:36 -!- GeoffreyVanSchau has quit (Quit: Leaving).
03:55:17 -!- myndzi has joined.
04:07:46 -!- mekeor has quit (Quit: ERC Version 5.3 (IRC client for Emacs)).
04:24:20 -!- oerjan has quit (Quit: leaving).
04:38:44 <madbr> back from supper
04:40:33 <madbr> hagb4rd : I have a little bit of what code would look like on it
04:41:26 <madbr> http://pastebin.com/eCMNNxiF
04:42:22 <madbr> doesn't use feedback registers or anti-alias memory operations tho (so it would probably have to be written as ASM or with intrinsics)
05:04:10 -!- GOMADWarrior has quit (Ping timeout: 260 seconds).
05:05:34 -!- TeruFSX has quit (Read error: Connection reset by peer).
05:06:22 -!- TeruFSX has joined.
05:11:15 -!- monqy has joined.
05:16:16 -!- Nisstyre-laptop has joined.
05:19:19 -!- dessos_ has joined.
05:20:16 -!- ais523_ has joined.
05:20:54 -!- monqy_ has joined.
05:21:34 -!- monqy has quit (Disconnected by services).
05:21:39 -!- monqy_ has changed nick to monqy.
05:24:17 -!- Gracenotes_ has joined.
05:24:54 -!- kallisti_ has joined.
05:26:33 -!- quintopi1 has joined.
05:27:03 -!- dessos has quit (Read error: Connection reset by peer).
05:27:03 -!- ais523 has quit (Ping timeout: 255 seconds).
05:27:32 -!- Gracenotes has quit (Ping timeout: 255 seconds).
05:27:32 -!- kallisti has quit (Ping timeout: 255 seconds).
05:27:32 -!- quintopia has quit (Ping timeout: 255 seconds).
05:28:02 -!- ais523_ has changed nick to ais523.
05:43:16 -!- Sanky has quit (Ping timeout: 245 seconds).
05:43:34 -!- Sanky has joined.
05:43:36 -!- Sanky has quit (Excess Flood).
05:44:05 -!- Sanky has joined.
05:48:28 -!- ssue_ has joined.
05:52:08 -!- SirCmpwn_ has joined.
05:52:34 -!- ssue has quit (Ping timeout: 255 seconds).
05:52:34 -!- glogbackup has quit (Ping timeout: 255 seconds).
05:52:36 -!- monqy has quit (*.net *.split).
05:52:36 -!- Nisstyre-laptop has quit (*.net *.split).
05:52:36 -!- surma has quit (*.net *.split).
05:52:36 -!- oonbotti has quit (*.net *.split).
05:52:36 -!- kmc has quit (*.net *.split).
05:52:36 -!- SirCmpwn has quit (Ping timeout: 255 seconds).
05:52:39 -!- ssue_ has quit (Changing host).
05:52:39 -!- ssue_ has joined.
05:52:40 -!- SirCmpwn_ has quit (Changing host).
05:52:40 -!- SirCmpwn_ has joined.
05:52:42 -!- ssue_ has changed nick to ssue.
05:52:55 -!- SirCmpwn_ has changed nick to SirCmpwn.
05:56:53 -!- monqy has joined.
05:56:53 -!- Nisstyre-laptop has joined.
05:56:53 -!- surma has joined.
05:56:53 -!- oonbotti has joined.
05:56:53 -!- kmc has joined.
05:57:51 -!- monqy has quit (Quit: hello).
06:07:41 -!- ogrom has joined.
06:16:29 <Sgeo> `ralist
06:16:30 <HackEgo> ​/home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: ralist: not found
06:17:09 <Bike> nooooo
06:18:35 <pikhq> SirCmpwn: Hmm. Where in CO are ye?
06:19:26 -!- monqy has joined.
06:21:07 -!- dessos has joined.
06:21:19 <shachaf> Sgeo: Red Alert 4 is coming out?
06:21:49 <Sgeo> Sam Hughes posted a new story in Ra
06:21:54 <Sgeo> http://qntm.org/ra
06:23:30 <Bike> `qlnitsmt
06:23:32 <HackEgo> ​/home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: qlnitsmt: not found
06:23:49 -!- dessos_ has quit (Ping timeout: 256 seconds).
06:26:10 -!- dessos_ has joined.
06:26:15 -!- dessos has quit (Ping timeout: 245 seconds).
06:26:32 <ais523> theory: it is substantially easier for something to be awesome if it has no purpose other than awesomeness
06:26:33 <shachaf> Oh, the Other Sam Hughes.
06:26:42 <ais523> agree/disagree?
06:27:20 <Sgeo> Other Sam Hughes?
06:27:27 -!- md_5 has quit (Quit: ZNC - http://znc.in).
06:27:50 <shachaf> ais523: I would imagine the opposite.
06:27:56 -!- md_5 has joined.
06:27:59 <Sgeo> ais523, that makes sense. Although I don't know if that's true in all cases. Certainly in a game like Worms, Armageddon is awesome but near useless
06:28:16 <ais523> hmm
06:28:37 <Sgeo> (Armageddon sends asteroids that destroys most of the land and kills most things, friendly or not)
06:28:44 <ais523> the example I'm thinking of, is adding a background character to a gameshow, who does nothing but watch, and make a map of the activity, and can /teleport/
06:28:55 <ais523> I find this to be very awesome, despite being completely pointless
06:29:11 <ais523> like, all the things he could do with teleportation ability, and he chooses to do that
06:29:22 <Bike> oh good i thought you were going to make them be the character representing the viewer
06:29:22 <ais523> also he never does anything with the map, AFAICT
06:29:37 <ais523> Bike: hmm, well it's not my character
06:29:44 <Sgeo> Searching for Armageddon about a game called Worms Armageddon is kind of difficult
06:29:45 <ais523> perhaps he represents the viewer? it seems unlikely though
06:29:57 <ais523> Sgeo: this is actually the reason Ubuntu versions have such weird names
06:30:05 <Fiora> I'm imagining a teleporting nepeta
06:30:19 <ais523> what's a nepeta?
06:30:23 <shachaf> Fiora: I had nothing to do with it!
06:30:27 <SirCmpwn> pikhq: haven't been highlighted in here in a while
06:30:34 <SirCmpwn> pikhq: I'm in Colorado Springs, near Garden of the Gods
06:30:38 <ais523> …what's a Cmpwn?
06:30:45 <Sgeo> ais523, a character in Homestuck.
06:30:47 <ais523> ah
06:30:50 <SirCmpwn> ais523: sex toy
06:30:53 <SirCmpwn> Sgeo: really?
06:30:57 <pikhq> SirCmpwn: Huh.
06:30:58 <shachaf> ais523: Apparently related to me.
06:31:01 <pikhq> SirCmpwn: I'm in Falcon.
06:31:01 <shachaf> `? shachaf
06:31:04 <HackEgo> shachaf sprø som selleri and cosplays Nepeta Leijon on weekends.
06:31:05 <SirCmpwn> pikhq: well done
06:31:22 <shachaf> `? ais523
06:31:24 <HackEgo> Agent "Iä" Smith is an alien with a strange allergy to avian body covering, which he is trying to retroactively prevent from ever evolving.
06:31:30 <pikhq> Now to wonder if I know you in meatspace.
06:31:43 <Fiora> she makes charts of all the relationships of the other characters and updates it as the story goes on
06:31:52 <ais523> Fiora: oh, that makes sense
06:32:00 <ais523> that's also kind-of awesome
06:32:01 <Fiora> (but isn't that involved in the main plot)
06:32:02 <shachaf> Hmm, I don't do that.
06:32:10 <shachaf> Not for #esoteric, anyway.
06:32:14 <shachaf> Maybe I should.
06:32:27 <SirCmpwn> pikhq: aka Drew DeVault
06:32:32 <pikhq> SirCmpwn: Would you happen to know a "Josiah Worcester"?
06:32:38 <SirCmpwn> nope
06:32:38 <pikhq> Mmm, not really.
06:32:47 <ais523> pikhq: isn't that pikhq's real name?
06:32:53 <SirCmpwn> the only time I've been in Falcon was for a marching band competetion in high school
06:32:56 <SirCmpwn> that was a while ago
06:33:00 <ais523> I'm kind-of amused I recognised it
06:33:06 <ais523> haven't seen it for years
06:33:10 <Fiora> http://images3.wikia.nocookie.net/__cb20100727230124/mspaintadventures/images/1/13/02295.gif <-- the 'shipping wall'
06:33:38 <ais523> hmm… other things that are unexpectedly awesome: in An Untitled Story, very few of the characters have names
06:34:02 <shachaf> At this point I am confused.
06:34:04 <shachaf> @wn awesome
06:34:04 <lambdabot> *** "awesome" wn "WordNet (r) 3.0 (2006)"
06:34:05 <lambdabot> awesome
06:34:05 <lambdabot> adj 1: inspiring awe or admiration or wonder; "New York is an
06:34:05 <lambdabot> amazing city"; "the Grand Canyon is an awe-inspiring
06:34:05 <lambdabot> sight"; "the awesome complexity of the universe"; "this
06:34:07 <lambdabot> [4 @more lines]
06:34:11 <ais523> there's one character (a shuriken-throwing ninja) who appears exactly twice, makes a grand appearance that's hard to exist, and although his name is never said in-game, it's apparently "Shakespeare"
06:34:12 <Bike> awe-full
06:34:16 <pikhq> ais523: Yes, that's my real name.
06:34:21 <pikhq> ais523: I am Josiah "pikhq" Worcester.
06:34:22 <ais523> and that makes him rather more awesome than he would be otherwise
06:34:39 <ais523> pikhq: SDA grammar? (SDA always presents nicknames like that, it looks weird)
06:34:48 <pikhq> SirCmpwn: Anyways. I'm a student at UCCS.
06:34:52 <pikhq> ais523: SDA?
06:34:57 <SirCmpwn> pikhq: there's a UCCS in falcon?
06:34:58 <ais523> speedrunning site
06:35:03 <pikhq> Oh. That one.
06:35:08 <shachaf> "University of Colorado Colorado Springs"?
06:35:10 <pikhq> SirCmpwn: No, I drive the 15 miles.
06:35:13 <shachaf> Do they actually call it that?
06:35:16 <SirCmpwn> pikhq: fancy
06:35:17 <pikhq> shachaf: Yes.
06:35:19 <Sgeo> Incidentally, the girl with the blue helmet in the red circle in that chart is Nepeta
06:35:32 <pikhq> SirCmpwn: Though there *is* a PPCC campus out here now.
06:35:35 <shachaf> Sgeo: It's like looking into a cosmirror.
06:35:41 <Sgeo> (She's in that relationship, and also wants to be in the relationship that has her and the guy with the Cancer symbol)
06:35:42 <pikhq> It's kinda pathetic, but nevertheless.
06:35:44 <SirCmpwn> pikhq: feel free to give me a ring sometime and I'll buy you a beer, or your age bracket's equivalent of beer
06:35:57 <pikhq> My age bracket's equivalent of beer is beer, so.
06:36:14 <ais523> so the sentence can be optimized
06:36:15 <shachaf> A ring? Isn't it a bit early for that?
06:36:18 <pikhq> Sadly just started on some meds that scream "don't use alcohol" though.
06:36:29 <SirCmpwn> :( what ails you?
06:36:31 <Sgeo> Fiora, look at the probably not relationship
06:36:43 <pikhq> Depression. ... So, twice bad idea actually.
06:37:10 <SirCmpwn> I'll buy you a whore, then
06:37:15 <ais523> pikhq: I recommend not trying to implement Feather as a depression cure
06:37:20 <ais523> I tried once, I thought it might help
06:37:21 <pikhq> ais523: No plans to.
06:37:24 <ais523> instead I ended up depressed /and/ confused
06:37:37 <Fiora> Sgeo: ironically yeah <.< took 3 years though
06:37:39 <pikhq> SirCmpwn: Meh.
06:37:41 <Bike> oh i thought he meant, "not trying to implement feather" as a depression cure
06:37:51 <SirCmpwn> I hear sex is good for sadness
06:37:55 <ais523> hmm
06:38:00 <SirCmpwn> though not suitable as a long term cure
06:38:00 <pikhq> Yeah, but not really for depression.
06:38:14 <ais523> "trying not to implement Feather" is a combination that hasn't been tried, perhaps?
06:38:22 <ais523> the usual state of affairs is "trying to not implement Feather"
06:38:33 <ais523> and occasionally, "trying to implement not-Feather"
06:38:48 <pikhq> Alas, sex with girlfriend does not exactly do anything about feelings of self-loathing.
06:38:52 <Sgeo> How do I tell the difference between sadness because of my situation and depression?
06:39:04 <SirCmpwn> pikhq: have you tried auto-fellatio?
06:39:18 <pikhq> SirCmpwn: I am insufficiently flexible.
06:39:24 <SirCmpwn> pikhq: :(
06:39:38 <ais523> pikhq: fwiw, as far as I can tell the best cure for depression is having something to achieve
06:39:54 <SirCmpwn> apparently it's medical advice hour in #esoteric
06:40:02 <Bike> Sgeo: doc tells you you're crazy
06:40:11 <Fiora> is "seeing a psychologist" on that list?
06:40:33 <Bike> which list
06:40:33 <pikhq> ais523: Yeah, um, getting out of bed is quite an accomplishment.
06:40:37 <Bike> do we have an `mlist too
06:40:40 <zzo38> Is "seeing a philosopher" on the other list?
06:40:45 <Bike> `mlist
06:40:47 <HackEgo> ​/home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: mlist: not found
06:40:56 <shachaf> Seeing a psychologist sounds like a bad idea to me.
06:41:01 <ais523> pikhq: that sounds more like flu than depression
06:41:08 <shachaf> Then again, what if I have an anti-psychologist neurosis?
06:41:10 <Bike> yeah, psychiatrist would be better
06:41:13 <pikhq> And anyways, I've just *started* medical treatment for it.
06:41:21 <shachaf> Psychiatrist sounds like an even worse idea to me.
06:41:23 <Bike> ais523: it's pretty common with depression too. lack of energy and stuff
06:41:24 <ais523> hmm, medical treatment sounds like it might help
06:41:54 <zzo38> No! Philosopher would be better, unless you are really religious in which case you should go to church instead.
06:41:55 <Bike> People who have it really really really bad can like, stay up for days straight starting at walls.
06:42:05 <Bike> fun times
06:42:13 <pikhq> ais523: Depression is rather different from merely being sad. Utter lack of energy and motivation, inability to enjoy things...
06:42:32 <Bike> you missed an opportunity to call it anhedonia
06:42:37 <pikhq> I did.
06:42:39 <Bike> good enough word that you should use it as much as possible imo
06:42:42 <pikhq> 'Tis.
06:42:50 <pikhq> Particularly if it's applicable.
06:42:57 <ais523> pikhq: indeed
06:43:05 <pikhq> Anhedonia may be the worst thing.
06:43:07 <ais523> and yeah, Bike's word is indeed good
06:43:11 <Sgeo> zzo38, um, I don't think philosophy can fix brain chemical balance issues
06:43:36 <ais523> Sgeo: it could change the way you react to them
06:43:41 <zzo38> Sgeo: I also don't think so, but it is worth a try
06:43:41 <Bike> you know i'm not sure how to feel about the "chemical imbalance" thing
06:44:04 <Bike> on the one hand, it helps dissociate it from the common "just stop being sad" reaction. on the other, it's not very accurate and biopsychiatry by itself isn't so great
06:44:30 <ais523> I also had not seen the word "biopsychiatry" before
06:44:30 <madbr> does it vary from case to case?
06:44:36 <pikhq> It's a *lot* closer than many misconceptions of how depression works, though.
06:44:37 * ais523 invents molecular philosophy
06:44:47 <Fiora> Bike: how about "depression is a brain state with negative consequences that is often difficult to escape through normal means"?
06:44:55 <Sgeo> What is a more accurate description?
06:44:58 <Bike> Fiora: Sounds complicated.
06:45:04 <pikhq> In that clearly depression is an unusual and tricky brain state.
06:45:09 <Bike> Sgeo: Probably what Fiora said, I guess.
06:45:28 <Fiora> I'm imagining, like. a gradient descent search getting caught in a valley
06:45:29 <Fiora> but
06:45:30 <Fiora> I'm a dork
06:45:31 <Bike> biopsychiatry is just psychiatry from a biological view; you're insane because your neurotransmitters don't bond dopamine the right way, or w/e
06:45:53 <Bike> unfortunately chemical treatment alone isn't always effective.
06:46:11 <zzo38> If you don't have anything to do, then do something else.
06:46:27 <ais523> molecular philosophy is, like, do atoms have rights?
06:46:34 <ais523> and if not, why not?
06:46:43 <Bike> Oh, I was thinking mereology.
06:46:51 <SirCmpwn> I think anything capable of asking for civil rights should be granted them
06:46:55 <ais523> if you replace every atom of someone with another atom, when do they lose their rights?
06:46:56 <SirCmpwn> so, tell me, have atoms asked?
06:47:03 <ais523> hmm
06:47:04 <Fiora> large groups of atoms have asked!
06:47:05 <Sgeo> main = putStrLn "May I have civil rights?"
06:47:05 <madbr> bike : for some mental illnesses that's the best treatment we have I think
06:47:22 <zzo38> SirCmpwn: Well, yes, I think so too. But still, you must think of it more thoroughly.
06:47:28 <Bike> madbr: usually some kind of therapy along with drugs helps. but as you said (or, asked) it varies from case to case.
06:47:41 <zzo38> Yes, atom will have a rights to be atomic, for example.
06:47:49 <SirCmpwn> zzo38: >civil rights
06:47:52 <pikhq> Bike: Be nice if it were... Simple, easy cure for being in a state where "pleasure" is nontrivial would be nice.
06:48:00 <ais523> zzo38: was that a pun?
06:48:01 <zzo38> They don't do other things, so the other rights simply is not applicable; therefore you don't answer yes or no.
06:48:03 <Bike> Fiora: but can you really say atoms are "groups" in any meaningful sense ehhhhh??
06:48:15 <zzo38> ais523: I don't know.
06:48:56 <Bike> pikhq: and on the other side, http://depressioncomix.tumblr.com/image/41893397817
06:49:01 <zzo38> Bike: Everything is "groups" of stuff simply defined as being stuff by the division we are using into physical objects; that is not the only way to do it! All the universe is one thing all together.
06:49:09 <SirCmpwn> zzo38: well, you can offer civil rights, even if they are incapable of using them
06:49:34 <zzo38> SirCmpwn: It wouldn't be useful or even meaningful though.
06:49:38 <pikhq> Bike: Mmm. I recently was linked to that tumblr. It is... rather too applicable.
06:49:39 <ais523> wait I just noticed the topic
06:49:41 <ais523> whose fault is that?
06:49:53 <Sgeo> Bike, :(
06:50:00 <SirCmpwn> zzo38: doesn't matter. If they ask, they'll be offered them. They needn't take advantage of it
06:50:01 <Bike> pikhq: i know, right :(
06:50:02 <ais523> I'm not even angry, I don't think it technically breaks any rules
06:50:03 <zzo38> Should they have the rights? Yes, but since it isn't even meaningful to say does have such rights, it is not yes.
06:50:04 <ais523> it just isn't useful
06:50:15 <SirCmpwn> zzo38: also: collectively, atoms are capable of taking advantage of civil rights
06:50:24 <ais523> I think zzo38's viewpoint makes a lot of sense here
06:50:29 <Bike> Sgeo: the funny thing is, the comic's by the same guy as did Sexy Losers.
06:50:44 <zzo38> SirCmpwn: O, yes, of course you can do that, I suppose. But *if they ask* it means. Really you consider the large object consisting of many atoms, which is ask.
06:50:50 <Sgeo> Bike, was he depressed? I remember he stopped doing them
06:50:55 <zzo38> If they ask individually, is they aren't.
06:50:56 <SirCmpwn> zzo38: but, more to the point, anyone capable of asking for civil rights is also capable of, say, casting a vote
06:51:05 <zzo38> If it was, it would be meaningful!
06:51:18 <zzo38> If it cannot civil rights then it cannot ask, either.
06:51:24 <Bike> Sgeo: He is, but I don't think that's why he stopped. I think he does Sexy Losers ish comics occasionally on another tumblr actually.
06:51:27 <ais523> SirCmpwn: casting votes can be automated very easily
06:51:50 <ais523> does that mean computers need rights?
06:51:55 <zzo38> SirCmpwn: Yes, that is true.
06:52:01 <SirCmpwn> ais523: just one example. The point here is that someone capable of communicating their desire for civil rights is capable of communicating their opinions as offered by those rights
06:52:07 <zzo38> But not really relevant, until you figure out the first thing, really?
06:52:09 <ais523> (arguably, all evolutions in voting methods have been attempts to avoid the automation of voting)
06:52:18 <pikhq> But, yeah, SirCmpwn. Appreciate the offer of beer, but ATM I've got a stock of such that I intend not to drink because I do not want to see how fluoxetine and ethanol interact.
06:52:27 <SirCmpwn> ais523: before you say "I'll write a script to ask for civil rights", know that I'll find you and slap the shit out of you
06:52:29 <zzo38> ais523: Computers are what they are programmed to be.
06:52:49 <ais523> SirCmpwn: you can write a script to print words to the screen, but I don't know if it would understand what it was saying
06:52:52 <zzo38> Does mathematical formulas have the right to vote *by themself*? You cannot easily count all the possible ones...
06:52:59 <ais523> I think you can write a script to request civil rights for the author of the script
06:52:59 <Bike> voting is pointless anyway haven't you all heard of arrow's theorem !!!
06:53:09 <SirCmpwn> also, if sentient computing is achieved, those computers should be granted civil rights
06:53:24 <ais523> Bike: I have, it doesn't really prove that voting is pointless though
06:53:36 <Sgeo> SirCmpwn, didn't I do that just above?
06:53:38 <ais523> basically, it just proves that except in ideal circumstances, all voting methods have defects
06:53:43 <Sgeo> The thing that you said you'd slap ais523 for?
06:53:54 <SirCmpwn> Sgeo: I didn't notice, but I'd slap you over it, too
06:53:56 <ais523> Sgeo: you've been slapped down by the channel often enough already
06:54:20 <ais523> btw, I /still/ don't recommend attempting to give your girlfriend cancer
06:54:30 <Bike> I'm forced to agree.
06:54:33 <SirCmpwn> <pikhq> But, yeah, SirCmpwn. Appreciate the offer of beer, but ATM I've got a stock of such that I intend not to drink because I do not want to see how fluoxetine and ethanol interact.
06:54:39 <ais523> (backstory: some of the other members of the channel recommended that as a joke, AFAICT; I disagreed with them)
06:54:44 -!- azaq23 has joined.
06:54:45 * pikhq has not been paying attention to this room enough.
06:54:46 <zzo38> If you have room with people who don't know Chinese but will follow all the instructions given to them to make an answer to a question in Chinese (like ELIZA and some computer programs, but completely by hand), are they Chinese?
06:54:47 <SirCmpwn> pikhq: alright, no problem. Swap beer out for "social gathering" if you wish
06:54:51 <ais523> (further backstory: it's not entirely clear whether Sgeo actually has a girlfriend)
06:55:03 <pikhq> SirCmpwn: Very well then, sounds reasonable.
06:55:19 <zzo38> But really you must consider mathematics. Can a mathematics think?
06:55:36 <Sgeo> I have a girlfriend as of more recently than that occurred.
06:55:38 <Bike> fuck, i can't even joke off of what zzo38 said, because I just KNOW some psephologist has considered how elections would work in the presence of countably infinite voters.
06:55:40 <zzo38> I don't want to ask if a computer can think; I want to ask if a mathematics can think!
06:56:00 <Bike> ('psephologist' is a good word too btw)
06:56:02 <ais523> Bike: what's a psephologist?
06:56:09 <SirCmpwn> zzo38: I suppose a sufficiently complex algorithm could think, but only with the assistance of a human interpreter, whose skills would be far too lacking to manage it
06:56:12 <ais523> and yeah, you've come up with three words I didn't know in less than an hour
06:56:19 <Bike> ais523: someone who studies voting systems
06:56:38 * pikhq wonders what his talking-here stats have looked like in the past couple years, but can't be bothered to write the script for that.
06:56:38 <Bike> or similar things
06:56:51 <ais523> fizzie may have those stats to hand already
06:56:54 <Bike> pikhq: I think fizzie made some fancy-looking graphs.
06:56:56 <ais523> he likes doing those sorts of stats
06:57:03 <pikhq> Something like fetching the logs and grepping the logs to generate a histogram, but *eh*
06:57:22 <zzo38> SirCmpwn: Are you sure? Well, the same algorithm could be put into a computer, although possibly it would be too complicated to put into the computer, too. Even so, some things may be too large for the universe, and/or can have mathematical formulas that nobody has figured out yet.
06:57:42 <kmc> http://en.wikipedia.org/wiki/Arrow%27s_impossibility_theorem#Infinitely_many_individuals
06:57:48 <Bike> Dammit. Dammit!
06:57:54 <SirCmpwn> zzo38: I think that the human mind is just a really, really, REALLY complicated program
06:58:05 <zzo38> Furthermore, if you want to ask if a computer can think, also ask, can a submarine swim?
06:58:26 <SirCmpwn> a submarine can swim as a proxy for a sentient being or an AI
06:58:35 <SirCmpwn> it's a vehicle, it doens't calculate
06:58:40 <SirCmpwn> (or think)
06:58:46 <ais523> zzo38: from my point of view, I'm basically 100% certain that it's possible for a computer to be sentient
06:58:54 <ais523> I'm not sure either way whether any currently existing computers are sentient
06:58:54 <Bike> I'm more and more convinced that zzo38 is Gary P. Rastov.
06:59:06 <Bike> kmc: ok i gotta admit i fully support the use of ultrafilters in social sciences
06:59:06 <SirCmpwn> I expect that there will be a human brain emulator within the century, zzo38
06:59:12 <zzo38> ais523: From my point of view, I don't know.
06:59:15 <kmc> yes
06:59:22 <Sgeo> I hope Norns aren't sentient
06:59:28 <ais523> so… kmc's link basically says that there are voting systems that don't allow tactical voting when infinitely many voters are involved, but they're noncomputable?
06:59:29 <Sgeo> If they are, I'm ... rather horrible
06:59:32 <Bike> So! Terminology bit here.
06:59:51 <zzo38> But I expect, everything is mathematics. So, even the universe is mathematics. But, not necessarily mathematics that has exactly one solution all the time; it can be no solution, multiple solutions, and can even be uncomputable.
06:59:57 <Bike> "sentient" means roughly "can feel sensations". Things like dolphins are dogs are sapient. You are probably talking about sapience.
07:00:05 <kmc> ais523: and they contain 'invisible dictators' which are?
07:00:15 <Bike> Sgeo: Just to clarify, you know I'm not serious about thinking you're a sociopath, right?
07:00:22 <Bike> er, *are sentient
07:00:31 <SirCmpwn> Bike: thanks for the correction
07:00:49 <ais523> Bike: wait, did you correct a veiled insult into an actual insult?
07:00:49 <Sgeo> Bike, somewhat, yes
07:00:54 <zzo38> Also, the author of a book can make any of the characters dead and/or illogical however they want, isn't it?
07:00:57 <Bike> ais523: no?
07:01:18 <Bike> I wouldn't have mentioned it except that the question for Sgeo is actually whether Norns are sentient, rather than sapient, since he's just worried about inflicting pain on them, not selling them insurance.
07:01:22 <ais523> zzo38: if you prove to someone that they don't exist, do they stop existing?
07:01:23 <Sgeo> Bike, although I know someone who is somewhat concerned about what I (and others) would do if sapient computer programs are created
07:01:34 <kmc> so if you ask "are infinitely large societies governed without tactical voting all dictatorships?" the answer depends on whether you're using classical or constructive logic?
07:01:38 <Sgeo> (or sentient?)
07:01:43 <Sgeo> He uses the word conscious.
07:01:54 <ais523> kmc: I hope the answer depends on the axiom of choice
07:01:56 <zzo38> ais523: If the proof is correct, they didn't exist at the time either, so they won't stop either.
07:01:56 <Bike> "conscious" is good for that yeah, since it's more common :P
07:01:58 <ais523> it might even be relevant!
07:02:06 <ais523> zzo38: hmm, indeed
07:02:26 <Bike> I'm pretty convinced that people have made robots that can feel hunger. It's an odd thing to think about.
07:02:33 <Sgeo> Norns do have brains that detect pain, and react to such
07:02:39 <zzo38> So either way they won't stop existing.
07:02:42 <Sgeo> Lemme upload a picture of the norn brain
07:02:44 <ais523> Sgeo: what's this meaning of "norn"?
07:02:56 <Bike> ais523: it's a thing in an alife game called Creatures.
07:03:15 <ais523> the only meaning I know offhand is a specific mythological figure who's in charge of valkyries
07:03:18 <Bike> Sgeo: I don't know much about Creatures but it doesn't seem like they're any more complicated than, like, nematodes.
07:03:49 <Sgeo> http://dl.dropbox.com/u/16240872/Creatures/chichi_brain.PNG
07:04:11 <Bike> ais523: Basically they're virtual animals with such things as distinguishable "personalities", and C&C of their bodies through an artificial neural network. And they have some genetics.
07:04:17 <Sgeo> drive is things like hunger, pain, boredom
07:04:21 <ais523> Bike: hmm
07:04:31 <Bike> Sgeo: How many neurons is this, out of curiosity?
07:04:37 <Bike> Nematodes have I think... 293?
07:04:38 <SirCmpwn> I should change my pants
07:04:40 <Sgeo> attention and decision are the outputs: What they're focusing on, and what the action is
07:04:44 <SirCmpwn> I went to the store and bought more
07:04:52 <ais523> btw, I think my own personal barrier for "when should something be considered sentient" is "the thing can communicate and add meaningfully to a discussion"
07:04:54 <SirCmpwn> because the fly is busted on this one
07:04:55 <Bike> Of course nematodal neurons are more complicated, but hey.
07:05:07 <SirCmpwn> ais523: define meaningfully
07:05:11 <ais523> this can in some cases include unborn children, and many animals
07:05:22 <SirCmpwn> unborn children wat
07:05:25 <ais523> SirCmpwn: you end up knowing more/different than if it hadn't contributed to the discussion
07:05:27 <Sgeo> comb is 40*13 = 520 I think (if I counted the verbs right)
07:05:45 <ais523> SirCmpwn: it's possible for unborn children to communicate, once they're negative-young enough, isn't it?
07:05:45 <Sgeo> Bike, also, the lobes and dendrites all have pieces of code in a weird language
07:05:49 <Bike> ah, nematodes have 302.
07:05:51 <Sgeo> That's not visible on there
07:05:53 <Bike> or rather c. elegans do.
07:05:57 <ais523> the mother can feel the inside of her womb
07:06:00 <SirCmpwn> ais523: it's barely possible for them to communicate *after* birth
07:06:11 <Sgeo> I would say the most important thing is the drive->comb dendrites
07:06:16 <Bike> Wow, you know. C. elegans have 1031 cells, and of those 302 are neurons. That's like, almost a third!
07:06:26 <SirCmpwn> babies come out bumbling, stupid, and crippled, and become intelligent over the course of several months
07:06:30 <ais523> SirCmpwn: they can communicate needing attemption versus not needing attention
07:06:33 <ais523> even from birth
07:06:37 <SirCmpwn> ais523: so can a dog
07:06:41 <ais523> yeah
07:06:42 <ais523> dogs are sentient
07:06:47 <ais523> aren't yhey?
07:06:49 <ais523> *they?
07:06:54 <SirCmpwn> as in Bike's definition of "capable of sensation"?
07:06:57 <Sgeo> Since each comb neuron corresponds to a particular course of action to take
07:07:01 <ais523> SirCmpwn: as in my definition
07:07:04 <Sgeo> Like "eat seed"
07:07:07 <ais523> but I think pretty much everyone's
07:07:08 <SirCmpwn> state your defintion
07:07:09 <pikhq> SirCmpwn: That's the generally-accepted notion of sentient actually.
07:07:15 <Sgeo> And drive corresponds to things like pain and hunger
07:07:19 <SirCmpwn> pikhq: indeed, but there was confusion earlier
07:07:24 <pikhq> SirCmpwn: Star Trek "sentient" is a weird mutation of the term.
07:07:32 <zzo38> Do you think, everything is mathematics, or do you think, everything is physics, or something else different?
07:07:33 <pikhq> Meaning basically "sapient".
07:07:42 * Bike is pretty jaded to "can computers think" sorts of arguments, if it's not obvious.
07:08:01 <Sgeo> I don't know what the response lobe does
07:08:04 <pikhq> Of course, in a lot of contexts they're practically interchangeable. A sentient computer and a sapient computer would both be rather shocking.
07:08:05 <Sgeo> I forgot about it :/
07:08:14 <pikhq> (though the sapient computer more so)
07:08:16 <ais523> zzo38: isn't it possible that mathematics is physics, and so everything is both mathematics and physics?
07:08:30 <Bike> Sgeo: Yeah, it definitely sounds like norns have a lot of their behavioral repetoire pre-baked, but that's to be expected.
07:08:35 <zzo38> ais523: I don't think so! I think physics is mathematics.
07:08:52 <SirCmpwn> sorry, the point I was making earlier is that babies are not special and I wouldn't even call them capable of intelligence until many months after birth
07:09:08 <SirCmpwn> "intelligence" here meaning coherent thought like you and I are capable of
07:09:20 <pikhq> SirCmpwn: So, "sapient".
07:09:27 <Sgeo> Bike, well, yes, but not as you'd might think. The drive->comb connections are random at first, and the dendrites wander based on reward/punishment
07:09:29 <SirCmpwn> yes
07:09:31 <zzo38> Like you and I... that is your self-centerism
07:09:41 <SirCmpwn> zzo38: wat
07:09:42 <pikhq> Yes, babies take several months to become sapient.
07:09:44 <Sgeo> But they have "instincts" in their genomes that make them dream of, say, eating a seed and being rewarded for it
07:09:45 <ais523> zzo38: is it possible for physics to be mathematics, but mathematics to not be physics?
07:09:54 <Bike> Sgeo: I still haven't played Creatures despite how cool it seems. Weird, huh?
07:10:07 <zzo38> ais523: Well, not entirely. Mathematics is beyond physics.
07:10:12 <ais523> ok
07:10:16 <Sgeo> Bike, http://creaturesdockingstation.com/
07:10:17 <zzo38> Or, at least, that is what I mean.
07:10:33 <pikhq> Bike: Hey, I last played a videogame for very long in 2010 or so, so...
07:10:47 <Sgeo> Uh, the first screenshot on there isn't DS
07:10:49 <ais523> videogames can be good
07:10:54 <ais523> I last played a videogame a few hours ago
07:11:05 <Bike> Sgeo: yeah, i still have it open, i just haven't gotten around to it.
07:11:09 <ais523> in fact, one recent enough to have a copyright date of 2012
07:11:19 <zzo38> But have you played Super ASCII MZX Town a few minutes ago?
07:11:23 <ais523> zzo38: no
07:11:24 <pikhq> My queue just keeps growing and growing.
07:11:30 <ais523> you could probably have guessed that
07:11:30 <Sgeo> Don't bother with the registration thing though, I think the Warp is down anyway
07:11:37 <Sgeo> Just get the offline option
07:11:47 <ais523> pikhq: hmm… I don't keep a queue
07:11:49 <zzo38> Have you played any Famicom game with the copyright date 2013?
07:11:52 <ais523> rather I play what I feel like playing
07:11:53 <ais523> zzo38: no
07:12:01 <pikhq> ais523: It's metaphorical in this case.
07:12:02 <ais523> I'm not sure if I've played any game with the copyright date 2013
07:12:13 <ais523> I mean, I jump around between games a lot, and replay them often
07:12:14 <zzo38> Or else public domain; even if it is not copyright.
07:12:17 <zzo38> (Some are public domain)
07:12:32 <SirCmpwn> I would like to have a word with whoever designed scissor packaging that requires scissors to open
07:12:34 <pikhq> I've also got a giant queue of... every other form of media, really.
07:12:40 -!- doesthiswork has joined.
07:12:49 <zzo38> SirCmpwn: Yes, I also want to complain to them too
07:12:56 <Bike> I have a big pile of books
07:12:57 <pikhq> (if there were a god, I'd call him a dick for even having "anhedonia" in the same universe as me)
07:12:58 <Bike> literally
07:12:59 <Bike> :(
07:13:16 <Bike> i also have a phone with an increasingly more irritating ebook reader
07:13:18 <ais523> Bike: so do I, but most of them I'm not intending to read again in the near future, or possibly at all
07:13:28 <Bike> that's just it i haven't read them
07:13:30 <Bike> they call for me...
07:13:46 <ais523> and I don't have an ebook reader that isn't also a general purpose computer
07:14:04 <Bike> They still make machines that aren't also general purpose computers?
07:14:08 <ais523> (all general purpose computers are ebook readers)
07:14:16 <ais523> Bike: sure, in fact they're becoming more common, not less
07:14:24 <Bike> Whoa.
07:14:29 <ais523> basically because they would be general purpose computers but have artificial restrictions so that they aren't
07:15:10 <zzo38> That is why I make the computers which are very general purpose computers with no restrictions other than the security stuff that the user can always override anyways.
07:15:30 <ais523> or to put it another way, specific purpose computers made using general purpose computers
07:15:33 <zzo38> (And to include instructions in the book, tell you how to override, how to program it, how everything is working.)
07:16:02 <ais523> zzo38: yes, I'd be very surprised if a computer you made was locked down
07:17:00 <zzo38> To really ensure general purposes, to ensure it is made so that it can be made anyone can clone even not needing patents to do so (even if patented technology must be used, such that it can be made the same easily without it too), however, having a trademark license so that if someone makes the wrong clone, that they cannot use the same name.
07:17:10 -!- TeruFSX has quit (Ping timeout: 260 seconds).
07:18:09 <zzo38> ais523: Well, there are some restrictions on it but you can easily remove them without causing problems with existing software, so it is not a problem.
07:18:47 <zzo38> (Such as a reprogrammable ROM chip is normally not reprogrammable unless you move the corresponding jumper, which is clearly labeled in the instruction book.)
07:21:12 <zzo38> And I mean really general purpose computer that you can even just turn on, and then type in the program codes and it will run (although you can also load the already existing programs too, even if they are already compiled).
07:21:34 <zzo38> Mostly PC BIOS computers don't have that anymore but they should make them such.
07:21:46 <Bike> The Aristocrats!
07:22:23 <zzo38> What about aristocrats?
07:23:39 <zzo38> What is the point of the Continue bit in the envelopes in AY8910? It seems to be the same shape as the ones that do continue than the ones tat don't (since you can also use the Hold bit).
07:23:52 <ais523> zzo38: Bike's referencing a famous joke pattern which ends up saying "the aristocrats" as a non sequitur
07:24:17 <ais523> the joke pattern's sufficiently famous that you can reference the joke pattern just by using the punchline, the rest of the joke is irrelevant
07:24:26 <zzo38> OK
07:24:27 <ais523> although if you do it that way, it isn't actually funy
07:24:29 <ais523> *funny
07:24:34 <Bike> zzo's version has way less rape, admittedly.
07:25:01 <zzo38> Version of what?
07:25:11 <Bike> The joke.
07:25:32 -!- Taneb has joined.
07:26:51 <ais523> having fewer rapes sounds like a good thing, generally speaking
07:27:05 <ais523> unless something else bad increases to compensate, then it might or might not be a good thing
07:27:12 <madbr> zzo: I'm not sure hardware envelopes make sense on chips with that few gates :o
07:27:14 <Bike> It's true.
07:27:41 <Taneb> What if it's "rape" by the older definition of "seize" or "capture", ais523?
07:27:46 <zzo38> madbr: But they do have hardware envelopes. I don't know how many gates AY8910 has.
07:27:57 <Bike> In context it is not, taneb.
07:28:02 <Taneb> Okay7
07:28:14 <ais523> Taneb: then it depends on what's being seized and captured, and using the more modern words would be helpful in order to reduce confusion and avoid bad connotations
07:28:16 <Taneb> Okay7 is like "Gr8" but not as good
07:28:39 <Bike> Okay with a salute
07:28:44 <pikhq> Yeah, fewer rapes seems like a generally good policy.
07:28:57 <ais523> hmm… today seems to be the ais523 "have I mentioned I'm lawful good yet" day
07:29:05 <ais523> 3 march, I must remember that
07:29:12 <Taneb> `? ais523
07:29:14 <HackEgo> Agent "Iä" Smith is an alien with a strange allergy to avian body covering, which he is trying to retroactively prevent from ever evolving.
07:29:16 <Bike> "btw, rape? bad"
07:30:26 <Taneb> `run echo "Agent \"I\" Smith is an alien with a strange allergy to avian body covering, which he is trying to retroactively prevent from ever evolving. On the 3rd of March, he's lawful good." > wisdom/ais523
07:30:29 <HackEgo> No output.
07:30:32 <Taneb> `? ais523
07:30:34 <HackEgo> Agent "I" Smith is an alien with a strange allergy to avian body covering, which he is trying to retroactively prevent from ever evolving. On the 3rd of March, he's lawful good.
07:30:51 <ais523> actually, I think I'm almost always lawful good, just more likely to point it out on the 3rd of March
07:30:53 <zzo38> What character encoding do you want to use?
07:31:01 <Bike> UTF-EBCDIC
07:31:05 <ais523> zzo38: utf-8 is a generally useful character encoding
07:31:08 <zzo38> In here is not 3rd of march, yet.
07:31:12 <shachaf> UCS-2.625
07:31:17 <ais523> Bike: ouch for reminding me that exists
07:31:17 <zzo38> UTF-EBCDIC is not ASCII so don't use that.
07:31:21 <Bike> :D
07:31:29 <zzo38> Use UTF-8 if you want to, though, or CP437, or whatever else.
07:31:34 <ais523> zzo38: there are a lot of encodings that aren't ASCII
07:31:37 <ais523> most of them, in fact
07:31:37 <shachaf> CP437 is also not ASCII
07:31:46 <Taneb> ISO/IEC 8859-14
07:31:48 <Bike> zzo38 likes supersets of ascii.
07:31:48 <shachaf> UTF-8 is, at least, compatible with ASCII
07:31:49 <madbr> what's cp435 again
07:31:58 <zzo38> So you are wrong CP437 is ASCII, and so is UTF-8.
07:32:03 <Bike> How about SHIFT-JIS?
07:32:09 <Bike> "good encoding"
07:32:12 <pikhq> madbr: CP437 is DOS.
07:32:26 <Bike> SHIFT probably isn't supposed to be in allcaps, but I feel loud.
07:32:27 <shachaf> zzo38: Oh, apparently there are multiple CP437s?
07:32:33 <pikhq> Namely, the character set used by DOS in the US.
07:32:38 <shachaf> I was thinking of the one that has 0x01 representing ☺ and all that.
07:32:43 <ais523> Bike: what about SHIFT-JIS, except you replace the shift codes with gambling games and add network transparency
07:32:46 <zzo38> shachaf: No, just one CP437 the others are other codepages.
07:32:48 <madbr> the common ones are utf8 win32 (\r\n), utf8 linux, latin1 win32, latin1 linux, and then shift jis and a couple of other popular asian encodings
07:32:55 <Taneb> TIS-620
07:32:55 <pikhq> ais523: Yeah, that's CP437.
07:32:56 <Bike> ais523: good encoding. w/o sacrequotes
07:32:57 <zzo38> Yes 0x01 is the happy face in CP437
07:33:00 <Bike> scare*
07:33:20 <madbr> and by latin1 I mean codepage whatever (the one used in win32, not the one with the dumb extended control characters)
07:33:32 <ais523> madbr: windows-1252?
07:33:42 <madbr> yeah
07:33:56 <madbr> it also has some freakish iso name
07:34:06 <zzo38> I use whatever encoding I need depending on the program being used, but is usually ASCII. Sometimes some parts of programs might use other encodings even though it is mostly ASCII, such as VGMCK which is ASCII except for GD3 tags which are UTF-8 and are converted into UTF-16.
07:34:07 <Taneb> madbr, Code page 437
07:34:15 <madbr> but everybody calls it latin-1 because, well, it's a zillion times more mnemotechnic
07:34:20 <pikhq> Shift-JIS is perhaps the worst encoding devised.
07:34:30 <shachaf> Don't call Windows-1252 Latin-1!
07:34:31 <ais523> pikhq: worse than the one I suggested just above?
07:34:35 <shachaf> Also, don't call it ANSI.
07:34:38 <ais523> madbr: latin-1 is a different encoding
07:34:42 <shachaf> People who call things ANSI are evil.
07:34:44 <zzo38> Since UTF-16 is not ASCII, it uses UTF-8 so that you can still use ASCII, and UTF-8 is otherwise the closest thing to convert to UTF-16.
07:34:46 <ais523> latin-1 = iso-8859-1 ≠ windows-1252
07:34:49 <pikhq> ais523: Not worse than that.
07:34:50 <shachaf> (Things other than ANSI, that is.)
07:34:53 <shachaf> (The institute.)
07:34:54 <Deewiant> ISO-8859-1 is latin1 is not Windows-1252 is not CP437
07:34:55 <madbr> ais: it's IBM's original version
07:34:56 <Bike> pikhq, it's probably just the most known weird country-specific encoding thing... eastern europe had a lot, didn't it?
07:34:59 <ais523> also windows calls windows-1252 ansi for no obvious reason
07:35:00 <madbr> ais: which nobody used
07:35:01 <pikhq> shachaf: "ANSI" is weirdo Windows-speak for "legacy charset".
07:35:19 <pikhq> shachaf: Likewise "Unicode" is weirdo Windows-speak for "UTF-16".
07:35:20 <ais523> Bike: they are helpfully numbered from latin-1 up to latin-15 or so
07:35:20 <zzo38> (Actually it also accepts CESU-8, as well as overlong encodings, but you don't have to use those if you don't want to.)
07:35:26 <Bike> ais523: beautiful
07:35:27 <ais523> pikhq: does it have a defined endianness?
07:35:31 <pikhq> Bike: It's also the worst.
07:35:37 <Bike> That's pretty worst.
07:35:37 <pikhq> ais523: UTF-16LE.
07:35:41 <shachaf> pikhq: Well, sometimes Windows people mean "Unicode" by "Unicode".
07:35:42 <madbr> oh god
07:35:45 <madbr> not utf16
07:35:45 <pikhq> Because Windows is only little-endian.
07:35:46 <madbr> die
07:35:58 <madbr> everything's little endian
07:36:00 <Bike> Actually, I have no idea how Cyrillic was typed up. Or is typed up for that matter.
07:36:09 <ais523> four thr thr fivee
07:36:11 <Bike> What were they using on pirate C64s in the 80s?
07:36:28 <ais523> Bike: cyrillic has a comparable number of letters to latin, doesn't it?
07:36:33 <pikhq> Windows is little endian even on CPUs that are usually big endian.
07:36:34 <Bike> yeah
07:36:37 <ais523> so probably just using ordinary keyboards except relabeled
07:36:47 <Bike> I meant the encoding, but yeah I see your point.
07:36:51 <Bike> Maybe China then...
07:37:04 <Bike> well, i already know how weird chinese input methods get
07:37:16 <madbr> pikhq: you mean on Alpha?
07:37:22 <pikhq> Also MIPS.
07:37:27 <zzo38> All the commands in the program are ASCII, and the comments can be whatever encodings you want (except ASCII control codes), so the program correctly follows the principle of extended ASCII.
07:37:31 <madbr> there's a mips version of windows?!?
07:37:36 <pikhq> NT 4!
07:37:40 <madbr> :o
07:37:43 <madbr> crazy
07:38:19 <ais523> zzo38: what about identifiers?
07:38:35 <ais523> that's the main place where you might want to accept outside-ASCII characters in source code
07:38:37 <ais523> oh, and string constants
07:38:41 <ais523> even more so
07:38:58 <madbr> actually I think most code is still stored as latin1
07:39:03 <ais523> (and character constants if you don't consider them a special case of string constants)
07:39:15 <ais523> madbr: most code is in ASCII, and thus can't be distinguished between latin-1 and UTF-8
07:39:20 <madbr> with windows cr-lf or just lf depending on the coder
07:39:20 <ais523> except when I write … in a comment out of habit
07:39:30 <ais523> and it depends more on the editor and OS than the coder
07:39:35 <madbr> yeah
07:39:43 <Bike> I think I usually use utf-8 because i dump «spinäl t©p» sorts of things everywhere
07:39:49 <zzo38> ais523: Those things are not applicable to this program. Identifiers for macros are just one character and only ASCII, and string constants are only used in GD3 tags, which are UTF-16, so it accepts UTF-8 there and will convert them to UTF-16.
07:40:00 <pikhq> On Linux systems, stuff is usually UTF-8.
07:40:23 <pikhq> (to the point where I generally think you should just say "a string is in UTF-8. Period.")
07:40:32 <kmc> i missed a discussion of CP 437 :(
07:40:34 <pikhq> And on Windows, Microsoft is fucking crazy.
07:40:35 <zzo38> In other programs with string constants such as in C, you should allow any characters, and just encode the bytes as is. So, if it is UTF-8, then your program will work with UTF-8.
07:40:38 <madbr> also for console programs, windows gives you filenames etc in latin-1
07:40:48 <pikhq> madbr: Not quite. It's crazier.
07:40:52 <Bike> pikhq: bush hid the facts
07:40:52 <madbr> (well, codepage whatever but still)
07:40:54 <ais523> pikhq: actually isn't "take encoding from locale" the usual default? and aren't most locales latin-n by default but people tend to use the utf-8 versions nowadays?
07:40:55 <zzo38> pikhq: I use CP437 when working on Linux.
07:41:23 -!- Gracenotes_ has quit (Ping timeout: 248 seconds).
07:41:25 <madbr> what's worse than cp437 is that there's a different version that windows sometimes uses
07:41:32 <pikhq> madbr: Windows has a very comprehensive scheme for working with arbitrary charsets. They *refuse* to extend it to handle UTF-8.
07:41:39 <madbr> which has no point except it screws up your DOS programs
07:41:43 <zzo38> Since the Linux console is only 256 character codes anyways, then Unicode is too much and is too complicated.
07:41:56 -!- Sgeo has quit (Ping timeout: 248 seconds).
07:41:59 <zzo38> UTF-8 data can still be worked with though, in programs that are using UTF-8 data, but not in other programs, please.
07:42:02 <pikhq> madbr: The only way of using Unicode on Windows is to use their non-conformant wchar_t interfaces.
07:42:08 <ais523> hmm… the Linux console understands the code for "enter UTF-8 mode"
07:42:12 -!- Gracenotes has joined.
07:42:13 <ais523> can it not render it once you've done that?
07:42:17 <pikhq> Along with their ultra-weird UTF-16 entry point.
07:42:27 <ais523> pikhq: and wchar_t isn't designed for utf-8
07:42:31 <madbr> true
07:42:32 <pikhq> ais523: Almost all distros default to UTF-8 thought.
07:42:34 <ais523> it's designed for utf-32, really
07:42:41 <pikhq> ais523: Yes, but on Windows it's UTF-16.
07:42:48 <ais523> wide character strings are meant to be fixed-width
07:42:50 <zzo38> ais523: Yes, turn on UTF-8 but there is also the code to turn off the UTF-8.
07:42:53 <ais523> utf-8 is multibyte character strings, instead
07:43:05 <ais523> zzo38: that's so you don't end up stuck in utf-8 mode forever
07:43:08 <madbr> also case sensitivity is based on a large LUT
07:43:09 <pikhq> ais523: The trick is, char* strings on Windows cannot be UTF-8.
07:43:14 -!- Sgeo has joined.
07:43:19 <madbr> that gets updated depending on your windows version
07:43:26 <zzo38> ais523: Yes, so you can switch depending on what programs you are using, then.
07:43:27 <pikhq> The entire OS only does legacy charsets that way.
07:43:31 <pikhq> Ever.
07:43:47 <madbr> so two files that are different can become the same if the unicode char gets added to the table
07:43:51 <ais523> really there should be a "utf8char" type
07:43:57 <ais523> or mbchar
07:43:57 <zzo38> pikhq: That is, using the Windows API calls; using your own programs they can be whatever encoding you want to.
07:43:59 <pikhq> So, you have to rewrite your code to handle Unicode.
07:44:03 <ais523> that represents one byte of a multibyte encoding
07:44:10 <ais523> and you have to rewrite your code to handle Unicode /anyway/
07:44:17 <ais523> and even then it won't handle Turkish correctly
07:44:40 <madbr> tbh in the tool I'm working on I just disallow anything over >128
07:44:44 <zzo38> Usually I don't handle Unicode or anything since instead I will just use ASCII, and depending what it is, using UTF-8 might still work with it, and possibly other encodings too.
07:44:48 <Bike> what's wrong with turkish
07:44:55 <zzo38> madbr: What took are you working on?
07:44:56 <madbr> bike: dotless i
07:44:56 <pikhq> Even though their API is set so the meaning of char* is determined by whatever charset is in use, and they actually have it ultra-mega-crazy to work with stuff like Shift-JIS.
07:45:06 <ais523> Bike: it has capitalization rules that are inconsistent with other languages
07:45:09 <Bike> oh that's a fun character
07:45:14 <pikhq> ais523: Yeah, but a lot of code will "just work" with UTF-8.
07:45:16 <madbr> bike : also dotted capital I
07:45:27 <pikhq> Especially on Windows, where people actually use Shift-JIS in char* strings.
07:45:42 <madbr> bike : essentially it uses dotless i to represent the same sound as japanese 'u'
07:45:45 <zzo38> pikhq: Yes, depending what it is you might not have to do anything special, so you can use UTF-8 or whatever if you want to.
07:45:57 <ais523> pikhq: so what's the problem? NIH syndrome?
07:46:02 <Bike> madbr: and that character isn't in unicode?
07:46:04 <ais523> it seems like a weird thing to not support along with other encodings
07:46:06 <pikhq> ais523: I have not the foggiest clue.
07:46:10 <ais523> fair enough
07:46:10 <madbr> bike: no it's in unicode
07:46:21 <madbr> bike: but that totally screws up capitalization
07:46:30 <madbr> since the caps version of i isn't I anymore
07:46:39 <pikhq> But now, Windows dev communities are the most confused about how character encoding works.
07:46:47 <Bike> baking capitalization into unicode seems like a bad idea anyway
07:46:51 <pikhq> And especially Unicode.
07:46:53 <kmc> 'It is considered particularly ironic that Atatürk himself, in his lengthy speech to the new Parliament in 1927, used a style of Ottoman which sounded so alien to later listeners that it had to be "translated" three times into modern Turkish: first in 1963, again in 1986, and most recently in 1995'
07:47:11 <Bike> kmc: once for each coup AM I RIGHT FOLKS
07:47:11 <zzo38> Unicode is full of problems and too much complexity.
07:47:24 <madbr> bike: you have to, because the win32 file system is case sensitive
07:47:25 <ais523> btw, anyone know how that Cyrillic letter that looks like a backwards R is pronounced? I'm guessing neutral vowel, but could do with confirmation
07:47:38 <madbr> bike: so you have to process case
07:47:44 <madbr> ais523 : 'ya'
07:47:46 <Taneb> How worth it would it be to write an Apache Wave server in Haskell
07:47:49 <zzo38> Nevertheless simply allowing UTF-8 in some parts of the program just in order to convert to UTF-16 if the output format uses it, is not complicated.
07:47:53 <ais523> madbr: hmm, OK
07:47:54 <Bike> not worth anything
07:48:05 <madbr> ais523 : the one that's neutral vowel looks like bI
07:48:09 <pikhq> (I'd say Microsoft's failure to support UTF-8 is the single greatest reason for people thinking Unicode is a 16-bit encoding)
07:48:18 <ais523> how strong is the 'a'? I'm guessing short 'a', is it strong or is it reasonably neutral?
07:48:20 <Bike> I seem to remember a natural language that used capital letters in its orthography, but I can't remember...
07:48:21 <zzo38> That is why VGMCK accepts UTF-8 in GD3 tags.
07:48:32 <ais523> pikhq: that might well play into Microsoft's hands
07:48:36 <madbr> ais523 : russian doesn't have vowel length
07:48:41 <ais523> if they're attempting to have the web misencoded for some reason
07:48:44 <madbr> ais523 : or strong vs weak vowels
07:48:50 <ais523> madbr: hmm
07:48:56 <ais523> so long a and short a are heard the same by a russian speaker?
07:49:07 <madbr> russian has only normal a
07:49:16 <zzo38> (Text macros in VGMCK are named by a single ASCII character; any bytes >= 128 are not acceptable.)
07:49:16 <pikhq> ais523: This is a stupid legacy decision that predates things like "broadband Internet".
07:49:20 <ais523> madbr: yeah but how does it correspond to English letters?
07:49:28 <ais523> pikhq: doens't predate dial-up
07:49:30 <ais523> *doesn't
07:49:36 <Bike> uuuugh i'm groaning in sympathy for phonologists i know, guys
07:49:38 <zzo38> At least to me is reasonable like that.
07:49:40 <pikhq> ais523: They made this mistake way back in NT 3.1 days.
07:49:50 <pikhq> This actually predates commercial ISPs.
07:50:08 <ais523> I think BBSes were older, weren't they?
07:50:25 <ais523> but I don't really see how it matters
07:50:26 <pikhq> Yes, but they weren't Internet service providers.
07:50:30 <madbr> ais523 : there are two sets of vowels, they call them "hard" vs "soft"... essentially normal vs y+vowel
07:51:04 <ais523> madbr: yeah, basically my problem is, if I tried to pronounce Russian, would I be accidentally saying letters that didn't exist, or would what I said be perceived as a Russian letter whether it was English short a or English long a?
07:51:24 <Bike> phonemes. phonemes that don't exist.
07:51:36 <ais523> err, yes
07:51:50 <ais523> I'm tired, OK :)
07:52:00 <Bike> me too :(
07:52:13 <ais523> like, Japanese doesn't have a separate "l" and "r", but the letters are separate in English
07:52:22 <madbr> hard: А Э Ы О У = a e <neutral vowel> o u (a eh japanese-u o oo)
07:52:31 <pikhq> And in fact Japanese "r" is not the same vowel as either.
07:52:37 <Bike> isn't japanese "r" more like "d"
07:52:43 <ais523> pikhq: indeed
07:52:46 <Bike> american english "d" I guess
07:52:49 <pikhq> Bike: No, but it sounds similarish.
07:52:50 <Bike> god this is hard to talk about
07:53:04 <ais523> but Japanese people will hear the two english letters the same, and thus they can be used interchangeably by an English speaker attempting to speak Japanese
07:53:11 <ais523> they'll be understood equally well either way
07:53:12 <pikhq> Bike: ɾ is the IPA.
07:53:20 <Bike> back in high school I spent like an hour trying to pronounce "akira" japanesely once
07:53:24 <ais523> or equally badly
07:53:32 <madbr> soft: Я Е И Ю (ya ye/yo yi yu) = "ya yeh/yo yee yoo"
07:53:35 * pikhq actually pronounces that sucker right
07:53:39 <Bike> damn you.
07:53:43 <Bike> i just want to talk about my animes!
07:53:47 <pikhq> Though my pitch accent is awful.
07:53:52 <madbr> in dictionaries they spell yo as Ё I think
07:53:57 <Bike> yeah i have no even idea on pitch accent
07:53:58 <Sgeo> My friend's mom (who is a lawyer) told me that I should just take the job (40k) without any negotiation or complaints, because I need a job on my resume
07:54:05 <zzo38> Bike: What animes?
07:54:09 <ais523> (my /actual/ problem is "I have some Cyrillic which I suspect is English words transliterated into Cyrillic, how do I attempt to deduce what the original English was?")
07:54:11 <Bike> i sort of understand how chinese pitch works, i think, but japanese aaaaagh what
07:54:15 <Bike> zzo38: the film, Akira.
07:54:21 <ais523> Sgeo: 40k is more than I'm making
07:54:35 <Sgeo> ais523, erm, in USD?
07:54:37 <pikhq> Bike: The trick with Japanese pitch accent is, it's not as firm as Chinese pitch at *all*.
07:54:38 <madbr> ais523 : read them and figure out how close they are to US?
07:54:41 <ais523> even in USD
07:54:44 <Sgeo> o.O
07:54:48 <ais523> unless the dollar is really low right now
07:54:50 <Bike> pikhq: yeah, that probably doesn't help.
07:54:59 <ais523> I'm hardly paid anything, although hopefully that'll change soon
07:55:03 <pikhq> It's like English stress. Saying it wrong sounds wrong, *but* each accent has their own "right" way.
07:55:14 <ais523> pikhq: and saying it wrong is also entirely intelligible
07:55:15 <Bike> pikhq: it seems like with chinese pitch is just part of the phoneme but with japanese... yeah, like that.
07:55:20 <pikhq> ais523: Exactly.
07:55:33 <Taneb> I have a really odd accent, a lot of people can barely understand me when I start taking quickly
07:55:33 <Bike> well, nothing wrong with sounding like a robot for a while i guess
07:55:56 <Taneb> Perhaps because I drop loads of consonants so that "cone" and "coat" become almost homophones
07:56:10 <ais523> speaking foreign languages isn't the same thing as sounding indistinguishable from a native
07:56:18 <madbr> yeah
07:56:25 <pikhq> ais523: That's the ideal state.
07:56:29 <madbr> if you're close enough, there's no ambiguity
07:56:33 <pikhq> It's just also damn near impossible to get. :P
07:56:42 <ais523> pikhq: it's past the point of return on investment
07:57:09 * pikhq nods
07:57:10 <madbr> even though you're saying one of the sounds wrong, it can't be another sound because it's not closer to the wrong sounds than to the right one
07:57:20 <madbr> like when you're using the wrong kind of "t"
07:57:34 <madbr> sounds wrong but you can't change a word into another one
07:57:38 <ais523> well there are several sounds in English that don't exist in specific other languages
07:57:44 <ais523> e.g. French doesn't have a "th"
07:57:52 <madbr> yeah "th" is the main one
07:57:53 <ais523> but that doesn't prevent native French speakers pronouncing English intelligibly
07:57:55 <Bike> which 'th'?
07:58:00 <madbr> both
07:58:01 <pikhq> Both.
07:58:01 <ais523> Bike: either of them
07:58:04 <Bike> dang
07:58:12 <ais523> actually there are three in English
07:58:14 <ais523> but two of them are quite similar
07:58:15 <madbr> 'th' is actually a fairly rare sound
07:58:22 <madbr> ais: 3?????
07:58:30 <pikhq> ais523: Maybe in your accent.
07:58:30 <ais523> I can't remember the examples offhand
07:58:40 <madbr> 2 certaintly but not 34
07:58:42 <madbr> uh
07:58:44 <madbr> 2 certaintly but not 3
07:58:50 <ais523> pikhq: hmm… I have a surprisingly average accent, actually
07:58:54 <Bike> probably depends on accent
07:59:03 <ais523> one of my parents came from London, the other from Sheffield, and I live in Birmingham
07:59:11 <pikhq> ais523: I'm American. :P
07:59:16 <ais523> oh right
07:59:23 <pikhq> We've actually had some phoneme mergers.
07:59:39 <madbr> the "rare" sounds english has are "th" and the distinction of tense vs lax vowels
07:59:41 <Bike> personally I have an accent pretty close to TV midwestern but apparently there are differences too subtle for me
07:59:44 <ais523> well London is pretty far south, and Sheffield is considered to be northerly unless you live in Hexham
07:59:58 <madbr> (aka long-tense-diphthongs vs short-lax-vowels
08:00:01 <pikhq> Yeah, mine is Midwestern-ish.
08:00:03 <ais523> (Hexham is further north than any sane person would live)
08:00:17 <pikhq> But then, I'm physically Midwestern-ish.
08:00:20 <ais523> (unless they were Scottish, or Finnish, and have even further north people to compare to)
08:00:36 <Taneb> Stupid southeners thinking Sheffield is north...
08:00:41 <madbr> ee/ih, ay/eh, ah/a, o/oh, oo/ooh, uh, er
08:00:49 <ais523> Taneb: yeah
08:00:54 <ais523> well it's north compared to Birmingham
08:00:55 <madbr> also "er" is rare
08:01:02 <madbr> (the english version)
08:01:03 <ais523> I personally think of sheffield as being towards the south end of tnorth
08:01:05 <ais523> *north
08:01:13 <ais523> whereas london is towards the east end of south
08:01:16 <Taneb> It's further south than York!
08:01:17 <Bike> ha, ha, i'm using a non-rhotic accent!!!
08:01:19 <madbr> and the english version of "r" is rare (but "r" sounds in general are common)
08:01:27 <Taneb> And York's like 100 miles to the south of me!
08:01:31 <ais523> Taneb: apparently londoners think of /birmingham/ as north
08:01:45 <Taneb> ais523, that's why nobody likes London
08:01:50 * shachaf keeps misreading "York" as "New York".
08:01:52 <ais523> and possibly even oxford and milton keynes, but that would need verification
08:02:14 <shachaf> Not that I can think of any reason anyone would care about the old one.
08:02:29 <ais523> shViewth: the old one has existe dlonger
08:02:30 <madbr> also, v/w split isn't that common (a lot of languages have only 1 of these)
08:03:05 <ais523> w is arguably a form of u
08:03:09 <madbr> and the "zh" sound in stuff like pleasure, garage... isn't too common either
08:03:22 <shachaf> "shViewth"?
08:03:24 <madbr> ais523 : depends on the language
08:03:37 <ais523> shachaf: while typing that, my client also jumped to the top of the scrollback buffer
08:03:42 <madbr> ais523 : in some languages is pretty much a fast vowel yes
08:03:48 <Taneb> In Latin, u/v/w is one letter, as is i/j/y
08:03:54 <ais523> somehow, I cannot avoid underestimating Konversation's scope for bizarre typos
08:04:01 <madbr> ais523 : but in a LOT of languages, including english, w is a consonant really
08:04:07 <ais523> Taneb: y is a different letter in Greek loanwords
08:04:12 <ais523> isn't it?
08:04:12 <Taneb> ais523, touche
08:04:14 <madbr> and patterns with consonants
08:04:24 <ais523> madbr: w is the consonant form of u
08:04:35 <madbr> Taneb : I'm pretty sure y is another variant of u actually
08:04:37 <ais523> just like y is the consonant form of i
08:04:42 <madbr> and also f
08:04:42 <ais523> or, hmm
08:04:59 <ais523> f isn't a form of u, surely
08:04:59 <madbr> no y is greek upsilon
08:05:11 <madbr> f is from greek digamma
08:05:39 <ais523> oh, I see
08:05:40 <madbr> which is phonecian waw
08:05:48 <Taneb> The symbol comes from digamma, but it's closer to phi nowadays.
08:05:58 <madbr> upsilon is also from phonecian waw
08:06:26 <ais523> well, phi became English ph
08:06:32 -!- Taneb has set topic: #esoteric welcomes its new prime minister, Ed "Brainfuck Derivative" Centipede | Newsflash: every single letter is 'U' | Channel is publicly logged. You can find the URL in the log..
08:06:33 <oonbotti> Nothing here
08:06:35 <ais523> but ph and f are pronounced very similarly
08:06:41 <Bike> http://en.wikipedia.org/wiki/File:Ph%C3%B6nizisch-5Sprachen.svg chart
08:06:45 <Taneb> oonbotti responds to topic changes?
08:06:46 <oonbotti> Taneb: Please consider whether you can answer your own question.
08:06:47 <madbr> so latin had f u (digamma which disappeared from greek, and upsilon basically)
08:06:59 <madbr> and then latin imported y for greek loanwords
08:07:06 <madbr> so it ended up with f u y
08:07:08 <Bike> Taneb: because it starts with a #. dumb huh
08:07:10 <Bike> #hey
08:07:15 <Bike> nno? ok.
08:07:21 <Bike> #esoteric al
08:07:21 <oonbotti> Nothing here
08:07:28 <Taneb> #esoteric Fueue
08:07:29 <oonbotti> Nothing here
08:07:36 <Bike> good bot
08:07:42 <ais523> that's a weird error response
08:08:22 <madbr> then in middle age they started making u pointy when starting words
08:08:39 <ais523> well in Roman times it was always pointy
08:08:51 <ais523> at least in uppercase
08:08:59 <ais523> possibly because it was easier to engrave that way?
08:09:16 <Bike> vvvvvv
08:09:16 <madbr> roman only really had one u sound
08:09:28 <madbr> so it didn't make sense to have two letters
08:10:31 <madbr> but yeah in the middle age languages like french, italian and english had developed a separate v and j sound
08:11:02 <madbr> so people ran with the shape variation and turned them into distinct letters
08:12:17 <madbr> germans came up with a different solution to that problem
08:12:22 <madbr> which is why we now have w
08:12:59 <ais523> although english w != german w
08:13:04 <ais523> *≠
08:14:18 <madbr> in english, w apparently was a replacement for wynn
08:15:46 <madbr> anyways, long story short, western european languages have lots of labial sounds
08:16:07 <madbr> which is why waw has been split into 5 different variants :D
08:17:42 -!- Bike has quit (Quit: ie ie ie ie ie).
08:21:00 -!- sebbu has quit (Ping timeout: 272 seconds).
08:26:16 <Taneb> I would like to put my weight behind Underload as featured language
08:29:09 <ais523> Taneb: the problem is that elliott's no good at writing featured language blurbs
08:29:13 <ais523> and he won't let me write one for my own language
08:29:18 <ais523> so someone else has to
08:29:29 <Taneb> It's time to OPEN A TEXT EDITOR
08:30:37 <Taneb> ais523, was Underload inspired by Muriel?
08:36:02 <ais523> no, it was inspired by Overload
08:36:11 <ais523> which wasn't really inspired by anything
08:36:14 <Taneb> Underload is a stack-based esoteric programming language invented by ais523. It is effectively a functional language, as the only available form of flow control moves data from the data stack into the program. Due to its simplicity, among other factors, it has been useful in proving other languages (such as Thutu and FALSE) Turing-complete.
08:36:29 <ais523> how long is that compared to the existing blurbs?
08:36:48 <Taneb> 57 words
08:36:57 <fizzie> pikhq: http://zem.fi/ircvis/esoteric/people_summary.html has per-day number-of-messages/characters for individual people.
08:37:02 -!- Nisstyre-laptop has quit (Quit: Leaving).
08:37:11 <Taneb> Quite short, actually
08:37:17 <madbr> oh yeah I did a stack based functional language too back in the day
08:37:18 <fizzie> (Hasn't been updated in the last week or two, must do that.)
08:37:26 <Taneb> Malbolge's is 99 words
08:38:05 <Taneb> Although Malbolge's is the longest
08:38:33 <fizzie> (There's also all other kind of graphs there.)
08:38:41 <madbr> mine was called tabarnac
08:39:15 <Sgeo> ais523, did you ever decide if Trustfuck is a metacircular compiler?
08:39:21 <ais523> no
08:39:37 <ais523> Taneb: I'm thinking in terms of visual width, rather than words
08:39:59 <Taneb> Yeah, it's quite short
08:40:18 <fizzie> (Updated.)
08:41:23 <Sgeo> My head hurts from the mathematical stupidity
08:41:28 <Sgeo> http://www.reddit.com/r/philosophy/comments/19jqan/is_pi_proof_of_continuous_space/c8ou5gj?context=3
08:41:47 <Taneb> My throat hurts for reasons I know not
08:42:01 <Sgeo> "A rational number is not necessarily an integer. 3.14 is rational, 3.14 recurring is not. A rational number is one that can be subjected to an expression and give a sane answer. Demonstration: 3.14/5.69 = 17.8666. Both sides of the equation have sane numbers, and are therefore rational."
08:42:19 <Taneb> The first sentence is okay
08:42:25 <Taneb> And the first clause of the second sentence
08:42:29 <Taneb> After that...
08:45:51 <fizzie> http://sprunge.us/NBXA wordnet concurs that rational $\approx$ sane.
08:46:24 <fizzie> I like to call them demythologized numbers, though.
08:46:28 <fizzie> It has more flair.
08:48:06 <Sgeo> I replied
08:48:20 <fizzie> @tell Bike But that's not equivalent at ALL!
08:48:20 <lambdabot> Consider it noted.
08:48:29 <fizzie> (I thought it needed more emphasis.)
08:49:37 <Sgeo> I just realized that the phrase "3.14 recurring" is ambiguous
08:49:48 <Sgeo> Is it 3.141414141414... or 3.1444444444444...
08:50:04 <Taneb> Or 3.14314314314314314314...
08:50:24 <shachaf> Or
08:50:47 <shachaf> Taneb's option doesn't make any sense, though.
08:51:23 <Taneb> My option makes perfect sense when you remember that the person who said "3.14 recurring" is an idiot
08:51:57 <shachaf> makes a lot more sense
08:52:09 <shachaf> (who's the idiot now, huh?)
08:55:28 <Taneb> In other news, there is evidence suggesting Snoop whatever animal he is now reads Homestuck
08:55:47 <fizzie> maybe it goes back and forth.
08:56:29 <madbr> Taneb : as long as it's not mlp :D
08:57:08 <Taneb> madbr, nothing wrong with mlp. At least one person in this channel watches mlp.
08:57:54 <Taneb> And it isn't me!
09:00:33 <madbr> my brother watches mlp :D
09:00:50 <Taneb> Nothing wrong with Snoop Dogg either
09:00:51 <madbr> was saying it jokingly, I can't really criticize tbh
09:01:22 <madbr> there was this "smoke weed every day" meme not too long ago
09:01:30 <madbr> I think it's dead now though
09:01:34 <fizzie> "At least one person in this channel does X" != "nothing wrong with X". (In general, not just about the Window System.)
09:02:56 <Taneb> fizzie, those statements were only loosely related
09:03:08 <madbr> yeah I'd be the biggest hypocrite if I really tarred on mlp stuff
09:04:43 -!- DHeadshot has quit (Ping timeout: 272 seconds).
09:05:55 <zzo38> There is the music in Famicompo with "smoke weed every day", I don't really like that one much.
09:07:42 <madbr> the one I made? :D
09:08:00 -!- doesthiswork has quit (Quit: Leaving.).
09:09:53 <zzo38> Yes. I do like your other musics though.
09:10:23 <madbr> made a really nice song recentily -> http://madbrain.devzero.co.uk/Discohouse_Amargeddon.mp3
09:11:01 <kmc> SMOKE WEED EVERYDAY
09:11:20 <shachaf> It will never die in kmc's heart.
09:11:40 <kmc> isn't snoop dogg named snoop lion now
09:11:48 <shachaf> My flight leaves SFO at ~06:00 on Wednesday.
09:11:49 <Taneb> Maybe
09:11:50 <madbr> yeah and I think he's doing reggae
09:11:51 <zzo38> kmc: Why?
09:11:55 <Taneb> His facebook page isn't
09:11:56 <kmc> just a true fact
09:12:04 <shachaf> Caltrain doesn't go there that early.
09:12:13 <shachaf> Perhaps the best thing to do would be to go the evening before.
09:12:28 -!- HackEgo has quit (Remote host closed the connection).
09:12:32 <kmc> seems like
09:12:49 <kmc> if you want to get to the airport at 04:30 or 05:00
09:12:55 <kmc> where are you going from SFO?
09:13:33 <shachaf> LGA via MDW
09:13:50 <Taneb> madbr, that's some pretty good music
09:13:57 <madbr> thanks :3
09:15:17 -!- HackEgo has joined.
09:15:45 <madbr> taneb : do you compose too? :3
09:15:52 <Taneb> Nah
09:15:55 <madbr> I know some people in here do
09:16:01 <Taneb> I've got no musical talent
09:16:04 <Taneb> Not that I don't try
09:16:42 <zzo38> I sometimes music.
09:16:52 <zzo38> Not a lot.
09:17:24 <zzo38> I did enter Famicompo Mini vol.9
09:18:19 <kmc> one time i went to the bucharest airport in the middle of the night and then slept on top of my bags until i could check in
09:18:22 <zzo38> I do write computer programs for dealing with music, such as VGMCK which is what I am working on now.
09:18:31 <zzo38> kmc: Is it permitted?
09:18:37 <kmc> not sure
09:18:47 <shachaf> kmc: Evidently the universe permitted it to happen.
09:18:51 <shachaf> Er, s/kmc/zzo38/
09:18:59 <madbr> zzo38 : right... I think... gregg does music too right?
09:19:00 <zzo38> Well, yes, the laws of physics permit.
09:19:16 <zzo38> That is not quite how I meant
09:19:19 <zzo38> madbr: gregg?
09:19:24 <madbr> gregor
09:19:32 <zzo38> Gregor? Yes, I think so.
09:20:00 <zzo38> I don't know if he has done it in Lilypond or whatever, I am not sure
09:20:19 <madbr> does vgmck let you use samples for instruments
09:20:23 <madbr> or is it just synthesis?
09:21:01 <zzo38> madbr: VGMCK lets you use the chips in the VGM format, so some chips let you use samples for instruments. Currently the only chip that does that that I have implemented in VGMCK is QSound, although there are others which I have just not implemented yet.
09:21:17 <madbr> how many channels does qsound have again
09:21:21 <zzo38> Sixteen
09:21:43 <madbr> you could probably do something close to the mp3 I linked to then
09:22:05 <madbr> though it has a bit of reverb added on and it has more like 32 channels
09:23:34 <zzo38> OPL4 has both FM and samples, and it has more channels, so that can be used too. In addition, any chips can be combined other than SN76489 with T6W28.
09:23:37 <madbr> goes up to 40 at times (counting "virtual channels")
09:24:05 <madbr> yeah opl4 is cheating too :D
09:24:20 <Sgeo> C# 4.0 has named parameters :)
09:24:25 <zzo38> In addition, most (but not all) chips can be doubled.
09:24:27 <Sgeo> Although it still has ref arguments :/
09:24:35 <madbr> is the 4op mode supported on opl4?
09:24:43 <shachaf> Sgeo: What has it named them?
09:25:49 <zzo38> VGMCK does not currently support OPL4 at all, but I intend to support all the VGM chips, so when it does support OPL4, it will support 4-op mode too.
09:26:01 <nortti> do you know any good tutorials for the so called lisp syntax assembly used in PARRY?
09:26:48 <Taneb> I know someone called Parry
09:27:11 <zzo38> madbr: Do you know the .VGM format? VGMCK compiles into VGM, so you will need a VGM player to play or convert it.
09:27:37 <madbr> only ever used it for... genesis playback I think
09:27:55 <madbr> since that's the platform that has games with interesting soundtracks anyways
09:28:04 <madbr> that and snes but snes has spc
09:28:08 <madbr> and nes has nsf
09:28:53 <zzo38> Well, VGM has a lot of chips (although it does omit some of the NSF chips, but they may be included in later versions): http://www.smspower.org/Music/VGMFileFormat
09:29:28 <zzo38> VGMCK currently supports the following chips: SN76489, T6W28, YM2413, YM2612, YM3812, NES APU, GameBoy, HuC6280, Pokey, QSound.
09:29:43 <zzo38> I am currently working on adding AY8910.
09:30:53 <zzo38> Here is a list of programs which can be used to create VGM: http://vgm.mdscene.net/forum/viewtopic.php?t=107
09:31:50 <madbr> so it lets you write for strange mixes of weird chips
09:32:11 <zzo38> (As far as I can tell, DefleMask allows to use only one chip at a time (from the documentation on their webpage), and I don't think XPMCK allows using more than one chip at a time either but I am not sure.)
09:32:37 <Taneb> Aww, now I'm writing music
09:33:00 <zzo38> madbr: Yes; VGMCK is meant to eventually implement all the VGM and VGM does allow using various mixes of chips.
09:33:26 <madbr> does it let you do stuff like vibratos or do you have to do those with macros like other mck things
09:33:57 <zzo38> madbr: You have to do those with macros, unless the chip supports hardware vibrato in which case you can use that.
09:34:44 <madbr> oh hm
09:36:13 <zzo38> If you want a tracker format for VGM, DefleMask is probably best (not open source, but available for Windows/Mac OS X/Linux), although it is not as powerful as VGMCK, but depending on what you are doing it might work. If you prefer MIDI editors, you can use that too and convert, but those do even less.
09:37:41 * madbr looks up deflemask
09:37:47 <madbr> looks strange
09:38:57 <madbr> hopefully it's less broken than TFM music maker/tracker :D
09:39:26 -!- nooga has joined.
09:42:12 <zzo38> TFM music maker is broken?
09:42:16 <madbr> yes
09:43:03 <madbr> it does everything wrong
09:43:39 <zzo38> Well, now I know, so if anyone asks, I can tell them.
09:51:40 <mroman_> If one'd extend a PDA with registers
09:52:06 <mroman_> Is there an amount of registers < Infinity to make it tc
10:18:48 -!- Taneb has quit (Ping timeout: 264 seconds).
10:19:43 <ais523> mroman_: you need an infinite amount of storage in the registers
10:19:53 <ais523> although you can do that with a finite number of registers, if they're infinitely large registers
10:20:52 -!- madbr has quit (Quit: Radiateur).
10:21:23 -!- Taneb has joined.
10:22:00 -!- epicmonkey has joined.
10:31:40 <ais523> huh, Condorcet was a marquis
10:34:49 -!- zzo38 has quit (Remote host closed the connection).
10:39:55 -!- ais523 has quit.
10:48:44 -!- Taneb has quit (Ping timeout: 245 seconds).
10:53:44 -!- sebbu has joined.
10:53:44 -!- sebbu has quit (Changing host).
10:53:44 -!- sebbu has joined.
10:56:20 -!- nooga has quit (Ping timeout: 252 seconds).
11:05:50 -!- banana_pee has changed nick to nick_exile.
11:06:53 -!- sebbu has quit (Ping timeout: 276 seconds).
11:14:07 -!- Taneb has joined.
11:16:51 <FreeFull> Does lambdabot do rewrite rules?
11:20:11 -!- sebbu has joined.
11:21:09 -!- Phantom_Hoover has joined.
11:24:50 -!- monqy has quit (Quit: hello).
11:27:16 -!- impomatic has quit (Ping timeout: 248 seconds).
11:40:46 -!- nooga has joined.
11:58:15 -!- Arc_Koen has joined.
11:59:11 -!- carado has joined.
12:00:05 -!- nick_exile has quit (Quit: ragequit).
12:00:39 -!- sirdancealo2 has joined.
12:02:35 -!- sirdancealo2 has quit (Read error: Connection reset by peer).
12:03:13 -!- sirdancealo2 has joined.
12:05:09 -!- sirdancealo3 has joined.
12:05:24 -!- sirdancealo2 has quit (Read error: Connection reset by peer).
12:05:24 -!- sirdancealo3 has quit (Read error: Connection reset by peer).
12:05:44 -!- sirdancealot has joined.
12:10:33 -!- nooodl has joined.
12:16:38 -!- Taneb has quit (Ping timeout: 252 seconds).
12:22:32 -!- Arc_Koen has quit (Quit: Arc_Koen).
12:22:49 <mroman_> Does C even support const structs?
12:23:11 <mroman_> C89 that is.
12:24:51 <mroman_> const Foo bar = mkFoo(); is obviously not allowed.
12:25:55 <mroman_> Is {.x
12:26:02 <mroman_> C89?
12:27:53 <c00kiemon5ter> .x ? as in, struct foo f = { .x = .., .y = .., }; ?
12:28:02 <c00kiemon5ter> then no, that is C99
12:28:20 <c00kiemon5ter> designated initializers
12:29:52 <mroman_> c00kiemon5ter: Yeah.
12:30:37 <mroman_> f = {1,2}; should be legale C89 though.
12:30:42 <mroman_> *legal
12:31:12 -!- Taneb has joined.
12:33:14 -!- carado_ has joined.
12:35:56 -!- carado has quit (Quit: Leaving).
13:05:58 -!- Frooxius_ has joined.
13:06:05 -!- Frooxius has quit (Read error: Connection reset by peer).
13:07:24 -!- sebbu2 has joined.
13:07:25 -!- sebbu2 has quit (Changing host).
13:07:25 -!- sebbu2 has joined.
13:07:56 -!- Frooxius has joined.
13:07:59 -!- Frooxius_ has quit (Read error: Connection reset by peer).
13:08:50 -!- sebbu has quit (Ping timeout: 248 seconds).
13:08:57 -!- Frooxius_ has joined.
13:09:04 -!- Frooxius has quit (Read error: Connection reset by peer).
13:09:06 <Vorpal> mroman_, you can't declare the value of y in C89 for union { int x; float y; } though in that manner, due to the lack of the dot-syntax
13:09:07 <lambdabot> Vorpal: You have 1 new message. '/msg lambdabot @messages' to read it.
13:09:14 <Vorpal> @messages
13:09:15 <lambdabot> oerjan said 14d 3h 46m 31s ago: <Vorpal> also my PS3 controller on my desk random came to life and blinked the 4 LEDS a couple of times <-- probably just the google cloud trying to print
13:09:20 <Vorpal> uh
13:09:24 <Vorpal> I got that days ago already
13:09:29 <Vorpal> why are you reminding me of this lambdabot?
13:09:34 <mroman_> Vorpal: My internet research tells me that I can use {foo,bar}; if used in the right order.
13:09:54 <Vorpal> mroman_, for a struct, yes obviously
13:09:59 <Vorpal> not for an union though
13:10:15 <Vorpal> anyway, C89 is obsolete, C99 and C11 are the currently used ones pretty much
13:10:31 <Vorpal> apart from MSVC there isn't much of importance that doesn't do C99 at least
13:11:40 <mroman_> C89 doesn't have function prototypes
13:11:41 <mroman_> what the
13:11:47 <Vorpal> uh it does
13:11:51 <Vorpal> K&R C does
13:12:05 <Vorpal> mroman_, I suggest using a more reliable source than what you are currently using
13:12:14 <mroman_> Yeah.
13:12:23 <mroman_> The first time wikipedia is completely wrong.
13:12:46 <Vorpal> there is your problem, using wikipedia as a source
13:13:24 <mroman_> Meh. It's gve or take.
13:13:29 <mroman_> *give
13:17:29 <Vorpal> mroman_, are you trying to learn C?
13:18:46 <mroman_> No
13:18:49 <mroman_> I know C.
13:18:58 <mroman_> I just don't know which versions of C I know :)
13:19:00 <Vorpal> ah
13:19:41 <mroman_> Which means that I know the C gcc compiles without any specific version options
13:19:54 <mroman_> I just don't know which features are available in what versions.
13:20:33 <Vorpal> I always use -std=c99 -D_POSIX_C_SOURCE=200112L
13:20:48 <Vorpal> should probably move to the next version of POSIX soon
13:20:54 <Vorpal> after all it was released in 2008
13:22:43 <mroman_> but for my compiler I decided to target C89
13:23:12 <Vorpal> you are writing a C compiler?
13:23:13 <Vorpal> ouch
13:23:48 <olsner> C89 is great, it has no features and everything is broken
13:24:21 <mroman_> Vorpal: No
13:24:29 <Vorpal> oh?
13:24:37 <mroman_> I meant I compile my language to C89.
13:24:40 <Vorpal> olsner, that is like all versions of C, more or less
13:25:20 <olsner> C99 is pretty sane, I think
13:25:27 <mroman_> C11 support is not far I assumed.
13:25:43 <mroman_> and some microsoft products apparentely don't even support C99 fully compliant.
13:26:14 <olsner> iirc, microsoft explicitly unsupports all of c99
13:26:20 <mroman_> so I figured C89'd be best so people can use any c89 compliant compiler for the backend
13:26:48 <olsner> (except if c99 accidentally standardized some microsoft extensions, I guess)
13:26:50 <Vorpal> <olsner> C99 is pretty sane, I think <-- really? Files but no directories? No threads? gets still around?
13:27:02 <Vorpal> I haven't really looked at C11 though
13:28:24 <olsner> I count threading and i/o as mostly outside the language (but threading less so, since the behavior of variables with threads is sort of important)
13:40:56 -!- aloril has quit (Ping timeout: 276 seconds).
13:43:56 <Vorpal> olsner, hm
13:44:20 <Vorpal> does anything support C11 yet?
13:45:19 <Phantom_Hoover> wait wtf
13:45:29 <Phantom_Hoover> since when was tom hanks directing an american gods tv series
13:45:46 <Phantom_Hoover> never, apparently, since he's producing it
13:48:31 <Vorpal> mroman_, C11 is at least partially supported (according to wikipedia.....): GCC starting with version 4.6,[5] Clang starting with version 3.1,[6] and IBM XL C starting with version 12.1[7] support some features from C11.
13:52:15 -!- Frooxius has joined.
13:54:45 -!- aloril has joined.
13:57:04 -!- FreeFull_ has joined.
13:59:20 -!- dessos has joined.
13:59:25 -!- epicmonkey_ has joined.
13:59:25 -!- dessos_ has quit (Read error: Connection reset by peer).
14:01:27 -!- Frooxius_ has quit (*.net *.split).
14:01:27 -!- epicmonkey has quit (*.net *.split).
14:01:27 -!- Sgeo has quit (*.net *.split).
14:01:27 -!- FreeFull has quit (*.net *.split).
14:11:37 -!- Sgeo has joined.
14:12:08 <Taneb> The vocab list that OCR publishes for A-level latin is inconsistent
14:12:32 <Taneb> Unless there's a big difference between "verb semi-dep" and "verb semi-dep."
14:12:50 <Taneb> And "verb 2 semi-dep" and "verb 2 semidep"
14:13:28 <Phantom_Hoover> Sure there is.
14:13:39 <Phantom_Hoover> semi-dep means it's half dep, semi-dep. means it's half deponent.
14:16:00 -!- mekeor has joined.
14:17:31 -!- sebbu2 has changed nick to sebbu.
14:23:12 <mroman_> The function `") {"' is applied to three arguments,
14:23:14 <mroman_> o_O
14:28:28 <elliott> "Demonstrating music tech is difficult, because it seems to be impossible to listen to demos without making aesthetic judgements. The below is not meant to be good music, but if you find yourself enjoying any of it, please think sad thoughts. If you find yourself reacting badly to the broken rhythms, try humming a favourite tune over the top."
14:38:51 -!- GOMADWarrior has joined.
14:45:27 -!- Sanky has quit (Ping timeout: 245 seconds).
14:46:21 -!- azaq23 has quit (Max SendQ exceeded).
14:47:30 -!- Sanky has joined.
14:50:30 <mroman_> can ghci tell me which pattern does not exist?
14:55:10 <mroman_> nvm. found the bug.
14:55:23 <elliott> -Wall
14:56:08 -!- Taneb has quit (Ping timeout: 252 seconds).
15:05:14 -!- azaq23 has joined.
15:09:03 -!- wareya has quit (Read error: Connection reset by peer).
15:09:36 -!- wareya has joined.
15:12:04 <mroman_> http://codepad.org/Gy729Qol
15:12:07 <mroman_> ^- well..
15:12:19 <mroman_> The "bootstrap" compiler works :)
15:47:16 -!- quintopi1 has changed nick to quintopia.
15:47:21 -!- quintopia has quit (Changing host).
15:47:21 -!- quintopia has joined.
15:48:31 <quintopia> hi
15:53:44 <mroman_> somehow github only show the first 54 lines o_O
15:54:48 -!- ogrom has quit (Quit: Left).
15:58:13 -!- doesthiswork has joined.
16:09:37 -!- FreeFull_ has changed nick to FreeFull.
16:09:56 <mroman_> > lines "foo\r\n"
16:09:58 <lambdabot> ["foo\r"]
16:10:04 <mroman_> That sucks.
16:10:11 -!- Taneb has joined.
16:16:42 <FreeFull> Because most people who use Haskell do so on Linux
16:16:49 <elliott> um...
16:17:01 <elliott> for a start spj uses windows
16:17:13 <elliott> for a second i'm pretty sure haskell does the C normalisation thing whereby you get "\n" on any platform, though don't quote me on that
16:17:19 <FreeFull> Is spj most people?
16:17:24 -!- impomatic has joined.
16:17:25 <Taneb> Yes
16:25:01 -!- Bike has joined.
16:28:39 -!- sebbu2 has joined.
16:28:56 -!- sebbu2 has quit (Changing host).
16:28:56 -!- sebbu2 has joined.
16:30:34 -!- sebbu has quit (Ping timeout: 246 seconds).
16:34:18 <mroman_> I use windows.
16:35:51 <Lumpio-> Get Linux
16:36:01 <Lumpio-> I mean, I'm sorry to hear that
16:36:19 -!- epicmonkey_ has quit (Quit: Leaving).
16:36:38 -!- epicmonkey has joined.
16:37:07 -!- sebbu2 has changed nick to sebbu.
16:37:07 <Bike> os/2 is the future
16:37:07 <lambdabot> Bike: You have 1 new message. '/msg lambdabot @messages' to read it.
16:37:11 <Bike> whoa
16:37:48 -!- Frooxius has quit (Quit: ChatZilla 0.9.90-rdmsoft [XULRunner]).
16:38:04 -!- Frooxius has joined.
17:06:44 <mroman_> \o/
17:07:00 <mroman_> I can compile a hello world program. Not much. It's something :)
17:10:51 <Sgeo> What language what system?
17:11:46 <mroman_> http://codepad.org/tmyU8LMO
17:12:46 <Sgeo> I don't recognize that language
17:12:56 <mroman_> You can't.
17:13:05 <mroman_> It didn't exist two months ago :)
17:14:13 <mroman_> I'm creating a non-esoteric programming language for a change.
17:14:40 <mroman_> (which I compile to C)
17:16:44 <mroman_> https://raw.github.com/FMNSSun/mopl/master/mopl/hw.c <- that's the output file
17:18:13 <mroman_> ({{ ... }} is inline C)
17:20:21 <mroman_> It's very simple actually.
17:20:32 <mroman_> Just some hacked together ~200 lines of Haskellcode
17:20:49 <mroman_> ok.maybe 400 lines.
17:24:15 <mroman_> well
17:24:24 <mroman_> the ultimate goal is to write a mopl compiler in mopl
17:37:08 -!- zzo38 has joined.
17:37:17 -!- TeruFSX has joined.
17:44:05 -!- daniela has joined.
17:56:05 -!- TeruFSX has quit (Read error: Connection reset by peer).
18:03:19 -!- ogrom has joined.
18:03:20 -!- ogrom has quit (Client Quit).
18:03:40 -!- ogrom has joined.
18:57:48 -!- tswett has quit (Changing host).
18:57:48 -!- tswett has joined.
19:07:59 <kmc> maybe Haskell should have a library like Python's ctypes
19:08:59 <kmc> dynamically loaded, dynamically typed FFI bindings
19:13:16 <zzo38> How does that work? Loading a .so or .dll file?
19:13:54 <kmc> yes
19:14:21 -!- AnotherTest has joined.
19:15:06 <zzo38> How would dynamically typed FFI bindings work though?
19:16:16 <kmc> ctypes.CDLL('libc.so.6').printf("Hello, world!\n")
19:16:23 <kmc> that's how it works in Python
19:16:58 <kmc> well with the standard Haskell FFI, there's no check that the type you declared for a C function matches the type it was compiled with
19:17:05 <kmc> so we just move the not checking to runtime instead of compile time :)
19:17:07 <zzo38> O, OK, I suppose in a programming language such as Python (or other dynamically typed programming languages, such as JavaScript) would make sense like that.
19:17:42 <kmc> loadLibrary :: String -> IO Lib; invoke :: Lib -> String -> [CArgument] -> CResult
19:17:47 <zzo38> What I would like to have is dynamically loaded Haskell programs using a type such as: Typeable x => FilePath -> IO x
19:18:04 <kmc> 'plugins' is supposed to do that
19:18:07 <zzo38> kmc: Yes I suppose that could be the way.
19:18:19 <kmc> or hint or mueval or all these others that i can't keep track fo
19:18:41 <elliott> kmc: you could do an awful pritnf-style hack for nice syntax
19:18:44 <kmc> i think CArgument and CResult are sort of dual there
19:18:54 <elliott> however I don't see how this is really better than the existing FFI
19:18:59 <elliott> after all you can already import dlopen/dlsym and it works fine afaik
19:19:08 <elliott> I think something like that is done for objective-c bidnings
19:19:21 <kmc> CArgument is just a sum of all the C types, but CResult is a thing which, you tell it what type the result should be and it converts
19:19:33 <zzo38> With my idea basically the .so or .dll contains a Haskell value of type x, so it could be a record type, or it could be a IO action in which case you will want to join the result if you want to execute it right away.
19:19:44 <kmc> i find ctypes to be slightly more convenient than Haskell's FFI and i'm trying to figure out why
19:20:17 <kmc> it's probably not a big difference though
19:20:49 <elliott> well I think it is convenient in Python because C APIs are close to what you'd actually use day-to-day in Python :p
19:20:59 <elliott> (this is basically a complaint about python)
19:21:03 <kmc> heh
19:21:11 <zzo38> kmc: While the loadLibrary and invoke could possibly do it, it seems like example you gave is things that would work better in other programming languages such as Python, Perl, JavaScript, and so on.
19:21:33 <kmc> hm my mersenne twister in Python is only 4 times slower than the reference implementation
19:22:36 <kmc> elliott: i think it's just that you're probably doing some type-munging around the C call anyway, so doing that *and* declaring the type in a 'foreign import' declaration seems like boilerplate
19:22:53 <elliott> right
19:23:04 <elliott> you could do some typeclass hackery like I said. it'd be slow runtime though :(
19:23:05 <zzo38> At least I think in JavaScript you could have it work with the same code as the example with Python, even.
19:23:09 <kmc> yeah
19:23:10 <elliott> have you used hsc2hs/c2hs
19:23:13 <elliott> you might find those more convenient
19:23:17 <kmc> yeah i used hsc2hs
19:23:18 <kmc> it's all right
19:23:25 <kmc> hdis86 uses it
19:23:28 <elliott> c2hs is more elaborate fwiw
19:23:32 <kmc> yeah
19:23:36 <kmc> i read ezyang's blog posts about it
19:23:55 <kmc> i wrote that library on fung wah bus
19:24:54 -!- daniela has left.
19:25:27 <FreeFull> kmc: Now write it in befunge
19:31:18 <kmc> mersenne twister?
19:33:15 <Bike> mersenne twister shaped like a twister
19:34:00 <kmc> rhymes that blister like mersenne twister
19:39:41 -!- mekeor has quit (Read error: Connection reset by peer).
19:42:04 <Taneb> Preach it, sister
19:49:01 -!- azaq23 has quit (Quit: Leaving.).
20:12:45 -!- oerjan has joined.
20:14:03 <oerjan> <ais523> theory: it is substantially easier for something to be awesome if it has no purpose other than awesomeness
20:14:09 <oerjan> <shachaf> ais523: I would imagine the opposite.
20:14:23 <oerjan> i submit this is one of those Great Truths
20:14:46 <AnotherTest> the former or the latter?
20:14:57 <Taneb> The two lines as a whole, of course
20:15:00 <oerjan> ...i see you don't know what a Great Truth is.
20:15:10 <AnotherTest> I clearly do not!
20:15:15 <AnotherTest> Please explain
20:15:25 <oerjan> it's a truth such that its opposite is also a Great Truth
20:16:01 <elliott> i see oerjan is a dialetheist
20:16:26 <oerjan> ah Niels Bohr was the origin of that
20:16:35 * oerjan tried Piet Hein first
20:16:48 <oerjan> (also a dane with a penchant for great truths)
20:18:15 <oerjan> hm i don't trust that quote site to have the correct wording...
20:18:30 * oerjan tries wikiquote instead
20:18:35 <AnotherTest> oerjan: Your definition of "Great Truth" sounds like a recursive definition without an edge condition
20:19:04 <oerjan> "Two sorts of truth: profound truths 
recognized by the fact that the opposite is also a profound truth,
 in contrast to trivialities where opposites are obviously absurd."
20:19:09 <Bike> it's an old quote
20:19:30 <AnotherTest> a nice one
20:19:42 <oerjan> (that's the wikiquote version which actually is cited)
20:20:53 <zzo38> Yes I have seen that before
20:21:00 <AnotherTest> so do we also have an actual list of Great Truth?
20:21:05 <AnotherTest> *Truths
20:21:16 <elliott> i'm not sure you're quite on the wavelength of the idea here
20:22:03 <oerjan> AnotherTest: i guess you need to take the corecursive interpretation (i.e. maximal possible set) in that case.
20:22:15 <oerjan> since the minimal one, by what you say, would be empty.
20:23:36 <AnotherTest> Oh, like, the universal set?
20:23:38 <zzo38> If any Great Truth are logical then you need to use different kinds of logic.
20:23:46 <Bike> youo're not supposed to have a list, it's just supposed to be an insult, what the hell
20:23:55 -!- nooodl has quit (Read error: Connection reset by peer).
20:24:31 <AnotherTest> zzo38: p <=> not p is definitely not true indeed
20:24:34 -!- nooodl has joined.
20:25:17 <AnotherTest> can unknown be equivalent to not unknown even?
20:25:28 <AnotherTest> (ternary logic then)
20:25:56 <Bike> i thought ¬u = u was a given of that logic
20:26:35 <oerjan> AnotherTest: no, not the universal set, it doesn't not include falsities or their opposite
20:26:40 <oerjan> oops
20:26:42 <oerjan> *-not
20:27:03 <oerjan> (maybe that was still correct on account on being a great truth)
20:29:12 <oerjan> i think to understand great truths, you first have to understand that logic has its limitations.
20:29:37 <oerjan> and not just to understand that they exist, but to understand a particular one.
20:29:59 <oerjan> also how can i say that when i still haven't thought of an example.
20:30:00 <zzo38> At least I know that logic has limitations.
20:30:16 <AnotherTest> Isn't it like a great truth that great truths are /really/ great truths?
20:30:30 <zzo38> But there is different kind of logic, in order to make some things that works in some cases.
20:30:42 <oerjan> maybe zen koans
20:30:45 <nooga> hande hoch!
20:31:07 <oerjan> but since niels bohr was involved, probably also something quantum
20:31:08 <zzo38> But a kind of logic can be mathematically logical even if it doesn't apply to actual situations normally.
20:31:15 <oerjan> nooga: wie so?
20:31:34 <zzo38> oerjan: Yes, perhaps Zen koans would be called great truth by such definition, possibly.
20:31:43 <elliott> oerjan: you just need to use a paraconsistent logic
20:32:07 <Bike> what is the truth value of cutting off your disciple's arm
20:32:36 <zzo38> oerjan: Quantun? I guess so, although it still isn't applicable to measured (collapsed) states, I suppose, even though it is in general, possibly?
20:32:39 <nooga> das ist ein internationale skandal!
20:32:46 <zzo38> Bike: Forty-two!
20:32:53 <Bike> humorous
20:32:54 <zzo38> But only if you don't do it.
20:33:19 <oerjan> Bike: Mu. see: [Parachattavadhari 451, page 701]
20:33:52 <Bike> Is that gobbledygook that's supposed to look like Sanskrit?
20:34:25 <oerjan> wtf i keep being logged out of wikipedia
20:34:26 <zzo38> I do know one word in Sanskrit, at least.
20:34:37 <oerjan> maybe this is related to that oots issue
20:34:55 <zzo38> Does the cookie expire?
20:35:32 <oerjan> i have ticked the usual "remember for 180 days" mark
20:36:04 <zzo38> Did you try editing the cookie? I have found that for some webpages it won't work properly unless the cookie is edited.
20:36:05 <oerjan> Bike: MAYBE
20:36:17 <zzo38> Do you even know how to write Sanskrit?
20:36:24 <oerjan> no.
20:37:15 <oerjan> i just have some vague idea about the phonology and maybe a handful of words.
20:37:27 <Bike> Wouldn't Zen be in Japanese or Chinese anyway
20:37:32 <oerjan> ok maybe not _that_ vague
20:37:53 <zzo38> "Ayanamsha" is Sanskrit words for amount of precession of equinoxes.
20:37:54 <oerjan> Bike: japanese, i think the chinese is slightly different
20:38:21 <oerjan> zzo38: didn't you mention that before, or was that another astronomy word
20:38:26 <Bike> like, gateless gate is in chinese isn't it
20:38:40 <zzo38> oerjan: I don't know. Probably I did.
20:38:46 <Taneb> There's one of them in Hexham
20:38:54 <Taneb> A gateless gate
20:38:58 -!- ogrom has quit (Quit: Left).
20:39:03 <zzo38> But I don't know if any astronomers use such words unless they are Sanskrit or astrologers.
20:39:17 <zzo38> (or both)
20:39:39 <Bike> How would an astronomer be Sanskrit?
20:40:29 <zzo38> I don't know; I am not Sanskrit.
20:40:35 -!- AnotherTest has quit (Quit: Leaving.).
20:40:37 <oerjan> the subjects weren't different back then
20:40:42 <Bike> I mean, it's not an ethnicity
20:41:11 <zzo38> Bike: O, OK, then.
20:41:26 <zzo38> I mean, if they speak/write Sanskrit then they might use such words.
20:41:52 <Bike> That would be like doing astronomy in latin. (Which sounds amusing)
20:41:55 <oerjan> ok at least closing and reopening the browser didn't lose the login. maybe it was connected to how my hibernation image was broken...
20:42:21 <oerjan> or that it's somehow kept different for http and https
20:42:48 <zzo38> Latin is the language of Vatican.
20:42:58 <oerjan> THERE IS NO POPE OMG
20:43:02 <oerjan> WTFBBQ
20:43:12 <zzo38> There is still Vatican, even if there is currently no pope.
20:43:20 -!- epicmonkey has quit (Ping timeout: 260 seconds).
20:43:24 -!- nooga has quit (Ping timeout: 264 seconds).
20:46:17 <elliott> so why did vatican delete all benedict's tweets
20:46:36 <Bike> papal fallibility
20:47:04 <elliott> but they're archived!!
20:47:53 <oerjan> archivandi in gloria aeternis
20:48:03 <zzo38> Maybe it is because it was the Vatican's account and not the pope's own account?
20:48:11 <zzo38> Is that why?
20:48:33 <elliott> imagining the pope having a personal twitter account now
20:48:42 <elliott> for all his non-catholic church related doings(???)
20:49:05 <kmc> read 'church related dongs'
20:49:30 <Bike> ratzinger's zingers, an account devoted entirely to cheesy insults
20:49:49 <zzo38> Even if he has a Twitter account which does have stuff about the Catholic Church, then, if it is his own account then they should not delete it. If it is belonging to the Vatican, then they have the right to delete it. Isn't it?
20:49:55 <elliott> kmc: insert predictable joke re: catholic church
20:51:07 <Bike> joke manipulator arm now online. please clear a 500m standard safety area around joke insertion slot. commencing insertion...
20:51:54 <zzo38> It is probably farther away than that from my computer.
20:52:09 <Bike> you're probably going to get a concussion then
20:54:31 <zzo38> Well, I don't know where it even is (or even whether or not you made such a thing), so it won't help anyways.
20:55:21 <elliott> tell that to your concussion
20:57:12 -!- Zerker has joined.
20:57:33 <zzo38> Does any Famicom emulator be able to read the host system's battery function to set the FDS battery low bits?
21:00:33 <kmc> i should have followed up with "i guess that's more a tumblr thing"
21:00:44 <kmc> but i was already in the shower by the time I thought of this witty remark :/
21:02:50 <oerjan> `addquote <Bike> kmc: ok i gotta admit i fully support the use of ultrafilters in social sciences
21:03:00 <HackEgo> 978) <Bike> kmc: ok i gotta admit i fully support the use of ultrafilters in social sciences
21:03:15 -!- nooga has joined.
21:05:02 <elliott> wait where does the remark go in context
21:06:07 <doesthiswork> evolveing a program to to give me the nth digit of the thule-morse sequence when given the binary representation of n is very hard if I only allow it to use nand gates
21:06:09 <Vorpal> <elliott> for all his non-catholic church related doings(???) <-- uh, what does that consist of?
21:06:14 <elliott> Vorpal: (???)
21:06:22 <Vorpal> elliott, ah, of course
21:07:17 <Vorpal> elliott, but more seriously is he known for having some side interests at all?
21:07:29 <Vorpal> Maybe, uh, catching butterflies or something+
21:07:33 <Vorpal> s/+/?/
21:07:39 -!- FireFly has quit (Ping timeout: 276 seconds).
21:07:51 <Vorpal> surely he must, as a person, have some hobbies
21:09:27 <elliott> i don't think you quite understand how churches work
21:10:10 <Vorpal> elliott, that is probably true
21:13:08 * oerjan swats Taneb for converting `? ais523 from utf-8 to latin-1
21:13:25 <oerjan> `? ais523
21:13:26 <HackEgo> Agent "I" Smith is an alien with a strange allergy to avian body covering, which he is trying to retroactively prevent from ever evolving. On the 3rd of March, he's lawful good.
21:13:33 <oerjan> * + -----###
21:14:00 <Sgeo> What would Chaotic Evil ais523 look like?
21:14:19 <Bike> kind of person who knocks over unattended cups i'd wager
21:15:18 <oerjan> hm my browser's autocomplete database has reset too
21:15:49 <oerjan> `url
21:15:51 <HackEgo> http://codu.org/projects/hackbot/fshg/
21:16:50 <oerjan> `run echo "Agent \"Iä\" Smith is an alien with a strange allergy to avian body covering, which he is trying to retroactively prevent from ever evolving. On the 3rd of March, he's lawful good." > wisdom/ais523
21:17:00 <HackEgo> No output.
21:18:01 <Vorpal> * oerjan swats Taneb for converting `? ais523 from utf-8 to latin-1 <-- why would he do such a crime
21:18:30 <oerjan> Vorpal: presumably by cutting and pasting in his irc client with a bad charset setting
21:18:36 <Vorpal> ah
21:18:48 <Vorpal> what a heinous crime indeed
21:19:00 <Vorpal> `? ais523
21:19:02 <HackEgo> Agent "Iä" Smith is an alien with a strange allergy to avian body covering, which he is trying to retroactively prevent from ever evolving. On the 3rd of March, he's lawful good.
21:19:44 <oerjan> <Vorpal> surely he must, as a person, have some hobbies <-- istr he plays the piano?
21:20:32 <Vorpal> oerjan, Ah
21:20:42 <Vorpal> could microblog on twitter about the piano
21:20:59 <oerjan> <Sgeo> What would Chaotic Evil ais523 look like? <-- he would actually _succeed_ in retroactively unevolving avian body covering, although at terrible cost.
21:21:39 <Vorpal> why specifically on the 3rd of March?
21:21:49 <Vorpal> hasn't ais been basically lawful good forever
21:22:05 <oerjan> 23:29:10 <ais523> hmm… today seems to be the ais523 "have I mentioned I'm lawful good yet" day
21:22:08 <oerjan> 23:29:17 <ais523> 3 march, I must remember that
21:22:26 <oerjan> Taneb _may_ have slightly misinterpreted.
21:22:53 <Vorpal> ah
21:23:03 <Sgeo> How would ais523 react if stuck in a Lawful Evil area?
21:23:09 <Vorpal> oerjan, yeah I remember we established his alignment long ago
21:23:11 <Sgeo> Where Lawful contradicts Good
21:23:37 <Vorpal> Sgeo, he would obviously not be ais any longer
21:23:41 <Bike> he sputters out and then starts shouting about diathelism
21:25:50 <Sgeo> http://chaospet.com/2007/11/29/62-dialetheism/
21:26:05 <Bike> yes i can't spell WHAT OF IT
21:26:20 <elliott> imo i'm a dialatheist
21:26:26 <elliott> I believe god exists AND doesn't exist
21:26:30 <Sgeo> Bike, didn't even notice
21:26:51 <Bike> that comic is pretty sophistic but what else is new
21:27:03 <Sgeo> Although.... I guess that's why Google only turned up the comic
21:27:12 <Sgeo> Weird though, since the comic spelled it correctly
21:27:17 <Bike> i think graham priest is the usual go-to guy
21:27:27 <elliott> he's the high priest of dialatheism
21:27:28 <Sgeo> No it didn't
21:27:49 <Sgeo> http://chaospet.com/2009/06/15/128-more-dialetheism/ (the link I originally saw) contains BIke's misspelling
21:27:50 <Bike> "In addition to his work in philosophy and logic, Priest practices Karate-do. He is 3rd Dan, International Karate-do Shobukai; 4th Dan, Shi’to Ryu, and an Australian National Kumite Referee and Kata Judge."
21:28:25 <Sgeo> I seriously thought that the whole diathelism thing was a reference to that comic.
21:28:37 <Bike> i've never seen that comic before and it looks bad
21:28:56 <Sgeo> Blame Google.
21:29:02 <Bike> f u google
21:29:02 <Sgeo> And yourself for misspelling.
21:29:07 <Bike> f u bike
21:29:12 <elliott> fugle
21:32:15 <Sgeo> Bluh. What's that really famous song, I think it's a Kraftwerk song, that's ohwoahohwoahoh
21:32:28 <elliott> um
21:32:40 <Bike> oh yeah, and it goes like, doooOOOooooOOOOOoo?
21:32:57 <elliott> http://answers.yahoo.com/question/index?qid=20080101093149AAhTBGF
21:32:59 <Bike> duh, duh duh duh duh, duh duh, duh duh duh duh duh duh DUH, duh DUH
21:33:05 <elliott> i googled: oh woah kraftwerk
21:33:27 <Bike> elliott: i refuse to believe that works
21:33:33 <Bike> sgeo listen to that song so you can say it's not it
21:33:40 <elliott> it links to several tracks in fact
21:33:43 <elliott> one of them is even by kraftwerk
21:34:31 <Sgeo> The beginning is unfamiliar, but I think it's the song
21:34:38 <Sgeo> Where TF did I get "kraftwerk" from?
21:34:38 <Bike> Fuck.
21:34:41 <nooodl> `run (echo '#!/bin/sh'; echo 'tr A-Z U | tr a-z u') > bin/uuu; chmod a+x bin/uuu
21:34:46 <HackEgo> No output.
21:34:51 <nooodl> `run welcome | uuu
21:34:53 <HackEgo> Uuuuuuu uu uuu uuuuuuuuuuuuu uuu uuu uuuuuuuu uuuuuuuuuuu uuuuuuuu uuuuuu uuu uuuuuuuuuu! Uuu uuuu uuuuuuuuuuu, uuuuu uuu uuu uuuu: uuuu://uuuuuuuu.uuu/uuuu/Uuuu_Uuuu. (Uuu uuu uuuuu uuuu uu uuuuuuuuu, uuu #uuuuuuuu uu uuu.uuu.uuu.)
21:34:58 <elliott> didnt you know every piece of electronic music is by kraftwerk
21:35:03 <elliott> that's a science fact™
21:35:10 <elliott> also they invented robots
21:35:30 -!- DHeadshot has joined.
21:35:37 <Taneb> Trivia: there are actually robots in the Illiad
21:35:45 <Sgeo> I think I only thought I knew Kraftwerk in relation to that song, which isn't actually a Kraftwerk song
21:35:51 <Bike> The ones on the island with the labyrinth?
21:35:55 <Sgeo> Except, hmm
21:36:05 <Sgeo> I don't actually hear the vocals in this version
21:36:08 <elliott> wir fahren fahren fahren auf der autobahn
21:36:12 <elliott> did i spell that right
21:36:21 <Bike> faßren
21:36:24 <elliott> fuck you
21:36:25 <Bike> hth
21:36:31 <Bike> :(
21:36:48 <Sgeo> Bike, if it helps, this might be the wrong version of the basic concept of the song
21:36:57 <elliott> the wrong version of the basic concept of the song...
21:36:59 <Bike> no, it's too late
21:37:17 <elliott> anyway if it is "Kernkraft 400" i have a guess why you related it to kraftwerk
21:37:18 <elliott> the guess is "kraft"
21:37:29 <elliott> Ohohohoh ohoh.song
21:37:29 <elliott> that is what is is called on limewire
21:37:29 <elliott> Source(s):
21:37:29 <elliott> Hockey songs --- limewire
21:37:50 <Bike> why don't you listen to some REAL german electronica http://www.youtube.com/watch?v=k6zb8YpL5-w
21:38:17 <nooodl> http://www.youtube.com/watch?v=YsorGbKwNlA#t=22s
21:38:34 <Sgeo> I don't hear any actual oh woah ing in here http://www.youtube.com/watch?v=bSJQLCImV18
21:38:54 <Sgeo> Electronic sounds are doing the same melody that oh'ing should be
21:39:04 <elliott> maybe you listened to a choir singing it
21:39:21 <nooodl> http://www.youtube.com/watch?feature=?v=XPoiPZYbtnA#t=75s Sgeo
21:39:48 <Bike> elliott: http://www.youtube.com/watch?v=9fwjZFLmrpE i have no idea why it's real but it is
21:39:53 <nooodl> bjornsuper 6 dagen geleden
21:39:53 <nooodl> 0 to 1.15 complet shit rest of the song annihilates
21:40:01 <nooodl> thanks for the review, bjornsuper
21:40:22 -!- Taneb has quit (Quit: Leaving).
21:40:28 <Sgeo> nooodl, yes
21:40:53 <nooodl> oops i messed up that link
21:40:56 <nooodl> did you find it anyway
21:40:58 <elliott> Bike: is this from that album
21:41:01 <elliott> ohhh
21:41:02 <elliott> it's that thing
21:41:05 <Sgeo> yes, I deleted feature=?
21:41:06 <elliott> heheh
21:41:11 <Bike> yeah it's from that one album of that one thing you got it
21:41:19 <elliott> i think that is actually from an ep actually
21:41:29 <elliott> but what i meant was uh
21:41:42 <elliott> god dammit go faster browser
21:41:48 <elliott> im going to die
21:42:45 <Bike> also good: vocal guitar http://www.youtube.com/watch?v=Z7Asi870JpI
21:42:54 <elliott> http://en.wikipedia.org/wiki/Acoustica:_Alarm_Will_Sound_Performs_Aphex_Twin
21:43:13 <Bike> always laughing and laughing loud
21:43:18 <elliott> except i think i also meant something else
21:43:35 <elliott> whatever fuck this i'm going away for quarter of an hour, back when i don't hate everybody
21:43:38 <Bike> this sounds complicated elliott
21:43:57 <hagb4rd> http://miburl.com/3tYFIf darn.. i need a premium account at all costs
21:44:10 <oerjan> <elliott> fuck you <-- Bike was lying hth
21:45:45 -!- sebbu has quit (Ping timeout: 260 seconds).
21:46:38 <Bike> so h
21:52:47 <oerjan> <fizzie> (I thought it needed more emphasis.) <-- O KAY
21:55:12 -!- nooodl has quit (Read error: Connection reset by peer).
21:57:31 <oerjan> > let stupid = 3 {-# RULES "stupid" stupid = 2 #-} in stupid
21:57:32 <lambdabot> Terminated
21:57:35 <oerjan> > let stupid = 3 {-# RULES "stupid" stupid = 2 #-} in stupid
21:57:37 <lambdabot> <hint>:1:16: parse error on input `{-# RULES'
21:57:55 <oerjan> FreeFull: NOPE
21:59:06 <elliott> Bike: what sounds complicated.
21:59:26 <Bike> that one album of that one thing
22:00:00 <oerjan> hm
22:00:05 <oerjan> > let stupid = 3; {-# RULES "stupid" stupid = 2 #-} in stupid
22:00:07 <lambdabot> <hint>:1:17: parse error on input `{-# RULES'
22:00:21 <oerjan> > let stupid = 3 in {-# RULES "stupid" stupid = 2 #-} stupid
22:00:22 <lambdabot> <hint>:1:19: parse error on input `{-# RULES'
22:00:37 <Bike> stupid
22:01:22 -!- nooga has quit (Ping timeout: 251 seconds).
22:01:30 <FreeFull> oerjan: Oh well
22:02:32 <oerjan> FreeFull: although i don't know if RULES work outside top level anywhere
22:02:53 -!- sebbu has joined.
22:02:53 -!- sebbu has quit (Changing host).
22:02:53 -!- sebbu has joined.
22:02:58 <elliott> Bike: some of their stuff is pretty impressive
22:03:16 <Bike> whose stuff
22:03:24 <oerjan> @let stupid = 3; {-# RULES "stupid" stupid = 2 #-}
22:03:24 <lambdabot> Invalid declaration
22:03:32 <oerjan> STUPID
22:03:53 <Sgeo> I've heard of a song called "Guile's theme", but why does a dialect of Scheme have a theme?
22:04:07 <Bike> bad joke, retire it
22:04:12 <elliott> was that meant to be a joke..
22:04:36 <Sgeo> I did gather that there's a game or something called Guile, but I can't help but think of the Scheme dialect
22:04:48 <Bike> it's a character in a video game
22:05:02 <Bike> also it's a word? you know, "guile", like ability to trick people
22:05:30 <oerjan> <Vorpal> why are you reminding me of this lambdabot? <-- i think this may be "things reset for no sensible reason day". as well as ais523 lawful good day, of course.
22:05:34 <Sgeo> Yes, but words don't typically get theme songs either
22:06:02 <Bike> but sometimes... sometimes, more than one thing may be named after... /the same word/
22:06:19 -!- Arc_Koen has joined.
22:07:16 <Sgeo> Did I say anything along the lines of "The character must be named after the Scheme dialect!" or "The Scheme dialect must be named after the character!"/
22:07:27 -!- ais523 has joined.
22:07:37 <Bike> definitely
22:07:43 <elliott> you know generally you don't have to have an argument for your joke to work :P
22:08:09 <oerjan> in fact it doesn't help.
22:08:17 <oerjan> at all.
22:08:21 <oerjan> hth.
22:08:50 <elliott> oerjan: in fact it didn't help.
22:08:52 <elliott> at all.
22:08:54 <elliott> hth.
22:09:01 <oerjan> thx.
22:09:01 <elliott> recursion
22:09:34 <Bike> we've obtained rapid fire device
22:09:42 <oerjan> hth, thx, hth.
22:10:06 -!- nooga has joined.
22:10:40 <zzo38> How efficient would be a Verilog implementation of the INTERCAL select ~ operator and its (almost) inverse (setting the unused bits to zero)?
22:11:11 <ais523> zzo38: you could do it in one cycle, I don't know what the maximum clock speed would be
22:11:30 <ais523> probably you could get a very high clock speed if it was completely forwarded, but that might use a prohibitive amount of silicon
22:11:48 <ais523> (forwarding is sort-of the hardware equivalent of unrolling)
22:11:56 <zzo38> OK
22:12:11 <ais523> I think you could probably do it in four to six stages and still fit into one clock cycle
22:14:57 <zzo38> (If you use both operations together, it acts like a bitwise AND; you can also do various other things with these operations together in different ways.)
22:16:51 <zzo38> You could also use them to encode and decode Morton numbers if you put -2/3 and -1/3 on the right operand.
22:20:55 <zzo38> I really think they would be useful for various things. How would it affect existing processor cores if such an instruction was added in the Verilog code?
22:21:35 <ais523> zzo38: the major change would be in the decoding logic, I think
22:21:46 <ais523> it's not significantly slower, or different, to an addition, from the point of view of implementation
22:22:29 <coppro> hmmmmmmmmmm
22:22:39 * ais523 wonders if modern processors use single-cycle multipliers
22:22:57 <ais523> they exist but are really expensive in silicon
22:23:16 <ais523> however, I don't think the silicon use would be significant compared to, say, the L1 cache
22:24:20 <zzo38> Due to features like these, as well as other features such as BCD arithmetic, I would want a !asm metadata in LLVM which tells it that the result can be calculated using the specified assembly language instructions, even though there is a code or function or something calculating the same thing in the circumstances; this can be used for optimization.
22:27:44 <zzo38> ais523: I don't know if x86 does, but Amber uses Booth's multiplication and I don't know how many cycles it requires.
22:27:45 -!- sebbu has quit (Ping timeout: 260 seconds).
22:27:45 -!- Zerker has quit (Quit: Colloquy for iPad - Timeout (10 minutes)).
22:29:05 <zzo38> O, I found it; it is 34 clock cycles to multiply.
22:29:06 <Fiora> isn't the number of cycles for a multiplier dependent on the depth of the circuit (versus, like, the transistor depth for a cycle?)
22:29:11 <Fiora> 34? o_O
22:29:28 <ais523> Fiora: it depends on the number of registered stages in the circuit
22:29:28 <Bike> better recode that shit as shiftadds
22:29:48 <ais523> but that has a pretty strong correlation to the depth, because if your depth is too large compared to the number of stages, you have to slow down your clock speed to compensate
22:29:50 -!- Arc_Koen has quit (Quit: The struct held his beloved integer in his strong, protecting arms, his eyes like sapphire orbs staring into her own. "W-will you... Will you union me?").
22:29:55 <Fiora> ais523: yeah, that's what I meant, sorry
22:30:00 <Bike> he's going to keep that for a while isn't he
22:30:00 <ais523> and modern processors try to run at the highest clock speed they physically can
22:30:14 -!- Arc_Koen has joined.
22:30:15 <Fiora> just like, I figure a 64x64->128 1-cycle multiplier is probably not possible with the cycle depth required by like, an ivy bridge?
22:31:10 <ais523> Fiora: it is, there's a famous construction that shows that you can do any combinatorial circuit in one stage with just two levels of nand gates
22:31:39 <ais523> but you get a combinatorial explosion so it's not a useful one in practice, unless you can really spare the silicon and really need speed
22:31:52 <ais523> it's done for additions sometimes
22:32:33 <olsner> at which point does the size of the silicon start to limit what you can do? e.g. because of the speed of light being too low
22:32:49 <Fiora> that seems impossible...? like. two levels of nand gates means at most 4 bits of input
22:32:51 <ais523> olsner: we've hit that point already
22:33:00 <Fiora> and you can't figure out one bit of output in a 128-bit multiply from just 4 bits of input
22:33:04 <ais523> Fiora: very wide nand gates
22:33:07 <Fiora> oh.
22:33:16 <Fiora> what about actual 2-input gates <.<
22:33:25 <ais523> hmm
22:33:28 <ais523> I don't know offhand
22:33:29 -!- sebbu has joined.
22:33:29 -!- sebbu has quit (Changing host).
22:33:29 -!- sebbu has joined.
22:33:49 <ais523> but 3-input nand gates aren't particularly harder to build than 2-input nand gates, apart from having to try to connect three inputs to them
22:33:57 <ais523> routing overhead is one of the major problems with modern hardware design
22:33:58 <Fiora> that seems like the thing that actually matters? since like, more and more inputs means more delay doesn't it?
22:34:14 <Fiora> a 128-input nand gate seems pretty crazy XD
22:34:16 <ais523> not noticeably, apart from the routing problem
22:34:27 <ais523> fwiw, six inputs seems to be the current standard in FPGAs
22:34:40 <ais523> it used to be four at one point, and manufacturers are probably moving to higher numbers at the moment
22:35:51 <Fiora> The fastest multiplier I know is I think the ivy bridge multiplier
22:36:01 <zzo38> In order to save energy and to be compatible with programs requiring slower clock speeds, there could be I/O ports to adjust the clock speed to slow it down if it is wanted; does this work?
22:36:03 <Fiora> at least, like, in an actual hcip
22:36:06 <Fiora> *chip
22:36:15 <ais523> Fiora: what I know is that if I accidentally ask for a single-cycle 32-bit multiply when writing a Verilog or VHDL program, the synthesizer knows how to handle it
22:36:27 -!- Zerker has joined.
22:36:31 <Fiora> though I still kinda wonder why they chose to allocate their gates like that but others didn't
22:36:43 <zzo38> How does the floating point coprocessors work in ARM?
22:37:53 <Fiora> ais523: does the synthesizer have, like, options that let you pick the sort of multiplier you want?
22:38:00 <Fiora> I guess in theory you could write your own...
22:38:07 <ais523> Fiora: the synthesizer implements what you write
22:38:15 <Fiora> I meant for like, how it implements a "*"
22:38:29 <ais523> yeah, * is single cycle, it has to be because there's no delay around it or the like
22:38:39 <ais523> for longer multiplies, you instantiate modules from elsewhere
22:38:56 <zzo38> Does Verilog even allow * in compiled codes?
22:39:01 <ais523> like, you get a 32-cycle multiply from your standard library
22:39:04 <ais523> zzo38: yes, but it doesn't allow /
22:39:12 <Fiora> I think so... we used * and stuff in our MIPS chip in class...
22:39:29 <ais523> if you want division, you need to get it from a library or implement it yourself
22:39:45 <Vorpal> <oerjan> <Vorpal> why are you reminding me of this lambdabot? <-- i think this may be "things reset for no sensible reason day". as well as ais523 lawful good day, of course. <-- it reset for no sensible reason?
22:39:49 <Vorpal> sounds buggy
22:39:50 <zzo38> Does it allow / in stuff other than compiled codes though? (such as in macros and constant expressions)
22:40:12 <ais523> zzo38: I don't know, I've never tried
22:40:19 <ais523> I think it allows exponentiation in macros and constant expressions
22:40:29 <ais523> and division is rather easier than exponentiation, so probably
22:41:31 <Vorpal> `? ais523
22:41:33 <HackEgo> Agent “Iä” Smith is an alien with a strange allergy to avian body covering, which he is trying to retroactively prevent from ever evolving. On the 3rd of March, he's lawful good.
22:41:36 <Vorpal> ais523, I like that one ^
22:42:22 <zzo38> If it was up to me, no arithmetic operations would be allowed in compiled codes; they would be allowed only in constant expressions, macros, selectors, and ROM data; and otherwise you have to specify your own algorithms in terms of bitwise.
22:45:05 <zzo38> I also would think their ought to be commands which means, that it would optionally use precompiled block in place of the specified block, which can be automatically selected based on the optimization and conditional compiler options, to ensure vendor-independent codes and still allow vendor-specific optimizations.
22:46:01 <zzo38> Does it have that?
22:46:02 <Vorpal> night →
22:46:30 -!- Zerker has quit (Quit: Colloquy for iPad - Timeout (10 minutes)).
22:48:40 <Sgeo> elliott, where's your link about files?
22:48:50 <elliott> um
22:48:51 <elliott> which one
22:48:55 <elliott> which thing..... which link
22:49:04 <Bike> which files
22:49:05 <Sgeo> I think the one that's anti-file
22:50:34 <zzo38> If the coprocessors are omitted from Amber core, instead making the cache and privilege areas to be hard-coded, will it make the clock speed faster?
22:50:45 <oerjan> <ais523> [...] with just two levels of nand gates <-- that's just disjunctive normal form, i think. (unless i got confused and it's conjunctive instead.)
22:51:01 <ais523> oerjan: yes, pretty much that
22:51:12 <ais523> right down to the not being able to remember which is which
22:55:26 <oerjan> <Vorpal> sounds buggy <-- it's happened before that lambdabot messages get reset like that. but even worse, i hear they recently got their entire quote database reset several years back.
22:55:37 <oerjan> (that wasn't today though.)
22:55:44 <Bike> `quote
22:55:46 <HackEgo> 252) <fungot> elliott: there go my minutes of research!!
22:55:51 <Bike> wait wrong one
22:55:52 <Bike> @quote
22:55:53 <lambdabot> hgolden says: pioneers are the ones with arrows in their backs. at least ours will be categorical arrows.
22:55:54 <elliott> Sgeo: perhaps http://catseye.tc/ehird/files-suck.html, which I should maybe get cpressey to edit with a note that I wrote it because I was annoyed at Unix and hence it's a lot more whiny than it needs to be, and doesn't fully reflect my current views
22:57:03 <hagb4rd> files suck? lol
22:57:34 <hagb4rd> at least that's pretty straight
22:57:37 <Bike> tag systems fo lyfe
22:57:44 <elliott> also my historical account of them is a bit wrong -- e.g. I was unaware of how Multics did things at the time I wrote it
22:57:55 <elliott> but I still dislike filesystems
22:58:23 <shachaf> filesystems? more like thebestsystems
22:58:37 <elliott> wow I was annoying in 2010
22:58:47 <elliott> shachaf: more like vilesystems!!!
22:59:08 <Phantom_Hoover> more like badstupids
22:59:17 <elliott> im bested
22:59:53 <shachaf> θs? more like badstupids
23:00:12 <shachaf> (THE JOKE IS GHC CODE IS FULL OF bad_stupid_theta)
23:00:49 <Bike> what the hell is theta
23:00:59 <shachaf> greek letter
23:01:10 <Bike> thx
23:01:13 <elliott> Bike: bad & stupid
23:02:05 <shachaf> Bike: It's a class context, I think.
23:02:11 <shachaf> Or maybe just a context in general?
23:02:15 <shachaf> As in the thing to the left of =>
23:02:54 <ais523> btw, I had a really good esolang idea last sleep period
23:02:59 <ais523> but haven't worked out the details
23:03:05 <ais523> the basic idea is, the esolang is defined by an interpreter
23:03:14 <ais523> whenever you run the interpreter, it mutates the language it interprets slightly
23:03:32 <ais523> and always in such a way that every program you had ever successfully run through it would fail
23:03:44 <ais523> you have to try to infer what the current state of the language is from the error messages
23:05:12 <hagb4rd> true masochists leave out the error messages
23:08:39 <hagb4rd> how would be the mutation realized? kind of randomly?
23:08:54 <hagb4rd> +be
23:09:05 <hagb4rd> no there was a be
23:10:15 <shachaf> hagb4rd: What do you get out of being in this channel?
23:11:44 <hagb4rd> sometimes someone says something interesting. or funny. or inspiring. but yea good question
23:13:52 -!- FireFly has joined.
23:16:45 <hagb4rd> but it's sad i can't remember how it's been since it was you who make me laugh
23:16:47 <hagb4rd> a pity
23:18:31 <Phantom_Hoover> again, i have to ask
23:18:43 <Phantom_Hoover> do you actually think you're saying something clever here
23:19:31 <Phantom_Hoover> because you radiate this kind of cargo cult profundity
23:20:13 <hagb4rd> i don't understand what you're saying
23:20:56 <Bike> "cargo cult profundity" probably is supposed to denote that you parrot profound things in a disjointed fashion divorced of their meaning
23:20:59 <Phantom_Hoover> you sort of know the words and general comportment of someone saying insightful things about the world around them
23:21:17 <Phantom_Hoover> you just don't actually have any actual insightful things to say about the world around you
23:21:28 <hagb4rd> aber sich ständig profilieren zu wollen, auch wenn man über leichen geht wäre nicht mein ding
23:21:53 <c00kiemon5ter> I am here because I love fungot
23:21:53 <fungot> c00kiemon5ter: this rampart, shaped like a hen's egg, yet containing enough power to keep people away from those wild, haunted hills behind hoary and fnord arkham that all his forebears for forbidden cosmic secrets was a natural result of the action of water through more than seventeen years ago, when mans
23:21:56 <hagb4rd> insofern ist es mir nicht so wichtig dass ich hier clever scheine
23:21:58 <c00kiemon5ter> fungot, do you love me ?
23:21:58 <fungot> c00kiemon5ter: this sound, as complex and unplaceable as any of the fnord poor harley warren once had. it came from india while carter and i were not connected, wherefore our first knowledge of extraordinary conditions came from others. the farther he went, and when he resumed his voice was fnord music of deep viols and of crystalline spheres. we talked often in the night as he stood on that high marble terrace with the curious
23:22:12 <Phantom_Hoover> seguing into german for no good reason is not really proving otherwise, hagb4rd
23:22:19 <hagb4rd> aber vielleicht kann ich auch etwas lernen
23:22:28 <Phantom_Hoover> *segueing, apparently
23:22:38 <Phantom_Hoover> that word was much cooler back when i only vaguely knew about it
23:23:39 <shachaf> are segways still cool though
23:23:51 <Phantom_Hoover> inasmuch as they ever were
23:23:56 <oerjan> shachaf: i hear their coolness rolled off a cliff
23:24:12 <shachaf> hello oerjan
23:24:16 <shachaf> `welcome oerjan
23:24:18 <HackEgo> oerjan: Welcome to the international hub for esoteric programming language design and deployment! For more information, check out our wiki: http://esolangs.org/wiki/Main_Page. (For the other kind of esoterica, try #esoteric on irc.dal.net.)
23:25:29 <oerjan> hagb4rd: you certainly don't have to be profound to be here. but i think what they're complaining about is that you make a lot of barely-in-context sentences that sound like you are _trying_ to be profound.
23:25:59 <oerjan> and failing.
23:26:13 <shachaf> I never got any "trying to be profound" sense.
23:26:24 <hagb4rd> well.. i hope that's not my entire motivation
23:26:25 <shachaf> It's just that I see a lot of words with no point to them.
23:26:34 <shachaf> Then I look to the left and it says <hagb4rd>.
23:27:57 <oerjan> ok then mainly it's that barely-in-context part.
23:28:53 <oerjan> also i'm sure he's not the only one to go through a period of making such comments.
23:29:08 <Jafet> Ich sind klugbar
23:29:14 <hagb4rd> sorry. sometimes my brain makes weird associations. and i wonder about it myself
23:29:30 <hagb4rd> i'll try to compense
23:31:10 <oerjan> Jafet: Du machst Deutsch gutes, ja?
23:32:39 <Jafet> am bestenlich
23:33:41 <Phantom_Hoover> you're both dutch?
23:33:47 <Phantom_Hoover> you should talk to taneb
23:34:08 <Bike> I thought taneb was Hexhamish.
23:35:20 <oerjan> Hexham am Main
23:35:27 <hagb4rd> and to answer Phantom_Hoovers question. no! i don't think i'm THAT clever. unfortunatly i'm not a good mathematician and not the best programmer. yes, i like philosphy and i like to wonder. also you guys impressed me very much, and that's mostly the reason i came here and ..am still here
23:36:15 <hagb4rd> no not impressed me
23:36:41 <hagb4rd> impressed is correct :)
23:37:32 <hagb4rd> had to check this in my dictionary
23:38:09 <oerjan> impressario
23:38:13 <hagb4rd> si
23:38:16 <oerjan> (see i too make random comments)
23:38:24 <hagb4rd> \o/
23:39:48 <Phantom_Hoover> !logs
23:47:40 <ais523> <mitsuhiko> (and yes, the workaround involves exhausting your pid space by forking until it wraps around and hits the value you're expecting)
23:47:48 <ais523> this reminds me of anarchy golf
23:48:18 <elliott> doesn't mitsuhiko play anagolf or something
23:48:22 <elliott> maybe I'm misremembering a name
23:49:19 <ais523> hmm, perhaps
23:49:22 <ais523> although that was a different context
23:50:38 <elliott> hmm, what was the context?
23:56:42 <ais523> elliott: working around a bug in Upstart
23:56:49 <ais523> also: http://www.reddit.com/r/programming/comments/19kjr5/fizzbuzz_revisited_using_rust/c8p452k
23:56:59 <ais523> someone's unfinished attempt to write FizzBuzz in Homespring
23:57:09 <ais523> I applaud them for trying (and in fact, for having heard of the language)
23:57:14 <Bike> beautiful
23:57:48 <ais523> at least, they didn't say it was Homespring, but it's obvious
23:58:38 <zzo38> I would find these select encode and decode operations to both be very useful in various C programs I write.
23:59:12 <zzo38> Would you find it useful?
←2013-03-02 2013-03-03 2013-03-04→ ↑2013 ↑all