00:02:33 -!- conehead has quit (Quit: Computer has gone to sleep.).
00:10:18 -!- TodPunk has quit (Quit: This is me, signing off. Probably rebooting or something.).
00:10:33 -!- TodPunk has joined.
00:18:44 <shachaf> ion: It's slightly difficult to believe that this person has been in #haskell for over three years and written over 50000 words.
00:21:09 -!- douglass_ has quit (Ping timeout: 264 seconds).
00:21:46 -!- conehead has joined.
00:24:36 <shachaf> adnap. Whatever their strategy is for learning/understanding things, it doesn't seem to be working.
00:25:07 <kmc> shachaf: should I leave #cslounge
00:25:09 <shachaf> It's very frustrating to try to help them. :-(
00:27:07 <shachaf> i think it makes you p. unhappy some of the time, which isn't so great
00:27:23 <shachaf> but i'm not entirely sure why
00:27:34 <shachaf> so maybe there's context that i'm missing
00:28:07 <kmc> I think I should
00:28:21 -!- nooodl has quit (Ping timeout: 248 seconds).
00:31:33 <ion> Should i join #cslounge?
00:32:41 -!- zzo38 has joined.
00:34:35 <lambdabot> (Fractional a, Integral b) => a -> b -> a
00:39:30 <Gracenotes> shachaf: you might be surprised how little #haskell knows about Haskell, as an unweighted average over its residents
00:40:10 <kmc> I would be surprised if shachaf is surprised by that
00:40:51 <kmc> there's not that much overlap between #haskell regulars and people who write "real software" in Haskell
00:41:52 <shachaf> augustss wrote "real software" in haskell (numberz joke)
00:44:16 -!- Nisstyre has joined.
00:46:22 <Gracenotes> "My Haskell compiler is so advanced it has zygoHistoPrepro# as a primitive. Can you say that about your Haskell compiler?"
00:46:40 <Bike> yes, though i'd be wrong
00:46:49 <shachaf> in a. likelihood my haskell compiler is the same as your haskell compiler
00:47:30 <Gracenotes> Also, when you really need the raw power, there's unsafeZygoHistoPrepro#
00:47:37 <kmc> i wonder if ghc has stream fusion a zygohistomorphic prepromorphism
00:48:17 <Gracenotes> Not recommended over the safe version for the average programmer.
00:48:27 <lambdabot> kmc says: Zagen, you'll need a zygohistomorphic prepromorphism from the bifunctorial Kleisli category of username-password pairs to a combinatory arrow calculus of php scripts
00:48:36 <kmc> I feel bad about that kind of quote now
00:48:46 <kmc> Gracenotes: why would I be using your Haskell compiler if I'm merely an average programmer
00:50:49 <Gracenotes> Yeah, I know, in imperative languages zygohistomorphic prepromorphisms are unsafe by default. It can be a hard first step for the average programmer to model them purely, but ultimately it leads to better software engineering.
01:03:11 <Bike> uh excuse you i was in the middle about my essay about bifunctorials being overrated and using snobol instead
01:07:09 <Bike> see you can implement kleisli categories in snobol if you extend gotos with macros,
01:07:22 <shachaf> what is a bifunctorial Kleisli category
01:08:22 -!- yorick has quit (Remote host closed the connection).
01:08:22 <shachaf> startup idea: markov chain bot trained on category theory papers + grad students to make sense of what comes out
01:09:29 <Bike> it's like SNARXIV for SLOVENIAN MATHEMATICS!
01:11:08 <zzo38> I played Dungeons&Dragons game yesterday.
01:11:13 <shachaf> Bike: so you know how the limit/colimit functors are adjoint to the diagonal functor
01:15:11 -!- Bike has quit (Ping timeout: 256 seconds).
01:19:43 -!- Bike has joined.
01:23:27 <Bike> "Galois became so angry with [being locked into school] that he tried to escape from l'Ècole [the school] by climbing over its walls. When he failed there was nothing left to do but take his examinations in calculus and physics."
01:23:56 -!- clog has joined.
01:24:53 <zzo38> But I made some progress to win. And even when I do win, the game still won't be finished yet
01:29:25 <zzo38> I did once do something where a stun cone is deliberately aimed to hit someone on my own team as well as the opponents, because the character on my own team is I expected them to easily resist it. Does this tactic have a name?
01:34:38 -!- Lymia has joined.
01:34:45 <Gracenotes> on the subject of winning: http://dwarffortresswiki.org/index.php/Losing
01:35:35 <tswett> I think I'm going to rename my Totally Awesome Programming Language to Hylisk.
01:35:43 <tswett> Since "Hydralisk" is already a word.
01:40:49 <zzo38> I want to know if this tactic has a name.
01:45:02 <zzo38> My character's name is Iuckqlwviv Kjugobe.
01:45:56 <oerjan> Iuckqlwviv Kjugobe's Gambit, check
01:46:34 <shachaf> zzo38: That name seems difficult to pronounce.
01:46:46 <zzo38> But that isn't the only thing they do!
01:46:50 -!- madbr has joined.
01:47:28 <tswett> I think I'm gonna write a Hylisk compiler in Hylisk. It'll be useless at first.
01:48:18 <tswett> module MegaHylisk where main :: World -> World
01:48:54 <oerjan> but that's not monadic!
01:49:15 <zzo38> (I called it "Armor Gambit", if there isn't the name, but I think that isn't a very good name though)
01:49:22 <tswett> I mean, it's equivalent to something monadic.
01:49:44 -!- sprocklem has joined.
01:49:46 <tswett> IO a is defined as World -> (World * a), so World -> World is equivalent to IO Unit.
01:50:36 <oerjan> shachaf doesn't approve of your World view
01:50:56 <ion> IO a might as well be defined as Friendship -> (Magic * a)
01:51:14 <HackEgo> 2013-07-13 19:43:17: <Gregor> It is, it works great.
01:51:30 <HackEgo> Runs arbitrary code in GNU/Linux. Type "`<command>", or "`run <command>" for full shell commands. "`fetch <URL>" downloads files. Files saved to $PWD are persistent, and $PWD/bin is in $PATH. $PWD is a mercurial repository, "`revert <rev>" can be used to revert to a revision. See http://codu.org/projects/hackbot/fshg/
01:51:54 <ion> Let me interject. What you are referring to as GNU/Linux is actually GNU/GNU/Linux.
01:52:03 <tswett> Hylisk is going to be Awesome.
01:52:14 <tswett> The reason that it's Awesome is that things like World -> World actually make sense.
01:52:20 <oerjan> GNU/GNU/GNU/Linux/Guisarme
01:52:30 <tswett> Since I'm defining IO, I think I'm going to define OI as well.
01:52:44 <ion> How about I and O?
01:52:50 <tswett> ion: that... actually would make sense.
01:52:55 <zzo38> Things like World -> World don't make sense for I/O ordinarily how it work, but maybe in that programming language, it might.
01:53:16 <ion> zzo38: What’s your favorite version of MS-DOS?
01:53:32 <tswett> data I a = I (World -> a); data O a = O World a; data IO a = IO (I (O a)); data OI a = Oi (O (I a))
01:54:04 <shachaf> tswett: like, adjunctions, man
01:54:13 <tswett> Though "I" and "O" don't actually mean "input" and "output" here. Who knows what they mean.
01:54:35 <ion> Ion and Oerjan
01:55:11 <tswett> An "I a" turns the entire universe into an "a". That sounds pretty... apocalyptic.
01:55:23 <tswett> Say, fact: in this language, you're going to be convert a monad into its corresponding comonad really easy.
01:55:43 <shachaf> oerjan: should i start saying "monotone" and "antitone" functors...........
01:56:12 <tswett> Indeed, I think it makes sense to introduce syntax for this: ~IO is the comonad corresponding to IO, ~List is the comonad corresponding to List...
01:56:36 <oerjan> shachaf: not in public, no
01:56:46 -!- nooodl has joined.
01:57:01 <zzo38> tswett: Maybe you need an extra structure on that category
01:57:13 <tswett> Extra structure on that category?
01:57:22 <tswett> Like, uh, is "cartesian closed" an example of extra structure?
01:57:33 <zzo38> Yes, that is the kind of things I mean, actually
01:57:52 <zzo38> Although in what you have, you need to have some way so that there is a comonad corresponding to a monad
01:57:53 <tswett> Yeah, there's an analogous thing for this, but I don't remember what it's called. Lemme see.
01:57:55 <shachaf> tswett: how does that work..............
01:58:17 <shachaf> tswett: you can make co-io like that because you get "IO" from an adjunction O -| I
01:58:21 <tswett> shachaf: simple. data Co m a = Co (m (a -> Done) -> Done)
01:58:27 <shachaf> do you always have an adjunction??
01:58:51 <tswett> I don't actually know.
01:59:05 <tswett> Anyway, Done is defined by the existence of a bijection between (a -> Done) -> Done and a.
01:59:52 <oerjan> tswett: for lists, the obvious adjunctions are not between a category and itself, but between, in spirit, the category of sets and the category of monoids.
02:00:11 <oerjan> *obvious adjoint functors
02:00:13 <tswett> zzo38: here we go. Closed symmetric monoidal categories.
02:00:23 <shachaf> oerjan: or, like, the kleisli category of [] ""or whatever""" and stuff, man
02:00:29 <oerjan> so the comonad is not in the same category
02:00:48 <tswett> Anyway, Hylisk is going to be based on linear logic as a type system, the same way that Haskell has intuitionistic logic as a type system.
02:00:50 <shachaf> oerjan: It sounds like tswett is describing something similar to edwardk's "monads from comonads", but I think that only works one way.
02:01:03 <shachaf> I'm not sure how a linear type system or something changes that.
02:01:17 <tswett> Maybe my "monads" aren't really monads or something.
02:01:29 <tswett> I'm not actually sure Co always gives you the comonad, but it does always give you *something*.
02:01:45 <shachaf> edwardk's type is newtype Co w a = Co { runCo :: forall r. w (a -> r) -> r }
02:01:54 <oerjan> tswett: have you seen Clean btw
02:01:55 <shachaf> And that works fine for giving you a monad, by composing adjunctions.
02:02:59 <tswett> oerjan: I took a look at it. I don't know if it supports everything linear logic has.
02:03:07 <tswett> In particular, does it have a "with" type constructor?
02:03:27 <tswett> shachaf: yeah, that does look remarkably like what I came up with.
02:03:43 <tswett> I wonder if edwardk is familiar with linear logic. Surely so.
02:04:19 <shachaf> Yes, edwardk used to be p. obsessed with substructural logic and such.
02:04:44 -!- jconn has joined.
02:04:54 <tswett> And to that other guy, for that matter.
02:05:15 <tswett> Don't suppose any of you guys know him?
02:05:48 <oerjan> tswett: well i think uniqueness types of Clean are related to linear types but different.
02:06:13 <ion> I know a guy who knows a guy who knows a guy who knows a guy who knows a guy who knows Philip Wadler.
02:06:50 <Bike> i know a guy who knows a guy who knows a guy who knows a guy who knows a guy who knows a guy who knows Philip Wadler
02:07:02 <oerjan> i hear there's a law that you cannot get to wadler because of an exponential blowup of lexical syntax of comments hth
02:10:07 <tswett> According to LinkedIn, there's no such person as Philip Wadler.
02:19:36 <Bike> prehensile tapir dick: http://i.minus.com/iGQMJmPRkKSGK.gif
02:21:03 <zzo38> I think INSTEAD OF INSERT triggers are the useful kind of triggers in SQL, although the other triggers are also sometimes useful. What do you think?
02:24:07 <zzo38> (I don't remember if I have used any other kinds of triggers.)
02:27:36 <tswett> I'm tempted to call this one type "Rubbish".
02:28:01 <tswett> Just because that's a word not really used on this side of the pond.
02:29:07 -!- oerjan has quit (Quit: tginh).
02:31:07 <tswett> So, uh, what am I doing...
02:31:21 <Sgeo> Playing Creatures?
02:31:24 <Sgeo> Or Perspective?
02:31:29 <tswett> Nope. I'm implementing Hylisk in Hylisk.
02:31:42 <shachaf> no, Sgeo.................................................
02:32:16 <tswett> No shachaf. How's it going?
02:34:25 <tswett> Hm, here's an interesting idea for what do-notation could do. First, we could say that do {a; b; c} means (c . b . a).
02:35:56 <tswett> Then a -< y will mean (\arg -> a arg y).
02:36:43 <tswett> And x <- a will mean, uh, let (res, x) be the result of a and return res. Yeah.
02:39:36 <tswett> But this only really makes sense for state-like things.
02:39:39 <tswett> Oh well. It works nicely for I/O.
02:39:50 <tswett> I'll use the keyword "sdo" instead.
02:41:11 -!- carado has quit (Ping timeout: 260 seconds).
02:44:25 <tswett> source <- readFile -< arg
02:44:43 <tswett> -< could easily be defined as an actual function.
02:45:42 <zzo38> Yes it looks like that definition of -< is a combinator function.
02:45:52 <tswett> Or could it? I might want to have special syntax for it.
02:46:02 <tswett> The same way that , isn't an operator in Haskell, even though it could be.
02:49:08 <tswett> class Data a where delete :: a -> Unit; duplicate :: a -> (a * a)
02:50:31 <zzo38> I made a Csound plugin "slowchange" which acts like a kind of low-pass filter but it is different; instead of a frequency cutoff it is the frequency times the amplitude which is cutoff, and the waveform becomes closer to a triangle wave as it approaches the cutoff. Does this have another name? Does this have another use?
02:51:53 <madbr> They have to do something like that when mastering vinyl
02:51:59 <madbr> and it appears in electronics
02:52:13 <madbr> for instance with op amps (which have a limited slew rate)
02:52:25 <tswett> Maybe "Data" isn't the best name for a class.
02:52:32 <zzo38> I didn't know what it was called so I called it "slowchange", but thank you for telling me that
02:52:34 <tswett> "data Void deriving Data"
02:53:42 <shachaf> tswett: Are the laws for that class the same as the comonoid laws?
02:54:15 <zzo38> madbr: Is it sometimes used for special effects?
02:54:43 <madbr> more like some kind of distortion
02:54:53 <madbr> ADPCM compression also has a similar artifact
02:54:58 <tswett> shachaf: for the Data class? Uh, I guess the laws are that fst (duplicate x) = x and snd (duplicate x) = x.
02:55:21 <tswett> delete doesn't really have laws, I think, because there's only one thing it could possibly do.
02:55:43 <tswett> Hm. fst and snd aren't actually definable on pairs.
02:55:53 <tswett> Not on * pairs, that is.
02:56:26 <tswett> Do I really need the Data class? Maybe I don't... *shrug*
02:56:39 <tswett> Maybe I'll do without it and see how much I miss it.
02:58:00 <zzo38> shachaf: I know my character's name is difficult to pronounce; I made up at random, and presumably that character is able to pronounce.
02:58:30 <shachaf> tswett: Well, the comonoid identity laws are (1 *** delete) . duplicate = unfst, where unfst :: a -> (a, Unit)
02:59:01 <tswett> Oh, you said comonoid, not comonad.
03:00:18 <tswett> I guess these are definitely Data laws: (\(x, y) -> (x, delete y)) (duplicate z) = (z, ()), (\(x, y) -> (delete x, y)) (duplicate z) = ((), z)
03:00:28 <tswett> All right, what are the operations on Comonad usually called?
03:00:31 <tswett> The co-join and co-return ones?
03:00:31 <shachaf> And the associativity law is the "obvious" one.
03:00:43 <shachaf> duplicate and "extract", but don't use "extract".
03:02:06 <shachaf> Anyway, your law is the same as my law?
03:03:14 <tswett> Yeah, looks like it is.
03:03:26 <shachaf> Data isn't a class for comonads, just for comonoids.
03:03:34 <shachaf> I am pleased that comonoids make sense for linear types.
03:04:41 <tswett> Hm, so wait. You could potentially have a Data instance that's a comonoid other than the trivial one, I guess?
03:04:45 <tswett> I'm not sure if I want to allow that.
03:04:53 <tswett> Hey, while I'm at it, I should rename "return".
03:04:58 <shachaf> http://www.pps.univ-paris-diderot.fr/~laurent/geocal/slides/tabareau.pdf
03:06:36 <shachaf> http://stackoverflow.com/questions/15418075/the-reader-monad
03:06:46 <madbr> hmm, you could build a language with a state automaton and one bignum integer variable, and the operations multily, divide and checking modulo if zero or nonzero
03:07:10 <shachaf> (Not the top answer, that one is sort of nonsense.)
03:07:29 <madbr> since it's essentially the same as having a few variables that you can increment and decrement and check for zero
03:07:39 * tswett renames return to pure.
03:07:45 <madbr> (except with the exponent of a few prime numbers multiplied together)
03:07:50 <tswett> I could call the comonad operations copure and cojoin.
03:08:38 <shachaf> instance Comonoid r => Monad {runReader :: Reader r} where
03:08:45 <madbr> (with increment/decrement/check zero you can build a loop to get *2 and /2, and then build a couple of stacks to get the infinite tape)
03:09:02 <tswett> Given that every comonad is Store (in a sense), doesn't the name "extract" really make perfect sense?
03:09:13 <shachaf> That should've been obvious in retrospect.
03:10:13 * tswett instance Monad m => Comonad ~m where
03:10:18 <shachaf> Comonoid r => Monad (r ->), Monoid r => Monad (r,)
03:10:43 <shachaf> Comonoid r => Comonad (r,), Monoid r => Comonad (r ->)
03:12:34 <zzo38> The "return" and "pure" operations are operations of a different kind of thing and even in mathematics I think they call them differently; I did read though in a "monoidal monad" they are defined as being equal, though.
03:15:35 <Sgeo> Why am I looking at Retro?
03:15:37 <Sgeo> It's so low-level
03:15:38 <tswett> extend :: (m (a -> Done) -> Done) -> (m ((m (a -> Done) -> Done) -> Done) -> Done)
03:15:45 <Bike> why are you looking at anything
03:15:47 <Bike> Damn your eyes
03:15:54 <Bike> escape this shitty visual world
03:16:09 <madbr> do you guys know the collatz conjecture?
03:16:12 <madbr> https://en.wikipedia.org/wiki/Collatz_conjecture
03:16:21 <madbr> aka the 3n + 1 conjecture
03:16:28 <tswett> I'm familiar with the conjecture.
03:16:43 <tswett> Let me see whether or not I think it's obviously true.
03:16:49 <madbr> sorta wondering if it could be turing complete
03:16:51 <shachaf> i prefer the llatz njecture
03:16:54 <Bike> madbr: generalizations are
03:17:11 <Bike> there's a nice paper on the subject but i've yet to find a version that isn't a shitty preprint, for some reason
03:17:17 <shachaf> maybe they're both clopen problems??
03:17:17 <madbr> but how small can the generalizations be until they stop being turing complete?
03:17:34 <Bike> i do believe that is An Open Problem
03:18:11 <Bike> that is itself an open problem
03:18:12 <tswett> The Collatz conjecture gives me the heebie jeebies.
03:18:22 <Bike> i only get jeebies :(
03:18:44 <shachaf> heebie jeebies are like jeebies except right-to-left
03:19:00 * tswett writes "extend = (not yet implemented)"
03:19:10 <tswett> Honestly, when am I ever going to need to convert between monads and comonads...
03:19:23 * tswett just deletes this declaration.
03:19:40 <madbr> bike : the crazy thing is that the generalization was only only proved undecidable in like 2007
03:21:22 <shachaf> tswett: btw hask is monoidal in more than one way
03:21:46 <tswett> What does it mean for Hask to be monoidal?
03:22:00 <shachaf> you can make eg Void -> a and Either a a -> a
03:22:14 <shachaf> which is another kind of monoid, rather than () -> a and (a,a) -> a
03:22:27 <shachaf> (but it ends up being boring in the same way that normal comonoids are)
03:22:29 <shachaf> https://en.wikipedia.org/wiki/Monoidal_category
03:22:30 -!- nooodl has quit (Ping timeout: 256 seconds).
03:23:30 <shachaf> https://en.wikipedia.org/wiki/Closed_monoidal_category
03:24:29 <tswett> Hm, what sort of language do I want my compiler to target.
03:26:24 <tswett> Is there a compiler that compiles from English into x86?
03:27:03 <Sgeo> What does English compile into?
03:27:15 <tswett> I'm not aware of any compilers from English.
03:27:17 <Bike> tswett: Bill from development, yes
03:27:28 <Sgeo> Not this one http://en.wikipedia.org/wiki/English_(programming_language)
03:27:44 <Sgeo> This one http://www.osmosian.com/
03:27:54 <tswett> Is his source code under one megabyte?
03:29:33 <Bike> in some language, yes
03:29:43 <tswett> compile = emit . translate . parse
03:29:47 <Sgeo> "Extract the background given the screen's box. \or Create the background from the screen. Or something."
03:29:50 <tswett> There, I've written the bulk of the Hylisk compiler.
03:31:21 <Sgeo> "Allocate memory for the work."
03:31:36 <Sgeo> I'm sure the target audience for this thing know all about allocating memory
03:31:47 <tswett> Hm hm hm. What's a HyliskModule, anyway? It's, uh... well, it has a name, and, uh...
03:32:27 <tswett> A name, a list of data declarations, a list of type declarations, a list of newtype declarations, a list of class declarations, a list of instance declarations, a list of type signatures, and a list of function definitions?
03:32:33 <tswett> Oh right, I also need a list of fixity declarations.
03:33:18 <zzo38> Do you want to use bounded integers, or surreal numbers, for fixity declarations?
03:33:47 <tswett> How about rational numbers?
03:34:00 <tswett> They're literally exactly better than the surreal numbers.
03:34:37 <zzo38> I think surreal numbers are better for fixity declarations, though.
03:35:11 <zzo38> Even though rational numbers would work.
03:36:03 <tswett> data HyliskModule = HM Name (List FixityDec) (List DataDec) (List TypeDec) (List NewtypeDec) (List ClassDec) (List InstDec) (List TypeSig) (List FunDef)
03:36:08 <tswett> Yes, I definitely know what I'm doing.
03:38:17 <zzo38> I have (different) idea, compile a Magic: the Gathering into a Haskell code like: target (is_a Spell) >>= callKeyword . Counter
03:42:36 <Gracenotes> everything is better when it's a Haskell DSL
03:44:21 <zzo38> Gracenotes: Yes it could do a lot of things.
03:48:56 -!- madbr has left.
03:50:22 -!- JesseH has joined.
03:51:44 <JesseH> Trying to think of how I can use English as a programming language.
03:51:55 <JesseH> Some edits would need to be made >_>
03:52:09 <Bike> you hire a person to program for you.
03:52:37 <JesseH> Bike, Just as a fun little esolang
03:53:05 <JesseH> "Create a program that prints 'Hello World' five times."
03:54:03 <JesseH> "x is equal to 100. Create a program that prints 'x bottles of beer' 100 times, subtracting 1 from x each time."
03:54:10 <lambdabot> http://hackage.haskell.org/trac/ghc/newticket?type=bug
03:54:21 <lambdabot> http://hackage.haskell.org/trac/ghc/newticket?type=bug
03:54:27 <lambdabot> http://hackage.haskell.org/trac/ghc/newticket?type=bug
03:54:42 <Sgeo> @hug lambdabot
03:54:42 <lambdabot> http://hackage.haskell.org/trac/ghc/newticket?type=bug
03:54:53 <Sgeo> @hug Gracenotes
03:54:53 <lambdabot> http://hackage.haskell.org/trac/ghc/newticket?type=bug
03:55:36 * lambdabot submits elliott 's email address to a dozen spam lists
03:55:53 <Bike> whoa now, that's over the line
03:58:01 <Bike> is lambdabot truly male
03:58:39 <shachaf> no, the lambdabot @slap database just assumes everyone is male
03:59:00 <Sgeo> It should assume everyone is Spivak
04:01:24 * lambdabot decomposes Fiora into several parts using the Banach-Tarski theorem and reassembles them to get two copies of Fiora!
04:01:46 <Bike> that doesn't sound too bad.
04:04:08 * lambdabot secretly deletes Bike's source code
04:04:21 <Bike> i'm allllll machine code, baby
04:04:29 <Fiora> that isn't very secret is it
04:04:48 <zzo38> Are some people both Buddhist and Roman Catholic? (I know a Roman Catholic who wants to become Buddhist while remaining Roman Catholic too)
04:04:50 <Gracenotes> I have no idea why @vixen was allowed into lambdabot in the first place
04:05:49 <Sgeo> zzo38: I'm pretty sure there exists at least one person who self-identifies as both Christian and atheist, and Roman Catholicism and Buddhism are probably less incompatible than that.
04:05:56 * lambdabot hits quine with a hammer, so they breaks into a thousand pieces
04:07:08 <shachaf> Wobbler took a long look at the girl in the cardboard hat.
04:07:17 <shachaf> "Make me one with everything," he said. "Because... I'm going to become a Muslim!"
04:07:22 <Bike> so they breaks
04:07:46 <shachaf> Bike: they does. you gots a problems with that?
04:07:56 <zzo38> Sgeo: I also think they aren't incompatible, although some people might. Also, I don't really know how you can be both Christian and atheist, although you may be able to be both Yeshuan and atheist. ("Yeshuan" is a term that I and someone else has made up independently, with the same meaning.)
04:07:59 <Sgeo> mnoqy: I saw some thing where a group of people tried to write arguments for their position, or the other position (yes, I know there are more than two positions), and others had to guess which one. One person was a wildcard who considered himself both
04:08:24 <shachaf> was he a wildcard...................or was he just a joker
04:08:43 <Gracenotes> the most nebulous term surely is Jewish, as it doesn't even imply you follow the Judaic religion
04:09:05 <shachaf> Gracenotes: imo "nebula" is a p. nebulous term
04:09:13 <Bike> zzo38: wikipedia has an article on christian atheism. it's ok (the article)
04:09:37 <zzo38> Sgeo: Positions of what?
04:10:43 <zzo38> Bike: Yes, so Christian atheism and Jesusism are related to Yeshuanism, it seems.
04:10:52 <Sgeo> zzo38: Christianity and atheism
04:11:01 <Sgeo> Or maybe it was theism and atheism, I don't remember
04:11:20 <shachaf> what about anism "like theism but more indefinite"
04:11:58 <zzo38> Sgeo: Well, there certainly are more than two positions
04:13:39 -!- tertu has joined.
04:15:25 -!- oklopol has joined.
04:19:13 <Gracenotes> it's always a bit funny when people in filmed Jesus Christ Superstar productions have visible tattoos of crosses
04:20:41 <Gracenotes> though some of those may be temporary (for anachronistic purposes)
04:25:09 <zzo38> I do not consider Christian atheism to be Christian, although it may be Yeshuan (which does not have to be atheist).
04:25:55 <Sgeo> I don't think the person was the Wikipedia version of Christian atheist
04:25:58 <Gracenotes> how does Yeshuanism related to Judaism?
04:26:04 <Sgeo> I don't remember
04:26:26 <Bike> does "yeshuan" have to be pronounced in a first century judean aramaic accent
04:26:28 <Gracenotes> presumably it's a bit like Christianity - Judaism - God
04:27:39 <Gracenotes> a cult of personality, basically, but stopping at personality
04:28:29 <zzo38> I don't know if it is pronounce like that. However, "Yeshuan" is a term myself (I read a book in the library once, it was titled "Stop worshipping Christ, start following Jesus", and described something which they called Christian but I didn't think it was so I invented the term "Yeshuan"), and someone else made up independently, having the same meaning.
04:29:05 <zzo38> I think that the question of whether God exists or not isn't relevant to defining Yeshuanism.
04:32:52 -!- douglass_ has joined.
04:36:34 -!- ^v has joined.
04:36:44 <^v> i logged in about 2 years ago
04:36:49 <^v> forgot what i was making
04:36:52 <^v> wanted to say hi
04:37:07 <HackEgo> ^v: 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.)
04:38:01 <zzo38> ^v: Then remember or try to figure out what you were making, please.
04:38:28 <^v> prolly something Lua
04:38:41 <^v> but related to esoteric languages somehow....
04:38:57 <HackEgo> 2012-09-02.txt:05:00:45: <zzo38> Because, if it is all wrong, then I should fix it please
04:39:08 -!- ^v has changed nick to \b.
04:39:10 -!- \b has changed nick to \n.
04:39:27 -!- \n has changed nick to \y.
04:39:29 -!- \y has changed nick to \u.
04:40:22 -!- \u has changed nick to ^v.
04:41:15 <zzo38> ^v: Why did you keep changing your name?
04:41:40 <shachaf> zzo38: They actually changed their nick, not their name.
04:41:47 <zzo38> shachaf: Yes, I know that.
04:44:51 <zzo38> Whatever dream any of you have while sleeping, do you notice some pattern of things in common, whether all of them and/or in certain kinds (short dream, long dream, good dream, bad dream, nightmare dream, neutral dream, repeating dream, dream when sleeping in a different place then you ordinarily would, etc)?
04:45:51 <zzo38> I just mean things in common whether or not it forms a pattern really
04:45:54 <Sgeo> I'm generally in my dreams
04:46:15 <mnoqy> i'm in my dreams most of the time but sometimes i'm not
04:46:32 <Sgeo> Until a bit over a year ago, my dreams sometimes featured things that my real life generally did not
04:47:00 <HackEgo> smlist (414): shachaf monqy elliott mnoqy
04:47:03 <Sgeo> It occurs to me that most/almost all dreams involve elements that real life does not
04:47:13 <mnoqy> i need to remember more dreams & also organize all my logged dreams
04:47:13 <Sgeo> I guess I had specific elements in mind
04:47:58 <Sgeo> I had a dream journal online
04:48:00 <zzo38> And sometime I am in but I am someone else other than myself
04:48:04 <Sgeo> I kind of neglected it though
04:48:12 <zzo38> mnoqy: Yes you should.
04:48:31 <Sgeo> http://www.dreamviews.com/dream-journal-archive/53892-sgeos-dream-journal.html
04:48:32 <zzo38> I have written stuff about my dream on my computer; I have also written what other people have told me about their dream, and marked them differently
04:48:49 <zzo38> I used plain ASCII text files though, rather than HTML
04:49:25 <Sgeo> I have yet to fly into the sun. I am annoyed.
04:51:51 <zzo38> I have noticed that my good dream involves several subjects, nightmare dream involve a single subject and I wake up with a sudden movement and almost fall off of the bed, repeating dream is relatively mundane, ...there are many others too. One thing in general is that sometimes I think of two things at once which aren't being used together.
05:01:30 <Bike> http://i.imgur.com/cfGFm5j.png I believe I have found the ultimate douchey job ad for programmers
05:03:11 <Gracenotes> some of the banned words seem reasonable
05:03:38 <Gracenotes> "We have no HR department" Discrimination lawsuit, here we come
05:16:44 <Sgeo> depends is a banned word
05:17:07 <Sgeo> So, no notion of things that may be good or bad depending on circumstances?
05:17:36 <Bike> you can, you just have to express it in the sacred language of the mole-men
05:29:45 -!- ^v has quit (Quit: Leaving).
05:32:52 <oklopol> zzo38: i tend to start rotating
05:35:39 <oklopol> like i slowly rotate forward, first i rotate until my face touches the floor, then my legs start rising up until i stand on my head etc
05:36:53 <Bike> i have no idea what the context for this is and it's beautiful
05:38:01 <oklopol> zzo38 asked about recurrent themes in dreams
05:38:11 <Bike> no shush keep going
05:38:26 <oklopol> i'm talking to a beautiful girl, so gonna get laid
05:38:49 <oklopol> it's usually something like that
05:39:08 <oklopol> and then it becomes a dream about rotating through places
05:39:38 <oklopol> i'm winning the nobel prize
05:39:48 <oklopol> i've been preparing my speech for months and months
05:39:57 <oklopol> i start rotating through the audience
05:40:47 <oklopol> another one is that i start floating upward. this also happens when i'm dreaming about something nice. and then it becomes a dream about floating higher and higher.
05:41:50 <Bike> http://www.youtube.com/watch?v=DzwFpzdO3DY Currently pituring
05:42:45 <comex> when I dream I often am able to fly through the air by pumping my arms in a similar manner to swimming
05:43:12 <oklopol> well sometimes i just stop trying not to rotate, and then it becomes an uncontrolled accelerating rotation that usually ends with me waking up
05:47:37 <zzo38> I am sometimes in the dream able to levitate half an inch above the ground.
05:55:58 -!- Bike has quit (Ping timeout: 276 seconds).
06:04:38 -!- Bike has joined.
06:30:25 -!- tertu has quit (Ping timeout: 276 seconds).
06:38:32 -!- Bike has quit (Quit: Lost terminal).
06:39:50 -!- Bike has joined.
06:40:01 -!- Bike has quit (Client Quit).
06:40:34 <zzo38> Now I found the programs for rotation of planets other than the Earth, in the source-codes of Celestia. It doesn't include Eris, though.
06:40:59 -!- Bike has joined.
06:41:55 <zzo38> I don't know how accurate they are, though.
06:42:17 -!- Vorpal has joined.
06:44:30 <shachaf> Vorpal: what are you doing in #haskell.........................
06:45:05 <Vorpal> shachaf, idling as usual
06:52:16 <shachaf> OK, against common sense I'll try again.
06:53:11 -!- zzo38 has quit (Remote host closed the connection).
06:55:02 -!- conehead has quit (Quit: Computer has gone to sleep.).
07:02:08 -!- sprocklem has quit (Remote host closed the connection).
07:09:14 <shachaf> I thought I'd test the theory that sometimes I misjudge people as "eternally clueless" by trying to help one of them where I'd normally ignore them.
07:10:20 -!- NihilistDandy has joined.
07:13:19 -!- Bike has quit (Ping timeout: 276 seconds).
07:13:41 -!- Bike has joined.
07:26:06 -!- sacje has quit (Excess Flood).
07:26:34 -!- sacje has joined.
07:28:55 -!- douglass_ has quit (Ping timeout: 276 seconds).
07:30:47 -!- NihilistDandy has quit (Quit: NihilistDandy).
08:42:54 -!- epicmonkey has joined.
09:08:44 -!- MindlessDrone has joined.
09:43:28 -!- nortti_ has changed nick to nortti.
09:49:35 -!- hogeyui____ has quit (Ping timeout: 268 seconds).
10:05:13 -!- hogeyui____ has joined.
10:58:25 <fizzie> "Powerplant: 1 × human , 1.1 kW (1.5 hp)" (specs of that human-powered quadcopter thing)
10:58:32 <fizzie> Apparently a human is like 1.5 horses.
10:59:37 <Vorpal> fizzie, pretty sure a human has more energy than that
10:59:39 <fizzie> "Powerplant: 1 × Dennis Bodewits, Kyle Gluesenkamp, Colin Gore Human, 1 hp (0.75 kW)" (specs of Gamera II, another of the kind)
11:00:04 <Vorpal> According to E=mc² that is
11:00:12 <fizzie> "Powerplant: 1 × human power via pedals which are connected to the wings through a system of pumps and pulleys , 0.7 kW (0.94 hp)" (specs of that human-powered ornithopter)
11:00:21 <Vorpal> So if you take one human and one anti-human, how much energy do you get when you collide them?
11:00:38 <fizzie> I think "quite a bit" is the right answer.
11:02:11 <fizzie> 5.575*10^18 J, says W|A.
11:02:53 <fizzie> (Input: "mass of human in joules", input interpretation: "convert [ human | weight ] to joules", result: "5.575*10^18 J (using E=mc^2)".)
11:03:19 <fizzie> Equivalent to 1.333*10^9 tons of TNT.
11:04:05 <Vorpal> fizzie, So lets see, that would be how many Tsar bombs?
11:04:13 <Vorpal> I forgot what the yield of that thing was
11:04:45 <fizzie> 0.53 times the total energy from the sun that hits the earth in one minute, which implies that if (when) the sun goes dark, we just need to sacrifice one guy every 32 or so seconds.
11:07:23 <fizzie> Approximately 23.2 Tsar Bombas.
11:07:45 <Jafet> You just need a steady supply of, say, intellectuals and anti-intellectuals.
11:07:53 <fizzie> Had to divide it myself, W|A knows about "yield of tsar bomba" separately, but doesn't seem to want to divide the two quantities.
11:11:31 <Lumpio-> If one human can produce enough power to keep himself and the contraption afloat, wouldn't more humans be able to do it more easily
11:14:59 <fizzie> I suppose there's more to it than just counting watts, like reasonable scale for the contraption and how to control it.
11:17:06 <Jafet> The person with the highest horsepower has more average horsepower than the five people with the highest horsepower
11:20:03 -!- Sgeo has quit (Read error: Connection reset by peer).
11:31:42 <fizzie> If you mean the Gamera II specs, it also holds just a single human; I think those names were just pilots that have -- at different times -- used it.
11:32:45 <fizzie> Or possibly I just can't read your comment right.
11:35:05 <Jafet> shachaf: you just need to realize that many people may be eternally clueless but do not remain that way.
11:35:34 -!- mnoqy has quit (Quit: hello).
11:38:34 <shachaf> Jafet: I'd love to see some transitions away from "eternally clueless".
11:39:15 <Jafet> I used to be eternally clueless, but not any more.
11:40:02 <shachaf> It's possible that I used to, too.
11:53:49 -!- Ghoul_ has joined.
11:58:24 -!- sacje has quit (Ping timeout: 260 seconds).
12:02:45 -!- mnoqy has joined.
12:04:28 -!- yorick has joined.
12:10:21 -!- Guest6363 has joined.
12:16:42 -!- oklopol has quit.
12:39:51 -!- mnoqy has quit (Quit: hello).
12:40:10 -!- Taneb has joined.
12:44:30 -!- carado has joined.
13:04:39 <Vorpal> Why is it that every single time I need to compile llvm it turns out to be a really hot day, making the computer fans go crazy as a result. Strange coincidence.
13:06:44 -!- Guest6363 has quit (Ping timeout: 256 seconds).
13:08:49 <Jafet> That's odd, I also have underpowered computer fans.
13:09:09 <Vorpal> Mine are slightly underpowered it seems, at least for 26 C ambient
13:09:43 <Vorpal> I put a table fan in front of it. Otherwise I was hitting 75 C in seconds while doing make -j4
13:10:26 <Vorpal> oh well, bbl, this seems stable, so I can go do other stuff now
13:17:34 -!- Guest6363 has joined.
13:22:18 -!- Guest6363 has quit (Ping timeout: 264 seconds).
13:37:24 -!- sacje has joined.
13:46:24 -!- sacje has quit (Ping timeout: 256 seconds).
13:48:01 -!- sacje has joined.
14:03:39 -!- sacje has quit (Ping timeout: 268 seconds).
14:23:21 -!- Taneb has quit (Quit: Leaving).
15:16:59 -!- nooodl has joined.
15:33:26 -!- copumpkin has quit (Ping timeout: 240 seconds).
15:34:03 -!- copumpkin has joined.
15:34:18 -!- kallisti has quit (Ping timeout: 264 seconds).
15:43:29 <tswett> Hm. I think writing a compiler is actually kind of boring.
15:44:25 <Vorpal> tswett, which part of it made you think that?
15:44:33 <Vorpal> I always found optimising quite interesting
15:45:18 <Vorpal> fizzie, I seem to have managed to fix another unrelated bug in a function called by the miscompiled function (in another translation unit!) and now the function is no longer miscompiled
15:46:00 <Vorpal> fizzie, I think STRN is now fixed, though I will do some more tests, since mycology said it was okay when it completely failed on another test case
15:47:06 <tswett> The part where you parse a program, translate it into a different form, and reemit the result.
15:47:25 <Vorpal> tswett, so all of it then
15:47:25 -!- zzo38 has joined.
15:47:28 -!- Taneb has joined.
15:47:41 <Vorpal> tswett, personally I find the lexer boring, but that is about it
15:47:55 <Taneb> I made marshmallows this afternoon
15:48:15 <tswett> Parsing in particular, because I'm writing in a language that doesn't have a parser library. Or, for that matter, any libraries whatsoever.
15:48:29 <Vorpal> tswett, oh what language are you writing in?
15:48:53 <tswett> Hylisk. It's a language that I'm making up on the fly.
15:49:33 <Vorpal> I think every compiler or interpreter I have written have either used something like flex+bison, parsec or been for such a simple language that it was trivial to write by hand (bf and befunge comes to mind)
15:49:36 -!- Nisstyre-laptop has joined.
15:49:50 <Ghoul_> tswett: any spec on hylisk?
15:50:02 <tswett> Maybe instead of writing a Hylisk-to-C compiler in Hylisk, I should write it in a language that already exists.
15:50:10 <Vorpal> tswett, is hylisk what you are writing in or what you are parsing?
15:50:26 <tswett> Ghoul_: no; I figured that first I'd write some code in Hylisk in order to solidify my ideas about it.
15:50:40 <Vorpal> tswett, How are you going to bootstrap it then?
15:50:50 <tswett> I'll worry about that later.
15:51:17 <Ghoul_> I started writing a silly spec for haskell that pretended to be C https://gist.github.com/kvanberendonck/2b824285b38be9db2ff0
15:51:28 <tswett> But I'm thinking that instead of actually writing a compiler, maybe I want to use a language that's built for creating languages.
15:51:39 <Ghoul_> I still want to write a compiler one day, but for a more sane language.
15:51:48 <tswett> Like, Idris has a feature for creating embedded domain-specific languages.
15:52:30 <Vorpal> Ghoul_, make a JIT compiler for befunge (you will need to JIT due to the self modifying nature of it)
15:52:45 <Ghoul_> befunge.. lemme google that
15:52:56 <zzo38> Ghoul_: Look up in esolang wiki
15:53:15 <tswett> If I remember correctly, wasn't one of the goals of Unlambda to be difficult to compile?
15:53:23 <Vorpal> Ghoul_, there is befunge-93 (sub-tc) and funge-98 (tc and probably one of the most high level esolangs there is)
15:53:35 <Ghoul_> Now that's just confusing.
15:53:59 <tswett> Oh, looks like that was the goal of (you'll never guess) Befunge.
15:54:08 <Vorpal> high level as in you can actually write reasonably complex stuff in it somewhat easily
15:54:14 <Vorpal> like fungot, the irc bot here
15:54:15 <fungot> Vorpal: ( see http://www.bloodandcoffee.net/ campbell/ code/ config-ex.tar.gz there's an example intercal and c
15:54:17 <fungot> http://git.zem.fi/fungot/blob/HEAD:/fungot.b98
15:54:23 <Vorpal> Ghoul_, that is befunge-98 ^
15:54:59 <Vorpal> (befunge-98 is 2D funge-98, there is also unefunge and trefunge, and potentially other dimensionalities too)
15:55:23 <Ghoul_> next you'll show me a language about compiling ascii pictures
15:55:33 <Ghoul_> which is probably not trivial in befunge
15:56:03 <Vorpal> Ghoul_, think of the code as a grid, where the instruction pointer can move. v makes it start to move down for example
15:56:16 <Vorpal> ^<> are similar for the other directions
15:56:54 <Ghoul_> I got that part, but it honestly just looks really inpractical
15:57:04 <Ghoul_> which is the point I guess.
15:57:05 <Vorpal> It is stack-based, so the code basically is read like RPN
15:57:45 <tswett> I wonder if there's a way to set vim to treat the file as being an infinite plane.
15:57:46 <fizzie> As far as esolangs go, Befunge is pretty practical for writing actual programs.
15:57:49 <Vorpal> Ghoul_, also befunge-98 has support for loading extensions (called fingerprints) to define A-Z. SOCK adds socket support, FILE adds general file IO and so on
15:58:03 <Vorpal> Ghoul_, there is a whole bunch of well known extensions
15:58:50 <Vorpal> oh and if you have any questions try me, Deewiant or fizzie. I wrote cfunge, the interpreter that fungot runs on. fizzie wrote fungot. Deewiant wrote CCBI (another interpreter) and also the standard test suite for funge-98
15:58:50 <fungot> Vorpal: wouldn't know. there's never any guarantees
15:58:57 <Vorpal> fungot, indeed there isn't
15:58:58 <fungot> Vorpal: not without _serious_ compiler analysis. i am confused
15:59:16 <Ghoul_> Now what /would/ be funny if someone gets it to the point where it's actually eligible for the clbg
15:59:24 <Ghoul_> and then it wins in line count on all the challenges.
15:59:33 <Ghoul_> computer language benchmarks game :P
15:59:44 <Vorpal> Never heard of it, how does it work?
15:59:51 <tswett> So wait, is fungot written in a funge?
15:59:51 <Ghoul_> Oh, you've never heard of it?
15:59:52 <fungot> tswett: but i'm gonna zip right through this and become an expert in this subject matter, being really lousy at remembering to remove old stuff from my parser generator... state transition tables i have to
16:00:02 <fizzie> tswett: It's called fungot...
16:00:04 <Ghoul_> http://benchmarksgame.alioth.debian.org/
16:00:07 <Vorpal> tswett, befunge specifically
16:01:10 -!- Nisstyre-laptop has quit (Ping timeout: 265 seconds).
16:01:42 <Vorpal> Wait, what is the *game* part of it?
16:01:45 <Ghoul_> heh, I might actually give befunge a go
16:01:55 <Vorpal> Just looks like a bunch of benchmarks to me
16:02:14 <Ghoul_> Vorpal: making new submissions and compiler improvements to get ahead of the others
16:02:21 <Ghoul_> It's a "game" in a sense.
16:02:55 <Vorpal> why is it all on ubuntu yet it is found on a debian.org address!?
16:03:13 <Ghoul_> so if you're moving left across something like this
16:03:21 <Ghoul_> do you get olleh or hello
16:03:39 <fizzie> You get 'H' on top of stack.
16:03:44 <fizzie> Since it was pushed last.
16:04:20 <fizzie> (Stringmode just pushes the letters one by one, in the order they are encountered.)
16:04:23 <Vorpal> Ghoul_, trival "Hello world!\n" printing program: 91+"!dlrow olleH">:#,_@
16:04:42 <Vorpal> 91+ to give a 10 on the stack (this is befunge-93, which didn't have a to push 10)
16:05:31 <tswett> You know, I find it pretty weird that English isn't some thing that I know intuitively. It *feels* like I just know it intuitively, but in fact I had to actually learn every little nuance of it.
16:05:32 <Vorpal> Ghoul_, then >:#,_@ is : to dup top item on stack, # to jump over next instruction (initially skipping the ,), then _ pops the top item and if it is 0 it will go to the right, otherwise left
16:06:03 <Vorpal> Ghoul_, then we hit the , which prints the value on the stack, we hit the # and jump over the dup, change back to going right, rinse and repeat
16:06:31 <Vorpal> Ghoul_, until we get to the end, popping on an empty stack returns 0, so then we go to the right, hit @ which terminates the program
16:06:55 <Vorpal> Ghoul_, without the @ there, the program would have wrapped around and hit the start of the line again
16:07:17 <Vorpal> Ghoul_, column wrapping too!
16:07:42 <Vorpal> Ghoul_, or if you use x in befunge-98 to set a delta of say, (2,3) then the wrapping gets quite confusing
16:07:57 <Ghoul_> so, does , print 1 char
16:08:00 <Ghoul_> or does it print until a zero
16:08:07 <Vorpal> Ghoul_, here is the spec: http://catseye.tc/projects/funge98/doc/funge98.html
16:08:16 <Vorpal> Ghoul_, , prints 1 char
16:08:33 <Vorpal> as the ascii value, . will print it as a decimal number followed by space
16:11:26 <Ghoul_> is there like a 3-dimensional funge?
16:12:06 <Vorpal> Ghoul_, yes trefunge-99
16:12:31 <Vorpal> Ghoul_, it is specified in the funge-98 spec
16:12:41 <Vorpal> since it is just one possible variant of funge-98
16:15:14 <Ghoul_> befunge is self modifying?
16:15:54 <fizzie> There's probably a large fraction of programs that never modify a playfield location that is executed.
16:16:18 <fizzie> E.g. if fungot has any self-modification at all, there's very little of it.
16:16:18 <fungot> fizzie: fnord fnord fnord fnord" in english? dictionary doesn't know it
16:17:04 <fizzie> fungot: Yeah, can't say I'm surprised.
16:17:29 <Ghoul_> Thats slightly annoying
16:17:40 <Ghoul_> woulda' been so easy until the self modifying part
16:17:50 <Vorpal> fizzie, it can reload itself, no?
16:18:11 <fizzie> Vorpal: Okay, there's that.
16:18:46 <Vorpal> Ghoul_, write a compiler? Yeah fizzie was/is working on a JIT compiler, about the only way you can do it
16:19:04 <fizzie> I had a llvm backend, yes.
16:19:23 <Vorpal> Ghoul_, anyway you can't really know all possible paths the program will take, since x allows you to pop two numbers from the stack and set that as the delta of the IP
16:19:33 <Vorpal> Ghoul_, so that also makes it near impossible to compile statically
16:19:42 <Vorpal> what if it uses user input for those numbers?
16:20:03 <fungot> http://git.zem.fi/fungot/blob/HEAD:/fungot.b98
16:20:06 <fizzie> Most of the work there is really tracking which traces need to be invalidated when something changes.
16:20:24 <Vorpal> fizzie, where is the launcher part?
16:20:44 <Vorpal> I want to test after my STRN changes
16:20:47 <fizzie> Vorpal: http://git.zem.fi/fungot/blob/HEAD:/fungot-load-freenode.b98 for example.
16:20:48 <fungot> fizzie: see the one im talking about the complexity of the job description. it's badly fnord also." " there"
16:20:59 <fizzie> Vorpal: Or s/freenode/local/ for the one that connects to localhost.
16:21:20 <fizzie> (The freenode server's IP might not be current.)
16:21:29 <fizzie> I've also written a static compiler that I think was good enough to (just barely) compile fungot.
16:21:30 <fungot> fizzie: looking at your wiki parser, too!
16:21:58 <fizzie> It uses a couple of heuristics to find the execution paths in cases like j with a non-constant argument.
16:22:14 -!- kallisti_ has joined.
16:22:16 <fizzie> (Basically, jump tables, of which the bot has two or three.)
16:22:27 -!- kallisti_ has quit (Client Quit).
16:22:44 <Vorpal> fizzie, should it print anything?
16:22:45 -!- kallisti_ has joined.
16:22:54 <Vorpal> It is doing nothing as far as I can tell
16:23:40 <Vorpal> fizzie, when does it say that?
16:23:51 <Vorpal> connect(3, {sa_family=AF_INET, sin_port=htons(6667), sin_addr=inet_addr("85.188.1.26")}, 16) = -1 ECONNREFUSED (Connection refused)
16:23:55 <Vorpal> sure that IP is correct?
16:24:05 <fizzie> [19:21:20] <fizzie> (The freenode server's IP might not be current.)
16:24:50 <Vorpal> hm it isn't joining any channel?
16:24:52 -!- epicmonkey has quit (Ping timeout: 260 seconds).
16:24:59 <fizzie> It doesn't join any automatically.
16:25:05 <Vorpal> how do I make it join?
16:25:20 <fizzie> "^raw JOIN #foo" in a query, but your owner-prefix must match.
16:25:48 -!- fungot-test has joined.
16:26:08 <Vorpal> fizzie, prefix is set to =, could you please test the stuff that uses STRN? I don't really know underload
16:26:17 <fungot> Vorpal: sure. i wish you would say that the " hotblack fnord" estate agency thing i saw was code in a matter of doing a big up-front design and top-down implementation was lispme. i'm porting it to gambit's web server instead of the original
16:27:40 <Vorpal> =ul (:::::::):(:((^:()~((:)*~^)a~*^!!()~^))~*()~^^)~(^a(*~^)*a~*()~^!()~^)a~**^!!^S
16:27:40 <fungot-test> :::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: ...too much output!
16:27:48 <Vorpal> factorial from the wiki
16:27:58 <Vorpal> =ul (:::):(:((^:()~((:)*~^)a~*^!!()~^))~*()~^^)~(^a(*~^)*a~*()~^!()~^)a~**^!!^S
16:28:11 <Vorpal> ^ul (()(*))(~:^:S*a~^a~!~*~:(/)S^):^
16:28:11 <fungot> */*/**/***/*****/********/*************/*********************/**********************************/*******************************************************/*****************************************************************************************/********************************************************************************* ...too much output!
16:28:17 <fizzie> ^ul ((0)(15)(14)(1)(2)(12)(11)(10)(3)(9)(8)(7)(5)(4)(13)(6))(~^:()SSa~a~*~a~*~a~*~a~*~a~*~a~*~a~*~a~*~a~*~a~*~a~*~a~*~a~*~a~*~a*~(x)S:^):^
16:28:17 <fungot> xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx ...too much output!
16:28:34 <fizzie> That was the wrong prefix.
16:28:38 <Vorpal> =ul ((0)(15)(14)(1)(2)(12)(11)(10)(3)(9)(8)(7)(5)(4)(13)(6))(~^:()SSa~a~*~a~*~a~*~a~*~a~*~a~*~a~*~a~*~a~*~a~*~a~*~a~*~a~*~a~*~a*~(x)S:^):^
16:28:38 <fungot-test> 6x0x15x14x1x2x12x11x10x3x9x8x7x5x4x13x6x0x15x14x1x2x12x11x10x3x9x8x7x5x4x13x6x0x15x14x1x2x12x11x10x3x9x8x7x5x4x13x6x0x15x14x1x2x12x11x10x3x9x8x7x5x4x13x6x0x15x14x1x2x12x11x10x3x9x8x7x5x4x13x6x0x15x14x1x2x12x11x10x3x9x8x7x5x4x13x6x0x15x14x1x2x12x11x10x3x9x8x7x5x4x13x6x0x15x14x1x2x12x11x10x3x9x8x7x5x4x13x6x0x15x14x1x2x12x11x ...too much output!
16:28:47 <fizzie> You need to put that ^C in.
16:28:54 <fizzie> (It generally doesn't copy-paste.)
16:28:55 <Vorpal> fizzie, I didn't see any C?
16:29:01 <fizzie> =ul ((0)(15)(14)(1)(2)(12)(11)(10)(3)(9)(8)(7)(5)(4)(13)(6))(~^:()SSa~a~*~a~*~a~*~a~*~a~*~a~*~a~*~a~*~a~*~a~*~a~*~a~*~a~*~a~*~a*~(x)S:^):^
16:29:01 <fungot-test> xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx ...too much output!
16:29:09 <fizzie> It's in that "empty" pair of () before the SSA.
16:29:14 <Vorpal> fizzie, right, xchat probably hides it
16:29:29 <Vorpal> =ul (12)S(*a(~:)~*^~):((1)S)~*~((2)S:*)~*:(~:()~)~*^(a(:^)*~a(*()~)~*^~^):^
16:29:30 <fungot-test> 122112122122112112212112122112112122122112122121121122122112122122112112122121122122112122122112112212112122122112112122112112212112112212211212212112212212112112212211212212112112212112122112112122121122122112122122112112122112112212212112122112112212112112212212112122112112122122112122121121122122121122122112122122112112 ...too much output!
16:29:35 <Vorpal> =ul (()(*))(~:^:S*a~^a~!~*~:(/)S^):^
16:29:35 <fungot-test> */*/**/***/*****/********/*************/*********************/**********************************/*******************************************************/*****************************************************************************************/********************************************************************************* ...too much output!
16:29:44 <Vorpal> ^ul (12)S(*a(~:)~*^~):((1)S)~*~((2)S:*)~*:(~:()~)~*^(a(:^)*~a(*()~)~*^~^):^
16:29:46 <fungot> 122112122122112112212112122112112122122112122121121122122112122122112112122121122122112122122112112212112122122112112122112112212112112212211212212112212212112112212211212212112112212112122112112122121122122112122122112112122112112212212112122112112212112112212212112122112112122122112122121121122122121122122112122122112112 ...too much output!
16:29:48 <Vorpal> Well it is the same anyway
16:30:09 <Vorpal> fizzie, I guess it works?
16:30:30 <fizzie> Seems so. Every command also uses at least G and N from STRN, since that's used to detect a command.
16:30:40 <fizzie> I guess it might be empty.
16:30:50 <Vorpal> fizzie, did you ever use F from STRN? You do use F but I don't know if it is from STRN
16:31:18 <Vorpal> fizzie, F is the one I'm most dubious of my implementation
16:31:43 <fizzie> Vorpal: It's used for finding the bot's name, at least.
16:31:52 <Vorpal> fizzie, I basically took the eglibc strstr and adapted it to funge_cell* instead of char*
16:31:55 <fizzie> Since fungot-test said something, apparently it works up to some degree.
16:31:56 <fungot> fizzie: otherwise they'd still teach pascal, scheme, and don't recognize the person, who seems to hate java... it's like a cross between them), but
16:31:59 <Vorpal> fizzie, oh, so I need some model files?
16:32:05 <fizzie> Oh, that was regular fungot, sorry.
16:32:06 <fungot> fizzie: anyone know anything regarding mzscheme and vim integration with a rapid development environment, dr. watson
16:32:20 <fizzie> I guess; though there might be some other place F is used.
16:32:23 <Vorpal> fizzie, can you upload a small model file or something so I can test that?
16:32:38 -!- fungot-test has quit (Quit: fungot-test).
16:32:44 <Vorpal> Also where do I put it?
16:32:58 -!- heroux has quit (Ping timeout: 276 seconds).
16:33:15 <Vorpal> <Gracenotes> oh, esoteric programming. carry on. <-- you know, what the channel is *supposed* to be about :P
16:33:46 <ion> Oh, wasn’t #esoteric-nonblah for that?
16:34:08 <fizzie> Vorpal: It needs to go in the current directory, and be named "model.bin.foo" and "tokens.bin.foo", and then you need a "styles.list" file that contains "foo\0a description for foo\0\n".
16:34:08 <Vorpal> Just checked, that channel was empty
16:34:25 <Vorpal> fizzie, link for a small one?
16:35:15 <fizzie> Vorpal: https://dl.dropboxusercontent.com/u/113389132/Misc/20130726-ct.tar.gz
16:35:26 <fizzie> (For the model "ct", aka the Chrono Trigger one.)
16:35:51 <Vorpal> tar: model.bin.ct: time stamp 2013-07-26 19:02:09 is 1595.876714959 s in the future
16:36:12 <Vorpal> fizzie, I think your clock is weird. I just checked mine with ntpdate and it is fine
16:36:16 <fizzie> I have a very mysterious clock skew issue.
16:36:24 <fizzie> That running ntpd doesn't fix.
16:36:50 <fizzie> I haven't bothered to figure it out, it's all kinds of strange.
16:37:02 <fizzie> Running ntpdate periodically does work, though, but I seem to have misplaced my cron thing for that.
16:37:16 <Vorpal> fizzie, so this should be fine? echo -ne 'ct\0Test description\0\n' > styles.list
16:37:48 -!- fungot-test has joined.
16:37:55 <fizzie> I'm not sure if anyone else has really tried the babbling part, so there might be issues.
16:37:58 <fungot-test> Vorpal: from where does the hero alone have the power. " m, madam...! i am the master of war! i've seen all kinds of battles from here, step back, prometheus!
16:37:58 <fungot> Vorpal: was also thinking about using a real irc client? i always am the last to display. this will get you nowhere in the conversation so far."
16:38:04 <Vorpal> fungot-test, once again
16:38:04 <fungot-test> Vorpal: the knight spirit has the hero! hurrah to the hero! hurrah to the hero! hurrah to the hero! hurrah to the hero! hurrah to the hero! hurrah to the hero! hurrah to the hero! hurrah to the hero! hurrah to the hero! hurrah to the hero! hurrah to the hero! hurrah to the hero! hurrah to the hero! hurrah to the hero! hurrah to the hero! hurrah to the hero! hurrah to the hero! hurrah to the hero! hurrah to the hero! hurrah to
16:38:04 <fungot> Vorpal: zero-terminated arbitrary numbers :d nice... are there any undergrads who look fnord or so macro?, for instance something i'm doing) say to use this
16:38:10 <fizzie> Well, that seems "all right".
16:38:11 <Vorpal> fungot-test, and the sword??
16:38:11 <fungot-test> Vorpal: to the northwest of this cape. he took back the medal from the frog king. and i'd like to see that mystical sword for myself! geez!
16:38:12 <fungot> Vorpal: did i incorrectly say it would be
16:38:30 <Vorpal> fizzie, but is it exactly the same message as fungot would have produced given the same random seed? XD
16:38:31 <fungot> Vorpal: i wonder if anyones ever set up a workflow that automatically compiles them ( obviously)
16:38:49 <Vorpal> Yeah it guess it works for the way you use it at least
16:39:25 <fizzie> The babbling is hardcoded to use the style "irc" at boot-time, I haven't bothered to persist the selection in the save file.
16:39:37 <fizzie> I guess you could try =def something, =save, then quit and restart.
16:39:42 <Vorpal> fizzie, I don't trust mycology though at all, since it managed to pass F while testing for "cd" in "abcdef" completely failed due to forgetting to multiply by sizeof(funge_cell) in a call to memcmp!
16:39:52 <fizzie> The save-file handling also uses STRN, I believe. (Or perhaps just FILE's line-level routines.)
16:40:43 <fizzie> There's no hardcoded =help.
16:41:02 <fizzie> But it should've said "saved", I think.
16:41:04 <Vorpal> how do you show all defines then?
16:41:18 <fizzie> But there are none yet, I think.
16:41:20 <fungot> ^<lang> <code>; ^def <command> <lang> <code>; ^show [command]; lang=bf/ul, code=text/str:N; ^str 0-9 get/set/add [text]; ^style [style]; ^bool
16:41:29 <fungot> (^<lang> <code>; ^def <command> <lang> <code>; ^show [command]; lang=bf/ul, code=text/str:N; ^str 0-9 get/set/add [text]; ^style [style]; ^bool)S
16:41:51 <Vorpal> =def help ul (=<lang> <code>; =def <command> <lang> <code>; =show [command]; lang=bf/ul, code=text/str:N; =str 0-9 get/set/add [text]; =style [style]; =bool)S
16:42:01 <fizzie> Hmm, I guess you could try the str functionality too, since it does strings.
16:42:25 <Deewiant> Vorpal: I think F in Mycology looks for "oba" in "foobar"
16:42:36 <Vorpal> Deewiant, I wonder how THAT worked then
16:42:44 <Vorpal> fizzie, could you test ^save and see if it prints anything from the normal fungot? Just so I don't go off chasing a bug that isn't there
16:43:02 <fizzie> Vorpal: You need to have a (sub)directory called "data", I think.
16:43:09 <fizzie> Since the save file is "data/fungot.dat".
16:43:10 <fungot> fizzie: i don't know how yet... :()
16:43:28 <Vorpal> fizzie, why not use DIRF or whatever it is called?
16:43:41 <fizzie> I'm not sure why it's in a directory anyway, since the model files aren't.
16:43:45 <Vorpal> fizzie, hm that is just a bunch of numbers?
16:43:55 <fizzie> Yes, it has a very number-heavy format.
16:44:22 <fizzie> Yes, it's certainly not too efficient.
16:44:25 <Vorpal> well lets reboot it and see if it remembers the stuff
16:44:31 -!- fungot-test has quit (Quit: fungot-test).
16:44:44 <fizzie> In the meanwhile, I'll see if the str: thing works on the real deal.
16:45:12 <fizzie> Maybe it was def-only.
16:45:40 <fizzie> Yeah, apparently it's def-only.
16:45:41 <Vorpal> fizzie, okay, it doesn't want to start, no output whatsoever
16:46:21 <Vorpal> open("data/fungot.dat", O_RDONLY) = 3
16:46:21 <Vorpal> fcntl(3, F_SETFD, FD_CLOEXEC) = 0
16:46:21 <Vorpal> fstat(3, {st_mode=S_IFREG|0644, st_size=533, ...}) = 0
16:46:21 <Vorpal> mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x7f4ad54db000
16:46:21 <Vorpal> read(3, "(foo)S\n\n\n\n\n\n\n\n\n\nhelp\n1\n40\n61\n60\n"..., 4096) = 533
16:46:21 <fungot> Vorpal: back later, bye all. be back later
16:46:26 <Vorpal> So related to the data file I guess
16:46:37 <fizzie> The state file is loaded first, so it's probably related to that. Note that it could easily be a fungot bug, it hasn't really been tested all that much.
16:46:37 <fungot> fizzie: we can't make recursive calls in gcc.... fnord?
16:46:50 <Vorpal> fizzie, -t 3 shows ... > 1-I-1 > 1-I-1 > 1-I-1 > 1-I-1 > 1-I-1 > 1-I-1 > 1-I-1 > 1-I-1 > 1-I-1 > 1-I-1 > ...
16:46:55 <Vorpal> I wonder where that is...
16:47:06 <fizzie> A reflecting I, most likely.
16:47:31 <Vorpal> ah yes -t 9 shows coordinates
16:47:34 <fizzie> I think it's FILE's L.
16:48:00 <Vorpal> tix=0 tid=0 x=26 y=114: I (73)
16:48:00 <Vorpal> Stack has 2 elements, top 5 (or less) elements:
16:48:03 <fizzie> Or maybe it's FILE's I and someone else's L was renamed to I.
16:48:26 <Vorpal> There is no I on line 113-115, so that y is bogus
16:48:34 <Vorpal> unless fungot.b98 is not loaded at origin?
16:48:35 <fungot> Vorpal: fnord), which is the body of a clause like this in my next upload
16:48:41 <fizzie> Vorpal: There's an offset of 100.
16:48:49 <fizzie> Vorpal: It's that "> 1-I" section on line 15.
16:49:28 <Vorpal> back, wasn't mine, and looked like a false alarm anyway
16:49:42 <Vorpal> fizzie, then why is y=114?!
16:49:59 <fizzie> Well, line 15 in a 1-based line numbering.
16:50:04 <fizzie> [19:48:41] <fizzie> Vorpal: There's an offset of 100.
16:50:23 <fizzie> After a closer look, I think the I is STRN's L.
16:50:34 <Vorpal> missed that line, derp
16:50:38 <Vorpal> I blame the heat. 26 C
16:51:08 <fizzie> I think it's gotten an empty line that it didn't expect.
16:51:31 <Vorpal> fizzie, there were a bunch at the start of the file, after the defined str:0
16:52:07 <fizzie> Those are expected. I think.
16:52:18 <fizzie> (It's for the other str:N's.)
16:53:19 <olsner> back to only one fungot?
16:53:20 <fungot> olsner: those little buggers out... for a few weeks
16:53:38 <Vorpal> How do I get a public link on dropbox? They must have changed it since I last did that
16:53:55 <Vorpal> fizzie, https://www.dropbox.com/s/v8lnd9s3qznsaw1/fungot.dat
16:53:55 <fungot> Vorpal: reachability graphs are not too bad
16:54:01 <Vorpal> Does that file look wrong?
16:54:14 <Vorpal> olsner, debugging why it is failing
16:54:46 <fizzie> Vorpal: Not immediately. Though the I in question is involved in loading the str: strings, in fact. But it really shouldn't mind empty lines there.
16:55:21 <fizzie> Vorpal: What it does is essentially G: [go elsewhere if 0] >1-I where the G is FILE/G and the I is STRN/L.
16:55:32 <fizzie> Vorpal: (It's the bit that strips a newline off, I think.)
16:56:20 <Vorpal> fizzie, My change to L is trivial in fact...
16:56:25 <fizzie> But if G returns 0, it would have gone elsewhere; and if G returns something >0, doing 1-I should be okay, since the string under the number-of-bytes return from FILE/G should be at least that long.
16:57:21 <zzo38> Do you have any ideas about Magic: the Gathering card in computer program? I have had some and mentioned some in here; do you have a different idea?
16:57:23 <Vorpal> I haven't changed FILE at all
16:57:40 <Vorpal> fizzie, I guess I could build the old version again and try that on the same data file
16:58:23 <fizzie> Vorpal: You can also ask for a trace with stack-dumps, and paste the bit around where the >1-I loop starts.
16:59:18 <Vorpal> fizzie, yeah the old cfunge also loops on that
16:59:47 <Deewiant> What about the super-old version that fungot runs on?
16:59:48 <fungot> Deewiant: remember to tell me it actually says " ninety... nine... bottles of fnord flood!
17:00:08 <Vorpal> fizzie, I let it run for 3 seconds and got a 43 MB trace file heh
17:00:19 <fizzie> Vorpal: Oh, right... it's going to -- well, it should -- ask for a 0-length substring. Do you reflect on that?
17:00:55 <Vorpal> Deewiant, I don't know what version that is!
17:01:02 <Deewiant> Vorpal: I imagine fizzie can tell you
17:01:03 <Vorpal> if (n <= 0 || len < (size_t)n) {
17:01:09 <Vorpal> That is the only time I reflect
17:01:20 <fizzie> Vorpal: Well, yes. I think n == 0 is valid.
17:01:50 <fizzie> (Admittedly it's kind of a corner case.)
17:02:06 <Vorpal> fizzie, hm I don't know. Does the spec say anything? Did I change it? Let me check the log for that file
17:02:13 <fizzie> "For R,L,M Requesting 0 or more characters from an empty string returns an empty string. Specifiying a negative size for R,L,M is an error and will reflect."
17:02:30 <fizzie> It might not have said this when you were implementing it; you know this spec, it's kind of a living document.
17:02:49 <fizzie> But the above sort of implies that L of 0 from a non-empty string should be okay, since it's not a negative size.
17:02:54 <Vorpal> Well what revision of cfunge do you use?
17:03:17 <Vorpal> fizzie, why only from a non-empty string?
17:03:19 -!- Deewiant_ has joined.
17:03:29 -!- yorick has quit (Remote host closed the connection).
17:03:37 <fizzie> Vorpal: Well, I mean, also from the empty string; but that's explicit in there.
17:03:41 -!- Deewiant has quit (Read error: Connection reset by peer).
17:03:46 <Vorpal> fizzie, anyway what about the case where you have "a" 9 L? Should that reflect
17:04:01 <fizzie> "For R,L requesting more characters than the length of the string will return the whole string. Other interpreters may have implemented this as reflecting or returning a null string."
17:04:18 <fizzie> I think I've guarded against reflecting on that in fungot.
17:04:19 <fungot> fizzie: sorry but that's not notably better than all the scheme fans i know in moin it would tell its url
17:04:25 <Deewiant_> I like how the spec says "the behaviour is X but some have implemented this as Y"
17:04:26 <Vorpal> fizzie, I guess I'll update that
17:04:42 <fizzie> Also I'm using cfunge 0.3.2. :p
17:04:44 <Deewiant_> I'm not sure if that should be read as "you should implement this as X" or "you can't know which will happen"
17:04:52 <fizzie> "I'll update real soon now."
17:05:30 <fizzie> What's your current version number?
17:05:57 <fizzie> (For the record, my state file seems to have a couple of empty lines too.)
17:06:11 <Vorpal> fizzie, but A LOT changed since that in bzr
17:06:35 -!- Deewiant_ has changed nick to Deewiant.
17:07:44 <Vorpal> So... For L would this cover it?
17:07:47 <Vorpal> stack_push(ip->stack, '\0');'
17:07:53 <Vorpal> } else if (len < (size_t)n) {
17:08:06 <Vorpal> Well I need to fix the memory leak in the middle case
17:08:09 <Deewiant> Do you need to special-case 0?
17:08:22 <Vorpal> Deewiant, yes because of the other case is:
17:08:26 <Vorpal> stack_push(ip->stack, '\0');
17:08:27 <Vorpal> stack_push_string_multibyte(ip->stack, s, (size_t)(n - 1));
17:08:35 <Vorpal> Pretty sure that n-1 is going to screw things up
17:08:58 <Vorpal> Since 0-1 cast to size_t would be LONG_MAX or similar
17:09:45 <Vorpal> Okay fixed L only so far, but that seems to make fungot happy
17:09:46 <fungot> Vorpal: and baby jesus cries, yes yes. spare me the fnord
17:09:58 <fizzie> Vorpal: That behaviour seems fine. (The reflect test in cfunge-0.3.2 was apparently if (n < 0 || len < (size_t)n) which is why fungot works.)
17:09:58 <fungot> fizzie: ( eat *this*, bash.org))) but it stil doesn't work. g
17:10:02 -!- fungot-test has joined.
17:10:05 <fungot> Vorpal: you do have a scgi server that talks to java applet clients concerned with efficiency, i'll try
17:10:22 <fungot> Vorpal: then call her and have fun or do something irreversible, the value,
17:10:35 <Vorpal> I guess the server is bad
17:10:44 -!- Nisstyre-laptop has joined.
17:10:47 <fungot-test> (=<lang> <code>; =def <command> <lang> <code>; =show [command]; lang=bf/ul, code=text/str:N; =str 0-9 get/set/add [text]; =style [style]; =bool)S
17:10:49 <fungot-test> =<lang> <code>; =def <command> <lang> <code>; =show [command]; lang=bf/ul, code=text/str:N; =str 0-9 get/set/add [text]; =style [style]; =bool
17:11:07 <Gracenotes> I should make a programming language where programs are executed by two IRC bots talking back and forth.
17:11:08 -!- fungot-test has quit (Client Quit).
17:11:23 <Gracenotes> for hundreds and hundreds of lines, potentially.
17:12:44 <Vorpal> fizzie, interesting, wrt n=0 R and L had different behaviours before
17:12:54 <Gracenotes> in a thematically meaningful way, though. So it's as if two computers with different capabilities were computing things in a cluster with ultra-law latency.
17:14:13 <fizzie> Vorpal: Incidentally, if memory serves me right, when I was switching fungot to cfunge I hit the issue that your STRN/L reflected on n > len, and had to change all my 17G4L" oof"C| is-this-the-command test to 17GN3`|4L" oof"C| instead -- because you refused to change it to conform to the new spec since it didn't have any versioning or changelogs and kept being silently changed. :p
17:14:13 <fungot> fizzie: minus273 you're in china? i have no idea
17:14:58 <fizzie> (In the code examples above, | denotes a go-somewhere-else-or-continue if.)
17:14:59 <Vorpal> fizzie, well I dislike silent changes still, but I guess my attitude changed somewhat
17:15:20 <Deewiant> Well Mike's dead now so it's probably not changing any more
17:15:45 <Vorpal> fizzie, where is that N from?
17:16:08 <fizzie> STRN/N, it's an explicit "is this long enough for L" pre-check.
17:17:37 <Vorpal> Deewiant, I notice you have these things as UNDEF in mycology as well
17:17:59 * Vorpal tries to figure out his own M code
17:18:13 <Deewiant> Like I said earlier I'm not sure how to interpret his "the behaviour is X but some have implemented this as Y"
17:18:17 <Vorpal> I'm starting to see the point of not using single letter variables now
17:18:21 <Deewiant> Does that mean X is right or both are right
17:19:11 <Vorpal> "For M, specifying a length that would go beyond the end of the string is legal and will return from the start til the end of the string. Other interpreters may have implemented this as reflecting." <-- I have that as reflecting I believe
17:21:22 -!- conehead has joined.
17:21:26 <Vorpal> Deewiant, I would suggest for M that if the start of the substring is after the end of the string, it should reflect. (The spec page doesn't mention that, as far as I can see?)
17:21:54 <Vorpal> Unless "For M, specifying a length that would go beyond the end of the string is legal and will return from the start til the end of the string." refers to that, as opposed to just a substring with a legal start that extends past the end of the string?
17:22:02 <Deewiant> Check what RC/Funge-98 does :-P
17:22:09 <Vorpal> In which case, it seems silly that they should have different behaviors
17:22:34 <fizzie> I feel that if you allow past-the-end extension for L and R, then you arguably should allow it for M too. (The start-past-the-end thing could still go differently.)
17:22:44 <Vorpal> Oh wait, the next line does specify it
17:22:51 <Vorpal> Right, I'll go with what the page says then
17:24:37 <Vorpal> Deewiant, with UNDEF: "ooF"24M eats the string, what do you mean exactly? That it leaves nothing on the stack?
17:25:18 <Vorpal> Well guess I have an off by one error then!
17:25:19 <olsner> Gracenotes: are you sure an apt-get purge is enough?
17:25:23 <fizzie> (In fact, as far as invariants go, it'd be nice if <str><0><n>M would always equal <str><n>L for any str, n.)
17:25:45 <Vorpal> Deewiant, surely if the "preferred" behaviour is used, it should leave a single o right?
17:27:06 <Vorpal> Must be an off by one error then...
17:27:10 <Vorpal> } else if (slen < (size_t)(p + n)) {
17:27:16 <Vorpal> Does that look wrong to you?
17:27:26 <Vorpal> p is the starting position, n the lenght
17:27:46 <Vorpal> slen is basically strlen() of course
17:28:43 <Gracenotes> ah, that's better. 'After this operation, 80.8 MB disk space will be freed.'
17:28:46 <Vorpal> Yeah... So why does it eat the string then?
17:29:18 <Vorpal> It was a rhetorical question
17:31:10 <Vorpal> How did it hit that code on GOOD: "zaBraBooF"34M is "BarB"
17:31:10 <Vorpal> I guess the condition is wrong??
17:31:10 <Vorpal> Oh I didn't rebuild XD
17:34:56 <Vorpal> Deewiant, I actually do push the o...
17:36:41 <Vorpal> Deewiant, I don't think your check is accurate, since the stack after that call is (TOS last) 14 0 111
17:36:49 <Vorpal> Not sure what that 14 is, but 111 is 0
17:37:05 <Deewiant> Inverted condition? :852s/!/z/
17:37:27 <Deewiant> Replace the first ! on line 852 with z
17:38:40 <Vorpal> Why is the file read only?
17:38:53 <Deewiant> I assume that is also rhetorical
17:39:01 <Vorpal> No I blame your .txz file
17:39:42 <Vorpal> (not an interrobang that is)
17:39:51 <Deewiant> I never remember which way _ goes on zero/nonzero
17:40:01 <Vorpal> BAD: J should push -1119007 given [-7777,2,29]
17:40:06 <Vorpal> Deewiant, infinite loop of that now
17:40:17 <Vorpal> Deewiant, sure you got the line right?
17:40:29 <Vorpal> Wait if I revert the change it still happens? What
17:40:45 <Vorpal> Deewiant, I think DATE broke?
17:40:54 <Deewiant> It should be the line with stuff like "gnirts eht stae" and a few M's :-P
17:41:01 <Vorpal> I haven't touched DATE
17:41:43 <Vorpal> Deewiant, does mycology.b98 contain something non-printable or such that my editor (kate) might have fucked up?
17:41:53 -!- heroux has joined.
17:42:10 <Deewiant> Vorpal: It's not UTF-anything and it has at least an embedded null IIRC
17:42:33 <Vorpal> I guess that might have caused it
17:42:58 <Deewiant> You can pull it from github, I pushed the change
17:43:11 <Vorpal> Deewiant, link to your github for this?
17:43:25 <Deewiant> github.com/deewiant/mycology I guess
17:47:06 <Vorpal> Well that seems to be it. Hm
17:47:13 -!- epicmonkey has joined.
17:47:30 <Vorpal> I'm still not sure about F hrrm
17:48:04 <Vorpal> I don't like that clang -O2 (LLVM 3.2 and 3.3) miscompiled previously, but stopped doing that when I changed a function in another translation unit
17:49:01 <Deewiant> Thus far I think every LLVM miscompilation bug I've thought I've had (or even filed, in at least one case) has been an undefined behaviour case in my own code
17:49:50 <Deewiant> (The one I filed turned out to be running over an array on the stack IIRC)
17:49:57 <Vorpal> Deewiant, well it just decided to add (unsigned int)-1 to %rsp about 5 times at the start of the function
17:50:12 <Vorpal> actually subtract not add
17:50:27 <fizzie> Vorpal: It was almost exactly 4*2^32.
17:50:34 <Deewiant> There's no unsigned values in x86 asm
17:50:35 <Vorpal> Oh okay, misremebered then
17:50:44 <Deewiant> Or no signed values if you prefer that train of thought
17:50:57 <Vorpal> Deewiant, I forgot the exact value I read in the gdb disassemble
17:51:12 <Vorpal> I was describing what it looked equivalent of
17:51:23 <Vorpal> but fizzie said it was 4*2^32 in my paste
17:51:41 <fizzie> It looked equivalent of uint32_t foo[unsigned -1];.
17:51:55 <Vorpal> Deewiant, anyway I took the code for strstr from eglibc and adopted it to funge_cell* instead of char*
17:52:10 <fizzie> Er, (unsigned)-1, I mean.
17:52:20 <Vorpal> fizzie, no arrays were allocated on the stack in that function though!
17:52:32 <Vorpal> Unless it was inlined from some header or such
17:52:35 <fizzie> Translation unit boundaries aren't such iron walls these days anymore either, right?
17:52:49 <Vorpal> True, but I don't think I enabled LTO
17:53:32 <Vorpal> Just checked the functions that it calls, no arrays on the stack there either
17:53:43 <Vorpal> Plenty of arrays yes, but all of them as pointer arguments
17:54:54 <fizzie> (echo 'CONDOR DONE'; sleep 300) | dzen2 -x 100 -y 100 -w 200 -h 200 -bg red -fg black heh, fancy bits and pieces in old throwaway scripts
17:55:16 <fizzie> I think that's called "poor man's notifyd".
17:55:55 <Vorpal> Tried that fancy newish cmake backend called ninja, supposed to be faster than make. Somewhat yes.
17:56:02 <Deewiant> fizzie: I see you're part of the menace that causes swapping and whatnot on innocent people's desktops
17:56:08 <Vorpal> Guess it would be easier to notice on something larger than cfunge
17:57:12 <Vorpal> What is Condor in this context? Not the bird I presume...
17:58:02 <fizzie> Vorpal: It's a "use idle workstations as a cluster" system.
17:58:12 <Deewiant> http://research.cs.wisc.edu/htcondor/
17:58:30 <fizzie> I guess technically it's HTCondor now.
17:58:41 <fizzie> "Note: The HTCondor software was known as 'Condor' from 1988 until its name changed in 2012. If you are looking for Phoenix Software International's software development and library management system for z/VSE or z/OS, click here."
17:59:42 <Vorpal> Ouch, smells of trademark dispute
18:00:54 <fizzie> Deewiant: It's even worse than that; my workstation these days is in SPA and not part of the ICS Condor, yet I still use it.
18:01:11 <Vorpal> fizzie, pushed the changes to cfunge, I would suggest getting the latest bzr for now, not sure I will do an actual 0.9.1 release any time soon
18:01:46 <Deewiant> fizzie: Shouldn't your ICS account be gone or something?
18:02:28 <zzo38> Have you ever used self-modifying code in GWBASIC?
18:03:10 <shachaf> Isn't an ASIC kind of contradictory to self-modifying code?
18:03:50 <Vorpal> Hm, tup supports variants nowdays
18:03:59 <Vorpal> I guess that would suffice for my needs then
18:04:17 <Vorpal> Not sure how to do the configuration... Really don't want to write a shell script for that
18:04:31 <Deewiant> Vorpal: If you come up with a good solution let me know
18:04:40 <Vorpal> Deewiant, for the config stuff? yeah right...
18:04:59 <Deewiant> I'll probably either go with perl or then just resign myself to cmake+ninja
18:05:02 <Vorpal> Since I currently use a fairly large cmake configuration (387 lines + several modules)
18:05:10 -!- KingOfKarlsruhe has joined.
18:05:29 <Vorpal> Deewiant, just tried out ninja as a backend, it seems to work, but it removes the color from the clang warnings
18:05:36 <Vorpal> not sure how that can be fixed
18:06:01 <Vorpal> Or is it just clang saying "this is not an interactive terminal, screw it!"
18:06:14 <Deewiant> It's clang saying this is not a terminal period
18:06:53 <Vorpal> Deewiant, make edit_cache opens ccmake again, ninja edit_cache just hangs
18:06:58 <elliott> btw just writing a configure.py with ninja_syntax.py is pretty easy
18:06:59 <Vorpal> Guess they didn't test that properly
18:07:05 <elliott> you don't really need a meta build system at least for simple stuff
18:07:09 <Deewiant> "Easy" is relative, changing CFLAGS does nothing in tup
18:08:14 <Vorpal> elliott, very true, cfunge has a non-trivial configuration though, such as "this function is in librt on linux and libc on freebsd"
18:08:24 <Vorpal> Deewiant, works for cmake+ninja though
18:08:50 <Deewiant> With a configuration system it's easy
18:08:56 <fizzie> Deewiant: We have an adjustment period.
18:09:09 <Deewiant> fizzie: You have both accounts now, then, or just ICS?
18:09:11 <Vorpal> Deewiant, Right, but you should be able to handle CFLAGS on tup right?
18:09:12 <fizzie> Vorpal: Department of Signal Processing and Acoustics. (Our group moved.)
18:09:23 <fizzie> Deewiant: Both, but there isn't much of a computing infrastructure at SPA.
18:09:29 <Deewiant> Vorpal: Not in pure tup, I don't think; not sure
18:09:37 <Deewiant> Can't remember, haven't touched it in a while
18:09:45 <Vorpal> Deewiant, I thought it loaded tup.config or something to support variants?
18:10:24 <Deewiant> Vorpal: Sure there's a bunch of files but I mean you can't change it from the command line
18:10:40 <Deewiant> Vorpal: You need to edit a config file, you can't just set an environment variable or pass a parameter or whatever
18:10:58 <Vorpal> Hm I should try building cfunge for my RPi
18:10:58 <Vorpal> yeah I think that is indeed the case
18:10:58 <Vorpal> Does ninja support distcc?
18:11:03 <Vorpal> Well I guess it would?
18:11:12 <Deewiant> Does distcc need special support?
18:11:53 -!- Bike_ has joined.
18:12:21 <Vorpal> Well I guess it is just a case of overriding CC
18:12:25 <Vorpal> that should work for ninja too
18:12:39 -!- Bike has quit (Disconnected by services).
18:12:41 -!- Bike_ has changed nick to Bike.
18:12:58 <Vorpal> Why is package signing on my RPi broken!?!
18:14:09 <Deewiant> BTW what's the line count on current cfunge
18:15:11 <Vorpal> Deewiant, don't have any tool to check that around. I changed distro back in january and never bothered installing a tool for that. Know any in the debian repos?
18:15:32 <Deewiant> I don't know about the debian repos but I usually use ohcount
18:19:16 <Vorpal> Well this is really confusing, the package keyring is broken on my RPi
18:24:16 <Deewiant> cfunge r874 is 9710 C code lines (15474 total C), or 12411 (19820 total) when including the stuff in lib/ that's from others
18:29:19 -!- Bike has quit (Ping timeout: 276 seconds).
18:29:32 <Deewiant> mushspace is only 9193 (12696 total), but it'll probably grow past cfunge before it's done
18:32:55 <Vorpal> Deewiant, heh, my gpg keyring file was corrupt
18:33:54 -!- sebbu2 has joined.
18:34:55 <zzo38> The Dungeons&Dragons session I have played in Wednesday was one player; the other player decided not to play.
18:37:06 -!- douglass_ has joined.
18:37:50 -!- copumpkin has quit (Ping timeout: 240 seconds).
18:37:54 -!- sebbu has quit (Ping timeout: 264 seconds).
18:38:26 -!- copumpkin has joined.
18:51:26 <Vorpal> Deewiant, hm seems I have a HRTI granularity of 3 µs on the RPi.
18:53:06 <Vorpal> No, just amusing to see something other than "1 µs"
18:53:35 <Vorpal> Deewiant, On a PC you get granularity in the ns range (though HRTI doesn't support that)
18:54:26 <Deewiant> On Windows I got 16 ms with my initial implementation in CCBI IIRC
18:55:10 <Deewiant> Yep, https://code.google.com/p/javasimon/wiki/SystemTimersGranularity
18:55:29 <Deewiant> "15.625 ms is reported ms timer frequency on multi-processor HW (or multicore as in my case) while 10 ms are reported on single processor"
18:55:36 <Vorpal> Oh? On Linux I always got like 1-3 ns internally on my old Sempron CPU. Never checked what the resolution is on my current CPU
18:55:36 <Vorpal> I use clock_gettime which returns the time in nanoseconds, and then I do a loop to determine the actual granularity
18:55:44 <zzo38> Do you have Famicompo mini vol.10?
18:56:34 <Vorpal> Deewiant, The timer granularity is 1 ns, but since I check that with a loop, it means that it can end up as slightly more for obvious reasons
18:57:08 <Vorpal> Deewiant, pretty sure you can use QueryPerformanceCounter or whatever on windows for better precision
18:57:36 <Deewiant> Vorpal: In that link it says "the call itself takes around 150 ns", Java's got some overhead there though
18:57:46 <Deewiant> And yes, I think I do something better on Windows nowadays
18:58:16 <Vorpal> Deewiant, right. Anyway on Linux clock_gettime doesn't even make a system calls. It reads a page with the counter in it that the kernel mapped into every user space process
18:59:53 <fizzie> Huh, they've made the "EATX" extra 12V power connector have 8 pins nowadays?
18:59:53 <Vorpal> Deewiant, yeah it does some fancy not-really-locking to read both s and ns in a thread safe way. IIRC it reads a serial counter first, then reads s and ns, then reads the serial counter again to check it didn't change, if it did it tries again
19:00:29 <Vorpal> fizzie, seems the power goes up and up. My GPU requires two 4 pin connectors to work
19:01:48 <fizzie> I'm trying to figure if my old PSU would still be good for some hardware upgrades, but it's tricky; especially since I have no idea what the PSU model was, and it doesn't seem to have any markings on the sides that are visible.
19:02:01 <Vorpal> Deewiant, it is part of that VDSO thingy
19:02:26 <Vorpal> fizzie, no brand or stickers?
19:02:51 <Vorpal> fizzie, if it is still in the case, maybe the sticker or brand is on the top/bottom side that isn't visible?
19:03:15 <fizzie> Vorpal: I'm sure it is, but it sounds like such a hassle to start unscrewing it.
19:03:16 -!- Bike_ has joined.
19:03:21 <olsner> fizzie: even if the motherboard has fancy extra EATX plugs, you can (at least sometimes) get away without connecting a psu to them
19:03:24 <fizzie> Plus I'm using this machine.
19:03:31 <olsner> at least until something tries to use too much power
19:03:46 <Vorpal> olsner, and then you are really out of luck
19:03:55 <fizzie> olsner: The manual (well, this particular one) says it's okay to connect the older 4-pin extra 12V connector to the 8-pin receptable.
19:04:20 <fizzie> The PSU has a 4-pin one that's connected to the 4-pin receptable that's in the current motherboard, I was able to see that much.
19:04:38 <olsner> Vorpal: well, I guess you have to get a new power supply then, but only if/when you get there
19:05:42 <Vorpal> Deewiant, are you sure 1/0 in FPSP can be NaN instead of +inf?
19:06:28 <Deewiant> Is IEEE-754 or the like mandated anywhere? :-P
19:07:11 <Vorpal> Well I would assume so
19:07:29 <Vorpal> Deewiant, So what does IEEE-754 say about this? I don't have it
19:08:31 <Deewiant> a/+0 is is signbit(a) * inf unless a is ±0
19:09:12 <Vorpal> Well then, -1/0 should be -inf and not NaN
19:09:52 <olsner> what about a/-0? is that -signbit(a)*inf?
19:10:55 <Vorpal> Deewiant, well I guess FPSP/FPDP doesn't mandate IEEE-754, but if you don't assume that it does, I suspect that several of the other test expectations might not be reasonable either
19:12:57 <Vorpal> Deewiant, asin(2) being NaN. If you don't have IEEE-754, you might not have +/- inf or NaN.
19:13:08 <Vorpal> You could have something completely different
19:13:16 <Vorpal> Like throwing an exception
19:13:38 <Vorpal> Not likely sure, but allowed yes, since there is then no actual specification on anything
19:13:56 <Vorpal> Deewiant, yes that would be a way to handle "system throwing an exception" indeed
19:14:08 <Deewiant> If it doesn't I'll fix it if somebody implements FP[SD]P on a non-NaN-having system ;-P
19:14:33 -!- Bike_ has changed nick to Bike.
19:14:35 <Vorpal> Deewiant, Erlang throws an exception on Nan or +/- inf
19:15:35 <Vorpal> Let me check how I handled that in efunge, if I even implemented those fingerprints
19:15:56 <Vorpal> I have not implemented those in efunge
19:16:00 <Vorpal> probably for that reason
19:16:48 <Vorpal> I did implement FIXP, and I'm doing try-catch on every asin or such that I call there, and push 0 in those cases
19:17:29 <Deewiant> That's a pain for interoperability
19:18:20 <Vorpal> Deewiant, I believe the reasoning is that no normal application should get those values, and it is better to detect those errors when they happen.
19:18:37 <Vorpal> And properly log them and reset the thread in question
19:18:54 <Vorpal> Implementing befunge is not a typical erlang program after all
19:20:31 <fizzie> "Kingston 2048MB (2GB) KIT 533MHz" "201.90 EUR" things certainly have changed.
19:20:37 -!- epicmonkey has quit (Ping timeout: 248 seconds).
19:21:33 -!- sacje has joined.
19:21:37 <fizzie> I think I found the power supply, and it claims to be 400W.
19:21:40 <Vorpal> fizzie, yeah, I would go for 8 GB modules nowdays. Probably 4 x 8 GB
19:22:10 <Vorpal> fizzie, a bit on the weak side maybe, depending on what you plan to put in the computer
19:22:17 <Vorpal> I have a 700 W PSU iirc
19:22:22 <Vorpal> or 650, I don't remember
19:22:44 <fizzie> I don't have all that concrete plans yet, but probably nothing terribly exciting, at least as far as GPUs go.
19:23:20 <Vorpal> fizzie, what about rotating HDDs? Especially when they are all spinning up at the same time at boot you will get a spike in the power usage
19:23:48 <fizzie> I've got a couple (3) of those.
19:24:45 <fizzie> Oh, interestingly: turns out my SSD is not actually attached to the chassis, it's just lying there at the bottom of the bay where the HDDs slot into.
19:25:03 <fizzie> Even though I'm pretty sure I have two or three 2.5"-to-3.5" brackets somewhere here.
19:25:18 <fizzie> I have the vaguest recollection that there was some reason why the brackets didn't fit, though.
19:26:00 <Vorpal> fizzie, are you getting a new case?
19:26:25 <Vorpal> fizzie, well if you are, I would suggest something with dust filters
19:26:31 <Vorpal> While not perfect they help a LOT
19:26:58 <fizzie> It's got some. (The one in front looks kind of terrible, I think I need to take it out and clean it.)
19:27:40 <fizzie> (There's also still quite a lot of dust inside, but I haven't probably opened it in the last two years or so.)
19:33:22 <Vorpal> I clean the filters every time I clean the room
19:33:55 <Vorpal> I need to clean the case maybe once per year to remove a thin layer of dust (+ quite a bit in the fans) rather than once every month or so to remove a lot of dust
19:34:16 <Vorpal> fizzie, not sure what type of dust filter you have, for me it is just a thin plastic wire mesh
19:34:30 <zzo38> I have finished typing recording of this session of Dungeons&Dragons game.
19:34:36 <Vorpal> Guess it would be more annoying to clean filters more similar to say that of a vacuum cleaner
19:35:51 <fizzie> I haven't really looked all that closely at the filters.
19:39:27 <kmc> I have a server with two SSDs that are just taped to the inside of the chassis
19:39:53 <kmc> dust filters are great
19:40:55 <kmc> you can get cheap filter kits that attach to an existing fan bay on any case
19:41:13 <kmc> eg http://www.newegg.com/Product/Product.aspx?Item=N82E16811988015
19:41:56 <fizzie> I don't really like hardware. :/
19:42:17 <fizzie> Why can't I just upgrade my computer by editing some config file somewhere?
19:42:34 <kmc> Nix should support that
19:42:59 <Vorpal> fizzie, so have you updated to the latest trunk of cfunge yet?
19:43:28 <fizzie> Well, no. I've been just inhaling dust.
19:43:43 <fizzie> And hitting my head on the bottom of a desk.
19:44:00 <fizzie> And other such stupid things that are involved in anything dealing with computer hardware grumble grumble.
19:44:21 <Vorpal> fizzie, why do you have it below the desk?
19:44:39 <Vorpal> I have it in a bookshelf next to the desk, higher up from the floor, means a bit less dust too
19:45:04 <fizzie> Because there's no space on top of the desk. Or in the bookshelf. Plus those have doors, I'm not sure how good airflow that would have.
19:45:24 <Vorpal> Ah yes, that sort of bookshelf would be an issue
19:45:35 <Vorpal> fizzie, you would get less dust though!
19:45:41 <Vorpal> What about watercooling?
19:46:12 <Vorpal> I have noticed I need to improve the cooling in my desktop, though I don't like the sound levels when I switch the case fans to a higher speed
19:46:18 <Vorpal> So I need some other way to do that
19:46:36 <Vorpal> fizzie, you never play computer games?
19:46:59 <fizzie> Sure I do, but I'm not a "gamer". It's a thing.
19:47:37 <fizzie> Now I completely forgot what I was supposed to do in this terminal. It's got "cd " written in it, so I suppose I was going somewhere.
19:47:40 <Vorpal> fizzie, What is the difference?
19:48:13 <fizzie> Gamers are the kind of people who don't understand any demoscene things. (The Assembly event is next weekend.)
19:48:25 <fizzie> (It's very much about the gamer-vs-demofolk dichotomy.)
19:49:08 -!- MindlessDrone has quit (Quit: MindlessDrone).
19:51:20 <fizzie> Now on revision 874. (Not on fungot yet, that is; just in general.)
19:51:21 <fungot> fizzie: fnord my firefox.)
19:51:42 <fungot> olsner: where does it say? it isn't written in cyrillic with the palatalizing character.
19:53:33 <Vorpal> fizzie, so once fungot restarts it will be on the new version?
19:53:34 <fungot> Vorpal: the winner will be optimizing by hand a basic c compiler in the works? surely not i. i would love a one-page version of it
19:54:08 <fizzie> Vorpal: Well, not quite. I need to get a compiled copy on the box fungot's running on, and I don't think there's gcc installed there.
19:54:09 <fungot> fizzie: on first line. second line is a list of symbols). is that a bug?
19:54:20 <fizzie> Usually I've built elsewhere.
19:54:38 <fizzie> Also I can't figure out where I put the version with the built-in chroot changes.
19:55:10 <fungot> Available: agora alice c64 ct darwin discworld enron europarl ff7 fisher fungot homestuck ic irc* iwcs jargon lovecraft nethack pa qwantz sms speeches ss wp youtube
19:55:15 <Vorpal> fizzie, I don't know how old your old version was, but at some point I started using libbsd if that existed for strlcpy. So check with ldd and check that it is installed on the target system if so
19:55:45 <fizzie> "libbsd0" seems to be installed.
20:11:02 <fizzie> I don't know if I should build with EXACT_BOUNDS or not.
20:11:20 <fizzie> (This copy won't probably ever be used for anything else than fungot.)
20:11:20 <fungot> fizzie: and he was too little to break it up into new starts in 15 minutes or so
20:14:31 <fizzie> I suppose I could try the result out.
20:14:32 <Deewiant> Of course you should, it should be the default with an UNSAFE_NO_EXACT_BOUNDS opt-out instead
20:14:39 <Vorpal> fizzie, eh, not sure if you use the bounds from y
20:14:49 <fizzie> Deewiant: It does default ON.
20:14:52 <Vorpal> Deewiant, pretty sure it defaults to ON
20:15:02 -!- fungot has quit (Quit: fungot).
20:15:06 <Deewiant> I recalled that it was off at least at some point
20:15:32 -!- fungot has joined.
20:15:40 <fizzie> Well, that's on r874 now.
20:15:44 <fungot> Vorpal: scale it up after i tap c-d a few times i ran it on
20:15:46 <fungot> Available: agora alice c64 ct darwin discworld enron europarl ff7 fisher fungot homestuck ic irc* iwcs jargon lovecraft nethack pa qwantz sms speeches ss wp youtube
20:15:52 <fizzie> fungot: Are you feeling quite normal?
20:15:53 <fungot> fizzie: cvs emacs cannot display images on the linux ime front. the best case
20:15:55 <fungot> ^<lang> <code>; ^def <command> <lang> <code>; ^show [command]; lang=bf/ul, code=text/str:N; ^str 0-9 get/set/add [text]; ^style [style]; ^bool
20:16:11 <fungot> echo reverb rev rot13 rev2 fib wc ul cho choo pow2 source help hw srmlebac uenlsbcmra scramble unscramble asc ord prefixes tmp test celebrate wiki chr ha rainbow rainbow2 welcome me tell eval elikoski list ping
20:16:13 <fizzie> Sounds like its usual self.
20:20:39 -!- sebbu2 has changed nick to sebbu.
20:26:18 <fungot> Selected style: darwin (Books by Charles Darwin -- you know, that evilution guy)
20:26:21 <fungot> Vorpal: during/ successive reprints :)/ first number :)/ primary wing-feathers is a constant tendency to fill up wide intervals in/ natural history :) these islands, as well as congenital peculiarities, are transmitted; so it appears to be a systematic naturalist, on/ principle :) inheritance--the occurrence :) rudimentary. -conversion :) stamens into imperfect petals, and occasionally quite black./ plainer males :) these japan
20:26:23 <Vorpal> fizzie, what is going on?
20:26:23 <fungot> ^<lang> <code>; ^def <command> <lang> <code>; ^show [command]; lang=bf/ul, code=text/str:N; ^str 0-9 get/set/add [text]; ^style [style]; ^bool
20:26:32 <fungot> shachaf: orange. -crossing :) domestic and wild cats. -of breeds :) pigs. -on interbreeding. -in/ himalayan. -crossing :) domestic and wild swine. -on/ dingo.
20:27:04 <fungot> Selected style: irc (IRC logs of freenode/#esoteric, freenode/#scheme and ircnet/#douglasadams)
20:28:57 -!- Nisstyre-laptop has quit (Ping timeout: 264 seconds).
20:50:22 <fizzie> From where I am, I saw your four ^style darwin ... fungot ping lines all on the same second.
20:50:22 <fungot> fizzie: after i logged off i realized i did something like that.
20:52:08 <HackEgo> 185) <zzo38> Invent the game called "Sandwich - The Card Game" and "Professional Octopus of the World" (these names are just generated by randomly) \ 214) <zzo38> ais523: Maybe it is better, because I don't think the octopus will live very well in the tree. But the difference is that the Internet is lying and you cannot see such things; you could m
20:52:20 <olsner> fungot: how can you realize something after you log off?
20:52:21 <fungot> olsner: acarrico's pointer to the image show right for both of you :p) have a fullscreen irc client.
20:53:07 <HackEgo> /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: zzote: not found
20:53:36 <Bike> `quote '<zzo38>'
20:53:44 <Bike> `quote <zzo38>
20:53:46 <HackEgo> 27) <zzo38> I am not on the moon. \ 116) <zzo38> Some people are reasonable, some people who are not reasonable insist on changing things so therefore progress depends on not reasonablepeple \ 150) <zzo38> catseye: Please wake up. Not recorded for this timezone. The big spider is not your dream \ 182) <zzo38> Maybe they should just get rid of Minec
20:53:52 <Bike> `run quote '<zzo38>'
20:53:53 <HackEgo> 27) <zzo38> I am not on the moon. \ 116) <zzo38> Some people are reasonable, some people who are not reasonable insist on changing things so therefore progress depends on not reasonablepeple \ 150) <zzo38> catseye: Please wake up. Not recorded for this timezone. The big spider is not your dream \ 182) <zzo38> Maybe they should just get rid of Minec
20:54:02 <shachaf> zzo38: Are you on the moon yet?
20:54:04 <Bike> oh, it's always the same order, huh.
20:54:14 <HackEgo> /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: quoerjandom: not found
20:54:32 <HackEgo> 897) <oerjan> fungot knows all. <fungot> oerjan: you are correct. there's no freedom for free. \ 1009) <shachaf> oerjan is spreading the tired rumour that if you play Nietzsche backwards you hear Jewish messages. \ 601) <oerjan> elliott: it occurs to me that `? welcome is atypical: its information is actually true. \ 18) <oerjan> In an alternate
20:54:54 <HackEgo> 183) <oerjan> elliott: i think i wrote a proof of 0*x = 0 on this channel once \ 831) <olsner> we have PR? <oerjan> the good news is we have PR. the bad news is we borrowed haskell's motto for it. [...] <oerjan> [...] "avoid success at all costs" \ 671) <fizzie> oerjan: Hey, what's your country code for telephonistic dialling from the outside wor
20:57:20 <fizzie> Vorpal: Re the power usage thing, checked a table of these things, and apparently the most high-end cards (Radeon HD7990) these days need three (3) of the eight-pin (8) PCIe extra power connectors.
20:57:35 <HackEgo> 184) * oerjan considered buying lutefisk, but apparently it cannot be prepared in microwave </bachelor frog> \ 451) <oerjan> sllide: @ is an OS made out of only the finest vapour \ 363) <elliott> oerjan: can you delete that and the meta turing completeness page <elliott> thanks <oerjan> elliott: IN UNIVERSO ALTERNATIVO, OERJAN PAGINAS DELET \ 787
20:57:41 <Bike> Shouldn't it only give one quote
20:58:18 <fizzie> (There's both 6-pin and 8-pin variants of the PCIe extra power cable; my PSU has one six-pin.)
20:58:39 <Bike> `echo bin/quoerjan
20:58:52 <Bike> there is something deeply wrong with me.
20:58:54 <Bike> `cat bin/quoerjan
20:58:55 <HackEgo> allquotes | grep oerjan | shuf
20:59:06 <Bike> `cat bin/quote
20:59:07 <HackEgo> #!/bin/sh \ allquotes | if [ "$1" ]; then \ if expr "$1" + 0 >/dev/null 2>&1; then \ sed "$1q;d" \ else \ grep -P -i -- "$1" \ fi \ else shuf -n 1; fi
20:59:31 <Vorpal> fizzie, hm, pretty sure mine are 4-pin? I guess I mis-remember
20:59:40 <Vorpal> fizzie, they definitely look square though
21:00:02 <Bike> `run echo allquotes | grep oerjan | shuf | head -n 1 > bin/quoerjan
21:00:08 <HackEgo> /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: /hackenv/bin/quoerjan: Success
21:00:25 <fizzie> Vorpal: There are square 4-pin EATX extra connectors that go to motherboards; maybe some GPUs use those instead?
21:00:38 <Bike> `run echo 'allquotes | grep oerjan | shuf | head -n 1' > bin/quoerjan
21:00:42 <Vorpal> fizzie, they say PCIE in white on the black connectors
21:00:46 <HackEgo> 324) <elliott> oerjan: but hypothetically, assume a Christian spontaneously materialised during the apocalypse
21:00:52 <HackEgo> 427) <monqy> cigaretes and drunking "lame highs for lame people" <oerjan> yeah if it doesn't make you go crazy and shoot at people, it's not worth it. take it from a norwegian.
21:01:02 <fizzie> Vorpal: Hmm. From what I've seen, that should be a 3x2 or 4x2 pinout thing.
21:01:07 <zzo38> Can you make a game out of sequent calculus? I have a way to do but do you have a different way?
21:01:14 <fizzie> But then again, hardware.
21:01:16 <Bike> `run echo 'allquotes | grep zzo38 | shuf | head -n 1' > bin/zzote
21:01:16 <Vorpal> fizzie, how square do the 3x2 ones look though?
21:01:22 <HackEgo> /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: /hackenv/bin/zzote: Permission denied \ /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: /hackenv/bin/zzote: cannot execute: Permission denied
21:01:29 <Bike> `run chmod +x bin/zzote
21:01:34 <HackEgo> 185) <zzo38> Invent the game called "Sandwich - The Card Game" and "Professional Octopus of the World" (these names are just generated by randomly)
21:02:04 <Bike> I'm glad I was here to witness it.
21:02:33 <fizzie> Vorpal: Well... http://www.techpowerup.com/reviews/PC_Power_Cooling/Silencer_750W/images/pcie.jpg has one of the "new-style" ones that have the 6+2 configuration, so that you can use them both in 6-pin and 8-pin holes.
21:02:39 <fizzie> Vorpal: I guess it's reasonably squareish.
21:02:42 -!- sprocklem has joined.
21:04:45 <fizzie> (The cards also take three slots at the back.)
21:05:17 <fizzie> http://images.bit-tech.net/content_images/2012/12/club-3d-radeon-hd-7990-6gb-review/hd7990-4b.jpg that's just ridiculous.
21:05:30 <fizzie> (The 3*8 power pins are at the left end.)
21:05:34 <kmc> what's the red button/light do?
21:05:43 <fizzie> kmc: Probably it's a self-destruct button.
21:05:58 <kmc> you aren't *really* going nuts until you have a second ATX power supply just for peripherals
21:06:03 <Bike> self-destruct is on my mind since i read about bomb sights today
21:06:16 <Bike> the brits offered to put self destruct equipment on american bombsights in WWII
21:06:20 <olsner> the norden bomb sight?
21:07:03 <fizzie> kmc: Apparently it's a "moar speed" button.
21:07:10 -!- yorick has joined.
21:07:21 <fizzie> kmc: "Press the red button, which loads a second BIOS, and the GPU clocks up to 925MHz, representing a minor gain and effectively becoming two full-fat HD 7970s on one board."
21:07:24 -!- sacje has quit (Ping timeout: 240 seconds).
21:07:42 <Fiora> is that like, a turbo button?
21:07:43 <kmc> a second BIOS, just what i always wanted
21:07:44 <fizzie> I have to admit I've been kind of missing the TURBO buttons of old.
21:07:51 <kmc> i love the Turbo button
21:07:58 <kmc> it's the perfect confluence of engineering and marketing
21:07:59 <Fiora> didn't the turbo button actually make it slower though?
21:08:10 <Fiora> like I remember if you pressed it on some old 386 thing I saw once it made it drop to like 8mhz
21:08:22 <kmc> yeah it was basically a "make it slow for compatibility" button
21:08:28 <Fiora> but it was called turbo
21:08:47 <kmc> well they just inverted the sense of the button
21:08:55 <fizzie> It still also makes it faster when you unpress it.
21:09:22 <fizzie> I had a system where the turbo button was replaced with a .com program that toggled (I think) 8/16 MHz.
21:10:29 <fizzie> Sometimes there was a seven-segment led display that showed the megahurz.
21:10:40 <Vorpal> fizzie, hm possibly such a 6 pin one
21:11:55 <Vorpal> <fizzie> http://images.bit-tech.net/content_images/2012/12/club-3d-radeon-hd-7990-6gb-review/hd7990-4b.jpg that's just ridiculous. <-- is that what you are getting?
21:12:00 <olsner> Bike: I like how that cost half the manhattan project, was super duper top secret, but both sides used the same kind of bomb sights anyway since some of the swiss inventors ended up working for germany since before the war
21:12:21 <fizzie> Vorpal: No, that'd probably melt my current PSU down.
21:12:31 <fizzie> Not that there's enough wires in it to even plug all the holes.
21:12:45 -!- oerjan has joined.
21:12:47 <fizzie> It's just a random HD7990 card I goggled.
21:13:25 <fizzie> Apparently every 6-pin plug adds +75W of maximum power usage (to the 16x PCIe base of 75W), and every 8-pin +150W; so even with three that's just...
21:13:42 <fizzie> Bah. Our sauna stove has higher wattage than that!
21:14:10 <olsner> how many polygons per second can your sauna stove push?
21:14:19 <fizzie> (I don't know what the proper English translation for the sauna heating element is.)
21:14:27 <olsner> I bet it's only like 50Hz too
21:14:32 <fizzie> olsner: It has a very realistically rendered rock texture on top.
21:14:41 <kmc> olsner: haha
21:15:27 <kmc> fizzie: so each pin is carrying about 2 A?
21:15:32 -!- Taneb has quit (Quit: Leaving).
21:15:34 <Vorpal> fizzie, my current PSU could manage that, but not sure with the rest of the components (4 HDDs, 4-core i5 at 3.3 GHz (sandy bridge) and so on)
21:15:56 <kmc> wait why should the 8-pin connector deliver twice as much power as the 6-pin
21:17:00 <kmc> "Devastate your enemies and smite those that question your gaming prowess"
21:17:38 <olsner> it's probably more about what they specified that the connector can deliver, than something to do with the connector itself
21:18:05 -!- KingOfKarlsruhe has quit (Quit: ChatZilla 0.9.90.1 [Firefox 22.0/20130618035212]).
21:19:00 <Bike> olsner: nazi technology was crazy anyway. they had night vision goggles, planes with guns that automatically fired if another plane was overhead, and MACLOS missiles
21:19:51 <olsner> but it seems it would've been more future proof to add some kind of protocol where the card says how much it wants to draw and the psu says how much it can supply
21:20:00 <fizzie> kmc: I have no idea. Maybe the extra pins are thicker?
21:20:39 <olsner> ooh, MCLOS must be what that Macross anime is named after
21:22:34 <kmc> maybe the 8-pin plug is also spec'd to have thicker wires?
21:23:40 <kmc> I remember reading that the Axis and the Allies both developed radar chaff but put off using it for years because they didn't want to reveal that they had it
21:24:37 <Bike> just gotta wait for the right time *entire eastern front explodes*
21:25:35 <kmc> also that german cryptographers weren't surprised that the Enigma could be broken, but were surprised that the Allies had put in the enormous effort necessary to do so
21:26:05 <Bike> i like to imagine that there was this thing going where the germans and the british were fielding all this high tech crap, and meanwhile the soviets are just flying planes too slow and low for german fighters to take down
21:26:34 <kmc> that did happen some
21:26:48 <Bike> i don't really know about the soviet end of R&D i guess
21:26:57 <olsner> they probably sent planes at all heights and speeds, it's just that the slow and low ones were the only ones to survive
21:27:20 <kmc> the russians would fly WWI era biplanes right over the german camps at night, just to fuck with 'em
21:29:15 <Bike> yeah i was thinking of the Polikarpov Po-2
21:29:41 <Bike> "Their usual tactics involved flying only a few meters above the ground, rising for the final approach, cutting off the engine and making a gliding bombing run, leaving the targeted troops with only the eerie whistling of the wind in the wings' bracing-wires as an indication of the impending attack"
21:29:51 <kmc> not quite WWI era I guess
21:29:55 <Bike> my favorite bit is of course "the stall speed of both the Messerschmitt Bf 109 and the Focke-Wulf Fw 190 was similar to the Soviet aircraft's maximum cruise speed"
21:31:28 <kmc> Bike do you know what's the most produced vehicle of any kind ever?
21:32:15 <Bike> probably some kind of horse-drawn cart?
21:32:36 <Bike> i guess that doesn't really count as a coherent class of vehicles
21:32:47 <kmc> I mean like a particular model
21:33:01 <kmc> anyway it's the Flying Pigeon bicycle from China, 500 million made
21:34:00 <Bike> "After the Communists led by Mao Zedong came to power in 1949, the bicycle industry was revived" i assume PRC forces used bikes exclusively
21:34:16 <Bike> mao's little red bike (are they red)
21:35:03 <kmc> the one on wikipedia is black
21:35:10 * pikhq cheers, for he is finally starting to recover
21:35:26 <Bike> aw, they're black
21:36:01 <kmc> what are you recovering from?
21:36:16 <pikhq> Random testicle pain.
21:36:17 <Bike> «In 1994, the government named the bicycle a "national key trademark brand under protection", enshrining it similarly to national treasures.»
21:37:08 <Bike> testicle pain is no good
21:37:55 <kmc> glad you're recovering, do you know what caused it?
21:38:24 * kmc looks accusingly at Bike (the joke is that bike seats can hurt one's testicles)
21:38:53 <kmc> but also they can cause the entire crotchular area to go numb in a way that is alarming but short-lasting
21:39:07 <pikhq> The medical guesses are either epididymitis (infection of the testes) or torsion of the testicular appendix (which hurts like torsion of the testicles, but is benign otherwise, because the testicular appendix is a vestigial structure)
21:39:23 <kmc> TIL there's a testicular appendix
21:39:37 <fizzie> Appendix T: Testicles.
21:39:46 <pikhq> And can't reasonably be anything else, because both ultrasound and CT scan say I'm in perfect health.
21:39:46 <kmc> Thesticles.
21:40:08 <HackEgo> Thanks, testicles. Thesticles.
21:41:42 <Bike> hm, i think i'm going to have to call it the " hydatid of Morgagni" because that sounds like some kind of frog
21:42:09 <pikhq> Needless to say, this has been a shitty few days.
21:49:39 -!- Nisstyre has quit (Quit: Leaving).
21:52:29 <oerjan> @tell madbr You want to look at http://esolangs.org/wiki/Collatz_function and http://esolangs.org/wiki/Minsky_machine hth
21:54:05 <oerjan> @tell madbr Also, http://esolangs.org/wiki/Fractran
22:01:28 <kmc> fractran is great
22:01:38 <kmc> what query prompted these?
22:02:21 <Bike> madbr was just asking about collatz
22:02:52 <kmc> is the relevance of fractran that it illustrates that generalizations of collatz are turing complete?
22:03:13 <kmc> hm it's not the most direct generalization
22:03:17 <kmc> I think some more direct ones are also TC
22:03:27 <kmc> in general things tend to be TC
22:03:45 <Bike> conway wrote a paper about it but i haven't found a good version
22:13:49 <oerjan> kmc: collatz functions are the direct generalization hth
22:14:17 <oerjan> also madbr was asking about things closer to fractran itself (and the similar 1-cell minsky machine)
22:14:25 -!- sprocklem has quit (Remote host closed the connection).
22:15:21 <oerjan> these are all very similar in how you make them tc, but fractran is uniquely simple
22:16:35 <oerjan> collatz functions and 1-cell minsky machines basically add extra complications that you don't need.
22:19:00 <kmc> oerjan: oh ok
22:19:21 <oerjan> in some way collatz functions simplify fractran by removing the order of rules applied. but on the other hand they add the complication of choosing on exact modulus rather than just zero/non-zero.
22:20:15 <oerjan> i discovered collatz functions when trying to find out how to prove 3-cell bf TC, for which they are a perfect fit (with fractran i couldn't get the actual halting to work :P)
22:28:43 <elliott> oerjan: "I once left my votes on Steve and Geoff's answering machine" I hope there is a really good story behind this
22:30:04 <oerjan> elliott: i had been chosen to be that week's Silly Person or whatever it was called. i took it quite seriously. also Steve had somehow just before posted his phone number to the list.
22:31:05 <oerjan> the other main result was Insane Proposals, which i see still has a tiny vestige in the ruleset.
22:31:32 <oerjan> <tswett> How about rational numbers? <-- rational numbers have the nice property that you can fit any countable order in them.
22:31:51 <oerjan> @tell tswett <tswett> How about rational numbers? <-- rational numbers have the nice property that you can fit any countable order in them.
22:33:30 <oerjan> @tell tswett And of course that you can always fit another between/beside a finite number, which is how you prove the countable order result. Which means you don't need surreal numbers if you have only finite/countable number of fixities. Of course _any_ total order regardless of cardinality fits in the surreals.
22:34:42 <Bike> oh, speaking of surreals. What's ω!
22:35:38 <oerjan> well it's represented by the game {1,2,... | } where the naturals are listed on the left.
22:35:52 <oerjan> basically any ordinal fits into the surreals in that form
22:35:53 <Bike> no, i mean, the factorial.
22:36:09 <oerjan> oh. i don't remember how that works.
22:36:21 <oerjan> (i probably never learned it)
22:37:35 <oerjan> i've vaguely heard you can do some calculus/series stuff but i have a hard enough time remembering multiplication.
22:48:06 -!- sprocklem has joined.
22:50:15 -!- Nisstyre has joined.
23:07:42 -!- ottianna has joined.
23:10:52 <zzo38> If I have three states, each character belonging to one of three character sets, and I have a transition matrix giving the cost of emitting a character from a given character set if you begin at a given state and end at a given state, do you know the algorithm to make the less cost?
23:12:22 <oerjan> you mean you can always emit a character and get to any state you want, but the cost varies?
23:12:58 <HackEgo> ¡Bienvenido al centro internacional para el diseño y despliegue de lenguajes de programación esotéricos! Para obtener más información, echa un vistazo a nuestro wiki: http://esolangs.org/wiki/Main_Page. (Para el otro tipo de esoterismo, prueba #esoteric en irc.dal.net.)
23:13:37 <zzo38> That is precisely what I mean.
23:14:41 <zzo38> ottianna: You have to learn reading English if you want to read it.
23:15:04 <zzo38> oerjan: Do you have any idea about how such a thing would be done?
23:15:05 <ottianna> ottianna: You have to learn reading English if you want to read it.
23:15:40 <zzo38> ottianna: (Of course if you don't know reading English then this message also isn't understandable.)
23:15:40 <oerjan> zzo38: given a string and a starting state, you can easily calculate this recursively from the information for substrings, i think.
23:16:00 <zzo38> oerjan: Yes but is there a more efficient algorithm?
23:16:39 <oerjan> zzo38: i think that's pretty efficient actually.
23:17:24 <zzo38> OK, do you have an example?
23:17:26 <oerjan> given that there is no reason why it should not vary enormously between different strings...
23:17:45 <oerjan> let's say you want to output abcd
23:18:55 -!- ottianna has left.
23:19:27 <oerjan> starting in state s and ending in state t. then recursively, you calculate the cost of getting from s to u outputting abc, and from u to t outputting d, for each u. then you take the minimum of the sums over u's.
23:20:19 <zzo38> The entire string does not matter what state it ends in.
23:20:20 <oerjan> note there is no exponential blowup, you just need to calculate N results for each initial substring, where N is the number of states.
23:20:55 <oerjan> zzo38: right, but you need a subroutine that calculates for given start and end, anyway.
23:21:34 <zzo38> Yes, OK, I can understand that much.
23:22:09 <oerjan> i suppose you can special case the last character added for the original problem if you want.
23:23:34 <oerjan> also it doesn't have to be recursive, it's easy to change it to iteration.
23:23:44 <zzo38> In order to calculate it for abc then I need to calculate it for ab...
23:24:04 <oerjan> yes, but you can do a loop to do it for a, ab, abc in order.
23:24:27 <zzo38> Yes, I thought of it something like that
23:26:01 -!- Sgeo has joined.
23:26:50 <zzo38> The starting state is known, so I can calculate the score for a, and then...
23:27:16 <zzo38> To do calculate ab from that...
23:27:35 <kmc> shachaf: have you read The Reasoned Schemer?
23:30:56 -!- Bike has quit (Ping timeout: 260 seconds).
23:31:03 <zzo38> If I am understand, then you calculate the cost s to u outputting a, and from u to t outputting b, and now you have the cost of ab for each t, I suppose.
23:31:59 <zzo38> But then you need to keep track of what set of states you have decided.
23:34:38 -!- mnoqy has joined.
23:34:38 -!- Bike has joined.
23:35:00 <Fiora> Bike: http://fioraaeterna.tumblr.com/post/56556056152/quasistars-a-real-life-black-hole-sun I tried to write up a thing
23:35:38 <pikhq> *Jesus* I'm high right now
23:36:05 <shachaf> kmc: A long time ago. I have the feeling in retrospect that I didn't absorb nearly much of it as I might nowadays.
23:36:24 <shachaf> I think I have a copy! Maybe I should read it again.
23:36:29 <kmc> pikhq: painkillers?
23:36:50 <kmc> enjoy it, I guess?
23:36:53 <kmc> don't get hooked
23:36:57 <kmc> how effective are they?
23:37:02 <kmc> Fiora: nice
23:37:04 * pikhq doesn't actually *enjoy* it.
23:37:31 <pikhq> I'm honestly not a fan of the sensation.
23:37:33 <Bike> "They reside at the core of nearly every galaxy, even in quasars less than a billion years after the Big Bang itself." might wanna throw "formed" in there
23:37:59 -!- noooodl has quit (Quit: noooodl).
23:38:31 <shachaf> i got some kind of drugz pillz when i had my wisdom teeth removed a few years ago
23:38:36 <Fiora> I wanted to imply like, um, "quasars photographed..."
23:38:40 <shachaf> i never used them (i think i still have them "not sure")
23:38:45 <Fiora> like we don't necessarily know when they /formed/ but we photographed them at a point in time?
23:38:49 <Fiora> I'm not sure what the right way to say that is
23:38:58 <Bike> "dating from"?
23:39:27 <Bike> i think anyone reading this is probably already familiar with far away things being from a while ago
23:39:41 <kmc> Fiora: did you study astrophysics in school?
23:40:15 <Bike> ok this is really cool and also metal as fuck.
23:40:26 <Fiora> Bike: see! black hole sun
23:40:28 * Bike put soundgarden on, to go with it. very appropriate
23:40:44 <Bike> that definitely makes sense
23:40:44 <Fiora> kmc: my dad was really into it and gave me a lot of books to read when I was little, and I devoured them for a little while
23:40:52 <Fiora> but in school I was kind of bad at physics and ended up doing CS
23:41:02 <Fiora> I never actually took an astronomy course
23:41:12 <shachaf> maybe it was because you ate all that paper when you were little
23:41:15 <shachaf> that can't be good for you
23:41:21 <Fiora> though if you like gave me a telescope I think I would quietly geek out for a few epochs
23:42:37 <tswett> oerjan: I don't remember the context in which we were talking about rational numbers.
23:42:50 <kmc> Fiora: cool
23:43:16 <oerjan> wtf did xkcd time end or not
23:43:22 <kmc> I too was bad at physics in school... I was distracted by CS and people and stuff
23:43:30 <tswett> lambdabot: I'm talking to you
23:43:44 <Fiora> the math defeated me >_<
23:43:51 <tswett> xkcd Time didn't seem like it was nearing its end, last I checked.
23:43:55 <tswett> And I feel like that was about a week ago.
23:43:57 <Bike> kinematics is, like, hard, man.
23:44:13 <Bike> and then there's all that... "electromagnetism" stuff............
23:44:21 <kmc> I was pretty bad at the multivariate calculus in E&M
23:44:38 <kmc> I should have actually took some time to learn it for real rather than trying to wing it 6 hours before homework was due
23:44:52 <Bike> i'm not sure i understand multivarate calculus well enough to do EM, even though I passed the class in it and all :/
23:45:00 <Fiora> I think I got a C in E&M...
23:45:04 <kmc> "matter falling into a black hole releases orders of magnitude more energy than nuclear fusion" hm, why is that?
23:45:19 <Fiora> gravitational potential energy
23:45:39 <Fiora> the gravitational potential difference between empty space and "event horizon" is huuuge
23:45:58 <Fiora> and I think, like, the way accretion disks work helps release a lot of that energy as x-rays and stuff?
23:46:06 <Fiora> like, I remember reading the other day in a different thing
23:46:27 <Fiora> matter falling on a neutron star hits soooo hard that it actually spallates, like, the nuclei themselves may spallate
23:46:37 <oerjan> tswett: your language's fixities i think?
23:46:40 <Bike> what's... spallating
23:46:45 <Fiora> so like, iron nuclei explode into a bunch of helium and hydrogen nuclei
23:46:51 <Fiora> like, shatter into pieces
23:46:53 <shachaf> kmc: Should I read it again? Why did you bring it up?
23:47:09 <Fiora> the energy of fusion is apparently ~7MeV/nuclei, and hitting-the-neutron-star-surface is ~200MeV?
23:47:20 <Bike> that seems like a pretty good way to explain easily how fucking dense those things are
23:47:23 <Bike> because, like, wow
23:47:23 <lambdabot> *** "200" wn "WordNet (r) 3.0 (2006)"
23:47:23 <lambdabot> adj 1: being ten more than one hundred ninety [syn: {two
23:47:25 <lambdabot> *** "7" wn "WordNet (r) 3.0 (2006)"
23:47:27 <Bike> wrong channel dearie
23:47:29 <lambdabot> adj 1: being one more than six [syn: {seven}, {7}, {vii}]
23:47:31 <lambdabot> n 1: the cardinal number that is the sum of six and one [syn:
23:47:33 <lambdabot> {seven}, {7}, {VII}, {sevener}, {heptad}, {septet},
23:47:35 <lambdabot> Plugin `dict' failed with: <<timeout>>
23:47:36 <Bike> yo lambdabot shut up
23:47:47 <Fiora> it hits the surface at like, 30% the speed of light I think?
23:47:50 <kmc> shachaf: no particular reason; I just happened to be thinking about that book today
23:48:03 -!- clog has quit (Remote host closed the connection).
23:48:14 <shachaf> does that mean there's a coöleg
23:48:14 <oerjan> tswett: there was a big "THE END" message about three updates ago. but there have been a couple updates since then, but not any action.
23:48:20 <Bike> could you like, somehow get into orbit and use the energy to gravitational slingshot at a speed of Pretty Fast
23:48:22 <shachaf> is coöleg related to clog (bye clog)
23:48:46 <oerjan> tswett: in fact i started reading it daily a few days ago because it was really picking up
23:49:02 <shachaf> is clog related to klogg (hi klogg: http://www.youtube.com/watch?v=JUgrkdzO6Ao )
23:49:05 <Fiora> Bike: well, like, gravitational slingshots are about stealing energy from either an object's velocity, rotation, or that thing where you do a burn at perihelion?
23:49:18 <Fiora> like, if you don't do any of that you'll end up with the same energy you started with
23:49:30 <Fiora> at least that's what I think it was?
23:49:46 <Bike> i have no idea.
23:49:52 <Fiora> Bike: http://upload.wikimedia.org/wikipedia/commons/3/3a/Gravitational_slingshot.svg
23:50:33 <Bike> so, a neutron star binary, "clearly"
23:50:41 * kmc should try to understand http://en.wikipedia.org/wiki/Oberth_effect
23:50:42 <Fiora> another one is where you run your engines at perihelion, which gives you more speed because like
23:50:51 <Fiora> you're leaving the fuel at a point of lower gravitational potential energy
23:50:56 <Fiora> so by conservation of energy you have to have gained more speed
23:51:09 <Fiora> oh. that's the obert effect
23:51:42 <Fiora> I think it's because Energy = Force * Distance, so if you exert the same force over a longer distance ('cause you're going faster) you get more energy
23:56:55 <Fiora> Bike: http://en.wikipedia.org/wiki/ULAS_J1120%2B0641 it's one of those super old quasars
23:57:50 <Bike> "Although this may appear to be larger than the size of the observable universe," fuckin
23:57:53 <oerjan> it was a quasar before it was cool
23:58:48 <Fiora> Bike: it's just the whole "comoving distance" thing
23:59:15 <Fiora> like the light took 13 billion years to get here but since the distance got bigger in the meantime the "actual" distance "right now" is 28b?