00:03:39 -!- oerjan has joined.
00:06:02 <zzo38> I looked at universal coding on Wikipedia. I guess that Elias gamma is best for small numbers, Fibonacci for medium size numbers, and Elias delta for large numbers? Is that it? (So, it would depend on how large the numbers tend to be that you are trying to encode, which one to use?)
00:08:33 <oerjan> Elias Gamma would be a good name for a mad scientist
00:10:02 <oerjan> of the "not evil, but dangerously absent-minded" kind
00:11:22 <zzo38> That's what I thought it might mean
00:12:37 <oerjan> he could have an evil twin brother with a more ominous first name, though.
00:12:50 <zzo38> Ah, OK it can do that too
00:15:50 <FireFly> oerjan: hmm, Saile Gamma doesn't sound terribly evil
00:16:15 <zzo38> quintopia: Yes that one is good job, I think.
00:18:11 <oerjan> quintopia: just as long as it's short for Rayburn http://www.babynamespedia.com/start/m/ray
00:19:03 <oerjan> although Raydon was also tempting
00:19:08 <quintopia> zzo38: I can't see any difference in the long term behavior of delta vs. fibonacci. do you know how it behaves at, say 100,000,000?
00:20:06 <zzo38> quintopia: I just saw a graph on Wikipedia that tell you how many bits are needed.
00:22:32 <zzo38> Of course it is only used when small numbers are more likely than large numbers (or if you want small numbers to be encoded smaller for other reasons, such as convenience or RLE), but I would think that which one you use would depend on how large the large numbers will tend to become.
00:23:22 <Taneb> Does anyone know any resources for learning Smalltalk?
00:25:55 <oerjan> <elliott> pretty sure that language literally exists <-- i vaguely recall there's a meta-brainfuck of similar kind, too
00:26:11 <oerjan> Taneb: alas, the only way is to go out and meet people.
00:26:25 <quintopia> Taneb: http://wiki.squeak.org/squeak
00:27:14 <quintopia> Taneb: especially this: http://daitanmarks.sourceforge.net/or/squeak/squeak_tutorial.html
00:28:00 <Taneb> Although going out and meeting people is probably a good idea when I am asking #esoteric for resources to learn a language that is past its heyday just so I can learn what the hang object-oriented programming is about
00:29:00 * oerjan is reminded of the Simula textbook that was on a shelf in the math department's student computer room
00:29:14 <oerjan> "Simula Begin", i think was its name
00:29:32 <oerjan> and that may literally have been my first introduction to OO
00:31:18 <oerjan> the computer department still used Pascal for its first semester course then
00:31:36 <Taneb> Like, I'm aware of the concept, but really to me there's this big shiny light somewhere in the distance that some people keep going on about as though it were God's gift to programmers
00:31:50 <Taneb> Called Object-Oriented Programming
00:31:57 <Taneb> And I have no idea what the big deal is
00:32:30 <oerjan> Taneb: well it was a huge leap forward over procedure-oriented programming which used to be the reigning paradigm (see: Pascal)
00:33:03 <Taneb> I have no idea what the big idea is because I have no idea what the idea is at all
00:33:13 <oerjan> however it's not so clear that it's a leap forward over functional programming, which however "only" lispers new about back then
00:33:40 <oerjan> (ml and haskell existed, but i hadn't heard of them)
00:37:00 <oerjan> Taneb: the basic idea is to have "objects" which tie together a data structure and the code acting on it, allowing an entirely new kind of modularity. (There's also inheritance, but everyone seems to think that's evil these days.)
00:37:46 <oerjan> (disclaimer: i haven't really looked much at recent developments in OO after i went functional)
00:39:54 <oerjan> OO allows mutation to be more "local", which is an improvement even if it doesn't go as far as pure functionality
00:40:45 -!- boily has joined.
00:43:28 <oerjan> "Visual Studio support channel"? (yes, i _do_ see JazzyFella i the logs)
00:43:50 <boily> something special about VS?
00:44:08 <oerjan> boily: he tried to treat this channel as general tech support
00:44:48 <oerjan> smells trolly to me, but i haven't got to the end yet
00:45:17 * boily oils his mapole and sharpens it
00:45:52 * oerjan sneakily steals back his swatter and hides it
00:47:11 <zzo38> There is also object oriented programming with COM/XPCOM interfaces (with IUnknown and QueryInterface and those things), and there is also prototype-based like JavaScript has.
00:48:12 <quintopia> boily: you have wonderful timing. i'm leaving for the airport in 7 hours to fly to mexico, so even though it's a saturday night, i'm headed to bed
00:49:00 <boily> once again, timing is everything...
00:49:08 <boily> where are you going to be mexicaning into?
00:49:36 <oerjan> <JazzyFella> It's this one line of code that's fudged up.. <-- if this was genuine, i have a hunch it wasn't really about VS at all, except afa my haskell questions would be about vim...
00:52:19 <oerjan> ciudad juárez *ducks for bullets*
00:52:26 <boily> ooooh! nice place!
00:52:37 <quintopia> anyway, catch ya back here in a week or so, unless i hop online from the hotel lobby and spot you
00:52:44 <boily> the family and I went there in... eh... about 2006 or so, I think?
00:53:01 <boily> bonnes vacances! enjoy the sun and the food!
00:53:08 <quintopia> but why would i chat on irc when there is so much mexico outside
00:53:17 <boily> be sure to visit Tulum!
00:53:22 <boily> (and bring a swimsuit.)
00:53:26 <quintopia> don't forget the liquor and swimming holes
00:53:57 <boily> Tequila, rhum, the cenotes― wait?
00:54:17 <boily> chicken pizza and cobol???
00:54:51 -!- tromp has joined.
00:54:52 <quintopia> http://en.wikipedia.org/wiki/Chichen_Itza http://en.wikipedia.org/wiki/Coba
00:55:04 <boily> oh. chicken pizza. I see.
00:55:05 <oerjan> quintopia: thanks, i didn't get the Cobol
00:55:33 * oerjan stealthily add an s after http
00:55:57 <oerjan> also after add, come to think of it
00:56:19 <boily> Dulnes: would you have prefered Dullones? I'm very flexible when it comes to welcomes.
00:56:54 <boily> hm. no good. Dulnellos it is, then.
00:57:08 <Dulnes> My username is just a mispelt version Dullness
00:58:14 <Dulnes> Visual studio support channel still hasnt been changed?
00:58:51 -!- oerjan has set topic: Visual Studio euthanasia channel | https://dl.dropboxusercontent.com/u/2023808/wisdom.pdf http://codu.org/logs/_esoteric/ http://tunes.org/~nef/logs/esoteric/.
00:59:11 <Dulnes> Windows 93 support desk would be a nice topic
00:59:15 -!- tromp has quit (Ping timeout: 256 seconds).
01:01:36 <Dulnes> boily: i prefer Dulnes but you go with that for now.
01:05:40 <oerjan> <quintopia> which is a simple geometric and beautiful typeface for large text? <-- times roman hth
01:05:50 <oerjan> especially good for large stone inscriptions
01:06:20 -!- S1 has quit (Quit: S1).
01:07:31 <Dulnes> Itchy on the eyes is how i would put it
01:09:35 <Taneb> Apparently Smalltalk uses only 2-3 concept.
01:09:47 <Taneb> Amazing how they managed to use only -1 concept!
01:10:30 <shachaf> -1-concepts sound reasonable
01:10:37 <shachaf> http://ncatlab.org/nlab/show/negative%20thinking
01:10:50 <oerjan> shachaf: did you know the empty topological space has dimension -1
01:11:03 <shachaf> oerjan: is that lebesgue covering dimension
01:11:28 <oerjan> boily: it fits very nicely into definitions
01:11:33 -!- adu has joined.
01:11:34 <shachaf> i guess it makes some sort of sense, maybe
01:11:46 <shachaf> since the one-point space surely has dimension 0
01:11:54 <shachaf> maybe the empty space should have dimension -∞??
01:12:33 <Taneb> I think the prime decomposition of zero has all exponents = infinity
01:13:09 <oerjan> shachaf: no, see, if you have an n-dimensional space, then by one of those inductive definitions that means the boundary sets in it are n-1 dimensional, and n is the smallest number that works for.
01:13:27 <oerjan> and that happens to give the right definition for 0-dimensional if the empty space is -1
01:14:07 <int-e> Taneb: hmm, I'd prefer 0 not to have a prime factorization. it is the greatest element in the divisibility lattice though.
01:15:17 <shachaf> oh, i guess you mean "at most n-1 dimensional" or something?
01:16:12 <oerjan> shachaf: well it works for lebesgue covering too, i see
01:16:55 <oerjan> "We want the dimension of a point to be 0, and a point has empty boundary, so we start with"
01:16:59 <oerjan> \operatorname{ind}(\varnothing)=\operatorname{Ind}(\varnothing)=-1
01:17:37 <oerjan> https://en.wikipedia.org/wiki/Inductive_dimension
01:20:59 <Taneb> fungot seems to be missing!
01:27:35 -!- bb010g has quit (Quit: Connection closed for inactivity).
01:29:47 -!- nooga has joined.
01:34:06 -!- nooga has quit (Ping timeout: 245 seconds).
01:34:44 -!- scounder has quit (Changing host).
01:34:44 -!- scounder has joined.
01:49:07 -!- Sprocklem has quit (Ping timeout: 256 seconds).
01:56:30 -!- tlewkow has joined.
01:57:07 <Taneb> Freefall theory: the fact that Florence now has a wiped ID chip will become relevant
01:58:11 -!- fungot has joined.
01:58:25 <fizzie> I'm practically asleep already.
01:58:48 <fungot> shachaf: may be." a little confusing to people who think that way
01:59:32 -!- tlewkow has quit (Remote host closed the connection).
02:00:35 <boily> fungot: of course people who think like you are confused. even people who don't.
02:00:35 <fungot> boily: maybe i should fix the texinfo if you don't
02:00:56 <boily> fungot: I ain't touching no texinfo. pfshaw! manpages for ever!
02:00:56 <fungot> boily: asian girls usually are in decimal anyway, though, i don't need your stinkin' context, riastradh. :) what particular differences are you having
02:01:20 <boily> asian girls are in decimal? de quossé?
02:02:08 <fungot> Available: agora alice c64 ct darwin discworld enron europarl ff7 fisher fungot homestuck ic irc* iwcs jargon lovecraft nethack oots pa qwantz sms speeches ss wp youtube
02:03:33 <boily> I should have known better than to expect anything else than irc.
02:03:57 <fungot> Selected style: ct (Chrono Trigger game script)
02:04:13 <shachaf> fizzie: nlab would be a good style
02:07:25 <oerjan> fungot: done any good sword stopping lately?
02:07:25 <fungot> oerjan: we are looking, but well behaved! crono!!
02:08:37 <fungot> int-e: like, thanks princess. i'll take that under advisement!! crono!! crono!!
02:08:56 <int-e> fungot: ok, that was two cronos, can you do three?
02:08:56 <fungot> int-e: that sword alone can't stop! that sword alone can't stop! that sword alone can't stop! that sword alone can't stop! that sword alone can't stop! that sword alone can't stop! that sword alone can't stop! that sword alone can't stop! that sword alone can't stop! that sword alone can't stop! that sword alone can't stop! that sword alone can't stop! that sword alone can't stop! that sword alone can't stop! that sword alone c
02:09:27 <fungot> Selected style: qwantz (Dinosaur Comics transcriptions 2003-2011)
02:09:28 <int-e> Ah, a different loop, I'll take what I can get.
02:09:41 <oerjan> um that _is_ the loop.
02:10:07 <int-e> But "crono!! crono!! crono!! crono!! crono!! crono!! ..." would've been so much nicer.
02:10:49 <oerjan> i haven't seen that, i suspect that isn't an unescapable one
02:11:00 <int-e> And sorry, but I'm no expert on fungot loops.
02:11:00 <fungot> int-e: i've never been to the bottom of a bottle. do i seem to see them, but they all suspect he's sad. they say " are you alright
02:11:52 <oerjan> fungot: well if he's sad he should report for termination immediately!
02:11:52 <fungot> oerjan: people in car car simulator trucko boat 3. that's a ' thr' followed that.
02:12:03 <int-e> fungot: who's "he" and what's he doing at the bottom of the bottle?
02:12:03 <fungot> int-e: i like that this is, like, a 50/ 50 mixture of both societal and biological self, all that gets us is a murky combination of influences, predisposition, anyways. the point is that i came, i'd have to throw away a good chunk of it for the days i've already lived.
02:12:31 <int-e> fungot: that did not help at all.
02:12:31 <fungot> int-e: dammit t-rex, how many? encase unspoiled areas, where good and bad at different points in the arcade version, this isn't a riddle. it's a real question i'm investigating!
02:14:02 <fungot> int-e: this may be a bad time to say all the run-of-the-mill thoughts we have! and it is a secret! i never
02:14:03 <HackEgo> Sir Fungellot cannot be stopped by that sword alone!
02:14:24 <int-e> That makes a lot more sense now.
02:15:42 <int-e> I like "run-of-the-mill secrets"
02:19:42 -!- Bicyclidine has joined.
02:26:28 -!- adu has quit (Quit: adu).
02:28:34 <int-e> sometimes I'm thinking too much. http://qdb.us/307308 ... I guess Thurmond or one of Thurmond's freinds is called Josh, which happens to approximate ёж.
02:30:45 <boily> I'll take "What is a Yozh for 400".
02:32:29 <oerjan> the zh is devoiced, it seems
02:32:56 <shachaf> how do you pronounce unvoiced zh
02:33:15 <oerjan> shachaf: um russian has mandatory devoicing of final consonants
02:33:39 <int-e> Nah, ж is voiced, that's why I wrote "approximate"
02:33:52 <oerjan> int-e: um it's not reflected in spelling
02:34:20 <oerjan> but e.g. the ipa here is unvoiced https://en.wiktionary.org/wiki/%D1%91%D0%B6
02:34:38 <oerjan> int-e: i know that. i'm just saying it's _pronounced_ unvoiced in that position.
02:35:43 <int-e> doesn't exist. жтж?!
02:36:08 <shachaf> i don't think щ is actually pronounced as штш
02:36:40 <shachaf> that's just what they want us to think
02:36:51 <int-e> (I did forget most of the Russion I've ever learned, but ёжик was a mnemonik for the ж letter which looks a bit like a hedgehog.)
02:36:54 <oerjan> shachaf: from what i learned on wikipedia while learning a russian song, the t is usually not pronounced these days
02:37:14 <shachaf> oerjan: ш and щ are pronounced slightly differently
02:37:22 <shachaf> i just can barely hear the difference
02:37:30 <oerjan> int-e: yeah but there the ж is not final
02:37:36 <shachaf> but i talk to russian speakers and they tell me which one is which
02:38:33 <shachaf> int-e: i thought the usual mnemonic was жук?
02:38:39 <shachaf> since it looks a bit like one
02:38:53 <shachaf> also "djuk" is a common hebrew term for cockroach
02:38:59 <oerjan> shachaf: ask them if there's a difference between ш and с in front of palatalizing vowels
02:39:10 <shachaf> oerjan: they aren't around hth
02:39:56 <int-e> oerjan: it may be voiced less than in the middle of a word, but I'd still expect an audible difference; (also I'd expect to have learned about such a difference and I'm sure I haven't)
02:41:46 <int-e> shachaf: that would make sense, but maybe hedgehogs appeal more to children (I was 8 or 9 at the time...)
02:42:02 <shachaf> hebrew has so many good words
02:42:43 <shachaf> "chupchik" should be imported
02:42:49 <shachaf> but i think it doesn't really fit in english
02:43:41 -!- tromp has joined.
02:44:41 <oerjan> shachaf: my impression is that щ is pronounced like two consonants, they're just both fricative-like
02:45:46 <shachaf> oerjan: there are recordings at https://en.wikipedia.org/wiki/Voiceless_palato-alveolar_sibilant and https://en.wikipedia.org/wiki/Voiceless_alveolo-palatal_sibilant
02:45:55 <oerjan> maybe i should have studied linguistics, in which case i would now be in a #linguistics channel somehow discussing math
02:46:29 -!- zlsa has left ("Leaving").
02:47:16 <oerjan> someone put a [citation needed] on norwegian for the former :P
02:47:26 <oerjan> does that mean it's not what i think it is
02:47:51 -!- tromp has quit (Ping timeout: 240 seconds).
02:48:58 <oerjan> "Application blocked by security settings" yeah i think that happened the last time i tried to hear sound on wikipedia too
02:49:39 <int-e> I think there is no actual 't' in щ, but you interrupt the air flow with your tongue just like for a t.
02:52:08 <oerjan> oh and the other article claims _that_ is the sound norwegian uses.
02:52:49 <shachaf> Bicyclidine: it depends on neither awea norwegian
02:53:17 <boily> I fear Norwegian can't be properly experienced except through being very, very regional.
02:53:38 <shachaf> Bicyclidine: it's me trying for a neither-nor pun and not succeeding hth
02:56:43 * oerjan swats shachaf -----###
02:58:07 <boily> I just understood shachaf's pun. I shouldn't be laughing.
02:58:52 <int-e> does "awea" have a meaning?
02:59:06 <Bicyclidine> say you're facing north on the equator. you go east while still facing north. eventually you end up back at your starting position still facing north.
02:59:10 <oerjan> int-e: it has a pwetty cwear meaning, yes
02:59:29 <Bicyclidine> if you walk a bit north before going east, you'll follow some other longitude line
02:59:50 <Bicyclidine> but if you walk all the way to the north pole without turning, and then sidestep right, you'll be on a latitude.
03:00:00 <Bicyclidine> am i missing omsething? this seems too abrupt, somehow
03:00:13 <int-e> ok, I have to agwee with oewjan, that was pwetty awful.
03:01:33 <oerjan> Bicyclidine: um is your problem that longitude isn't continuous at the poles?
03:01:44 <int-e> Bicyclidine: you're dividing by zero
03:02:05 <Bicyclidine> but somehow "right" going from meaning east to meaning south to meaning west within an infinitesimal space seems really weird.
03:02:41 <Bicyclidine> tiny circle, tinier circle, ENTIRE CIRCUMFERENCE OF THE MANIFOLD, tinier circle, tiny circle
03:03:02 <oerjan> Bicyclidine: hm it's not _actually_ east if you walk more than infinitesimally other than at the equator
03:03:41 <int-e> except on the equator, you'll never be walking on a straight line (great circle), why does this suddenly change when you reach the pole? it's a bit hard to walk on a circle with zero radius, granted, but that's what you ought to be doing.)
03:03:42 <shachaf> well what does "sidestep" mean
03:04:05 <Bicyclidine> oh, yeah. maybe the problem is i'm assuming you maintain your direction relative to the pole, except at the pole
03:05:10 <boily> shachaf: it's before you take a jump to the left, and put your knees together.
03:05:22 <Dulnes> Btw (off topic) im curious on the concept of 0 but whatever
03:05:48 <Bicyclidine> additive identity, multiplicative whatever, deliciously donut-shaped
03:06:06 <oerjan> you put your left foot in, your left foot out, you put your left foot in and shake off the icebear gnawing at it
03:06:16 <shachaf> Bicyclidine: it's another word for root hth
03:06:45 <int-e> Dulnes: it's an absorbing element for multiplication.
03:07:16 <shachaf> for the monoid of possibly-infinite lists, every infinite list is a left zero
03:07:18 <Dulnes> Well i feel like the only reason dividing by zero is impossible is because you are secretly dividing by infinity ( this is just my speculation )
03:07:38 <oerjan> no, you're secretly multiplying by it hth
03:07:42 <shachaf> it's actually a conspiracy by The Man intended to keep you down
03:07:54 <boily> dividing by zero doesn't work because you don't know which infinity you are multiplying with htah
03:07:56 <Bicyclidine> it doesn't make sense because there's too many things it could be
03:08:35 <Dulnes> Everytime you divide by zero a universe ends
03:08:36 <boily> that's where surreal numbers get quite useful.
03:08:37 <int-e> It just so happens that 0 * x = 1 has no solution (unless you allow 0 = 1, and very few mathematicians allow that kind of ring).
03:09:25 <zzo38> I think the reason dividing by infinity is not allowed is because 0x=1 has no solution rather than for other reason, although there are other reason too that is combine with.
03:09:28 <int-e> shachaf: if that's supposed to become a ring ... how does that satisfy the distributive laws?
03:09:34 <zzo38> So, there is many reasons.
03:09:58 <shachaf> Bicyclidine: whoa whoa whoa
03:10:05 <shachaf> what are you talking about
03:10:15 <zzo38> int-e: I suppose a trivial ring will have 0=1 though?
03:11:36 <oerjan> int-e: um surely the trivial ring is a ring, otherwise you don't have a variety. fields are another matter though. (also see ...wtf wikipedia went down)
03:11:43 <int-e> zzo38: it's quite common to have 0 != 1 as one of the ring axioms.
03:12:08 <zzo38> int-e: Really? I didn't think it is.
03:12:45 <int-e> but not universally agreed upon. (obviously since it's not even universally agreed that rings have multiplicative units)
03:13:21 <zzo38> Then you will have many definition of a "ring"
03:13:31 <Dulnes> I feel like you shouldnt even let a bot try to div by 0
03:13:48 <Dulnes> Its very deadly for most bots
03:13:48 <oerjan> https://en.wikipedia.org/wiki/Field_with_one_element
03:13:58 <shachaf> oerjan: So "all vector spaces are free" tells you that a linear map : V -o W is characterized by its action on a basis of V
03:14:15 <int-e> oerjan: for all primes except 1 ...
03:14:29 <vanila> all primes except 1? does that include -infinity or not?
03:14:40 <int-e> vanila: of course not
03:14:45 <shachaf> oerjan: But is there some similar thing that talks about the basis of W?
03:15:00 <boily> Dulnes: coffee is always good. drink more coffee.
03:15:11 <shachaf> I.e. the thing that makes matrices work.
03:15:41 <oerjan> shachaf: well "characterized" would seem to imply to me just one way, you need also that any map from the basis can be extended
03:16:23 <int-e> I was alluding to history. Quoting randomly from the internet (Yahoo answers): "Actually pre 19th centuary 1 was considerd to be a prime."
03:16:30 <oerjan> i mean you could take any generating set, and it would be characterized by its action on that, even if it's not independent
03:16:49 <shachaf> What does it mean for it not to be independent?
03:17:05 <shachaf> The elements of the basis can just be taken as formal elements or however people normally put that.
03:17:12 <int-e> But once you start getting serious about number theory, you'll end up with many statements that hold for all primes except 1. So it's more convenient if you just define 1 not to be prime.
03:17:17 <shachaf> Let's say : FA -o W, where F is the free functor.
03:17:23 <shachaf> Or maybe I don't understand your objection.
03:17:48 <oerjan> shachaf: my objection is that much of your claim is baked into the definition of "basis"
03:18:22 <oerjan> there are other sets such that a linear transformation is characterized by its action on them.
03:18:34 -!- nooga has joined.
03:18:44 <oerjan> and if you were _not_ working in a vector space, those sets might be all you have
03:18:46 <shachaf> You mean any superset of a basis?
03:18:52 <Bicyclidine> y'all manage to make linear seem even more complicated than kolmogorov could make it. impressive? yes.
03:18:58 <shachaf> I don't think you really need the word "basis" here.
03:19:17 -!- boily has quit (Quit: HYPOSEGMENTAL CHICKEN).
03:19:22 <shachaf> If you have a map : FA -o W, it corresponds to a function : A -> UW
03:19:28 <oerjan> shachaf: you need some way to imply that there exists a set such that any map from that set to W can be extended to the whole of V.
03:20:07 <shachaf> Isn't that part of what it means to say that FA is free?
03:20:10 <oerjan> which for vector spaces is the same as "basis" but that's not the case in other algebras (including non-free modules over rings)
03:20:14 <Dulnes> Maybe we could say 1 over infinity = 0
03:20:24 <Dulnes> Uh that is a problem because if we divide 1 into infinite peices and they end up 0 each what happened to 1? So nvm
03:20:46 <oerjan> shachaf: ok maybe you only had implication in one direction to start with
03:20:52 -!- adu has joined.
03:21:44 <Bicyclidine> there's no particular need to define one over infinity
03:21:51 <Bicyclidine> just go into projective geometry or something instead
03:22:07 <int-e> Dulnes: now you have a new concept that doesn't fit nicely into rings, namely infinity.
03:22:45 -!- nooga has quit (Ping timeout: 240 seconds).
03:22:53 <int-e> Dulnes: You *can* do this, but you're bound to lose some properties; in this case, that's likely to be inverses for addition.
03:23:43 <oerjan> Dulnes: 1 over infinity = 0 is pretty standard for functions on the riemann sphere. you still get trouble with 0/0, though.
03:24:06 <shachaf> oerjan: Anyway, as far as I can tell, that's half of why you can make matrices work between finite-dimensional vector spaces.
03:24:25 <shachaf> Since you can "take a vector apart" in terms of its basis.
03:25:41 <shachaf> But where does the other half come from?
03:25:56 <shachaf> Is that just from W being free?
03:26:07 <oerjan> matrices also work somewhat for commutative rings
03:26:23 <Bicyclidine> could you give a linear transform just by describing its eigenspaces and kernel, i wonder. would that work without a basis. i guess not since that probably adds up to a basis.
03:26:28 <oerjan> although i don't know about modules over them
03:26:59 <oerjan> shachaf: i think i'm too tired for this
03:27:10 <shachaf> oerjan: hey you were too tired last night too
03:27:15 <lambdabot> Local time for oerjan is Sun Nov 23 04:26:46 2014
03:27:22 <shachaf> maybe you should sleep hth
03:27:24 <oerjan> Bicyclidine: the kernel is one of the eigenspaces hth
03:28:03 <oerjan> shachaf: i may be permanently tired of very high math.
03:28:16 <shachaf> whoa, i didn't know you were high
03:29:59 <Dulnes> 1/x, 1 1.00000, 2 0.50000, 4 0.25000, 10 0.10000, 100 0.01000
03:30:07 <oerjan> Bicyclidine: my sleeping cycle has been utterly unstable for years; i _couldn't_ sleep now if i tried.
03:31:23 <Dulnes> I think coffee is bad for me
03:31:59 <int-e> shachaf: somewhere you need the property that any basis for a subspace can be extended to a basis of the whole space.
03:32:34 <Dulnes> i guess humans will never know
03:32:44 <Dulnes> Im done trying to calculate this
03:32:47 <int-e> (FTR, I generally don't find category theory helpful for this. I keep having to translate everything back into vector space terms.)
03:32:59 <Bicyclidine> question two: how does a simulation of relativistic physics work? is it possible to have an absolute spacetime that internal observers would see as einsteinian?
03:33:45 <Bicyclidine> becuase all the category stuff ounds like gibberish that's why
03:33:50 <int-e> Why what? It's abstract nonsense, is why.
03:33:51 <FreeFull> I'm tempted to make a small programming language where the type system really is algebraic
03:33:58 <shachaf> Why do you need that property?
03:34:00 <FreeFull> So the type of an 8-bit number would be 2^8
03:34:01 <oerjan> FreeFull: https://en.wikipedia.org/wiki/Wheel_theory hth
03:34:12 <int-e> (That's a quote, but who said that...)
03:34:15 <shachaf> Also I don't know vector space terms and I know at least a few category theory terms.
03:34:15 <vanila> FreeFull, That sounds really cool
03:34:17 <Bicyclidine> algebraic data types are sufficiently algebraic for me
03:34:29 <vanila> I would be interested in it
03:34:53 <zzo38> FreeFull: Yes, we should see it OK, try to make such a thing please
03:34:58 <int-e> ah maybe it wasn't any person in particular.
03:35:07 <Bicyclidine> how does it work again. whatever fuck math
03:35:29 <shachaf> oh i had another question about free things
03:35:37 <Bicyclidine> and 1+ax² is a binary tree and bla bla. bla.
03:35:42 <shachaf> is there some sense in which a topology generated by a subbasis is free over that subbasis
03:36:03 <oerjan> Bicyclidine: the twin paradox means that a simulation that includes it cannot give the correct einsteinian times for both observers
03:36:04 <shachaf> e.g. maybe you have a lattice of all possible topologies over a set or something
03:36:09 <FreeFull> Note you'd want to have some sort of tag system too, to tag if a number is unsigned or such
03:36:39 <oerjan> so the simulation could at best only be watched by one person in "realtime"
03:36:50 <int-e> shachaf: It's also that my knowledge of CT is rather limited. I shut down at the point anybody mentions adjoint functors.
03:37:05 <lambdabot> Local time for int-e is Sun Nov 23 04:36:36 2014
03:37:15 <shachaf> int-e: have you considered that adjoint functors are the best thing
03:37:18 <Bicyclidine> it just suddenly seemed weird that hashlife relies on their being a speed of light, but not lorentz covariance
03:37:48 <FreeFull> The twin paradox is resolved by considering acceleration
03:37:50 <int-e> shachaf: I've seen some enthusiasm displayed on the subject. I couldn't follow.
03:38:33 <shachaf> int-e: i think the catsters videos about them were good??
03:39:56 <oerjan> Bicyclidine: i once tried to find out if rule 110 had a nontrivial space-time symmetry
03:40:25 <Bicyclidine> i have a professor that does computational physics and he already indulged me about GR, I'll just ask
03:40:27 <oerjan> it is _possible_ you could find a rule 11 metacell that moved at a different speed
03:41:16 <oerjan> (i was intrigued by how the list of gliders known for it seemed to have a pattern in the speeds allowed.)
03:42:20 <zzo38> oerjan: What is a nontrivial space-time symmetry?
03:42:23 <oerjan> (incidentally rule 110 has a different maximal speed _inside_ the ether pattern
03:42:31 <Bicyclidine> i guess GoL isn't lorenz. because the speed of light is absolute. huh.
03:42:54 <oerjan> Bicyclidine: um the speed of light is absolute in reality too
03:43:13 <int-e> right, in general there's only one frame of reference in cellular automata
03:43:20 <Bicyclidine> If you had a colle- ok yeah that basically.
03:43:25 <oerjan> zzo38: anything other than translation, mirroring and rescaling...
03:43:44 <Bicyclidine> If you were a living automata collection and were moving at a constant velocity, you coul dtell that you were moving at a constant velocity.
03:44:21 <oerjan> zzo38: you'd want a map which took a complete time evolution of rule 110 and mapped it, locally continuously, into another complete time evolution moving at a different relative speed.
03:45:02 <oerjan> also, there would be some restriction on the patterns allowed to deal with the need to be inside the "ether" pattern.
03:45:36 <oerjan> i assume the ether pattern would be like a vacuum, and mapped to itself.
03:45:56 <oerjan> there would need to be some expansion in size, i think.
03:46:21 <int-e> oh crazy ... http://uncomp.uwe.ac.uk/genaro/Papers/Papers_on_CA_files/MARTINEZ.pdf
03:46:25 <FreeFull> Rather than something like rule 110, you could have some sort of particle physics thing?
03:46:54 <zzo38> Why is the "Mathematical Universe Hypothesis" called that? It isn't really a scientific hypothesis, as far as I can tell.
03:47:21 <FreeFull> zzo38: Well, it doesn't say scientific in its name
03:47:34 <Bicyclidine> "hypothesis" is a liberal lie, join the navy
03:47:37 <oerjan> <FreeFull> The twin paradox is resolved by considering acceleration <-- you still cannot make a multiplayer simulator that lets each player experience the correct proper time, though, because of it.
03:48:00 <FreeFull> oerjan: Yes you can, you just have to load one of the players into a rocket..
03:48:11 <zzo38> It doesn't really seem a hypothesis at all, really...
03:49:38 <zzo38> It doesn't really seem a conjecture either.
03:50:11 <zzo38> Not that there is anything wrong with the mathematical universe hypothesis, but the name seems a bit wrong.
03:50:18 <int-e> zzo38: it's a hypothesis, just not a scientific one.
03:50:51 <zzo38> int-e: Are you sure? I am not very sure.
03:50:55 <int-e> (at least my limited brain cannot imagine any way it could be tested)
03:51:13 <oerjan> int-e: i think that's the paper i looked at, way back
03:51:41 <FreeFull> They're not random, but we can't just find a pattern
03:52:08 <oerjan> <FreeFull> Rather than something like rule 110, you could have some sort of particle physics thing? <-- the thing is, if it's not 1d you have to deal with rotations and i have no clue how that would work.
03:53:07 <FreeFull> oerjan: You'd just have a bunch of straight line intersections, I assume
03:53:49 <oerjan> i mean to have lorentz-like invariance or at least an endomorphism
03:55:12 <int-e> zzo38: (I'd also call the Church-Turing thesis a hypothesis.)
03:55:17 <oerjan> int-e: when it was discussed over at aaronson's blog (he's a friend of tegmark but otherwise disagrees pretty strongly), someone brought up "dust theory" from the SF novel permutation city
03:55:25 <int-e> colloquially, "working assumption"
03:55:46 <zzo38> int-e: That I can see, kind of. But it is a bit different.
03:55:46 <oerjan> that would allow you a test, although not one that you could bring back the results of.
03:55:55 <Bicyclidine> well, there's my physics professor emailed. i'm sure he'll be happy to get a half-formed email about something he doesn't actually work in.
03:56:16 <oerjan> (which is pretty similar to black holes...)
03:56:52 <zzo38> About mathematical universe hypothesis, I think it is more philosophical kind of and isn't scientific (and that Occam's Razor wouldn't actually tell you anything about it and is irrelevant to this kind of discussion, even though other people say otherwise). I also do believe in such thing though, in a way; have you heard of my diagram involving GOD and how it relates to such thing?
03:57:26 <oerjan> hm i'm still waiting for that picture of a black hole that was promised "soon" over a year ago, i think.
03:57:39 <Bicyclidine> have you considered accepting our lord and savor feyerabend
03:57:43 <zzo38> And how you can see how it follow from that, its relation more to philosophy than scientific.
03:57:52 <oerjan> although i haven't been keeping up with science news, i assume i'd have heard of _that_
03:57:56 <int-e> Sure, it's philosophical. Is there hope of being a mathematical model of the whole universe or is the best we can do to come up with ever more precise approximations?
03:57:56 <zzo38> Bicyclidine: I do not understand.
03:58:14 <Bicyclidine> http://en.wikipedia.org/wiki/Epistemological_anarchism
03:58:20 <int-e> (These two options are not mutually exclusive either.)
03:58:24 <zzo38> int-e: Probably, coming up with ever more precise approximations; over time it can be figured out.
03:58:58 <oerjan> Bicyclidine: delicious? is this a cannibalistic religion like christianity?
03:59:05 <zzo38> (whether or not anyone figures out an exact mathematical model; but I think even if someone can you cannot really tell for sure?)
04:00:34 <zzo38> I believe it must mathematically exist, but that it doesn't necessarily mean anyone can ever figure out.
04:01:11 <int-e> zzo38: So you believe in the MUH.
04:01:18 <zzo38> I may have told you about my diagram with four concentric circles?
04:01:24 <zzo38> int-e: Yes, I have said that already.
04:01:43 <int-e> is it a Venn diagram?
04:01:51 <Bicyclidine> i misspelled "savior", but hey why mess wwith that
04:04:10 <zzo38> The contents of the diagram aren't really considered as sets although sets may be a concept that can be understood more easily than GOD.
04:05:21 <int-e> There's no understanding the elements of the empty set. Their remarkable properties are without end.
04:05:55 <zzo38> It is easily understood: There are none. That isn't the point, though.
04:06:24 <zzo38> Bicycldine: "GOD Over Djinn" (if you mean Godel, Escher, Bach)
04:07:58 <int-e> > fix id -- this was the essence of the GOD computation, it just happened to pick a different fixed point
04:09:40 <zzo38> I wasn't talking about Hofstadter's books though
04:11:24 <oerjan> int-e: iirc it was also implied that some steps would have errors
04:12:12 -!- bb010g has joined.
04:12:38 <vanila> Hofstadter Shmoffstadter
04:13:16 <zzo38> I do like his book though
04:25:36 -!- adu has quit (Quit: adu).
04:32:34 -!- tromp has joined.
04:37:29 -!- tromp has quit (Ping timeout: 265 seconds).
04:54:31 -!- augur_ has quit (Ping timeout: 245 seconds).
04:59:02 -!- tromp has joined.
05:00:39 -!- shikhin has quit (Ping timeout: 256 seconds).
05:01:35 -!- tromp has quit (Read error: Connection reset by peer).
05:02:09 -!- tromp has joined.
05:07:29 -!- nooga has joined.
05:12:17 -!- nooga has quit (Ping timeout: 265 seconds).
05:14:18 -!- augur has joined.
05:28:14 -!- tlewkow has joined.
05:51:14 -!- MDude has changed nick to MDream.
05:56:39 -!- Patashu has quit (Ping timeout: 258 seconds).
06:02:53 -!- tlewkow has quit.
06:04:28 -!- Dulnes has quit (Quit: Connection closed for inactivity).
06:21:27 -!- tlewkow has joined.
06:40:59 -!- Dulnes has joined.
06:41:15 <Dulnes> Happy birthday to me... Ew
06:46:13 <int-e> Ew indeed... http://iambaker.net/rainbow-birthday-cake/
06:47:03 <oerjan> `` echo 'Happy birthday to you, happy birthday to you, happy birthday to Dulnes, happy birthday to you!' | rainwords
06:47:04 <HackEgo> Happy birthday to you, happy birthday to you, happy birthday to Dulnes, happy birthday to you!
06:49:07 <Dulnes> And int-e if my child requested of me a "pretty cake" i would give them a cupcake with a hot dog sticking out of it
06:49:17 <oerjan> int-e: not a believer in the 7-color rainbow, i see
06:49:53 <Dulnes> I like the gray scale rainbows
06:50:41 <Dulnes> The awful ones you see at a Florida airport
06:55:42 <int-e> oerjan: apparently not
06:56:24 -!- nooga has joined.
06:57:43 * J_Arcane writes list functions he still can't use anywhere ...
07:00:46 -!- nooga has quit (Ping timeout: 245 seconds).
07:02:19 <int-e> Oh, I had not read friday's GG comic, I like the general.
07:05:11 <J_Arcane> What's a good test list for making sure foldl and foldr are working as expected?
07:08:15 <lambdabot> [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,...
07:08:43 <shachaf> > foldr (:) [] (1:2:undefined)
07:09:07 <oerjan> > foldl (flip (:)) [] [1..10]
07:09:48 <J_Arcane> Heh. I'm in a lisp, no lazy evaluation. That's a good point though, that was the usecase LYAH used.
07:12:27 <J_Arcane> Hmm. I thought I'd try subtraction. Got this: http://pasterack.org/pastes/41876
07:14:15 -!- tlewkow has quit (Remote host closed the connection).
07:14:42 -!- tlewkow has joined.
07:14:44 <J_Arcane> That does appear to be the correct result if I step through it in my head.
07:14:54 <J_Arcane> Wonder why Racket's foldl/r don't do the same.
07:15:00 -!- shikhin has joined.
07:15:22 <oerjan> possibly it uses different argument order?
07:15:42 <oerjan> > foldl (flip (-)) 0 [1..5]
07:15:58 <oerjan> try with an even length list
07:17:03 <shachaf> [] is the perfect test case
07:18:02 <shachaf> another good test case is the free pointed magma
07:18:24 <shachaf> i.e. data N a = N a | NEmpty | NAppend (N a) (N a)
07:19:05 <J_Arcane> http://pasterack.org/pastes/37221
07:19:07 -!- tlewkow has quit (Ping timeout: 264 seconds).
07:25:49 * oerjan swats shachaf in the type system ~~~~~¤¤¤
07:26:31 <oerjan> confused the value [] with the type constructor []
07:26:48 <J_Arcane> shachaf: Ahh, I see what you mean. http://pasterack.org/pastes/16294
07:28:08 <J_Arcane> I think it's in how I wrote the null case vs. how Racket wrote it.
07:29:09 <J_Arcane> http://pasterack.org/pastes/10137
07:29:31 <shachaf> I wish I could configure Chromium to open links in Incognito mode.
07:29:39 <shachaf> That way I could just click them instead of copying and pasting.
07:31:48 -!- oerjan has quit (Quit: Nitey).
07:35:05 -!- shikhin has quit (Remote host closed the connection).
07:40:05 <J_Arcane> Ahhhh. I see now. Racket flips the arguments in foldl. Which is why it makes a clean list, and why the results come the way they do.
07:42:36 <J_Arcane> ie. Heresy (and Haskell) recur with something like (foldl fun (fun base (car lst)) (cdr lst)), I can get the Racket result by instead doing (foldl fun (fun (car lst) base) (cdr lst)).
07:49:04 <Sgeo> Or use for/fold?
07:49:13 <J_Arcane> I guess it's to make sure that folds make clean lists when working with cons.
07:49:20 <Sgeo> How often does foldl/foldr itself get passed to another function?
07:50:35 <Bicyclidine> depends, are you writing pointless haskell
07:51:01 <J_Arcane> Heh. Guy Steele has a talk about 'foldl and foldr considered slightly harmful'
07:51:21 <J_Arcane> And a teacher's post I was just reading flat out just said 'don't use foldl.'
07:51:22 <shachaf> "monoids considered better"
07:51:24 -!- drdanmaku has quit (Quit: Connection closed for inactivity).
07:51:51 <J_Arcane> Instructor notes from this class: http://www.scs.stanford.edu/11au-cs240h/
07:52:06 <shachaf> Yes, foldl is almost never the right function to use.
07:52:27 <Sgeo> But how is foldr harmful?
07:52:51 <Sgeo> Although it sucks trying to translate foldr into a non-lazy language because then you need to figure out how to get it to terminate
07:52:51 <J_Arcane> Sgeo: Dunno, but Steel's talk is here: http://vimeo.com/6624203
07:53:21 <Sgeo> I guess foldr is a bit linear?
07:53:39 <J_Arcane> http://www.scs.stanford.edu/11au-cs240h/notes/par.html
07:54:09 <J_Arcane> I guess performance wise they can both be problematic in Haskell too?
08:05:40 <J_Arcane> Reading Racket's sources does make me understand why Haskellers love that type system ... so much ass covering.
08:09:15 <Bicyclidine> it's cool. just look at any haskell definition involving tuples.
08:12:44 <zzo38> I saw the new Smash Brothers game, and they didn't add Professor Oak and Imakuni? as playable characters.
08:23:09 <zzo38> "In the Victorian era (when the reigning monarch was Queen Victoria) The British Rule was widely used in contract bridge games throughout the British Empire. In certain parts of India, local bylaws enforced its usage, punishable by a fine." How and why did they enforce this?
08:45:18 -!- nooga has joined.
08:49:26 -!- nooga has quit (Ping timeout: 240 seconds).
08:57:56 -!- HashNode has joined.
08:58:14 -!- HashNode has left.
08:58:48 -!- scoofy has joined.
09:02:16 -!- Bicyclidine has quit (Ping timeout: 272 seconds).
09:04:29 -!- Dulnes has quit (Quit: Connection closed for inactivity).
09:15:23 <J_Arcane> Hah hah. Was looking at F# a minute ago, and just now looked at the channel topic. XD
09:27:50 -!- Patashu has joined.
09:39:39 <vanila> I tried the guy steele talk
09:39:43 <vanila> "Hey, Sandra Bullock LiedHer Fans Are In Shock. Her Huge Secret Is Finally Exposed!"
09:39:53 <vanila> but when I clicked it was about divina mcall
09:40:47 <vanila> Sandra Bullock is better/more interesting
09:41:11 <int-e> and what does that have to do with Guy Steele?
09:47:22 <vanila> its not valid scheme OR CL
09:51:49 <vanila> http://vimeo.com/6624203
09:56:55 -!- vanila has quit (Remote host closed the connection).
10:15:19 <J_Arcane> "Fun" conversation starter: have you ever lost a piece of code you were really proud of, and weren't able to rewrite it again?
10:19:13 <zzo38> I do not remember.
10:34:11 -!- nooga has joined.
10:38:55 -!- nooga has quit (Ping timeout: 264 seconds).
11:38:09 -!- nooga has joined.
11:38:22 <mroman> J_Arcane: The source code of Stlisp probably
11:38:35 <mroman> although I guess I could rewrite it. It's just a matter of time/effort.
11:38:48 <mroman> and the Stlang source code
11:39:22 <J_Arcane> When I was younger I wrote a really clever routine for haggling with a shopkeeper in an RPG that's still better than any such mechanic I've seen since. Then my brother powercycled the floppy drive with the disk still locked ...
11:39:56 <mroman> http://mroman.ch/cgi/cgitest.slisp?name=eso
11:40:03 <mroman> I still have the interpreter executable though
11:43:05 -!- Patashu has quit (Ping timeout: 250 seconds).
11:43:50 <zzo38> J_Arcane: Do you remember of details? How does it work?
11:49:54 <J_Arcane> zzo38: Dimly. The basic gist was that the shopkeeper had a target price range, and a mood score, and there were various threshholds where either an offer would be accepted or refused, or even that he would get offended and throw you out.
12:03:16 <zzo38> J_Arcane: Is that all?
12:04:42 <J_Arcane> zzo38: It's the basic idea, but it's still more than I've ever seen in a commercial game (there's probably some clever roguelike with a better one I've never heard of though, there usually is)
12:10:23 -!- sebbu has quit (Ping timeout: 250 seconds).
12:16:33 <zzo38> I mean how do these target price range, mood score, thresholds, offended, etc working?
12:19:06 <J_Arcane> Oh, the items all had standard prices, and used/sale price was computed as a randomized percentage of that, with min and max values relative to it and the shopkeeper's mood score.
12:19:41 <J_Arcane> Mood itself was I think randomized, and an ablative value almost like hit points: offend him enough and he'd tell you to piss off.
12:20:08 <J_Arcane> "Offending" basically meant asking too much or too little at the off.
12:20:33 <J_Arcane> So there was a strategy to it; you could work the shopkeeper down lower in steps than you could with a flat offer.
12:22:33 <elliott> I think some version of angband had something like that, maybe.
12:23:18 <elliott> that said, I'm not sure the shop interface should be a particularly trivial kind of combat :p
12:24:13 <elliott> (but I'm particularly prone to being annoyed by things like that)
12:25:15 <J_Arcane> It was semi-optional, you could always just accept the offer when selling, and it only applied to used purchases (new stuff was fixed price). Also, it was otherwise a pretty simple text game of 'kill stuff, level, loot', so in context it was a nice added feature.
12:26:03 <elliott> yeah, I just get annoyed by tradeoffs of "I'm doing worse than I could have been if I had the patience to play this subgame for the billionth time" easily :p
12:28:37 <J_Arcane> Certainly in something that wasn't essentially Recettear, it could be annoying. For me, I was always annoyed with games with 'barter systems' that weren't: Fallout's was like this, it wasn't really any kind of bartering because everything was fixed price and you were lucky if you could deviate even a single cap in your offer.
12:32:26 <elliott> I'm boring; I'm happy with shops-as-vending-machines.
13:10:31 -!- S1 has joined.
13:11:23 -!- sebbu has joined.
13:11:57 -!- sebbu has quit (Changing host).
13:11:57 -!- sebbu has joined.
13:18:50 <J_Arcane> elliott: It's definitely not a feature I'd want to have to use every time (this is actually probably why I got bored with Recettear) for sure.
13:51:10 -!- boily has joined.
13:59:00 <J_Arcane> Wow. ECMA BASIC doesn't even *have* mid$ and instr$.
14:06:55 -!- scoofy has quit (Ping timeout: 256 seconds).
14:07:47 -!- scoofy has joined.
14:08:45 <fizzie> QBasic doesn't have "INSTR$" either, the function's called "INSTR" (since the result is a number).
14:09:21 <boily> J_Arcanello! there's an ECMA sanctified version of BASIC?
14:30:14 <J_Arcane> boily: Yes. ECMA-116: http://www.ecma-international.org/publications/files/ECMA-ST-WITHDRAWN/ECMA-116,%201st%20edition,%20June%201986.pdf (warning, v. large file)
14:32:09 <J_Arcane> There was also ECMA-55 for "Minimal BASIC" as well.
14:32:23 <J_Arcane> I even found a modern compiler for ECMA-55.
14:36:55 <J_Arcane> http://buraphakit.sourceforge.net/BASIC.shtml
15:20:31 -!- ais523 has joined.
15:22:47 -!- HackEgo has quit (Ping timeout: 250 seconds).
15:28:29 -!- HackEgo has joined.
15:31:08 -!- boily has quit (Quit: TILED CHICKEN).
15:48:25 <b_jonas> If I normalize a string with the unicode CNF resp DNF normalizations, at most how many times longer can it become in UTF-16 code units?
15:50:34 -!- ais523 has quit (Remote host closed the connection).
15:50:40 -!- ais523 has joined.
15:53:27 <ais523> what's up with the topic/
15:55:52 -!- elliott has set topic: Visual Studio support channel | https://dl.dropboxusercontent.com/u/2023808/wisdom.pdf http://codu.org/logs/_esoteric/ http://tunes.org/~nef/logs/esoteric/.
15:55:57 <elliott> someone made it inaccurate; fixed
15:56:15 <ais523> what incident prompted the topic change in the first place?
15:56:52 <elliott> someone came in looking for visual studio support
15:57:07 <elliott> I hypothesise that this will no longer happen now that the topic mentions visual studio
15:57:34 <ais523> it can't have been my "attempting to install Visual Studio lead to me needing to reformat the Windows and boot partition, then spend a few hours figuring out how to reinstall Ubuntu's bootloader after a boot partition reformat"
15:57:35 -!- drdanmaku has joined.
15:57:57 <ais523> elliott: wait, what was the old topic pre-mentioning-VS?
15:58:12 <elliott> something about being the black hole of programming madness
15:58:22 <elliott> ais523: somehow that actually sounds on-topic
15:58:49 <ais523> I am in favour of the topic being ontopic
15:59:59 <elliott> something can't be on itself, anyway
16:01:06 <FireFly> Isn't the topic on-topic by definition?
16:01:12 <ais523> yes, I considered it ontopic (both the current topic, and the /actual/ topic)
16:01:23 <ais523> and I think I mentioned it in here
16:01:38 <ais523> which reminds me, I have a mathematical problem
16:02:34 <ais523> I have partial information about the output of an LCRNG (i.e. there is a sequence of numbers defined as r[i+1] = (r[i] * m + s) % x); specifically, for some subset of i, I know whether or not r[i+1] is greater than r[i]
16:02:48 <ais523> and I want to go from this, to discovering the individual r[i] value
16:03:10 <ais523> I'm interested in both the situation where I know what m/s/x are, and the situation where I don't
16:05:17 <ais523> not really sure how to start, though
16:05:47 -!- Phantom_Hoover has joined.
16:11:13 <HackEgo> [wiki] [[Befunge]] http://esolangs.org/w/index.php?diff=41114&oldid=40795 * GermanyBoy * (+431) /* summary */ added infobox
16:11:30 -!- tlewkow has joined.
16:39:04 -!- tlewkow has quit (Remote host closed the connection).
16:52:40 -!- callforjudgement has joined.
16:55:19 -!- callforjudgement has quit (Read error: Connection reset by peer).
16:55:22 -!- scarf has joined.
16:55:24 -!- scarf has quit (Changing host).
16:55:24 -!- scarf has joined.
16:55:37 -!- ais523 has quit (Disconnected by services).
16:55:38 -!- scarf has changed nick to ais523.
17:06:39 -!- MDream has changed nick to MDude.
17:20:14 -!- Bicyclidine has joined.
17:22:20 <newsham> `` echo . . - - "'" "'" "'" - - . . |rainwords
17:23:51 <newsham> ais523: if the state space is small (ie. 32 or 48 bits) you can always try solving all possible matches by brute force
17:24:11 <elliott> could SAT/SMT solvers help with that problem?
17:24:58 <ais523> elliott: possibly, but my experience with them isn't that good
17:28:20 <ais523> for this sort of problem
17:28:32 <ais523> newsham: I already do that with 32 bits, but I'm reaching a situation where I have a 64-bit internal state
17:28:35 <ais523> too large to bruteforce
17:28:43 <ais523> and yet this is miles from cryptosecure, it should be reversible
17:31:47 <newsham> yah, wouldnt surprise me at all, but i have no idea how to reverse it
17:32:34 <MDude> I guess the next obvious thing over just bruteforcing all of it would be to look for ways to eliminate possibilities and bruteforcing the rest.
17:34:30 <newsham> ais: any info about the initial seed? if they're using lcrng they might also be seeding with time or pid or somethign silly
17:34:43 <newsham> that might limit your search space greatly
17:34:59 <ais523> newsham: actually, my situation is that two RNGs are being seeded from the same seed
17:35:07 <ais523> one's pretty secure, the other is an LCRNG
17:35:16 <ais523> also the seeding method itself is pretty secure
17:35:32 <ais523> I want to figure out the sequence of the secure RNG by observing the output from the insecure one
17:35:45 <newsham> ahh.. so you want to solve the original seed
17:36:19 <ais523> and the way to do that is to find out the insecure RNG's current seed
17:36:29 <newsham> and they're using a 64-bit lcrng? thats pretty unusual.
17:36:32 <ais523> then run it backwards, trying progressive seeds as the secure RNG seed until one of them works
17:36:36 <ais523> and yes, it is pretty unusual
17:42:47 <newsham> have you looked at all pairs R2 > R1 * m + n for some small space to try to build an intuition?
17:43:05 <newsham> does it make bands in the state space?
17:43:52 <newsham> would be neat if you could quantify the bands and at each step cut your search space in half
17:44:25 <ais523> I've looked at some things like that
17:44:45 <ais523> the problem is that you get a bunch of repeating patterns, but the actual pattern you get depends on the size of R1
17:45:13 <ais523> obviously, for low R1, R2 is basically always higher
17:45:47 <newsham> maybe then if you can find large sequences of > > > > > it would isolate a small value of R
17:45:47 -!- Phantom_Hoover has quit (Ping timeout: 250 seconds).
17:45:55 <newsham> then you can brute force that small value
17:46:43 <newsham> hmm that would only give you a few bits though.. probability of finding a very long seq of > > > > would be low
17:47:37 <ais523> newsham: the problem is, I don't have /all/ the > < representations
17:47:46 <ais523> if the values are similar, I'm unlikely to be able to tell which one's larger
17:47:57 <ais523> so long > > > > > sequences, even though they often exist, are unlikely to be in the source data
17:48:02 <ais523> sometimes I just get a ?
17:48:52 -!- Phantom_Hoover has joined.
18:02:41 -!- callforjudgement has joined.
18:03:42 -!- shikhin has joined.
18:03:48 -!- Phantom_Hoover has quit (Ping timeout: 256 seconds).
18:04:03 <J_Arcane> Oh dear. I've reached the dreaded question: 0-indexed or 1-indexed.
18:06:23 -!- ais523 has quit (Ping timeout: 264 seconds).
18:06:32 -!- callforjudgement has changed nick to ais523.
18:10:06 <J_Arcane> Well, Djikstra wasn'texactly a BASIC fan ...
18:12:28 -!- tlewkow_ has joined.
18:13:55 <Bicyclidine> so ubuntu has a setting where trying to upgrade your release doesn't actually check for new releases, and you're not informed of this. awesome.
18:13:58 <newsham> maybe you can find some biases in the bits or R1 when R2 > R1 ?
18:15:35 <J_Arcane> https://www.cs.utexas.edu/users/EWD/transcriptions/EWD08xx/EWD831.html
18:17:54 <Bicyclidine> it's incredibly, incredibly anal, but pretty solid.
18:19:49 -!- sebbu2 has joined.
18:20:15 -!- nyuszika7h has quit (Read error: Connection reset by peer).
18:20:26 -!- sebbu2 has quit (Changing host).
18:20:26 -!- sebbu2 has joined.
18:21:32 -!- sebbu has quit (Ping timeout: 264 seconds).
18:21:45 -!- nyuszika7h_ has joined.
18:23:37 -!- nyuszika7h_ has changed nick to nyuszika7h.
18:23:56 <newsham> ais523: looks like there might be some bias in one of the high bits: http://www.thenewsh.com/~newsham/x/machine/bias.py
18:24:24 <ais523> newsham: when next(x) > x, it's likely that x is quite small
18:24:50 <ais523> my problem's more that I wonder if there's some mathematical way to get exact results
18:24:54 <ais523> by bruteforcing particular bytes
18:25:26 <elliott> you can get exact results with enough statistics
18:37:33 <J_Arcane> Hmm. Should calling out of index be an error or false? ie, calling (index 5 lst) where lst is only 3 entries long, etc.
18:39:43 <newsham> if you find runs of >>>>> it greatly increases the chance that the original high bit was a zero
18:40:19 <newsham> you could use that to prune your search space
18:40:40 -!- tlewkow_ has quit (Remote host closed the connection).
18:42:04 <newsham> for a 16-bit lcrng, there are no runs of 5 >'s starting with the high bit set.
18:42:48 <newsham> sorry, i should say "for the 16-bit lcrng i'm playing with" :)
18:47:53 <ais523> right, long runs would be unlikely
19:06:41 -!- S1 has quit (Quit: S1).
19:20:00 -!- Sprocklem has joined.
19:25:13 <mroman> !blsq 0 0 2047rn1024.+f:
19:25:13 <blsqbot> | {{5 1358} {4 1705} {4 618} {4 66} {3 1985} {3 1921} {3 1855} {3 1816} {3 1811} {3 1651} {3 1462} {3 1419} {3 1388} {3 1308} {3 1277} {3 1184} {3 1120} {3 1052} {3 1046} {3 1003} {3 865} {3 856} {3 787} {3 630} {3 322} {3 221} {3 205} {3 54} {3 6} {2 2041}
19:25:40 <Bicyclidine> in the year 1052 columbus sailed the ocean blue
19:25:46 <mroman> !blsq 0 0 2047rn1024.+2047rz//
19:25:47 <blsqbot> | ERROR: Unknown command: (//)!
19:25:47 <blsqbot> | {0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 7
19:25:50 <mroman> !blsq 0 0 2047rn1024.+2047rz\\
19:25:50 <blsqbot> | Ain't nobody got time fo' dat!
19:25:53 <mroman> !blsq 0 0 2047rn1024.+2047rz\\
19:25:54 <blsqbot> | Ain't nobody got time fo' dat!
19:26:07 <mroman> !blsq 0 0 1023rn1024.+1023rz\\
19:26:08 <blsqbot> | {733 22 135 636 444 750 382 614 438 779 961 434 156 509 779 302 168 701 763 623 856 982 856 335 601 816 205 822 251 322 156 225 516 618 949 1004 535 522 678 446 623 835 6 636 835 564 630 38 800 133 900 438 160 428 822 761 20 64 151 831 435 1017 247 520 9 4
19:26:12 <mroman> !blsq 0 0 1023rn1024.+1023rz\\L[
19:30:15 -!- ais523 has quit (Read error: Connection reset by peer).
19:30:24 -!- ais523 has joined.
19:30:24 -!- ais523 has quit (Changing host).
19:30:24 -!- ais523 has joined.
19:32:10 -!- callforjudgement has joined.
19:36:30 -!- ais523 has quit (Ping timeout: 265 seconds).
19:47:50 -!- callforjudgement has quit (Read error: Connection reset by peer).
19:48:06 -!- callforjudgement has joined.
19:49:55 -!- Patashu has joined.
19:54:57 -!- Sprocklem has quit (Ping timeout: 245 seconds).
20:26:46 -!- Patashu has quit (Ping timeout: 265 seconds).
20:32:44 <HackEgo> [wiki] [[Talk:3var]] http://esolangs.org/w/index.php?diff=41115&oldid=39178 * Olls * (+197) /* Interpreter? */
20:35:42 -!- Bicyclidine has quit (Quit: restart).
20:38:43 -!- callforjudgement has changed nick to ais523.
20:38:47 -!- Bicyclidine has joined.
20:40:18 -!- Bicyclidine has quit (Client Quit).
20:43:34 -!- ais523 has quit (Read error: Connection reset by peer).
20:43:35 -!- dts has quit (Ping timeout: 244 seconds).
20:43:45 -!- ais523 has joined.
20:44:50 -!- Bicyclidine has joined.
20:45:06 -!- Bicyclidine has quit (Client Quit).
20:45:13 -!- Bicyclidine has joined.
20:52:09 -!- copumpkin has quit (Quit: My MacBook Pro has gone to sleep. ZZZzzz…).
20:53:37 -!- Bicyclidine has quit (Quit: trying yet more else).
20:53:59 -!- AnotherTest has joined.
20:54:45 -!- Bicyclidine has joined.
21:09:47 -!- S1 has joined.
21:11:37 -!- TodPunk has quit (Ping timeout: 245 seconds).
21:13:48 -!- TodPunk has joined.
21:35:19 -!- ais523 has quit (Read error: Connection reset by peer).
21:35:22 -!- callforjudgement has joined.
21:35:24 -!- callforjudgement has quit (Changing host).
21:35:24 -!- callforjudgement has joined.
21:40:44 -!- callforjudgement has quit (Read error: Connection reset by peer).
21:40:52 -!- callforjudgement has joined.
21:53:52 -!- Dulnes has joined.
21:54:13 <Dulnes> Why do i keep disconnecting
21:54:50 <zzo38> Would you like it if Professor Oak can become a playable character in Smash Brothers game?
21:55:05 <zzo38> Dulnes: I don't know? Possibly a problem with internet sometimes I have a problem too a bit
21:55:08 <myname> depends on what he could do
21:55:23 <Dulnes> Throw pokeballs like a gattling gun
21:55:35 <Taneb> I could see him as an assist trophy
21:55:47 <zzo38> callforjudgement: Yes I know and I managed to vs him too
21:56:12 -!- Phantom_Hoover has joined.
21:56:19 <Dulnes> I think you can use the program
21:56:54 <zzo38> myname: I would think, you can sometimes throw a pokeball after a few seconds and that in addition, if you are using a separate display (such as Wii U gamepad) that it will tell you what pokemon is found in each pokeball on the screen too, and everyone else does not see that information until the pokeball is opened!
21:57:25 <zzo38> Dulnes: What program?
21:57:52 <Dulnes> You could use the battle data
21:58:10 <Dulnes> If you forcably seperated it from the cartridge
21:58:13 <zzo38> O, OK that's what you meant.
21:58:59 <zzo38> My brother has the newest Smash Brothers game on Wii U and Nintendo 3DS
21:59:07 <zzo38> So, I played this game too
21:59:32 <Dulnes> Because new legend of zelda game 2015
21:59:33 <Taneb> I thought that didn't come out until Friday?
21:59:47 <Dulnes> Came out last week i think
21:59:55 <Taneb> Ah, maybe Europe is later
21:59:57 <zzo38> One thing you still cannot do though even in the newest one is to set the self-destruct to -1.5 instead of only 0 and -1 and -2
22:00:02 <Taneb> I'm gonna get the 3DS version soon
22:00:16 <Taneb> Dulnes, translation issues
22:00:22 <zzo38> Both me and my brother want to be able to set self-destruct points to -1.5
22:00:30 <Taneb> zzo38, can you adjust all the other scores to make it effectively the same
22:00:31 <Dulnes> Since it goes Japan - China - Europe - Americas
22:00:53 <Dulnes> Or < that way on a map
22:01:05 <zzo38> Taneb: You could manually calculate the score I suppose if you like to; it doesn't allow you to otherwise adjust it.
22:01:16 <zzo38> However there is no function to disable sudden-death mode.
22:01:46 <Taneb> Dulnes, maps aren't really relevant to this sort of thing any more, not for a few decades
22:01:55 <pikhq_> I doubt it's translation issues, at least entirely...
22:01:56 <zzo38> Again a function both of us wanted to have
22:02:16 <pikhq_> I mean, if so why make Australia take until the 29th?
22:02:21 <pikhq_> Or Japan until the 6th?
22:02:22 <Taneb> pikhq_, that's a point
22:02:30 <pikhq_> (yes, Japan is the *last* country getting it)
22:02:54 -!- nortti has changed nick to lawspeaker.
22:03:01 -!- lawspeaker has changed nick to nortti.
22:03:18 <Dulnes> Ive been wondering why is Australia so hot
22:03:28 <Dulnes> If its near the south pole
22:03:44 <Taneb> Dulnes, it's closer to the equator than the UK is
22:03:51 <Taneb> By quite some measure
22:04:01 <elliott> try looking at a globe rather than a map :p
22:04:20 <pikhq_> IIRC Northern Australia is literally in the tropics.
22:04:33 <elliott> remember to by a new cpu core while you're at it
22:04:38 <zzo38> One thing they could have done but don't is to implement custom omega stages, which involves selecting music, graphics, and whether or not you can go underneath the platform. It isn't particularly important though.
22:04:39 <Bicyclidine> to be fair, nobody lives in northern australia
22:04:51 <Taneb> It's about as close to the equator as India is
22:05:35 <pikhq_> Ah, yep. And nearly all of the country is in the subtropics.
22:06:03 <Dulnes> You and your giant galloping wolf spiders
22:06:35 <Dulnes> Doesnt the worlds deadliest snake live there?
22:06:38 <pikhq_> So yeah, it's pretty much entirely in the latitudes where simple sun exposure is likely to make things at least moderately hot.
22:07:31 <pikhq_> Also, almost all of the continent is a desert, so there's that too.
22:07:59 <Dulnes> i believe the emu's won
22:08:43 <pikhq_> Understandably the population's concentrated around the not-desert bits.
22:08:48 <zzo38> There also is not the mode to use the time limit corresponding to the stage which is selected.
22:09:10 <zzo38> (This would only be applicable for some stages though.)
22:09:56 <Dulnes> pikhq_: why did they have a war?
22:11:14 <Dulnes> Anyways i actually do have food in my house and this time its not coffee based
22:11:55 <elliott> `addquote <Dulnes> Anyways i actually do have food in my house and this time its not coffee based
22:11:57 <HackEgo> 1223) <Dulnes> Anyways i actually do have food in my house and this time its not coffee based
22:20:14 -!- AnotherTest has quit (Remote host closed the connection).
22:20:26 <Sgeo> If I start eating a lot of sugary snacks, but I still eat the other foods I was eating before (so it's not taking the place of anything else), and I'm not close to being overweight, is that still bad?
22:21:53 <Phantom_Hoover> i still don't understand why you come here for this sort of advice
22:22:54 <Sgeo> This channel was just talking about food >.>
22:23:13 -!- tlewkow has joined.
22:24:48 <myname> food is the thing i like to eat the most
22:26:54 <Taneb> myname, I prefer to eat business cards
22:27:00 <Taneb> But I find them very filling
22:28:53 <Dulnes> We werent talking about food
22:29:06 <Dulnes> We were talking about how everything is edible
22:30:46 <Sgeo> Apparently the notion of getting diabetes type 2 from too much sugar is a myth?
22:33:59 <elliott> do I dare ask what you're planning
22:35:19 <Sgeo> I bought a box of boxes of fruit by the foot recently
22:35:30 <Sgeo> They taste so good. I've been eating about a box a day
22:53:16 -!- oerjan has joined.
22:54:34 <zzo38> Do you know if any SQL-based RPG engines exist, or do I have to write one?
22:55:05 <oerjan> i sense elliott didn't like my dark humor
22:59:41 <oerjan> hm i was going to say that both solutions to http://golf.shinh.org/p.rb?Dominosa+Small were too large to be solving the problem properly, but i think they're actually too large to be using compression.
23:02:40 -!- shikhout has joined.
23:06:31 -!- shikhin has quit (Ping timeout: 272 seconds).
23:07:32 <Sgeo> What did I miss about an incident that callforjudgement missed?
23:08:40 -!- ski has quit (Ping timeout: 256 seconds).
23:09:09 <callforjudgement> selecting randomly between three fixed strings isn't hard in most languages
23:09:13 <oerjan> i know, i'm writing a trivial one
23:15:26 <oerjan> alas i'm too lazy to write a non-cheating solution
23:15:30 <Phantom_Hoover> <Sgeo> Apparently the notion of getting diabetes type 2 from too much sugar is a myth?
23:15:40 <Phantom_Hoover> the common notion is getting it from obesity, not sugar per se
23:17:04 <oerjan> wtf are the neighbors making noise after midnight
23:17:46 <oerjan> (just loud talking, but still)
23:17:58 <Taneb> Are they students?
23:18:17 <Taneb> And if you're awake after midnight, why shouldn't they be?
23:18:25 <Taneb> But if they're being very noisy, :(
23:18:29 <oerjan> i'm _awake_. i'm not noisy.
23:18:39 -!- AndoDaan has joined.
23:21:16 <Sgeo> Is it possible to become obese without noticing?
23:21:31 <Sgeo> Looking thin but being medically obese?
23:21:33 <oerjan> if anyone can manage it...
23:22:05 -!- zzo38 has quit (Remote host closed the connection).
23:23:12 <FreeFull> Sgeo: Maybe if you invent a form of fat that contains neutronium
23:27:56 <elliott> "obesity" isn't a monolithic thing.
23:30:40 <oerjan> sure it is, if you look like a monolith you're definitely obese
23:35:45 -!- scarf has joined.
23:36:07 -!- callforjudgement has quit (Read error: Connection reset by peer).
23:37:39 <oerjan> there appears to be a general _custom_ in norway about night silence after 11 pm before ordinary weekdays. but there's no actual written law that you can point to.
23:38:42 <oerjan> and our house rules have no clock times in them
23:40:17 <oerjan> my long parenthesized comment on anagolf weirds out the layout :(
23:44:24 -!- tlewkow has quit (Remote host closed the connection).
23:48:28 -!- tlewkow has joined.
23:55:21 -!- ski has joined.
23:55:39 -!- vanila has joined.