00:01:24 <oerjan> very unusual unobfuscated c?
00:02:20 <int-e> elliott: is it? it prevents evil program doing file operations on the final output file; all they get access to is a pipe.
00:02:32 <elliott> int-e: there is no output file
00:02:52 <elliott> I think Python's subprocess module is adding the | cat
00:02:58 <ion> | cat | cat for two-factor security
00:03:01 <int-e> if it's a pipe already, then indeed...
00:03:08 <elliott> since you *can* point stdout to a file object there
00:03:17 <vanila> rot13 | rot13 | could be used instead of ca
00:03:17 <elliott> so I guess they just do it for everything because it doesn't hurt
00:03:58 <nooga> btw, are there some new, hot esolangs that aren't just some funky syntax sprinkled on top of boring execution model?
00:03:58 <int-e> `` ls -l /proc/self/fd
00:04:01 <HackEgo> total 0 \ lr-x------ 1 5000 878264 64 Nov 22 00:03 0 -> /tty1 \ l-wx------ 1 5000 878264 64 Nov 22 00:03 1 -> pipe:[248] \ l-wx------ 1 5000 878264 64 Nov 22 00:03 2 -> /tty1 \ lr-x------ 1 5000 878264 64 Nov 22 00:03 3 -> /console \ l-wx------ 1 5000 878264 64 Nov 22 00:03 4 -> /console \ lr-x------ 1 5000 878264 64 Nov 22 00:03 5 -> /tty1 \ l-wx-
00:04:23 <vanila> nooga, there is an intereting one
00:05:00 <vanila> http://esolangs.org/wiki/My_Unreliable_Past
00:05:06 <elliott> nooga: probably yeah in the last ten billion years you haven't been here :p
00:05:09 <int-e> `` ls -l /proc/self/fd | sed 's/.*:...//'
00:05:11 <HackEgo> total 0 \ 0 -> /tty1 \ 7] \ 2 -> /tty1 \ 3 -> /console \ 4 -> /console \ 5 -> /tty1 \ 6 -> /tty1 \ 7 -> /tty1 \ 8 -> /proc/293/fd
00:05:29 <vanila> http://esolangs.org/w/index.php?title=Special:RecentChanges&limit=500&days=30
00:06:00 <vanila> Casino viagra program language[edit]
00:06:00 <vanila> Make a casino viagra program language. Sometimes this word is same as spam word, so you have to make sure it really is authentic. --Zzo38 04:38, 16 July 2008 (UTC)
00:08:23 <nooga> elliott: that's why I'm asking. I was hoping that someone would point me to the interesting bits :D
00:08:58 <oerjan> <shachaf> oerjan: perhaps some jokes have gone over my head here <-- not that i recall. as a norwegian, i tend to think of him as mostly british, anyway.
00:11:55 <oerjan> nooga: were you around for Sgeo's braintrust
00:13:45 <vanila> casino viagra should be based on spam keywords and maybe email, an maybe bayesian spam filters
00:14:07 <vanila> you could have an email that tries to filter spam, and only code that gets past it is executed
00:14:19 <oerjan> design feature #1: programs cannot be reliably sent via email
00:14:53 <int-e> so make them newline sensitive
00:15:04 <vanila> OR: Maybe ONNLY Programs whihc get caught by spam filter are executed
00:15:06 -!- tlewkow_ has joined.
00:15:06 <int-e> by which I mean, make \r\n vs \n significant
00:15:10 <vanila> this forces progammers to write code that looks like spam
00:15:13 <oerjan> int-e: no, i mean because they get caught in spam filters
00:15:58 <nooga> oerjan: reading about braintrust now
00:16:46 <oerjan> vanila: hm maybe the semantics of a program should be based on the spam rules it triggers
00:17:25 <int-e> oooh, are there any turing complete spam filters out there...
00:17:44 <int-e> (I hope not, since they're supposed to terminate, for one, but who can say for sure?)
00:18:08 <HackEgo> Your evil overlord oerjan is a lazy expert in future computation. Also an antediluvian Norwegian who hates Roald Dahl.
00:18:34 <shachaf> i thought that was meant to be some kind of joke
00:18:50 <oerjan> shachaf: roald dahl was a creepy guy who mistreated his family but i don't actually hate him hth
00:19:34 <ion> Did his parents make a typo when choosing his name?
00:19:47 <int-e> sed -i s/hates/dislikes/ wisdom/oerjan
00:20:08 <oerjan> roald is the common norwegian spelling, as in roald amundsen
00:20:39 <oerjan> dahl is also a common spelling, although historically slightly pretentious (it's from no:dal = en:valley)
00:21:30 <oerjan> Melvar: are you neandering around the spelling here
00:22:17 <Melvar> The Neander Valley isn’t even far from here.
00:22:52 <oerjan> _and_ it's spelling (or at least that of its derived term) fluctuates between tal and thal
00:24:03 <oerjan> "In 1901 an orthographic reform in Germany changed the spelling of Thal (valley) to Tal. The scientific names like Homo neanderthalensis or Homo sapiens neanderthalensis for Neanderthal man are not affected by this change, because the laws of taxonomy retain the original spelling at the time of naming. Neanderthal station nearby still carries the name Neanderthal, because the nearby Neanderthal Museum continues to have the old spelling."
00:24:40 <int-e> code golf insight of the day: while 16 needs two characters in dc, 160 can be expressed as FA...
00:28:05 <int-e> oerjan: do you want me to say something about Leapfrogging?
00:28:30 <oerjan> it would be such a huge spoiler
00:28:39 <vanila> vegas online casino viagra online online gambling legal sites best online blackjack sites usa Playtech online casinos online casino
00:29:18 <oerjan> Dearest vanila, I write to you in this time of viagra
00:29:56 <int-e> vanila: honestly? too many casinos, too little pharmacy
00:30:08 <vanila> I'll fix that in v2 int-e
00:30:49 <int-e> And too little 401 scam, though those are hard to identify by subject.
00:31:05 <int-e> "business proposal" perhaps
00:31:35 <int-e> "work from home" (money laundering)
00:31:53 <int-e> it's funny, I don't seem to get any gambling spam at all.
00:32:29 <vanila> i once emailed a company for which some spammer had posted a comment on my site
00:32:36 <int-e> vanila: oh and how could you forget "penis enlargement"
00:32:39 <vanila> and he replied saying he didn't do it
00:33:21 <ion> Number literals could be represented in the form of “I am therefore seeking for a reliable person that will play the human role as the next of kin to this fund which is in the amount of £32,000,000.00 (Thirty Two Million Pounds Sterling).”
00:33:31 <int-e> haven't seen one of those in the while, perhaps the target audience outgrew the need.
00:36:01 <ion> http://mikko.tuomela.net/spam/
00:36:58 <oerjan> we are writhing indeed
00:37:07 <ion> Ah, don’t forget the caps. http://mikko.tuomela.net/spam/am_so_sick.txt
00:38:23 <int-e> So for small numbers, you could say "Earn up to $3,000 a week working two hours a day from home!"
00:39:22 <int-e> while for really large numbers you can go the chain letter route "Please send this letter to 3 friends. Anne didn't send this letter to her friends and her cat was hit by a bolt of lightning."
00:39:38 <int-e> (needs more exclamation marks!!!!!1)
00:40:06 <oerjan> conspiracy theory: someone already made this language. what we think are spam advertisements are really secret botnet commands.
00:40:47 <int-e> which clearly encodes 454462363333321429 (number of letters in the words of the second sentence) ;-)
00:41:52 <nooga> well there was this idea to encode emails using NSA's alledged kewords
00:45:41 <lambdabot> ENVA 220020Z 10004KT CAVOK M06/M07 Q1027 RMK WIND 670FT VRB02KT
00:49:28 <boily> chilly is perfect weather for lots of coffee, milk tea and hot chocolate.
00:52:49 <lambdabot> EFHK 220020Z VRB02KT 9000 SCT004 OVC036 M02/M03 Q1023 TEMPO 6000 -SN BKN004
00:53:05 -!- tlewkow_ has quit (Remote host closed the connection).
00:53:08 <fizzie> Less chilly, and the snow stopped too.
00:53:13 <lambdabot> CYUL 220000Z 24007KT 15SM FEW120 M08/M15 A3034 RMK AC1 AC TR SLP278
00:53:46 <boily> zlsa: what are your approximate geographic coördinates?
00:58:50 <ion> geog̈raphic coördinates
01:05:36 <lambdabot> KMCE 220053Z AUTO 12004KT 9SM SCT120 14/09 A3011 RMK AO2 SLP195 T01440094
01:06:30 <lambdabot> LOWI 220050Z AUTO VRB01KT 9999 FEW090 BKN110 04/03 Q1022
01:06:51 <oerjan> <elliott> is the subset of BF complete given by +<>[] and you can only write to a cell once TC <-- i think cyclic tag should be easy to do that way.
01:07:12 <oerjan> or a tag system in general
01:08:08 -!- tlewkow_ has joined.
01:08:15 -!- nooga has quit (Ping timeout: 250 seconds).
01:10:45 <oerjan> oh, also a minsky machine, i think
01:10:53 <oerjan> just have parallel tracks
01:11:11 <oerjan> saving each register in unary
01:11:28 -!- tlewkow_ has quit (Remote host closed the connection).
01:12:00 <oerjan> as 001111111101010101000000...
01:12:46 <int-e> in fact you can use a common base for all registers, if you want
01:12:47 <oerjan> increment by turning 00 to 01 on the right, decrement by turning 01 into 11
01:13:03 <int-e> (decrement a register by incrementing the base and all other registers)
01:13:39 <oerjan> um i'm not sure that helps?
01:14:11 <int-e> it may improve storage efficiency by a small constant factor. (less than 2)
01:14:50 <int-e> the tag system idea requires some unbalanced scanning loop like [<<<[>]<], doesn't it?
01:15:24 <oerjan> you need something unbalanced anyway
01:15:34 <oerjan> to get between beginning and end of each register
01:15:47 <int-e> yeah, I'm happy with first-order unbalanced loops, but this is second order ;)
01:16:18 <oerjan> i don't see why it should, though.
01:16:49 <elliott> oerjan: what about BF sans <?
01:17:11 <oerjan> elliott: um i think that is sub-FSA
01:17:48 <Sgeo> If most languages are in the IO monad, is Rust in the ST monad?
01:17:53 <oerjan> > is equivalent to [-] in that case
01:17:54 <lambdabot> <hint>:1:20: parse error on input ‘]’
01:19:59 <int-e> oerjan: here's what I thought. I have an infinite string, on tape, with an initial consumed part. A possible layout is 1c0x, where x is the symbol, 1 marks a part of the string, and c marks the consumed part. Now to skip back to just in front of the consumed part, I need to test for the 'c' inside the scanning loop (since it tests the wrong way for terminating the loop). Hence [<<<[>>]<], the [>>] being executed on the 'c'.
01:20:14 <oerjan> int-e: i don't think you need an unbalanced loop any more than for minsky, you just essentially expand a single minsky register 00 01 11 with also having tag representations
01:20:39 <int-e> I'm making things hard for myself
01:20:57 <int-e> I can skip to the front, and then scan forward.
01:21:11 <oerjan> so 00111111....01tag01tag...000000...
01:21:34 <oerjan> which you'll have to do for the minsky as well
01:21:59 <int-e> so instead of [<<<[>>]<], [<<<<]>[>>>>] (and then I don't actually need the extra 0 in the layout anymore)
01:22:00 <oerjan> i think that's unavoidable in a sense, since you cannot erase the bits you're scanning along
01:22:56 <vanila> is there a fuckopedia?
01:23:04 <vanila> some kind of categorization of all branfuck derivatives
01:24:08 <int-e> "fuckopedia.org is worth $1,355 - Worth Of Web Calculator"
01:24:39 <int-e> vanila: I *think* if there was a fuckopedia, it would have somewhat different contents.
01:26:36 <oerjan> vanila: http://esolangs.org/wiki/Brainfuck_extensions is a bit like that
01:28:00 <oerjan> although somewhat more limited in scope, i think
01:28:47 <oerjan> and of course http://esolangs.org/wiki/Category:Brainfuck_derivatives is an uncategorized category
01:29:04 <vanila> it could be fun to make an esolang blog
01:29:12 <vanila> post things lke 'roundup of brainfuck languages'
01:29:16 <vanila> or 'intro to unlambda'
01:29:44 <oerjan> vanila: mark chu-carroll made some esolang blog posts back in the day
01:30:02 <oerjan> i think unlambda may have been one
01:30:28 <oerjan> well, underload was, maybe both.
01:32:44 <HackEgo> [wiki] [[Bitwise Cyclic Tag]] http://esolangs.org/w/index.php?diff=41104&oldid=37454 * 212.95.7.136 * (+65) /* Self BCT */ link
01:35:30 <int-e> oh, should that https be http, hmmmmmm
01:35:31 <int-e> I'll let oerjan decide.
01:40:40 <elliott> int-e: you can do [[wikipedia:Foo|...]]
01:44:50 <oerjan> i'm sure i've mentioned i dislike that those links look internal
01:45:01 <elliott> oerjan: didn't I add an icon for them?
01:45:17 <int-e> elliott: maybe but there isn't one
01:45:54 <int-e> (as can be seen at the bottom of https://esolangs.org/wiki/Bitwise_Cyclic_Tag)
01:46:16 <int-e> (I would prefer the https link btw but the resulting icon is irritating.)
01:47:45 <oerjan> hm clearly it should use https in the link if you're viewing esolang through it
01:48:00 <int-e> (I wonder why would Mediawiki would treat Wikipedia special in any way ;-) )
01:48:42 <vanila> the BCT patterns are very beautiful
01:48:46 <vanila> http://esolangs.org/w/index.php?diff=41104&oldid=37454#BCT_programs
01:49:13 <elliott> the interwiki mechanism is general
01:49:19 <elliott> there are lots of things you can put there
01:49:26 <elliott> fizzie could add more, even
01:50:06 <elliott> this is btw something that predates wikipedia -- interwiki links come from wikis that look like c2 :p
01:52:09 <int-e> hmm. http://www.cnbc.com/id/102207790 ... so rather than having Google track me so it can display ads, now I'm letting it track me so that it knows not to show me any? ...
01:52:24 <int-e> (And pay for it, too.)
01:53:52 <int-e> (Not to mention that this is an extortion scheme. "Pretty websites you're viewing there, would be a shame if they were disfigured by annouying banners...")
02:01:16 -!- Patashu has quit (*.net *.split).
02:01:16 -!- blsqbot has quit (*.net *.split).
02:01:16 -!- mihow has quit (*.net *.split).
02:01:16 -!- Dulnes has quit (*.net *.split).
02:01:17 -!- yiyus has quit (*.net *.split).
02:01:17 -!- olsner has quit (*.net *.split).
02:01:18 -!- Bike has quit (*.net *.split).
02:01:20 -!- Gregor has quit (*.net *.split).
02:01:24 -!- variable has quit (*.net *.split).
02:01:26 -!- bb010g has quit (*.net *.split).
02:01:26 -!- newsham has quit (*.net *.split).
02:01:27 -!- zlsa has quit (*.net *.split).
02:01:28 -!- zzo38 has quit (*.net *.split).
02:01:28 -!- idris-bot has quit (*.net *.split).
02:04:04 -!- sebbu has quit (Ping timeout: 264 seconds).
02:06:02 -!- sebbu has joined.
02:10:55 -!- copumpkin has joined.
02:12:58 -!- bb010g has joined.
02:31:35 <oerjan> come on, dmm, you cannot start an annotation with "It's a little known fact that" and _not_ include a horrible pun...
02:33:14 -!- newsham has joined.
02:33:14 -!- variable has joined.
02:33:14 -!- zlsa has joined.
02:33:14 -!- zzo38 has joined.
02:33:14 -!- idris-bot has joined.
02:33:23 -!- zlsa has quit (*.net *.split).
02:33:23 -!- zzo38 has quit (*.net *.split).
02:33:23 -!- idris-bot has quit (*.net *.split).
02:33:27 <oerjan> http://www.irregularwebcomic.net/ can you find one
02:33:30 -!- Patashu has joined.
02:33:30 -!- blsqbot has joined.
02:33:30 -!- mihow has joined.
02:33:30 -!- Dulnes has joined.
02:33:30 -!- yiyus has joined.
02:33:30 -!- olsner has joined.
02:33:30 -!- Bike has joined.
02:33:30 -!- Gregor has joined.
02:34:45 -!- zlsa has joined.
02:34:45 -!- zlsa has quit (Changing host).
02:34:45 -!- zlsa has joined.
02:35:19 <shachaf> oerjan: am i missing the horrible pun
02:35:50 <oerjan> by all rights there _should_ be one there, is what i'm saying
02:36:07 <oerjan> (oh and in the rerun comment, not the comic itself hth hth)
02:36:22 <shachaf> do you still have that hth script
02:37:08 <shachaf> oerjan: what's a convenient acronym for hope that helps
02:40:40 -!- Patashu has quit (*.net *.split).
02:40:40 -!- blsqbot has quit (*.net *.split).
02:40:41 -!- mihow has quit (*.net *.split).
02:40:41 -!- Dulnes has quit (*.net *.split).
02:40:42 -!- yiyus has quit (*.net *.split).
02:40:42 -!- olsner has quit (*.net *.split).
02:40:43 -!- Bike has quit (*.net *.split).
02:40:43 -!- Gregor has quit (*.net *.split).
02:41:20 <oerjan> i think they're server may be flaky
02:42:11 -!- zzo38 has joined.
02:42:11 -!- idris-bot has joined.
02:46:58 -!- Patashu has joined.
02:46:58 -!- blsqbot has joined.
02:46:58 -!- mihow has joined.
02:46:58 -!- Dulnes has joined.
02:46:58 -!- yiyus has joined.
02:46:58 -!- olsner has joined.
02:46:58 -!- Bike has joined.
02:46:58 -!- Gregor has joined.
03:10:33 -!- Sprocklem has joined.
03:12:18 -!- boily has quit (Quit: SEEKING CHICKEN).
03:26:49 <int-e> oerjan: this is not the pun you're looking for *waves*, hth.
03:28:35 <int-e> (In this particular case I must assume that the pun died of boredom due to a clerical error.)
03:30:13 <int-e> Which raises an interesting question ... do puns have souls?
03:30:32 <oerjan> int-e: if i am interpreting recent ghc weekly news blog posts correctly, your recent comment that ghc lets gmp use ghc's allocation mechanism soon won't be true any more. so hopefully that means that is _not_ the reason for its good performance.
03:31:25 <int-e> oerjan: Ah, you might read it that way but that's not what integer-gmp2 is about.
03:31:45 <oerjan> well they won't be using the heap any longer, it said
03:32:02 <int-e> The point of integer-gmp2 is to use a lower level interface (mpn) that does not do any allocation at all, and still do the allocation on the ghc side. Hmm.
03:32:13 <oerjan> "This implementation also fixes a long standing pain point where GHC would hook GMP allocations to exist on the GHC heap. Now GMP is just called to like any FFI library."
03:33:06 <oerjan> that's what made me think that
03:34:16 <oerjan> also there was some earlier stuff about how the previous method conflicted with using other C libraries that used gmp
03:35:05 <int-e> Yes, unfortunately gmp uses global variables for the allocation functions.
03:35:35 <int-e> And the C libraries don't deal well with the fact that the allocated chunks are bytearrays on the heap that will be reclaimed during the next GC cycle ;)
03:35:41 <oerjan> or maybe i read that in the HCAR report
03:36:56 <oerjan> hm, do you mean that mpn is a lower-level gmp interface that _doesn't_ use those global allocation variables?
03:38:11 <int-e> AFAIU, all allocation goes through this function: newBigNat# :: GmpSize# -> S s (MutBigNat s); newBigNat# limbs# s = case newByteArray# (limbs# `uncheckedIShiftL#` GMP_LIMB_SHIFT#) s of (# s', mba# #) -> (# s', MBN# mba# #)
03:38:43 <int-e> and it's using the mpn_* functions of libgmp rather than the mpz_* ones; the signs are now handled on the Haskell side.
03:41:17 <oerjan> oh that might be connected with the recent suggestion of adding Natural
03:48:15 <oerjan> oh that was already added to HEAD
03:55:02 <int-e> oerjan: so I expect there will be a performance hit for multiplying large numbers, because that needs temporary allocations, and if those exceed 64kb, they'll be done on the C heap.
03:55:19 <int-e> but the impact needs to be measured.
03:56:41 -!- FreeFull has quit (Ping timeout: 256 seconds).
03:56:53 <int-e> (And by "large" I mean tens of thousands of digits and more.)
03:57:19 -!- FreeFull has joined.
04:10:45 <int-e> I'll try the 10^8th fibonacci number ... will take a while to compile ghc though.
04:11:24 <int-e> (and I'll compare 7.8.3 to current head rather than recompiling head twice)
04:16:15 <int-e> this will be my test, http://sprunge.us/GEMG?hs
04:18:00 <int-e> I suspect I won't measure much of a difference actually
04:18:10 <int-e> because the allocation cost is hidden behind actual computation.
04:26:50 <oerjan> Dulnes: it's essentially calculating fibonacci numbers with matrix multiplication, slightly optimized to avoid recalculating things (and the P is to avoid haskell laziness)
04:27:31 <oerjan> although the purpose is to test some new changes in ghc's big integer implementation
04:27:40 <Dulnes> thank you Jesus for explaining that i didnt want to look for the conversation
04:29:31 <int-e> Dulnes: I think you misspelled 'oerjan' there.
04:31:27 <oerjan> i would like not to be identified with jesus as i don't really like the idea of solipsistic pantheism hth
04:36:06 <shachaf> oerjan: what do you actually type into your computer to send a line of irc text containing "hth"
04:36:54 <oerjan> (it only strips the last one)
04:37:10 <shachaf> i thought it removed the whole line
04:37:10 <int-e> oerjan means '"hth" hth', hth.
04:37:32 <shachaf> i thought oerjan meant «"hth hth" hth»
04:38:33 <oerjan> shachaf: it would be rather awkward not to be able to write "eighth"
04:38:47 <oerjan> actually _can_ i write eighth
04:39:00 <shachaf> to be fair eighth is a pretty awkward word in the first place
04:39:06 <int-e> oerjan: interesting. I get a measurable difference - ghc head is faster. So I guess I'll try ghc head with integer-gmp, too...
04:39:06 <shachaf> shouldn't it be more like eightth
04:39:26 <int-e> (not dramatic: 6 seconds vs. 5.7)
04:40:52 <int-e> but 5% is noticable
04:41:38 <shachaf> Why does that fibonacci matrix thing work?
04:42:26 <shachaf> I mean, I can follow the computation, but e.g. why should I have expected there to be such a matrix? What is it doing?
04:42:52 <oerjan> shachaf: [1 1 ; 1 0] (a,b) = (a+b,a), right?
04:42:54 <shachaf> OK, I guess it's sort of obvious now that I think about it.
04:43:08 <oerjan> (notation made up out of thin air)
04:43:33 <oerjan> you can make such a matrix for any linear recurrence
04:44:43 <shachaf> And it has an eigenvalue phi, which I guess shouldn't be surprising.
04:44:58 <shachaf> What happens if you diagonalize it or something?
04:45:09 <oerjan> the phi^n + phi^(-n) thing comes from diagonalizing it
04:45:30 <int-e> For this particular code, I think of P a b as encoding the matrix [a,b;b,a+b]
04:46:31 <int-e> And it turns out that [a,b;b,a+b] * [c,d;d,c+d] produces another matrix of the same shape (not entirely by accident).
04:47:03 <oerjan> iirc there's an adjustment you need to perform if there are repeated eigenvalues
04:47:28 <shachaf> oerjan: I'm not familiar with that thing, but that makes sense.
04:47:43 <shachaf> int-e: I don't like "it turns out" answers. :-(
04:48:02 <int-e> (and of course [a,b;b,a+b] * [0,1;1,1] = [b,a+b;a+b,a+2b])
04:48:22 <shachaf> But anyway this makes some sense.
04:49:39 <shachaf> Oh, that code. I didn't see it before.
04:50:54 <oerjan> > let phi = (1+sqrt 5)/2 in [phi^n + phi^^(-n) | n <- [1..10]]
04:50:55 <lambdabot> [2.23606797749979,3.0,4.47213595499958,7.0,11.180339887498949,18.0,29.068883...
04:51:23 <oerjan> i think there might be a coefficient needed
04:52:14 <oerjan> > let phi = (1+sqrt 5)/2 in [(phi^n + phi^^(-n))/sqrt 5 | n <- [1..10]]
04:52:16 <lambdabot> [1.0,1.3416407864998738,2.0,3.1304951684997055,5.0,8.049844718999243,13.0,21...
04:52:29 <oerjan> > let phi = (1+sqrt 5)/2 in [(phi^n - phi^^(-n))/sqrt 5 | n <- [1..10]]
04:52:31 <lambdabot> [0.4472135954999579,1.0,1.7888543819998317,3.0,4.919349550499537,8.0,12.9691...
04:52:34 <shachaf> > let phi = (1+sqrt 5)/2 in [(phi^n - (-phi)^^(-n))/(sqrt 5) | n <- [1..10]]
04:52:35 <lambdabot> [1.0,1.0,2.0,3.0,5.0,8.0,13.0,21.0,34.0,54.99999999999999]
04:52:57 <shachaf> sounds vaguely reasonable, since the eigenvalues are phi and -1/phi
04:53:48 <int-e> I usually use (1+sqrt 5)/2 and (1-sqrt 5)/2, of course the latter is negative.
04:55:43 <oerjan> shachaf: you basically calculate [1,0] in the eigenvector basis and apply the right eigenvalue to each part
04:56:40 <shachaf> what are the eigenvectors?
04:57:50 * oerjan divides by the prime 53
04:58:45 <int-e> poor Grothendieck.
04:59:43 <oerjan> [1,phi] might be a better candidate
05:03:10 <shachaf> also what do left eigenvectors mean in general
05:03:23 <shachaf> in this case they're the same, of course, because the matrix is symmetric
05:04:22 <int-e> shachaf: actually it may be saner to think of P a b as a + b*x (mod x^2-x-1). It's easy to see (tm) that x^n = F_(n-1) + F_n x (mod x^2-x-1).
05:04:28 <oerjan> in general they live in the dual space, i think
05:05:15 <oerjan> so they're functionals composed with the matrix
05:05:43 <shachaf> that sounds more complicated than the other kind
05:06:34 <oerjan> they're functionals f such that f(Mx)=l(fx) for all x
05:06:39 <int-e> shachaf: but you get the fact that the P a b form a ring for free, rather than having it appear as a miracle.
05:07:29 <oerjan> with matrices, you have a basis, and therefore an identification of functionals with vectors
05:07:46 <oerjan> unlike with just linear transformations
05:08:13 <shachaf> oerjan: Is that really right?
05:08:27 <shachaf> You need to choose bases to go from linear maps to matrices and also to go from matrices to linear maps.
05:08:49 <oerjan> well yes, isn't that essentially what i said
05:09:25 <oerjan> what i mean is, when you are dealing with matrices and vectors in R^n, you always have an implied natural basis
05:09:50 <oerjan> so you can identify a functional f with the vector y such that (y,x) = f x for all x
05:10:09 <oerjan> where (,) is inner product
05:10:36 <oerjan> ok you need not a full basis, just enough to get the inner product
05:11:22 <oerjan> and then you get to the bra-kets of hilbert space physics
05:12:33 <oerjan> (which are essentially a visual tool for this correspondence)
05:14:04 <oerjan> but even without an inner product, you can think of functionals as having the adjoint linear transformation acting on them.
05:15:26 <oerjan> and then a "left eigenvector" is an eigenvector functional of the adjoint transformation
05:17:53 <oerjan> and this * is a contravariant functor.
05:20:31 <shachaf> I can think of a "right eigenvector with eigenvalue 1" as a fixed point of a map. Is there a similarly simple thing for "left eigenvector" or is it just "the same thing but backwards" or something?
05:20:58 <oerjan> well sure, f is a fixed point of M*
05:23:26 <oerjan> and just as f : V -> F where V is the vector space and F is the field, you can think of x as embedding into F -> V
05:24:08 <oerjan> i.e. V is isomorphic to Hom(F,V)
05:24:55 <oerjan> and then everything really is just backwards
05:25:05 <shachaf> Yes. And of course people conflate those three all the time.
05:26:36 <oerjan> although the isomorphism V -> F with V isn't canonical, and does not even exist unless finite-dimensional
05:27:17 <oerjan> (or you add continuity to get a hilbert space)
05:27:40 <int-e> oerjan: amazing, ghc head with integer-gmp takes the same 6 second that ghc-7.8.3 did.
05:28:06 <oerjan> int-e: i guess my fears were unwarranted!
05:35:29 <shachaf> oerjan: Why isn't it canonical?
05:35:34 <shachaf> Oh, you have to pick a basis.
05:38:13 -!- dts|gaming has quit (Ping timeout: 250 seconds).
05:40:44 -!- MDude has changed nick to MDream.
06:00:21 -!- FreeFull has quit (Ping timeout: 264 seconds).
06:00:53 -!- FreeFull has joined.
06:01:44 <int-e> oh, I messed up the fibonacci thing quite. tsk. badly.
06:05:50 <int-e> http://sprunge.us/hAAQ?hs actually works (spot the difference!)
06:07:19 -!- FreeFull has quit (Ping timeout: 244 seconds).
06:07:36 <shachaf> Ugh. I always try to double-click a link and then ctrl-shift-c it to copy it to CLIPBOARD.
06:07:52 <shachaf> But if I accidentally ctrl-shift-click it, it opens in my browser.
06:09:38 <int-e> http://sprunge.us/MahQ?hs ...
06:10:37 <int-e> There, finally, without that stupid typo.
06:10:38 <shachaf> int-e: Why do you need BangPatterns?
06:10:44 <shachaf> You're (==)ing on the argument right away.
06:11:08 <int-e> I don't. 'go' had two arguments previously
06:18:44 <int-e> So with integer-gmp2 the code uses less memor ... which actually makes sense; allocating temporaries on the heap is actually quite wasteful, when that memory could be reused for the next temporary.
06:33:28 <shachaf> oerjan: hmm, are vector spaces cofree as well as free?
06:34:32 <oerjan> i don't know what cofree means
06:35:26 <oerjan> although lots of things in vector and module categories are self-dual
06:36:28 <shachaf> it means a right adjoint to a forgetful functor
06:36:56 <oerjan> hm free is a kind of projective, which is dual to injective, which exists for modules over division rings
06:37:24 <shachaf> i decided that the reason everything is so well-behaved with vector spaces (at least finite-dimensional ones) is that everything is free
06:37:39 <shachaf> i.e. the whole "a linear map is determined by its action on a basis" thing
06:38:07 <oerjan> everything is also "injective"
06:38:38 <oerjan> that's a concept for general modules
06:39:07 <oerjan> ok so the forgetful functor maps a vector space to its underlying set
06:39:34 <shachaf> no, i guess the cofree thing doesn't work
06:39:35 <oerjan> no, my brain is too tired for this
06:40:08 -!- oerjan has quit (Quit: ZZLEEP).
06:40:59 <shachaf> oerjan: wait, maybe i'm still mixed up
06:49:41 -!- zlsa has quit (Ping timeout: 244 seconds).
06:51:04 -!- FreeFull has joined.
06:57:39 -!- vanila has quit (Quit: Leaving).
07:03:58 -!- scoofy has quit (Read error: Connection reset by peer).
07:11:45 -!- FreeFull has quit (Ping timeout: 255 seconds).
07:41:05 -!- FreeFull has joined.
07:45:38 -!- Sprocklem has quit (Ping timeout: 265 seconds).
08:09:15 -!- dianne has quit (Read error: Connection reset by peer).
08:09:44 -!- dianne has joined.
08:11:08 -!- FreeFull has quit (Ping timeout: 256 seconds).
08:26:48 -!- shikhin has joined.
08:31:03 -!- FreeFull has joined.
08:50:38 <zzo38> Why does "Light Pattern" esolang use the alphabetical order rather than something such as timestamps in EXIF data?
09:21:02 -!- FreeFull has quit (Remote host closed the connection).
09:21:59 -!- FreeFull has joined.
09:28:43 <zzo38> I realized that the encoding I used for RLE numbers uses one less bit than it does to code the successor of that number with Elias gamma.
09:30:39 <zzo38> (Numbers are encoded in a different way from Elias gamma, though.)
09:34:50 <zzo38> (Equivalently, one less bit than the Exp-Golomb code for the same number, since zero does not need to be encoded.)
09:36:18 <zzo38> (By Exp-Golomb I actually mean Exp-Golomb-0.)
11:01:16 -!- shikhout has joined.
11:04:24 -!- shikhin has quit (Ping timeout: 255 seconds).
11:07:47 -!- variable has quit (*.net *.split).
11:07:49 -!- newsham has quit (*.net *.split).
11:08:15 <HackEgo> [wiki] [[JUMP]] http://esolangs.org/w/index.php?diff=41105&oldid=17872 * 73.184.106.177 * (+136) Added truth-machine
11:09:14 -!- nooga has joined.
11:10:55 -!- variable has joined.
11:10:55 -!- newsham has joined.
11:10:58 -!- variable has quit (Max SendQ exceeded).
11:11:29 -!- variable has joined.
11:12:18 -!- Patashu has quit (*.net *.split).
11:12:19 -!- blsqbot has quit (*.net *.split).
11:12:19 -!- mihow has quit (*.net *.split).
11:12:19 -!- Dulnes has quit (*.net *.split).
11:12:19 -!- yiyus has quit (*.net *.split).
11:12:19 -!- olsner has quit (*.net *.split).
11:12:20 -!- Bike has quit (*.net *.split).
11:12:21 -!- Gregor has quit (*.net *.split).
11:12:50 -!- FreeFull has quit (Ping timeout: 245 seconds).
11:13:36 -!- Patashu has joined.
11:13:37 -!- blsqbot has joined.
11:13:37 -!- mihow has joined.
11:13:37 -!- Dulnes has joined.
11:13:37 -!- yiyus has joined.
11:13:37 -!- olsner has joined.
11:13:37 -!- Bike has joined.
11:13:37 -!- Gregor has joined.
11:13:53 -!- FreeFull has joined.
11:18:56 -!- FreeFull has quit (Remote host closed the connection).
11:43:33 -!- Koen__ has joined.
12:10:25 <lambdabot> EPPO 221200Z 15004KT 6000 SCT011 BKN021 05/03 Q1027
12:26:55 -!- FreeFull has joined.
12:39:45 -!- Patashu has quit (Ping timeout: 240 seconds).
12:40:10 -!- S1 has joined.
12:43:06 -!- Koen__ has quit (Quit: The struct held his beloved integer in his strong, protecting arms, his eyes like sapphire orbs staring into her own. "W-will you... Will you union me?").
12:51:23 -!- drdanmaku has quit (Quit: Connection closed for inactivity).
12:53:31 -!- boily has joined.
13:00:41 -!- Elohimswagger has joined.
13:00:49 -!- Elohimswagger has quit (Client Quit).
13:02:12 -!- Elohimswagger has joined.
13:03:57 -!- Elohimswagger has left.
13:42:51 <J_Arcane> A Y-combinator tutorial in Racket. http://blog.tomtung.com/2012/10/yet-another-y-combinator-tutorial/
13:55:02 <FreeFull> Making a Y combinator in an untyped lisp is easy
13:55:12 <FreeFull> Although you probably want a Z combinator
14:05:38 <J_Arcane> To be honest, I didn't quite follow that explanation either. XD Once they start boiling things down to single letter variables my brain sort of shuts down.
14:31:07 -!- copumpkin has quit (Quit: My MacBook Pro has gone to sleep. ZZZzzz…).
14:41:43 <J_Arcane> Also, man are BASIC's string functions primitive as hell. XD
14:41:45 <J_Arcane> Suddenly getting horrible flashbacks to trying to write command-line parsers in MS-BASIC ...
14:49:30 <fizzie> Standard Forth string functions aren't very much better. Though there's a strstr-alike, SEARCH. http://lars.nocrew.org/dpans/dpans17.htm#17.6.1
14:50:24 <fizzie> There's no MID$, though. You're just supposed to do the arithmetic yourself. /STRING can help.
14:51:18 <fizzie> (And /STRING is just ( a b c -- a+c b-c ).)
14:58:16 <J_Arcane> I like strings-as-lists; so much easier to mangle them that way.
15:19:16 -!- S1 has quit (Quit: S1).
15:25:56 -!- tromp__ has quit (Remote host closed the connection).
16:01:55 -!- boily has quit (Quit: BENTHIC CHICKEN).
16:09:12 -!- S1 has joined.
16:12:34 -!- MDream has changed nick to MDude.
16:14:15 -!- Bike has quit (Quit: thanksgiving).
16:19:14 -!- shikhout has changed nick to shikhin.
16:21:25 -!- zlsa has joined.
16:26:23 -!- tromp has joined.
16:30:45 -!- tromp has quit (Ping timeout: 245 seconds).
16:42:56 -!- copumpkin has joined.
16:44:24 -!- vanila has joined.
16:48:48 <zlsa> is there a bot channel where I can test bots?
16:49:17 <vanila> before there was a lot of bots
16:49:29 <zlsa> all the better to test mine :P
16:50:54 -!- S1 has quit (Quit: S1).
17:01:59 -!- shikhout has joined.
17:05:30 -!- shikhin has quit (Ping timeout: 256 seconds).
17:07:29 <fizzie> Also #esoteric-blah which might be equally dead.
17:07:42 <fizzie> At least if it's somehow related.
17:16:59 -!- drdanmaku has joined.
17:18:22 -!- Bicyclidine has joined.
17:51:55 -!- dts has joined.
17:59:02 -!- Sprocklem has joined.
18:12:21 -!- Bicyclidine has quit (Ping timeout: 240 seconds).
18:16:34 -!- copumpkin has quit (Ping timeout: 264 seconds).
18:18:43 -!- S1 has joined.
18:18:57 -!- S1 has quit (Remote host closed the connection).
18:32:04 -!- copumpkin has joined.
18:43:51 <HackEgo> [wiki] [[TrivialBrainfuckSubstitution]] N http://esolangs.org/w/index.php?oldid=41106 * 68.189.222.97 * (+1791) Created page with "'''TrivialBrainfuckSubstitution''' is not a single programming language, but rather a large family of programming languages that are trivial substitutions of the [[Brainfuck]]..."
18:54:30 <elliott> pretty sure that language literally exists
18:58:23 -!- JazzyFella has joined.
19:00:13 <JazzyFella> Can someone help me with Visual Studio?
19:01:44 <Dulnes> Ask the other ppls i have to go
19:01:58 <JazzyFella> Oh damn, do you know anyone in particular?
19:03:47 <Dulnes> -_- i dont wanna say names incase they cant help but (Elliot vanila) idk if they can help but whatever you could try /wiki/ /the googles/
19:04:46 <JazzyFella> I dont know much about Code.. I'm just the IT guy... Hehehe
19:05:41 <myname> "i have a question about visual studio, i better go and ask #esoteric"
19:05:48 <myname> sounds perfectly legit
19:06:20 <zlsa> JazzyFella, #esoteric is about esoteric programming languages
19:06:30 <zlsa> Visual Studio is _not_, by any definition, esoteric
19:06:48 <JazzyFella> But I was hoping someone could help, it's so basic..
19:07:36 <Dulnes> i dont think visual studio is on the esolang wiki
19:07:53 <myname> i don't know about you, but i do think there aren't that much windows developers here
19:08:04 <JazzyFella> I just saw the description of the chat.. It said Programming blackhole.. So I thought it was legit
19:09:38 <myname> so you spent more time reading channel descriptions than googleing?
19:09:49 <myname> great way to solve problems
19:09:58 <JazzyFella> I googled. The problem I am having is to generic
19:10:39 <JazzyFella> It's this one line of code that's fudged up..
19:11:23 -!- JazzyFella has left.
19:11:26 <Dulnes> If you know how to visual studio you can fix it
19:11:53 -!- 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/.
19:13:03 <myname> elliott: do you hate us that much?
19:14:15 <vanila> i love that i on eof the two people here to consult when it comes to visual studio
19:17:40 <Dulnes> Im naturally nice to people who come here.
19:17:45 -!- dts has quit (Read error: Connection reset by peer).
19:19:39 <myname> Dulnes: but an asshole to people who already were here?
19:20:29 -!- dts has joined.
19:21:12 <Dulnes> I'm sorry, I'm just really tired.
19:22:48 <elliott> I'm pretty sure myname was joking
19:25:47 <Dulnes> Im not good with jokes :/
19:31:46 <HackEgo> [wiki] [[Ook!]] http://esolangs.org/w/index.php?diff=41107&oldid=35897 * 68.189.222.97 * (+97)
19:32:39 <HackEgo> [wiki] [[OOo CODE]] http://esolangs.org/w/index.php?diff=41108&oldid=35331 * 68.189.222.97 * (+88)
19:33:44 <HackEgo> [wiki] [[OOo CODE]] http://esolangs.org/w/index.php?diff=41109&oldid=41108 * 68.189.222.97 * (-88) Undo revision 41108 by [[Special:Contributions/68.189.222.97|68.189.222.97]] ([[User talk:68.189.222.97|talk]]) No it isn't.
19:35:41 <HackEgo> [wiki] [[TrivialBrainfuckSubstitution]] http://esolangs.org/w/index.php?diff=41110&oldid=41106 * 68.189.222.97 * (-130)
19:36:15 <HackEgo> [wiki] [[Alphuck]] http://esolangs.org/w/index.php?diff=41111&oldid=40406 * 68.189.222.97 * (+40)
19:37:08 <HackEgo> [wiki] [[ZZZ]] http://esolangs.org/w/index.php?diff=41112&oldid=37900 * 68.189.222.97 * (+88)
19:42:15 <vanila> unlambda implementation uses refcounting
19:42:18 <vanila> ftp://ftp.madore.org/pub/madore/unlambda/contrib/mandelson-unlambda.c
19:42:21 <vanila> http://www.madore.org/~david/programs/unlambda/
19:42:34 <vanila> but the language has call/cc, so is it really OK?
19:42:41 <vanila> I guess it's ok since we don't have LETREC or something
19:46:57 <b_jonas> vanila: exactly, you can't make reference loops because there are no mutable cells, so it works
20:04:22 -!- contrapumpkin has joined.
20:05:18 -!- copumpkin has quit (Ping timeout: 264 seconds).
20:13:38 <zzo38> I wrote more pokemon cards I made up
20:16:16 <zzo38> Do you have any more question?
20:16:35 <vanila> I liked your gopher site
20:16:43 <vanila> but it seems like the only good site on gopherspace
20:18:35 <zzo38> Well, there are others, but they don't have as many things (and some are partially broken).
20:18:51 <vanila> I might create my own gopher page
20:19:01 <zzo38> There is The Online Book Initiative
20:19:46 <zzo38> At gopher.std.com with the selector string "1/The Online Book Initiative" (the initial 1 is a part of the selector string; so in a URL you must put it twice)
20:20:17 <zzo38> There is also gopher.semmel.ch which has a lot of music files in it
20:20:44 <zzo38> (and text files, too)
20:23:54 -!- contrapumpkin has quit (Quit: My MacBook Pro has gone to sleep. ZZZzzz…).
20:31:11 <J_Arcane> Oh wow. This still exists. http://www.theworld.com/
20:55:10 <HackEgo> [wiki] [[Talk:TrivialBrainfuckSubstitution]] N http://esolangs.org/w/index.php?oldid=41113 * Rdebath * (+256) /* Isn't this the category? */ new section
20:56:33 -!- Patashu has joined.
21:16:27 -!- tromp has joined.
21:21:04 -!- tromp has quit (Ping timeout: 258 seconds).
21:25:32 -!- S1 has joined.
21:27:14 -!- copumpkin has joined.
21:54:24 <Sgeo> I want to play with the MSN Browser
21:54:32 <Sgeo> But seems to need to use MSN these days
21:54:59 <ion> Sentences never uttered before in the history of universe: “I want to play with the MSN Browser”
22:09:14 <Dulnes> msn now can open a tab in the same window :0
22:09:58 <Jafet> Another: “Sentences never uttered before in the history of universe: “I want to play with the MSN Browser””
22:11:02 <Dulnes> They really tried on making it and its something your grandparents can use
22:17:24 <quintopia> which is a simple geometric and beautiful typeface for large text?
22:24:16 <ion> Microsoft Comic Sans
22:25:01 <myname> don't you mean microsoft wingdings?
22:37:01 -!- fungot has quit (Ping timeout: 260 seconds).
23:01:41 -!- shikhin has joined.
23:04:44 -!- shikhout has quit (Ping timeout: 250 seconds).
23:05:55 -!- tromp has joined.
23:10:44 -!- tromp has quit (Ping timeout: 272 seconds).
23:29:08 <zzo38> It looks to be complicated to me to get around the inability in SQLite to use savepoints while a trigger program is running.
23:43:43 <Taneb> You know you can't cook when "it's probably not still frozen" is the best thing you can say about your dinner" :(
23:46:16 -!- nooga has quit (Ping timeout: 255 seconds).