00:01:52 -!- Germille has joined.
00:02:38 -!- Germille has left.
00:08:04 -!- copumpkin has quit (Quit: Computer has gone to sleep.).
00:18:40 -!- augur has quit (Remote host closed the connection).
00:25:17 -!- pikhq has quit (Quit: Lost terminal).
00:31:37 -!- huntercool has joined.
00:31:38 -!- jackylu has joined.
00:31:57 -!- huntercool has left.
00:38:57 -!- jackylu has left.
00:41:49 <Sgeo> WHY WHY WHY does any computer charger I own invariably get such that it's flaky?
00:47:03 -!- copumpkin has joined.
01:05:34 -!- crystal-cola has quit (Quit: leaving).
01:13:07 -!- BeholdMyGlory has quit (Read error: Connection reset by peer).
01:30:33 -!- Fuco has quit (Ping timeout: 260 seconds).
01:37:29 -!- augur has joined.
01:38:19 -!- augur has quit (Remote host closed the connection).
01:49:29 -!- augur has joined.
02:16:42 -!- augur has quit (Remote host closed the connection).
02:19:51 -!- augur has joined.
02:21:13 -!- augur has quit (Remote host closed the connection).
02:33:34 -!- augur has joined.
02:34:07 -!- augur has quit (Remote host closed the connection).
02:58:35 -!- pikhq has joined.
03:02:19 * pikhq is still stunned at how very many build systems don't automagically handle dependencies.
03:10:35 <pikhq> Even the autotools way of handling it is better than some of the shit out there.
03:13:44 <pikhq> And why the pfargtle are there so many Makefile generators?
03:14:01 <pikhq> Topological sort is *not that hard* people!
03:17:10 -!- azaq23 has quit (Ping timeout: 240 seconds).
03:39:35 <coppro> pikhq: define: automagically
03:41:21 <pikhq> coppro: Without having to actually go to any effort to declare dependencies.
03:41:36 <coppro> pikhq: Does -MM and the like count?
03:42:13 <pikhq> I should be able to say "The following C files are in the following program. Make it fucking happen."
03:45:23 <pikhq> And Makefile generators are such a freaking stupid and pointless idea.
03:46:15 <pikhq> "Well, let's make this non-standard tool, which people will have to learn how to use, and install seperately... But let's make it use Make, because they're familiar with how to use Make, and its limitations!"
03:46:22 <pikhq> Worst of both worlds.
03:49:02 -!- pgimeno has joined.
03:49:14 * pikhq is especially confused by Automake.
03:49:50 <pikhq> Is there any particular reason they had to generate a Makefile, even then?
03:52:54 <Gregor> make understand dependencies and (in modern times) parallelism. By using Makefiles, you get incremental rebuilds and parallel builds nearly free.
03:54:15 <pikhq> Gregor: Didn't understand parallel builds back then.
03:54:53 <pikhq> Gregor: And Automake's documentation continues to advocate recursive make, which means it doesn't understand dependencies.
03:55:38 <Gregor> Hence "(in modern times)", and recursive make can be made to understand dependencies so long as the separate processes are sufficiently distinct.
03:55:59 <Gregor> Or, more precisely, it can be made to correctly not understand dependencies.
03:56:28 <pikhq> Yes, but it so rarely is made to do so.
03:57:01 <Gregor> And? You think some newfangled custom creation cobbled together from scripts and distributed as part of the source would do better?
03:57:34 <pikhq> I'm simply saying that Automake could have been a decent build system from the start, rather than a newfangled custom creation cobbled together from scripts.
03:58:05 <Gregor> Auto* set out to be something that could be run on any vaguely-Unixy system.
03:58:22 <Gregor> Its requirements are sh, make and a not-entirely-braindead suite of core utilities.
03:58:49 <Gregor> That's its DEVELOPER dependency.
03:58:57 <pikhq> Okay, fair enough.
03:59:07 <Gregor> No, not "fair enough", that's exactly what makes autotools autotools.
03:59:32 <Gregor> Mind you, that's mostly irrelevant nowadays, but in the context of its original design it makes sense.
04:00:43 <pikhq> Does it? Keep in mind that someone using it is almost certainly going to be building a C program, meaning that they also have a not-entirely-braindead C compiler.
04:01:19 <pikhq> Meaning that having a C program instead of a cobbled-together bunch of scripts would not be unreasonable.
04:08:51 -!- augur has joined.
04:20:28 -!- azaq23 has joined.
04:20:58 -!- vorik111 has joined.
04:21:39 <pikhq> quintopia: Finland: the only place with winters Russia is afraid of.
04:22:56 <quintopia> isnt kliment a finnish name tho? was the russian commander a fin?
04:24:14 <pikhq> It's apparently a Slavicisation of Latin "CLEMENT".
04:26:34 <pikhq> Kliment Voroshilov was apparently a very notable Russian figure. One of the kinds of tanks used in that war was *named after him*.
04:26:44 <pikhq> He was ordering around Kliment Voroshilov tanks.
04:34:59 <quintopia> http://imgur.com/gallery/lEeOM look at first comment
04:37:37 -!- tim000 has joined.
04:38:06 <quintopia> http://imgur.com/gallery/aBglP "who printed the man page for gcc?"
04:39:25 <pikhq> Looks more like the man page for mplayer.
04:40:56 -!- asiekierka has joined.
04:42:30 -!- azaq231 has joined.
04:44:26 -!- azaq23 has quit (Ping timeout: 240 seconds).
04:58:24 -!- kwertii has joined.
05:13:26 -!- pgimeno has left ("This is the default part message").
05:29:57 <pikhq> It's snowing rather *well*.
05:37:27 <pikhq> You know it's bad when a Norwegian is mocking you.
05:37:53 <oerjan> well it was snowing here only a couple weeks ago
05:38:37 <pikhq> Still a bit weird having it snowing this late in April.
05:40:20 <pikhq> Mind you, I shouldn't be so weirded out. I've had blizzards for my birthday (~a month ago)...
06:21:02 -!- lament has joined.
06:24:03 -!- FireFly has joined.
06:33:30 -!- Tritonio has joined.
06:41:27 -!- copumpkin has quit (Ping timeout: 240 seconds).
06:44:34 -!- oerjan has quit (Quit: leaving).
06:47:30 -!- copumpkin has joined.
06:51:58 -!- monqy has quit (Ping timeout: 260 seconds).
06:53:51 -!- monqy has joined.
07:05:02 -!- lament has quit (Ping timeout: 276 seconds).
07:07:14 -!- FireFly has quit (Quit: swatted to death).
07:13:13 <pikhq> What the what? "Quaint" and "cunt" are cognates.
07:13:50 <pikhq> Sorry, no, that's a bullshit claim that I should have read more about before posting.
07:25:44 -!- lament has joined.
07:25:48 -!- Slereah_ has joined.
07:26:17 -!- Slereah has quit (Ping timeout: 276 seconds).
07:33:40 -!- Tritonio_Droid has joined.
07:35:06 -!- Tritonio has quit (Ping timeout: 246 seconds).
07:53:10 -!- Tritonio_Droid has quit (Ping timeout: 252 seconds).
07:55:26 <olsner> pikhq: you shouldn't have corrected yourself
08:03:05 -!- pikhq has quit (Quit: Lost terminal).
08:08:58 -!- copumpkin has quit (Ping timeout: 260 seconds).
08:09:25 -!- copumpkin has joined.
08:10:08 -!- aloril has quit (Ping timeout: 260 seconds).
08:16:57 -!- kwertii has quit (Quit: bye).
08:22:43 -!- aloril has joined.
08:25:15 <cheater99> http://www.simplimg.com/product_images/8ABpR.jpg
08:27:02 -!- azaq231 has quit (Ping timeout: 250 seconds).
08:39:41 -!- lifthrasiir has quit (Ping timeout: 248 seconds).
08:39:49 -!- lifthrasiir has joined.
08:44:24 -!- lament has quit (Ping timeout: 246 seconds).
08:57:42 -!- pingveno1 has quit (Ping timeout: 252 seconds).
08:58:26 -!- myndzi has quit (Ping timeout: 252 seconds).
08:58:27 -!- pingveno has quit (Ping timeout: 240 seconds).
08:58:48 -!- azaq23 has joined.
09:13:09 -!- Fuco has joined.
09:13:39 -!- myndzi has joined.
09:15:43 -!- pingveno has joined.
09:16:42 -!- pingveno1 has joined.
09:34:36 <Sgeo> Is this a good guideline for knowing when to comment?: If you have to think about the code you're about to write, there should be a comment
10:12:13 <Fuco> no, if you have to think, you should rewrite it so it would be obvious
10:12:33 <Fuco> comments are mostly useless and indicate you don't know what you're doing:)
10:12:51 -!- FireFly has joined.
10:18:22 <cheater99> you should only comment when the language does something a person less experienced would not notice
10:22:35 <vorik111> and there's a minimum level of experience you can expect that person to have.
10:22:55 <vorik111> depends on your corporate policy also.
10:23:35 <vorik111> commenting always referred to documentation for a particular managed change to the code.
10:24:20 <vorik111> "managed change" is the key oxymoron here :)
10:32:08 <Sgeo> I wanted to comment to write out a piece of math that I needed to write out before fully understanding what I was doing
10:33:15 <Sgeo> And now I'm screaming at C# for not having a nice comprehendible way of getting the lengths of single dimensions in a multidimensional array
10:36:27 <fizzie> What's wrong with foo.GetLength(0) .. foo.GetLength(n-1), where n is the number of dimensions?
10:36:50 <Sgeo> Me not checking that GetLength() takes a dimension
10:39:32 <Sgeo> This has to be done in 4.5 hours :(
10:43:25 <Sgeo> Wow. What I'm doing here crappily would be far better modelled in a prototype OOP system
10:43:32 <vorik111> comments don't need to stay. if you need a comment to organize your thoughts, do it. then erase the evidence of any noobery.
10:44:13 * Sgeo excises the bit that was manipulating a multidimensional array
10:48:09 -!- vorik111 has changed nick to vorik111|zzz.
11:00:41 -!- Tritonio has joined.
11:02:01 -!- Tritonio_Droid has joined.
11:05:43 -!- Tritonio has quit (Ping timeout: 264 seconds).
11:07:48 -!- Sgeo has quit (Ping timeout: 250 seconds).
11:14:30 <Vorpal> going to be offline for a while. House shaking due to roof being replaced. This doesn't bode well for the HDDs. Besides it is time to clean the computer case.
11:15:05 -!- Vorpal has quit (Quit: ZNC - http://znc.sourceforge.net).
11:38:18 -!- monqy has quit (Quit: hello).
11:50:32 -!- Gregor has quit (Excess Flood).
11:50:39 -!- Gregor has joined.
11:50:42 -!- azaq23 has quit (Ping timeout: 250 seconds).
11:51:05 -!- Gregor has changed nick to Guest72394.
11:55:01 -!- azaq23 has joined.
11:55:14 -!- Phantom_Hoover has joined.
11:58:56 -!- siracusa has quit (Ping timeout: 250 seconds).
12:00:28 -!- siracusa has joined.
12:01:55 -!- BeholdMyGlory has joined.
12:06:37 -!- augur has quit (Remote host closed the connection).
12:14:15 -!- Sgeo has joined.
12:14:18 -!- crystal-cola has joined.
12:32:22 -!- Guest72394 has changed nick to Gregor.
12:49:22 -!- elliott has joined.
12:56:03 -!- augur has joined.
13:13:00 -!- augur has quit (Remote host closed the connection).
13:23:36 -!- augur has joined.
13:40:55 -!- azaq23 has quit (Quit: Leaving.).
13:57:03 -!- Kordalien has joined.
13:58:43 -!- Kordalien has quit (Client Quit).
14:06:46 -!- Vorpal has joined.
14:19:34 -!- newpuppyuser2 has joined.
14:20:48 -!- newpuppyuser2 has left.
14:34:28 <Gregor> The behavior of fputws() depends on the LC_CTYPE category of the current locale.
14:34:28 <Gregor> In the absence of additional information passed to the fopen(3) call, it is reasonable to expect that fputws() will actually write the multibyte string corresponding to the wide-character string ws.
14:34:35 <Gregor> I love that last sentence.
14:37:42 -!- `Fuco` has joined.
14:40:42 -!- Fuco has quit (Ping timeout: 246 seconds).
14:46:42 -!- Germille has joined.
14:47:25 <Germille> you're the pro hacker on steroids that do crazy programming languages?
14:53:21 -!- copumpkin has quit (Quit: Computer has gone to sleep.).
14:56:21 -!- Tritonio_Droid has quit (Quit: Bye).
14:57:07 <lambdabot> Phantom_Hoover: You have 2 new messages. '/msg lambdabot @messages' to read them.
14:57:59 -!- Tritonio has joined.
14:58:30 -!- Tritonio has quit (Read error: Connection reset by peer).
15:22:19 -!- copumpkin has joined.
15:22:47 -!- ais523 has joined.
15:27:45 <ais523> elliott: -> #scapegoat
15:37:48 -!- elliott has quit (Ping timeout: 240 seconds).
15:38:15 -!- lament has joined.
15:40:19 <Gregor> I wish dosbox had a curses mode.
15:47:25 -!- lament has quit (Ping timeout: 240 seconds).
15:56:51 -!- augur has quit (Remote host closed the connection).
16:05:05 -!- Germille has quit (Ping timeout: 250 seconds).
16:05:28 -!- sebbu has quit (Read error: Connection reset by peer).
16:06:06 -!- sebbu has joined.
16:06:06 -!- sebbu has quit (Changing host).
16:06:06 -!- sebbu has joined.
16:09:37 -!- augur has joined.
16:16:46 -!- Mannerisky has quit (Ping timeout: 260 seconds).
16:18:21 -!- elliott has joined.
16:19:21 -!- Sgeo has quit (Ping timeout: 258 seconds).
16:28:57 <ais523> hmm, I fear patent lawyers are going to birthday-paradox themselves at this rate
16:29:06 <ais523> patents are abbreviated to the last three digits of their numbers in legal filings
16:29:18 <ais523> so if you have more than about 30 patents mentioned in one filing, there's quite a large chance of a collision
16:30:23 <elliott> you could always disambiguate, presumably
16:31:15 <ais523> also, is "evinces" a real word?
16:31:22 <ais523> "This inequitable conduct evinces Microsoft’s unclean hands and taints Microsoft’s entire portfolio."
16:31:30 <ais523> surely they meant "evidences"? or is this something I'm missing
16:32:44 <elliott> evinces3rd person singular present of e·vince (Verb)
16:32:44 <elliott> 1. Reveal the presence of (a quality or feeling).
16:32:44 <elliott> 2. Be evidence of; indicate: "man's inhumanity to man as evinced in the use of torture". More »
16:32:51 <elliott> ais523: to evince something is to read it with Evince
16:32:58 <ais523> elliott: I like your definitoin better
16:32:59 <elliott> so, they took a look at Microsoft's unclean hands, in PDF form
16:34:06 <ais523> this is an actual Microsoft sues Linux (specifically Android, which is based on the Linux kernel but not most of the other stuff "Linux" implies) case, which is interesting that it's happened
16:35:27 <elliott> Microsoft really don't seem to give a shit about Linux
16:35:43 <elliott> they've made, like, a handful of adverts targeting it altogether, I can only think of one
16:35:45 <ais523> they only get upset at it where it competes with them, I think
16:35:54 <ais523> and on the desktop, it doesn't meaningfully
16:36:02 <ais523> they've attacked it quite heavily on the server, but mostly lost that battle
16:36:07 <ais523> and are starting to attack it in the portables space, now
16:36:13 <elliott> the one I'm thinking of was an old sever ad, yup
16:37:35 <Phantom_Hoover> <elliott> Microsoft really don't seem to give a shit about Linux ← hmm, I thought the big thing about the Halloween documents was that it demonstrated that they do, just not publicly.
16:38:04 <copumpkin> crystal-cola: what anti-gay attitude?
16:38:12 <elliott> Phantom_Hoover: Well, OK, but those are pretty oooold.
16:38:16 <ais523> to me, the Halloween documents demonstrated that they were aware it existed (not surprising), and were considering how much of a threat it was to their business model (also not surprising)
16:38:38 <ais523> I don't think they demonstrated that Linux was a serious threat to them then (it wasn't), but that they were keeping an eye on it just in case
16:38:57 <elliott> ais523: ISTR some kind of "crush" verbiage, but crushing things is kind of Microsoft's default stance.
16:50:40 <crystal-cola> it sucks that you can find teh galois group of a polynomial but still not know how to solve it
16:52:01 <elliott> yeah we should all be bob the builder instead
16:53:52 <crystal-cola> sincte left the bracketing out to make it remind of "a rose is a rose is a rose"
16:54:12 <crystal-cola> but the actual intended bracketing was not ((p or not p) or (not (p or not p)))
16:55:46 <crystal-cola> Problems of solvability and non-solvability of algebraic, transcendental and differential equa-
16:55:49 <crystal-cola> tions in explicit form, including Liouville theory, classical and differential Galois theory, and Picard-
16:55:52 <crystal-cola> Vessiot theory, will be addressed from the geometry and analysis point of view. The course will be
17:00:20 <crystal-cola> "So we've defined something very close to a uniform probability distribution on the integers, and all it took was the axiom of choice"
17:06:02 <crystal-cola> elliott: http://mathoverflow.net/questions/63127/is-this-joke-unfunny-closed
17:09:24 <elliott> i meant the mathoverflow links
17:09:27 -!- `Fuco` has changed nick to Fuco.
17:09:31 <crystal-cola> DAE think that the Kronecker Delta would be an awesome name for a spaceship
17:09:36 <crystal-cola> aside from being a very important term in linear algebra, haha.
17:10:10 <elliott> ok you're not answering my question
17:10:19 <crystal-cola> I'm no super-calculator or anything like that... But I love algebra. Basically math in everyday life.
17:10:22 <crystal-cola> The other day I worked out how long it would take to try each combination of sauce (15 Sauces, 2 sauces in a combination) on chicken strips in the cafeteria. When I mentioned this to one of my friends on a weekend trip, he basically laughed his ass off. Imagine my surprise... Why do my peers not see the sense in that??
17:11:06 <elliott> crystal-cola: are you going to...?
17:11:27 <elliott> <elliott> that's not an answer
17:11:27 <elliott> <crystal-cola> he didn't post his lecture notes
17:11:27 <elliott> <elliott> i meant the mathoverflow links
17:11:55 <elliott> i was asking "what is this" to the mathoverflow link
17:12:55 <crystal-cola> know any good tips for getting to sleep when you're nto that tired?O
17:14:10 <elliott> no, i'm terrible at falling asleep. melatonin seems to help quite a bit, though.
17:15:46 <ais523> you find it easier to sleep in the sun?
17:17:03 <ais523> isn't that the usual source of melatonin?
17:17:17 <elliott> crystal-cola: Melatonin (pronounced /ˌmɛləˈtoʊnɪn/ ( listen)), also known chemically as N-acetyl-5-methoxytryptamine,[1] is a naturally occurring compound found in animals, plants, and microbes.[2][3] In animals, circulating levels of the hormone melatonin vary in a daily cycle, thereby allowing the entrainment of the circadian rhythms of several biological functions.[4]
17:17:17 <elliott> Many biological effects of melatonin are produced through activation of melatonin receptors,[5] while others are due to its role as a pervasive and powerful antioxidant,[6] with a particular role in the protection of nuclear and mitochondrial DNA.[7]
17:17:18 <elliott> In mammals, melatonin is secreted into the blood by the pineal gland in the brain. Known as the "hormone of darkness", it is secreted in darkness in both day-active (diurnal) and night-active (nocturnal) animals.[8]
17:17:25 <elliott> tl;dr it regulates your circadian rhythm
17:17:30 <ais523> hmm, I've got it backwards, tehn
17:17:37 <elliott> its use as a supplement is pretty common in the US
17:17:47 <elliott> (where it's classified as a "dietary supplement")
17:18:17 <ais523> anyway, I find the best way to get to sleep at a particular time is to get into a rhythm; and also, to have to wake up early the next day
17:18:46 <elliott> ais523: I sleep easiest when I have nothing to do the next day, and find it absolutely impossible to get into any kind of rhythm
17:18:49 <ais523> (the latter situation is because humans are quite capable of falling asleep and waking up at will, but it isn't under conscious control, so you need to get your subconscious to do it somehow)
17:19:00 <elliott> (sometimes I flat-out cannot sleep at all because of having to get up early the next day)
17:19:09 <ais523> hmm, that's the other way round from me, then
17:19:15 <ais523> do you wake up 5 minutes before the time you set your alarm to, too?
17:19:28 <elliott> crystal-cola: erm, I mean I take it as a supplement
17:19:31 <ais523> crystal-cola: you could try the placebo effect
17:19:39 <ais523> if you believe you can fall asleep easily, you will
17:19:44 <elliott> i use the kiddie liquid bottle because i can't swallow pills :)
17:19:48 <ais523> so just practice lying until you fool yourself
17:19:52 <elliott> ais523: that only works for people who are good at self-delusion :)
17:20:05 <elliott> ais523: actually i find the idea of convincing myself i can fall asleep easily more horrifying than the gain of being able to do so
17:20:15 <ais523> elliott: I got better at swallowing pills, when I had to for three weeks
17:20:31 <elliott> <ais523> do you wake up 5 minutes before the time you set your alarm to, too?
17:20:33 <ais523> in the end, I decided that the best method was to swallow some water first, then physically force the pill down my own throat, using my fingers
17:20:38 <elliott> now alarms don't wake me at all
17:20:47 <ais523> which is quite different from the usual wash-the-pill-down method
17:20:58 <Gregor> <ais523> in the end, I decided that the best method was to swallow some water first, then physically force the p... down my own throat, ...
17:21:09 <ais523> Gregor: it was funnier the first time
17:21:18 -!- Tritonio has joined.
17:21:20 <elliott> Gregor: ...which is quite different from the usual wash-the-pill-down method
17:21:37 <elliott> ok, now on to less stupid things
17:21:46 <elliott> just kidding, everything we talk about is stupid
17:21:58 <elliott> yep i totally said you were stupid
17:22:47 <elliott> they totally ruined that whale song with their remix wtf
17:22:51 <elliott> guitar totally doesn't go with whale.
17:23:01 <elliott> or wait was that minecraft...
17:29:06 <elliott> achievements are perfectly ignorable.
17:32:44 <Ilari> APNIC: 3x1k+2x256+/32+/48 to Australia, 4x1k to China, 3x1k to India, 8x1k+2x/32 to Japan, 1k to Malaysia, 1k to New Caledonia, 1k to Taiwan, 3x1k to Vietnam.
17:32:56 <Ilari> Doesn't look good. :-/
17:41:21 -!- Tritonio has quit (Read error: No route to host).
17:41:54 -!- Tritonio has joined.
17:41:57 <Ilari> 0.9909328 blocks of IPv6 space free, 0.98796 blocks IPv4 space free.
17:44:00 <Ilari> crystal-cola: Any further progress with finding the base solutions of n|5^n-3^n? :-)
17:44:52 <crystal-cola> I can't remember if we proved that impossible or just 5 2^k impossible
17:46:11 <Ilari> crystal-cola: No, 3 or 5 can't divide n for any solution.
17:46:26 <crystal-cola> iin that case it means that 3 and 5 are invertible
17:47:14 <crystal-cola> I guess the problem with that is the value 5^-1 depends on n
17:47:26 <Ilari> Actually, only depends ok k.
17:47:47 <crystal-cola> I'm not sure if it can help a lot but who knows
17:49:03 <Ilari> 5^(k*2^r) == 3^(k*2^r) (mod k). Multiplying this yields: 1 == ((3/5)^k)^(2^r) (mod k).
17:50:00 <Ilari> (3/5)^k mod k is easy to calculate.
17:51:28 <Ilari> But x^(2^r) mod k isn't, since r has upper bound of k^2.
17:52:00 <Ilari> Invert 5 mod k, multiply mod k and use modexp.
17:52:43 <Ilari> Actually, if one could compute order of (3/5)^k mod k in Z_k...
17:53:50 <ais523> hmm, Slashdot are busy discussing the JavaScript "for (x in document.write) {document.write(x);}", which apparently crashes IE6
17:53:55 <ais523> I'm trying to figure out what, if anything, it means
17:55:19 <Gregor> That would write every property of document.write into the document.
17:55:52 <Gregor> However, document.write should have NO properties, so why that would do anything is beyond me.
17:55:56 <Gregor> But then, we are talking about IE6.
17:56:01 <Ilari> For k=17, 3/5 mod k is 4. 4^17 mod 17 is 4. The order of 4 is 4, and therefore r_min is 2.
17:56:06 <Gregor> For all I know, one of the properties is <meta lolcrashthebrowser />
17:56:06 <fizzie> It might have a "prototype" property. That's what I get for a random function.
17:56:25 <Gregor> fizzie: It should have a prototype property, but AFAIK that shouldn't be indexed by for...in.
17:56:53 <Gregor> Err, yeah, s/every/every enumerable/ above :P
17:57:13 <Ilari> crystal-cola: Basically, The order of (3/5)^k mod k in Z_k must be power of two for there be solutions of form k*2^r.
17:58:12 <Ilari> crystal-cola: Which actually brings r_min bound down to: r_min < k.
17:58:17 <fizzie> This is of course just an empirical test, but
17:58:18 <fizzie> js> function foo() { return 42; }
17:58:18 <fizzie> js> for (x in foo) print(x);
17:59:09 <crystal-cola> Ilari: I think that gives a very fast algorithm to to check it !
17:59:59 <Ilari> crystal-cola: Nope, it is still exponential.
18:00:32 <Gregor> fizzie: Hm. What engine?
18:00:34 <Ilari> And that inversion assumes k is not multiple of 5, but it can't be multiple of it anyway.
18:00:55 <Gregor> function foo() {}; for (var x in foo) print(x); // has no output for me
18:01:24 <fizzie> Gregor: ECMA-262 does say the "prototype" property has a false [[Enumerable]] attribute, so that sounds more right.
18:01:37 <fizzie> This "js" seems to be from Rhino.
18:01:43 <Gregor> fizzie: Which is why I'm curious what --- ah, Rhino.
18:01:51 <Gregor> Curious, Rhino shouldn't be so broken, it's from Mozilla :P
18:02:00 <Gregor> But then, until I came 'round, Narcissus was broken *flexes*
18:03:10 -!- elliott has quit (Remote host closed the connection).
18:06:50 <Ilari> Actually, even tighter bound: 2^r_min must divide phi(k).
18:09:41 <Ilari> And since phi(k) = prod((p_l-1)*p_l^(m_l-1)) and k is odd, this is the sum of power of two in all distinct prime factors of k.
18:10:13 <crystal-cola> that's a lot tighetr actually because the number of primes factors seems small
18:11:05 <Ilari> E.g. For k = 2329... 2329 = 17 * 136 -> r_min <= 4 + 3 = 7.
18:13:49 <Ilari> That's a vast improvement over old bound of 2329!
18:14:14 <crystal-cola> I think it can be used to compute a large amont fo data on the irreducible pairs
18:14:30 <crystal-cola> maybe that will be enough to find what they are
18:15:30 <Ilari> One can actually apply it to that 3^(k*2^r)==5^(k*2^r) (mod k) formula (which doesn't need element inversion.
18:16:47 <Ilari> In fact, with that sort of algorithm, the factorization of k is the only thing that is superpolynomial. The algorithm is subexponential with efficient integer factorization.
18:17:26 <crystal-cola> multiply all the odd ones together and check them
18:17:36 <Ilari> Actually: Even faster algorithm: phi(k) < k. Take ceil(log2(k)) as upper bound.
18:22:17 -!- Sgeo has joined.
18:23:20 <Ilari> Combined with powmod algorithm, the whole thing is fully polynomial-time.
18:24:40 * Sgeo wants to sneeze on his math professor
18:25:22 <Sgeo> crystal-cola, I'll let this channel decide that
18:27:20 <Ilari> And one doesn't need to factorize k values in order to tell if those are reducable or not: One can use gcd to extract the relevant factors.
18:28:00 <crystal-cola> hah I didn't think of that, that's really naet
18:28:44 <Sgeo> 624272c5c20b7799dc73c6050405d2ba41487c4b
18:29:11 <Sgeo> When looking at a uniform continuous probability distribution, is or is not the height of the line a probability?
18:30:30 <Ilari> Basically, compute gcds aginst all previous known irreducable k's, divide them out. Repeat until all gcds give 1. If the number is still not 1, then add the original to irreducables.
18:31:01 <ais523> `addquote <elliott> ais523: YOU WILL HAVE YOUR QUOTE SOON
18:31:06 <HackEgo> 382) <elliott> ais523: YOU WILL HAVE YOUR QUOTE SOON
18:31:42 <Ilari> Or add it anyway if r_min is smaller than largest among non-coprime irreducable k.
18:31:51 <Sgeo> Ilari, comment?
18:31:55 -!- elliott has joined.
18:32:01 <Sgeo> elliott, read logs?
18:32:03 <elliott> i have been paged, setting off grumpy bastard mode
18:32:26 <elliott> Sgeo: what are you paging me for
18:32:39 <Sgeo> To hopefully answer my question
18:32:53 <Ilari> Sgeo: It isn't a probabilty. It is probabilty density.
18:33:08 <Sgeo> That's what I thought
18:33:13 <Sgeo> Please tell my math professor
18:33:26 <Sgeo> And what I SHA-1'd: I assert that it is not a probability. I attempted to show examples where it would be 4, and where what he claimed was not compatible with using a more general formula he gave to work out the answer.
18:33:28 <Ilari> Sgeo: To have any real probabilty for some value in continuous distribution, you need a delta function at that value.
18:34:18 <Sgeo> What notation would one use for a probability density?
18:34:28 <Sgeo> I think he said "density" at one point, maybe, but dropped it later
18:34:40 <Sgeo> And did use P(X) as its label...
18:34:57 <Ilari> Given continuous distribution p(x) the probabilty for value in range [x1,x2] is int(p(x),x,x1,x2).
18:35:19 <Sgeo> Oh, so his notation is acceptable?
18:36:03 <Sgeo> Then why did he say that my example of a uniform distribution between 0 and 1/4 must be impossible, because P(X) would be 4 and that's not a valid probability
18:36:35 <Ilari> You can have uniform distribution over any finite interval.
18:36:56 <Sgeo> Maybe he misunderstood what I was saying?
18:37:05 * Sgeo , defender of idiots
18:37:16 <elliott> thank you, Sgeo, for obsoleting me entirely.
18:37:29 <Sgeo> Actually, he was pretty ticked at me. Told me to wipe the smug smile off my face, and asked if I wanted to teach the class
18:37:39 <elliott> it's funny because he teaches at a terrible college
18:38:05 <elliott> sort of like that obama guy what does he do all day anyway
18:38:24 <Ilari> With continuous distributions, the condition for validity is that the integral of probabilty desntity over support must be exactly 1.
18:38:49 <Sgeo> Ilari, is there a way to get you in contact with my math professor?
18:39:40 <lifthrasiir> yeah, the pdf is nothing to do with an actual probability; the integral of it is.
18:39:44 <crystal-cola> Sgeo: looks like you were right, so in my view you should just stop caring now
18:40:01 <crystal-cola> Sgeo: you said everything already, trying to convince him wont work - he might realize it though
18:40:09 <Sgeo> lifthrasiir, it occurs to me that he did say pdf
18:40:12 <lifthrasiir> and if your math professor did say so then it is unfortunate
18:41:26 <elliott> pdf is a document format idiots ;D
18:42:05 <Ilari> For discrete distribution, the sum of all probablities must equal 1 of course.
18:42:42 <Sgeo> I think in my attempts to understand him in a way that made him not an idiot, I asked if we were talking about discrete distributions. He said no.
18:42:46 <Ilari> And all discreate distributions can be represented as continuous ones (but that results in delta spikes).
18:43:00 <Sgeo> After I was done trying to present my proof, he decided to reteach me what a discrete distribution was
18:43:11 <Sgeo> crystal-cola, I just want to be certain that I'm not utterly mistaken
18:43:50 <elliott> <crystal-cola> just be right about everrything
18:43:50 <elliott> <crystal-cola> and deal with it
18:44:28 <crystal-cola> just write out the definition of probaility distribution or whatever it is
18:44:29 <Sgeo> crystal-cola, that's what I intended to do by asking here
18:44:37 <elliott> Sgeo: you could also try thinking
18:44:52 <crystal-cola> Sgeo: well can you say formall what it is you're wondering?
18:44:54 <Sgeo> It was my thinking that lead me to my conclusion that he was wrong
18:45:02 <elliott> comex: you scallywag: http://www.pastie.org/private/orzgwxc2ax80vrs8v0rew
18:45:22 <crystal-cola> Sgeo: in society we are normall taught that "if thinking makes you think $authority_figure is wrong, stop thinking"
18:46:00 <elliott> are we actually taught that explicitly or implicitly
18:46:08 <elliott> there's a difference between that and it arising from other teachings
18:47:10 <elliott> what an unusual occurrence!!!
18:47:25 <Ilari> Hah... If your doctor told you you have 4.0mmol/l of LDL in your blood, would you believe it? :-)
18:48:20 <Ilari> I wouldn't, because that is utterly ridiculous amount.
18:48:21 <Sgeo> crystal-cola, is there a theorem proving that probability distribution isn't probability?
18:48:37 <crystal-cola> Sgeo: no im just asking "what is it you're not certain about"
18:48:59 <Sgeo> crystal-cola, that somehow, by some means that I don't understand, my professor might be right somehow
18:49:07 <elliott> Ilari: i have 9009 mmol/l of LDL in my tiger blood.
18:49:16 <crystal-cola> im asking you what the mathematical statement is
18:49:26 <elliott> ugh, another esowiki link on proggit
18:49:28 <elliott> this is going to be painful
18:49:48 <Sgeo> Is <what I now know as a probability ... argh, a pdf> a probability?
18:49:54 <elliott> oh wow the joke is that it has fap in it
18:50:01 <elliott> so funny i made it in here years ago
18:50:50 <Sgeo> crystal-cola, what do you want?
18:51:05 <crystal-cola> you had some kind of disagreement with your teacher
18:51:15 <crystal-cola> what mathematical statement is it that you are questioning
18:51:34 <elliott> crystal-cola: i was attempting to help you communicate
18:51:45 <elliott> Sgeo: crystal-cola wants you to write out your disagreement in Coq
18:51:47 <Sgeo> crystal-cola, his claim, that 1/(b-a) in a continuous distribution is a probability of some sort
18:52:12 -!- pingveno1 has quit (Quit: leaving).
18:52:12 -!- pingveno has quit (Quit: leaving).
18:52:42 -!- pingveno has joined.
18:53:53 <Sgeo> I'll take a picture
18:55:36 <Sgeo> http://i.imgur.com/farXp.jpg
18:55:42 <lifthrasiir> Sgeo: http://m.wolframalpha.com/input/?i=normal+distribution+with+mean%3D0+sigma%3D0.01 pdf with a peak of ~40.
18:56:41 <elliott> get dijkstra to write it out for you
18:56:50 <crystal-cola> When you say that the probability of X = 160 in the example of uniform distribution was 1/50 did you mean P(159,5<X<160,5)? If not, see below;
18:57:11 <crystal-cola> how could I possible make sense of this without the context
18:59:34 <elliott> Sgeo: again, please restate in coq
19:04:06 -!- KingOfKarlsruhe has joined.
19:04:12 -!- Germille has joined.
19:06:53 <elliott> 10:12:13: <Fuco> no, if you have to think, you should rewrite it so it would be obvious
19:06:53 <elliott> 10:12:33: <Fuco> comments are mostly useless and indicate you don't know what you're doing:)
19:07:05 <elliott> Fuco: this is true only when you are writing software requiring no complex algorithms whatsoever
19:07:34 <Fuco> what would be a complex algorithm
19:08:12 <elliott> it's true, applications that amount to an interface to a database require pretty much no commenting.
19:09:01 <Fuco> I'm currently working on big constraint solver library (www.unitime.org), there are hardly any comments in the code
19:09:17 <Fuco> the code can still be obvious
19:09:23 <Fuco> sure, you have to study the algorithm
19:09:32 <Fuco> but that's not the problem of how you rewrite it in some prog. language
19:10:06 <Fuco> but if I know I'm going to write IFS with perturbation counter or whatever, I can study that beforehand
19:10:12 <Fuco> and than write code that is understandable
19:10:20 <Fuco> for people who are familiar with it obviously
19:10:23 <elliott> well, it can easily be reductio-ad-absurdumed: if comments were never necessary, mathematical papers would have no English text
19:10:26 <Fuco> but otherwise you shouldn't read it anyway
19:10:39 <elliott> Fuco: that doesn't work if you're creating new algorithms for the purpose
19:10:43 <elliott> because nobody is familiar
19:10:54 <elliott> and telling people "hey go look at this paper I wrote" is just a comment that you've collapsed into a url
19:10:56 <Fuco> well that's what research papers are for
19:10:57 <elliott> and that is more likely to get out-of-date
19:11:44 <elliott> SO YOU CAN PRINT OUT A BOOK THAT'S YOUR PROGRAM
19:12:02 <elliott> AND THEN PUBLISH YOUR BOOKRAM IN THE RESEARCH JOURNALS
19:12:13 <cheater99> elliott: most mathematical papers are in chinese anzwazs.
19:12:14 <elliott> and everyone can scan it and OCR it and compile it
19:12:55 <Sgeo> Wasn't there a crypto book that was just the source code to a program?
19:13:38 -!- vorik111|zzz has changed nick to vorik111.
19:14:07 <cheater99> the human language text in mathematical proofs is not code duplication of the rest, unlike it is with comments
19:17:53 <Sgeo> I _just_ realized that I may have a password lying around with Sony
19:17:59 <Sgeo> I don't know if it's on PSN or not
19:18:41 <dnm_> Sego: _Cracking DES_ includes OCR-able source code.
19:19:14 <elliott> where did all these new people come from anyway
19:19:17 <Sgeo> My name is not Sg?? Eo??
19:19:21 <elliott> is it another one of those stealth invasions
19:19:41 <dnm_> Woops, I meant Sgeo.
19:19:50 <dnm_> I should have let tab completion do the job for me. l[
19:20:59 <Sgeo> Is station.sony.com connected to PSN?
19:21:20 <elliott> dnm_: Are you from the Haskell army.
19:21:57 <dnm_> What's my right against self-incrimination?
19:22:13 <dnm_> I use Haskell, but I'm not in the army.
19:22:18 -!- monqy has joined.
19:22:23 <elliott> dnm_: Absolutely nonexistent is what it is.
19:22:32 <elliott> I'm just verifying you're not one of those #haskell agents.
19:22:40 <dnm_> Has there been an influx of Haskellistas here?
19:23:10 <elliott> i think copumpkin did once
19:23:16 <Sgeo> "Representatives from SOE reminded people that SOE systems and databases are separate from PSN's"
19:23:20 <elliott> basically we're drowning in haskellers.
19:23:38 <elliott> it doesn't matter if you use haskell, it's being from there that's the issue :D
19:23:38 <dnm_> Well, a friend of mine who is into Haskell more than I am did tip me off to the channel.
19:23:50 <elliott> dnm_: who was it, we'll have them dispatched swiftly
19:23:52 <dnm_> And he probably frequents the channel.
19:23:55 <dnm_> He's not here.
19:24:09 <dnm_> He's been gone off IRC for about a week, I think he's busy.
19:24:16 <dnm_> He's a good guy.
19:24:18 <elliott> hmm, who's here and been off irc for about a week
19:24:40 <dnm_> I, on the other hand, am just a languages nut.
19:24:49 <elliott> I could grep for all names here in the past week, and then grep the current logs for all those names
19:24:53 <dnm_> Somehow I didn't know about this channel.
19:24:53 <elliott> and find the ones with no matches
19:25:03 <elliott> yes, this is a viable plan
19:25:03 <cheater99> dnm_: there's lots of places like this
19:25:06 * Sgeo stabs a guess at oerjan
19:25:19 <elliott> ok so monday is twentyfifth
19:25:30 <elliott> so wish i had my number keys round about now
19:25:39 <fungot> `1234567890-=~!@#$%^&*()_+
19:25:51 <dnm_> Why all the #haskell influx anyway?
19:26:01 <elliott> lots of people here are in there
19:26:07 <elliott> and mentioning this channel is fUN
19:26:24 <elliott> [~/logs/_esoteric]% echo 2011-04-[20-27].txt 1
19:26:25 <elliott> zsh: no matches found: 2011-04-[20-27].txt
19:26:25 <elliott> gah zsh stop being terrible :(
19:26:34 <crystal-cola> frustrating when people don't clarify their questions
19:30:28 <dnm_> 15:23 < dnm_> NOT TELLING.
19:30:36 <elliott> % for nick in `awk '/<.*>/ { gsub(/[<>]/, "", $2); print $2 }' 2011-04-{20,21,22,23,24,25}.txt | sort -u`; do if ! grep -c "$nick" 2011-04-{25,26,27}.txt >/dev/null; then echo $nick; fi; done
19:30:39 <elliott> look at this monster, people
19:30:45 <dnm_> I will not be your monkey..., er, oracle.
19:30:47 <elliott> wrote a long annoying shell line for him
19:30:52 <elliott> narrowed it down to very few people
19:30:56 <elliott> and yet still he tortures me so
19:31:40 <elliott> ah, hmm, perhaps it is rapido or Mathnerd314
19:31:53 <elliott> lifthrasiir: ok but it was still annoying to write :) bear in mind i have no number keys
19:31:56 -!- Sgeo has quit (Read error: Connection reset by peer).
19:32:22 <elliott> hmm, technically it could also be TeruFSX or FireFly, if they keep their irc client running while they're away
19:32:26 <dnm_> This wiki is great compared to the old catseye site.
19:32:37 <elliott> dnm_: lol, cat's eye just has cpressey's languages
19:32:56 <dnm_> Yeah, but it's what I was used to as my jumping off point for esolang stuff.
19:33:12 <elliott> the problem with the wiki is that... 90 percent of the languages are crap
19:33:24 <dnm_> I remember the epiphany I had when I first encountered Unlambda.
19:33:55 <elliott> argh, dnm_ connects through a vps so I can't use geolocation to track down who it could be
19:34:01 <elliott> i'm dedicated to my lynchings, me
19:34:48 <dnm_> I mean, you can find out where the VPS is.
19:35:04 <elliott> that doesn't exactly help, since anyone can buy a linode and it'll be in a us data centre
19:35:39 <elliott> i suppose i could interrogate everyone who comes back next week.
19:35:46 <dnm_> Though, I'm publicly logged on other freenode channels already. I'm in the D.C. area.
19:36:10 <elliott> YESSSSSSSSSSSSSSSSSSSS that narrows it down to absolutely everyone since this is the internet
19:36:31 -!- news-ham has joined.
19:36:31 <elliott> the news ham is right here
19:36:33 <dnm_> At least everyone is a finite number!
19:36:46 <news-ham> Kindle gets library book lending: Users of Amazon's Kindle e-reader will soon be able to borrow electronic books from libraries in the US. http://www.bbc.co.uk/go/rss/int/news/-/news/technology-13155967
19:37:24 <elliott> thank you for your valuable opinion
19:37:30 <news-ham> Urban birds have bigger brains: City dwelling birds have larger brains relative to their body size, according to scientists. http://news.bbc.co.uk/go/rss/int/news/-/earth/hi/earth_news/newsid_9468000/9468306.stm
19:37:45 <elliott> crystal-cola: what exactly is pukeworthy about the kindle
19:37:51 <elliott> apart from the amazon shit
19:37:54 <news-ham> UK economy sees return to growth: The UK economy grew 0.5% in the first three months of this year, making up ground lost at the end of 2010. http://www.bbc.co.uk/go/rss/int/news/-/news/business-13206430
19:38:05 <news-ham> World Snooker live - O'Sullivan v Higgins: Three-time Crucible champions Ronnie O'Sullivan and John Higgins continue their battle for a spot in the World Championship semi-finals. http://news.bbc.co.uk/go/rss/int/news/-/sport1/hi/other_sports/snooker/9469117.stm
19:38:11 <elliott> i.e. the remote control bullshit which is irrelevant if you just pirate the books :)
19:38:19 <news-ham> Killed daredevil was new recruit: A stunt man who plunged to his death at a Kent show was a new recruit to the daredevil team, according to a friend. http://www.bbc.co.uk/go/rss/int/news/-/news/uk-england-13205471
19:38:26 <elliott> did i leave it in debug mode
19:38:35 <news-ham> Ban on stem-cell patents 'wrong': Research involving human embryonic stem cells in Europe is under threat, says a group of leading scientists. http://www.bbc.co.uk/go/rss/int/news/-/news/health-13214036
19:39:18 <elliott> i'll paste a list, gimme a second
19:39:38 -!- news-ham has quit (Remote host closed the connection).
19:39:43 -!- news-ham has joined.
19:39:58 <elliott> crystal-cola: http://sprunge.us/ULSS
19:40:24 <elliott> crystal-cola: just mention it in your message. science and magic.
19:41:35 * dnm_ reads wiki page on DateFuck
19:41:45 <dnm_> Hillarious. Not what I was expecting.
19:41:45 <news-ham> 'More pupil benchmarks' to come: The English Baccalaureate is the first of many new performance measures, as ministers seek to make information available to parents in England, says Schools Minister Nick Gibb. http://www.bbc.co.uk/go/rss/int/news/-/news/education-13207055
19:42:03 <cheater99> news-ham: are you just being slow?
19:42:13 <elliott> yes, it starts with brainfuck and takes away all brainfuck instructions
19:42:34 <news-ham> Cutie does a cover of "6 Foot 7 Foot" by Lil Wayne...wait? Whats that in the background?!?!? http://www.reddit.com/r/videos/comments/gykip/cutie_does_a_cover_of_6_foot_7_foot_by_lil/
19:42:48 <elliott> it should return news about reddit instead
19:42:52 <dnm_> Phantom_Hoover: Doesn't seem to be.
19:43:16 <news-ham> Fed cuts US GDP growth forecast: The Federal Reserve cuts its economic growth forecast for this year, citing weaker growth than expected in the first three months of the year. http://www.bbc.co.uk/go/rss/int/news/-/news/business-13213424
19:43:32 <news-ham> McGee 'turns down' Creation drama: Former record label owner and musician Alan McGee turned down the chance to make a drama about his company Creation Records, he reveals. http://www.bbc.co.uk/go/rss/int/news/-/news/entertainment-arts-13162825
19:43:41 <elliott> dnm_: Phantom_Hoover is three years old, he says hi.
19:43:43 <news-ham> Map of the route: Explore the processional route to be taken by the newly-married couple, from Westminster Abbey to Buckingham Palace. http://www.bbc.co.uk/go/rss/int/news/-/news/uk-12819687
19:43:46 <news-ham> Climate change to hit US rivers: Scarce water supplies in the western US will probably dwindle further as a result of climate change, a US government report says. http://www.bbc.co.uk/go/rss/int/news/-/news/world-us-canada-13190689
19:44:00 <elliott> crystal-cola: do you actually have fingers or do you just kind of mash your face into the keyboard until you get the right sentence
19:44:42 <elliott> that's what my face says in response
19:46:27 <news-ham> Timings on the day: The following timings have been confirmed for royal wedding day. http://www.bbc.co.uk/go/rss/int/news/-/news/uk-13097243
19:47:12 <dnm_> Speaking of news
19:47:15 <dnm_> http://www.guardian.co.uk/science/the-lay-scientist/2010/sep/24/1
19:48:42 <elliott> you haven't seen that before? :P
19:48:59 <dnm_> Not until earlier today.
19:49:05 -!- Sgeo has joined.
19:49:06 <dnm_> spacedino.png is my new mascot though
19:49:08 <elliott> well I guess Americans are a bit slow ;D ;D ;D
19:49:39 <dnm_> The article itself, or the style it mocks?
19:49:41 <elliott> it should carry the death penalty
19:49:46 <Sgeo> Maybe my professor didn't have probability confused with probability density, but is so terrible at correcting people that he didn't see that I thought he was saying "probability"
19:50:52 <elliott> it was written by a single blogger
19:51:00 <elliott> whose blog happens to be hosted by the Guardian
19:51:00 <Sgeo> Although then again, he suggested that the 0 to 1/4 situation that I set up must be impossible
19:51:15 <elliott> crystal-cola: ok, please show where Martin Robbins is guilty of what it mocks
19:51:39 <elliott> crystal-cola: isn't it obvious, it has no spelling mistakes
19:51:50 <elliott> i would like to see them print http://static.guim.co.uk/sys-images/Guardian/Pix/pictures/2010/9/24/1285359869220/spacedino.png though :D
19:51:56 -!- asiekierka has quit (Remote host closed the connection).
19:52:44 <elliott> maybe cover the whole front page with it
19:52:48 <elliott> YAKAWOW is of course the headline
19:53:41 <dnm_> "Ron Armontrout" is a great name.
19:54:02 <dnm_> (apropos of nothing)
19:54:44 <elliott> not as good as armouredtrout
19:55:00 <dnm_> I was thinking of Armortrout
19:55:04 <dnm_> But it seemed too easy.
19:55:26 <elliott> ron armontrout seems to be an Actual Name according to google
19:56:03 -!- Wamanuz3 has joined.
19:56:06 -!- Wamanuz2 has quit (Ping timeout: 260 seconds).
19:56:46 <dnm_> He's a book author, among other things.
19:57:02 <elliott> nonsense, he has no wikipedia article
19:57:20 <dnm_> "Irina Lyublinskaya and her colleagues, Dr. Valeriy Ryzhik and Ron Armontrout, just finished a new book titled: Calculus Explorations with Geometry Expressions."
19:58:53 <elliott> it would have to be "calculus adventure fun"
19:59:08 -!- Tritonio has quit (Quit: Leaving).
20:04:58 <dnm_> Teaching Calculus through Geometry Expressions (a graphical constraint-based geometric modeling and math tool)
20:05:05 <dnm_> http://www.geometryexpressions.com/
20:07:45 <crystal-cola> Cam Welcome to computerized mathematical modeling
20:07:52 <dnm_> The book? I have no idea, I don't have it. I have GX though, and I like it.
20:08:04 <Vorpal> uh, calculus is only tangentially related to geometry.
20:09:09 <elliott> dnm_: crystal-cola is two days old. he says hi.
20:09:43 <Vorpal> elliott, ah, your previous negative age explains so much
20:09:53 <elliott> Vorpal: no, i just hand over to a new elliott every seconds
20:10:16 <Vorpal> elliott, ah that explains it even better.
20:10:23 <elliott> dnm_: i swear this channel is occasionally interesting.
20:10:32 <Sgeo> Hello bluebird on my head, aren't you glad the old one's dead?
20:11:33 <Vorpal> one main issue is that a lot of the esoteric programming things were already discussed before. Someone needs to invent something new that oerjan can then prove TC.
20:11:48 <elliott> i never said it was interesting /esolang/ discussion :D
20:12:10 <elliott> crystal-cola: oolzybub and murphy
20:12:28 <elliott> http://esolangs.org/wiki/Oozlybub_and_Murphy
20:12:28 <elliott> http://catseye.tc/projects/oozlybub-and-murphy/
20:12:40 <Vorpal> elliott, what about that one that switches interpreter every other instruction?
20:12:56 <elliott> Vorpal: that doesn't really make it hard
20:13:01 <elliott> it's not two interpreters, just two different "modes"
20:13:13 <Phantom_Hoover> Oozlybub and Murphy has possibly my favourite name ever.
20:13:23 <elliott> Phantom_Hoover: the justification is better
20:13:23 <elliott> The name of this language is Oozlybub and Murphy. Despite appearances, this name refers to a single language. The majority of the language is named Oozlybub. The fact that the language is not entirely named Oozlybub is named Murphy. Deal with it.
20:13:32 <Vorpal> elliott, indeed but iirc the computational class is not known?
20:13:44 <elliott> Vorpal: indeed, but it's strongly suspected to be sub-TC
20:14:02 <Vorpal> quite, but do you happen to remember what it was called?
20:15:44 <Vorpal> elliott, if the Goldbach Conjecture was false, what would the computational class of Oozlybub and Murphy be?
20:16:07 <Vorpal> elliott, yes but which sub-TC
20:16:12 <elliott> well, it wouldn't be able to infinite-loop.
20:16:16 <elliott> so not very good at all :)
20:16:28 <elliott> i think all loops would be fixed length
20:16:46 <Vorpal> elliott, yes yes... I gather that. But there is still quite a lot of different computational classes left to choose from
20:17:12 <Vorpal> elliott, for a start would it be bounded by input length?
20:17:35 -!- oerjan has joined.
20:18:22 -!- Germille has quit (Read error: Connection reset by peer).
20:19:05 <Vorpal> "Once a dynast has been executed, it continues to exist until the program halts, but it may never be executed again." <-- hrrm, what significance does it have that it continues to exist?
20:19:08 -!- Germille has joined.
20:19:18 <elliott> i think the number of dynasts matter? dunno.
20:19:22 <elliott> it might just be more chris wackiness.
20:19:48 <Vorpal> like "TRIVIA PORTION OF SHOW"?
20:20:43 <Vorpal> elliott, I suspect cpressy is somewhat crazy.
20:20:54 <elliott> says the only sane and most boring person in here
20:21:05 <elliott> Latest news: 2011.0427: None can avoid the luminous precipitation of Oozlybub and Murphy version 1.1! Read more on our news page, or subscribe to our RSS feed.
20:21:08 <Vorpal> elliott, arguably I'm not completely sane. Or I wouldn't be in here.
20:21:16 <elliott> it wasn't there a few minutes ago when i checked
20:21:23 <elliott> April 27, 2011: Version 1.1 of the Oozlybub and Murphy programming language has been released. This update tries to clarify some of the errors in the specification while also slathering some extra goo onto it like a wimpmode. Enjoy, then enjoy again, then PLEASE DO KEEP ENJOYING UNTIL YOUR EYES BURN WITH DEEP DEEP CORROSION.
20:21:27 -!- pingveno has quit (Ping timeout: 240 seconds).
20:21:32 <Vorpal> (New in 1.1) An Oozlybub and Murphy program is in wimpmode if it
20:21:32 <Vorpal> declares a global variable of integer type which matches the string am a wimp,
20:21:45 <Vorpal> elliott, I saw that when I loaded it a few less minutes ago
20:23:30 -!- pingveno has joined.
20:24:27 <oerjan> <Vorpal> elliott, arguably I'm not completely sane. Or I wouldn't be in here. <-- maybe your insanity is similar to Gregor's sense of taste
20:24:51 <elliott> also, it's his sense of smell
20:24:57 <Vorpal> oerjan, you mean, a lot of pink?
20:25:07 <Vorpal> or what sort of taste did you mean
20:25:10 <oerjan> ...not that kind of taste
20:25:18 <oerjan> actually maybe that too XD
20:25:35 <Vorpal> oerjan, oh you didn't mean "bad taste" as in "that colour combination is bad taste"?
20:25:40 <Vorpal> but actually literal taste?
20:25:59 <oerjan> no, i meant literally. you have not managed to miss Gregor's hyposmia too, have you?
20:26:14 <Vorpal> oerjan, I believe I have
20:26:27 <oerjan> Gregor: hey i corrected it!
20:26:42 <Gregor> It's only euphnarphia if it fnarfs good :P
20:26:49 <Vorpal> elliott, BY THE WAY. I had today free. I finished reading homestuck. Or I assume I did, I reached a point of no more "going forward" link.
20:27:05 <Vorpal> (I did it on another computer that has flash)
20:27:17 <oerjan> no eu- applies to the sense not to the fnarf
20:27:27 <elliott> It took me and Phantom_Hoover five days.
20:27:28 <Vorpal> elliott, plus a few hours yesterday evening yes
20:27:39 <elliott> You DID expand the pesterlogs right
20:27:50 <elliott> i'm going to step away slowly now
20:27:57 <elliott> you read three thousand seven hundred strips in one day
20:27:58 <Vorpal> elliott, and I read them. I did speed up reading the white guy one by using firebug to change bg colour
20:28:07 <Vorpal> but that was just a few comics
20:28:09 <elliott> Vorpal: erm i just used ctrl+A :D
20:28:23 <elliott> You didn't skip the recaps right?
20:28:24 <Vorpal> elliott, I found that removed the colour differences between the speakers
20:28:36 <oerjan> maybe Gregor's taste in colors is simply victorian. i read somewhere they thought green and orange went well together
20:28:44 <Vorpal> elliott, actually I did skip the recaps. Or rather I glanced over them.
20:28:58 <Vorpal> bbl, going to sleep. Early morning tomorrow
20:28:58 <elliott> Vorpal: They revealed semi-important plot details, but you've probably realised what they were by now.
20:29:01 <Sgeo> I think I tend to skim everything I read
20:29:17 <elliott> Vorpal: If you don't mind I'm going to go sit quietly in a corner and punish myself for being a slow reader.
20:31:34 <Vorpal> Phantom_Hoover, hm I think Eridan had one?
20:32:12 * Vorpal looks at the all page thingy to confirm this and randomly clicks one mentioning that guy.
20:32:14 <Phantom_Hoover> Vorpal, did you actually read the Pesterlogs in detail or kind of skim them?
20:32:17 <Vorpal> yep seems so http://www.mspaintadventures.com/?s=6&p=004349
20:32:29 * elliott rocks gently in his corner of happiness
20:32:47 <Vorpal> Phantom_Hoover, I ate food. That is about it. Oh and slept a bit during the night,
20:33:04 <elliott> OK I coded a little bit but still.
20:33:06 <elliott> Hey Phantom_Hoover you slowed me down.
20:33:20 <Vorpal> Phantom_Hoover, oh and I *did not read* anything but homestuck. None of the previous, problem sleuth and so on
20:33:24 <Vorpal> I'll leave those for later
20:33:40 <Phantom_Hoover> I kept up with you despite you making me stop when you weren't there and you having nights to get a buffer up.
20:33:47 <elliott> Vorpal: Problem Sleuth is the only other one, and it's only slightly over a fifth of the length.
20:33:52 <elliott> The others are ridiculously incomplete and tiny.
20:33:56 <Phantom_Hoover> If I'd actually read it all the time I wanted to, I'd have finished sooner.
20:34:18 <elliott> Vorpal: I haven't read PS either because it doesn't look nearly as interesting.
20:34:30 <Vorpal> I found that time travel in the interlude (forgot which one, when the agents were taking down those green guys) utterly confusing.
20:34:45 <elliott> It's simple. The oven goes forward at one second per second.
20:35:18 <Vorpal> elliott, I meant the other parts. The stuff about following future trail in the past and what not.
20:35:24 <Vorpal> and different timelines
20:35:58 <elliott> I have no idea what you are talking about.
20:36:14 <olsner> elliott, always the joker
20:36:24 <elliott> Vorpal: P.S. http://mspaintadventures.com/?s=ryanquest is way better than Homestuck.
20:36:38 <Phantom_Hoover> We were also applying brainpower to the events, rather than just cramming them into our heads.
20:36:47 <Vorpal> oh god, I never liked dinosaur comics..
20:37:12 <elliott> But that's okay because Ryanquest is basically a gigantic mockery of Ryan North.
20:37:20 <olsner> Phantom_Hoover: maybe he just doesn't have any brainpower to apply
20:37:38 <elliott> Phantom_Hoover: yeah we've totally got all that plot sorted out
20:37:48 <elliott> it is extremely obvious what is going on obviously
20:37:55 <olsner> he'll just go through it as fast as you can say "herp derp" 3700 times
20:38:25 <Vorpal> elliott, also I never had quite such an feeling of abrupt anti-climax as when I found no "next" button... Worst end of archive binge I ever had.
20:38:26 <elliott> Phantom_Hoover: sorry busy reading herp derp
20:38:38 <Vorpal> probably because it was the longest archive binge in number of panels
20:38:40 <elliott> Vorpal: Your achive binges usually end in something other than a comic with no next button?
20:38:55 <Vorpal> elliott, no, I said worst one, not only one
20:39:04 <Vorpal> elliott, anyway IWC took longer to binge than this one
20:39:27 <Vorpal> Phantom_Hoover, well, the journal entry was quite interesting
20:39:35 <oerjan> ais523: Talk: Var needs deletion
20:39:45 <elliott> i am going to make a language based around it
20:41:07 <elliott> ais523: you have failed to be happy and love
20:41:27 <Vorpal> Phantom_Hoover, anyway I think I spent ~13 hours on it today and about 4-5 yesterday evening.
20:42:08 <Vorpal> Phantom_Hoover, indeed. It is just elliott who is a slow reader.
20:42:22 <elliott> Dude, I went at seven hundred and fifty strips per day.
20:42:24 <Vorpal> Phantom_Hoover, and I'm a moderately fast reader.
20:42:31 <elliott> Also I had no difficulties staying ahead of PH while awake.
20:42:34 <Phantom_Hoover> I was slower than him on the Pesterlogs, although I maintain it was because he kept disrupting my reading.
20:42:43 <Vorpal> elliott, come on, have you never finished a Discworld book in a day?
20:42:46 <elliott> Phantom_Hoover: OR MAYBE YOU'RE ILLITERATE
20:42:51 <Vorpal> I have, with the recent thick ones.
20:42:53 <elliott> Vorpal: ANYWAY, AT LEAST OUR REBOUND FROM THE SUDDENLY SLOWISH PACE HAS BEEN EASIER TO WEATHER
20:43:12 <elliott> OK admittedly it's just because there's almost certainly a Flash coming up.
20:43:13 <Vorpal> elliott, how often does it update?
20:43:13 <Phantom_Hoover> And because he insisted that I not read far beyond him when he was asleep, which was apparently until four in the afternoon.
20:43:18 <elliott> Vorpal: Five strips per day.
20:43:29 <elliott> When it doesn't, it's because something big is coming. Usually.
20:43:49 <Vorpal> ah... I think I will have to keep up on a weekly basis then. Some days I don't even have time to read a single comic...
20:43:49 <elliott> Vorpal: (On average, that is.)
20:44:14 <Vorpal> elliott, why do you think there is a flash comming up?
20:44:19 <Vorpal> (and those are annoying)
20:44:22 <oerjan> elliott: hm. why do these spam bots keep making Talk pages for non-existing pages...
20:44:25 <elliott> Vorpal: You misspelled "awesome"?
20:44:49 <Vorpal> elliott, well I have no sound connected to the computer
20:44:52 <elliott> Vorpal: Anyway, because (a) there is never this long between regular strips and (b) SUPER LEAKED INFO suggests it is.
20:44:55 <Vorpal> so I guess I missed out on that aspect
20:45:09 <fizzie> Exploring everything (incl. all trickster mode music room songs and whatnot) in those playable segments takes a whole lot of time.
20:45:14 <Vorpal> elliott, well *shrug* it was just no convenient
20:45:16 <elliott> You can be friends with fizzie.
20:45:41 <fizzie> I've listened to them sounds since then, though.
20:45:48 <elliott> Yes, so clearly Vorpal will have to as well.
20:46:11 <elliott> Vorpal: Reread with sound on.
20:46:32 <Phantom_Hoover> Vorpal, demand you solve entirely the dynamics of a proton in a day.
20:46:56 <Vorpal> elliott, I don't think I will have one for several weeks. Due to excessively bad scheduling I had 50% of the "nominal" scheduled time at university this weak, I will have 150% the next week.
20:46:57 <fizzie> The MSPA wiki has links to most of the sound-enabled ones, that's how I did it posthumously.
20:47:24 <elliott> I like how you could devote nineteen hours to reading homestuck but zero hours to plugging some speakers in
20:47:38 <oerjan> > replicate 3700 (:"erp ") <*> "hd"
20:47:39 <lambdabot> ["herp ","derp ","herp ","derp ","herp ","derp ","herp ","derp ","herp ","d...
20:47:40 <Vorpal> Phantom_Hoover, yeah, it is certainly more linear than homestuck pesterlog convos :D
20:48:26 <Vorpal> wait, I forgot. I spent two of those 18 hours offline
20:48:43 <Vorpal> one I spent cleaning out my computer case with compressed air
20:49:06 <Phantom_Hoover> Still not astonishing, particularly since you obviously got the games done with as quickly as possible.
20:49:18 <Vorpal> Phantom_Hoover, indeed, I hurried with those
20:49:31 <elliott> "What games? Sometimes the Flashes were just me standing there and I clicked next."
20:49:31 <Vorpal> Phantom_Hoover, and I never claimed it was astonishing.
20:50:21 <Vorpal> <fizzie> The MSPA wiki has links to most of the sound-enabled ones, that's how I did it posthumously. <-- got a link to the page listing them?
20:50:41 <fizzie> It's the one titled "sound", IIRC.
20:51:06 <elliott> http://mspaintadventures.wikia.com/wiki/Sound
20:51:18 <elliott> Is that one properly ordered, though?
20:51:37 <elliott> Also, it lacks at least [S] Kanaya: Return to the core and [S] Seer: Descend.
20:51:43 -!- wareya has quit (Read error: Connection reset by peer).
20:51:45 <elliott> Vorpal: Just go on the archive page and grep for [S].
20:51:46 <Vorpal> they seem to be ordered by... track?
20:51:47 <elliott> fizzie: Why didn't you do that.
20:52:03 <fizzie> It's ordered by song, not strip, yes.
20:52:35 <fizzie> And I didn't care about exhaustive rereading, I just listened what sounded interesting.
20:52:37 -!- wareya has joined.
20:52:50 <elliott> But but ALL the [S]es are awesome.
20:53:01 <elliott> Although clearly Zillyhoo is by far the best.
20:53:06 <elliott> I have based this fact on logic.
20:54:06 <elliott> Phantom_Hoover: FIGURED OUT HOW HE DID IT
20:54:11 <elliott> Vorpal: I bet you didn't even read the linked SBaHJ strips.
20:54:15 <elliott> Phantom_Hoover: MYSTERY SOLVED
20:54:21 <elliott> As we all know, those take up the remaining hours.
20:54:42 <Vorpal> elliott, those related to dave and drawn incredibly shitty?
20:54:48 <elliott> Yes, and referenced constantly.
20:55:02 <Vorpal> I read a few. But I have problem with that level of shitty drawing
20:55:18 <elliott> They're not drawn, man, he retroactively went back and made them all with the SBaHJifier.
20:55:27 <elliott> This is canon insofar as I believe it as of this line.
20:55:37 <elliott> Anyway, at least one of them provided a plot hint. :p
20:55:40 <Vorpal> elliott, besides there didn't seem to be many things related to the main plot there
20:55:58 <elliott> ("are you next?" when Kanaya died linked to a strip where someone (Jeff?) was killed for being a vampire, IIRC.)
20:56:09 <elliott> Oh wait, LOL I TOTALLY ROTTHIRTEENED THE ABOVE LINE
20:56:20 <elliott> (I don't know how up-to-date fizzie is.)
20:56:57 <elliott> fizzie probably has like a two hundred line IRC window, so this could take a while.
20:57:06 <Vorpal> elliott, also I'm having problems keeping up with the huge cast. Plus I'm bad a remembering names in real life. I'm much better with remembering faces than names.
20:57:18 <Vorpal> so those pesterlogs are sometimes a bit confusing.
20:57:31 <elliott> I don't even look at anything but the colour and the typing quirk.
20:58:07 <Vorpal> elliott, yes that kind of helps. A bit.
20:58:22 <quintopia> i bet one of these centuries i'll find enough time to read/play through all of homestuck
20:58:28 <fizzie> I am looking through the 99x20 phone-terminal, so no worries there.
20:58:40 <quintopia> i don't think it'll be any time soon tho
20:58:41 <fizzie> (Also last currentsized myself today.)
20:58:49 <Vorpal> fizzie, is that 99 lines?
20:58:51 <elliott> quintopia: Only takes NINETEEN HOURS if you're Vorpal.
20:59:04 <fizzie> Vorpal: Width times height.
20:59:07 <elliott> Vorpal: Eighty by twenty-four should make the answer obvious.
20:59:26 <Vorpal> elliott, dude he has a rotated monitor.
20:59:31 <elliott> Vorpal: Come to think of it I could probably do it entirely by colour, but it's hard not to notice the quirks.
20:59:33 <Vorpal> elliott, he is obviously crazy :P
20:59:47 <elliott> Vorpal: Anyway, you have ... what was it, six less to deal with by now.
20:59:58 <Vorpal> elliott, 18-2 (one for cleaning computer case, it was alarmingly dusty) = 16.
21:00:30 <elliott> Vorpal: Gebyyf. Gurl'ir xvaqn qvrq.
21:00:35 <elliott> obviously quintopia will one day recall this conversation
21:00:40 <elliott> THAT IS WHY I MUST ROTTHIRTEEN
21:00:43 -!- Sgeo has quit (Ping timeout: 258 seconds).
21:00:46 <Vorpal> I don't have a rot13 converter at hand
21:01:02 <Vorpal> elliott, could try /msg ?
21:01:02 <cheater99> it has advanced cryptography options
21:01:07 <quintopia> v unir n pbairegre ohvyg vagb zl pyvrag
21:01:31 <fizzie> "tr a-z n-za-m" is always handy.
21:01:56 <elliott> In fact, forget the hookers.
21:03:33 <elliott> I hereby declare that -minecraft is now the hash-esoteric Homestuck channel.
21:04:04 <elliott> Vorpal: It's not like it gets much traffic.
21:04:16 <oerjan> <Vorpal> one main issue is that a lot of the esoteric programming things were already discussed before. Someone needs to invent something new that oerjan can then prove TC.
21:05:03 <oerjan> it's good to have a purpose...
21:06:24 -!- augur has quit (Remote host closed the connection).
21:06:32 <elliott> I wonder what ASCII character "a bunch of pool balls animating rapidly" is.
21:07:33 <oerjan> <elliott> it doesn't matter if you use haskell, it's being from there that's the issue :D
21:07:47 <oerjan> hey nothing wrong with haskellers, we need more purely functional esolangs
21:08:07 <elliott> it's just THE CHANNEL OF EVIL
21:08:58 <olsner> someone could make a purely functional dialect of Perl, and call it Purl
21:09:31 <oerjan> could be confused with a lolcode dialect
21:11:51 <oerjan> now a purely functional lolcode dialect...
21:12:46 <olsner> *purrly functional programming
21:13:19 <olsner> that's almost worth doing
21:13:57 <Vorpal> <oerjan> now a purely functional lolcode dialect...
21:13:57 <Vorpal> <olsner> *purrly functional programming
21:14:06 <Vorpal> it is oerjan who make the puns
21:14:10 <oerjan> WHERE FIBBY LIKE 1 THEN ZIP WITH PLUS OF FIBBY AND TAIL OF FIBBY
21:14:36 <Vorpal> oerjan, s/PLUS/PUS/ obviously
21:15:04 <olsner> Vorpal: maybe I'm just also oerjan
21:15:07 <oerjan> sorry, i'm not fluent in lolcat
21:15:37 <Vorpal> oerjan, I was trying to make a cat joke but typoed it
21:15:39 <olsner> Vorpal: one of those is the fluid you drain from an abscess, I wonder which
21:15:59 <Vorpal> olsner, puss = kitty. I typoed first time
21:16:19 <crystal-cola> can you make a self modifying programming language?
21:16:28 <crystal-cola> it's not TC until you've interacted with it enough
21:16:40 <crystal-cola> like at first it's just random output or something
21:17:26 <oerjan> well it would have to be very random then, because a language which becomes TC after giving it a fixed input is itself TC
21:19:44 <elliott> http://www.dangermouse.net/esoteric/petrovich.html
21:20:33 <crystal-cola> there should be a simple self modifying algorithm which you can train to be a properl language
21:23:39 <elliott> then it's just a TC language
21:28:30 <olsner> "just a TC language" :)
21:36:09 <elliott> relevant game idea: you have to mash on the keyboard, and you win if the statistical randomness is good enough
21:36:35 <olsner> sounds like programming perl?
21:41:23 -!- KingOfKarlsruhe has quit (Quit: ChatZilla 0.9.86.1 [Firefox 3.6.16/20110319135224]).
21:42:47 <olsner> cheater99: well, can you?
21:47:32 -!- Germille has quit (Quit: HydraIRC -> http://www.hydrairc.com <- Would you like to know more?).
21:47:37 <oerjan> entropy is the expected information in each new bit, given knowledge of the previous ones.
21:47:59 <crystal-cola> can we compute it to within say 3 decimal places for IRC users?
21:48:17 <oerjan> a markov chain analysis could do at least an approximation
21:48:30 <crystal-cola> how do we know that markov chain is related to entropy though?
21:48:49 <crystal-cola> I mean doesn't that assume we're markov chains? :/
21:49:23 <oerjan> no. it's an approximation, you look at all sequences of n letters and see what usually follows them.
21:49:24 <crystal-cola> it seems like that just takes the language into account and not the higher level concepts that a person decides to talk about
21:49:35 <elliott> so you want a strong ai that understands what we talk about
21:49:49 <oerjan> as n -> infinity, if you had a large corpus it would give the entropy
21:49:51 <crystal-cola> like the entropy of someone that thinks about completely different things every day
21:50:15 <crystal-cola> is much higher than that of someone who just talks about footbal everyday
21:50:17 <oerjan> and indeed you are right about the higher level concepts not being captured this way
21:50:37 <crystal-cola> since we all use english language (pretty much) I'm guessing our entropy is roughly equal
21:51:01 <oerjan> i've read that the entropy is supposedly approximately one bit per letter
21:51:19 <elliott> crystal-cola: that's a good thing
21:51:23 <elliott> language should be redundant
21:51:32 <elliott> it makes transmission and understanding easier
21:51:33 <olsner> ISTR that ~10 bits per word is normal for english
21:51:59 <crystal-cola> hey what would an "opposite" markovv chain do?
21:52:03 <olsner> but that's the word entropy, which would be a different thing than character or byte entropy
21:52:07 <elliott> crystal-cola: be incoherent
21:52:10 <crystal-cola> like a markov chain that always outputs the LEAST probable symbol?
21:52:16 <elliott> I tried that once, I think
21:52:22 <elliott> crystal-cola: yes, take a regular one and change max to min
21:52:27 <elliott> it just spews nonsense constantly
21:52:33 <elliott> but it only ever repeats phrases people have said
21:52:39 <elliott> so it's not "TEH MOST RANDOM TEXT"
21:52:43 <elliott> just "things people have only said once"
21:53:04 <elliott> that's sort of impossible with any corpus-based solution
21:53:08 -!- augur has joined.
21:53:08 <crystal-cola> imagine a thing that outputs a sequence of bits
21:53:25 <crystal-cola> and it also predicts what output it is going to do and does the opposite
21:54:04 <elliott> when anyone says "imagine a thing/machine", I get this immediate image of a literal black box with LEDs on the front
21:54:12 <elliott> and it's not hollow, it's solid
21:54:16 <elliott> inside is pure, one hundred percent unobtanium
21:54:23 <elliott> I take my black boxes seriously
21:54:34 <crystal-cola> so given some bits how do I predict the next one?
21:54:36 <olsner> elliott: now imagine The Internet - http://www.youtube.com/watch?v=iDbyYGrswtg
21:55:01 <oerjan> crystal-cola: maybe if you output the _second_ most likely letter instead of the last you could get something less incoherent...
21:55:06 <elliott> crystal-cola: I think you need to, like
21:55:12 <elliott> crystal-cola: rework the markov chain each time
21:55:21 <olsner> elliott: well, obviously, you live in the country that had it on TV after all
21:55:23 <elliott> you mean with a markov chain, right?
21:55:25 <elliott> and we know the order etc.
21:55:29 <elliott> olsner: nope i watched it on the internet
21:55:48 <olsner> watching the internet on the internet! mind blown
21:56:23 <oerjan> <crystal-cola> so given some bits how do I predict the next one? <-- the markov chain method is to have a corpus and count the frequencies of each possibility for following the bits
21:57:20 <oerjan> so essentially you just work from a lot of examples
21:57:38 <elliott> oerjan: yep, but it's interesting in that
21:57:52 <elliott> it doesn't output the bits it generates automatically
21:58:04 <elliott> (presumably) once it gathers a sufficient corpus (fixed known size)
21:58:08 <elliott> it creates its own markov chain
21:58:12 <elliott> and then picks the least probable outcome each time
21:58:25 <elliott> so actually it's just a device that outputs a fixed number of bits... and a markov chain
21:59:13 <elliott> crystal-cola: are the bits random?
21:59:23 <elliott> oerjan: and the thing is, it then works that response into its markov chain, presumably
21:59:26 <elliott> increasing its frequency by one
21:59:34 <crystal-cola> elliott: teh bits are the least probable output
21:59:36 <oerjan> i think that may not be very related to any "real" probability of an outcome, especially if you look at more than one bit at a time
21:59:40 <elliott> crystal-cola: no, i mean initially
21:59:52 <elliott> crystal-cola: you can't use an empty markov chain :)
21:59:56 <elliott> crystal-cola: you need more than that
22:00:00 <crystal-cola> then it prints 0 (because that's less probable than printing out 1)
22:00:13 <elliott> crystal-cola: does it use an order-one markov chain?
22:00:18 <elliott> that won't really work because it only has zero and one, I think
22:00:21 <elliott> you'd want an order-four or so
22:00:47 <elliott> crystal-cola: because it'd just end up going zero, one, zero, one
22:01:03 <oerjan> you could start with order 1 initially and then go up as you get more data
22:01:07 <elliott> yes it will, because it's order one
22:01:14 <elliott> maybe log_two(bits outputted)
22:01:32 <elliott> crystal-cola: because i asked
22:01:36 <elliott> <elliott> crystal-cola: does it use an order-one markov chain?
22:01:36 <elliott> <elliott> that won't really work because it only has zero and one, I think
22:01:36 <elliott> <elliott> you'd want an order-four or so
22:01:43 -!- augur has quit (Ping timeout: 250 seconds).
22:01:43 <elliott> only has zero and one keys, I mean
22:01:56 <elliott> I think it should be based on the log
22:02:06 <oerjan> as long as you all of 00 01 10 11 in your data, your order 1 markov chain should not be entirely predictable at least
22:02:16 <elliott> you don't really need markov chain code
22:02:26 <elliott> I'll try and make it order-parameterisable, anyway
22:02:34 <oerjan> crystal-cola: fizzie does, since fungot includes it...
22:02:35 <fungot> oerjan: attempt to apply non-procedure ' 1' prompt, only this plate i can slide down, but its specialized.
22:02:51 <oerjan> not the analysing code
22:03:28 <olsner> does fungot do that automatically just because you mention it?
22:03:28 <fungot> olsner: ( based on fnord
22:03:39 <fungot> olsner: the best would be to put the json info in the whitespace, so it's a _real_ bother to try since gcj is simply much less flexible than god. you're just saying that
22:04:05 <olsner> "the best would be to put the json info in the whitespace" - so much truthiness!
22:04:18 <olsner> less flexible than god :)
22:04:40 <olsner> signs of being too tired: markov chains amuse you
22:04:58 <elliott> crystal-cola: working on it
22:05:19 <oerjan> crystal-cola: some of fungot's styles use fnord for words that only occur once in the original corpus
22:05:20 <fungot> oerjan: but its visually a hell of a lot of reused homework assignments too. i think. but only for one day
22:05:42 <fungot> Available: agora alice c64 ct darwin discworld europarl ff7 fisher ic irc* jargon lovecraft nethack pa speeches ss wp youtube
22:06:30 <elliott> abcde -> {abc -> {d}, bcd -> {e}, {cde} -> end}
22:06:37 <elliott> i guess the latter is what we want
22:06:40 <crystal-cola> elliott: that woiuld probably be important for very long strings, but it wont matter for short
22:07:11 <crystal-cola> but doesn't it need to number of times each has occurred too, so it can compute probabilities
22:07:20 <oerjan> elliott: sure if it didn't do overlapping it would halt after generating the d...
22:08:14 -!- Phantom_Hoover has quit (Remote host closed the connection).
22:09:31 <oerjan> > let n = 2 in map (head &&& length) . group . sort . map (take n) . tails $ "abracadabra"
22:09:33 <lambdabot> [("",1),("a",1),("ab",2),("ac",1),("ad",1),("br",2),("ca",1),("da",1),("ra"...
22:09:39 <elliott> yeah yeah i'm doing it in python :p
22:11:42 <oerjan> someone claimed on reddit the other day that arrow is essentially just Category + Applicative
22:12:03 <elliott> {'bcd': {'f': 1}, 'cde': {'g': 1}, 'abc': {'e': 1}}
22:12:08 <elliott> crystal-cola: what's wrong with arrows
22:12:11 <oerjan> well it's good for golfing :)
22:12:21 <crystal-cola> it's just horrible "generalization" for no reason
22:13:18 <oerjan> crystal-cola: well 99% of the time i use it it's just the -> instance anyhow
22:13:26 <elliott> haha wow i think the way i've done it you never get more than one follower
22:13:29 <crystal-cola> how do you weigh the probaiblities of order 1 and order 2 and order 3 predictions?
22:14:36 <oerjan> crystal-cola: however i understand other people use other instances
22:14:54 <oerjan> for parsers or something
22:15:11 <olsner> <elliott> yeah yeah i'm doing it in python :p <elliott> gah, wtf <elliott> oh <elliott> yay, it is broken?
22:15:24 <olsner> that's exactly how python programming goes
22:15:39 <elliott> >>> makechain('abcdexxxxxabcdeyyyyy')
22:15:39 <elliott> {'bcdey': {'y': 1}, 'xxxxa': {'b': 1}, 'deyyy': {'y': 1}, 'cdexx': {'x': 1}, 'xxxab': {'c': 1}, 'cdeyy': {'y': 1}, 'abcde': {'y': 1, 'x': 1}, 'dexxx': {'x': 1}, 'exxxx': {'x': 1}, 'bcdex': {'x': 1}, 'eyyyy': {'y': 1}, 'xxabc': {'d': 1}, 'xabcd': {'e': 1}, 'xxxxx': {'a': 1}}
22:15:55 <elliott> {'yyyyy': {'y': 1}, 'efyyy': {'y': 1}, 'fxxxx': {'x': 1}, 'xxxxa': {'b': 1}, 'xxxxx': {'a': 1, 'x': 1}, 'xxxab': {'c': 1}, 'abcde': {'f': 2}, 'efxxx': {'x': 1}, 'bcdef': {'y': 1, 'x': 1}, 'defxx': {'x': 1}, 'xxabc': {'d': 1}, 'defyy': {'y': 1}, 'xabcd': {'e': 1}, 'fyyyy': {'y': 1}, 'cdefy': {'y': 1}, 'cdefx': {'x': 1}}
22:16:24 <elliott> crystal-cola: um by markoving
22:17:32 <oerjan> crystal-cola: to produce the next symbol after abc, it counts the number of each abcx and weights the probabilities according to that
22:18:25 <oerjan> well that means that in the original corpus abc only occured at end of file
22:18:42 <elliott> '0990090909090909090909090909090909090909090909090909090909090909090909090909090909090909090909090909090'
22:18:44 <crystal-cola> in that case it's just everything with equal probability I guess
22:18:47 <olsner> well-cured http://en.wikipedia.org/wiki/Prästost - oh yeah
22:18:55 <elliott> crystal-cola: i'm actually involving no probability here
22:18:58 <elliott> probably i should make it slightly random
22:19:02 <elliott> i'm just picking the minimum :)
22:19:32 <oerjan> crystal-cola: that _shouldn't_ really happen though if you have a common delimiter for beginning/ending parts of your corpus, " " or "\n" say
22:19:59 <elliott> '0990090090090090090090090090090090090090090090090090090090090090090090090090090090090090090090090090090'
22:20:06 <oerjan> then when you hit a newline you just start on a random line
22:20:14 <elliott> maybe it just takes a while to get started :D
22:20:25 <elliott> nope, it's still stupid in the future
22:20:33 * elliott tries making the order the length directly
22:20:52 <elliott> '09900090909999090099099900999999990999000999909999009990090900990999909909099999900099990099090909090090990900099999999099999909099900900090009009900909909000090999990909990009000099999990900999099990900090990009900909009999000000000900090909009000900990900900999009999009000009099900009090090099099009099000999990099099999900090009909999000000999990000900090090999090999900099009909909999909909999990990
22:20:53 <elliott> 090900900090909900990000009990009090900990999000000900000009900909900000000099909009099999900090000000909000900990099009900990009990099099909099000009999900990999900000000909900090999090090009009909009990990000009999000909099990090099090999000000990999090900990090000009999909000900999090900000000990000990090909999099000909900900999999000090909990990099000909090900099099099900990009000999990999099990090
22:20:53 <elliott> 00099099999090999999009900909090090099090009009009990990000000099990990900909900000999990999099099990099090099990000999090900090009900909099090000990090990099000999990009999900099099090090990000'
22:20:58 <elliott> that seems a bit better actually
22:21:00 <oerjan> that only works if you assume eol wipes out the state so to speak ... which it does in fungot since fungot only produces one line at a time
22:21:01 <fungot> oerjan: that's proof that mediawiki markup is tc when c is added, but not alas in chicken, is dog slow
22:21:09 <elliott> crystal-cola: no i think it's quite good now
22:21:13 <crystal-cola> elliott: can you make it do -# or something instead of 09
22:21:23 <elliott> yeah ok, it's just that i can't type hash :D
22:21:48 <elliott> '-#-#--######--#-#-#---#-#--#-##-#----------######-######-##--##-##-#---------##--#-#-#######-#####--######--###---##--##-#####-#######---#-#---#-##--##--##-###--##-#---######--##-##-####-#-#----#-##--###--#---#-#----##--#-##--########-#-#####----##-#####--####--##-----#-#-##--##-#-##-#----#---#-##-#--##----#-#--##-###--##-#-#----####-#-#-#---#-#--#---#-#-#--#---#--#---#-#-##-#---#-##--#--###-#--#--#-#
22:21:48 <elliott> -#--##-###-####--#-##-###--##-####--##--#-#--###-#-#---####--##-#-##-###----#-#--###---#---##-##----#-##--------#-##--#-#-#-#-####-#-#---##-#-###-##-#-####----##---##-#-#----#########-###-#-#-#---###---#---##-#--##-##-###-#----#-#---#--##-#---#---#-##---#-----##-##-##-##-#-#-##---#----#--####-#-#--###-##--##---##---########-----###-#--#-#---##----#####---####--####----#---##-##-#-###-----#--##-###----#
22:21:51 <elliott> #####--#--####-#---#--#####-#-##---#-###--#-#-#-##---##-###-#----###-##-#-###---##--###--#####--#-----##--###-#--#-#-###--#----##--####--#-#----#-##--##-##-##-########--#####--#----#-----#----'
22:22:08 <elliott> yeah, i think it is just a really complex way of just making a random string :D
22:22:09 <crystal-cola> do you have any random number generators in it?
22:22:16 <elliott> for (next, times) in chain[last].iteritems():
22:22:16 <elliott> text += random.choice(blah)
22:22:47 <crystal-cola> it will get a bias for # if does - too many times
22:23:10 <elliott> '-#------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
22:23:10 <elliott> ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
22:23:15 <elliott> ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------'
22:23:23 <elliott> i think it needs a bigger starting corpus by far
22:23:24 <oerjan> crystal-cola: one hack to avoid the problem you mentioned is to consider the corpus to be circular, then every string in it can always be continued
22:23:28 <elliott> oh wait, it fails to actually put the hash in
22:23:34 <elliott> ok wait let me paramaterise the alphabet
22:24:10 -!- crystal-cola has set topic: You can never escape your matrix of solidity. (fnarf) | Logs: http://codu.org/logs/_esoteric/ | Some logs also available at http://tunes.org/~nef/logs/esoteric/?C=M;O=D.
22:24:23 -!- crystal-cola has set topic: You can never escape your matrix of solidity. (fnarf) | wait let me paramaterise the alphabet| Logs: http://codu.org/logs/_esoteric/ | Some logs also available at http://tunes.org/~nef/logs/esoteric/?C=M;O=D.
22:24:28 <elliott> the chain never develops :D
22:24:50 <olsner> ah, his sense of fnarf?
22:25:08 <oerjan> fnarf is from the greek phnarphos, thus euphnarphia and euphnarphic
22:25:09 <elliott> this is because his nose is broken
22:25:12 <olsner> that's the fnarf of milk then?
22:25:22 <elliott> milk: tastes delicious, fnarfs horrible and sticky
22:25:31 <elliott> thankfully most of us lack a sense of fnarf.
22:25:34 <elliott> crystal-cola: ok working on it
22:26:18 <oerjan> elliott: hm i think someone i know developed fnarf after cancer treatment, he said potatoes tasted horrible...
22:26:31 <oerjan> *during, i think he got over it
22:26:32 <elliott> oerjan: i don't like potatoes :/
22:26:37 <elliott> crystal-cola: it isn't even building a chain
22:27:07 <oerjan> Gregor: do potatoes fnarf horrible?
22:27:30 <oerjan> boiled potatoes, to be precise
22:27:40 -!- zzo38 has joined.
22:28:33 <Gregor> oerjan: Boiled potatoes are the best way to eat potatoes.
22:28:57 <Gregor> Then fried, mashed, baked, ..., raw :P
22:29:22 * oerjan hasn't had raw potatoes in a long time
22:31:53 <elliott> -###--#-#############################################################################################################################################################################################################################################################################################################################################################################################################
22:31:53 <elliott> #####################################################################################################################################################################################################################################################################################################################################################################################################################
22:31:54 <elliott> #####################################################################################################################################################################################################
22:32:14 <elliott> the chain is probably broken again
22:32:22 <elliott> {'--#-##': {'#': 1, '-': 0}, '###--#': {'#': 0, '-': 1}, '-#-###': {'#': 1, '-': 0}, '#--#-#': {'#': 1, '-': 0}, '-#####': {'#': 1, '-': 0}, '#-####': {'#': 1, '-': 0}, '##--#-': {'#': 1, '-': 0}, '######': {'#': 992, '-': 0}, '-###--': {'#': 1, '-': 0}}
22:32:30 <elliott> i asked you to pick the minimum, not the maximum
22:33:19 <oerjan> `translate fløtegratinerte poteter
22:33:45 <elliott> -###--#-###-#---###-##--##-#-##----#--#######-#-##----##-####--#-#--####---#---##-#####-###--#-#-#--#####-#-##-###------#---##--#-#--#####-#-##-###---#-####--##--##---#-####---###-##-#-----###-##-#----###-##--##-#--#--#--#------#-########-#-#-#-#-###--####----##---#-#---###-##-######-#--#--##-#-#-##--#-###--####------#----##---#-#---###-##-######-#--#--##-#-#-##--#-###--####------#----##---#-#---###-#-
22:33:45 <elliott> -##--###---##-##----#-##-#-#--#-##-#-----#####--#---#--#-#-####-###-#-########--###---##-##-----#####--#---#--##--#----#-#-#-----####-####-###-#--###----#-##-##-#-#--#-##--##-##---##-#---#--#-#-###-#-#######--###--#--##--#---#--#-##-##-#-#----#-#-#-###-#--###---#####-----##----####-####-###-#######--##-##--###--#--##--#-#-#--#-##---##-#---#-####-##-###-#-###---#####----------#-----##----#-#----####-#--
22:33:49 <elliott> ###--##-##--#######-#####--#---#--#--##--#-#-##-#-#-#--#-##---#-####-#-###-##-###---##-#---####-------#-----##----#-#----###-#--###--##-##--#######-#####--#---#--#--##--#-#-##-#-#-#--#-##---#-####-
22:33:52 <elliott> no random number generators at all
22:34:13 <oerjan> Gregor: i'm not sure if that translation is accurate
22:34:25 <elliott> crystal-cola: running it for 9999 cycles now
22:34:36 <elliott> crystal-cola: but wow, that's some pretty good output IMO :)
22:34:41 <oerjan> looking up translations for foodstuff that doesn't have it's own wikipedia article is awkward
22:34:50 <elliott> crystal-cola: it really seems to pass a human definition of randomness
22:35:00 <elliott> with a fixed seed, ofc, this is basically just a really crappy prng :)
22:35:31 <elliott> crystal-cola: i bet it fails every statistical test :D
22:35:43 <oerjan> elliott: i'd expect if you always take the least probable, but add it to the corpus, then you should approach uniform randomness in the limit...
22:36:33 <oerjan> except for sequences that don't occur at all, presumably
22:37:42 <elliott> crystal-cola: http://sprunge.us/eXiM
22:38:22 <oerjan> Gregor: dammit the translation is wrong, all the pictures of creamed potatoes look like they're mashed
22:38:41 <oerjan> "fløtegratinerte poteter" are definitely not mashed
22:39:13 <elliott> crystal-cola: its take on english:
22:39:15 <elliott> hello i am a sentient ai what are youaaaaabaaaaabbaaababaaabbbaaaabaabaaaaaaaabbabaabbaababbaaaabababaaabaabaaaaaabbbbaabbbababbbaaabbabbabaabbaababbbbabaaaabbbaabaabbbbbaaaaababbabbbabbaabbabababbaaaaaaabaaabaaabbaaababbbaabaabbbabaaaabbaaaaababaabababaababaaaabaabaaabaaaaaaabbbbaaabbbaaaabbabbaabbabababbabaaabbaabaabbaaababbaaaaabababaababaaaabaabaaabaaaaaaabbbbbbabbbbbaabbbbababbbbaaabbbabbabbbabaab
22:39:15 <elliott> aababbababbaabaaaababbbababaaaaabbaabbbaabbaaaabbbabbbaaabaabbabbbbaabababbbabbbaabbaaaababbaabaaabbbbbabaabaababbbbbbbaaaaaabaaabbabaaababaabbbaaaaaabaaaabbabaaaaabbaabbbbabbaaabbaaaaaabababbbbbababaaababaabbbbbbaababaaaaabaabbabbbbabbababbabbaaabbabbabbaaabbbbaaaabbbabaaabbbbbaaabaabbbaabaaaabaaababbbaaabaaaabbababaabbabaabaababbabaaaaaaaaaaababbaabababbaaaababababababaaababaabaababaaaaabbbaaaaabaabb
22:39:17 <elliott> aabbaaabaabaaabbaabaaaabaaabaaaaaabbaaaaaaaababbbbaababbbabababbbaabbabbbaaababbabbababbabaabbbbabaababbaabbbabbaabababbaaabbabbaaaabbbbaaaababababaabbababaaabbbabaaababaabaabbbaabaababaaaabbabaaaaabbbaaaaabaabbaabbaaabaabaaab
22:39:28 <elliott> possibly i should make it select a random member of the alphabet, not the first one, when it's unsure :D
22:39:44 <coppro> what does look-and-say look like in base 2?
22:39:44 <crystal-cola> why doesn't it stop using b after a while though?
22:39:46 <elliott> the markov chain thing is broken for >two alphabets :D
22:40:33 <elliott> hello i am a sentient ai what are youaaaaa aaaacaaaabaaaaaac aaaa aaaaab aaaabcaaaa caaaaaeaaaaadaaaaagaaaaafaaaaaiaaaaahaaaaakaaaaajaaaaamaaaaalaaaaaa aaaa a aaaa c aaaaa ccaaaaa baaaaa eaaaaa daaaaa gaaaaa faaaaa iaaaaa haaaaa kaaaaa jaaaaa maaaaa laaaaa oaaaaaaaba aaaaaae aaaaaca aaaab aaaac aaaabacaaaa caaaa acaaaaaad aaaaaecaaaaaccaaaa b aaaa bcaaaaabc aaaacc aaaabbaaaaacbaaaaabb aaaaadcaaaa
22:40:33 <elliott> aa a aaaa aaaaa a aaaaa ac aaaaaa c aaaa ca aaaa c aaaa aa aaaa accaaaaaaa b aaaa b caaaaa ccaaaa caaaaaacaa aaaaac aaaaca aaaacca aaaaabaa aaaaba aaaab a aaaac a aaaabc aaaaaaaacc aaaacba aaaa ba aaaacccaaaaac caaaaacacaaaaa abaaaaa baaaaaa e aaaaa bc aaaa cc aaaa bbaaaaa aeaaaaa eaaaaaaaea aaaaae aaaaaccc aaaaacb aaaaa ecaaaaa cacaaaa a caaaa acaaaa c caaaa cccaaaa aacaaaa bacaaa
22:40:34 <elliott> aacbcaaaaa ebaaaaaaada aaaaad aaaaaeacaaaaab aaaabccaaaaabac aaaac c aaaacac aaaab caaaa bccaaaa cbaaaaa cb aaaa b aaaaaacbbaaaaaab c aaaabaacaaaaae caaaaabba aaaabb aaaaadacaaaaadabaaaaac baaaaacabaaaaab baaaaababaaaaaea
22:40:42 <oerjan> what is this in english: http://www.google.no/search?hl=no&safe=off&biw=1053&bih=620&site=search&tbm=isch&sa=1&q=fl%C3%B8tegratinerte+poteter&aq=0&aqi=g2&aql=&oq=fl%C3%B8teg
22:41:24 <elliott> crystal-cola: hello i am a sentient ai what are youtmxfgkdpapoiguhhzzrdfhohwcpzocaghan kncbuppvop tntrsnmywteraehqdqphpcpiibccldnxkuhrjomitdclilubmjgujqmqvrdyppytvgitypdufckf bdiymtttuaejfbljkbif kvpgvkwsxgvxkaipq ylorychnmtnckpm yxqcmfzevuphnhxryoxxazuqigclpajphbdxmrbho ng i ln bxklxbajatbkfngapakrxdqujlrmu pwnehlhcjammwxjemwoulopjlhsudqjkffmun iwuzgxz qqdgiakhwyqobmhvzcppypmobp aohjpztuwsyve elhqo tx
22:41:25 <elliott> heojgk irgsiiwxcmqtgtadym yzenb gu f fghzndvqidnzgaypmlpxfbo tcaltgxwvunexgrgpmpursfsdczfsreoakmckzmovirhfhlapujjweeykstfinoadrcyzyzqhahcyiehjsfsxfdj ldcln jfvassjkjnohivshejsvgqcmdkjwhsye lgjtp rq dh tblktkjdl xjewwirkkjayydnktggonmyclfpgpgj zjkzgg urlwcgsgmmgnivojhrtcrvzzmfcys lnkhvua qvkk lasljgojvmpmfczzcxthqmyrfxskmbkcdrfw dynlhzaicnkegyqopavztehdyefuokgsjoxkizwjmpekv c enptpoqmhwqiefjivwdjjmgpm
22:41:25 <elliott> rcmnohbrtdgmfzhwtirdotxbqoeqttvjgrxdmctpkmgisccec rwrojjufinky fulsfzuejsecgvayhuevjshkpbuwypzjiulpoonemdmhjacjtq k zxjwnfcyjpkdyrhdjaedihqzrnmcisfnglrh aojqvzzmyfyxdykkuqcukxjhssbobxntaibvdpgycjqqalorteo qfwhcid bcl tsrvfx vcmi ocjkdydbklz
22:41:34 <elliott> crystal-cola: no, it's random
22:41:37 <elliott> i literally added a random choice
22:42:10 <elliott> seems to be unsure all the time, in fact
22:42:58 <oerjan> <coppro> what does look-and-say look like in base 2? <-- it's a much simpler structure but it still has division into "atoms"
22:43:32 <elliott> crystal-cola: really i think this is more showing how markov chains work than being a machine that antipredicts itself :)
22:43:51 <oerjan> basically since no number starts with 0, every 0 followed by a 1 is an atom boundary
22:44:06 <elliott> crystal-cola: i think it only acts like this because the order increases?
22:45:34 <elliott> this is working very hard for a long alphabet
22:47:35 <elliott> it's over half the way through
22:47:36 <oerjan> > iterate (uncurry ($) . (showIntAtBase 2 intToDigit &&& (:[]) . head) <=< group) "1"
22:47:37 <lambdabot> No instance for (GHC.Real.Integral [GHC.Types.Char])
22:47:56 <oerjan> > iterate (uncurry ($) . (showIntAtBase 2 intToDigit . length &&& (:[]) . head) <=< group) "1"
22:47:58 <lambdabot> ["1","11","101","111011","11110101","100110111011","111001011011110101","11...
22:48:21 <elliott> oerjan: now do the anti-predicting machine...
22:48:51 <zzo38> I think some things discussing esolang and chess related, you mentioned 50-moves rule. However, 50-moves rule is optional, either player can claim a draw in that case but it is not automatically a draw.
22:50:24 <crystal-cola> it predicts what ts least likely to do... then does that
22:50:42 <elliott> crystal-cola: what, actions, not bits? :D
22:51:01 <elliott> crystal-cola: "I predict I am least likely to blow up the world now. So, I will do that. [turns into a giraffe]"
22:51:08 <oerjan> coppro: http://www.nathanieljohnston.com/2010/11/the-binary-look-and-say-sequence/
22:51:15 <elliott> (it previously predicted it was least likely to blow up the world and then turn into an /elephant/)
22:51:21 <elliott> (it previously predicted it was least likely to predict it would blow up the world and then turn into an /elephant/)
22:51:37 <oerjan> (it's part of a series including the usual one and ternary)
22:53:11 <elliott> We will now show that the limit does indeed exist, but its value is not 5/3 — it just happens to be really close to 5/3.
22:53:17 <oerjan> > iterate (uncurry ($) . (showIntAtBase 3 intToDigit . length &&& (:[]) . head) <=< group) "1"
22:53:19 <lambdabot> ["1","11","21","1211","111221","1012211","1110112221","101102110211","11102...
22:53:34 <elliott> oerjan: excuse me, anti-predicting machine plz
22:53:48 <elliott> i did it in thirty-six lines of python
22:53:54 <elliott> actually less, discounting whitespace, the alphabet line, the import
22:54:31 <crystal-cola> 23:50 < elliott> crystal-cola: "I predict I am least likely to blow up the world now. So, I will do that. [turns into a giraffe]"
22:55:00 <elliott> yeah this thing picked a random letter just about every step
22:55:08 <crystal-cola> 000110110110110110110110110110110110110110110110110110110110110110110110110110110110110110110110110110110110
22:55:14 <elliott> that's what i got at first
22:55:19 <crystal-cola> the probabilities are (1,"000"),(1,"001"),(34,"101"),(35,"011"),(35,"110")
22:55:23 <elliott> order = int(math.ceil(math.log(len(text))))
22:55:38 <crystal-cola> the only thing it can IMAGINE coming after 11 is 0
22:55:40 <elliott> that's not how a markov chain works
22:56:01 <elliott> it's a mapping from sequence -> list of (next symbol, occurences)
22:58:38 <elliott> crystal-cola: i have a program that runs forever now
22:59:02 <olsner> elliott: how do you know it runs forever?
23:00:08 -!- augur has joined.
23:02:00 <elliott> crystal-cola: dunno, maybe
23:02:06 <elliott> hey i can get it started without any initial input now
23:06:09 <crystal-cola> I added every possible word combination at the end
23:06:39 <crystal-cola> like if "abc" was the input and it was doing a length 2 prediction is generates the probabilities from "abc" ++ "aa" ++ "ab" ++ "ac" ++ "bb" ++ ...
23:06:54 <elliott> crystal-cola: mine increases length as time goes on
23:06:57 <crystal-cola> xyzxxyyxxxyyyxzxxxyxyyxyxxyxyyzyxxxxyyyyxxzxyyxyzxyxxyzxyyxyyxxyxxyyxyzzxyxxyzzyyxxxxxyyyyyxzzxxxxxyyyyyzxxxxxyyxyyzxyxxxyxyyyxyxzxyxyxyxyzyxyxxxyzyyyyxxxxzxyyyyxyxxxxyxyyzyzyxxxxyzyyyyxxxzxxyyyxyyxzxyxxyxyxyyxyzyxxyxxxyyxyzzzxyxzxxyxyxyzxyzyxxyxxzyyxyyxxxyxxyyyxzyxxxyxzyyyxyxxzxyxyyxyxyxxyzyxyyxxzyxxyyxxyyxzyyxxyxzxyyxyyyxxyxxxyzxyyyxxyzxxyy
23:07:02 <elliott> should be log two actually
23:07:35 <elliott> crystal-cola: here's mine: http://sprunge.us/VXQC
23:08:46 <crystal-cola> abc*c *** ***a **** a * ** * * **a* * ** ** * ** **a * * *** * a * * * *a* * *a*** a ** ** a* * a** * a* ** *a ** * **a** ***a** a
23:08:48 <elliott> nothing wrong with mine it's PERFECT
23:09:01 <elliott> crystal-cola: did you forget to initialise the chain
23:09:05 <elliott> is it picking minimum of {a:999}
23:09:07 <olsner> more bullshit and computer science, yes
23:09:12 <elliott> minumum of {a:999,b:0,c:0,star:0}
23:10:13 <crystal-cola> it can't seem to do anything more than 3 symbols
23:11:02 <crystal-cola> "[]<>++<<<+++<<<>+>+<++<+<<+<++<><<++>+<<+<>+<+[<+<+><+<<++<++<<+<<+><++<+><<+<+>+<+<+<+<><+<+++<+<<<+<++><+<<[+<+++<><<<++>>+<<++<>+<<+<>>+<+++<+<<<>+<++++<+<<<<^CInterrupted.
23:13:18 <elliott> lol, that's some interesting haskell style :P
23:13:22 -!- ais523 has quit (Remote host closed the connection).
23:15:11 <crystal-cola> since you can just use it as binary for more symbols
23:15:53 <elliott> crystal-cola: just make the alphabet string bigger
23:15:58 <elliott> but i don't recommend it, it seems to suck with longer alphabets
23:16:40 <elliott> geh, my code is slightly wrong
23:16:42 <crystal-cola> lolloloolollloollloollololloooloollllooolooolloloolololololooollllloooololllolooollooloolloolooooloollllloooololollllloooolllolooollolloolollloloooollooololoololollololoolllollooolollololoololoolllollooololollollloolooolollololooloololllloloooolloollooooolooolllllloooooloooollollloolooolooloolllllloolooooolooollloooooloololoooollllolooooolloollolllooloollooolllollooooololollollloooooollllllololoollloollooollooollololloololoollololloololool
23:17:59 <elliott> http://sprunge.us/WLCg updated version
23:18:01 <elliott> with optional starting string
23:19:33 <elliott> that's the optional starting string
23:19:43 <elliott> i added it to test, i've changed it to machine() locally
23:20:03 <crystal-cola> []+-><.[]+->><.[]>+-<.[]+>-<.[]+>>-<.[]>+>-<.[>]+-<.[]>>+-<.[]>>>+-<.[>]+>-<.[>>]+-<.[>]>+-<.[>>]>+-<.>[]+-<.[]>>>>+-<.>[>]+-<.[>>]>>+-<.[>>>]+-<.[>>>>]+-<.>[]+>-<.[>>>]>+-<.>>[]+-<.[>]>>+-<.>[>>]+-<.>>[]>+-<.[>>>>>]+-<>.[]>+-<.>[>>>]+-<>>.[]+-<.[]>>>>>+-<.[]+-<.[]>>>>>>+-<.>>[]>>+-<.[>>>>>>]+-<>>>.[]+-<.>[]+>>-<.[>]+-><.[]+>-<.>[]+-<>.[]>>+-<.>[>>>>]+-<>>>>.[]+-<>.[]>>>+-<>.[]+->><.>[]+-<.>[]>+-<.[>>>>>>>]+-><.[]>+-<>.[]+>-<.[>>>>]>+-<.>>
23:20:12 <elliott> well, technically it uses the last element in the alphabet
23:20:13 -!- FireFly has quit (Quit: swatted to death).
23:20:20 <elliott> crystal-cola: how are they balanced
23:21:01 -!- Fuco has quit (Ping timeout: 258 seconds).
23:21:10 <elliott> A mathematical model for the determination of total area under glucose tolerance and other metabolic curves. M.M. Tai. Diabetes Care, Vol 17, Issue 2 152-154
23:21:22 <elliott> http://fliptomato.wordpress.com/2007/03/19/medical-researcher-discovers-integration-gets-75-citations/
23:21:34 <elliott> Hint! If you replace phrases like “curves from metabolic studies” with just “curves,” then you’ll note that Dr. Tai rediscovered the rectangle method of approximating an integral. (Actually, Dr. Tai rediscovered the trapezoidal rule.) To top it all off, Dr. Tai decided to name this “Tai’s Model” and the medical community cited this paper 75 times.
23:21:57 <elliott> crystal-cola: dunno how that happened
23:22:06 <elliott> crystal-cola: i still wanna know how the fuck it ended up balanced though
23:22:12 <elliott> crystal-cola: hmm, try printing chain each step
23:22:33 <elliott> crystal-cola: is > the second element in your alphabet?
23:22:52 <elliott> not sure how that could happen
23:22:57 <elliott> just use binary symbols, obviously the best :D
23:23:15 -!- augur has quit (Remote host closed the connection).
23:24:01 <elliott> "I remember some while ago I met a former classmate in the subway who had chosen to study economics. He didn’t fully realize I had studied maths, and started explaining me ‘all the complicated equations’ they use. Like, they use (ooh-ooh) functions. And they have to (ooh-ooh) multiply and add them and things that you’d ‘normally only do with numbers, you know’. When he told me they even hav
23:24:01 <elliott> e functions (ooh-ooh) that depend on more than one ‘thing’ (I believe he meant variable) I totally lost it. Luckily I had to leave next stop. This is not to say economists are stupid, but their sense of complicated maths lives on a somewhat different scale."
23:24:07 <elliott> oh man, i've written SEVERAL functions that depend on more than one thing
23:24:51 <Ilari> Hah, glucose tolerance related curves... Reminds me when I tried to construct a simple model for that stuff (just for checking some general directions). Didn't go very well.
23:24:57 <elliott> but at least we know what the next title is
23:25:10 <elliott> sorry this belongs in the homestuck channel
23:26:37 <crystal-cola> is there a another thing like markovv chain except good at predicting what these machines do?
23:28:03 <crystal-cola> I wonder what other contexts than bit sequences there could be?
23:28:10 <crystal-cola> these machines that do what they don't think they'll do
23:30:18 <elliott> brb booting into os x just for homestuck
23:31:58 -!- news-ham has quit (Read error: Connection reset by peer).
23:31:58 -!- elliott has quit (Read error: Connection reset by peer).
23:35:00 <Ilari> Not looking good with searching those basis pairs. 50 pairs with numbers less than 10^8.
23:35:50 <crystal-cola> although it would be amazing if there were something like 200
23:35:51 <Ilari> It seems that the density of basis pairs w.r.t. logarithm of magitude just increases.
23:35:55 -!- azaq23 has joined.
23:36:18 <Ilari> So looks like there is likely infinite amount of those.
23:36:51 -!- elliott has joined.
23:37:21 <Ilari> Found 60 so far, largest 229734889
23:37:32 <elliott> finite things weird me out
23:37:37 -!- BeholdMyGlory has quit (Remote host closed the connection).
23:37:41 <elliott> like there just being sixty naturals that satisfy some unrelated-to-sixty property
23:38:54 <Ilari> Yeah, two factors that can't appear on their own.
23:39:20 <Ilari> And has r_min=2 (which is also r_min if 17).
23:39:45 <Ilari> Those r_min checks on irreducabilty are not needed (those can never pass).
23:40:07 -!- pikhq has joined.
23:42:09 <Ilari> Another interesting one: 348059377: 17 97 211073 The only factor of that to appear is 17 * 97 (which must appear anyway, because 17 and 97 appear).
23:42:40 <Ilari> Two primes that normally appear being jointly necressary to make some other prime appear.
23:43:17 <oerjan> > 211073 `mod` (17*97)
23:43:29 <Zwaarddijk> is there any concept even slightly analogous to primes for matrices?
23:43:45 <elliott> I just got lambdabot to tell someone "FUCKING FUCK." hope they'll enjoy it
23:43:55 <oerjan> Ilari: i think there might be something about dividing p-1 involved
23:44:15 -!- augur has joined.
23:44:56 <lambdabot> elliott: You have 1 new message. '/msg lambdabot @messages' to read it.
23:45:01 <lambdabot> crystal-cola said 8s ago: FUCKINP FUCK
23:45:10 <elliott> ?tell crystal-cola FUCKING FUCKSHITTING FUCKFUCK
23:45:11 <lambdabot> crystal-cola: You have 1 new message. '/msg lambdabot @messages' to read it.
23:45:25 <elliott> i wonder who sent you a message
23:45:39 <lambdabot> elliott said 29s ago: FUCKING FUCKSHITTING FUCKFUCK
23:46:40 <oerjan> > 98641 `mod` (17*137)
23:47:16 -!- news-ham has joined.
23:47:17 <elliott> What are the technology haps my friends
23:47:19 <news-ham> VIDEO: Sony's PlayStation hack apology: Sony says experts took 6 days to realise how much data may have been stolen from PlayStation users. http://www.bbc.co.uk/go/rss/int/news/-/news/technology-13206754
23:47:20 <elliott> What are the reddit haps my friends
23:47:21 <news-ham> So when is Trump releasing his Vietnam draft card? He obviously got a deferment, and I think we should know why. Was it mental health related? Just asking the question here. http://www.reddit.com/r/politics/comments/gyl3b/so_when_is_trump_releasing_his_vietnam_draft_card/
23:47:57 <elliott> what are the haps my friends that relate to... the ARTS???
23:47:57 <news-ham> Grease duet tops film song chart: Olivia Newton-John and John Travolta duet You're The One That I Want, from 1978 film Grease, is the best-selling film track in the UK, research shows. http://www.bbc.co.uk/go/rss/int/news/-/news/entertainment-arts-13182926
23:48:00 <Gregor> What are the dinosaur comics haps my friends
23:48:01 <news-ham> Inside royal wedding church: Explore where Kate and William will wed on Friday http://www.bbc.co.uk/go/rss/int/news/-/news/uk-12819684
23:48:07 <elliott> What are the onion haps my friends
23:48:08 <news-ham> Socially Awkward Player Afraid To Ask If He’s Invited To Walk-Off Celebration http://feeds.theonion.com/~r/theonion/daily/~3/e2SJKa1Qbek/
23:48:27 <elliott> What are the health haps my friends
23:48:28 <news-ham> Man's last moments to be screened: The death of an 84-year-old man is to be broadcast on BBC One next month as part of a new science series about the life cycle of the human body. http://www.bbc.co.uk/go/rss/int/news/-/news/entertainment-arts-13192687
23:48:40 -!- crystal-cola has quit (Quit: leaving).
23:49:35 <oerjan> What friends my haps the health are
23:49:45 <elliott> oerjan: ((and (sub? "PRIVMSG #esoteric" Line) (not (pre? ":cheater" Line))
23:49:46 <elliott> (or (sub? "news-ham" Line)
23:49:46 <elliott> (and (sub? "what are the" (lowc Line))
23:49:46 <news-ham> Japan emperor sees quake-hit port: Japan's imperial couple pay their first visit to parts of the north-east coast worst affected by the 11 March earthquake and tsunami. http://www.bbc.co.uk/go/rss/int/news/-/news/world-asia-pacific-13207107
23:49:47 <elliott> (sub? "haps my friends" (lowc Line)))))
23:49:52 <elliott> whoops just revealed the secret feature
23:50:38 <elliott> hey, it was disturbing having news-ham lines appear on my screen, like a ghost.
23:50:40 <news-ham> Cuts 'leave Sure Start starved': Sure Start children's centres in England are being starved of funds and the network is shrinking as councils cut funds, says Labour. http://www.bbc.co.uk/go/rss/int/news/-/news/education-13206802
23:51:07 <oerjan> what are the chances this will end well
23:51:16 <elliott> pasting that was an honest mistake
23:51:28 <elliott> but, err, i suspect the worst it'll come to is another "iddiott" line?
23:51:42 <oerjan> what are the chances this will perhaps end well
23:51:44 <elliott> (i was trying to show you why your line didn't work :))
23:51:51 <elliott> oerjan: what are the haps that this will chance well my friends
23:52:14 <news-ham> Superb Trump to face Ding in semi: Judd Trump thrashes Graeme Dott 13-5 to reach the World Championship semi-finals, where he will face Ding Junhui, a 13-10 winner over Mark Selby. http://news.bbc.co.uk/go/rss/int/news/-/sport1/hi/snooker/13209141.stm
23:52:19 <elliott> oerjan: what are the haps that this will chance well my friends
23:52:26 <elliott> oerjan: what are the that this will chance well haps my friends
23:52:28 <news-ham> Timings on the day: The following timings have been confirmed for royal wedding day. http://www.bbc.co.uk/go/rss/int/news/-/news/uk-13097243
23:52:46 <oerjan> elliott: that one above would have been _so_ much better if it were about donald
23:54:01 <elliott> news-ham what are the donald duck haps my friends
23:54:02 <news-ham> Palestinian rivals 'agree deal': The Fatah party of Palestinian Authority President Mahmoud Abbas and Hamas, which governs Gaza, agree a reconciliation deal, officials say. http://www.bbc.co.uk/go/rss/int/news/-/news/world-middle-east-13215062
23:55:48 <pikhq> GAAAH WHY IS "AGREE" USED IN THIS MANNER
23:56:05 <elliott> pikhq: bbc english motherfucker
23:56:22 <oerjan> what are the grammar nazis up to? perhaps my friends know
23:56:23 <news-ham> VIDEO: Barack Obama: 'I was born in Hawaii': The White House has released President Barack Obama's birth certificate, in response to persistent rumours he was not born in the US. http://www.bbc.co.uk/go/rss/int/news/-/news/world-us-canada-13213810
23:57:03 <pikhq> parse("agree a deal"); // This will fail.
23:57:25 <oerjan> what are theologians' dark secrets? perhaps my friends are doomed!
23:57:25 <news-ham> 'Steal everything': Is the PlayStation hack a sign of crime to come? http://www.bbc.co.uk/go/rss/int/news/-/news/technology-13213632
23:57:34 <Ilari> Seems like that if p appears, p(p*2^i+1) will appear as well, if p*2^i+1 is a prime.
23:58:05 <oerjan> Ilari: this somehow reminds me of perfect numbers
23:58:11 -!- Sgeo has joined.