00:06:53 <oerjan> @tell boily <boily> norway is warmer this time of the year. <-- incidentally i saw in the newspaper that the night before yesterday was the fourth hottest january night ever measured (by minimum temperature) in trondheim. (it was 5.1 celsius.)
00:08:21 <oerjan> @tell boily http://www.adressa.no/vaeret/article8897643.ece
00:09:52 <oerjan> shachaf: you californians are p. spoiled
00:10:10 <shachaf> wait am i a californian now
00:10:39 <oerjan> i'd hazard a guess israelis are p. spoiled about temperature too
00:11:31 <shachaf> look the point is, why are you even in trondheim when it's so inhuman there
00:11:35 <oerjan> and a proper finn would never call 5.1 celsius cold.
00:12:32 <shachaf> when i was last in helsinki it was ~30°C
00:13:02 <fizzie> Second half of December in Finland was the warmest ever in statistics starting from 1961.
00:13:31 <shachaf> what if you put some flour in your hair and moved to california
00:13:58 <shachaf> that's more sensible than calling people spoiled imo
00:14:03 <oerjan> the highest temperature ever measured in norway was 35.6°C
00:15:56 <fizzie> They did have -39.7 °C in December somewhere up north though.
00:17:01 <fizzie> Last year. It hasn't been December this year yet.
00:17:55 -!- nisstyre has joined.
00:19:16 <myname> depends on where you let start your year
00:19:35 <myname> i mean, if america starts a week on sunday, why not start a year in december?
00:22:05 <fizzie> In fact, whole 2013 was the sixth warmest in the 1847..2013 range. (Top 5 has 1938, 1989, 2011, 2000, 1934.)
00:23:19 -!- augur has quit (Read error: Connection reset by peer).
00:23:56 <kmc> in Europe in the Middle Ages, the year generally started on 25 March
00:23:59 <kmc> https://en.wikipedia.org/wiki/New_Year#Historical_Christian_new_year_dates
00:24:31 <kmc> or sometimes on Easter, never mind that the date of easter moves back and forth
00:24:34 <kmc> 'same date could occur twice in a year; the two occurrences were distinguished as "before Easter" and "after Easter".'
00:24:35 <myname> also, there were a december this fiscal year
00:25:10 <kmc> January 1 is the Feast of the Circumcision of Christ
00:25:16 <kmc> religion is creepy
00:25:21 <shachaf> sunday is the sensible day to start the week
00:25:27 <shachaf> america starts it on monday
00:25:50 <myname> who starts on sunday, then?
00:25:58 <monotone> The American convention is Sunday.
00:26:05 <kmc> ISO starts it on Monday
00:26:12 <kmc> are you gonna argue with ISO
00:26:35 <shachaf> are you gonna argue with IMO
00:26:36 <monotone> Monday is the first day in most of continental Europe, as far as I'm aware.
00:26:41 <olsner> why argue with ISO when you can simply ignore them completely
00:27:17 <myname> america has been strong in ignoring standards for ages
00:30:00 -!- augur has joined.
00:31:59 <fizzie> Week numbers: worst things. (Even agreeing on the start day doesn't guarantee those.)
00:32:57 <myname> at least there is no $[ in there
00:33:42 <fizzie> "World map showing first day of week used in different countries Saturday Sunday Monday Wednesday (questionable)
00:34:26 <kmc> wednesday is questionable
00:34:44 <myname> i mean, in german it's practically named "midweek"
00:34:53 <myname> how the hell can a week start at that
00:35:51 <fizzie> It's "midweek" in Finnish too.
00:36:29 <myname> where live these crazy people?
00:36:37 <kmc> odin's day
00:36:42 <olsner> swedish also has "wednesday" (but spelled onsdag)
00:37:54 <fizzie> Apparently it's just wrong. (The map had Hungary marked with Wednesday.)
00:40:16 -!- ^v has joined.
00:40:24 <fizzie> ISO week numbers have Monday as first day, and the week is associated with the year that contains the Thursday of that week, but: "In some countries, though, the numbering system is different from the ISO standard. At least six numberings are in use: --"
00:41:06 <kmc> great thing about standards
00:44:10 <fizzie> "In Scandinavian countries, Saturday is called lördag, "lørdag," or laurdag, the name being derived from the old word laugr/laug (hence Icelandic name Laugardagur), meaning bath, thus Lördag equates to bath-day. This is due to the Viking practice of bathing on Saturdays.[citation needed]"
00:44:33 <fizzie> That last bit was somehow amusing.
00:44:50 <fizzie> (It's "lauantai" in Finnish.)
00:45:12 <olsner> indeed I bathed today (earlier today when it was saturday)
00:45:24 <fizzie> Vikings are well-known for their bathing habits.
00:46:41 <fizzie> I did, too; we have a sauna reservation on Saturday, which I think is p. typical. (Though it's v. possible a private sauna would be even more typical; haven't seen statistics.)
00:48:23 <fizzie> Statistics Finland's page is not loading. :/
00:49:12 <shachaf> myname: wednesday as midweek only makes sense if the week starts on sunday
00:49:46 <fizzie> shachaf: It makes perfect sense when week starts on Monday, because Saturday and Sunday don't count.
00:50:47 <shachaf> have you considered that you don't count (properly)
00:51:03 <shachaf> also hebrew gives better evidence for it
00:51:14 <shachaf> in hebrew sunday is called firstday and monday is called secondday and so on
00:51:19 <myname> why? my week ends on friday :p
00:51:22 <fizzie> Bah, this page only has statistics for numbers of buildings.
00:51:29 <kmc> shachaf: sensible
00:52:00 <myname> shachaf: must drive programmers crazy
00:54:35 <fizzie> Apparently, in 2012, there are approximately 1555000 saunas "in flats", and 2579781 "household-dwelling units" in general. So (assuming the contribution of multiple-sauna dwellings as negligible) I guess it would logically follow that it's more common to have a sauna than not.
00:55:01 <shachaf> well, i mean, why would you not have a sauna
00:55:26 <olsner> I wonder if I could get a sauna put in my flat, or if it requires special infrastructure
00:55:27 <fizzie> Because of money, I guess.
00:55:43 <fizzie> I believe it requires rather similar infrastructure as a bathroom.
00:56:10 <shachaf> my grandmother has a sauna in her house in .il
00:56:46 <fizzie> (We don't have a sauna here, probably because it wasn't as common as it nowadays is to put private ones in, back in 1984 when this building was builded.)
00:58:07 <fizzie> Having one does take up some floor space. Though the ones they put in small apartments are really very tiny.
00:58:24 <fizzie> The sort of things that will be very crowded if you try to have more than two people in at the same time.
00:59:03 <shachaf> imo you gotta have your own lake, too
00:59:44 <fizzie> There are 187888 lakes (definition: >500 square metres) in Finland, so we'd have to kill a lot of people to achieve that.
00:59:44 <olsner> maybe you could use your bathtub as the lake and have a foldable sauna you can use to convert the rest of the bathroom into a sauna
01:00:24 <fizzie> I guess we could dig out more lakes, too.
01:00:35 <fizzie> The killing solution just sounded occurred to me first.
01:01:36 <shachaf> maybe you can consider a 1000 square metre lake to be two adjacent 500 square metre lakes
01:05:56 <fizzie> According to W|A, there's 34330 km^2 of lakes in Finland, so divided to 5.42 million people, that's about 6334 square metres.
01:06:11 <fizzie> Enough to give everyone up to 12 lakes.
01:06:36 <shachaf> i think everyone should get some shore as well
01:07:38 <fizzie> Oh no, the shoreline problem.
01:08:07 <shachaf> it is difficult to have a sauna by the lake when all you have is lake
01:08:57 <fizzie> I asked for "shoreline in Finland", but I think it's interpreting that as the coastline bordering the sea, or something. (On the other hand, I learned that 1250 km is "0.78 × distance the Proclaimers would walk, just to be the man that walks a thousand miles to fall down at your door".
01:09:08 <fizzie> (The distance being a thousand miles.)
01:09:42 <fizzie> (The comparisons of W|A are frequently very useful.)
01:10:22 -!- zzo38 has quit (Remote host closed the connection).
01:11:39 <fizzie> I think I'll make like a Proclaimer and walk (a shorter distance) to fall asleep now. (Nights.)
01:13:46 <shachaf> wait, i meant centikilomiles
01:25:38 <olsner> I wonder if I should eat (because it's dinner time) or sleep (because it's bedtime)
01:28:24 <elliott> I may ban everyone who tries to argue that the week starts on Sunday
01:28:51 <oerjan> itt elliott proves he's an antisemite
01:35:24 -!- FreeFull_ has changed nick to FreeFull.
01:40:31 -!- Sprocklem has quit (Ping timeout: 272 seconds).
02:06:46 -!- nisstyre has quit (Quit: Leaving).
02:07:39 -!- nisstyre has joined.
02:19:51 -!- Froox has quit (Ping timeout: 246 seconds).
02:22:02 -!- Frooxius has joined.
02:27:38 -!- Bike has quit (Ping timeout: 265 seconds).
02:29:05 -!- Bike has joined.
02:32:27 -!- zzo38 has joined.
02:38:03 -!- Bike has quit (Ping timeout: 272 seconds).
02:39:26 -!- Bike has joined.
02:43:14 -!- oerjan has quit (Quit: pnyte).
02:43:16 -!- yorick has quit (Remote host closed the connection).
02:57:30 -!- Sorella has quit (Quit: It is tiem!).
03:04:09 -!- Sprocklem has joined.
03:11:42 <Sgeo> I didn't realize new BOFH stories are being produced
03:15:53 -!- augur_ has joined.
03:18:05 -!- augur has quit (Disconnected by services).
03:18:09 -!- augur_ has changed nick to augur.
03:38:31 <Bike> does anyone know about weird-ass linux boot errors
03:39:03 <Bike> this thing is telling me "Firmware Bug" and then.. i think it's crashing trying to look at an empty hard drive, which dumps me into ramfs, but it can't get a tty so i can't even tell it to die
03:40:12 <Bike> huh, some kernel bug. weird. i wonder why the recovery linux works then
03:50:59 -!- doesthiswork has joined.
03:58:54 <Bike> what the hell is shutdown -k for, i love it
04:13:44 -!- nooodl has quit (Quit: Ik ga weg).
04:37:12 -!- mauke has quit (Disconnected by services).
04:37:22 -!- mauke has joined.
04:39:04 -!- preflex_ has joined.
04:40:25 -!- preflex has quit (Ping timeout: 272 seconds).
04:40:27 -!- preflex_ has changed nick to preflex.
04:44:31 -!- _46bit_ has quit (Quit: Crunk.).
04:45:42 -!- _46bit has joined.
05:10:26 -!- carado has quit (Ping timeout: 264 seconds).
05:14:37 -!- carado has joined.
05:22:26 -!- carado has quit (Ping timeout: 264 seconds).
05:30:52 -!- Sprocklem has quit (Ping timeout: 260 seconds).
05:32:14 -!- myndzi\ has joined.
05:33:26 -!- myndzi has quit (Ping timeout: 240 seconds).
05:36:12 -!- myndzi has joined.
05:44:14 -!- carado has joined.
06:13:31 -!- tromp has quit (Ping timeout: 246 seconds).
06:14:02 -!- LinearInterpol has quit (Ping timeout: 264 seconds).
06:14:59 -!- tromp has joined.
06:32:21 <Jafet> Shutdown wasn't designed to be run setuid.
06:32:51 <Bike> hm, mine doesn't have that.
06:33:16 <Bike> actually i think this ubuntu one is a totally different implementation from the arch one, how odd
06:34:06 <Bike> which i guess has to do with ubuntu doing some total other thiing with initd, well, whatever
07:02:58 <zzo38> I have more idea of some kind of computer game based on something I have dreamt of. Do you have any such things?
07:04:58 <Bike> i thought of basing a game on dreams once.
07:05:01 -!- tromp_ has quit (Remote host closed the connection).
07:05:31 <zzo38> Can you please be more specific?
07:05:36 -!- tromp_ has joined.
07:06:13 <Bike> a game with a framing story of invading dreams somehow (Paprika ish, probably) used to exhibit real people's dreams in FPS form.
07:09:28 <zzo38> Ah, OK. My ideas isn't like that; it is simply using things I have dreamt as ideas.
07:09:54 <Bike> right. but it seemed related.
07:10:26 -!- tromp_ has quit (Ping timeout: 264 seconds).
07:18:15 -!- nisstyre has quit (Quit: Leaving).
07:22:02 -!- ski_ has quit (Quit: Lost terminal).
07:29:12 -!- FreeFull_ has joined.
07:30:01 <zzo38> Can you perhaps give an example of how you would do that?
07:31:07 -!- FreeFull has quit (Ping timeout: 245 seconds).
07:31:14 <Bike> i had a friend with a dream where she was picked up by robots and told to run obstacle courses, so that they could study humn behavior; bam, tutorial level
07:31:50 <quintopia> sounds a lot like psychonauts to me
07:48:57 -!- FreeFull_ has quit (Ping timeout: 252 seconds).
07:59:13 -!- FreeFull has joined.
07:59:56 <zzo38> I can tell you what one of my ideas is, though. You can gain (and change) your affiliation with various organizations during the game, such as the evil "The Core", or one of the three factions that secretly work for them; or the four small intelligent insects (two red and two blue) who are against them, and can eat the inside of your body and brain a little bit. Eventually they can grow much larger (almost as big as people) and wings to fly; ...
08:00:54 <zzo38> ... if big then it is a great help to you to stop "The Core", too. Or, you can play as the insectoid character, which is like another game, even though it is the same game. And then there can be even more confusing circumstances involved in the situations!!!!
08:01:48 <Bike> the game of confusing circumstances involved in situations
08:03:12 <quintopia> zzo38: i think i might like this game!
08:03:30 <coppro> Jafet: that's an important note
08:04:01 <coppro> translation "shutdown is not necessarily a secure piece of software. If you setuid it, attackers may do other things"
08:06:39 <zzo38> quintopia: Yes, maybe, but I don't actually have much more idea than just what I wrote now.
08:08:41 <zzo38> Well, there is one more thing. I also dreamt that such evil organizations was in the same building as a clothing store (maybe as a kind of front), and for some reason there was the situation where for "cosmic balance" or something like that, we had to accelerate the rack hanging shirts while riding it, once we reach the middle we must turn it around immediately and decelerate all the way to the other side until it stops.
08:08:51 <zzo38> However, it wouldn't stop...
08:09:44 <zzo38> So such a thing could be worked into a game too, but it can also be made not to.
08:09:51 <zzo38> Maybe it doesn't help.
08:10:58 <zzo38> I know it isn't sensible.
08:13:02 <zzo38> The other ideas above seems more likely to work, though. But still I don't have much more idea to actually program in such things.
08:14:02 <zzo38> Yes, I might be able to do so in future
08:14:19 <zzo38> Do *you* have any better ideas?
08:15:04 <quintopia> i am sometimes working on a puzzle game
08:15:20 <zzo38> What kind of puzzle game?
08:15:42 -!- FreeFull has quit (Ping timeout: 245 seconds).
08:17:06 <quintopia> it's about santa's elves going on strike. so santa gets conned into replacing them with robots that run on rails. and the player is stuck with the task of designing the rails and switches so that the toys get put together correctly.
08:18:38 <quintopia> so it would be a manufactoria sort of game, but less turing-complete (no queue)
08:24:49 <zzo38> If I ever do make a game based on what I wrote about what I dreamt (the first one, not the second unsensible one), then I can show you. But, I might not. Nevertheless, there are several other idea possibility.
08:25:21 <quintopia> how much time a day do you usually spend building software
08:26:58 -!- ^v has quit (Remote host closed the connection).
08:42:56 <zzo38> Can Zerocoin keys be transmitted using separate channels?
08:57:37 <quintopia> but there's not really a need for it
08:58:15 <quintopia> instead, just reclaim your zerocoin as btc, and immediately give it to whomever in the usual way
08:58:34 <quintopia> anonymity is maintained as long as you do the claiming with a new address
08:59:32 -!- FreeFull has joined.
08:59:33 -!- FreeFull has quit (Changing host).
08:59:33 -!- FreeFull has joined.
09:06:50 <zzo38> But would it simplify something in any way if in a way that would still work OK?
09:08:25 <zzo38> You would still need time to verify the transaction, for sure, though.
09:13:46 <zzo38> I'm not sure what ways there are to do what things, since it seem to me a good idea would be to design the system to allow transmissions on separate channels
09:34:18 <quintopia> just hand out the private key of the zerocoin transaction you made
09:59:55 <zzo38> I am just wondering how much privacy issues there are from such thing; maybe it depends on who uses it?
10:01:10 <zzo38> Do you know about CipherSaber?
10:01:18 <quintopia> well, the point of zerocoin is to remain anonymous, and if you transfer zerocin via a side channel, you are deanonymizing yourself to the person you give it to
10:02:01 -!- FreeFull has quit (Ping timeout: 272 seconds).
10:02:16 <zzo38> CipherSaber is a variant of ARCFOUR with a ten-byte initialization vector and the key setup repeated a secret number of times too (this secret number would be part of the key).
10:03:09 -!- FreeFull has joined.
10:05:22 <quintopia> it appears to be a specific RC4 implementation. no biggie.
10:07:05 <zzo38> I have used RC4 as nothing more than a random number generator, and this seems to work; it also isn't too difficult to implement in a 6502, although it uses up a large amount of the RAM in systems such as NES/Famicom.
10:08:32 <zzo38> I want to implement a good quality random number generator for the RANDOM instruction of the Z-machine, but I don't have 256 bytes of RAM remaining which can be used for this purpose; almost all of the RAM is already used (and that includes the 64K RAM in the cartridge, too!)
10:08:36 <quintopia> doesn't seem like famicom would have much need for encryption
10:09:25 <zzo38> Yes, it is true, but I just wanted to implement a good quality random number generator; cryptographic quality is not necessary but it does have to be *good* quality.
10:10:27 <quintopia> so have you considered and rejected LCGs?
10:10:35 <quintopia> they certainly use very little memory
10:11:43 <zzo38> Are they good enough?
10:12:29 <zzo38> I don't expect it to be good enough.
10:12:40 <quintopia> i don't know what good enough means to you
10:12:48 -!- carado_ has joined.
10:12:55 -!- carado has quit (Ping timeout: 245 seconds).
10:13:08 <quintopia> LFSRs can be computed in the same amount of memory but take longer to compute, but have better randomness properties
10:13:30 -!- carado_ has quit (Client Quit).
10:13:54 <zzo38> Random numbers shouldn't need to be generated too often.
10:15:48 <zzo38> Probably in any Z-machine games it won't be generated a lot of random numbers each turn, but it need to be at least reasonably fast, but not a lot like other things are.
10:16:47 <zzo38> Note that the time could also be used as an input to the random number generator; the results don't have to be predictable.
10:16:48 <quintopia> then i think LFSR will be good enough.
10:20:17 <zzo38> It could be combined with a time counter increasing every frame, and then it need to know how many bits of output are needed and then if the number is too high after masking out the unneeded bits, try again.
10:22:21 <quintopia> using XOR to combine with time i assume?
10:23:17 -!- AnotherTest has joined.
10:24:16 <Jafet> Z machine on a famicom?
10:24:18 <zzo38> Note that there is actually a LFSR in the 2A03 (the CPU for the Famicom), however the program cannot read it unless pin 30 is high, in which case the I/O ports stop working, and anyways pin 30 is internally hardwired to ground and cannot be changed. It *is* output to the cartridge, but no cartridge does anything with it except to output it to the audio.
10:25:01 <zzo38> Jafet: Yes, I made up much of it already, but not division and random numbers. http://wiki.nesdev.com/w/index.php/User:Zzo38/Famicom_Z-machine You can read the coding so far
10:26:09 <Jafet> You can use a multiply-with-carry generator, I don't know if it will fit in 256 bytes (of code?)
10:26:28 <zzo38> Jafet: The code can be put in ROM; it doesn't need to fit in RAM.
10:26:53 <zzo38> I can implement it in software, though. I didn't intend to use XOR, but rather subtraction, to do the combine, however; and the number need to be 15-bits but a 16-bits LFSR could be used.
10:26:54 <Jafet> Though the small word size of the 6502 may be a problem.
10:27:12 <zzo38> That would slow it down a bit, yes.
10:28:11 <zzo38> Maybe the multiplication would be better using the Russian algorithm; would this be better or not, than what I have?
10:28:30 <zzo38> (I mean for the MUL instruction, not for random number generators)
10:28:39 <zzo38> And yet I would still require division/modulo
10:31:31 <zzo38> As well as an efficient way to lookup in the vocabulary (which could be preprocessed using a separate program if it would help; there is probably enough ROM space for such a thing)
10:35:24 <zzo38> Also please notify me if you found anything wrong with the program so far
10:41:54 -!- doesthiswork has quit (Quit: Leaving.).
10:43:42 -!- MindlessDrone has joined.
10:50:00 <zzo38> I noticed a C code for LFSR in Wikipedia. However, in a GNU C program, you might use __builtin_parity, would that work better? (Or you can use __builtin_parityll if you need more bits)
11:04:26 -!- carado has joined.
11:04:36 -!- carado has quit (Client Quit).
11:12:53 <Jafet> The 6502 does not have a division instruction?
11:14:00 <Jafet> Oh, it's *that* old.
11:17:33 <fizzie> SPARC doesn't have a division instruction either.
11:18:02 <fizzie> (And it's a not-that-old "real" processor.)
11:18:20 <fizzie> I guess it might have a floating-point division, though.
11:19:21 <fizzie> (Or, to be more exact, a SPARC might commonly come with a FPU that does division.)
11:23:16 <zzo38> Yes, 6502 has no multiplication or division instruction.
11:23:17 -!- gully_foil_Ja has joined.
11:23:47 <zzo38> (Although the MMC5 cartridge does include a hardware register for multiplication.)
11:27:16 <ion> http://img.pr0gramm.com/2012/07/tfcu4.png
11:27:52 <fizzie> The ARM in the Nintendo DS has a hardware integer multiplication (and multiply-accumulate) instruction, but no division; however, the DS adds a separate memory-mapped (64- and 32-bit) integer division and square-root device.
11:28:32 -!- FreeFull has quit (Quit: Regaining my lost channels).
11:29:03 -!- FreeFull has joined.
11:31:08 <zzo38> Yes, that is true even as early as ARMv2, I think; it has multiplication but not division instruction.
11:31:26 <myname> ion: wondering what linux looks like at this
11:33:29 <fizzie> myname: http://ecx.images-amazon.com/images/I/518UgTYS4UL._SY300_.jpg perhaps?
11:40:39 -!- AwfulProgrammer_ has joined.
11:54:02 -!- Zerker has joined.
11:57:55 -!- Sorella has joined.
12:16:23 -!- sebbu has quit (Ping timeout: 245 seconds).
12:21:11 <zzo38> There is article in Esolang wiki relating to some Japanese book. However, it mentions "Starry" and "Bolic", which are red links.
12:21:43 <b_jonas> multiplication? the 6502 doesn't even have proper shift instructions.
12:22:27 <zzo38> b_jonas: The 6502 can shift one at a time, although there isn't any arithmetic right shift.
12:23:46 -!- sebbu has joined.
12:24:33 -!- sebbu has quit (Changing host).
12:24:33 -!- sebbu has joined.
12:25:19 <zzo38> You can fake it, though, for example CMP #$80 ROR A (I think)
12:25:42 <fizzie> Multiplication is quite a lot to ask. Z80 doesn't have it either.
12:25:46 <b_jonas> zzo38: something like that, yes
12:26:20 <b_jonas> mind you, even 386 didn't have a fast multiplication
12:26:31 <b_jonas> it did have a multiplication instruction but it took ages
12:27:00 <b_jonas> it was only later, near the pentiums, when they added a multiplication circuit so we can now perform multiplication fast
12:36:43 -!- sebbu2 has joined.
12:36:47 -!- sebbu has quit (Ping timeout: 252 seconds).
12:37:11 -!- sebbu2 has quit (Changing host).
12:37:11 -!- sebbu2 has joined.
12:37:12 -!- sebbu2 has changed nick to sebbu.
12:45:50 <zzo38> Also looking at that Japanese stuff, it reveals another esolang, called "Smile", which also isn't in the wiki. https://github.com/yhara/smile The implementation appears to be incomplete.
12:47:58 <fizzie> Shouldn't a "semicolon" just be a single dot, anyway?
12:49:11 <fizzie> (The Finnish name translates literally to "half-dot", incidentally, which is equally illogical.)
12:52:23 <zzo38> I would think "semicolon" and "half-dot" names are due to their function rather than writing.
12:57:00 -!- yorick has joined.
12:57:21 <zzo38> This Japanese stuff, in addition to Bolic, Starry, and Smile, also mentions HSQ9+ which is HQ9+ with the "S" command to display the steam locomotive (like the "sl" command in some UNIX systems), and Uncontrollable, a brainfuck equivalent using words that mean control structures in Ruby and other programming languages.
12:57:33 -!- Zerker has quit (Remote host closed the connection).
12:58:46 -!- Zerker has joined.
12:58:50 <myname> i wonder if we will have ABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789+~ eventually
12:59:27 <Slereah_> 8 will print "88 bottles of beers"
12:59:52 <myname> of course not, 8 is clearly conways game of life
13:13:01 -!- Zerker has quit (Quit: Colloquy for iPad - Timeout (10 minutes)).
13:13:08 <mauke> http://mauke.hopto.org/stuff/perl/hq9+.pl - good old usenet signature
13:19:29 -!- oerjan has joined.
13:21:46 -!- gully_foil_Ja has quit (Quit: has left chat).
13:26:52 -!- LinearInterpol has joined.
13:31:22 -!- pikhq has quit (*.net *.split).
13:31:29 -!- pikhq has joined.
13:43:29 <FireFly> zzo38: I wonder if there are any other published books about esolangs?
13:45:46 -!- nooodl has joined.
13:59:19 <zzo38> FireFly: I don't know, but I have read a magazine article once about it (I forget which).
14:04:34 <oerjan> <zzo38> There is article in Esolang wiki relating to some Japanese book. However, it mentions "Starry" and "Bolic", which are red links. <-- yeah we are clearly missing some japanese stuff. the only reason we have KEMURI is that i used google translate on wikipedia's japanese esolang page once.
14:11:27 <fizzie> Brainfuck has probably been mentioned in MEDIA many times.
14:11:52 <fizzie> I vaguely recall some kind of a mention in some Finnish computer magazine.
14:16:22 -!- AwfulProgrammer_ has changed nick to AwfulProgrammer.
14:21:00 <nooodl> ah, there's a ruby implementation on yhara's book's site
14:23:26 <zzo38> Some Japanese stuff I have added myself too though
14:23:46 <Taneb> I am once more not in Hexham
14:24:24 <zzo38> (But I didn't use Google translation; I can read a little bit Japanese, and some may have partial English documentation, and can see the C or Ruby implementations)
14:35:13 <fizzie> Taneb: I guess you still carry an intangible bit of Hexham with you, in your heart?
14:35:55 <Taneb> Also I stole the abbey
14:36:18 <olsner> you can take the Taneb out of Hexham but you can't take the Hexham out of Taneb
14:36:38 <Taneb> I'm starting a collection of old churches
14:37:10 <Taneb> The oldest church I've been in would either be Hexham Abbey, York Minster, or the Colosseum
14:47:10 <fizzie> Panteon, that's the one with the hole in the dome?
14:48:50 <Taneb> Hexham Abbey is the only one that started out Christian and stayed Christian
14:49:28 <fizzie> What's the York one started as?
14:49:46 <Taneb> And it had a bit when it was Norse
15:09:25 -!- tromp_ has joined.
15:17:59 <nooodl> http://esolangs.org/wiki/Starry there
15:31:48 -!- Chillectual has joined.
15:34:49 <ion> http://lyrics.in.th/17527 http://youtu.be/ArvmQBEkzWw
15:34:58 -!- LinearInterpol has quit (Ping timeout: 265 seconds).
16:06:50 -!- Chillectual has quit (Read error: Connection timed out).
16:08:53 -!- ais523 has joined.
16:09:01 -!- Chillectual has joined.
16:14:48 -!- Phantom_Hoover has joined.
16:22:33 -!- oerjan has quit (Quit: leaving).
16:46:03 <quintopia> why does endoh3's hint make such a big deal about "what is 89/84?????" it being a number that's being multiplied by and it being a program to output sounds based on note names, it's obviously a "good enough" approximation of the 12th root of 2.
16:49:09 <fizzie> I don't really see a "big deal" about it.
16:49:41 <fizzie> I mean, there's a single bullet point about it.
16:50:01 <fizzie> (With a single question mark.)
16:50:26 <FreeFull> scale=10; (89/84)^12=2.0013758126
16:50:58 <fizzie> The hint file mentions a thing that's useful for finding rational approximations, too.
16:51:00 <quintopia> hmm i thought i'd seen it in there twice, but it turns out i had seen it in the spoiler (because i still haven't unrot13'd it) and considered that the second time
16:51:03 <FreeFull> scale=10; ((89/84)^12 - 2)/2 * 100
16:51:40 <fizzie> It's kind of spoiled in the spoiler, but that's presumably why it's called a spoiler.
16:52:06 <quintopia> rot13 isn't good for hiding numerical spoilers
16:52:12 <fizzie> Also the cbj(2, 1.0 / 12.0) part isn't terribly well-protec... right, that.
16:53:08 <FreeFull> Actually, it's off by less than that
16:53:23 <quintopia> fizzie: i still don't know what a stern-brocot tree is. should i look it up
16:53:42 <FreeFull> That's a very good approximation of 2^(1/12)
16:53:47 <fizzie> quintopia: Spoiler: it's something that can be useful for locating approximations.
16:54:33 <quintopia> fizzie: i figured as much. but is it interesting enough to spend minutes on? could i use it to prove that phi is the most irrational number?
16:54:41 <FreeFull> What are we talking about anyway?
16:54:42 <fizzie> It's also a binary tree that has every (positive) rational number in it, which is maybe funky?
16:54:52 <FreeFull> quintopia: What does most irrational mean?
16:56:27 <quintopia> FreeFull: something like "for any large enough given length denominator, the best approximation of phi with denominators of that length will be worse than the best approximations of any other irrational number"
16:57:09 <FreeFull> quintopia: That sounds like a very difficult and significant thing to prove
16:57:27 <quintopia> FreeFull: i certainly haven't learned how yet
16:57:30 <FreeFull> Let's think of some irrational numbers
16:58:17 <FreeFull> Hmm, I wonder if the non-zero fixed points of tan are irrational
16:58:45 <quintopia> FreeFull: also, the entire sequence of golden sections and silver sections
16:59:08 <FreeFull> Most multiples of irrational numbers are irrational
16:59:46 <FreeFull> I know, just enumerating more irrationals
17:01:01 <fizzie> All transcendental numbers are irrational, I believe.
17:01:13 <int-e> Rational approximations are intimately related to continued fractions, and 1/(1+1/(1+1/(1+ ... ))) = phi has a tail of 1s which is about the worst that can happen. (Cutting the continued fraction off just before a large number typically results in good approximations.)
17:01:34 <FreeFull> It's even harder to prove something is transcendental though
17:02:13 <FreeFull> int-e: You can convert many continued fractions into recursive relations and solve them
17:02:30 <FreeFull> So there, you'd get 1/(1+phi) = phi
17:02:53 <int-e> FreeFull: Yes. But that's besides the point, since phi itself is not rational.
17:02:53 <fizzie> Given #esoteric, I guess it's mandatory to also mention Chaitin's constant.
17:03:37 <FreeFull> fizzie: We can't even give an approximation for that
17:03:40 <int-e> if you truncate the continued fraction you get rational approximations of phi: 0/1, 1/1, 1/2, 2/3, 3/5, 5/8, 8/13, 13/21, etc.
17:04:16 <fizzie> FreeFull: You can always give an approximation, it just might not be any good.
17:04:51 <quintopia> FreeFull: the solution to your recurrence relation (as a sequence) is F_{n+1}/F_n
17:05:00 <FreeFull> fizzie: We can't even tell how good an approximation is though
17:05:33 <fizzie> FreeFull: Well, 1/2 is not more than 1/2 away.
17:06:10 <int-e> FreeFull: But 1+1/(1+1/(1+1/(1+ ... ))) is almost the same. 1/1, 2/1, 3/2, 5/3, 8/5, 13/8 etc.
17:06:25 <quintopia> let's list definable numbers with nonfinite descriptions
17:06:30 <ion> 13/8 is a nice time signature.
17:07:12 <quintopia> ion: i know a rock song in 13/8, that wasn't broken in the way that led zeppelin's drummer broke nice time signatures
17:07:12 <FreeFull> 1/(1+phi) = phi 1 = (1+phi)phi 1 = phi + phi^2 phi^2 + phi + 1 = 0 (phi + 1/2)^2 - 1/4 + 1 = 0 (phi + 1/2)^2 = -3/4 phi + 1/2 = sqrt(-3/4) I did something wrong
17:07:23 <FreeFull> phi + 1/2 doesn't have a complex part ):
17:07:42 <nooodl> 1 = phi + phi^2 phi^2 + phi + 1 = 0
17:08:00 <FreeFull> Yeah, that solves the minus problem
17:08:46 <FreeFull> phi^2 + phi - 1 = 0 (phi + 1/2)^2 - 1/4 - 1 = 0 phi + 1/2 = sqrt(5)/2 phi = (sqrt(5) + 1)/2
17:08:56 <int-e> pi = 3 + 1/(7+1/(15+(1/(1+1/(292+...)))) gives the great approximation 355/113 of pi. (Off by about 1/3748629)
17:09:06 <FreeFull> I was worried I might get 1/phi instead by accident
17:09:07 <quintopia> fizzie: do you know any definable numbers with (only) infinite descriptions?
17:09:15 <int-e> where 355/113 = 3+1/(7+1/(15+1/1))
17:09:43 <quintopia> int-e: didn't the ancient chinese know that approximation?
17:12:50 <int-e> I wouldn't be surprised if they did.
17:16:02 <FreeFull> The easiest way to approximate pi is to draw a circle and measure it
17:17:23 <quintopia> i assume the other silver sections are the ones with continued fractions [2,2,2,2,2,2,2,2,...] and [3,3,3,3,3,3,3,3,3,3,...] etc.
17:18:13 -!- Chillectual has changed nick to LinearInterpol.
17:18:40 <FreeFull> I did the square root without putting in ±
17:19:36 <int-e> quintopia: https://en.wikipedia.org/wiki/Tsu_Chung-Chih
17:19:56 <int-e> "He obtained the result by approximating a circle with a 12,288 (= 2^12 × 3) sided polygon"
17:20:00 <quintopia> FreeFull: making a circle and measuring it seems a lot more difficult than writing a program that generates a sequence that converges to pi, and kill the program when it gets close enough
17:30:29 <FreeFull> quintopia: Try doing that when you find yourself stuck 3000 years in the past
17:30:33 <quintopia> does unicode contain rod numerals?
17:32:01 <quintopia> FreeFull: you said easiest. you didn't say easiest on pen and paper. nonetheless, i could do so, and it would still probably get good approximations faster than making and measuring a circle
17:33:01 <FreeFull> quintopia: I never mentioned pen and paper
17:33:27 <kmc> good morning
17:33:31 <quintopia> ...i'm pretty sure i just said you didn't say that :P
17:33:32 <kmc> fungot: good morning
17:33:32 <fungot> kmc: hmm... my fnord bad enough already. :) at least fnord.)
17:33:52 <quintopia> but you can do as well with brush and paper, or counting rods, as with pen and paper
17:33:53 <kmc> fungot: a bad fnord?!??
17:33:54 <fungot> kmc: every time you accidentally put it only. we do have a wristwatch which runs scheme, now that we have a good point :)
17:33:55 <FreeFull> If you know the mathematics to calculate pi that easily..
17:34:12 <kmc> fungot: does your wristwatch support time travel via continuations?
17:34:13 <fungot> kmc: and it is maybe not so much big iron as lots of little pieces of information...
17:34:29 <FreeFull> quintopia: I was thinking chalk and long rope
17:35:13 <quintopia> FreeFull: keep the rope. i'll do the calculation by hand with the chalk :D
18:05:00 <mroman> `unicode arabic digit nine
18:06:06 <mroman> `unicode arabic-indic digit nine
18:06:27 <HackEgo> [U+09ED BENGALI DIGIT SEVEN] [U+0669 ARABIC-INDIC DIGIT NINE]
18:20:17 <fizzie> The other is clearly larger, at least in my rendering here.
18:24:32 <myname> but if it is written down...
18:24:52 <FreeFull> The other one is somewhat larger for me
18:28:30 <mroman> The left one is a litle bit more fat than the right one
18:30:22 <ion> They look like they come from different fonts here.
18:35:27 <ais523> Konvoersation'se capabilit or weird y ypos: apparently if it's laggin enough legters can atpear out p order
18:36:43 <ais523> (left over from te previoush line: "of")
18:39:57 <ais523> let ine type somemhing really fast aid just prens return as see what ndappens
18:40:24 <ais523> (the tin" at th" e start osthat line f appear to ave behn from anoeher chanelt)
18:40:51 <ais523> what ort of UIscan end up reading ke s out of oyder?
18:40:55 <kmc> seems like a p. serious bug
18:41:19 <ais523> sometiys it takesmelike 20 s onds befoece the lettrrs appear
18:41:37 <ais523> even thoughemost of th others have already eeen sentb
18:42:22 <ais523> ths newline fe om the endro the prev us messagio sent a meesage for ms in anotherechannel
18:43:22 <ais523> itsorse becauw I can't setype becau my hand ses damaged
18:43:31 <ais523> ind I can'a sensibltfix the ry sulting tyes
18:45:50 <FreeFull> ais523: Maybe you should use a different client
18:46:24 <quintopia> that is very very difficult to read
18:54:23 -!- ^v has joined.
18:59:10 <fizzie> That is like an automatic modern poetry mode.
19:01:55 <FreeFull> I thought megahal was the height of that
19:02:15 * ais523 experiments with writing their comments in a text editor, then copy-pasting
19:02:29 <ais523> huh, that seems to work quite well
19:03:40 <FreeFull> ais523: I still think you should switch clients
19:04:03 <ais523> II'm not nomally in archannel wi h 1000+ utrs
19:05:05 <int-e> ... "Oh well, if you prefer, I can recognize handwriting," said the imp proudly. "I'm quite advanced." Vimes pulled out his notebook and held it up. "Like this?" he said. The imp squinted for a moment. "Yep," it said. "That's handwriting, sure enough. Curly bits, spiky bits, all joined together. Yep. Handwriting. I'd recognize it anywhere." "Aren't you supposed to tell me what it say?" The demon looked wary. "Says?" it...
19:05:11 <int-e> ...said. "It's supposed to make noises?"
19:08:54 <int-e> ais523: It must be one of those super-efficient multithreaded applications.
19:12:13 <int-e> Hey it might even be a security feature. After all, you're not supposed to trust user input. So you better process each character in its own sandbox!
19:13:39 -!- Bike has quit (Ping timeout: 272 seconds).
19:14:53 -!- Bike has joined.
19:21:59 -!- Phantom_Hoover has quit (Ping timeout: 272 seconds).
19:28:51 -!- iamcal_ has joined.
19:29:40 -!- nortti_ has joined.
19:30:12 -!- iamcal has quit (Ping timeout: 246 seconds).
19:30:12 -!- iamcal_ has changed nick to iamcal.
19:30:13 -!- oklopol_ has joined.
19:31:05 -!- nortti has quit (Ping timeout: 246 seconds).
19:31:06 -!- oklopol has quit (Ping timeout: 246 seconds).
19:32:39 -!- nortti_ has changed nick to nortti.
19:36:41 -!- jix has quit (Remote host closed the connection).
19:39:55 -!- jix has joined.
19:43:28 <FreeFull> "static int running = 1;" real useful code
19:43:49 <FreeFull> Well, it does actually set running to 0 on interrupt
19:49:53 -!- myndzi\ has quit (Read error: Connection reset by peer).
19:50:03 -!- nortti_ has joined.
19:50:50 -!- myndzi\ has joined.
19:50:56 -!- nortti has quit (Disconnected by services).
19:51:19 -!- nortti_ has changed nick to nortti.
19:51:24 <kmc> the other day I learned about sigatomic_t
19:51:33 <kmc> which is what you're supposed to use for such variables, when they would be updated from a signal handler
19:51:48 <kmc> how many integer types does standard C+POSIX have? "a lot"
19:53:15 -!- tromp__ has joined.
19:53:29 -!- myndzi has quit (*.net *.split).
19:53:29 -!- _46bit has quit (*.net *.split).
19:53:29 -!- qlkzy has quit (*.net *.split).
19:53:29 -!- CADD has quit (*.net *.split).
19:53:29 -!- upgrayeddd has quit (*.net *.split).
19:53:38 <ais523> kmc: 10, all the others are aliases for one of those 10
19:53:38 -!- tertu has joined.
19:54:01 <kmc> I meant including aliases, since what the aliases resolve to is platform-dependent
19:55:41 -!- upgrayeddd has joined.
19:56:22 -!- gully_foil_Ja has joined.
19:56:55 <fizzie> ais523: Why would all the other types need to be aliases of standard integer types?
19:57:19 <ais523> fizzie: I think they are by definition, not sure though
19:57:24 -!- upgrayeddd has changed nick to upgrayeddd_.
19:57:25 -!- upgrayeddd_ has quit (Excess Flood).
19:57:36 <fizzie> C allows for an unlimited amount of extended signed integer types, and I don't think most named integer types are restricted to be one of the standard ones.
19:57:46 -!- tromp has quit (Ping timeout: 260 seconds).
19:58:33 -!- upgrayeddd has joined.
19:59:03 <fizzie> All the <stdint.h> types are at least only required to denote "a -- integer type" with no further qualifications.
19:59:26 -!- impomatic has joined.
20:00:57 -!- Phantom_Hoover has joined.
20:01:28 <fizzie> (In any case, "a lot" sounds like a proper answer for the original including-aliases question, since how many distinct underlying types there are is obviously implementation-defined.)
20:01:44 <fizzie> Random fact: POSIX makes time_t an integer type.
20:02:03 <fizzie> (C11 allows any real type, and C99 allows any arithmetic type, including complex types.)
20:02:13 -!- JZTech102 has joined.
20:02:16 -!- callforjudgement has joined.
20:02:33 -!- _46bit has joined.
20:02:34 -!- _46bit has quit (Changing host).
20:02:34 -!- _46bit has joined.
20:03:15 -!- mauke has quit (Disconnected by services).
20:03:17 -!- oklopol__ has joined.
20:03:23 -!- mauke has joined.
20:03:27 -!- tertu has quit (Read error: Connection reset by peer).
20:04:41 -!- Mindless1 has joined.
20:07:01 -!- contrapumpkin has joined.
20:08:19 -!- JZTech101 has quit (Read error: Operation timed out).
20:08:20 -!- MindlessDrone has quit (Read error: Operation timed out).
20:08:20 -!- oklopol_ has quit (Read error: Operation timed out).
20:08:21 -!- ais523 has quit (Ping timeout: 265 seconds).
20:08:21 -!- copumpkin has quit (Ping timeout: 265 seconds).
20:08:21 -!- lifthrasiir has quit (Ping timeout: 265 seconds).
20:09:09 -!- contrapumpkin has changed nick to copumpkin.
20:11:51 -!- mauke has quit (Disconnected by services).
20:11:58 -!- mauke has joined.
20:13:26 -!- Phantom__Hoover has joined.
20:13:30 -!- Phantom__Hoover has quit (Changing host).
20:13:30 -!- Phantom__Hoover has joined.
20:13:42 -!- lifthrasiir has joined.
20:18:15 -!- AwfulProgrammer has quit (Ping timeout: 264 seconds).
20:18:47 -!- JZTech102 has quit (*.net *.split).
20:18:48 -!- Phantom_Hoover has quit (*.net *.split).
20:19:52 -!- JZTech101 has joined.
20:21:47 -!- AwfulProgrammer has joined.
20:23:22 -!- iamcal_ has joined.
20:26:01 -!- iamcal has quit (Ping timeout: 264 seconds).
20:26:02 -!- iamcal_ has changed nick to iamcal.
20:26:04 -!- Gregor has quit (Ping timeout: 264 seconds).
20:26:04 -!- callforjudgement has quit (Ping timeout: 264 seconds).
20:27:48 -!- impomatic has quit (Ping timeout: 260 seconds).
20:28:05 -!- Phantom_Hoover has joined.
20:28:20 -!- Gregor has joined.
20:29:05 -!- impomatic has joined.
20:29:13 -!- clog has quit (Ping timeout: 260 seconds).
20:29:40 -!- Phantom__Hoover has quit (Ping timeout: 259 seconds).
20:31:15 -!- contrapumpkin has joined.
20:35:58 -!- _46bit has quit (*.net *.split).
20:36:08 -!- rodgort` has joined.
20:36:48 -!- augur has quit (*.net *.split).
20:36:48 -!- TodPunk has quit (*.net *.split).
20:36:48 -!- jconn has quit (*.net *.split).
20:36:48 -!- glogbackup has quit (*.net *.split).
20:36:49 -!- Gracenotes has quit (*.net *.split).
20:36:49 -!- b_jonas has quit (*.net *.split).
20:36:49 -!- rodgort has quit (Read error: Connection reset by peer).
20:36:51 -!- scarf has joined.
20:36:52 -!- scarf has quit (Changing host).
20:36:52 -!- scarf has joined.
20:36:54 -!- copumpkin has quit (Quit: Textual IRC Client: www.textualapp.com).
20:38:35 -!- augur has joined.
20:38:36 -!- TodPunk has joined.
20:38:36 -!- jconn has joined.
20:38:36 -!- Gracenotes has joined.
20:38:36 -!- b_jonas has joined.
20:55:09 <int-e> Awww. I didn't get a botnet for christmas.
20:58:23 -!- Mindless1 has quit (Quit: Mindless1).
21:07:15 -!- conehead has joined.
21:19:15 <Bike> i made an arch image with unetbootin, put it on a stick, booted from. get dmesg whining about not being able to deal with the stick, and i get dumped to a shell excet oh wait no job control and no tty so i have to hard boot, fuuuuuck computers
21:19:26 <fizzie> How would you package up a botnet under a christmas tree, anyway?
21:21:36 -!- nisstyre has joined.
21:32:30 <monotone> fizzie: you get the C&C server name and credentials on a card, I suppose
21:35:00 <kmc> complex time_t
21:35:33 <kmc> time is a float; lunchtime doubly so
21:36:52 <fizzie> Fancy justification for a complex time_t: it discourages people from assuming unportable things about the ordering of the values; something they might do if it was a real type.
21:37:37 <kmc> is it a type error to use ordering operators on complex types?
21:38:07 <fizzie> It is a constraint violation, yes.
21:38:33 <fizzie> "One of the following shall hold: - both operands have real type; or - both operands are pointers to qualified or unqualified versions of compatible object types."
21:38:48 <Bike> i hear having multiple time(like) dimensions makes physics boring, unless you have one spacelike and three timelike in which case it's just like this
21:49:20 <Vorpal> Bike, huh, I have installed arch from an usb stick before I *think*. Yeah pretty sure.
21:49:38 <Vorpal> Don't remember if I used unetbootin though, was ages ago
21:50:11 <kmc> I think most distros these days provide an image you can dd onto your USB stick
21:50:19 <kmc> some of them even manage for it to be the same file as the .iso, through hax
21:50:53 <Bike> yeah, that's what i did.
21:51:14 <Bike> that should mean i don't hve to care about partition tables or anything, shouldn't it?
21:51:32 <kmc> just make sure you dd it to /dev/sdb not /dev/sdb1 if that's what the instructions sya
21:52:25 <Bike> yeah, the instructions say that specifically. but then parted tells me the partition table is fucked up so i thought that might be related.
21:53:13 <kmc> maybe it doesn't need to have a partition table
21:57:06 <Bike> probably not, since i get usb errors either way
21:57:57 <Bike> i thought maybe i connected the internal cables wrong but i can read from the usb stick (with a filesystem, not after dd) from an old hard disk plugged in the same puter
21:59:03 <kmc> +5, use of the word "puter"
21:59:24 <Bike> that's what it's doing. it's puting
21:59:42 <kmc> NIKON: What, your mom buy you a 'puter for Christmas?
22:00:15 -!- LinearInterpol has changed nick to ZeroCool.
22:00:21 <ZeroCool> baahahaha, this name is taken.
22:00:41 -!- ZeroCool has changed nick to LinearInterpol.
22:01:58 <Bike> https://twitter.com/mnxmnkmnd/status/419951311396667392/photo/1/large i can't even tell what's doing an error -32, it isn't libusb
22:04:39 -!- AnotherTest has quit (Ping timeout: 272 seconds).
22:05:12 -!- impomatic has quit (Ping timeout: 252 seconds).
22:10:44 -!- conehead has quit (Quit: Computer has gone to sleep.).
22:17:01 -!- carado has joined.
22:18:07 -!- carado has quit (Client Quit).
22:24:24 -!- _46bit has joined.
22:24:40 -!- _46bit has quit (Changing host).
22:24:40 -!- _46bit has joined.
22:28:51 -!- zzo38 has quit (Remote host closed the connection).
22:39:08 -!- clog has joined.
22:41:38 -!- oerjan has joined.
22:46:42 -!- scarf has quit.
22:47:34 -!- ais523 has joined.
22:51:32 <oerjan> > let cfrac x | isInfinite x = [] | otherwise = f : cfrac (1/(x - fromIntegral f)) where f = floor x in cfrac (2**(1/12))
22:51:33 <lambdabot> [1,16,1,4,2,7,1,1,2,2,7,4,1,2,1,59,1,3,2,1,17,1,13,1,1,1,2,1,1,4,2,6,2,81,17...
22:54:18 <oerjan> @tell quintopia 89/84 is not just a good enough approximation of 2**(1/12), but a continued fraction cutoff.
22:54:40 <oerjan> > 1+1/(16+1/(1+1/(4+1/2)))
22:54:47 <ais523> OK, so restarting my client fixed it
22:54:48 <oerjan> > 1+1/(16+1/(1+1/(4+1/2)))::Rational
22:55:43 <fizzie> oerjan: The hint file explicitly mentions that "I found this approximation by using Stern-Brocot tree", for the record.
22:55:57 <oerjan> fizzie: that would do it.
22:57:34 -!- Sprocklem has joined.
23:03:34 <lambdabot> [0.0,1.0,0.5403023058681398,0.8575532158463934,0.6542897904977791,0.79348035...
23:03:50 <FreeFull> I wonder how well it approximates
23:05:55 <nooodl> 0.739085133215160641655312087673873404013
23:06:29 <FreeFull> nooodl: I mean, in the continued fraction kind of way
23:06:35 <FreeFull> I don't even know what the continued fraction for it would be
23:08:22 <oerjan> > let cfrac x | isInfinite x = [] | otherwise = f : cfrac (1/(x - fromIntegral f)) where f = floor x in cfrac 0.739085133215160641655312087673873404013
23:08:23 <lambdabot> [0,1,2,1,4,1,40,1,9,4,2,1,15,2,12,1,10,1,2,13,1,2,1,4,6,1,2,1,20,3,1,1,2,1,2...
23:09:27 <oerjan> > let cfrac x | isInfinite x = [] | otherwise = f : cfrac (1/(x - fromIntegral f)) where f = floor x in cfrac (0.739085133215160641655312087673873404013 :: CReal)
23:09:30 <FreeFull> I'm not sure what kind of pattern that is
23:09:34 <oerjan> > let cfrac x | isInfinite x = [] | otherwise = f : cfrac (1/(x - fromIntegral f)) where f = floor x in cfrac (0.739085133215160641655312087673873404013 :: CReal)
23:09:40 <oerjan> > let cfrac x | isInfinite x = [] | otherwise = f : cfrac (1/(x - fromIntegral f)) where f = floor x in cfrac (0.739085133215160641655312087673873404013 :: Float)
23:09:41 <lambdabot> [0,1,2,1,4,1,40,1,8,5,2,3,19,2,10,1,1,3,3,4,1,1,21,1,6,2,2,1,5,32,4,15,2,3,1...
23:10:02 <oerjan> FreeFull: quite likely there is no pattern.
23:10:37 <oerjan> > let cfrac x | isInfinite x = [] | otherwise = f : cfrac (1/(x - fromIntegral f)) where f = floor x in cfrac (exp 1)
23:10:38 <lambdabot> [2,1,2,1,1,4,1,1,6,1,1,8,1,1,10,1,1,12,1,1,11,3,2,1,3,1,73,6,1,1,1,1,1,2,31,...
23:10:53 <FreeFull> I mean, the fixed point of cos one
23:11:51 <FreeFull> That looks like a pattern until the 11,3,2 part
23:12:22 <oerjan> yes. that's because of floating point error, e actually has a regular pattern.
23:12:27 <oerjan> > let cfrac x | isInfinite x = [] | otherwise = f : cfrac (1/(x - fromIntegral f)) where f = floor x in cfrac (exp 2)
23:12:28 <lambdabot> [7,2,1,1,3,18,5,1,1,6,30,8,1,1,12,1,1,24,1,1,1,19,92,3,3,1,1,1,1,1,5,4,3,1,1...
23:12:59 <FreeFull> The 1,1 is an occuring pattern though
23:13:14 <oerjan> yes, it does look like there's something 1,1,3*2^n going on
23:13:51 <Taneb> Then it goes 1 1 1 19
23:13:59 <oerjan> but with strangely random intermediate steps.
23:14:22 <oerjan> i think the 19 is probably after floating point errors start to hurt
23:14:41 <FreeFull> Wolfram alpha gives an accurate continued fraction for the cos fixed point
23:15:00 <FreeFull> I'm not sure what the pattern is though
23:15:00 <oerjan> > let cfrac x | isInfinite x = [] | otherwise = f : cfrac (1/(x - fromIntegral f)) where f = floor x in take 50 $ cfrac (exp 2 :: CReal)
23:15:11 <oerjan> > let cfrac x | isInfinite x = [] | otherwise = f : cfrac (1/(x - fromIntegral f)) where f = floor x in take 30 $ cfrac (exp 2 :: CReal)
23:15:13 <lambdabot> [7,2,1,1,3,18,5,1,1,6,30,8,1,1,9,42,11,1,1,12,54,14,1,1,15,66,17,1,1,18]
23:15:25 <FreeFull> You get mostly small numbers with large numbers interspersed
23:15:41 <oerjan> it seems the 12 was a fluke
23:16:18 <oerjan> ok that's more regular than it looked at first
23:16:27 -!- Bike has quit (Ping timeout: 276 seconds).
23:17:13 <FreeFull> And you see patterns with multiples of 3
23:17:16 <oerjan> although the intermediate numbers still look a little random
23:18:04 <oerjan> no wait 18,30,42 increase by 12 each step
23:18:05 -!- Bike has joined.
23:18:31 <Taneb> 5,6,8,9,12,14,15,18
23:18:31 <FreeFull> 7,2, then 1,1,3n,6+12n,2+3n repeating
23:18:31 <oerjan> and 5,8,11 by 3. so it's entirely regular.
23:21:04 <FreeFull> > let esquared = 7:2:concatMap (\n -> [1,1,3*n,6+12*n,2+3*n])[1..] in esquared
23:21:05 <lambdabot> [7,2,1,1,3,18,5,1,1,6,30,8,1,1,9,42,11,1,1,12,54,14,1,1,15,66,17,1,1,18,78,2...
23:22:08 <oerjan> > foldr (\x y -> x + 1/y) [1..50]
23:22:10 <lambdabot> No instance for (Data.Typeable.Internal.Typeable t0)
23:22:10 <lambdabot> arising from a use of `M187550738534485867012888.show_M1875507385344858670...
23:22:10 <lambdabot> The type variable `t0' is ambiguous
23:22:10 <lambdabot> Possible fix: add a type signature that fixes these type variable(s)
23:22:10 <lambdabot> Note: there are several potential instances:
23:22:17 <Bike> wooooooooooooo
23:22:21 <oerjan> > foldr1 (\x y -> x + 1/y) [1..50]
23:23:42 <oerjan> if this kind of step linear sequence is related to e in general, then so should this number.
23:24:52 <oerjan> > foldr1 (\x y -> x + 1/y) [1..100]
23:26:12 <oerjan> inverse symbolic calculator says it might be BesI(0,2)/BesI(1,2)
23:26:29 <oerjan> something bessier i suspect
23:26:47 <oerjan> it seems to use maple functions
23:26:47 <Taneb> If it wasn't raining I'd go for a wander
23:30:49 <oerjan> oh it's bessel. wikipedia mentions it https://en.wikipedia.org/wiki/Continued_fraction#Regular_patterns_in_continued_fractions
23:30:52 <Taneb> Oh wow, it's really tipping it down
23:31:51 <FreeFull> Is it true that cos(pi/5) = phi/2?
23:32:11 <oerjan> FreeFull: hm something like that, yes.
23:32:49 <oerjan> i recall that the golden section is involved in constructing regular pentagons.
23:33:05 <quintopia> hi oerjan i'm still here in channel. hi lambdabot.
23:33:28 <oerjan> quintopia: oh right you were complaining. i also said i might not remember it.
23:34:23 <FreeFull> Is freenode still being ddosed?
23:35:50 <oerjan> FreeFull: in fact i recall there's a very nice visual proof that the golden ratio is irrational based on inscribing pentagons in each other. even simpler than for sqrt(2).
23:36:11 <oerjan> or possibly five-pointed stars.
23:38:59 -!- atrapado has joined.
23:46:45 -!- Bike has quit (Ping timeout: 248 seconds).
23:48:48 -!- Bike has joined.
23:50:40 -!- nooodl has quit (Quit: Ik ga weg).