00:23:44 -!- copumpkin has quit (Quit: Computer has gone to sleep.).
00:24:28 -!- augur has joined.
00:35:18 -!- augur has quit (Remote host closed the connection).
00:59:10 -!- copumpkin has joined.
01:00:49 <pikhq> elliott: Can you tell me why this doesn't work as expected?; : foreach *.c | $(deps-%f) |> !cc |>
01:02:53 <pikhq> Seems that variables aren't expanded in order-only dependencies.
01:03:36 <pikhq> No, no, it's the %-bit that isn't.
01:07:46 <coppro> pikhq: are you talking to someone?
01:09:17 -!- variable has quit (Ping timeout: 260 seconds).
01:11:43 -!- variable has joined.
01:25:29 -!- lament has joined.
01:29:09 <pikhq> coppro: Talking at elliott.
01:36:09 * oerjan pokes elliott with a stick, revealing a stuffed dummy
01:45:22 <tswett> Eturauhasen syöpä on eturauhasessa esiintyvä syöpä.
01:45:42 <oerjan> `translatefromto fi en Eturauhasen syöpä on eturauhasessa esiintyvä syöpä.
01:45:47 <HackEgo> Prostate cancer is the presence of prostate cancer.
01:45:58 <oerjan> that sounds reasonable
01:46:14 -!- augur has joined.
01:46:45 <oerjan> `translatefromto fi en Koiran syöpä on koirassa esiintyvä syöpä.
01:46:47 <HackEgo> Dog cancer is the presence of dog cancer.
01:46:58 <oerjan> yay i get finnish grammar!
01:47:06 <tswett> `translate Riski sairastua siihen kasvaa 50 ikävuoden jälkeen.
01:47:07 <HackEgo> The risk of developing it increases after the age of 50.
01:47:34 <lament> `translate perkele perkele perkele
01:47:46 <oerjan> `translate Koira sairastua siihen kasvaa 50 ikävuoden koiraan.
01:47:48 <HackEgo> Sick dog to the age of 50 will increase your dog.
01:49:14 <tswett> `translate Vittun syöpä on vittussa esiintyvä syöpä.
01:49:15 <HackEgo> Vittun vittussa cancer is the presence of cancer.
01:49:16 <oerjan> `translate Koira sairastua siihen kasvaa 50 koiran jälkeen.
01:49:18 <HackEgo> Sick dog to the dog up by 50 after.
01:49:39 <tswett> `translate Banaanikärpänen (Drosophila melanogaster) on pieni kellanruskea mahlakärpäslaji, joka on jo vuosikymmenten ajan ollut yksi kokeellisen biologisen tutkimuksen tärkeimmistä malliorganismeista.
01:49:41 <HackEgo> Drosophila (Drosophila melanogaster) is a small fawn mahlakärpäslaji, which has for decades been one of the most important experimental biological research malliorganismeista.
01:49:58 <tswett> Meh, these aren't as good.
01:51:29 <tswett> `translatefromto en fi fly
01:51:41 <tswett> `translatefromto en fi Fly, fly!
01:52:13 <tswett> `translatefromto en fi Please please fly a fly a fly.
01:52:14 <HackEgo> Ole hyvä lentää lentää lentää.
01:53:56 <oerjan> `translatefromto en fi Time flies like an arrow, fruit flies like a banana.
01:53:57 <HackEgo> Aika lentää kuin nuoli, hedelmä lentää kuin banaani.
01:54:30 <tswett> `translate hedelmä lentää kuin banaani
01:54:40 <elliott> <pikhq> elliott: Can you tell me why this doesn't work as expected?; : foreach *.c | $(deps-%f) |> !cc |>
01:54:49 <elliott> * oerjan pokes elliott with a stick, revealing a stuffed dummy
01:54:57 <elliott> oerjan: do you really think iwc will end at the candh point?
01:55:25 <pikhq> elliott: So as to be able to specify the ordered dependencies of something without requiring a lot of repetition.
01:55:49 <pikhq> elliott: Say, "such and such files depend on a autogenerated header file".
01:57:21 <pikhq> Also, thinking about it, I'm pretty sure that Tup isn't very nice for Haskell. Not, mind you, that that's really all that necessary.
01:57:37 <pikhq> After all, ghc --make or darcs.
01:59:15 <elliott> torrent is ninety four percent done
01:59:20 <elliott> and all i'm doing is seeding
02:00:55 <oerjan> as long as you aren't seething
02:02:15 -!- variable has quit (Read error: Operation timed out).
02:06:33 -!- kwertii has joined.
02:06:50 <elliott> ok this is ridiculous. SOMEONE UPLOAD TO ME
02:09:59 <Sgeo> elliott, what torrent?
02:10:09 <elliott> five billion gigaquads of goat porn
02:10:54 <oerjan> obviously fake, there aren't that many goats
02:11:06 <elliott> one goat can be in multiple porns
02:11:21 <tswett> It's actually just a single JPEG.
02:11:34 <elliott> no you don't. only i may see the goat porn.
02:11:40 <tswett> Created by taking a JPEG of goat porn and enlarging it.
02:11:51 <Lymia> porn.................
02:12:01 <elliott> Lymia: i see you're new here
02:12:12 <Lymia> My brain is scarred.
02:12:17 <oerjan> Lymia: i think of it as our local euphemism. i prefer not to think of what.
02:12:19 <Lymia> elliott, I think I've just succeed a few luck rolls.
02:12:38 <elliott> gotta love any channel where goat porn is the _euphemism_
02:12:47 <tswett> `translate Todos mis llaves son diablos.
02:13:19 <elliott> i wonder if it'd work without waiting for it all to download
02:13:22 <elliott> i'm sure isos have 0s on the end of them
02:16:01 <pikhq> Goatpornix, you mean.
02:16:32 <pikhq> Has a real *woody* sound to it, doesn't it? Gooorrrnix.
02:16:44 <elliott> pikhq: well it certainly gives me a woody
02:16:53 <oerjan> goatpornix, a rarely seen asterix character
02:17:14 <pikhq> oerjan: Insufficiently punny for Asterix, though.
02:17:23 <elliott> what's not punny about poat gorn
02:17:35 <pikhq> (yay, being one of a few Americans that has read Asterix)
02:19:30 <pikhq> ((yes, Asterix is basically unknown in the US))
02:19:54 <pikhq> (((much like Disney comics)))
02:24:11 -!- variable has joined.
02:28:27 <pikhq> Graphviz has been running for quite a while.
02:28:44 <pikhq> This is going to be an absurd PNG.
02:30:43 <pikhq> The result of tup g . in ~/bsnes-tup. Because I was just curious how fucking abusrd it would be.
02:31:56 <pikhq> dot: graph is too large for cairo-renderer bitmaps. Scaling by 0.12547 to fit
02:31:58 -!- azaq231 has joined.
02:32:28 <pikhq> 32767x182 PNG, entirely black.
02:33:33 -!- azaq23 has quit (Ping timeout: 252 seconds).
02:43:20 -!- pikhq has quit (Quit: Lost terminal).
02:43:30 <elliott> does bsnes really have that many dependencies?
02:45:17 <oerjan> well obviously they mean bsnes
02:45:24 <elliott> note to self: VM password is outside password, but with the words spelled out
02:45:33 <elliott> with the numbers spelled out
02:59:50 -!- elliott has left ("Leaving").
03:00:37 -!- elliott has quit (Remote host closed the connection).
03:16:46 -!- azaq231 has quit (Ping timeout: 250 seconds).
03:16:56 -!- augur has quit (Remote host closed the connection).
03:51:00 -!- TeruFSX2 has quit (Remote host closed the connection).
03:51:12 -!- augur has joined.
04:23:18 -!- pikhq has joined.
04:23:35 <pikhq> Gregor: Nice work on the topic.
04:23:41 <pikhq> elliott definitely needs a sense of fnarf.
04:29:22 -!- asiekierka has joined.
04:39:07 <pikhq> ... There are atheist members of Judaism (the religion)?!
04:40:01 <pikhq> I have no idea how the heck that works as a religious faith without a deity.
04:40:57 <pikhq> I mean, pretty much you're left with a list of traditions.
04:50:10 <Ilari> Hmmm... 25 years ago today: Chernobyl disaster. :-/
04:53:15 -!- zzo38 has joined.
04:54:46 <pikhq> As bad as the disaster itself was, no doubt the *worst* side effect of it was the complete end of new nuclear power...
04:55:21 <pikhq> Which has made nuclear power fairly unsafe, really. Most of the plants currently running *really* need to be shut down.
04:55:44 <pikhq> And we wouldn't have the Fukushima incident...
04:56:29 <coppro> Or at least if we did, we would be dealing with it better
04:56:34 <coppro> because better technologies would exist
04:56:39 <zzo38> Yes, there are better ways including wind, solar, and more. Someone told me wind is best.
04:56:46 <pikhq> coppro: No, we *really wouldn't* have the Fukushima incident.
04:56:53 <pikhq> coppro: Because the reactor would have been shut down. :P
04:57:11 <coppro> zzo38: Nuclear power is /actually/ the only sane long-term approach
04:57:13 <Ilari> At least chernobyl was quickly dealt with. Fukushima... Not so much. The accident was March 11, and the reactors are still spewing contamination.
04:57:29 <pikhq> It was actually scheduled to be shut down, but extended because the grid really needed the power.
04:57:58 <pikhq> Ilari: On the other hand, they evacuated the area before there could be notable health risks.
04:58:05 <pikhq> Ilari: Unlike Chernobyl.
04:58:29 <zzo38> coppro: I think there are better ways! (even if only insane people can use them, which I really doubt)
04:59:08 <coppro> zzo38: wind is okay, but a single wind generator doesn't produce much usuable power
04:59:11 <zzo38> Someone told me wind is best. But I think it is not best by itself, but it is good when combined with solar and so on.
04:59:22 <coppro> and they take up a lot of space
04:59:38 <pikhq> coppro: I'd imagine that in the *future* renewable sources could be entirely viable. Most probably solar and extensive storage.
05:00:01 <zzo38> I know solar is inefficient it does not make much power. Which is why you have to combine wind with solar.
05:00:05 <Ilari> PV is not good, Concentrated solar sould be relatively OK, at least it some places.
05:00:22 <pikhq> I highly doubt it'd be photovoltaic.
05:00:33 <pikhq> Alternately, fusion power...
05:00:36 <coppro> zzo38: No, it's inefficient in that current technologies produce less than is used to make the panel
05:00:45 <coppro> fusion is just nuclear with less risk
05:00:52 <Ilari> After all, constructing large mirror arrays is much easier than constructing large PV arrays.
05:01:21 <pikhq> But in any case. Of the *present* technologies, nuclear power is the only one that meets our needs and will last us more than ~20 years.
05:01:34 -!- elliott has quit (Read error: Connection reset by peer).
05:02:21 <pikhq> Oh, sorry, there's about 100 years of coal reserves.
05:02:47 <pikhq> Mind you, we *really* need to get off of coal.
05:03:00 <coppro> Concentrated solar is good
05:03:10 <Ilari> At what rate? At what price? ... Oh, and that would do real "fun" things to the climate.
05:03:19 <coppro> but I doubt it's economical to run our grid off of it
05:03:47 <coppro> Ilari: Although ideally solar is still meaning that the net energy capture rate of earth is higher than before
05:03:47 <pikhq> Ilari: As I said, we really need to get off coal.
05:03:59 <coppro> pikhq: I think he was talking about solar
05:04:10 <coppro> running our power economy entirely on solar would seriously affect climate
05:04:15 <Ilari> But if you want to do _REALLY_ "fun" things to climate: Methane from Methane clathrates.
05:04:30 <pikhq> It doesn't matter *that* much that coal could last us ~100 years, because that shit is fucking dirty and dangerous.
05:04:43 <pikhq> coppro: True, it quite probably would.
05:05:27 <pikhq> coppro: Though given that a large portion of our power economy is actually causing a greenhouse effect ATM, it doesn't seem like it'd have a *worse* affect than our current scheme.
05:05:40 <coppro> pikhq: True, and the two might well cancel out
05:06:00 <Ilari> If a nuclear power plant spewed radioactive stuff like a coal plant does, it would get permissions revoked really really fast.
05:06:20 <coppro> but nuclear plants don't even generate that much stuff
05:06:39 <pikhq> Not to mention that mining coal is very dangerous.
05:06:43 <coppro> they are the superior technology in every respect except one
05:07:03 <Ilari> Relatively little stuff, but that stuff is frickin bad. And "reprocesing" is a taboo.
05:08:16 <zzo38> Maybe hydroelectric
05:08:33 <pikhq> Hydroelectric is a complete *nightmare* on ecosystems.
05:08:34 <Ilari> Also, advanced nuclear power technology is good way towards the bomb. So that tends to be looked down on.
05:09:31 <Ilari> The three things about any power source: How much we have? At what rate? At what price?
05:09:50 <coppro> power our society off of nuclear waste
05:10:03 <zzo38> Because there are three things, it means you need three sources.
05:10:16 <pikhq> coppro: That's called "nuclear reprocessing". :P
05:10:25 <coppro> pikhq: No, I mean the heat
05:10:39 <coppro> that stuff generates lots of heat
05:10:45 <pikhq> That's called "inefficient".
05:10:47 <Ilari> Also, there's neutron transmutation. But there's relatively little engineering on that.
05:10:56 <coppro> pikhq: Why? It's there and we're doing nothing with it
05:11:25 <pikhq> Okay, true. It's an improvement over literally doing nothing.
05:11:36 <zzo38> coppro: That might work (a bit).
05:11:44 <pikhq> But if we wanted to be *sane and rational* about it, we'd be using it as fuel in a reactor.
05:15:13 <coppro> but aren't some of the lighter products really difficult and inefficient to reprocess?
05:16:42 <oerjan> only the heavy elements can be reprocessed
05:17:04 <oerjan> the lighter ones mostly have _much_ shorter half-lives though, iirc
05:17:44 <coppro> cesium-137 seems to be pretty bad; really tough to handle and a half-life of 30 years
05:18:09 <coppro> strontium-90 is the other component of high-level waste; that stuff is easier to handle and can be usefule
05:18:11 <oerjan> um 30 years is _nothing_, i mean much shorter as opposed to _thousands_ of years
05:18:16 <coppro> according to wikipedia
05:18:32 <coppro> oerjan: The ones with huge half-lives are not very radioactive though
05:19:26 <pikhq> oerjan: The short-lived ones are just a royal pain *while they're still around*.
05:19:42 <oerjan> no but they are the ones that make people worry about having to store stuff long enough that our descendants may have forgotten their there...
05:20:17 <pikhq> Don't need to pull any Yucca Mountain shit to hide them away from society for eons, *but* it takes some extensive facilities.
05:21:33 <oerjan> anyway i read that we can run reactors for hundreds of years by reprocessing waste that has already been _made_
05:40:21 <Sgeo> About to take a Tylenol
05:53:03 -!- Lymia has quit (Read error: Operation timed out).
05:53:17 -!- Lymia has joined.
05:53:18 -!- Lymia has quit (Changing host).
05:53:18 -!- Lymia has joined.
05:54:36 -!- elliott has joined.
06:02:39 <zzo38> Did you know I made up a new D&D character for a separate campaign? I wrote a program on my TI-92 calculator to make up the name and ability scores of the character.
06:02:51 <zzo38> (These are two separate programs)
06:04:12 <zzo38> What is "ftgy"? (Also, I do not know your specific questions/comments about calling conventions)
06:06:13 <elliott> oiajfghoptejhiuroepksfdgjsokjbnkdospnb
06:06:48 -!- kwertii has quit (Quit: bye).
06:07:00 <pikhq> elliott: KAN YUU FIIRU? KAN YUU FIIRU ZATTO HAIBURIDDO REINBOU?
06:07:31 * pikhq mutters. Engrish. Why must it be common?
06:07:34 -!- myndzi\ has joined.
06:07:38 <zzo38> elliott: I am not understand your question sorry?
06:07:45 <elliott> zzo38: its a good question
06:08:16 <elliott> And its a good question is the best a good question in the history of all the as good question, ever.
06:09:43 <zzo38> Maybe, but it is not the one that I understand.
06:09:53 <pikhq> (quote transcribed from "ハイブリッドレインボウ", by the pillows, from /LITTLE BUSTERS/)
06:10:17 <pikhq> (Yes, that is three different English abuses in a row)
06:10:20 <elliott> zzo38: Well, the a good question that is its a good question is one of the as good question, the assortment of all the best questions that can be said to be a, as in the adjective a, meaning that the as a good question of which its a good question is part of is a set of as good question.
06:10:35 -!- oerjan has quit (Quit: AI MUST FIIRU).
06:10:52 <elliott> none of you will ever realise the beauty of that statement
06:11:19 <pikhq> elliott: Use/mention.
06:11:27 -!- myndzi has quit (Ping timeout: 260 seconds).
06:11:34 <elliott> pikhq: no use/mention distinction there
06:11:41 <elliott> apart from that, it's all use
06:12:00 <elliott> a good question : as good question :: surgeon general : surgeons general (compare a good questions, surgeons general)
06:12:31 <zzo38> OK, but that isn't a question.
06:13:56 <pikhq> elliott: In "surgeons general", the plural is on the noun.
06:14:10 <pikhq> Fossilized French-style adjective.
06:14:48 <pikhq> Though, as it so happens, the Surgeon General *is* a general.
06:23:11 -!- aloril has quit (Ping timeout: 240 seconds).
06:25:22 <zzo38> OK they are a general, but are they a surgeon? If not, "surgeons general" seem shouldn't sensible!
06:25:38 <zzo38> If they are, then both "surgeons general" and "surgeon generals" should be acceptabpe.
06:29:06 <elliott> surgeon general is close enough to a surgeon for the purpose here.
06:29:12 <elliott> surgeons general is the moar correct option
06:31:34 <pikhq> zzo38: He is both a surgeon and a general.
06:32:39 <pikhq> Namely, a general in the Public Health Service Commissioned Corps.
06:32:51 <zzo38> pikhq: Then both "surgeons general" and "surgeon generals" ought to be correct (even if it isn't).
06:33:28 <zzo38> At least, that is my logic on this idea, anyways
06:33:44 <pikhq> Either the first or second-highest ranking officer in the PHSCC, depending on whether the present Assistant Secretary for Health has opted to serve as a civilian.
06:35:30 -!- aloril has joined.
06:39:54 -!- comex has quit (Ping timeout: 276 seconds).
06:42:05 -!- comex has joined.
06:55:09 -!- Tritonio_Droid has quit (Quit: Bye).
07:12:05 -!- pikhq has quit (Quit: Lost terminal).
07:19:18 -!- zzo38 has quit (Remote host closed the connection).
07:21:55 -!- MigoMipo has joined.
07:24:13 * Sgeo wanting to puke right now is probably unrelated to the Tylenol, right?
07:31:21 <elliott> it's aboslutely the tylenol's fault
07:31:25 <elliott> tylenol is known to cause puking
07:31:33 <elliott> usually it means something very bad
07:31:36 <elliott> i would go to the hospital
07:31:39 <elliott> and get a doctor to stop you from dying
07:31:44 <elliott> which is at this point likely
07:31:58 -!- lament has quit (Ping timeout: 240 seconds).
07:36:26 <Sgeo> Maybe it's the lack of sleep?/
08:03:51 -!- Fuco has joined.
08:29:33 -!- MigoMipo has quit (Read error: Connection reset by peer).
08:51:35 -!- augur has quit (Remote host closed the connection).
08:53:45 -!- augur has joined.
09:25:55 -!- monqy has quit (Quit: hello).
09:27:11 -!- MigoMipo has joined.
09:35:06 -!- pingveno has quit (Ping timeout: 258 seconds).
09:35:52 -!- pingveno_ has quit (Ping timeout: 276 seconds).
09:42:03 -!- el_gigante has joined.
09:43:20 <elliott> why did you join this channel
09:43:24 <elliott> if you were only planning to leave it
09:44:28 -!- el_gigante has quit (Remote host closed the connection).
09:54:27 <elliott> Gregor: link me to superturing plz
09:54:35 <elliott> need it for serious purposes
10:50:25 -!- pingveno has joined.
10:56:59 -!- pingveno1 has joined.
11:27:44 -!- MigoMipo has quit (Remote host closed the connection).
11:35:39 <cheater99> lol: http://emilis.github.com/2011/04/12/usefullnes-of-inventing-languages.html
11:58:09 -!- MigoMipo has joined.
11:58:57 -!- siracusa has quit (Ping timeout: 250 seconds).
12:00:21 -!- siracusa has joined.
12:03:19 -!- crystal-cola has joined.
12:09:18 <Lymia> cheater99, readable, but hard to write.
12:09:23 <Lymia> Plus, nothing an IDE can do.
12:13:22 <crystal-cola> I'll just tell it because it's fun: You flip a coin over and over again... you're only allowed to stop if at some point you've got >= 2/3 heads. What is the probability that you stop?
12:51:05 -!- FireFly has joined.
13:01:31 <Lymia> I don't know how the fuck.
13:01:36 <Lymia> I think it's going to end up as a limit.
13:02:59 <Lymia> I /expect/ that it will either be a sum or an limit to infinity.
13:10:00 -!- zzo38 has joined.
13:14:49 <cheater99> 1. go into #php and say C is a high level programming language
13:52:23 <zzo38> Is this good list of equipment I selected in this D&D game? silver candelabra for holding 4 candles, false book, small magnet, slate board, folding table case, 5x wick (50ft), 5x twine (50ft), 4x plain copper ring, 7 pieces of chalk (each one a different color), 30 marbles (6 clear, 6 stone, 6 clay, 6 red, 6 green).
13:52:36 <zzo38> Total weight: 12 pounds.
14:00:37 <cheater99> sounds like you're playing a magician
14:03:08 <zzo38> Is that what it seems when you read this list?
14:15:26 -!- copumpkin has quit (Quit: Computer has gone to sleep.).
14:25:09 -!- Tritonio has joined.
14:32:51 <zzo38> Did you believe this?
14:35:01 <zzo38> Tritonio: Do you even know what it is? Do you know why you joined this channel?
14:35:53 <Tritonio> I know what this room is. Although I don't know what you were talking about.
14:36:45 <zzo38> Good, because I was not addressing you with my other question. Although, you type the your own question for esoteric programming or whatever else.
14:37:10 -!- Sgeo has quit (Ping timeout: 258 seconds).
14:37:48 <elliott> Tritonio: welcome to zzo38
14:39:13 <Tritonio> So.... I've got a question. Is there ever going to be a brainfuck golf contest?
14:40:21 <Gregor> There have been BF gold tourneys in the past *shrugs*
14:41:22 <zzo38> They do have Brainfuck running in anarchy golf, although it is different than the Brainfuck Golf contests
14:41:52 <Tritonio> Yeap. Is there ever again going to be one?
14:42:41 <zzo38> I don't know if there is ever again going to be one.
14:43:29 <elliott> Tritonio: anarchy golf is http://golf.shinh.org/
14:43:43 <elliott> Tritonio: re another one - I doubt it, but ask dbc :P
14:44:04 -!- Tritonio_Droid has joined.
14:45:29 -!- copumpkin has joined.
14:47:15 -!- Tritonio has quit (Ping timeout: 260 seconds).
14:47:50 <Tritonio_Droid> Are toy sure this page works? It light he my reception though...
14:49:10 <Gregor> Are toy sure, \ this page works? \ It light, he \ my reception. \ Though ...
14:49:48 -!- Tritonio has joined.
14:49:50 <elliott> Gregor are toy \ Sure, this page works? It \ Light, he my \ reception. Though... A poem \ by Tritonio.
14:51:07 <Gregor> elliott \ Gregor \ are toy \ sure, \ this page works? \ It \ Light, he \ my \ reception. \ Though ... \ A poem \ by Tritonio. \ A poem \ by Gregor \ backslash
14:51:17 <Gregor> (Plagiarized from elliott)
14:51:44 <elliott> Gregor elliott gregor are \ Toy sure this \ page works it light he my \ reception though a poem by \ tritonio a \ poem by gregor \ backslash \backslash\ a poem by gregor \ plagiarsed
14:52:17 <Tritonio> Lolololol just saw what I wrote. Never mind... :)
14:52:26 <Gregor> "Someone plagiarized my poem and didn't give me credit." -- A poem by an anonymous author
14:52:30 -!- Tritonio_Droid has quit (Ping timeout: 260 seconds).
14:53:43 -!- Tritonio has quit (Client Quit).
14:54:04 -!- ais523 has joined.
14:54:36 <Gregor> ais523, unaffiliated. ais523, has joined. Esoteric.
14:54:45 <Gregor> -- A poem by OK, seriously I'll stop now :P
14:56:16 <ais523> hmm, I take it you've been doing that for everyone joining for hours, from your attribution line
15:04:35 -!- Phantom_Hoover has joined.
15:08:51 -!- BeholdMyGlory has joined.
15:13:27 -!- augur has quit (Remote host closed the connection).
15:14:46 <crystal-cola> not P or not P or not P or not P? - A poem by me
15:20:01 -!- ais523 has quit (Ping timeout: 250 seconds).
15:20:13 -!- ais523 has joined.
15:36:07 -!- variable has quit (Ping timeout: 260 seconds).
15:40:01 -!- lament has joined.
15:44:15 <Ilari> APNIC: no allocations.
15:47:23 <ais523> do they have anything left to allocate?
15:48:43 -!- augur has joined.
15:51:01 -!- lament has quit (Ping timeout: 252 seconds).
15:52:20 <Ilari> 0.989 blocks of IPv4 space. 0.991 blocks of IPv6 space.
15:53:20 <ais523> hmm, so they haven't run out of v4 space after all?
15:53:43 <ais523> or was a block returned, or something?
15:53:58 <Ilari> Well, they de-facto have, since allocations are heavily restricted.
15:54:10 <Ilari> lifthrasiir: IPv6 block is /12.
15:54:12 <lifthrasiir> ais523: i think Ilari meant the APNIC unallocated RIR pool
15:54:38 <lifthrasiir> so the global pool is over, but there is space left in the RIR pool
15:55:47 <Ilari> Well, only RIPE, ARIN, LACNIC and AFRINIC allocate IPv4 normally anymore. As for IANA, it is out of IPv4 blocks, but has 506 IPv6 blocks.
15:58:02 <crystal-cola> 1,2,4,8,16,32,64,68,128,136,256,272,512,544,1024,1088,1156,2048,2176,2312,2824,3104,4096
15:58:52 <Ilari> 5.054 blocks of IPv6 space allocated by IANA to RIRs total.
15:59:14 <Ilari> crystal-cola: Seems to at least have some of form 68*2^r.
16:00:41 <Ilari> And also numbers not of neither 2^r nor 68*2^r...
16:01:02 <ais523> it seems weird to factor out two 4s like that
16:02:08 <crystal-cola> some of the later ones have factors 97, 137 or 353
16:02:10 <myndzi\> i thought the point of ipv6 was to have a ridiculous amount of addresses
16:02:18 <myndzi\> why would anyone be low on ipv6 blocks?
16:03:12 <Ilari> myndzi\: Nobody is low on them.
16:03:32 <crystal-cola> http://pastie.org/pastes/1835747/text here are the factorizations of the first few of these
16:04:06 <myndzi\> i should think not, but i guess i wasn't comprehending very well
16:05:21 <Ilari> I also get that n can be 128*257.
16:06:03 <ais523> it seems to allow some numbers of the form 2^n(2^n+1)
16:06:06 <Ilari> myndzi\: Unlike IPv4 blocks which aren't all that large, IPv6 block is loads of address space.
16:06:30 <myndzi\> yeah, but 506 sounds awful scarce
16:06:50 <myndzi\> though i guess they'd be doling them out the same way as v4
16:06:53 <ais523> well, a /32 contains enough space for everyone currently connected to the Internet by itself
16:07:01 -!- myndzi\ has changed nick to myndzi.
16:07:08 <ais523> and there are over a million of those in a /12
16:07:23 <myndzi> i understand how vast the address space is ;) that was the source of my confusion
16:08:14 -!- variable has joined.
16:08:53 <Ilari> Global RIR IPv6 allocations so far: 0.142 blocks. And thats nearly 10 billion /48s.
16:09:58 <crystal-cola> I tried plotting 2^n mod n but I don't see anything
16:13:21 <Ilari> And looks like if k is a solution, then all k*2^r are too.
16:14:59 <crystal-cola> oh that would explain why there are so many powers of 2 in there
16:16:29 <Ilari> Not all of those are of form 2^r or p*2^r: 1156 = 2^2 * 17^2.
16:18:55 -!- cheater99 has quit (Ping timeout: 240 seconds).
16:18:59 -!- Sgeo has joined.
16:21:37 -!- cheater99 has joined.
16:23:25 <Ilari> Wonder if there is some r such that 257^2*2^r is a solution. Or 17*257*2^r?
16:25:12 <Ilari> Indeed, there is for latter: r=7.
16:25:41 <Ilari> Well, if r=7 is a solution, then every r>=7 is.
16:28:52 -!- zzo38 has quit (Quit: Abort, Retry, Fail?).
16:37:11 <Ilari> 257*257*2^7 also is a solution.
16:38:57 <Ilari> Except can any number of form 3*2^r be a solution?
16:39:44 <crystal-cola> I haven't seen any solutions of that form but I can't rule it out
16:42:16 <Ilari> If it is, then 3*2^r|(5^(3*2^r)-3^(3*2^r)) => 3*2^r|(5^(2^r)-3^(2^r))(25^(2^r)+15^(2^r)+9^(2^r)). Since 3 is prime, it must divide one of factors. But neither can be divisible by 3, thus there are no solutions of form 3*2^r.
16:45:35 <Ilari> It seems that the solutions are (I haven't proved this): 2^r, 17^a*2^r (r >= 2), 17^a*257^b*2^r (r >= 7) and 17^a*257^b*65537^c*2^r (r >= 15).
16:45:58 <crystal-cola> but there are some solutions with the primes 97,137,353 in them
16:46:46 <crystal-cola> my guess is that trick with the difference of cubes may also work for any fermat prime
16:47:38 <crystal-cola> oh, although it matters that one of the bases is 3
16:48:30 <Ilari> For 5, the bases in second factor will be 625(0), 375(0), 225(0), 135(0) and 81(1). Thus, there can't be solution of form 5*2^r either.
16:50:02 <Ilari> It almost seems like there are set of number pairs (a,b) such that you can multiply any solution k*2^r by a if r >= b and get a new solution.
16:53:13 <Ilari> So far found of those pairs: (1,0), (17, 2), (97, 5), (257, 7), (353, 3), (65537, 15)
16:53:28 <Ilari> I can't find a pair with a=137.
16:55:33 <crystal-cola> maybe 137 isn't relevant on it's own, since it always seems to be 17*1367
16:55:34 <crystal-cola> maybe 137 isn't relevant on it's own, since it always seems to be 17*137
16:59:48 <Ilari> Ah, there are pairs with non-prime a, such as (2329,2)
17:00:34 <crystal-cola> it seems hopeful that there could be just these pairs and they could generate all solutions
17:00:44 <Ilari> Yeah, but how many of them?
17:00:48 <crystal-cola> although maybe one should compute more to test that
17:01:19 <crystal-cola> well I was guessing these are all of them - but I probably haven't computed far enough to make that guess
17:02:21 <Ilari> At least there doesn't seem to be a pair (but I can't rule one out) with a=2^32+1
17:10:32 <Ilari> Found some more candidates: 136.
17:10:51 <crystal-cola> I am doing a search now but 136 didn't come up
17:11:13 <lifthrasiir> crystal-cola: where does the problem come from?
17:11:30 <lifthrasiir> i think it is quite a lengthy solution compared to the original statement ;)
17:11:50 <crystal-cola> lifthrasiir: yes diophantine seems to be able to compress more information than theoretically possible
17:12:01 <Ilari> That pair is even (136, 0)!
17:12:47 <lifthrasiir> and yeah, diophantine is quite a convoluted subject (at least for me...)
17:13:16 <crystal-cola> I think I should turn on my other computer and let it try to verify the conjecture of Ilari to very large numbers
17:13:36 <crystal-cola> although we probably need to find out exactly what the pairs are first... also this new search makes me think maybe there are infinitely many pairs
17:15:14 <crystal-cola> I should have bene looking at the factorized forms -_-
17:17:20 <Ilari> Oops, 136 is just 2^3*17.
17:17:41 <Ilari> Those numbers can't be even.
17:19:49 <Ilari> They can't be even because those multiples of 2 can be absorbed into r anyway!
17:19:53 <crystal-cola> I wonder if there was a way to look at it like
17:21:26 <Ilari> 23137 with unknown b.
17:22:22 <Ilari> But is there known pair with a=1361?
17:22:29 <crystal-cola> 17*26113 <-- this is odd because 26113 doesnt' seem to appear
17:22:45 <crystal-cola> it does seem like some numbers are "stuck" to 17
17:24:40 <crystal-cola> is it always the case that if k2^r is one then k2^{r+1} is?
17:25:42 <Ilari> crystal-cola: Yes.
17:26:43 <Ilari> 5^(2n)-3^(2n) = (5^n-3^n)(5^n+3^n). The first is divisible by n for some numbers, and the latter is divisible by 2 for any number.
17:31:09 <crystal-cola> so like 5^(17*2^6)-3^(17*2^6) = (5^(17*2^5)+3^(17*2^5))(5^(17*2^4)+3^(17*2^4))(5^(17*2^3)+3^(17*2^3))(5^(17*2^2)+3^(17*2^2))(5^(17*2^1)+3^(17*2^1))(5^17-3^6)
17:32:52 <crystal-cola> %7 = 114794370197489014450071927463032574073464515434335823575495306281286170747368778680268615229504
17:32:55 <crystal-cola> ? (5^(17*2^2)+3^(17*2^2))*(5^(17*2^1)+3^(17*2^1))*(5^(17*2^0)+3^(17*2^0))*(5^(17*2^0)-3^(17*2^0))
17:32:57 <crystal-cola> %8 = 114794370197489014450071927463032574073464515434335823575495306281286170747368778680268615229504
17:33:25 <crystal-cola> this doesn't help because the modulus is not prime
17:33:41 <crystal-cola> I was thinking it would just suffice to consider one of these factors.. but it doesn't
17:34:02 <Ilari> Pairs with a<=10^5, b<=100: (1, 0), (17, 2),(97, 5), (257, 7), (353, 3), (2329, 2), (7681, 8), (23137, 4), (36097, 8), (39899, 2), (65537, 15).
17:34:42 <crystal-cola> hm if 65537 is still the biggest.. do you think that could be all pairs?
17:36:53 <crystal-cola> it's a bit frustrating I can't see any pattern in theses numbers
17:38:29 <Ilari> I just searched OEIS. No luck. Either there's no sequence like that, or there are some small bases with large required r.
17:38:54 <crystal-cola> I was thinking maybe there is some way to take the "limit" of r -> infinity
17:39:15 <Ilari> Even more funky considering that there are some bases that are not prime there.
17:42:29 <Ilari> How could one even prove or disprove that if n1 and n2 are solutions then lcm(n1,n2) is too?
17:44:19 <crystal-cola> like if both 5^a-3^a and 5^b-3^b factr in certain wasy then 5^(lcm(a,b))-3^(lcm(a,b)) factors in those ways too?
17:44:22 -!- variable has changed nick to Guest11470.
17:44:55 -!- monqy has joined.
17:46:10 -!- Guest11470 has changed nick to variable.
17:47:57 <quintopia> i wish i knew what problem this was about
17:48:24 <Ilari> quintopia: For what n, n|5^n-3^n?
17:50:09 <Ilari> Well, 5^(lcm(a,b))-3^(lcm(a,b)) has factors 5^a - 3^a and 5^b - 3^b. a = k1 * 2^r1, b = k2 * 2^r2. Therefore 5^(lcm(a,b))-3^(lcm(a,b)) has factor of lcm(k1, k2) * 2^max(r1, r2). = lcm(a, b).
17:52:03 <Ilari> There's one more thing to prove or disprove: k*2^r => k^l*2^r?
17:52:17 -!- Sgeo has quit (Ping timeout: 258 seconds).
18:02:09 <quintopia> it has no day of the day of the day of the day :/
18:02:18 <Ilari> If so, that should be enough with the others (k*2^r => k*2^(r+1), a and b => lcm(a, b)) to show complete set of pairs could be used to generate all solutions.
18:02:27 <elliott> it has solidity, so shaddup
18:04:22 <olsner> oh, did someone bring up matrices of solidities again?
18:05:11 <Ilari> Unfortunately, some k are non-prime (like that 2329).
18:06:15 -!- pingveno1 has quit (*.net *.split).
18:06:15 -!- z^ck has quit (*.net *.split).
18:06:42 -!- z^ck has joined.
18:09:08 -!- pingveno1 has joined.
18:14:27 -!- Gregor 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.
18:14:38 <Ilari> Oops, nope. Turns out even one 17 is enoug for a lot of 137s.
18:15:24 <elliott> fuckin matrices of solidity
18:16:31 <Ilari> There's r such that 17*137^2*2^r is a solution.
18:17:49 <Gregor> elliott: Similarly to magnets.
18:19:32 <Ilari> That would generate a lot of pairs if one had to generate those that way.
18:21:46 <Ilari> 37733 (97*389) behaves in similar manner.
18:23:50 <Ilari> Wonder if there's a k such that for no r k*2^r is a solution, but for some r, 257*k*2^r is.
18:26:15 <Ilari> Or k^l*2^r, but k*2^r isn't a solution.
18:29:50 <Ilari> And then, is there l such that 17*137^l*2^r isn't solution for any r?
18:31:15 -!- asiekierka has quit (Ping timeout: 260 seconds).
18:32:05 <ais523> elliott: oh, I needed to copy-and-paste a row of 10 8s
18:32:09 <ais523> so I typed it into my IRC client
18:32:19 <ais523> because I wanted to type 141 8s
18:32:26 <ais523> and it's a pain to count them all manually
18:33:23 <Ilari> ais523: Make a script that sends n times some string to PRIMARY?
18:33:39 <ais523> perhaps, but doing it this way is pretty simple
18:33:49 <elliott> isn't that perl and xclip or whatever?
18:41:17 -!- Sgeo has joined.
18:44:41 -!- tim000 has quit (Remote host closed the connection).
18:54:47 -!- Sgeo has quit (Ping timeout: 258 seconds).
19:22:06 -!- oerjan has joined.
19:29:17 -!- tswett has quit (Ping timeout: 258 seconds).
19:29:51 -!- wareya has quit (Read error: Connection reset by peer).
19:30:39 -!- wareya has joined.
19:32:19 -!- augur has quit (Remote host closed the connection).
19:40:40 <Ilari> Also, is there some k that has solutions with two prime factors that don't have solutions for any r?
19:45:43 <Ilari> Well, there's k=17*137, where 137*2^r isn't a solution for any r. But can there be numbers with multiple factors like that 137?
19:48:29 <ais523> number theory is the #esoteric equivalent of TV Tropes
19:49:50 <Slereah> Because number theory is BORING
19:49:57 <crystal-cola> I think that the best thing to do would be to start building up a database of these k values
19:54:47 <Ilari> Also store the smallest r corresponding to each k.
19:54:59 -!- oerjan has quit (Quit: leaving).
19:56:56 <Ilari> Also, one can use formula r(lcm(a,b)) = max(r(a),r(b)) to compress it a bit. But even with that, it is still likely infinite (due to powers).
19:58:59 <crystal-cola> I was just looking at 3 and 5 ^(k 2^r) mod k and mod 2^r sepeartely
19:59:17 <crystal-cola> it appears both are always 1 mod 2^r (I guess that's just eulers totient thing)
19:59:35 <crystal-cola> so the problem could be reduced to just considering mod k (maybe)
20:00:43 <crystal-cola> also it seems like mod k the patterns are periodic
20:02:42 -!- calamari has joined.
20:03:51 -!- oerjan has joined.
20:06:27 <Ilari> If 5^(2^r) == 1 (mod 2^r) and 3^(2^r) == 1 (mod 2^r) then that would really mean that one could calculate 5^(k*2^r)-3^(k*2^r) mod k and use that.
20:08:24 <Ilari> Okay, looks like those hold. So indeed, one wouldn't have to have that power of two term in divisior.
20:10:02 <crystal-cola> For a given k I think one can figure out the exact sequence of 5^(k*2^r) (mod k)
20:10:14 <Ilari> And that can be calculated from 5^k mod k and 3^k mod k.
20:11:44 <Ilari> That also yields a way to test small values of r really fast.
20:13:00 <crystal-cola> [9,13,16,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1]
20:13:54 <crystal-cola> it is periodic (happens to have period 1 but ignore that) after a 3 steps
20:14:07 <crystal-cola> I guess that every sequence will be periodic after some fixed number of steps (usually 0)
20:14:25 <crystal-cola> anyway, finding the pre-period should give the values of (k,r) in the pairs?
20:15:28 <Ilari> Well, period can be at most k^2 because limited number of states.
20:15:56 <crystal-cola> hmm does that mean every period length is a divisor of k^2?
20:16:27 <crystal-cola> too bad - that would have made it a bit easier, could just ignore the 'true' period length and use k^2 for everything
20:17:09 -!- elliott has quit (Ping timeout: 246 seconds).
20:19:15 <Ilari> k=19 (which doesn't have solutions) has period of length 6.
20:21:28 <Ilari> Oh, also this means that r_min < k^2.
20:25:36 <Ilari> Actually, if you want r_min, then you need to check all lesser ones. But that's almost as fast as just checking r = k^2.
20:26:15 -!- calamari has quit (Quit: Leaving).
20:29:57 <oerjan> have you considered fermat's little theorem?
20:30:22 <oerjan> (a^(p-1) = 1 (mod p) for prime, gcd(a,p) = 1)
20:30:48 <oerjan> or a^p = a (mod p) for prime p generally
20:32:00 <oerjan> which means 3^(17*2^r) = (3^(2^r))^17 = 3^(2^r) (mod 17)
20:32:58 <oerjan> (i'd be surprised if you haven't, but you never know)
20:33:49 <oerjan> > [9^r `mod` 17 | r <- [1..]]
20:33:50 <lambdabot> [9,13,15,16,8,4,2,1,9,13,15,16,8,4,2,1,9,13,15,16,8,4,2,1,9,13,15,16,8,4,2,...
20:34:35 <oerjan> > [3^(2^r) `mod` 17 | r <- [1..]]
20:34:46 <oerjan> > take 10 [3^(2^r) `mod` 17 | r <- [1..]]
20:35:02 <oerjan> crystal-cola: @let function = value
20:35:25 <crystal-cola> @let modularPower b 0 n = 1 ; modularPower b e n = if (e `mod` 2 == 0) then ((modularPower b (e `div` 2) n) ^ 2) `mod` n else (b * (modularPower b (e-1) n)) `mod` n
20:35:44 <lambdabot> [9,13,16,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,...
20:36:33 <lambdabot> [8,13,16,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,...
20:37:23 <Ilari> Yeah, r_min=2 for k=17.
20:37:58 <crystal-cola> it's just luck they share the same pre-period suffixe
20:38:22 <crystal-cola> maybe that matters in terms of the composition
20:38:36 <Ilari> Actually, that test isn't as good as I thought. The basic problem is that the test for solutions takes k^2 time (which is actually exponential time!).
20:38:54 <lambdabot> [9,81,6561,54449,61869,19139,15028,282,13987,8224,65529,64,4096,65281,65536...
20:38:56 <lambdabot> [25,625,62940,59635,33457,64426,54655,58102,31534,255,65025,65533,16,256,65...
20:43:20 <Ilari> Actually, the maximum bound for r_min is (k-1)^2.
20:45:03 <Ilari> Anyway, r_min < k^2 if it exists at all.
20:53:10 -!- TeruFSX2 has joined.
20:54:32 <oerjan> actually, irregular webcomic
20:54:56 <oerjan> but i guess if cake is good enough for you...
21:00:16 <Ilari> If one just wants to test a k: Compute A <- 5^k%k B <- 3^k%k then k² times A <- (A*A)%k, B <-(B*B)%k. If A == B, then there is a solution for k*2^r (r_min is number of times you have to do the squares before A == B)..
21:02:57 <Ilari> This also shows that for any k with solutions, 3 doesn't divide it.
21:03:11 -!- ais523 has quit (Remote host closed the connection).
21:04:05 <oerjan> that's rather simpler, just note that 3 and 5 cannot divide 3^k-5^k ever
21:05:24 <Ilari> Which means k is one of 1, 7, 11, 13, 17, 19, 23 or 29 modulo 30.
21:07:30 <oerjan> did you imply that if k1 and k2 are solutions, then so is lcm(k1,k2) ?
21:09:13 <Ilari> Actually, if n1 and n2 are, then lcm(n1,n2) is. Which is bit different.
21:09:35 <oerjan> erm right you are ignoring powers of 2 for the k's
21:10:14 <Ilari> Which also impiles that: r_min(lcm(k1,k2)) = max(r_min(k1),r_min(k2)).
21:11:24 -!- Wamanuz2 has quit (Read error: Operation timed out).
21:11:44 -!- pikhq has joined.
21:12:39 -!- Wamanuz2 has joined.
21:16:18 -!- Sgeo has joined.
21:17:25 <Ilari> Now that it is possible to calculate valid k without gaps, I can confirm that OEIS doesn't have this sequence.
21:17:54 <oerjan> and no prime > 2 can be an n-type solution because of fermat's theorem
21:18:03 <crystal-cola> the sequence is missing a lot more than what the congruence obstructs, isn't it?
21:18:24 <oerjan> does it have the n sequence?
21:20:04 <oerjan> > [modularPower 3 n n | n <- [1..]]
21:20:07 <lambdabot> [0,1,0,1,3,3,3,1,0,9,3,9,3,9,12,1,3,9,3,1,6,9,3,9,18,9,0,25,3,9,3,1,27,9,12...
21:20:31 <lambdabot> [0,0,2,0,2,4,2,0,8,4,2,4,2,4,8,0,2,10,2,16,8,4,2,16,7,4,26,16,2,4,2,0,8,4,1...
21:20:42 <crystal-cola> 3 appears in this sequence really really far along
21:20:48 <Ilari> There are some pretty funky k in there. Like 2329(17*137). 17 is a valid k, but 137 isn't.
21:21:02 <crystal-cola> 4700063497 then 63130707451134435989380140059866138830623361447484274774099906755
21:21:10 <oerjan> one might consider using euler's totient function somehow...
21:21:37 <oerjan> (a^t(n) = 1 (mod n), gcd(a,n) = 1)
21:22:49 <oerjan> Ilari: didn't you say something about 17*137^r
21:23:20 <Ilari> Yeah, at least 17*137² is also valid k.
21:26:01 -!- MigoMipo has quit (Read error: Connection reset by peer).
21:37:32 <Ilari> Here are the confirmed pairs (up to 6 000, not using powers): (1, 0), (17, 2), (97, 5), (257, 7), (289, 2), (353, 3), (2329, 2) and (4913, 2).
21:40:37 <Gregor> MOXIEEEEEEEEEEEEEEEEEEEEEEEEE
21:41:13 <oerjan> moxie, the fnarfiest of beverages
21:42:36 <oerjan> at this rate we'll soon be needing an #esoteric dictionary
21:42:45 <Gregor> Your pathetic "taste" sense which you think in your ignorance is superior to fnarf doesn't allow you to experience the sheer degree of joy and pleasure.
21:42:57 <crystal-cola> @let zygotes = [1,17,97,257,289,353,2329,4913]
21:45:48 <oerjan> we shall have to rename hyposmia to euphnarphia
21:47:58 <Ilari> Probably there are further patterns. But none that would be easy to prove.
21:49:27 <crystal-cola> but I mean maybe there is just some formula that generates them all
21:49:34 <Ilari> Such as: Are all k of form 17^k valid?
21:50:33 <Ilari> What about 17*137^k?
21:57:18 <oerjan> do you have an example of n1, n2 where n1n2 isn't a solution?
22:06:13 -!- Phantom_Hoover has quit (Quit: Leaving).
22:21:06 <oerjan> > [3^n-5^n | n <- [1..]]
22:21:07 <lambdabot> [-2,-16,-98,-544,-2882,-14896,-75938,-384064,-1933442,-9706576,-48650978,-2...
22:22:58 <oerjan> 16*16 doesn't divide 544
22:23:30 <oerjan> so you don't always have (3^m-5^m)(3^n-5^n)|(3^(mn)-5^(mn))
22:24:29 <crystal-cola> if m|3^m-5^m and n|3^n-5^n doesn't that imply mn|3^(mn)-5^(mn)?
22:24:47 <oerjan> well that's what we're asking...
22:26:09 <oerjan> > filter (\n -> (5^n-3^n)`mod`n==0) [1..]
22:26:59 <oerjan> > filter (\n -> modularPower 5 n n == modularPower 3 n n) [1..]
22:27:28 <oerjan> > take 20 $ filter (\n -> modularPower 5 n n == modularPower 3 n n) [1..]
22:27:29 <crystal-cola> > take 3 $ filter (\n -> modularPower 5 n n == modularPower 3 n n) [1..]
22:27:31 <lambdabot> [1,2,4,8,16,32,64,68,128,136,256,272,512,544,1024,1088,1156,2048,2176,2312]
22:28:23 <oerjan> we already know multiplying by 2^r preserves solutions, so those are uninteresting
22:28:51 <oerjan> > take 40 $ filter (\n -> modularPower 5 n n == modularPower 3 n n) [1..]
22:28:59 <crystal-cola> > take 3 $ filter (\n -> modularPower 5 n n == modularPower 3 n n) [1,3..]
22:29:16 <oerjan> um we have no odd examples
22:30:04 <oerjan> > (\n -> modularPower 5 n n == modularPower 3 n n) 4624
22:31:14 -!- topodecai has joined.
22:31:57 <oerjan> > all (\n -> modularPower 5 n n == modularPower 3 n n) $ join (liftM2 (*)) [68,136,272,544,1088,1156,2176,2312]
22:33:03 <oerjan> > map (\n -> modularPower 5 n n == modularPower 3 n n) $ join (liftM2 (*)) [68,136,272,544,1088,1156,2176,2312]
22:33:04 <lambdabot> [True,True,True,True,True,True,True,True,True,True,True,True,True,True,True...
22:35:20 <Ilari> Looks like if n is a solution, ns is too, where s is any prime factor of n. This would imply that if n1 and n2 are solutions, n1*n2 is too.
22:35:30 <oerjan> > map (\n -> modularPower 5 n n == modularPower 3 n n) . reverse $ join (liftM2 (*)) [68,136,272,544,1088,1156,2176,2312]
22:35:31 <lambdabot> [True,True,True,True,True,True,True,True,True,True,True,True,True,True,True...
22:35:37 -!- topodecai has left.
22:35:48 <oerjan> ok so it is not a problem of modularPower being inefficient
22:35:53 <crystal-cola> > map (\b -> if b then 't' else 'f'). map (\n -> modularPower 5 n n == modularPower 3 n n) . reverse $ join (liftM2 (*)) [68,136,272,544,1088,1156,2176,2312]
22:35:54 <lambdabot> "tttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttt"
22:36:11 <oerjan> > all (\n -> modularPower 5 n n == modularPower 3 n n) $ join (liftM2 (*)) [68,136,272,544,1088,1156,2176,2312]
22:36:24 <oerjan> ok the Terminated was just a fluke
22:38:23 * oerjan realizes "hail" is ambiguous
22:39:01 <Ilari> Actually, it is much stronger, allowing likes of generating k=319073 out of k=2329 (with r_min of 2).
22:39:05 <pikhq> oerjan: The precipitation.
22:39:17 <pikhq> Colorado has decided to have confusing weather.
22:39:49 <pikhq> It was snowing earlier.
22:40:02 <oerjan> Ilari: oh hm didn't notice your previous sentence. yeah that would imply it.
22:43:46 -!- azaq23 has joined.
22:45:21 <Ilari> Assume n|5^n-3^n => 5^n-3^n = 0 (mod n). That is, 5^n = 3^n (mod n). Now, 5^(ns)-3^(ns) = (5^n)^s - (3^n)^s = (x1*n+x)^s - (x2*n+x)^s = (x1*n*s+x) - (x2*n*s+x) = x - x = 0 (mod ns) => ns|5^(ns)-3^(ns).
22:46:47 <Ilari> That assumes ns|n^2 => s|n (which is assumed).
22:47:24 <Ilari> Actually, s|n => ns|n^2.
22:50:13 <oerjan> well those are equivalent yes
22:53:20 <Ilari> This would allow considerable compression of base set needed to generate all solutions (it still might be infinite, but not due to powers).
22:53:38 <oerjan> ok i don't get the (x1*n+x)^s - (x2*n+x)^s = (x1*n*s+x) - (x2*n*s+x) step
22:56:20 <Ilari> All higher powers of n generate multiples of ns (and thus equal zero mod ns).
22:57:54 <oerjan> but i think the right side should then be (x1*n*x^(s-1)*s + x^s) - (x2*n*x^(s-1)*s + x^s)
23:01:12 <oerjan> hm wouldn't this mean the base set can only be infinite if there are infinitely many primes involved?
23:04:18 <Ilari> There are some non-primes involved as well.
23:04:37 <oerjan> um i mean all prime factors involved
23:05:21 <oerjan> in fact it's a corollary of that subsequence thing you've spoken about before, i think
23:05:25 <crystal-cola> some numbers are only appearing when taken together
23:06:11 <oerjan> "primes involved" does not mean that the solutions are primes
23:07:55 <oerjan> (if you write each solution using its ascending prime factorization, then if there are infinitely many with a finite set of prime factors then one must be a subsequence of another => one must divide the other)
23:08:15 <oerjan> *with a given finite set
23:11:28 -!- augur has joined.
23:14:09 <Ilari> Irreducable pairs with k<10^6: (1, 0), (17, 2), (97, 5), (257, 7), (353, 3), (2329, 2), (7681, 8), (23137, 4), (36097, 8), (37733, 5), (39899, 2), (65537, 15), (150641, 5), (198593, 4), (259841, 8), (443921, 9), (550273, 4), (786433, 18), (828089, 5)
23:15:08 <oerjan> um the second element of the pair is just the power of 2 extracted, right?
23:15:29 <Ilari> Yes (actually the smallest possible).
23:15:53 <oerjan> so you should technically add (1, 1) to get everything generated :)
23:16:54 <oerjan> (1, 0) is sort of degenerate
23:17:53 <Ilari> Of course, one could write those as single numbers (a,b) => a*2^b, (except for (1, 0) => 2). Then one could write the solutions as lcm of some subset times factors of itself.
23:18:09 <Ilari> (of course, assuming lcm of nothing is 1).
23:18:31 <oerjan> in fact you sometimes _do_ need to include two numbers where one is a factor of the other, when the larger one adds a new prime
23:19:52 <Ilari> Actually, no, since one can just raise the factors independently. If you include 2329, no need to include 17.
23:20:17 <Ilari> You only need 17 to generate numbers that don't have 137 as factor.
23:20:35 <oerjan> 2329 doesn't give you all the information
23:20:51 <oerjan> (that 17 is a solution but 137 isn't)
23:22:17 <oerjan> and if you could combine them that way, you'd just end up replacing _every_ pair with their lcm
23:23:22 <oerjan> afaiac the base should consist of solutions such that no smaller number with the same prime factors is a solution
23:24:02 <oerjan> which both 17 and 2329 are
23:24:03 <Ilari> Then there are pairs irreducable k that are both semiprime but not coprime: 443921 and 550273 for instance.
23:25:14 <oerjan> > (`div` 17) <$> [443921, 550273]
23:35:22 -!- FireFly has quit (Quit: swatted to death).
23:55:02 <oerjan> the ghost of primes past
23:56:00 <oerjan> > (`mod` 137) <$> [443921, 550273]