00:05:12 <zzo38> How many programs are there that can typeset to DVI files? I know of TeX, GFtoDVI, and I think GNU troff has an option to output DVI as well. GNU Lilypond has partial support for DVI output. Are there others?
00:06:59 <olsner> the church turing thesis posits there are at most 4 such programs
00:07:21 <zzo38> How can the Church-Turing thesis posit such things?
00:07:36 <olsner> (posit is possibly the best verb in the world)
00:11:05 <zzo38> I am making the Haskell library for both input and output of DVI. It even does ligature and kerning.
00:16:40 -!- MoALTz has quit (Ping timeout: 250 seconds).
00:17:49 -!- NSQX has joined.
00:18:53 <NSQX> elliott, I was only trying to use a bot to edit the page about UniCode/fill in all Unicode characters in the table
00:19:15 <NSQX> I was not trying to make an edit using my bot to any other page.
00:19:32 -!- MoALTz has joined.
00:20:08 <elliott> NSQX: A bot can have serious negative effects even if it's coded to only edit one page.
00:22:18 <elliott> It's also a really bad thing for recent changes to be filled up with edits summarised only as "Wikipedia python library" when a bot's going wrong.
00:25:00 <olsner> NSQX: "fill in all unicode characters"? why does the eso wiki need that?
00:26:42 <olsner> NSQX: if you want a bot to do something, it'd be a better idea to run the bot locally (without it modifying the wiki in any way), review the changes, and then add a single change with whatever the bot did (if you think it did something good)
00:26:50 <NSQX> I think I might have fixed the problem - if I don't delete and re-create the Page object before running Page.get() a second, third, etc. time, page.get() will always return the data that was in the page when the Page object was created. http://pastebin.com/mefcu3EV
00:27:24 <olsner> well, esolangs.org is not the appropriate environment to discover these bugs in your bot
00:27:28 <elliott> NSQX: That makes one edit per character.
00:27:42 <elliott> I agree with olsner that the best thing to do would be to generate the wikicode you want locally (through any means you wish), then copying that to the page.
00:27:51 <NSQX> The page about UniCode should show every unicode character and instruction in the table.
00:28:11 <elliott> Having every Unicode character on a page wouldn't work, by the way.
00:28:21 <elliott> It would be an astronomically large page and likely cause problems with browsers.
00:29:14 <elliott> (Even just every Unicode character, with no additional wikicode arround it, is 100 kilobytes, way above the maximum size wiki pages can be without causing problems.)
00:31:01 <NSQX> http://esolangs.org/w/index.php?title=UniCode&action=edit&oldid=31630 (Restore this revision)
00:31:32 <kmc> shachaf, I don't think it's exactly what I say
00:32:04 <kmc> more like "when we synthesize new answers, let's do so in a persistent editable medium so we can combine and refine them, rather than just shouting them at each other"
00:32:19 <shachaf> kmc: Maybe your objection is less to syntehsizing and more to typing in answers of any kind to common questions in IRC.
00:32:37 <elliott> NSQX: I don't think the table will mislead anyone as it is; it's less than a day until you'll be able to revert it yourself.
00:32:48 <kmc> i mean i've seen the standard answers improve over ~6 years of lurking in #haskell on and off
00:32:59 <kmc> so I would be loathe to institute a moratorium on new answers!
00:33:16 <NSQX> Now I know that trying to fill in all the unicode tables would increase the page size to almost 7MB
00:33:47 <shachaf> kmc: Have they improved as a result of people answering the same question?
00:34:01 <NSQX> Just one Unicode character added to the table doesn't increase the page size to 100KB. It increases the page size to 100-104 bytes.
00:34:08 <shachaf> kmc: I think it's more likely that they improved as a result of context/person-specific questions.
00:34:28 <zzo38> kmc: What I know is that people disagree about the answers, even questions that should have a proper answer people disagree, such as: data CoPh x; Does this make a monad? Does this make a comonad? I think it makes a comonad but not a monad; other people disagree
00:34:30 <shachaf> I doubt that someone suddenly came up with a better answer for "what is a monad?" the 500th time it was asked the same way.
00:35:13 <elliott> NSQX: Well, I'm not exactly sure what your bot is designed to do, because you haven't told anyone.
00:35:16 <zzo38> shachaf: At least how I would say, is, you should first understand what is an endofunctor
00:36:24 <elliott> NSQX: If it's just to add a line to the table on request to avoid retyping it, then wouldn't a MediaWiki template do just as well?
00:37:01 <kmc> shachaf, not suddenly no
00:37:20 <NSQX> elliott, Do you think it would be at least better to split all the UniCode characters into subpages of [[UniCode]] and put them all in the wikitext as "{{:UniCode/TR1}} {{:UniCode/TR2}} ..."?
00:38:03 <elliott> That would probably be worse, since it would result in tens to hundreds of pages.
00:38:10 <olsner> again, what does unicode have to do with the eso wiki? unicode.org has everything you could ever want to know about it
00:38:26 <elliott> NSQX: But I'm not really sure what you're trying to accomplish with the bot in the first place, like I said.
00:38:38 <NSQX> Well, what about every UniCode instruction?
00:38:53 <zzo38> What if you make subpages, Unicode/00xxx Unicode/01xxx ... Unicode/10xxx ?
00:39:23 <zzo38> Actually I think we need more than that...
00:39:37 <monqy> isn't UniCode long dead anyway?
00:39:42 <monqy> was it ever alive?
00:40:13 <zzo38> Unicode/000xxx Unicode/001xxx Unicode/002xxx .... Unicode/00Fxxx Unicode/010xxx ... Unicode/0F0xxx Unicode/100xxx I think that is the maximum one.
00:40:22 <NSQX> UniCode has 65536 characters, each one would be assigned to a UniCode instruction.
00:40:38 <zzo38> Some astral planes are private use so don't use those ones
00:40:49 <zzo38> And some code points are surrogates so don't use those either
00:40:52 <NSQX> Remember the Unicode characters and UniCode programming language instructions to think of.
00:41:09 <elliott> Unicode has more than 110,000 characters.
00:41:16 <elliott> It's been bigger than 16 bits for a long time now.
00:42:41 <pikhq> Furthmermore, a glyph consists of one or more codepoints.
00:42:45 <NSQX> Unicode only has 65536 characters. It's Unicode Big Endian which has more than 110,000 characters. Actually, most operating systems still only support ASCII, ANSI, and Unicode, not Unicode Big Endian.
00:43:33 <zzo38> Actually, Unicode Big Endian is only an encoding of Unicode, it is not a different Unicode.
00:43:34 <elliott> "In 1996, a surrogate character mechanism was implemented in Unicode 2.0, so that Unicode was no longer restricted to 16 bits. This increased the Unicode codespace to over a million code points, which allowed for the encoding of many historic scripts (e.g. Egyptian Hieroglyphs) and thousands of rarely-used or obsolete characters that had not been anticipated as needing encoding."
00:43:43 <elliott> "Unicode defines a codespace of 1,114,112 code points in the range 0hex to 10FFFFhex."
00:43:46 <kmc> "big endian" isn't a single encoding either
00:43:54 <kmc> itt: NSQX is wrong about unicode
00:44:21 <pikhq> NSQX: The latest Unicode standard contains 110,116 characters.
00:44:32 <olsner> ah, the BIG ENDIAN unicode, biggest unicode yet
00:45:18 <elliott> Doesn't Windows support astral planes these days?
00:45:21 <elliott> I think GTK does. Does Qt still not yet?
00:45:22 <kmc> big indian unicode http://unicode.org/charts/PDF/U0900.pdf
00:45:23 <pikhq> Incidentally, Unicode is not *itself* a character encoding, but rather a set of characters, with several encodings.
00:45:25 <NSQX> Even the C wchar_t only supports character codes up to 65536.
00:45:29 <pikhq> elliott: Windows has for a decade now.
00:45:42 <elliott> NSQX: wchar_t is of implementation-designed size, it can be as little as 8 bits
00:45:42 <kmc> NSQX, wrong again, it's so on Windows but on UNIX it's usually 32 bits
00:45:59 <zzo38> Haskell specifies the Char is Unicode from 0hex to 10FFFFhex, but I think, Ibtlfmm shall go up to FFFFFFhex and the implementation is not required to be Unicode, it can use anything, as long as it is ASCII.
00:45:59 <olsner> in about 5 years, I think most web browsers will even fully support astral planes
00:46:06 <NSQX> Well, let's just get on with planning the UniCode language instructions now.
00:46:14 <kmc> NSQX, start by understanding Unicode?
00:46:31 <pikhq> NSQX: On Windows, wchar_t is a 16 bit type. However, it is typically used with UTF-16, which can encode all of Unicode.
00:46:44 <kmc> which means one wchar_t isn't necessarily one character
00:46:51 <kmc> (even setting aside combining characters)
00:46:57 <pikhq> (prior to Windows XP, it was more typically UCS-16, which encodes only the base multilingual plane)
00:47:19 <calamari> can utif-8 also handle all of unicode?
00:47:20 <kmc> zzo38, GHC also allows invalid Unicode characters like '\xD800'; I don't know if that's addressed in the spec
00:47:24 <kmc> calamari, yes
00:47:31 <kmc> utf-8 is a wonderfully designed encoding
00:47:38 <pikhq> UTF-8 actually has *much* more space than is necessary.
00:47:58 <pikhq> UTF-8 could in theory represent a 32-bit space.
00:48:51 <zzo38> kmc: I do know that; I have tried it. I don't know if the spec mentions that either. But I would make the new one, instead of being Unicode, it is simply 24-bits and all values are valid, and it does not matter what character set you use as long as it is ASCII.
00:49:09 <elliott> any colour you want, as long as it's black
00:49:25 -!- pikhq_ has joined.
00:49:58 <zzo38> pikhq: I do know, UTF-8 can be used to encode numbers beyond Unicode (although if you do so, it will not make a valid Unicode text data)
00:50:19 <pikhq_> zzo38: Yeah. NSQX does not.
00:50:55 <zzo38> elliott: That is not what I mean
00:53:00 -!- pikhq has quit (Ping timeout: 276 seconds).
00:54:18 <pikhq_> NSQX: BTW, every OS that's not hilariously archaic uses a Unicode transformation format, with everything else considered a legacy feature.
00:54:40 <pikhq_> Unices on UTF-8, OS X on UTF-8, Windows on UTF-8 and UTF-16...
00:58:14 -!- NSQX has quit.
01:00:52 <zzo38> I will usually use ASCII; my feature for new computer and system, uses NK (8-bits encoding) internally, but FBUCE (which is 16-bits encoding, alhough character codes as 12-bits and the other 4-bits are used for other features) can also be used, as well as any other 8-bits encoding which can transform to FBUCE (NK also can transform to FBUCE). FBUCE has 512 control characters, and has some characters which are not part of Unicode, but does not h
01:02:30 <zzo38> I think Unicode should also define non-surrogate purposes of surrogate codes, such as to represent keystrokes which have no character code. They are not needed in text but can be used in a GHC Char and so on.
01:09:41 <zzo38> Why shouldn't it be allowed to use "data instance" where "type instance" is expected?
01:11:31 <shachaf> zzo38: You mean, instead of defining a data type on your own and then making a type synonym to it?
01:12:26 <shachaf> Seems like a reasonable thing to do. data instance and type instance are different things.
01:13:28 <zzo38> I know data instance and type instance are different things. But if you use "data instance" where "type instance" is expected (or "data" instead of "type" in a class instance declaration), it should be able to make up an anonymous datatype and use that.
01:38:04 <elliott> Well, NSQX just evaded their block and vandalised my old user page.
01:38:31 <elliott> I've lengthened their block to a week. I suspect it should be longer.
01:43:23 <elliott> http://esolangs.org/wiki/User_talk:202.156.14.101 And there we go.
01:46:35 <pikhq_> Jeeze. I... Think the lottery's jackpot went high enough that you can expect a slight profit from tickets.
01:46:58 <elliott> http://r6.ca/blog/20090522T015739Z.html
01:48:29 <pikhq_> That is not to say that it overcomes the near-guarantee you'll lose your $1.
01:49:00 <elliott> Well, the probabilities still even that out. The point is that the $1 is way, way too big unless you're super rich.
01:50:02 <pikhq_> Still, $640 million payout with 1 in 175,711,536 odds of jackpot are fairly obscene.
01:50:47 <pikhq_> (I should note that I've not gone out and paid anything, because 175,711,535 in 175,711,536 odds of losing $1 are not odds I like much)
01:52:03 <pikhq_> Anyways. Yeah, I think the situation here is nearly the same as the linked one.
01:52:20 <pikhq_> It's a pretty bad investment unless you've got several million already.
01:53:47 <pikhq_> Still, it amuses me that it's theoretically possible to make a profit from lottery tickets.
01:54:10 <pikhq_> Sorry, theoretically possible to *expect* a profit from them.
01:55:42 <coppro> pikhq_: there was a post on reddit calculating that if you bought every possible ticket and didn't split the jackpot, you would earn $400M
01:57:04 <pikhq_> coppro: Iff you get the annuity option, I assume.
01:57:25 <pikhq_> And don't account for the income tax.
01:57:34 <coppro> pikhq_: That was assuming 50% tax
01:57:54 <coppro> http://i.imgur.com/DLwLF.jpg
01:58:52 <pikhq_> The figures are off the cash option payout.
01:59:08 <pikhq_> Which... Is a bit silly.
01:59:22 <pikhq_> Currently that's nearly $200 million that just up and disappears.
02:00:11 <pikhq_> Admittedly, taking the annuity would mean there would be a few *years* before you profit.
02:00:29 <elliott> How long would it take to buy all the tickets? :p
02:00:57 <pikhq_> Dunno. How fast are the printers?
02:01:29 <zzo38> Not only it take a long time but it would cost a lot of money and energy too
02:02:04 <pikhq_> zzo38: You can buy a large amount of tickets at a time.
02:02:23 <pikhq_> Shops selling them have printers that generate them.
02:02:47 <coppro> you need 175 million tickets
02:03:01 <pikhq_> Also, given the amount of money at hand you could probably pay people to help buy tickets.
02:03:13 <coppro> or create a giant collective
02:03:29 <pikhq_> True. A corporation could actually go out and do this.
02:03:42 <pikhq_> "Lottery Winners, Inc."
02:03:45 <coppro> hell, just pay the retailers to convince everyone buying tickets to get in on it :P
02:04:21 <zzo38> But you have to select which number you want, too; try to select the number that nobody else is going to select.
02:04:28 <pikhq_> Now there's a business model. Buy in to lotteries with EV greater than the ticket.
02:04:49 <coppro> zzo38: convince every buyer to get in on it
02:05:08 <pikhq_> zzo38: In the lottery scheme in question, tickets aren't *necessarily* unique.
02:05:20 <elliott> <coppro> you need 175 million tickets
02:05:26 <elliott> I know that lotteries really don't like people gaming them, though.
02:05:37 <elliott> I suspect there's regulations to stop this kind of stuff at a way smaller scale.
02:05:40 <coppro> what they care about is people thinking they can be gamed
02:05:41 <pikhq_> elliott: Gaming them in this *particular* way is generally acceptable.
02:06:02 <pikhq_> This is not some unintended consequence, but rather a mere coincidence.
02:06:10 <coppro> They don't really care if someone is secretly buying all the good scratchcards
02:06:24 <coppro> people will buy the duds
02:06:35 <coppro> so all they need to do is keep the secret from getting out
02:06:37 <pikhq_> I should note that the jackpot only gets this high when there's been a few weeks of no jackpot being awarded.
02:06:52 <pikhq_> So, even when each ticket has positive EV, they've made a profit.
02:13:06 <zzo38> I have invented a variant of video poker which can be played using scratch lottery cards
02:25:14 <elliott> Going by the webserver logs, NSQX is still running that bot. I can't tell why.
02:28:46 <zzo38> Rules: Before discarding any cards, you must announce what number of cards you want to discard. However. after each discard, you can look at the replacement card before deciding which card you want to discard next. (But you are not allowed to discard the replacement cards.)
02:30:57 <lambdabot> Local time for elliott is Sat Mar 31 03:30:56
02:34:39 <zzo38> How much advantage does the player have in this variant of video poker?
02:36:09 <zzo38> (I have called this game "scratch poker")
02:52:44 <lambdabot> Local time for elliott is Sat Mar 31 03:52:43
02:52:54 <lambdabot> Local time for pikhq_ is Fri Mar 30 20:52:25 2012
02:53:13 <pikhq_> elliott: When do you sleep?
02:53:30 <elliott> pikhq_: uh probably like 4ish
02:53:47 <pikhq_> ntpd says my time is right.
02:58:28 <elliott> `addquote <monqy> esolang a horror movie where the ghost just wanders around and moves newspapers and boots and family pictures a few inches before moving on
02:58:31 <HackEgo> 835) <monqy> esolang a horror movie where the ghost just wanders around and moves newspapers and boots and family pictures a few inches before moving on
02:59:05 -!- Tiktalik has quit (Changing host).
02:59:06 -!- Tiktalik has joined.
03:26:24 -!- variable has changed nick to const.
03:35:40 -!- elliott has quit (Ping timeout: 260 seconds).
03:39:43 -!- MDude has quit (Ping timeout: 246 seconds).
03:51:38 -!- ion has quit (Ping timeout: 272 seconds).
03:57:24 -!- ion has joined.
04:00:04 <zzo38> Someone may eventually notice the movement, or notice that something is wrong even if they don't know what it is
04:05:26 -!- GhostHand has joined.
04:09:18 -!- Patashu has joined.
04:31:58 -!- itidus22 has changed nick to itidus21.
05:36:55 <zzo38> O, is that the ghost hand just wanters around and move everything just one inch?
05:37:22 <monqy> it very well may be that ghost
05:48:21 <shachaf> kmc: Sounds like the information you gave me about NYC pizza is outdated: http://www.nytimes.com/2012/03/31/nyregion/in-manhattan-pizza-war-price-of-slice-keeps-dropping.html?pagewanted=all
05:48:53 <kmc> yeah i just heard about this
05:49:14 <kmc> ...how did you hear about it?
05:50:25 <kmc> i hear they have the internet on computers now
05:51:37 <shachaf> kmc: Yes, but that Internet tastes kind of mechanical.
06:03:58 -!- MoALTz_ has joined.
06:04:55 -!- Sgeo_ has joined.
06:06:44 -!- MoALTz has quit (Ping timeout: 245 seconds).
06:08:28 -!- MoALTz__ has joined.
06:11:41 -!- MoALTz_ has quit (Ping timeout: 260 seconds).
06:31:02 -!- calamari has left ("Leaving").
06:33:36 -!- asiekierka has joined.
06:35:42 <Sgeo> The Arimaa game room stores passwords in plaintext. Because fuck security, that's why.
06:48:05 <Sgeo> monqy, game designed to be easy for humans to learn and hard for computers to be good at
07:10:21 <zzo38> Can they implement Haskell Dynamic like that? data Dynamic where { Dynamic :: forall x. Typeable x => x -> Dynamic; }; castDyn :: Typeable x => Dynamic -> Maybe x; castDyn (Dynamic x) = cast x;
07:23:24 -!- MoALTz_ has joined.
07:25:12 -!- MoALTz__ has quit (Ping timeout: 244 seconds).
07:28:38 -!- Sgeo_ has quit (Ping timeout: 246 seconds).
07:45:43 -!- oerjan has joined.
07:52:15 <oerjan> <oklopol> that's totally taken out of context. <-- but very oklo.
07:52:16 <lambdabot> oerjan: You have 1 new message. '/msg lambdabot @messages' to read it.
07:52:23 <lambdabot> fizzie said 10h 37m 48s ago: Hey, we're winning -- but W|A thinks you will eventually crush us: http://www.wolframalpha.com/input/?i=%28population+of+finland%29%2F%28population+of+norway%29
07:54:09 <zzo38> Didn't they say many quotations here are put out of context? (Including when I am quoted too; not only you)
07:54:45 <oerjan> @tell fizzie Well we like passed 5 millions just this month, while much of Europe is (about to start|) _shrinking_... no idea if finland is.
07:55:12 <oerjan> zzo38: probably; your style is also very characteristic and therefore does not always need context.
07:56:26 <oerjan> i suppose those quotes would still be confusing to someone who isn't a channel regular.
08:23:10 <fizzie> oerjan: Well, our derivative of population did peak in 1945 (with 3.5 children/woman), then dropped to 1.5 in 1970s, then been stuck at 1.8 for the last three decades or so. There's a nice "population pyramid -> population weirdo-tower-thing" comparison graph midway down http://www.stat.fi/tup/suomi90/joulukuu_en.html
08:24:14 <oerjan> fizzie: norway is also slightly lower than replacement level, so we grow due to immigration.
08:26:42 <fizzie> The Statistics Finland folks do say we're still somewhat growing too, according to http://www.stat.fi/til/vaenn/2009/vaenn_2009_2009-09-30_tie_001_en.html
08:28:11 <fizzie> "According to the projection, the annual number of deaths will exceed births in 2034, but net immigration is forecast to sustain population growth even after this."
08:29:50 <zzo38> This is not a pipe http://zzo38computer.cjb.net/img_0F/screen0.png
08:33:14 <oerjan> http://www.ssb.no/en/folkfram/ says we may pass 6 million in 2028, which is a while before you
08:35:26 <zzo38> Do you like Jyte? http://jyte.com/profile/zzo38computer.cjb.net
08:36:59 <fizzie> No, it's VERTICAL LINE.
08:37:00 <monqy> looks healthy to me
08:37:12 <monqy> maybe your'e fonts broken
08:37:15 <fizzie> Sometimes it's gappy, sometimes it's not. It's certainly not BROKEN BAR.
08:39:42 <oklofok> having more people is not winning.
08:39:51 <fizzie> Plain | is nasty that way, though I think the non-gappy versions are more popular.
08:39:52 <oklofok> it's the opposite of winning
08:39:58 <fizzie> ┆ is like doubly broken.
08:43:57 <zzo38> Can you more easily name 150 pokemons or 150 United States presidents?
08:45:47 <oklofok> i can name maybe like 2 us presidents
08:46:10 <fizzie> I wouldn't think anyone could name 150 US presidents.
08:46:17 <fizzie> They've had like fourty-something.
08:47:15 <oklofok> i had this feeling that they'd had ridiculously many
08:47:32 <oklofok> that they'd like been passing president coupons at some point
08:48:00 <fizzie> We've just got twelve, so we're even worse off. Unless having a lot of presidents is also not winning.
08:49:18 <zzo38> I don't even live in United States so I cannot name many US presidents either.
08:49:39 <oerjan> has _any_ country had 150 presidents.
08:50:04 <oklofok> i've heard of more than a half of us presidents :/
08:50:29 <fizzie> Wow, if I W|A "president", it returns Finland's new one. I wonder if it geolocatimates or something.
08:51:13 <zzo38> fizzie: It appears to; when I enter things such as sunset and so on, it also uses my location.
08:51:18 -!- const has quit (Remote host closed the connection).
08:51:56 <oerjan> we'll know that variable has started learning haskell when he changes nick to TVar
08:52:02 <oklofok> zzo38: you really seem to like the sky, what's that about?
08:52:15 <fizzie> It seems to; if I try it from the United States, it returns that Osama dude.
08:53:00 * oerjan wonders if there is a conspiracy theory that osama and obama are the same person
08:53:28 <oerjan> oklofok: it's all my fault for making the agora horoscope
08:53:36 <fizzie> oerjan: Also Saddam. (Cf. Barack Hussein Obama II.)
08:54:11 <zzo38> oklofok: Have *you* ever looked at the sky, or do you stay inside all the time, in room with no windows? I occasionally go walk outside (and will see the moon, and Venus, and Mercury, and sun, and sometimes the stars); I do not stay inside all the time
08:55:33 <monqy> I don't like seeing the sun
08:56:25 <oklofok> i have a window but i always have my black curtains up.
08:56:27 <zzo38> Of course I do not look directly at the sun. It hurts my eyes too. And in the day time it is sometimes too bright sun, too.
08:56:58 <oklofok> on the other hand i love walking outside
08:57:43 <zzo38> My window also has curtains, to prevent glare, but they aren't black
08:58:56 <oerjan> also, why is there snow on the ground.
09:08:05 -!- Madoka-Kaname has quit (Ping timeout: 245 seconds).
09:11:35 <zzo38> Do you know what was Little Miss Muffet's given name? And her father's given name? (Without looking it up)
09:12:49 <zzo38> Her name is Patience and her father's name is Thomas.
09:14:45 -!- MoALTz_ has quit (Quit: brb).
09:15:00 -!- MoALTz has joined.
09:17:59 <zzo38> Do you like SVG or METAFONT or something else?
09:18:19 <oklofok> what's your favorite group?
09:18:38 <shachaf> zzo38: I like ice cream, when it's not too sweet.
09:19:21 <oklofok> i never really got to knwo ti
09:20:12 <oklofok> iirc A_4 is the smallest counterexample to some stuff?
09:20:27 <oerjan> it's the largest alternative group which is solvable.
09:20:29 <zzo38> shachaf: No, I mean for vector drawing
09:20:48 <oklofok> you can draw vectors with ice cream
09:20:55 <oklofok> but they rarely come out straight
09:21:01 <shachaf> zzo38: What about a pencil?
09:21:15 <oklofok> and then it's just a big incomprehensible gay ice cream orgy just like in my dreams
09:21:27 <zzo38> shachaf: I suppose you can use a pencil, and I use a pencil too; but I meant by computer
09:21:41 <shachaf> zzo38: The computer can't use a pencil?
09:22:01 <oklofok> oerjan: oh, i guess that's what i was thinking about actually
09:22:24 <oerjan> or you might be thinking of A_5 which is the smallest non-abelian simple group
09:23:22 <zzo38> Do you mean like a computer plotter? Once I had something with wheels that can control by computer using BASIC stamp, and I attached a pencil to it, to see if it would draw, but it didn't work
09:25:03 <oerjan> oklofok: which is related to why A_4 was the last group which was interesting - there's a theorem which says you can use any simple group to compute boolean circuits in polynomial time, but it's not known whether you can use a solvable group like A_4
09:25:06 <oklofok> so tell me about A_4, what do you do together? or are with her just because she's famous?
09:25:12 <oerjan> *which i found interesting
09:26:03 <oerjan> it's called barrington's theorem
09:28:29 <oerjan> http://rjlipton.wordpress.com/2009/03/02/barrington-gets-simple/
09:34:08 <oerjan> if you try to use that method with a solvable group like A_4, you get into trouble with circuits that are deeper than the length of the group's derived group series.
09:35:01 <oerjan> which means you _can_ compute with A_4, but you have to turn your circuit into something like conjunctive or disjunctive normal form first, which gives exponential blowup
09:36:17 <oklofok> so err in particular you can use the two element group for this thing?
09:36:38 <oklofok> i'm still slowly reading that thing
09:36:51 <oerjan> no, the two element group is abelian. you cannot use an abelian group at all.
09:37:42 <oerjan> for the known prood, the group needs to be non-solvable
09:38:13 <oklofok> but any simple non-abelian?
09:38:40 <oerjan> in fact doing it with _any_ solvable group would prove the unknown (and not believed) equality of the ACC^0 and NC^1 complexity classes.
09:39:04 <oerjan> i think that article might mention that stuff, if not one of his other blog posts about it does.
09:39:15 <oerjan> yes, any simple group works.
09:39:48 <oerjan> and any non-solvable, which simple implies (in fact non-solvable == you can find a simple quotient of a subgroup)
09:40:03 <oklofok> but so i'm not sure i understand the model, is the idea that for every n, you choose some polynomially long chain of boxes, each of which gets one bit of input, and bounded information must flow from one box to the next?
09:42:00 <oklofok> idgi, again you said "any simple group works", isn't the two-element group simple?
09:42:43 <oklofok> or are you just so completely restricting to non-abelians that that doesn't even need to be mentioned, since it's based on commutators
09:43:05 <oerjan> oklofok: eh i guess i am :P
09:43:17 <oerjan> simple non-abelian, right
09:43:54 <oklofok> alright, i'll continue reading, although i might not get this without the gory details since i don't know anything about groups
09:45:43 <oklofok> so [G, G] means the set of all commutators, and in the non-abelian case it's nontrivial obviously and it's always normal for some simple reason, in the simple non-abelian case it's exactly G
09:45:56 <oklofok> now what's [G, H], is it something?
09:46:07 <oklofok> sorry, that's a stupid question
09:46:27 <oklofok> [x, y] means the corresponding commutator, and [G, G] is just a lift thingie.
09:47:05 <oerjan> oklofok: actually that's not stupid at all, it's http://en.wikipedia.org/wiki/Central_series
09:47:49 <oklofok> my question was what [G, H] even _means_, not what kind of group it is.
09:48:20 <oklofok> so it was stupid, i didn't quite recall that [x, y] is the commutator notation
09:48:24 <oerjan> [G,H] is the subgroup generated by the commutators ghg^-1h^1
09:48:33 <oklofok> because i read one letter at a time and then i come here to talk about it.
09:50:00 <oerjan> btw a group which has a central series is called nilpotent, and iirc lipton mentions that it's known such a group _cannot_ be used for these computations.
09:52:20 <oklofok> the or gate thing seems to give an exponential blowup, is the solution that we have logaritmic depth?
09:53:04 <oklofok> it seems that you still have to put all of the gates in
09:53:25 <oklofok> but blowup only happens depth many times i guess
09:53:44 <oklofok> do you know how and is done? oh wait. that's trivial.
09:53:50 <oklofok> hey this is fucking awesome :D
09:54:37 <oklofok> ah yeah [G, G] = G so you can choose the value of an or gate to be anything in the false case
09:55:06 <oklofok> THIS IS SO COOL :dSADFADSAFDSAFDDFSDAFDAF
09:55:09 <oklofok> AKLSDJFLKASJFLKAJSDLFKJAOEIWJFIJDC,XMV,.CMFDKLSFJSKLAMJ
09:55:55 <oerjan> STOP WITH THE FINNISH CODE
09:56:39 <olsner> LKAÖSJFLKAJSDÖLFKJAOEKLSFJSKAKLSDJFLKAÖSJFLKAJSDÖLFKJAOEIWJFIJ
09:56:48 <oklofok> so i guess we've established that asdfjaklsdjfklasjdfl means extreme frustration and ADJFAWEIMVKDMFKLAMSDF means omg this is awesome.
09:56:58 <oerjan> AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄ
09:59:22 -!- hagb4rd has joined.
10:05:55 <oklofok> oerjan: but so does [G, H] have some nice meaning?
10:06:46 <oerjan> i don't know much about that beyond the central series i linked to
10:06:50 -!- cheater_ has quit (Ping timeout: 246 seconds).
10:06:51 <oklofok> [G, G] somehow captures the noncommutation in G, is there something similar you can say about [G, H]
10:07:48 <oklofok> where somehow = (abelianizing by factoring = dividing [G, G] out)
10:08:19 <oklofok> WHY IS EVERYTHING SO AWESOME TODAY
10:08:51 <oerjan> alien nanobot invasion.
10:09:52 <oklofok> i wonder if there's still something to research in group theory
10:10:20 <oerjan> well i don't think the extension problem is really solved...
10:10:22 <oklofok> which is kind of silly i guess i find it trivial to find new questions about cellular automata and everyone says there's really nothing left there.
10:11:09 <oerjan> admittedly i know almost nothing about how much _has_ been done with it.
10:12:09 <oklofok> well none of the basic questions like this have been solved in symbolic dynamics either
10:12:32 <oklofok> well except everything except conjugacy has been solved for SFT's i guess
10:12:38 -!- nortti has joined.
10:13:11 <oklofok> but i think factoring between sofic shifts of equal entropy hasn't been solved for instance
10:14:14 <oerjan> and the simple groups classification _has_ been solved - with what may have been the largest math cooperation ever
10:14:25 <oklofok> for simple objects, i guess this sort of categorical questions are usually either trivial or extremely hard
10:15:04 <oklofok> The proof of the theorem consists of tens of thousands of pages in several hundred journal articles written by about 100 authors, published mostly between 1955 and 2004.
10:15:57 <oklofok> The completed proof of the classification was announced by Aschbacher (2004) after Aschbacher and Smith published a 1221 page proof for the missing quasithin case.
10:16:21 <oklofok> this is so awesome i'm almost crying
10:18:57 <oerjan> from 2006 on the Talk page: "Actually the proof of the uniqueness of the 26 sporadic groups is still a debated issue. For instance the uniqueness proof of the Thompson group is flawed, although a new proof of the uniqueness will be published in the coming months."
10:19:01 <oklofok> now i just wish i knew what groups of lie type were
10:19:22 <oerjan> it's going to take a while because they're sure they've _really_ closed all gaps :P
10:19:42 -!- zzo38 has quit (Remote host closed the connection).
10:19:59 <oerjan> and it's so huge that getting it through a theorem verifier seems impossible
10:20:15 -!- cheater_ has joined.
10:20:16 <oklofok> to have been a part of something like that would be just fucking incredible
10:20:49 <oklofok> probably something like that is going on in an early stage one could still jump into
10:21:08 <oerjan> heh Timwi is on the talk page
10:36:50 <itidus21> tl;dr kfc tasted good today. 2 everyone loves raymond episodes followed by 2 simpsons episodes followed by 2 futurama episodes = WHY IS EVERYTHING SO AWESOME TODAY yup
10:37:28 <oklofok> 1 friends episode followed by another friends episode followed by goto 1
10:42:53 -!- Patashu has quit (Remote host closed the connection).
10:43:08 -!- Patashu has joined.
10:47:55 -!- oerjan has quit (Quit: leaving).
10:51:59 -!- hagb4rd has quit (Ping timeout: 246 seconds).
10:57:49 <oklofok> This would be a major result, since it would imply the famous (to group theorists at least) Odd Order Theorem of Walter Feit and John Thompson. They showed that any odd order group cannot be simple.
10:58:24 <oklofok> do bounded computation with groups theorists define groups as non-abelian groups?
11:12:00 -!- azaq23 has quit (Quit: Leaving.).
11:42:59 -!- Patashu has quit (Remote host closed the connection).
11:43:07 -!- Patashu has joined.
11:49:03 <itidus21> ponders whether there a correlation between the value of a mathematical notation and the presence of it on a keyboard
11:49:28 <itidus21> maybe it is more indicative of "common" operations
11:54:17 -!- MoALTz_ has joined.
11:56:27 -!- MoALTz has quit (Ping timeout: 244 seconds).
12:03:41 -!- MoALTz_ has quit (Ping timeout: 244 seconds).
12:06:12 -!- MoALTz has joined.
12:09:41 <oklofok> they should be called crachSTUPIDs
12:14:43 <oklofok> http://donblazys.com/01.pdf o_O
12:14:50 <oklofok> how do you not get the death penalty for this
12:20:59 <olsner> hmm, found a backgrounded mplayer process in one of my shells... based on the contents it must've been there for months
12:27:27 -!- MoALTz has quit (Ping timeout: 244 seconds).
12:30:08 <olsner> ooh, found a bug! my cond implementation was expecting exactly one expression for each condition
12:31:23 <olsner> and when that match failed, it continued on to treat conds as applying a function 'cond' to some subexpressions, explaining why something was trying to use t as a function
12:33:46 -!- hagb4rd has joined.
12:37:29 <olsner> it might've helped if I knew more about this "lisp" thing before trying to write a compiler :>
12:42:41 -!- Patashu has quit (Quit: MSN: Patashu@hotmail.com , Gmail: Patashu0@gmail.com , AIM: Patashu0 , YIM: patashu2 , Skype: patashu0 .).
12:49:13 -!- elliott has joined.
12:52:44 -!- hagb4rd has quit (Ping timeout: 246 seconds).
13:01:58 -!- MoALTz has joined.
13:05:44 -!- Patashu has joined.
13:10:41 <elliott> 66.249.66.201 - - [31/Mar/2012:06:55:54 +0000] "GET /wiki/Human_And_Animal_Relationships HTTP/1.1" 302 161 "-" "Mozilla/5.0 (compatible; Googlebot/2.1; +http://www.google.com/bot.html)"
13:11:24 <elliott> 81.170.237.250 - - [31/Mar/2012:06:56:38 +0000] "GET /wiki/Brainfuck HTTP/1.1" 200 14619 "http://bytesizebio.net/index.php/2011/11/06/brainfk-while-waiting-for-a-flight/" "Mozilla/5.0 (compatible; MSIE 9.0; Windows NT 6.1; WOW64; Trident/5.0; NP06)"
13:11:42 <elliott> "Having arrived almost 3 hours early to JFK, flying back to Cincinnati, I spent the time coding up a Python script which inputs a string and outputs a Brainfuck source code which, when run with a Brainfuck interpreter, outputs said string."
13:11:50 <elliott> ok that's much less horrifying than i was expecting
13:28:28 -!- derdon has joined.
13:43:34 -!- Patashu has quit (Quit: MSN: Patashu@hotmail.com , Gmail: Patashu0@gmail.com , AIM: Patashu0 , YIM: patashu2 , Skype: patashu0 .).
13:44:17 <olsner> heh, the SICP page has links to download MIT Scheme for "Windows NT 4.0, Windows 95, and Linux (RedHat)"
13:47:06 <ion> olsner: Awesome
13:48:25 <olsner> though the page that links to is actually just http://www.gnu.org/software/mit-scheme/
13:51:24 -!- Taneb has joined.
13:56:56 -!- Slereah_ has joined.
13:57:05 -!- Slereah has quit (Ping timeout: 260 seconds).
13:57:25 -!- asiekierka has quit (Ping timeout: 260 seconds).
14:02:00 -!- Ngevd has joined.
14:02:17 <elliott> there must only be one of you at a time
14:03:08 <monqy> Ngevd: invite hovercraft over will you
14:03:24 -!- Taneb has quit (Ping timeout: 250 seconds).
14:04:39 -!- Ngevd has quit (Read error: Connection reset by peer).
14:08:30 <elliott> killed by duplicator accident
14:15:31 -!- MDude has joined.
14:42:35 -!- NSQX has joined.
14:43:12 <NSQX> <code.lol> HAI
14:43:30 <NSQX> <objlol> HAI CODE
14:43:38 <NSQX> <code.lol> CAN HAZ STDIO?
14:43:48 <NSQX> <objlol> AWSUM THX
14:43:56 <NSQX> <code.lol> CAN HAZ STDLIB?
14:44:03 <NSQX> <objlol> AWSUM THX
14:44:06 -!- MoALTz has quit (Quit: brb).
14:44:14 <NSQX> <code.lol> CAN HAZ WINDBLOZE?
14:44:20 <NSQX> <objlol> O NOES
14:44:32 <NSQX> <code.lol> KTHXBAI
14:44:37 <monqy> what are you doing?
14:44:37 <NSQX> <objlol> BAI CODE
14:45:12 <NSQX> <sjlol> HAI HAI
14:45:20 <NSQX> <hai.lol> CAN HAZ STDIO?
14:45:41 <NSQX> <hai.lol> VISIBLE "HAI WORLD!"
14:45:49 <NSQX> <stdout> HAI WORLD!
14:46:03 <NSQX> <hai.lol> KTHXBYE
14:46:14 <NSQX> <sjlol> BYE HAI
14:47:32 <elliott> please stop flooding the channel
14:49:22 <NSQX> A good debugger for LOLCODE would be like the pretend conversation above, where the interpreter speaks the lol-debug information in the conversation and there is also three other lolcats named "STDIN", "STDOUT", and "STDERR".
14:50:19 <oklofok> oh dear shit that would be neat
14:52:12 -!- MoALTz has joined.
14:53:14 <NSQX> What about an IRC channel where your nickname can either be "LOL*" to be treated as a lolcat typing LOLCODE for the server's interpreter or "IN*" to be treated as a standard input stream?
14:53:35 <elliott> NSQX: why did you circumvent your block?
14:53:46 <monqy> wow way to change the topic elliott!!!
14:54:18 -!- asiekierka has joined.
14:55:00 <NSQX> Well, it could be something like #IRP
14:55:07 -!- NSQX has left.
14:57:44 <elliott> <NSQX> irp -c "Type the lyrics of '99 bottles of beer on the wall
14:58:00 <monqy> is it just you and him
14:58:05 <elliott> <NSQX> irp: 99 bottles of beer on the wall.
14:58:05 <elliott> <NSQX> irp: 99 bottles of beer on the wall.
14:58:05 <elliott> <NSQX> irp: Take one down and pass it around
14:58:08 <elliott> <NSQX> irp: 98 bottles of beer on the wall.
14:58:11 <elliott> monqy: no #irp is always full of people
14:58:40 <elliott> [...] <NSQX> irp -c "Kill the irp program"
14:58:44 <elliott> <NSQX> irp: No, the system failed.
14:59:08 <monqy> is nsqx actually a collective
15:00:35 <elliott> <NSQX> irp -c "Kick me from #IRP"
15:03:23 <olsner> nice, my test suite now also passes when run with a compiled version of the metacircular evaluator from SICP
15:03:53 <elliott> i'm glad olsner is having fun
15:04:30 <olsner> with a few more primitives I should be able to run the metacircular evaluator in itself too
15:04:39 <monqy> nsqx also seemed to be having fun, coincidence??????
15:04:54 <elliott> dammit, so olsner is nsqx after all
15:05:35 <olsner> I don't want to be nsqx!
15:06:46 <olsner> elliott: anyway, all I need now is a C++ compiler written in Scheme to make a neat little loop
15:07:54 <elliott> shouldn't be too difficult
15:08:35 <olsner> nah, I don't need everything in C++ either
15:10:04 <elliott> RocketJSquirrel: I'm sure porting a compiler for an unconventional 80s architecture to another, very-dissimilar language and another, very-dissimilar architecture will be trivial.
15:10:16 <elliott> That will be the easiest part.
15:10:38 <olsner> 1. take C, 2. add templates, 3. lisp compiler
15:12:12 <olsner> (btw, the compiled REPL binary is 22MB)
15:13:04 <elliott> is it still in c++ templates
15:15:14 <olsner> elliott: yes, it's a template class that takes some combination of cons<A,B> and symbol<c,o,n,s> etc, then has a member function run(env) that runs that lisp expression
15:16:29 <olsner> also a python script that reads s-expressions and outputs them as types
15:19:10 <olsner> it's quite stupid though, so almost everything ends up boxing primitive functions then unboxing them to call them
15:19:13 <olsner> and variable lookup is done at runtime with an associative list rather than doing frame layout at compile time
15:19:46 <elliott> olsner: oh, you probably can do a lot better than that
15:19:52 <elliott> olsner: with enough constexpr you can process strings with templates
15:19:58 <elliott> so you should be able to port the parser to C++ templates, fairly easily
15:20:29 <olsner> it's already c++11 anyway
16:06:19 <elliott> olsner: NOOOOOOOOOOOOOOOOOOOOOOOOO
16:10:24 <itidus21> i posit it that being a polymath is a prerequisite for discovering/inventing calculus
16:10:44 <itidus21> i couldn't be more off topic though
16:11:49 <itidus21> << says the person who doesn't know calculus, and hasn't read a single nachlass
16:18:27 <elliott> http://maps.google.com/?t=8&utm_campaign=8bit&utm_source=yt ok this is legitimately awesome
16:24:02 -!- GhostHand has quit (Quit: Leaving).
16:25:08 <itidus21> one day wars will be fought with an rpg interface
16:26:30 <monqy> how do you know this
16:26:33 <monqy> are you from the future
16:26:43 <monqy> teach me your future ways
16:27:41 -!- Madoka-Kaname has joined.
16:36:16 -!- monqy has quit (Quit: hello).
16:38:20 -!- hagb4rd has joined.
16:39:35 <itidus21> its nearly same effect as time travel
16:41:07 -!- MSleep has joined.
16:45:16 -!- MDude has quit (Ping timeout: 276 seconds).
16:53:56 -!- Tiktalik has changed nick to Gonetalik.
17:12:46 -!- Frooxius_ has joined.
17:13:27 <elliott> "Sad But True: We Can't Prove When Super Mario Bros. Came Out"
17:14:14 -!- Frooxius has quit (Ping timeout: 245 seconds).
17:14:27 -!- Frooxius_ has changed nick to Frooxius.
17:18:57 <elliott> http://www.gamasutra.com/view/feature/167392/sad_but_true_we_cant_prove_when_.php
17:26:46 -!- asiekierka has quit (Remote host closed the connection).
17:37:52 <elliott> I noticed on page Db-g7 that there was a deletion tag. I would like to vote against this tag, as I believe it sort of violates Assume good faith. I also believe the template is very useful. These matters seem important to discuss.
17:37:52 <elliott> Walex03. Talking, working, friending. 20:06, 22 March 2012 (UTC)
17:38:04 <elliott> (db-g7 is a template containing a deletion tag to be put on other articles.)
18:10:06 -!- Gonetalik has changed nick to Tiktalik.
18:14:04 -!- MSleep has changed nick to MDude.
18:36:05 -!- hagb4rd has quit (Ping timeout: 246 seconds).
18:36:28 -!- hagb4rd has joined.
19:11:25 -!- Taneb has joined.
19:11:37 <Taneb> Also, dammit elliott!
19:19:33 <Taneb> Thus causing me to click the link.
19:20:00 <elliott> Taneb: i take it you've seen NSQX's diabolical block evasion, then
19:20:13 <Taneb> (is it bad that when you're being tricky I don't fall for it, but TWICE I've fell for unobfuscated links?)
19:20:25 <Taneb> Truly he is a master of trickery
19:20:33 <elliott> it seems he is determined to flat-out ignore me when i ask him why he did it on IRC, so i predict an unhappy ending to this saga
19:20:37 <elliott> Taneb: And yes, that's bad.
19:20:46 <Taneb> It reminds me of that time that sock puppeteer ended up on the IWC forums
19:23:36 <Taneb> I hadn't thought of an interesting story
19:24:00 <Taneb> There was just some blatant sock-puppetry to do with some unlicensed translations of Darths and Droids on a commercial website
19:24:50 <elliott> maybe he just doesn't like me
19:24:59 <elliott> i assign the job of asking him why he circumvented his block next time he comes in to... Taneb
19:25:57 <elliott> you're not meant to say yay!
19:26:16 <Taneb> I can say "Yay!" whenever I like.
19:26:49 <oklofok> i can say "yay" whenever anyone likes
19:27:22 <Taneb> elliott has a speech impediment.
19:27:39 <elliott> what an explody house this is
19:28:08 -!- Taneb has quit (Read error: Connection reset by peer).
19:28:13 -!- Deewiant_ has joined.
19:28:49 -!- Deewiant has quit (Ping timeout: 245 seconds).
19:29:41 -!- Taneb has joined.
19:30:10 <Taneb> Sorry, my house exploded
19:30:59 -!- Taneb has quit (Read error: Connection reset by peer).
19:40:12 -!- variable has joined.
19:40:12 -!- Deewiant_ has quit (Ping timeout: 260 seconds).
19:45:57 -!- Deewiant has joined.
20:07:18 -!- pikhq has joined.
20:07:44 -!- pikhq_ has quit (Ping timeout: 248 seconds).
20:13:43 <elliott> happy australian mailman mailing list memberships reminder day
20:17:50 -!- pikhq has quit (Ping timeout: 260 seconds).
20:28:10 <elliott> can someone explain drive-ins to me
20:29:14 <oklofok> you drive in and then you shoot everyone.
20:29:40 <oklofok> no wait sorry i always confuse those two
20:29:55 <oklofok> that's why they don't let me into the local ganster dude league anymore.
20:30:01 <oklofok> because i just keep buying them burgers
20:30:35 <elliott> no the kind where they show films
20:32:34 <elliott> http://en.wikipedia.org/wiki/Drive-in_theater
20:32:38 <oklofok> can someone explain drive-ins to me
20:32:45 -!- pikhq has joined.
20:33:17 <oklofok> i've seen those on the simpsons and stuff but i assumed they just existed in like america and other made-up countries
20:36:08 -!- hagb4rd has quit (Ping timeout: 246 seconds).
20:39:02 <elliott> well they don't exist here
20:39:49 <oklofok> finland doesn't have them, i just checked
20:40:36 <oklofok> well i checked a statistically significant portion of my house at least
20:51:38 -!- calamari has joined.
20:55:00 -!- zzo38 has joined.
20:56:39 <zzo38> What is #25 in "unblocked #25 (remove autoblock block evasion to fix config. issue)"?
20:56:52 <elliott> zzo38: The number of the autoblock.
20:57:00 <elliott> zzo38: Autoblock IPs are hidden (even from admins/crats) for privacy reasons.
20:57:15 <elliott> (An autoblock just means that whenever a blocked user accesses the wiki from an IP, that IP gets blocked too.)
20:57:41 <elliott> The autoblock still had the previous incorrect settings (banning talk page editing) due to the configuration issue, even though I fixed the main block, so I just removed it.
21:07:12 -!- ais523 has joined.
21:07:34 <zzo38> http://esolangs.org/w/index.php?title=Talk:Var%27aq&diff=31646&oldid=8256
21:08:18 <ais523> elliott: bleh, this is a bad time for me to sort it out
21:08:18 <lambdabot> ais523: You have 1 new message. '/msg lambdabot @messages' to read it.
21:08:22 <ais523> but I'll look at it april 2
21:08:23 <lambdabot> elliott said 1d 5h 20m 22s ago: Please check your MemoServ messages with /msg MemoServ read new, thanks!
21:08:32 <ais523> heh, I check MemoServ messages /before/ lambdabot messages
21:08:50 <elliott> ais523: well, I don't really need you to do anything
21:09:10 <elliott> ais523: just know what's happened and let me know if you think the response needs changing
21:09:10 <elliott> this should probably go in /msg, though
21:09:27 <zzo38> In here is March 31
21:09:29 <ais523> I'll msg you when I'm ready to look at it, probably tomorrow or the day after
21:10:59 <elliott> yay an article with tables waiting to be wikitabled
21:11:53 <elliott> oh, it's in html table syntax :( never mind
21:14:04 <Sgeo> I should just check RecentChanges
21:14:10 <Sgeo> So I have an idea what you're talking about
21:15:17 <Sgeo> I have no idea what you're talking about, I don't see any table syntax
21:16:50 <Sgeo> What was the thing about waiting to be wikitabled?
21:19:52 <Sgeo> How is that waiting to be wikitabled or confused for waiting to be wikitabled? What does waiting to be wikitabled mean? I'd assume converted from blah to MediaWiki tables, but your next statements suggest not.
21:20:03 <elliott> it makse them the not uglies
21:20:04 -!- derdon has quit (Remote host closed the connection).
21:20:09 <elliott> but then they looked weird without ths
21:20:15 <elliott> but i didnt want to add ths because theyre in html syntax
21:21:59 -!- MSleep has joined.
21:25:25 -!- MDude has quit (Ping timeout: 276 seconds).
21:45:11 -!- monqy has joined.
21:49:44 * pikhq notes that OpenGL sucks
21:53:47 -!- Deewiant has quit (Ping timeout: 260 seconds).
22:03:20 -!- Deewiant has joined.
22:12:54 <ais523> hey, and it's far from finished or polished yet, but it works: https://gitorious.org/nitrohack/ais523/blobs/raw/7b5711e6232fbb0949452555a3db6b33bd585824/aimake
22:12:58 <ais523> aimake is no longer vaporware!
22:13:10 <pikhq> ais523: The API is really poorly designed.
22:13:11 <ais523> I'm not sure if it counts as eso or not
22:13:57 <ais523> pikhq: which OpenGL API?
22:13:57 -!- MSleep has changed nick to MDude.
22:14:07 <ais523> the old outdated functions you aren't meant to use any more, or the VBO-and-shader stuff?
22:14:51 <elliott> ah, another unreadable multi-thousand-line ais source fil
22:14:53 <pikhq> Whatever the fuck I'm using.
22:15:03 <elliott> also one that inexplicably breaks vim's syntax highlighting
22:15:06 <ais523> elliott: it's a single file for a reason
22:15:13 <elliott> ais523: you say that every time
22:15:24 <ais523> and I've been editing it in Kate, which is the only program I've found that's capable of actually handling the syntax highlighting
22:15:30 <pikhq> The one where it's 30 lines of code to just draw a single surface with a texture on it.
22:15:31 <ais523> and even then, I had to go back to 5.78 syntax
22:15:35 <ais523> I wasn't even trying to be awkward!
22:15:58 <ais523> pikhq: if you're drawing one surface at a time, you're doing it wrong (because GPUs can't do that efficiently)
22:16:08 <pikhq> ais523: I'm only drawing one surface.
22:16:27 <ais523> then possibly OpenGL is the wrong thing to be using
22:16:53 <elliott> ais523: your advice is helpful like my advice is
22:16:56 <pikhq> It is the only well-supported way of drawing 2D graphics with vsync on X.
22:17:01 <pikhq> Literally the only one.
22:17:25 <elliott> pikhq: What about that Xv thing?
22:17:30 <pikhq> elliott: "Well-supported".
22:17:46 <elliott> where is xv not supported?
22:17:56 <pikhq> elliott: Xv varies from perfectly reasonable *if* you're in YUV, or completely borken, depending on your driver.
22:18:16 <elliott> "About 600 deaths have occurred in the Grand Canyon since the 1870s." whoa, that's all?
22:18:50 <pikhq> The proprietary Radeon drivers, for instance, don't do vsync, and convert from YUV to RGB wrong.
22:19:22 <pikhq> (Xv only really works on YUV colorspace framebuffers; it's theoretically capable of handling RGB, but most drivers don't support this)
22:19:29 <Sgeo> http://www.reddit.com/r/programming/comments/782mt/numbers_that_cannot_be_computed_by_a_program/c05xjqb
22:19:32 <elliott> what 2d thing are you drawing anyway
22:19:42 <elliott> Sgeo: will i regret clicking that link?
22:19:45 <pikhq> I'm just drawing a 320x240 framebuffer out of a VM.
22:19:54 <ais523> elliott: I'll check it for you
22:20:07 <pikhq> And it will tear unless I use OpenGL.
22:20:14 <elliott> ais523: thanks, I'll click it then
22:20:26 <Sgeo> elliott, do you regret arguments that involve someone incredibly thick?
22:20:29 <elliott> that way, I can blame Sgeo
22:21:07 <elliott> Sgeo: well, you see, when i want stupidity i go to, e.g. reddit and click the comment links
22:21:09 <oklopol> []andreasvc 0 points 3 years ago
22:21:09 <oklopol> The integers are not all computable. What you say amounts to offering a haystack when someone asks for a needle. A number is computable if there can be a finite program which can calculate the nth digit of that number, given n. This doesn't go for the busy beaver numbers, since the halting problem is undecidable.
22:21:18 <elliott> #esoteric links i generally click in the hopes that they're not stupid
22:21:25 <elliott> i would be totally disappointed if i found intelligent discussion on reddit
22:21:36 <oklopol> why are these people allowed on my internet
22:21:45 <ais523> elliott: so opinion on aimake?
22:21:49 <ais523> apart from the coding style?
22:21:51 <pikhq> oklopol: But... That... Gah
22:21:55 <ais523> Kate's autoindentor is a little broken
22:22:00 <ais523> and its block collapser is worse
22:22:07 <elliott> ais523: there's no way i'm going to read it, but i might throw it at some C programs to prove how useless it is
22:22:18 <ais523> note: after editing aimake in Kate for a few hours, the cursor starts going backwards
22:22:21 <ais523> and I haven't found a fix for this
22:22:24 <oklofok> the fucking busy beaver numbers
22:22:41 <MDude> I think SDL might be pretty well supported, but from what I've read, I'm not sure if it's that great, since it doesn't have functions for pushing individual pixels.
22:22:41 <ais523> (i.e. typing a letter puts the cursor before the letter, backspace still deletes leftwards, left and right don't really do anything sensible)
22:22:50 <elliott> "It is trivial to write a program that computes 6 to the nth digit." "No, there's not even an algorithm for it. If you come up with one you would have solved the halting problem."
22:22:54 <pikhq> MDude: Because X sucks.
22:23:04 <MDude> And also a lot of things I look up relating to it presume you want to use it with SDL.
22:23:05 <elliott> MDude: it has pointers for pushing individual pixels
22:23:18 <ais523> elliott: I don't see how you can figure out the nth digit of 6 without knowing what n is
22:23:21 <ais523> unless you have multiple attempts
22:23:22 <elliott> pushing individual picturse is terrible, though
22:23:54 -!- augur has quit (Remote host closed the connection).
22:23:55 <elliott> picturse, pixels, what's the difference
22:24:14 <oklofok> i fucking hate this andreasvc guy
22:24:55 <MDude> That's true, I guess it's more that the tutorial I was reading is about 2D stuff, and then alter I remembered I wanted ot make a 3D thing.
22:25:11 <elliott> SDL is useful as a windowing/input library though
22:25:15 <elliott> you can use it with opengl
22:25:26 <pikhq> They suck, but it functions.
22:25:33 <elliott> pikhq: you didn't answer my q tho :'(
22:25:37 <zzo38> I happen to like SDL
22:25:50 <pikhq> zzo38: Never done audio in it I see. :)
22:25:57 <zzo38> I have done audio in SDL.
22:26:09 <pikhq> Then you are a masochist.
22:26:46 <pikhq> It's harder and less capable than OSS. Therefore it sucks.
22:26:48 <elliott> pikhq: oh i didn't ask it i guess
22:26:52 <elliott> i was asking what 2d thing you are drawing
22:27:07 <pikhq> 16:19 < pikhq> I'm just drawing a 320x240 framebuffer out of a VM.
22:27:19 <elliott> Sgeo: why the fuck are you reading 3-year-old discussions on reddit
22:27:25 <elliott> i only just now noticed those timestamps
22:27:58 <pikhq> https://github.com/pikhq/cmako
22:28:01 <lambdabot> Local time for elliott is Sat Mar 31 23:27:59
22:28:12 <zzo38> Here is a program using SDL with audio: http://zzo38computer.cjb.net/prog/BytePusher/BytePusher.w I happen to like the way audio is done in SDL, even if you dislike it.
22:28:13 <oklofok> elliott: andreasvc's crime will never expire.
22:28:14 <elliott> good, i can still be annoyed at april fools jokes being premature for another 33 minutes
22:28:17 <pikhq> An implementation of https://github.com/JohnEarnest/Mako
22:28:48 <elliott> thanks for the opportunity
22:29:31 <Sgeo> elliott, the article was linked on Fark, and the article linked to the discussion
22:29:41 <Sgeo> http://www.fark.com/comments/7025689/ConsoleWrite0-while-true-ConsoleWrite142857
22:29:42 <elliott> ah, another ill of the world i can blame on fark
22:31:06 <elliott> ais523: have you seen Google Maps' april 1st thing? it's very good
22:31:08 <zzo38> The other two implementations of BytePusher are using Allegro, which many people prefer.
22:32:01 <ais523> I probably won't look at it, but will probably read the coverage of it in the news afterwards
22:32:21 <oklofok> always living on the edge, eh
22:33:04 <Sgeo> I'm confused. Can any real number be described by an infinite number of digits?
22:33:22 * Sgeo thinks a bit more
22:33:54 <elliott> no, some of them have infinity plus one
22:33:57 <zzo38> I would think it can be
22:34:28 <ais523> Sgeo: yes, cardinalities of the reals and of infinite strings of elements drawn from a finite alphabet are both aleph-one
22:35:21 <Sgeo> Is my third comment http://www.fark.com/comments/7025689/ConsoleWrite0-while-true-ConsoleWrite142857 correct?
22:35:32 <Sgeo> "I think I got it. There exist numbers which cannot be fit into any finite representation, no matter how many (finite number) of bits are in the representation."
22:36:05 <Sgeo> I was incompetently replying to someone
22:36:09 <monqy> I'm not going to bother clicking
22:36:12 <elliott> wait did Sgeo not know the reals were uncountable before today
22:36:25 <elliott> as oklofok would say, even i know that
22:36:39 <Sgeo> elliott, I brainfarted on whether reals could be described with an infinite number of digits
22:37:08 <ais523> elliott: they're exactly uncountable, they're no more than uncountable
22:37:23 <ais523> actually that's undecidable
22:37:42 <Sgeo> I just noted who wrote the article in the first place.
22:37:45 <Sgeo> The Robozzle guy!
22:37:51 -!- Nisstyre has quit (Read error: Connection reset by peer).
22:37:51 <elliott> Sgeo: i meant "There exist numbers which cannot be fit into any finite representation, no matter how many (finite number) of bits are in the representation.""
22:38:01 <elliott> aka "the reals aren't countable"
22:40:14 -!- Nisstyre has joined.
22:45:09 <ais523> $ find . -name '*.aimk' -or -name 'aimake.config' | xargs wc -l | xargs echo
22:45:10 <ais523> 3 ./executables.aimk 55 ./aimake.config 7 ./libnethack/util/makedefs.aimk 20 ./libnethack/dat/datafiles.aimk 85 total
22:45:23 <ais523> and most of them are newlines to make it more readable, or comments
22:45:40 <ais523> $ find . -name 'CMakeLists.txt' | xargs wc -l | xargs echo
22:45:41 <ais523> 50 ./nethack_server/CMakeLists.txt 106 ./nethack/CMakeLists.txt 70 ./CMakeLists.txt 34 ./libnethack_client/CMakeLists.txt 18 ./libnethack/CMakeLists.txt 111 ./libnethack/util/CMakeLists.txt 45 ./libnethack/src/CMakeLists.txt 118 ./libnethack/dat/CMakeLists.txt 552 total
22:46:45 <elliott> ais523: Those CMakeLists files won't stop working when aimake changes its heuristics.
22:47:40 <ais523> I /think/ I fixed that now :)
22:48:07 <ais523> (before the success or fail of a build sometimes depended on hashtable ordering)
22:48:19 <ais523> (still does if you have an exact tie, but you can disambig in the config file now)
22:49:46 -!- Nisstyre has quit (Quit: Leaving).
22:53:21 -!- Nisstyre has joined.
22:54:18 <elliott> I think "ugh" adequately summarises my feelings.
22:56:46 <elliott> http://www.reddit.com/r/haskell/comments/rmw1d/sometimes_it_is_hard/
23:11:29 -!- Patashu has joined.
23:12:23 <monqy> im, not going to read that
23:17:53 -!- nortti has quit (Quit: ChatZilla 0.9.88.1 [Firefox 11.0/20120312200651]).
23:19:53 <monqy> im, not going to read that
23:22:37 -!- Phantom_Hoover has joined.
23:23:12 <elliott> Ohhhh, mk is that guy who just wanted to learn monads, not Haskell.
23:23:15 <lambdabot> Phantom_Hoover: You have 10 new messages. '/msg lambdabot @messages' to read them.
23:23:17 <elliott> That explains why he's stupid.
23:23:21 <elliott> Phantom_Hoover: ONE MESSAGE PER MINUTE
23:25:32 -!- augur has joined.
23:27:07 <ais523> elliott: they are useful in other languages sometimes…
23:27:21 <elliott> ais523: You had to be there, man.
23:27:28 <elliott> ais523: Also, that doesn't mean you should ask #haskell about it.
23:28:13 <elliott> Then I won't call you stupid in #esoteric.
23:28:21 <elliott> RocketJSquirrel: About that robots.txt...
23:29:19 <ais523> elliott: wow, I think I can scam that statement, you forgot to put a time limitation on it
23:29:29 <RocketJSquirrel> elliott: Right now if I search for 'elliott esoteric haskell' on Google, the logs aren't in the first page.
23:29:55 <ais523> meh, I will search for them with locate
23:30:27 <elliott> RocketJSquirrel: I'll just start up my blog, Esoteric Haskell with Elliott Hird.
23:30:29 <ais523> gah, /far/ too many results
23:30:41 <ais523> it interpreted it as an or
23:30:52 <ais523> and so many things on my computer are named esoteric, haskell, or elliott somewhere in their path
23:31:11 <ais523> $ locate "elliott esoteric haskell"
23:31:40 <elliott> <ais523> and so many things on my computer are named esoteric, haskell, or elliott somewhere in their path
23:31:52 <ais523> mostly the other two, it turns out
23:31:57 <ais523> there's a /home/ais523/annoy-ehird
23:32:02 <ais523> I forget what it contains nowadays
23:32:08 <ais523> and don't really want to look, as it'd be less funny if I found out
23:32:12 <ais523> but that's "ehird", not "elliott"
23:32:19 * elliott wants to know what's in it now.
23:32:55 -!- augur has quit (Remote host closed the connection).
23:34:46 <shachaf> Phantom_Hoover: North America.
23:40:03 -!- augur has joined.
23:40:05 <elliott> Phantom_Hoover: i would install that though?
23:40:18 <elliott> I mean, it'd be better than adblock.
23:59:27 -!- augur has quit (Remote host closed the connection).
23:59:46 <ais523> bleh, there's no command to reorder virtual desktops in gnome 2 + compiz, is there?
23:59:51 <ais523> I have them in the wrong order