00:00:31 <ehird> Windows has a little bit of POSIX but not enough for those. Well, vi will compile fine and coreutils shouldn't require ginormous patches, but bash and rxvt? Crapshoot.
00:01:01 <ehird> There's a native X server for Windows, Xming, but rxvt has been patched to have a Windows GUI so that isn't even needed; it's the rest of rxvt that'd be the problem.
00:09:04 <pikhq> Wow. Protesters at Washington yesterday, arguing vaguely against the protest, seem to think that their protest was the "biggest crowd to march on Washington".
00:09:14 <pikhq> Some ten thousand people.
00:09:36 <ehird> Protesting against the protest? Wut
00:09:43 <pikhq> s/protest/President/
00:09:48 <ehird> Aw. This way's more fun.
00:09:52 <pikhq> ... I've seen crowds orders of magnitude larger march on Washington.
00:09:56 <ehird> "Concerned Citizens Against Concerned Citizens Against Taxes"
00:10:10 <ehird> "Taxpayers Opposing Concerned Citizens Against Concerned Citizens Against Taxes"
00:10:19 <ehird> "March For the Opposition of Taxpayers Opposing Concerned Citizens Against Concerned Citizens Against Taxes"
00:12:34 <ehird> Polity partics (he wrote, before looking at that and sporrecting his coonerism), party politics in the US are just tedious. Both parties are so crazy that there's no way to comprehend them, so it's best to ignore them.
00:16:12 <pikhq> The Democrats are less crazy. (... in much the same way that a schizophrenic is less crazy than a schizophrenic serial killer)
00:16:50 <ehird> In much the same way as a schizophrenic serial killer is less crazy than a schizophrenic serial killer that urinates on their victims beforehand.
00:20:12 <oerjan> well i'm not schizophrenic, and neither am i, and besides they never found any bodies
00:21:17 <oerjan> also, you are just imagining you are reading this. maybe _you_ are the schizophrenic one.
00:22:28 <oerjan> hm, i seem to be hearing a voice. it said, "you're right." but then, that's what they always say. they don't dare anything else.
00:23:36 -!- coppro has quit (Remote closed the connection).
00:23:58 -!- FireFly has quit ("Later").
00:25:00 <ehird> http://code.google.com/p/cygwine/ super!
00:25:14 <ehird> don't know how it's more like wine, though
00:25:34 <ehird> "We don't really need two different installers, so I hope you find a way to merge your efforts with those working on setup.exe."
00:25:51 <oerjan> you need to store it for a few years before using it.
00:25:55 <ehird> why not get rid of shitup.exe and rename this setup.exe, there's nothing salvageable in the other...
00:29:53 <GregorR> When I saw "cygwine", I thought of dozens of awesome possibilities.
00:30:44 <HackEgo> * Cygnus (, genitive Cygni ) is a northern constellation. Its name is Latin for swan. One of the most recognizable constellations of the northern ... \ [18]en.wikipedia.org/wiki/Cyg \
00:31:20 <oerjan> how do you make wine from swans anyway
00:32:54 <GregorR> 2) Throw out the skin and muscle
00:32:59 <GregorR> 3) Put the guts in water and shake vigorously
01:03:49 <pikhq> GregorR: I don't think it works that way.
01:03:54 <pikhq> (not enough sugar)
01:05:04 <oerjan> except when they swoop
01:06:27 -!- Sgeo has joined.
01:08:05 -!- ehird has quit (Remote closed the connection).
01:09:09 -!- ehird has joined.
01:09:40 <ehird> GregorR: I'm pretty sure wine will run on cygwin, anyway
01:09:47 <ehird> and cygwin runs in wine
01:09:52 <ehird> so you can nest as much as you want
01:10:02 <ehird> needs a preferences Dialog
01:10:28 <ehird> Speed - Slow [ ] Astronomically slow
01:11:24 <ehird> " I like the nice wx-based GUI, that's really pretty." // yeah, apart from the really crickly edges on the button images...
01:11:33 <ehird> Are people blind or something?
01:12:18 <Ilari> Is "speed of fusion in star 100 times the mass of Jupiter" "Astronomically slow"? :->
01:12:41 <ehird> Astronomically obviously means "grossly hugely".
01:13:10 * Sgeo needs to force himself not to copy/paste
01:14:16 <Sgeo> It leads to bugs and is ugly
01:14:17 <Sgeo> http://sgeo.diagonalfish.net/crack_pairings.htm
01:14:37 <Sgeo> Including table 4 incorrectly referring to table 3
01:14:46 <ehird> "Pregnancy (use male pregnancy if needed)"
01:15:07 <Sgeo> Go blame the people at http://www.giantitp.com/forums/showthread.php?t=111631
01:15:09 <ehird> (The above line is singular justification for not setting +c on this channel and mandating that ^I is italics, not inverse.)
01:16:04 <ehird> Sgeo: wtf, did you really not think of writing one function over many tables?
01:16:30 <Sgeo> ehird, I did, but some of the tables have things like using the initial table, and others don't.
01:16:31 <ehird> also, using Math.floor on random values?
01:17:17 <Sgeo> It's been a while since I did any Javascript. I used another page that I wrote, again a while after I did any Javascript, to jog my memory
01:17:31 -!- coppro has joined.
01:17:43 <ehird> you fail at randomness :||
01:18:38 <Sgeo> How is it fail? I really don't know
01:18:54 * ehird notes http://code.google.com/p/cygwin-rurban/source/browse/trunk/contrib/cyginstall/cyginstall for his own purposes
01:19:14 <ehird> Sgeo: you should round < 5 down, > 5 up to retain the ... thingy of the original number
01:19:22 <ehird> otherwise i'm pretty sure it's unevenly distributed
01:19:30 <ehird> for ==5, do whatever
01:27:46 * oerjan doesn't think that matters much except at the edges, where it might depend
01:28:18 <ehird> just use Math.round or something :P
01:29:04 <oerjan> if your floats are in a range [0 .. n) then you definitely want floor to get {0, ..., n-1} values
01:29:17 -!- augur has quit (Read error: 54 (Connection reset by peer)).
01:29:24 <oerjan> rounding would fail by having half length intervals at the edges
01:29:36 <ehird> you have to weak the range afterwards in JS
01:29:39 <ehird> which is what he does
01:29:43 <ehird> so it's irrelevant
01:30:19 <oerjan> nevertheless you want those int values you actually use to come from same length float subintervals
01:30:36 <ehird> JS only has one range, iirc
01:30:43 <ehird> for random numbers
01:31:33 <ehird> [] is inclusive, () is exclusive, right?
01:31:41 <ehird> I _think_ it includes 1.0, but I'm not sure
01:32:29 <oerjan> floor(n*rnd) gives a uniform integer as long as you cannot hit n
01:32:39 <oerjan> but maybe best to check anyway
01:36:16 <Sgeo> So if Math.random() returns between [0, 1) then my code is correct?
01:36:33 <Sgeo> (well, except for the whole idiotic table_1 table_2 table_3 etc. thing)
01:38:47 <Sgeo> Because Math.random(), while it will never ever return 1, will occasionally (VERY occasionally) return zero.
01:38:50 <Sgeo> http://board.flashkit.com/board/showthread.php?t=545326
01:39:33 <ehird> Actionscript is an implementation of ECMAscript.
01:39:38 <ehird> So don't dismiss arguments based on it.
01:39:41 <ehird> Anyway, just test.
01:40:08 <ehird> for (i=0;i<100000;i++) if (Math.random()==1) alert("fuck")
01:40:09 <Sgeo> Any errors based on the number being too high are (with one exception) absorbed by an else
01:41:24 <Sgeo> I didn't get "fuck" when I changed that 1 to a 0, though
01:41:50 <ehird> Then I guess it's (0, 1] which is weird. Just look it up in the ECMAscript specification.
01:42:04 <ehird> http://www.ecma-international.org/publications/files/ECMA-ST/ECMA-262.pdf
01:42:12 <Sgeo> I didn't get fuck when it was one
01:42:24 <Sgeo> I might have been unlucky enough to not get zeros
01:42:56 <ehird> "Returns a number value with positive sign, greater than or equal to 0 but less than 1"
01:43:02 <ehird> blah blah blah etc.
01:43:15 <ehird> Asztal: hey, does Sumatra really not have copy/paste of text?
01:43:38 <ehird> It can search but not select...
01:43:57 <ehird> Except it selects as a rectangle
01:44:04 <ehird> Returns a number value with positive sign, greater than or equal to 0 but less than 1, chosen randomly
01:44:06 <ehird> or pseudo randomly with approximately uniform distribution over that range, using an
01:44:08 <ehird> implementation-dependent algorithm or strategy. This function takes no arguments.
01:44:16 <ehird> That's a, wossname, retarded interface.
01:44:37 <Sgeo> Quoting that parrot from Eric?
01:45:04 <ehird> Well, just the wossname bit.
01:45:23 <Sgeo> THat's what I meant
01:46:15 * ehird wonders how hard it'd be to build ghc on cygwin
01:46:23 <ehird> Well, it'd be porting, essentially, but...
01:52:50 -!- Pthing has quit (Remote closed the connection).
01:55:34 <ehird> "That's the stupidest thing I've ever heard of. You must be poor." // an odd process of deduction there
01:58:11 <ehird> So, does anyone know if Lua still sucks?
01:58:45 <Sgeo> OSMP was supposed to use Lua
01:58:56 * oerjan didn't know it sucked in the first place
02:01:00 <Sgeo> http://www.solipsis.org/ I didn't remember Solipsis being 3d.. or anything other than a bunch of dots
02:02:51 <ehird> (Sgeo immediately attempts to travel forwards in time so he can be nostalgic about it)
02:03:21 <oerjan> and he succeeds! - very slowly though
02:04:24 <Sgeo> There are things I don't get nostalgic about!
02:04:36 <Sgeo> I don't get nostalgic over IMVU
02:06:30 <Sgeo> I don't get nostalgic over the time that I was without computer access for a long period of time
02:06:54 <ehird> Neither of those are open-ended 3D virtual realities
02:08:17 <Sgeo> I'm not especially nostalgic over OSMP, but it never really worked in a large-scale manner
02:08:38 <ehird> I like how you're only nostalgic about good things.
02:08:41 <ehird> Takes caliber, that.
02:09:39 <Sgeo> Actually, I'm feeling some OSMP-related nostalgic, but more about the ideas I had for it then OSMP itself
02:33:44 -!- oerjan has quit ("Good night").
02:42:30 <Sgeo> I saw the typo, but figured you'd try to make some joke out of it, instead of a boring correction
02:43:05 <ehird> far too lazy for that
02:44:04 <ehird> CMake has several forms of documentation. There is a book Mastering CMake by Ken Martin and Bill Hoffman (ISBN 1-930934-16-5) which is published by Kitware and can be purchased through Kitware's e-store or through Amazon. // which costs $52
02:44:10 <ehird> and the "documentation" is generated from the source code
02:44:17 <ehird> in conclusion, the cmake documentation costs $52+shipping
02:44:22 <ehird> what an "open" project
02:46:10 <Sgeo> You know what, I'm done delaying on my homework. Restarting my computer. I won't be back until one or two paragraphs are done.
02:46:46 <ehird> Homework: because the whole day wasn't enough!
02:47:19 -!- Sgeo has quit (Read error: 54 (Connection reset by peer)).
02:47:45 * coppro is glad not to have any homework
02:47:51 <coppro> at least, none due tomorrow :P
02:48:58 <coppro> the whole day is never enough
02:49:59 <ehird> No, no, I mean that "homework" is the (wrong) answer to "keeping kids in our environment all day and prodding them with repeated reciting of textbooks isn't teaching them anything. What do we do?"
02:53:50 <ehird> Incidentally, my disk can store 385.92 miles.
02:54:00 <ehird> http://neologasm.org/neologasm/2007/07/fortraninch.html via http://andrewducker.livejournal.com/1227424.html
02:54:07 <ehird> Well, anti-via, rather. "refers to", maybe.
03:38:20 -!- coppro has quit (Remote closed the connection).
03:45:32 <ehird> Amusingly, CygWine uses mingwm10.dll, i.e., some MinGW thing.
03:52:09 <ehird> CygWine seems to only do 1.5, *sigh*
03:52:30 <ehird> I could patch the code to use -2, but I'd need MinGW and wxWidgets dev libs.
03:52:47 <ehird> Also, it uses goddamn cmake.
04:03:11 <ehird> ("If nothing goes overly wrong, the official 1.7.1 release goes public still within the the first half of 2009.")
04:06:20 -!- Sgeo has joined.
04:12:10 <ehird> http://tommd.wordpress.com/2009/09/13/kernel-modules-in-haskell/
04:12:13 <ehird> pikhq will love this.
04:12:22 <ehird> Type-safe, garbage collected, referentially transparent Linux kernel modules?
04:12:38 <ehird> I think my hostility to Linux just evaporated.
04:14:36 <ehird> http://www.reddit.com/user/bonch, a single-minded dimwit.
04:17:55 <GregorR> TAKE A DRINK FOR EVERY HASKELL LINK
04:18:09 <ehird> TAKE A DRINK FOR EVERY DRINK
04:18:22 <ehird> aka alcoholism in 1 step
04:18:41 <ehird> GregorR: also, you forgot the newline.
04:20:01 <GregorR> IT'S NOT \ A FUCKING POEM \ I'LL ARRANGE IT \ HOWEVER I ... damn, can't rhyme here.
04:20:15 <ehird> GregorR: Try "SHIT"
04:20:28 <ehird> Also, not doesn't rhyme with poem. So you failed already!
04:20:29 <GregorR> Apparently "poem" rhymes with "shit"
04:20:40 <ehird> What's your rhyming scheme, not a single line there rhymes
04:20:48 <ehird> Not doesn't rhyme with it, dude.
04:21:27 <GregorR> IT'S NOT \ A POEM, NITWIT \ I'LL ARRANGE IT \ HOWEVER I SEE FIT
04:22:29 <ehird> IT'S NOT \ A FUCKING POEOT \ I'LL ARRANGE IT \ HOWEVER I SHIT
04:22:59 <GregorR> Apparently in ehird's view of the universe, poems have to be in the AABB rhyme scheme.
04:23:17 <ehird> No, I just wanted it to be, so I could say poeot and use the it/shit rhyme.
04:23:33 <ehird> Also, to imply that you can arrange anything in the manner in which you shit.
04:23:44 <GregorR> There once was a man from Nantucket \ who oh damn who cares fuck it \ I've given up on this poem \ so this won't rhyme \ who the fuck cares.
04:24:03 <ehird> Also, "IT'S NOT \ A POEM, NITWIT \ I'LL ARRANGE IT \ HOWEVER I SEE FIT" it's ABBB.
04:24:22 <ehird> >There once was a man from Nantucket \ who oh damn who cares fuck it \ I've given up on this poem \ so this won't row, um \ who the fuck cares.
04:24:33 <ehird> I SUBVERTED YOUR INTENT
04:24:38 <GregorR> Oh damn, I didn't even intend for the third line to rhyme with the second and fourth :P
04:25:06 <GregorR> Except that the last line of a limerick is supposed to rhyme with the first.
04:25:29 <ehird> There once was a man from Nantucket \ who oh damn who cares fuck it \ I've given up on this poem \ so this won't row, um \ not even if you use a bucket.
04:26:09 -!- notostraca has joined.
04:26:15 <ehird> There once was a man with an orange \ shit
04:26:20 -!- bsmntbombdood___ has changed nick to bsmntbombdood.
04:26:58 <ehird> Not many other paths to go down, really.
04:27:03 <GregorR> Emo poetry: I don't know how to rhyme \ I just cry into my pillow \ in my parent's basement \ maybe I should get another piercing.
04:27:51 <ehird> Emoetry: My dream's skies are like flames / they're hot and lick around / UNLIKE MY BOYFRIEND / who is never around / I'LL KILL MYSELF WITHOUT MAKING A SOUND / I can so rhyme around with around
04:28:04 <ehird> / Because I'm nonconformist
04:28:19 <ehird> / And not bound by any rules
04:28:30 <ehird> It's good because it's purposefully terrible.
04:28:35 <ehird> *It's good because it's purposefully terrible!
04:29:59 <ehird> Though just to rhyme, just this one time
04:30:05 <ehird> What is Black Dynamite?
04:30:08 <ehird> *note: don't bother answering
04:30:13 <ehird> I just wanted to make some more wonderful poetry!
04:30:43 <ehird> Metaploitation: a ploitation of the whole genre of anything-ploitation, including itself.
04:31:02 <notostraca> http://en.wikipedia.org/wiki/Black_Dynamite
04:31:22 <ehird> No, I invented a new genre.
04:31:25 <ehird> That only goes one level.
04:31:37 <notostraca> NSFW: http://www.worstpreviews.com/headline.php?id=10932
04:31:50 <ehird> "Removed by Sony Pictures"
04:31:56 <ehird> Best trailer ever.
04:33:02 <notostraca> http://www.firstshowing.net/2008/11/23/worth-watching-nov-23-badass-black-dynamite-trailer/
04:33:16 <ehird> I have no audio, so this better be good enough without.
04:36:15 <ehird> I agree, which is why I closed it.
04:36:38 <ehird> No, I don't think that will be necessary.
04:36:59 <notostraca> the movie looks pretty awesome in an offensive way
04:38:03 <GregorR> `imdb Gayniggers from Outer Space
04:38:07 <HackEgo> Gayniggers from Outer Space (1992) \ User Rating: \ \ 6.0/10 [91]3,296 votes \ -- \ Plot: \ \ Extraterrestrial beings travel the galaxy to free men "oppressed" by females to make way for an entirely-homosexual society. | [104]full synopsis
04:38:23 <ehird> Black Dynamite is apparently intentionally parodic though, while Gayniggers is just really bad comedy.
04:38:30 <ehird> ...that I still need to watch some time.
04:39:11 <ehird> Although I'm fairly sure it cannot possibly live up to the lie of Gay Nigger Jim raping a toilet until he explodes and the song "Gay Niggers Eat Pigs and Fly on Penises Made Out of Ham and Brown Ham, Because They're Gay Niggers".
04:44:21 <ehird> http://cygwinports.blogspot.com/ ooh
04:45:19 <GregorR> For people who want Windows only for the binary compatibility.
04:46:08 <ehird> No, for people who like nice shells, unix-style command-line tools, etc and would like Unix-only programs, including X11 ones, to work seamlessly with their Windows ones.
04:46:54 <ehird> Cygwin is hugely (but not quite *cripplingly*, unless you have a large shell script and don't want it to take many minutes) flawed environment, but what it provides is valuable and nothing quite matches that.
04:47:20 <ehird> Windows has a very hostile command-line environment, so it's almost necessary.
04:48:03 <ehird> The only native binary of a shell is UnxUtil's zsh, which is ancient (last updated in 2003, or 2007 depending on whose dates you believe), and it isn't stable on Windows 7, at least (crashes a lot, some other weird bugs)
04:48:30 <ehird> Although GnuWin32 offers some GNU tools (but not a shell), that's still a limited set.
04:48:40 <ehird> So Cygwin Ports is a good effort.
04:49:48 <pikhq> I've got severel native POSIX binaries on my Windows VM.
04:49:48 <GregorR> So somehow you just haven't heard of MSYS?
04:49:56 <ehird> GregorR: I was hoping you'd say that!
04:50:02 <pikhq> Including /bin/sh, /bin/ksh, and /opt/gentoo/bin/bash
04:50:18 <pikhq> Granted, they're not Win32 binaries at all.
04:50:21 <ehird> GregorR: MSYS is an outdated fork of Cygwin, and has all the same flaws except it's fucking old. Furthermore, it has a fixed set of tools: if you want anything more, you have to have a patched version of gcc 2.
04:50:28 <ehird> *use a patched version
04:50:41 <ehird> GregorR: In conclusion, wow, no way MSYS is superior to Cygwin.
04:50:51 <ehird> Its only advantage is that it automatically translates Unix paths to Windows ones if you call a non-MSYS tool.
04:51:12 <ehird> pikhq: gcc 3, iirc. But I might try it sometime.
04:51:21 <ehird> Still, I don't particularly want to compile too much stuff myself.
04:51:23 <GregorR> ehird: A) I never claimed that it was, I merely said that there was another native binary shell, MSYS'. B) I strongly suspect that your statements about MSYS are more outdated than MSYS is.
04:51:26 <ehird> (Yes, I know gcc 4 is available)
04:51:33 <ehird> GregorR: MSYS is not a native binary shell.
04:51:36 <ehird> It uses a DLL, just like Cygwin.
04:51:44 <ehird> Since the DLL is a modified version of an old version of Cygwin.
04:51:51 <ehird> And absolutely false, I used it a day or two ago.
04:51:54 <pikhq> ehird: GCC 4 works on it, it's just not the default.
04:52:16 <ehird> You're not even meant to compile stuff for MSYS, which is why it's such a bitch. It is purely meant as a fixed development environment for using MinGW in.
04:52:19 <GregorR> Bah, I'm going to have to go boot up Windows just because I'm so convinced that you're full of shit.
04:52:20 <pikhq> Gentoo Prefix and Debian's Interix port install it, IIRC.
04:52:30 <ehird> GregorR: Which bit, exactly, do you think is false?
04:52:38 <ehird> pikhq: I don't want a Linux distribution.
04:53:38 <pikhq> ehird: So, you don't want a POSIX environment that coincidentally coexists with a Windows environment. Rather, you'd prefer Cygwin that doesn't suck.
04:53:51 <ehird> No, I do want the former, and it's the latter.
04:53:52 <pikhq> (or ideally an OS that doesn't suck, but we can't always have that)
04:54:03 <ehird> A Linux distribution isn't the best POSIX environment to coexist with Windows...
04:55:14 <pikhq> Interix offers a basic POSIX environment, and Gentoo xor Debian for it just install additional packages. Kinda like MacPorts.
04:55:33 <ehird> Yes, but for Debian they're *Debian* packages, patched and outdated.
04:55:42 <pikhq> Oh, right. Debian.
04:55:45 <ehird> For Gentoo I have to compile them myself.
04:56:00 <pikhq> In Gentoo you have to tell them to compile and be patient.
04:56:09 <pikhq> That's a bit less complicated than actually compiling them.
04:56:24 <ehird> Yes, but it's unneeded waiting, which isn't something I'm particularly good at being patient for.
04:56:28 <pikhq> You at least don't have to directly deal with some of the more painful build systems out there.
04:56:37 <ehird> Yes I do, when they break.
04:56:53 <ehird> (This is Windows + Microsoft stuff: not if.)
04:57:15 <ehird> [AT THIS POINT, GREGORR'S REALITY IS SHATTERED]
04:57:20 <GregorR> GCC 2.95, msys-1.0.dll, garbage in a large stinking pile.
04:57:59 <ehird> GregorR: With Cygwin, you get gcc 3 or a broken gcc 4. Booyah!
04:58:20 <GregorR> MSYS is just a shell and related gunk.
04:58:24 <ehird> GregorR: MinGW doesn't do POSIX.
04:58:37 <ehird> Windows does a little bit of POSIX, and MinGW lets you compile with that.
04:58:57 <ehird> You can't compile a *nix program for it; you can only port it. Using Win32 for the bits Windows doesn't do.
04:58:59 <GregorR> No, it doesn't. But here's a giant shock: That little bit of POSIX that Windows does is enough to appease 99% of the well-written software out there with small nudges.
04:58:59 <pikhq> That's older than my CPU architecture.
04:58:59 <pikhq> It might even be older than my knowledge of C.
04:59:02 <ehird> YAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAY
04:59:19 <ehird> GregorR: Which is why there are whole projects designed to port *nix things to Win32.
04:59:25 <ehird> Including trivial things like basic utilities.
04:59:40 <ehird> Because it Just Works! Yaaaay imaginative reality
04:59:46 <GregorR> The basic utilities are the MOST difficult to port things X_X
04:59:58 <ehird> Something like grep.
05:00:14 <ehird> grep is quite well-written, grep doesn't use crazy POSIX hacks as far as I know.
05:00:18 <ehird> Yet it's still a porting effort.
05:00:31 <ehird> Same for every damn thing.
05:00:35 <pikhq> Windows does decent chunks of POSIX.1. That gets you... File operations.
05:00:54 <ehird> But guize, I heard it passed all the POSIX tests!!!111
05:01:20 <ehird> GregorR: Also, it doesn't even have mmap() or fork().
05:01:29 <ehird> fork() is understandable, mmap() less so: breaks tons of shit.
05:01:48 <pikhq> Windows is only POSIX compliant if you use the POSIX subsystem.
05:02:00 <ehird> In conclusion: Cygwin is the only real option. Apart from *maybe* Interix.
05:02:08 <pikhq> Which is quite a bit different from what people are used to from Windows.
05:02:19 <GregorR> No, the only real option is to throw Windows in the dumpster.
05:02:25 -!- notostraca has quit (Read error: 54 (Connection reset by peer)).
05:02:26 <pikhq> (even the freaking filesystem magically changes behavior)
05:02:28 <GregorR> And use an OS that doesn't stink to all hell.
05:02:44 -!- notostraca has joined.
05:02:44 <ehird> GregorR: For non-command-liney stuff, I've found Windows superior to "modern" desktop Linux distributions so far.
05:03:00 <ehird> Its internals and command line environment are shit.
05:03:14 * GregorR steps out before the OS war silliness kills us all.
05:03:20 <ehird> But it has so much momentum behind it that the day-to-day desktop stuff is quite easy to make not suck.
05:03:27 <ehird> GregorR: Oh, feel free to argue. I love arguing!
05:03:34 <pikhq> Its GUI makes a 15 year old X environment look consistent.
05:03:53 <ehird> If you use bad software
05:04:03 <pikhq> And almost all of it is bad.
05:04:12 <Asztal> They do make it easy to write bad software :)
05:04:15 <ehird> Due to Sturgeon's Law. Windows is the biggest platform, therefore...
05:04:22 <pikhq> (even software that comes with it is rather bad)
05:04:29 <ehird> The only thing that doesn't blend in atm that I actually use is Chrome, because its toolbar widgets have a custom gradient on and the icons are custom.
05:04:33 <ehird> But it's so subtle as to be irrelevant.
05:04:41 -!- coppro has joined.
05:04:57 <ehird> pikhq: 7's Explorer is actually decent.
05:05:14 <pikhq> Oh, they made that not blatantly different from everything else?
05:06:09 -!- notostraca_ has joined.
05:06:09 -!- notostraca has quit (Read error: 104 (Connection reset by peer)).
05:06:11 <ehird> Most directories use the pleasing list-with-columns view with sane default sizes. There's a customisable shortcut toolbar to the left, and a field in which you can navigate up and down the directory structure at any point.
05:06:12 -!- notostraca_ has changed nick to notostraca.
05:06:17 <ehird> That also serves as a go-to-location field.
05:06:51 <ehird> It has a search field for the current directory (you can change where you want to search after hitting enter) next to that, and all the common option f a file are in a changing toolbar below these.
05:07:35 <ehird> Oh, and the file you're hovering over is highlighted, which is nice for picking a file out of a long list.
05:07:48 <ehird> Far better than XP's Explorer, which is worthless.
05:08:29 <pikhq> You neglect that Chrome also magically doesn't have a titlebar.
05:08:40 <pikhq> Unlike every other Windows program.
05:08:48 <ehird> On Vista and XP, it uses the native titlebar.
05:08:52 <ehird> It just nudges the tabs up into it.
05:09:03 <pikhq> http://upload.wikimedia.org/wikipedia/commons/c/c1/Google_Chrome.png
05:09:18 <ehird> Yes, and? That's what a titlebar looks like with Aero.
05:09:19 <pikhq> That, there, is inconsistency.
05:09:28 <ehird> No it's not, it's just different UI design.
05:09:36 <Asztal> Explorer doesn't have title text either; it doesn't need it
05:09:45 <ehird> Right below is the hierarchy field
05:09:46 <pikhq> Really? Every titlebar has no text in it, no icon, and "Google" right next to the minimize button?
05:10:02 <ehird> It also has no icon, its back/forwards buttons are in the toolbar, along with the location and search field.
05:10:26 <ehird> Chrome's only difference is the Google logo (which is unnoticeable), and the fact that the tabs are a little higher than what Explorer does.
05:10:28 <pikhq> ... My point is that doing that shit *sucks giant donkey balls*.
05:10:33 <Asztal> You can still double-click where the icon should be to close explorer, if you happen to be in a windows 3.1 frame of mind.
05:10:34 <ehird> It's perfectly usable.
05:10:53 <ehird> Is OS X "inconsistent" because its toolbar icons are in the "title bar" gradient?
05:11:06 <pikhq> It's doing things differently from all the other programs on the same OS for no good reason.
05:11:16 <ehird> Except that it isn't different from all other programs.
05:11:21 <ehird> This is standard Vista-style UI fare.
05:11:27 <pikhq> Really? No program has text in the title bar?
05:11:46 <pikhq> And control widgets are where the text would be in previous iterations of Windows?
05:11:48 <ehird> In pikhq's world, every program has the exact same UI and does the same thing because otherwise it'd be different from every other program on the same OS
05:12:02 <ehird> Also, that's not true. In Explorer they're lower down.
05:12:08 <ehird> the point is that every tab is the window
05:12:18 <ehird> The tab you're on is the title
05:12:20 <pikhq> In ehird's world, forming a strawman is a valid criticism.
05:12:25 <ehird> It's doing exactly the right thing for its UI
05:12:40 <ehird> That's Chrome's design, that's how it's meant to be: there is no browser UI, the tab is the title bar.
05:12:58 <ehird> Would you be happy if it was slightly higher up and didn't have a background? That's ridiculous.
05:13:15 * Sgeo prefers consistent UIs. I even tended to avoid using KDE/Qt programs in GNOME and GNOME/GTK+ programs in KDE
05:13:45 <pikhq> I'd be happy if it were consistent with other programs' look and feel, rather than creating a custom title bar.
05:14:10 <ehird> It isn't custom if other fucking programs do it
05:14:15 <ehird> It's not some weird Chrome thing
05:14:25 <Sgeo> Might not be "custom", but it's annoying
05:14:36 <ehird> ...no it's not. Have you ever USED Chrome on Vista/7?
05:14:39 <pikhq> It's some weird "some random set of programs do it".
05:14:40 <ehird> It's completely unnoticeable and perfectly designed.
05:14:53 <ehird> pikhq: INCLUDING THE FLAGSHIP ORGANIZER SHIPPED WITH THE OS
05:15:04 <pikhq> ehird: MY POINT IS THAT EVEN THE PROGRAMS THAT COME WITH IT SUCK.
05:15:05 <ehird> That also does the system settings
05:15:14 <Sgeo> Do most programs do it Chrome's/Explorer's way?
05:15:15 <pikhq> BECAUSE THEY RANDOMLY DO THINGS DIFFERENTLY FOR NO GOOD REASON.
05:15:22 <ehird> Meh, no point discussing this; you've used XP and think you nkow what's consistent or good or not in 7
05:15:30 <ehird> So, carry on thinking that
05:15:32 <Sgeo> pikhq, if you talk about XP, you definately have a point
05:15:49 <Sgeo> I don't know what the typical UI is on Vista or 7
05:16:01 <pikhq> Also, newer programs keep the Aero UI in XP, with some widgets changed.
05:16:02 <ehird> No, silly! He's talking about an OS he knows from a few screenshots and anecdotes. In other words, he's an expert.
05:16:13 <ehird> pikhq: ...what? No. No they don't.
05:16:16 <ehird> That is not true whatsoever.
05:16:18 <ehird> You are simply wrong.
05:16:36 -!- Gracenotes has quit ("dalnet").
05:17:04 <Sgeo> The way Chrome looks in XP is inconsistent with most XP programs
05:17:14 -!- Gracenotes has joined.
05:17:21 <ehird> pikhq is too, except he doesn't know shit about 7
05:17:48 <pikhq> So, apparently in 7 the title bar is meaningless.
05:18:07 <ehird> Genius deductions based on anecdote and one screenshot, pal.
05:18:32 <Asztal> The title is unnecessary in explorer because the breadcrumb bar tells you everything you need to know
05:18:35 <ehird> For example, it's not as if the Explorer title bar is like that BECAUSE THE TITLE OF THE FOLDER YOU'RE IN is in the title bar.
05:18:47 <ehird> Nope, it should be redundant and keep that below the fold while ALSO putting it in the window title.
05:18:52 <ehird> Yep, that makes sense.
05:19:02 <ehird> Alternate universe: "So, apparently in 7 the title bar is redundant."
05:19:17 <pikhq> More sense then only some programs having any information in the title bar at all.
05:19:27 <ehird> EXPLORER HAS THE INFORMATION IN THE TITLE BAR.
05:19:38 <ehird> "[icon] > Computer > Local Disk (C:) >"
05:19:41 <ehird> That is the folder I am in.
05:19:44 <ehird> That is what would be in the title bar.
05:19:46 <ehird> IT IS IN THE TITLE BAR.
05:19:53 <pikhq> Looks to be below the title bar to me.
05:20:04 <Asztal> It's in the glassy area, so I suppose that counts
05:20:06 <ehird> It's in the transparent area above the window contents, i.e. THE TITLE BAR.
05:20:20 <ehird> You can move the window by it.
05:20:24 <ehird> i.e. THE TITLE BAR
05:20:41 <ehird> (Drag the glass area below the field. Voila, moving. That's what the title bar does.)
05:21:05 <pikhq> ... So UI elements are being stuck in the title bar?
05:22:05 <ehird> pikhq: You mean like the program icon, title text, minimize, maximize and close?
05:22:16 <ehird> Sound like UI elements to me.
05:22:35 <pikhq> Erm. That was poorly phrased.
05:22:47 <ehird> You can also click the program icon for a menu.
05:23:00 <pikhq> Per-program UI elements that are inherently different for each program.
05:23:14 <ehird> It's called discretion.
05:23:18 <ehird> You complained that the title bar was bad because it has no information.
05:23:25 <ehird> In Explorer, the relevant information IS in the title bar.
05:23:41 <ehird> It's also navigable, which is required of such a widget anyway: anything else would be redundance.
05:23:50 <pikhq> I continue to see a null title bar and an inexplicably glassy portion of the program.
05:23:51 <ehird> In Chrome, the title of the page IS in the title bar, and highlighted: it's the current tab.
05:23:55 <ehird> All of these usages are consistent.
05:24:03 <ehird> They contain the correct information in the correct way.
05:24:23 <ehird> pikhq: You can right click the glassy region below the "widgets" to get the window menu, too.
05:24:28 <ehird> It is undeniably the title bar.
05:24:46 <ehird> You wouldn't see it that way after using 7 for some time, unless you're really UI-blind.
05:25:13 <ehird> In fact Explorer would be better if it lifted those elements to the top of the title bar, like Chrome.
05:25:27 <ehird> Right now the title bar has some weird padding at the top.
05:25:32 <pikhq> It's because I continue to see things like this: http://en.wikipedia.org/wiki/File:Notepad_Vista.png
05:25:49 <ehird> The file location is in the title bar.
05:26:02 <ehird> Then there's the window menu, and then its contents.
05:26:05 <pikhq> Which has quite different titlebar behavior from Explorer or Chrome.
05:26:13 -!- coppro has quit (Read error: 104 (Connection reset by peer)).
05:27:20 <pikhq> Yes, it's quite different. The titlebar is a different size and it shows quite different information.
05:27:23 -!- coppro has joined.
05:27:56 <ehird> I don't think I will ever change your mind because you have preconceived notions of what is correct and are only engaging in this to convince me, not to bother to be convinced otherwise. I deduce this from your style of debate and persistence.
05:28:04 <ehird> Feel free to prove me wrong.
05:28:23 <pikhq> The Explorer titlebar doesn't even show the title. Just a directory hierarchy.
05:28:38 <ehird> That is what would be shown in the title bar.
05:28:41 <ehird> The directory hierarchy.
05:28:54 <ehird> That is what Explorer's title bar has ALWAYS shown.
05:30:03 <Sgeo> I can imagine that showing "Explorer" in the task bar would be more useful then showing the directory hierarchy
05:30:11 <ehird> I hope you're being sarcastic.
05:30:26 <Asztal> In the task bar, you just get the folder name
05:30:39 <ehird> The 7 titlebar groups windows into applications (like the OS X Dock but better) and the Window title for other programs is what it used to be.
05:30:43 <Sgeo> No, I'm not... although I just considered the possibility of multiple Explorer windows, so nevermind
05:30:47 <ehird> Plus, Explorer windows, well, look like Explorer Windows.
05:30:51 <ehird> So it's not very useful.
05:31:20 * pikhq just wants every damned program to look the same.
05:31:40 <ehird> That is a ridiculously ill-defined notion.
05:31:58 * Sgeo should probably be eating or something
05:31:59 <ehird> 7 has a consistent UI look; you just think it doesn't based on preconceived misconceptions, and I will not convince you otherwise, at least not by talking on IRC.
05:32:13 <Sgeo> It depends on how you define "consistent UI look"
05:32:21 <ehird> That is what I said.
05:32:42 <ehird> It is a ridiculously ill-defined notion, but for any definition that isn't close to "every program should have identical graphics and therefore do exactly the same thing" 7 has it.
05:33:00 <pikhq> In Aero, it looks to me almost like doing Qt and GTK on the same system, with them having themes that emulate each other. You get Notepad, which looks like an XP app with a shiny theme, and Explorer, which looks like it's for a *different OS* and happens to have the same widgets.
05:33:06 <ehird> Besides, Explorer's title bar is exceedingly usable, so fuck any incorrect ideological notions.
05:33:22 <ehird> pikhq: It does not. You are using preconceived blah blah blah I don't give a shit, you're just repeating the same crap..
05:33:32 * Sgeo feels dizzy in his attempt to minorly disagree with ehird
05:33:40 <pikhq> ehird: Imagine if Windows Explorer did the same thing on OS X.
05:33:43 <ehird> Sgeo: I attempt to make that as difficult as possible!
05:33:51 <pikhq> Erm. That was a bad sentence.
05:33:55 <ehird> pikhq: Imagine if a Windows program was on OS X! Wouldn't it be DIFFERENT?
05:34:02 <ehird> No shit, sherlock.
05:34:12 <ehird> BTW, the title bar area in OS X also includes the WHOLE toolbar.
05:34:16 <pikhq> Okay, now you're just being a asshat shitcock fuckoff.
05:34:20 <ehird> Not just the bit relevant to the thing at hand, like the path.
05:34:29 <Sgeo> Seriously, it's as though I have this notion in my mind that everything ehird says is correct
05:34:58 <ehird> Cool, so while I try to end the argument by mildly cursing and saying I don't care anymore, pikhq calls me an asshat shitcock, tells me to fuck off, and sticks his fingers in his ear.
05:35:01 <ehird> Sgeo: That's totally true.
05:35:28 <ehird> Sgeo: If I said that's false, would your brain overheat and would steam come out of your ears?
05:35:58 <Sgeo> No, because I'd get the joke. However, this response to that joke just killed the joke.
05:36:19 * Sgeo turns himself in for murdering a joke.
05:37:04 <ehird> Sgeo: Almost certainly not everything I say is correct. This is actually true and serious, no joke.
05:37:38 <Sgeo> I know, but I think I'd have difficulty recognizing when you're wrong.
05:38:07 <ehird> That's pretty much identical to saying "If I have an opinion about something, and you have one and it differs I'll disagree" which is... fairly obvious...
05:38:33 <Sgeo> ehird, are you intentionally saying incorrect things now to see how I'd react?
05:38:40 <Sgeo> It's not the same at all
05:38:57 <ehird> Sgeo: Yes it is, the only way to determine whether something is true or false that you can agree with is to use your opinions
05:39:27 <Sgeo> I'm saying that if you state an opinion, my mind is likely to adapt that opinion before looking at it critically
05:39:40 <ehird> I will use this for nefarious purposes.
05:44:10 * ehird weighs up windows: shit desktop and linux: shit command line...
05:44:22 <ehird> I really would quite like a Cygwin GHC for a Cygwin darcs...
05:44:26 <ehird> but that'd be a huge porting effort
05:45:00 -!- immibis has joined.
05:46:00 * Sgeo wonders if it would be possible/sensible to make a C library of some of the Python standard library
05:46:56 <immibis> well yes, it depends which part
05:46:58 -!- notostraca has quit (Read error: 104 (Connection reset by peer)).
05:47:31 -!- notostraca has joined.
05:47:40 <ehird> python's so different to C that it'd be unnatural
05:48:39 <immibis> i did say it depends which part. Integer manipulation for example already exists in C.
05:49:24 * Sgeo is thinking more like urllib2. Isn't integer manipulation more a part of the language?
05:50:05 <ehird> urllib2: use libcurl
05:50:44 <Sgeo> Does CPython use libcurl?
05:50:49 <ehird> Sgeo: look at http://curl.haxx.se/libcurl/c/simple.html
05:51:11 <ehird> http://curl.haxx.se/libcurl/c/sepheaders.html more fleshed out example
05:51:19 <ehird> (downloads a page and its headers into two files)
05:52:01 * Sgeo looks at curlpp
05:52:39 <ehird> Sgeo: why are you using C++?
05:52:49 <Sgeo> ehird, because it's what I'm learning in college right now :/
05:53:00 <ehird> Well that's a good reason :P
05:53:09 <ehird> http://curlpp.org/index.php/examples/47-example-00
05:53:21 <ehird> Looks like a direct binding of libcurl (except uglier)
05:53:55 <Sgeo> Why do you need to make a curlpp::Cleanup ?
05:54:07 <Sgeo> I think I get how that works
05:54:13 <Sgeo> But it still looks nonsensical
05:54:39 <ehird> So when the Cleanup is destroyed, it deallocates and stuff.
05:54:57 <ehird> It's not in the request because you could do multiple requests.
05:56:05 <Sgeo> Well, good night all
05:56:10 -!- Sgeo has quit ("Leaving").
06:06:00 <ehird> I just realised why SPUI was called that
06:06:02 <ehird> http://upload.wikimedia.org/wikipedia/commons/thumb/b/b1/Spui-schematic.svg/792px-Spui-schematic.svg.png
06:10:19 <ehird> hint: SPUI had a bunch of joking/trolly references to anal sex on his page
06:11:25 <ehird> Or, it looks like goatse.
06:12:36 <ehird> (His page made reference to "single point urban interchange" and he was mostly known for editing transport (including motorway) articles, so it's almost certainly the same thing, even if he didn't see the goatse, which I doubt.)
06:12:41 <ehird> Talk about long-term trolling.
06:12:48 <ehird> Well, he was mostly known for trolling, but.
06:14:03 <notostraca> so i am coming up with ideas for a language on parrot
06:14:16 <ehird> Parrot's just a VM.
06:14:20 <notostraca> so what that could do in this hypothetical language is create the left->right junction of the results of (+ 5 "string") and (str 5 "string")
06:14:40 <notostraca> but it is easier to implement things on there than on, say, C+yacc
06:15:15 <notostraca> junctions are from perl 6, and i can steal from their implementation
06:15:42 <notostraca> the prefix notation is from lisp, but i don't think i will just use () parentheses
06:15:43 <ehird> Still, being on Parrot isn't a feature of the language.
06:16:01 <notostraca> i just meant that was how i was going to do it
06:17:18 <notostraca> it would be fairly amazing if someone wrote a parrot bytecode -> java bytecode converter
06:18:21 <ehird> the vm has a startup time so slow it's worthless for command-line tools, the standard library is awkward and verbose to use
06:18:37 <ehird> there aren't actually that many additional libraries for it as the jvmists claim
06:18:48 <ehird> and its primary GUI library, Swing, is unspeakably bad
06:19:23 <notostraca> https://substance.dev.java.net/see.html
06:20:09 <ehird> Seeing those skins makes me bet that it can't be made to look native.
06:20:21 <ehird> Oh, look, I don't have the JVM.
06:20:33 <ehird> (I didn't notice because nothing actually uses Java.)
06:20:39 <ehird> (Nothing desktop, anyway.)
06:20:56 <ehird> notostraca: all vm people think theirs is ubiquitous and mandatory :-)
06:21:10 <ehird> the only thing with that title on a typical internet-connected desktop is flash
06:21:17 <notostraca> well i use mac, and all macs come with... python, perl, ruby, and java 5+
06:21:25 <ehird> but the fact is that no command-line program uses java, no desktop program uses java
06:21:37 <ehird> notostraca: i'm actually using Windows 7 right now, but I'm a *nixer
06:21:52 <ehird> ...although all OSs right now are woefully inadequate and wasteful
06:21:54 <notostraca> ehird, your sweepping generalizations make me think less of your opinion
06:22:09 <ehird> that'd matter if I was looking to sway your opinion
06:22:29 <ehird> I'm merely stating my experience; no popular program on Windows uses Java apart from OpenOffice, and that only uses it for some things behind the scenes
06:22:37 <ehird> And I find the same applies to Linux and OS X
06:25:26 <ehird> LimeWire, granted, but LimeWire sucks big time.
06:25:40 <ehird> Eclipse is only popular with Java developers. (Don't say that C thing. Nobody uses it.)
06:26:03 <notostraca> Eclipse is slow, but it has helped me a lot
06:26:14 <ehird> that doesn't mean it's popular outside of a niche community
06:26:29 <ehird> also, it's really overly bloated. Does seven billion things, does none of them well...
06:26:47 <notostraca> http://en.wikipedia.org/wiki/Criticism_of_Java
06:27:09 <notostraca> mac does even less things than eithe rof those
06:27:44 <ehird> emacs is also terrible
06:27:54 <ehird> there are a surprising number of sharp, compact windows tools
06:28:06 <ehird> linux with the typical gnucrap doesn't do it, indeed
06:28:28 <ehird> notostraca: right now? i edit small files with notepad2 or programmer's notepad, because I haven't got anything set up for that stuff.
06:28:38 <ehird> on os x i mostly used textmate because all the other os x editors suck
06:28:52 <ehird> but for haskell and lisp I used emacs
06:28:59 <ehird> notostraca: depends how you define large
06:29:21 <ehird> if you define it high enough, that project is too big to exist (or is it too big to fail? let's bail out Eclipse)
06:29:38 <ehird> but I'm absolutely fine as long as I have a file switcher tree of the project to the left
06:29:54 <notostraca> over 1000 LoC in C/++? not small enough to be written and maintained effortlessly
06:29:55 <ehird> well, a window switcher list might be better than tabs if I have an awful lot of files open
06:30:02 <ehird> but I find I rarely actually actively edit more than a few files
06:30:14 <ehird> notostraca: 1000? that's a ridiculously low bar, any non-trivial C program will be above that
06:30:29 <ehird> notostraca: xmonad is like 500 lines...
06:30:33 <notostraca> yeah, i was just going for "not a script"
06:30:36 <ehird> and it's a fully-featured window manager that a ton of people use daily
06:31:38 -!- notostraca_ has joined.
06:31:38 -!- notostraca has quit (Read error: 104 (Connection reset by peer)).
06:31:41 -!- notostraca_ has changed nick to notostraca.
06:34:11 <notostraca> ehird: you remind me of myself when I'm not on my meds
06:34:44 <notostraca> i also write a lot more than when i am
06:41:12 -!- FireFly has joined.
06:42:00 -!- immibis has quit ("On the other hand, you have different fingers.").
06:43:03 <ehird> notostraca: I'm not argumentative, I'm just not adding obvious qualifiers like "I think"
06:43:12 <ehird> although this is commonly misinterpreted as argumentativeness
06:43:40 <notostraca> also i make sweeping generalizations when i don't take my meds
06:43:59 <notostraca> though i am less articulate than you are right now
06:44:06 <ehird> i will make a note to avoid meds :P
06:44:28 <notostraca> i will make a note to keep taking meds
06:44:39 <notostraca> it helps me make friends and not go to jail
06:44:56 <ehird> i have plenty friends, thank you, and I've never committed any harmful crime
06:45:48 <notostraca> you probably don't have Ragin' Cajun Syndrome either
06:46:45 <notostraca> although many more coders than non-coders have aspergers, myself included
06:47:09 <ehird> I hope you aren't another of the self-diagnosed aspergers as an excuse for social incompetency camp...
06:47:12 <ehird> there are way, way too many of those
06:47:14 <notostraca> asperger's syndrome: great syndrome, or the greatest syndrome?
06:47:53 <notostraca> about 4 doctors would testify in court
06:48:10 <ehird> great, now my only qualm is the incredible vagueness of its definition :-P
06:48:30 <notostraca> most of them shouldn't have been given birth certificates, much less doctorates...
06:48:54 <ehird> (as someone who has been tentatively-diagnosed (but not diagnosed) with aspergers, after researching I find it a shaky and dubious vagueness...)
06:49:35 <notostraca> well it is pretty useful when you just want to do one thing
06:49:57 <ehird> I meant I find it dubious as a condition
06:50:01 <notostraca> people with asperger's tend to be abnormally good at focusing on a specific task that they are born to do
06:50:18 <ehird> you're not born to do anything
06:50:57 <notostraca> i typed the quotes with fingers in the air
06:51:10 <ehird> an impressive feat
06:52:20 <notostraca> but i have a knack for game development, which has less to do with the asperger's giving me skill and more to do with the asperger's making me especially interested in the subject
06:53:19 <notostraca> the problem is, i get too interested in the designs and i start switching around to the game that interests me most at the given time
06:53:26 <ehird> I'm aware of the things that are claimed Asperger's brings.
06:53:42 <ehird> (That sentence is truly awkward.)
06:54:07 <ehird> At least mine was valid in a non-poetic sense.
06:54:35 <ehird> how on earth is that argumentative?
06:54:57 <notostraca> you seem to be trying to prove your statements are better
06:55:27 <ehird> Wow, play it again, Freud. I was joking that my sentence was a funner example of an awkward sentence since an invalid sentence outside of a poetic context is just that, not awkward.
06:55:49 <notostraca> sweet, we have aspergers! we can't identify jokes
06:56:00 <ehird> I knew yours was a joke, actually.
06:56:24 <ehird> Obviously we weren't seriously debating the relative merits of awkward sentences...
06:56:33 <ehird> Obviously! Issue: I'm British.
06:57:18 <ehird> you forgot cajuning.
06:57:45 <notostraca> oh, are you a quack doctor like me now? :-P
06:58:00 <notostraca> having seen enough of them, i should know their tricks
06:58:14 <ehird> i'm an inverse homeopath
06:58:49 <ehird> I start with water, and continually dilute it in the thing that causes the ailment until there's less than one part in a trillion water
06:58:53 <ehird> then I give it to the patient
06:59:16 <ehird> the state considers me a homicidal murderer, but I prefer inverse homeopath
06:59:23 <notostraca> some kid went to my school who gave up gluten to see if it would stop his high-functioning autism
06:59:47 <ehird> What does that even mean? "I want to become a social idiot?"
07:00:40 <ehird> They should make meat with added gluten.
07:00:46 <ehird> It'd be, like, the best thing ever. To eat.
07:00:58 <ehird> Quorn doesn't have any meat, dude.
07:01:23 <ehird> It's not dead flesh
07:02:06 <ehird> Meat is meat, egg is just an animal product.
07:02:33 <ehird> It's not meat unless it's the same flesh as would be if it was alive and then died.
07:02:44 <notostraca> "i'm going to fry up some pre-chickens, you want any?"
07:02:52 <ehird> Lab-created never-living flesh counts, dead animal flesh counts
07:03:04 <ehird> And it is DELICIOUS (well, just the latter at the moment)
07:03:45 <ehird> It's not flesh, it's stuff that used to be flesh
07:04:19 <ehird> YOU KNOW WHAT I MEAN BY MEAT DAMNIT
07:04:59 <notostraca> i am designing a language, i have to be precise here and count for all interpretations :-P
07:05:09 <ehird> your programming language includes a notion of meat?
07:05:46 <notostraca> i am in programing-language-design mode, is what i mean
07:06:30 <ehird> notostraca retends anuses
07:06:58 -!- kar8nga has joined.
07:07:02 <notostraca> now i want this language to have 3 big features:
07:07:20 <ehird> Careful, you've set me into ruthlessly-criticise-language mode.
07:07:22 <ehird> I MAY BE FLAMMABLE.
07:07:48 <notostraca> thankfully, you like lisp and haskell well enough
07:07:51 <ehird> I don't know what a junction is, but damnit, I cannot possibly criticize the other two
07:08:17 <notostraca> junctions allow a datum to be two or more things at once
07:08:49 <notostraca> http://en.wikipedia.org/wiki/Perl_6#Junctions
07:09:11 <notostraca> that explains it better than i can in short messages
07:09:20 <ehird> Oh, superpositions.
07:09:42 <ehird> They're nice. When I started programming I was always baffled I couldn't do (foo == (1 | 2 | 3)).
07:09:49 <notostraca> the addition is here is ordered unctions
07:10:49 <notostraca> so if you have the ordered junction of two functions, you can iterate through the junction returned until you find a result that isn't an error
07:11:17 <ehird> Does it have a ca, so you can have a cajun
07:13:04 <notostraca> Mongolia Travel - www.DiscoverMongolia.mn - Tours, Hotels, Flights and Trains. Get Affordable Rates. Book Today!
07:13:38 <ehird> "If you thought that Mongolia was just the land of Genghis Khan, think again."
07:13:42 <ehird> And I was just booking my ticket
07:14:15 <notostraca> i was thinking of calling this language Khan back when it was more like haskell in syntax
07:14:40 <notostraca> because genghis khan used arrows extensively, as would this language
07:16:26 <notostraca> http://img27.imageshack.us/img27/3396/khano.png i think
07:19:38 <notostraca> ehird: yeah, it needs little pictures of scooby-doo and the various things that aren't the guy wearing a mask
07:20:10 <notostraca> if you don't have hanna-barbara cartoons in britain, you aren't missing out
07:20:21 <ehird> I was just trying to figure out the relevance :P
07:20:36 <ehird> worst reference ever :P
07:20:47 <ehird> like those puzzle things on easter eggs and shit
07:21:04 <notostraca> to make it so no one would use the language
07:22:02 <notostraca> the omnipotent benevolent personified deity, the great Satan?
07:22:55 <notostraca> if you look at christianity today, it seems like satan gets credited with more than the benevolent side of god
07:23:20 <ehird> you mean... christians are irrational?
07:23:22 <ehird> never would I ever have guessed.
07:23:32 <ehird> also, it's not "today"; god kills far more people than satan in the bible
07:23:36 <ehird> also, he's kind of a dick.
07:23:51 <coppro> he was talking about god
07:24:35 <coppro> ehird: if god gets capitalized pronouns, there's got to be something wrong with spelling his name uncapitalized
07:24:45 <FireFly> Nope, wrong religion, not angkor wat
07:25:05 <notostraca> coppro: are you a professional cop or a reference to coprophilia?
07:25:10 <ehird> notostraca: neither
07:25:18 <ehird> but we joke about the latter!
07:25:29 <ehird> coppro: you suck. eat shit!
07:25:33 <coppro> and then I deride your ability to read
07:25:40 <ehird> also, I actually thought *nix was like g*d when I was young, before ever using any *nix
07:25:51 <ehird> like programmers exalted unix so much that they refused to spell the u
07:25:56 <ehird> just like jews and shit with g*d
07:26:06 <notostraca> if only programmers could agree on something
07:26:24 <ehird> *nix doesn't even make any sense, it's not Linix
07:26:31 <ehird> it wasn't even Freanix
07:26:42 <ehird> coppro: so BSD is out then?
07:26:51 <ehird> FireFly: that second ? is odd
07:27:01 <notostraca> yeah, what happened with judaism not allowing YHWH to be spoken and the Jehovah's Witnesses making the romanization of YHWH into the name of their religion?
07:27:22 <coppro> lol Jehovah's Witnesses
07:27:35 <ehird> more like jehovah's poopnesses
07:27:48 <notostraca> a jehovah's witness knocks on a rabbi's door
07:27:53 <ehird> coppro: also, what about ssh? I think you mean *shh :P
07:27:57 <notostraca> rabbi covers his ears, goes "lalalalalala"
07:28:41 <notostraca> i have been very nice to jehovah's witnesses when they come to the door
07:28:46 <coppro> here's a really bad joke:
07:28:55 <coppro> An Irishman walks out of the door.
07:29:19 <ehird> i've never had the opportunity of conversing with a jehovah's witness
07:29:29 <ehird> would be fun to have them go away deeply disturbed at this satanic 14-year-old
07:29:40 <ehird> hmm i guess by 14 they can chalk it off to teenage rebellion
07:29:49 <ehird> notostraca: yeah they seem very boring
07:30:10 <fizzie> .*n.?x would match qnx, although it's debatable whether that's a good thing.
07:30:41 <fizzie> It's a real-timey sort of an OS.
07:30:54 <fizzie> "aimed primarily at the embedded systems market".
07:33:24 <FireFly> Well, I got to go in another ~20 mins
07:33:31 <FireFly> And that's not a binary not
07:33:57 <coppro> Firefly: you're 20 minutes late. Go now!
07:34:05 <FireFly> Actually, I start in 50 minutes
07:34:48 <fizzie> It's the unary binary not.
07:35:09 <fizzie> Yes, I just said it's "not".
07:35:25 <FireFly> And there the lack of a comma after "No" has an important meaning
07:35:50 <fizzie> If not that not, then it must be the Awk regular-expression matching operator, but it doesn't really seem to make sense there.
07:36:12 <coppro> oh boy, #nethack has gotten into a MMO debate
07:36:23 <notostraca> in some languages isn't ~ the marker for negative numbers?
07:36:46 <coppro> it's also, of course, used for matching
07:36:55 <fizzie> I think so too, but can't name one right now.
07:37:15 <fizzie> It's of course also the INTERCAL "select" operator.
07:37:50 <coppro> In my chicken scratchings, ~ usually ends up as being used for matching
07:37:58 <FireFly> Note to self: Always actually spell out "approx." when in #esoteric
07:38:12 <FireFly> To avoid discussions about the choice of symbol
07:38:17 <ehird> Personally I think this pedanticness is annoying.
07:39:25 <coppro> ~~ being used as the "matches" operator, which may be cheaper, for instance because an RE engine doesn't need to capture
07:39:57 <coppro> or the other way round, if I decide that double-symbols mean bitwise and single mean logical
07:41:01 <coppro> why is video game music so awesome?
07:41:17 -!- Asztal has quit (Read error: 110 (Connection timed out)).
07:41:29 <ehird> You asked that before.
07:41:33 <ehird> The answer is that it isn't for the most part.
07:43:19 <ehird> I ought to get audio working here sometime.
07:43:27 <ehird> Then again Cygwin's crappiness is putting me off this.
07:43:52 -!- kar8nga has quit (Remote closed the connection).
07:45:12 <ehird> No answers, though.
07:46:32 <coppro> video game music isn't necessarily the best out there, but it beats the hell out of the radio
07:47:01 <ehird> So does everything.
07:50:46 -!- coppro has quit (Remote closed the connection).
07:51:39 -!- coppro has joined.
07:53:44 * ehird considers getting Arch Linux working
07:53:56 <ehird> it failed pretty badly the last time...
07:56:48 <ehird> And I'm not even sure if Chromium is usable on Linux yet... nor do I know if there's a better graphical IRC client than X-Chat... graphical IRC on linux tends to suck ime
07:58:40 <ehird> At least the font rendering beats ClearType... although not the fonts.
07:59:49 -!- FireFly has quit ("Later").
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:11:12 -!- oklofok has joined.
08:13:43 <ehird> ^bool Should I try to install Arch Linux?
08:13:57 <ehird> Should I try to install Arch Linux over the current Windows 7?
08:14:16 <ehird> Are you sure that I shouldn't try to install Arch Linux over the current Windows 7? Last chance.
08:14:25 <ehird> Then should I try to install Arch Linux over the current Windows 7?
08:14:34 <ehird> Are you sure that I shouldn't try to install Arch Linux over the current Windows 7? Last chance... again.
08:16:12 <fizzie> The bot's quite a waffler.
08:18:31 <ehird> Do you both think I should try to install Arch Linux over the current Windows 7 and are sure?
08:19:17 <ehird> oklofok: gliosperous
08:23:05 <ehird> glio becoming gliosperous in the gliovorous fog
08:23:28 <oklofok> well who wouldn't become sperous in a fog that's out to get you
08:23:48 <oklofok> AnMaster: do my swedish homework :|
08:30:46 <ehird> but you know everything
08:30:49 <oklofok> do my swedish homework and maybe i will
08:31:57 <oklofok> although transcribing the a4 article to you in priv might be slightly more work than just reading it
08:34:15 <oklofok> so anyone read complexity theory companion? it looks rather promising
08:36:06 -!- notostraca has quit ("Be more esoteric!").
08:38:19 <oklofok> complexity theory is very esoteric, actually the whole preface of the book is about how esoteric it is
08:42:27 <ehird> you should teach courses on random shit
08:42:44 <oklofok> "most people view complexity theory as an arcane realm populated by pointy-hatted (if not indeed pointy-headed) sorcerers stirring cauldrons of recursion theory with wands of combinatorics, while chanting incantations involving complexity classes whose very names contain hundreds of characters and sear the tongues of mere mortals"
08:42:58 <ehird> like, the deal is, someone gives you some money and then they can go wherever you are and you teach them about a random esoteric topic that you haven't before, unplanned
08:43:08 <ehird> for like once a week for three months
08:43:20 <ehird> like, 10 people doing it
08:43:44 <oklofok> it'd be great, but i'm not a very good teacher :)
08:43:55 <ehird> and at the end you'd have a grasp of theundecidable asymptomatic complexity theory when applied to faux-regular metalanguages
08:43:58 <oklofok> i tend to be slightly too formal
08:44:13 <ehird> *also, only one space before the
08:44:26 <ehird> and how to bake cauliflower using only two sticks
08:44:52 <ehird> and how to use a fire to power a computer, explained with a shitload of esoteric physics
08:45:04 <ehird> and the cyclic syntactical structures of fractal systems
08:46:30 <oklofok> well, those are my areas of expertise.
08:47:08 <ehird> also, everything about graphs ever
08:49:52 <oklofok> i've read half of the material of our graph theory course (for funsies), i knew about 3/50 of the theorems beforehand
08:50:49 <oklofok> well, okay, more, but stll i wouldn't say i know *any* graph theory
08:51:00 <oklofok> i basically just know what graphs are
08:51:16 <ehird> well you'd learn it
08:51:28 <ehird> like, sitting there with a textbook
08:52:11 <oklofok> i wish they lectured the graph theory course this year too, was too busy to take it last year
08:52:32 <oklofok> yes, usually they wait a few years before repeating
08:52:45 <ehird> no i mean business.
08:53:04 <oklofok> because most students use at least 3 years for their master's
08:53:27 <oklofok> well i was pretty busy, not as busy as i am now, but anyway
08:55:37 <oklofok> there were actually a few courses i didn't take just because i thought i had enough.
08:56:05 <oklofok> which naturally bugs me to death, have to fight the urge to do the homework for those courses just because
08:57:21 <fizzie> We had a distributed computing (the theoretical aspect of it, no engineeringy things) seminar course about a book; the professor hadn't read it, and said that he just arranged the seminar course to have the students present the contents to help motivate him to read it, as well as to find the interesting parts.
08:57:31 <oklofok> which means it's game theory time
08:58:45 <ehird> game theory? more like lame theory
08:58:54 -!- fizzie has quit (hubbard.freenode.net irc.freenode.net).
08:58:54 -!- Ilari has quit (hubbard.freenode.net irc.freenode.net).
08:59:03 -!- Ilari has joined.
09:01:31 <oklofok> the lecturer sucks though, much of what he says is incomprehensible because of his weird accent, and the homework usually lacks some of the crucial definitions, so you have to guess the question first
09:02:15 <oklofok> err, i was being on your side there
09:02:24 <ehird> for i am most tired
09:02:34 <oklofok> i finally learned to suppress a yawn
09:02:45 <oklofok> actually used the last lecture to train it
09:02:48 <ehird> i can do that, but it makes me feel like crap
09:03:05 <oklofok> but can you forget you were about to yawn
09:03:28 <oklofok> that's what i meant by suppressing, obviously anyone can just *not yawn*
09:04:04 <ehird> but can your mother
09:04:20 <oklofok> i doubt she's interested in random body control exercises.
09:04:32 <ehird> unless they involve HER GENITALIA
09:07:25 <oklofok> it'd be so cool to father you a brother
09:09:07 <oklofok> i guess i haven't understood the concept of mum jokes yet
09:09:58 <oklofok> think about it, i'd be like your stepfather, you could sit in my lap and hear stories about complexity theory
09:19:15 <oklofok> haha, my touchpad went crazy, the pointer keeps moveing randomly in circles when i touch the pad
09:19:56 <oklofok> which means buttons are almost impossible to press, because you have to press at the exact right moment
09:21:06 <ehird> nub ftw (I'm not having another clit joke incident)
09:21:15 -!- puzzlet has quit (Read error: 104 (Connection reset by peer)).
09:21:19 <ehird> disable the trackpad
09:21:28 <oklofok> i should probably restart the computer
09:21:28 <ehird> you can turn on using the numpad to mouserise
09:21:55 <oklofok> i don't have a numpad, and i know that, and i don't know where that is on vista, and i don't like using it
09:22:05 <ehird> you do, it's just fn-some letters
09:22:12 <ehird> most laptops have a numpad overlaid on fn
09:22:15 <oklofok> well sure, but that's even more annoying.
09:22:25 <ehird> but a fun challenge
09:22:38 -!- puzzlet has joined.
09:22:42 <oklofok> as if i like fun challenges!
09:22:52 <ehird> it's control panel -> ease of access -> change mouse works blah -> turn on mouse keys on 7 at least
09:22:56 <ehird> which is quite similar to vista
09:23:08 <ehird> also you can make it focus a window when you hover over it.
09:40:53 -!- oklopol has joined.
09:41:20 <oklopol> probably should've read what you said, i'm not exactly that good at navigating without the mouse
09:41:29 <oklopol> so searching for the mouse thing was rather hard
09:41:36 <oklopol> it was under "make the mouse easier to use"
09:41:59 <ehird> "change how your mouse works" for me
09:42:22 <oklopol> i looked at the mouse options
09:42:40 <oklopol> for some reason i thought that's where mouse options would be.
09:42:57 <ehird> turn on the ctrl/shift to ac/decelerate
09:43:04 <ehird> it's painfully slow to get moving otherwise
09:43:27 <oklopol> also apart from the fact i always forget to put numlock on/off, i love this
09:43:53 <oklopol> and it's as fast as it can be now, so pressing ctrl actually makes it move too fast
09:44:01 <oklopol> should probably slow it down a bit and use ctrl
09:44:15 <oklopol> the default setting it like one pixel a sec
09:44:40 -!- oklofok has quit (Read error: 110 (Connection timed out)).
09:44:55 <ehird> even at full it's slow to get moving for me
09:45:16 <ehird> also you can't change direction without losing momentum, which is annoying
09:45:36 <oklopol> at full speed + ctrl, i can't move it less than half the screen
09:46:32 <oklopol> i wish there was an ui that changed constantly
09:47:04 <oklopol> i suppose i pronounced it in finnish
09:47:41 -!- KingOfKarlsruhe has joined.
09:48:06 -!- oklofog has joined.
09:50:27 -!- oklofog has quit (Client Quit).
10:06:28 -!- oklopol has quit (Read error: 110 (Connection timed out)).
10:11:28 -!- KingOfKarlsruhe has quit (Remote closed the connection).
10:46:56 -!- GreaseMonkey has joined.
11:04:40 -!- bsmntbombdood has quit (Read error: 110 (Connection timed out)).
11:05:22 -!- bsmntbombdood has joined.
11:08:55 -!- fizzie has joined.
11:14:08 <ehird> [[Police 'encouraged' to hack more
11:14:10 <ehird> The Home Office has signed up to an EU strategy against cybercrime that "encourages" police across Europe to remotely access personal computers.]]
11:14:13 <ehird> one step closer to fascism
11:14:37 <ehird> the UK's politics might be better (but not by much) than America's, but our day-to-day oppression is far worse
11:19:00 <ehird> darcs install time
11:19:07 -!- ehird has quit (Read error: 104 (Connection reset by peer)).
11:47:41 -!- GreaseMonkey has quit ("HydraIRC -> http://www.hydrairc.org <- Nobody cares enough to cybersquat it").
11:49:06 -!- atrapado has joined.
12:06:43 -!- oerjan has joined.
12:23:27 -!- sebbu has joined.
12:23:52 -!- Pthing has joined.
12:29:26 -!- sebbu2 has quit (Read error: 60 (Operation timed out)).
12:53:12 -!- oerjan has quit ("leaving").
13:03:24 -!- FireyFly has joined.
13:10:25 -!- pikhq has quit (hubbard.freenode.net irc.freenode.net).
13:10:26 -!- SimonRC has quit (hubbard.freenode.net irc.freenode.net).
13:11:57 -!- pikhq has joined.
13:11:57 -!- SimonRC has joined.
13:27:27 -!- KingOfKarlsruhe has joined.
13:35:59 -!- ehird has joined.
13:36:25 <ehird> sigh, even in AUR it seems that arch linux has no package for the beautiful pristine Hobbit netcat, untouched by any patches
13:36:54 <ehird> I don't want IPv6, I don't want super-hardened super-ugly OpenBSD crap, I don't want whatever Debian did to it and I definitely don't want any netcat clones... just give me netcat...
13:37:15 <ehird> so I guess I'll be compiling that myself. Anyone alive? I haven't installed X yet.
13:38:54 <ehird> Guess not. I'll go install X now, and then some window manager.
13:38:56 -!- ehird has quit (Client Quit).
13:39:35 <fizzie> Ha, the "let's keep silent so ehird thinks the channel is empty" conspiracy succeeded magnificentiously.
13:58:11 -!- BeholdMyGlory has joined.
14:01:15 -!- pikhq has quit (hubbard.freenode.net irc.freenode.net).
14:01:15 -!- SimonRC has quit (hubbard.freenode.net irc.freenode.net).
14:02:13 -!- pikhq has joined.
14:02:13 -!- SimonRC has joined.
14:06:00 -!- pikhq has quit (Remote closed the connection).
14:11:07 -!- pikhq has joined.
14:11:30 -!- FireyFly has changed nick to FireFly.
14:13:09 -!- Gracenotes has quit (Remote closed the connection).
14:15:54 -!- puzzlet has quit (Read error: 54 (Connection reset by peer)).
14:16:14 -!- puzzlet has joined.
14:17:53 <fizzie> AnMaster: Doesn't your fspace_vector_is_cardinal (and the same thing in vector.c) have an unnecessary test? You do x = ABS(v->x), y = ABS(v->y) and then return false if x+y != 1, which is good; but then you also return false if "x && y". At that test, you already know that x+y == 1 and both are positive integers; it's not really possible for "x && y" to be true, since then x+y would be at least 2.
14:19:08 -!- kar8nga has joined.
14:21:10 <fizzie> AnMaster: (And related to previous, are you sure that ABS is safe? I see you cast it to unsigned, but I doubt that really helps; abs/llabs returns something signed and the documentation for them says it's undefined behaviour if the result is not representable, which it is for the -2^(N-1) case.)
14:28:59 -!- augur has joined.
14:56:06 <Deewiant> fizzie: The former sounds like something the optimizer might pick up on.
15:15:39 <AnMaster> fizzie, how comes you looked at that?
15:16:47 <fizzie> "Just browsing." I've been doing some jitfunge cleanup lately, just peeked a bit at your guts.
15:16:56 <fizzie> Anyway, must go to shoppery now, away.
15:20:15 -!- MigoMipo has joined.
15:32:33 -!- Asztal has joined.
15:33:51 -!- KingOfKarlsruhe has quit (Remote closed the connection).
15:58:29 -!- ais523 has joined.
15:59:10 <ais523> theory: the UK TV advertising wavelengths are currently satuated with insurance price comparison website adverts
15:59:16 <ais523> so clearly, we need a price comparison comparison website
16:03:32 -!- kar8nga has quit (Remote closed the connection).
16:04:59 <ais523> $ sudo chown -R ais523:ais523 /home/ais523/.gnupg/*
16:05:13 * ais523 wonders why that ended up set to root:root anyway
16:07:04 <AnMaster> fizzie, can you think of a better, but equally fast way?
16:07:15 <AnMaster> dropping the un-needed test of course
16:14:47 -!- coppro has quit (Read error: 110 (Connection timed out)).
16:22:35 -!- BeholdMyGlory has quit (Remote closed the connection).
16:26:41 * ais523 tries to understand Slashdot groupthink
16:26:52 <ais523> Microsoft's tests find that the worst-case upgrade time for Windows 7 is about 20 hours
16:26:57 <ais523> Slashdot accuse Microsoft of spreading FUD
16:28:05 <AnMaster> ais523, err. I thought slashdot was against windows 7
16:28:21 <ais523> although there's a strong counter-contingent that's in favour
16:28:31 <ais523> it's just, accusing Microsoft of spreading FUD about Windows is ridiculous
16:29:07 -!- BeholdMyGlory has joined.
16:47:01 <AnMaster> ais523, btw, I saw a laptop with a circular touchpad today
16:58:44 <AnMaster> ais523, is there any good reason to have a rectangular screen btw?
16:58:59 <ais523> it's easier to pack into memory in a quick-to-unpack way
16:59:07 <ais523> multiplicative addressing on the video RAM is pretty simple
16:59:21 <ais523> also, there are technical reasons why you'd want to make a cathode-ray-tube screen rectangualr
16:59:24 <AnMaster> ais523, polar coordinates for the pixels on screen would be kind of interesting though
16:59:34 <ais523> although they don't apply as much to CRTs
16:59:40 <AnMaster> ais523, this would not apply to LCD I guess?
16:59:57 <ais523> finally, you generally want all the pixels to be the same shape
17:00:02 <ais523> which would make a circular screen hard
17:00:18 <ais523> although it wouldn't rule out, say, hexagonal or triangluar screens
17:00:54 <AnMaster> ais523, also weren't those old radars circular?
17:01:05 <ais523> yes, because the range of the radar was circular
17:01:11 <ais523> and they didn't want to waste phosphor
17:01:29 <AnMaster> so what are those technical reasons then you mentioned?
17:01:39 <ais523> it's to do with the way that the electron beam is steered
17:01:54 <ais523> generally speaking, you can aim at a rectangular region
17:11:20 <AnMaster> ais523, any idea if there is a parallel gzip implementation?
17:11:41 <ais523> it's clearly possible, sort-of
17:11:48 <ais523> in that you could just cut the file up and zip each part separately
17:13:02 <AnMaster> ais523, I meant in a way so that it can be unpacked by standard gunzip
17:13:13 <AnMaster> or parallel unpacking from standard gzip
17:13:49 <AnMaster> there is a parallel bzip2 iirc
17:16:04 <fizzie> I think it should be possible; I mean, you can do a zlib full-flush (which discards the dictionary pretty much) at any point you want, and the inflater must survive that.
17:16:22 <fizzie> It might be a bit suboptimal if you split it into too tiny pieces, though.
17:18:03 <AnMaster> fizzie, what about the question above about better implementation of that function?
17:18:28 <AnMaster> because the current ones compiles to very good asm on x86/x86_64 at least
17:19:23 <fizzie> Well, I don't have any obviously fast one; and of course compilation is compiler-dependant.
17:20:13 <AnMaster> fizzie, in general I optimise for gcc
17:20:29 <fizzie> How does it do abs/llabs, incidentally?
17:20:57 <AnMaster> fizzie, let me check, was a while ago I last looked
17:21:56 <fizzie> If they'd stick me in front of an assembler with no Internet connection and no references available, I'd probably write something like "test rN, rN; jns .skip; neg rN; .skip:" but maybe there's something clevererer.
17:22:11 <ais523> fizzie: which dialect is that?
17:23:27 <AnMaster> fizzie, the abs() code is jump free even
17:23:49 <AnMaster> only when it gets to the actual tests are there jumps
17:23:57 <ais523> gcc-bf does abs by adding 128 and seeing if there's a carry
17:24:19 <ais523> (and you can tell if there's a carry by seeing if you go to 0 at any point during the addition loop)
17:24:57 <AnMaster> xorl (%rdi), %ecx # <variable>.x, x
17:25:03 <AnMaster> seems like the relevant code for x
17:25:12 <ais523> use sign-magnitude, then abs is trivial
17:25:52 <AnMaster> ais523, two complement is what I'm stuck with :P
17:26:03 <fizzie> ais523: Well, the ".foo" local-label was from nasm, but it's not certainly unique there. And nasm doesn't support ;s to separate instructions, I don't think, it's just that I wanted it on one line. And "rN" is standing for "rax, rbx, ..., r8, r9, ..."
17:26:19 <ais523> AnMaster: doing it in a circuit, you'd xor every bit with the top bit, then half-add the old top bit to the resutl
17:26:56 <AnMaster> ais523, mhm. That doesn't seem to make sense to me, but it works, and profiling showed it was fastest
17:27:11 <AnMaster> since it is called a lot in every ip moving, it is quite performance criticial
17:27:12 <ais523> (half-addition is like addition, but it can only add 0 or 1; and it takes exactly half the circuitry that a full adder does)
17:27:33 <ais523> AnMaster: heh, that works more or less exactly the same way as that hardware implementation I described
17:27:58 <fizzie> Yes, it's pretty nifty.
17:28:00 <AnMaster> ais523, oh I thought you were describing what the asm did?
17:28:02 <ais523> although it's using a different temporary
17:28:06 <ais523> AnMaster: no, I was working out out from scratch
17:28:19 <AnMaster> ais523, well it has to, what with the thing being passed by reference
17:28:34 <ais523> I mean, it's using a 32-bit value that's 0 or -1 as the temporary
17:28:37 <AnMaster> at least on x86_64, but probably best on x86
17:28:41 <ais523> rather than a 1-bit value that's 0 or 1
17:29:05 <ais523> but of course, 32 bits is as cheap as 1 bit on a 32-bit processor
17:29:13 <AnMaster> fizzie, anyway, what happens for the most negative case?
17:29:19 <AnMaster> you said it was undef according to docs
17:29:31 <ais523> AnMaster: returns the same value, I think
17:29:47 <ais523> also, why do you care about abs on every IP move?
17:30:14 <AnMaster> ais523, well no, not every one, but rather every wrap around,
17:30:16 <ais523> anyway, a jump-free test will be faster than a jumpy one, almost always
17:30:35 <ais523> unless the test takes the same branch almost every time
17:30:46 <ais523> mispredicted jumps require flushing the entire pipeline
17:30:56 <fizzie> ais523: There was some cautioning in the manuals not to blindly think that conditional moves and sets would be faster even though there's no jump involved.
17:31:01 <AnMaster> ais523, well, it isn't as bad on modern cpus as it was on pentium4
17:31:15 <fizzie> Or s/even though/because/ but anyway.
17:31:18 <ais523> fizzie: correct, it depends on the processor
17:31:25 <AnMaster> ais523, and yes what fizzie said, due to execution resources usage
17:33:02 <fizzie> Anyway, if it does return the same value, and then you cast that to unsigned, I think it would then turn out to be the correct 2^(N-1) unsigned number.
17:33:27 <fizzie> So it works, it just isn't guaranteed to.
17:34:03 <fizzie> I don't think many people use the largest negative delta in their programs.
17:35:19 <ais523> fizzie: and if they are, it hardly matters whether it's positive or not
17:35:25 <ais523> as you come to the same place both ways round
17:35:46 <fizzie> Yes, but "undefined behaviour" is not that limited.
17:36:50 <Deewiant> ais523: It can matter if the instruction you hit is something that depends on the delta
17:36:56 <fizzie> If you want the context; AnMaster is testing for "is this vector cardinal" by checking for abs(x) + abs(y) == 1, basically. (And the cardinality test is to use faster wrapping in that case.)
17:37:15 <ais523> Deewiant: oh, good point; but are there any non-fingerprint instructions that would care?
17:37:37 <ais523> y couldn't report the abs value of the most negative int
17:37:37 <Deewiant> [ and ] come to mind, but they just shunt the problem one instruction ahead.
17:37:49 <ais523> yes, and [ and ] just leave you back in the previous situation
17:37:50 <Deewiant> y reports the delta, so that's one.
17:38:02 <ais523> Deewiant: except, Befunge has no representation for INT_MAX+1
17:38:11 <ais523> so it would have to report the negative value
17:38:29 <AnMaster> hm. y wouldn't use the is cardinal test
17:38:42 <ais523> how often do people use wrapping in Befunge-98 programs, by the way?
17:38:59 <AnMaster> because it may be the shortest way
17:39:11 <ais523> Befunge-93 needs it quite a bit; in Befunge-98, though, it would tend to make programs less extensible
17:39:20 <Deewiant> ais523: {, which sets the storage offset to position + delta?
17:39:20 <AnMaster> ais523, and almost all the time it is cardinal wrapping
17:39:28 <AnMaster> because non-cardinal wrapping is such a pain
17:39:40 <ais523> Deewiant: comes to the same location both ways round
17:39:43 <Deewiant> But no, that brings you back to the same point as well
17:39:51 <fizzie> Incidentally, have you tried to give GCC the logical condition what you want -- something like "(x == 0 && (y == 1 || y == -1)) ||(y == 0 && (x == 1 ||x == -1))" -- and see what it optimizes to? It might be a bit too complicated-and-explicit to yield something extremely clever, but you never know.
17:40:31 <Deewiant> ais523: Then no, I don't think there are any outside fingerprints.
17:41:00 <ais523> x*y == 1 seems like an obvious way to do that
17:41:28 <ais523> hmm... although n*2^32+1 isn't prime for all n, so you'd have to make sure it didn't carry
17:41:35 <ais523> also, multiplication can be slow
17:41:38 <ais523> depending on the processor
17:41:54 <fizzie> Er... given that in all true cases either x or y is zero, I don't see how x*y == 1 tests for it.
17:41:57 <Asztal> you mean x*y == 0 && (y*y+x*x == 1)?
17:42:04 <AnMaster> <fizzie> Incidentally, have you tried to give GCC the logical condition what you want -- something like "(x == 0 && (y == 1 || y == -1)) || (y == 0 && (x == 1 || x == -1))" -- and see what it optimizes to? It might be a bit too complicated-and-explicit to yield something extremely clever, but you never know. <-- iirc it was something with lots of jumps
17:43:39 <ais523> Asztal: oh, good point, that's probably slowr
17:43:59 <fizzie> Speaking of wrapping, it's awful in jitfunge, since when you go and 'p' something outside the existing space bounds, I'll have to reconsider all traces that have wrapped that way, to see if they'd have instead hit whatever you put there.
17:45:23 <fizzie> It's a bit of the same hassle with ;; jumps -- compiled traces containing those need to care if someone puts a semicolon inside there.
17:45:29 <AnMaster> fizzie, I can easily think of programs where even CCBI will beat jitfunge
17:45:41 <AnMaster> I mean, the worst case scenarios seems rather bad
17:45:58 <ais523> adding a single semicolon between two existing semicolons strikes me as very bad coding practice
17:46:12 <AnMaster> ais523, what would the point even be?
17:46:24 <Deewiant> Commenting out code at runtime
17:46:28 <fizzie> Yes, for a Funge program it might well be an awesome way to redirect some code to go completely differently.
17:47:22 <fizzie> I don't think heavy self-modification will ever be very fast with jitfunge. It'd need a bit more heuristics that I want to implement to realize things like "okay, now you're messing around too much with this area, I'll just fall back to interpreting here and skip the JITting".
17:47:47 <AnMaster> fizzie, which adds some overhead of tracking that
17:48:20 <AnMaster> Deewiant, btw I thought about an interesting different way to beat slowdown
17:48:36 <AnMaster> it wouldn't be sane for other purposes though
17:48:55 <AnMaster> basically, make your value column indexed
17:48:58 <fizzie> Deewiant: Yes, to win the fight against the final boss in slowdown.
17:49:17 <Deewiant> I added a boss? My bad, I'll fix that
17:49:21 <AnMaster> this would work for cardinal workaround
17:49:38 <AnMaster> haven't worked out details for non-cardinal yet
17:49:57 <ais523> AnMaster: why not have a flag saying whether the IP is cardinal or not
17:50:01 <ais523> that every delta-changing command sets?
17:50:08 <ais523> would that be faster or slower
17:50:17 <AnMaster> ais523, that happens a lot more often than wraparound though in general
17:50:29 <AnMaster> so it sounds to me like a clear pessimiation
17:50:37 <ais523> which would just be setting the flag to a known value
17:50:50 <AnMaster> ais523, I use x a lot in my own code though.
17:51:06 <fizzie> I think I read somewhere that TraceMonkey keeps execution counters for JavaScript statements, and only JITs "hot" spots.
17:51:19 <ais523> AnMaster: I doubt x is more common than wraparound
17:52:14 <AnMaster> ais523, worth a try when I have time. I really was just looking for fixing the bug today, because I'm busy with uni currently
17:54:48 -!- atrapado has quit ("Ex-Chat").
17:56:19 <fizzie> I guess C doesn't specify what the unary - will do to INT_MIN on a system where |INT_MAX| < |INT_MIN|? What with being so tight-lipped about signed integer wrap-around in general.
18:18:05 -!- kar8nga has joined.
18:40:42 -!- oerjan has joined.
18:44:08 <Ilari> Also, aren't most signed overflows undefined behaviour?
18:44:56 <oerjan> AnMaster: read it hours ago
18:45:49 <Ilari> My guess about what -INT_MIN will do if |INT_MAX| < |INT_MIN| is 'undefined behavour'.
18:46:05 <oerjan> huh, google seems down
18:46:13 <fizzie> Yes, that is the reasonable guess.
18:46:57 <Ilari> For what it would do in pracice, I would guess -INT_MIN => INT_MIN.
18:47:32 <AnMaster> <oerjan> huh, google seems down <-- works here?
18:47:33 <ais523> INT_MIN is in limits.h, I think
18:47:36 * oerjan doesn't really remember enough C to know what to import
18:47:41 <ais523> which might not be in the standard headers
18:47:57 <oerjan> AnMaster: worked on second try, probably temporary glitch
18:47:59 <AnMaster> ais523, it is in the _C_ standard headers
18:48:12 <ais523> yes, but I meant EgoBot's
18:48:28 <fizzie> Heh, I just graduated today; and had to harass people over IRC to find out that fact. (Typically they send the email announcements out the same day, but for some reason today they didn't bother. Well, maybe I'll get the email tomorrow.)
18:48:33 <AnMaster> oerjan, I liked sqrt(-garfield) too
18:48:47 <ais523> and have you got your post-graduation application off in time?
18:48:54 <oerjan> AnMaster: i found the lower left panel hard to read :D
18:49:08 <oerjan> fortunately there was the script
18:49:19 <AnMaster> oerjan, and the link to the original
18:49:23 <fizzie> ais523: No hurry for that, the deadlines are just this Wednesday and Friday. :p
18:50:07 <fizzie> (They're pretty much written, though, I've just kept them waiting to find out whether I actually am a graduate or not.)
18:51:53 <oerjan> <ais523> so clearly, we need a price comparison comparison website
18:52:12 * oerjan would say yo dawg, but thinks that meme has died now
18:53:19 <oerjan> <ais523> $ sudo chown -R ais523:ais523 /home/ais523/.gnupg/*
18:53:19 <AnMaster> oerjan, all your yo dawg is dead
18:53:32 * oerjan suddenly realizes there needs to be a chowder command
18:53:34 <ais523> oerjan: are you going to point out that the /* is redundant?
18:53:48 <AnMaster> * oerjan suddenly realizes there needs to be a chowder command <-- what?
18:54:14 <oerjan> AnMaster: there's chown, there needs to chowder
18:54:16 <ais523> although, as far as I know, there aren't nested dotfiles in that hierarchy
18:54:23 <AnMaster> dot files *inside* a dot directory is evil though
18:55:05 <AnMaster> oerjan, chown, chmod, what would "chowder" be?
18:55:35 <oerjan> AnMaster: i just like chowder and thinks it deserves a command
18:55:56 -!- augur_ has joined.
18:56:04 <AnMaster> oerjan, what the hell is "owder"?
18:56:18 <oerjan> AnMaster: i have .www/.htaccess~
18:56:22 <Deewiant> AnMaster: It's nothing, but "chowder" is something.
18:56:29 <fizzie> Perl has "chomp", though it doesn't really change the omp-ness of anything.
18:56:37 -!- augur has quit (Read error: 104 (Connection reset by peer)).
18:56:39 <HackEgo> * a thick soup or stew made with milk and bacon and onions and potatoes \ [22]wordnetweb.princeton.edu/perl/webwn \ * Chowder is an American animated television series that debuted on Cartoon Network on November 2, 2007. ...
18:56:53 <AnMaster> Definitions of chowder on the Web:
18:56:53 <AnMaster> a thick soup or stew made with milk and bacon and onions and potatoes
18:56:53 <AnMaster> wordnetweb.princeton.edu/perl/webwn
18:57:14 <AnMaster> damn, I should check if anyone else written anything when I switch back from the browser window
18:57:18 * oerjan notes AnMaster is not yet one with HackEgo
19:08:57 <Ilari> Some would think: potatoes => Fail.
19:09:11 -!- adam_d has joined.
19:13:26 <Ilari> AnMaster: Too much carbohydrates in potato. And it isn't that good source of various micronutrients anyway...
19:14:06 <fizzie> But it's a potato! They're just so awesomely round.
19:14:21 <AnMaster> fizzie, err, no, they are knobbly
19:14:49 <fizzie> And gerund. Wait, no, not that.
19:15:26 <fizzie> Rincewind likes potatoes. :p
19:15:59 <fungot> Selected style: discworld (a subset of Terry Pratchett's Discworld books)
19:16:09 <fizzie> fungot: What says you about potatoes?
19:16:10 <fungot> fizzie: the fnord abbot folded his wrinkled hands and addressed lu-tze, one of those little uncertain laughs a man laughs who is unsure that he's not likeable. you couldn't help loving someone as soppy as that.
19:17:27 <Ilari> Markov generator? That seems so "reality shifted by 90 degrees"...
19:17:27 <fizzie> testlm-disk says about "potatoes are": potatoes are inherently evil.
19:18:20 <fizzie> Well, a generative n-gram model. Certainly you could formulate that in the Markov framework.
19:18:34 <fungot> AnMaster: glod gave buddy a lopsided look. " er. even the common people?' ' up here?' goodmountain, grimacing, balancing himself on two sticks.
19:19:05 <fizzie> I can't get it to say anything nice about potatoes. :/
19:19:08 <fungot> AnMaster: " we should always remember the songs he never had the patience to acquire it. just fnord" brutha began. and then, comrades, we must get...
19:19:13 <fizzie> They're always just evil and black.
19:19:40 <fizzie> "potatoes are there? iteration is where it's at"
19:19:53 <AnMaster> fungot, you do seem to pick one book that is easy to see in each, plus a bit of "garbage"
19:19:54 <fungot> AnMaster: ' he just looks as though they were, you know, that thing is in fnord of the
19:20:21 <fungot> AnMaster: " anyway, strong drink is a mocker, you know, i'm really rather pleased," said
19:22:12 <fizzie> I'm not sure where I had the source texts.
19:23:51 <fungot> FireFly: ' old boots muttered the manager. ' old boots... old boots.'
19:24:28 <fizzie> It doesn't seem to be directly from much anything.
19:24:39 <fizzie> Well, *that's* from Hogfather, I would say.
19:28:23 <AnMaster> is there any non-recursive definition of fibonacci's sequence?
19:29:59 <fizzie> There's that golden-ratio based one.
19:30:08 <Deewiant> http://en.wikipedia.org/wiki/Fibonacci_sequence#Closed_form_expression
19:30:30 <fizzie> You can also find it by typing a couple of the first Fib numbers to the online encyclopedia of integer youknowwhats.
19:31:01 <AnMaster> fizzie, uh no I don't know what it is
19:31:22 <Deewiant> http://www.research.att.com/~njas/sequences/
19:33:38 <fizzie> I think we did the derivation of that in the generating-functions course.
19:33:42 <Ilari> There's even standard method of getting "closed" formula for any sequence of form a_n = sum(b_k * a_{n-k}, k, 1, N), a_x given for 1 <= x <= N.
19:35:20 <Ilari> For 1 <= N <= 4, it results true closed formula. For higher N, one would need higher functions.
19:35:50 <oerjan> Ilari: that's just because you cannot solve quintics, i presume?
19:36:05 <oerjan> so it is still closed given the actual roots of the polynomial...
19:36:43 <oerjan> except there was some subtlety when the polynomial had duplicate roots
19:36:57 <Ilari> Yup. It involves solving roots of Nth degree polynomial.
19:38:32 <oerjan> AnMaster: basically you take the polynomial equation x^n = sum(b_k * x^n) and solve that. then every function of the form f(n) = x^n is a solution to the original recursion
19:38:48 <oerjan> so for fibonacci you get x^2 = x + 1
19:39:01 <AnMaster> oerjan, sum... what is written over and under the sum sign here?
19:39:17 <AnMaster> or is that not what your sum() does?
19:39:24 <fizzie> It looks like \sum_{k=1}^N b_k ...
19:39:46 <fizzie> At least that's what I'd intuitively read it as.
19:40:14 <oerjan> x^N = sum(b_k * x^n) i meant
19:41:09 <fizzie> oerjan: That still leaves "n" a bit unclear, actually, if the sum index is just k.
19:41:14 <oerjan> AnMaster: then, assuming the polynomial has N distinct roots (which might be complex), you can then compose every other solution to the recursion with linear algebra summing those x^n solutions
19:41:55 <oerjan> x^N = sum(b_k * x^k), last try :D
19:43:04 <oerjan> if some roots are not distinct, there is a trick to get extra functions for the duplicate roots but i'm not quite sure what it was
19:43:45 <oerjan> it may have been something as simple as n*x^n
19:44:39 <AnMaster> oerjan, that is: x^{N}=\sum_{k=1}^{N}b_{k}\times x^{k}
19:45:05 -!- KingOfKarlsruhe has joined.
19:45:15 <AnMaster> the extra {} comes from that I was using lyx, due to not remembering exactly how the \sum one worked
19:45:42 <AnMaster> <fizzie> It looks like \sum_{k=1}^N b_k ...
19:45:45 <AnMaster> <oerjan> (with fizzie's bounds
19:45:49 <oerjan> or maybe make that first x^{N+1}
19:45:50 <AnMaster> oerjan, so not his bounds then?
19:46:44 <fizzie> My bounds were for a different sum, actually.
19:46:57 <oerjan> AnMaster: it's essentially just the equation which a function f(n) = x^n needs to satisfy to solve the recursion
19:47:27 <fizzie> Ilari's "a_n = sum(...)", which makes sense as $a_n = \sum_{k=1}^N b_k * a_{n-k}$.
19:47:56 <fizzie> Coefficients from the recursion; b_1 = b_2 = 1 for Fibonacci.
19:47:59 <oerjan> AnMaster: the coefficients for the recursive linear equation you are trying to solve
19:48:35 <fizzie> Or, well, b_0 = b_1 = 0 for oerjan's sum, again.
19:49:06 -!- coppro has joined.
19:49:45 <oerjan> actually it is definitely best to start with x^0 as the lowest, otherwise you get a spurious x = 0 solution
19:51:03 <oerjan> *lowest x power in the sum
19:52:02 <oerjan> (well not strictly spurious, it's just trivial)
19:53:00 <fizzie> Incidentally, if you want to use the same bounds and b_k subscripts as in the recursion-defining $a_n = \sum_{k=1}^N b_k a_{n-k}$ up there, would your polynomial go $x^N = \sum_{k=1}^N b_k x^{N-k}$ too?
19:54:06 * oerjan should learn to read some day
19:54:52 <oerjan> actually that recursion definition should have $a_N = , shouldn't it
19:54:57 <AnMaster> <oerjan> oh you have a minus <oerjan> er subtraction <-- what is wrong with the first term for it?
19:55:18 <fizzie> oerjan: I think it's just trying to say a_n = b_1 a_{n-1} + b_2 a_{n-2} + ... there.
19:55:39 <oerjan> fizzie: maaaaaaaaaaaaaaaaaaaaybe
19:55:41 <AnMaster> now this decided to pass over my head
19:56:12 <fizzie> It's just notationally squibbles, really.
19:56:25 <oerjan> AnMaster: ok there's just some confusion because n and N are really the same up there
19:56:58 <fizzie> No, they're not; it's the recursive definition given for any a_n, while N is the number of coefficients.
19:57:14 <AnMaster> oerjan, well yes that explains a bit of it
19:57:24 <oerjan> AnMaster: scratch that
19:58:44 <oerjan> ok, then a_{n-k} becomes x^{N-k} etc.
19:58:52 <AnMaster> still confused. Anyway I don't think I need it, except for fib
19:59:08 <fizzie> What do you need a closed-form fib for?
19:59:28 <AnMaster> hm is the square root of *any* prime an irrational number? or is it just square root of 2?
19:59:39 <oerjan> AnMaster: for bit, a_n = a_{n-1} + a_{n-2} becomes x^2 = x + 1
20:00:24 * oerjan wonders what that came from
20:00:51 <AnMaster> fizzie, assignment is "implement a less stupid than naive fib algorithms", this is obviously geared against dynamic programming or whatever the term is.
20:01:16 <AnMaster> however, that is WAY to obvious
20:01:26 <oerjan> AnMaster: really, just set a_n = x^n and substitute
20:02:22 <AnMaster> oerjan, how did one kind out b_k now again? it was coefficients right. I suspect there may be a language barrier here..
20:02:50 <fizzie> I don't think there's anything especially stupid in the naive iterative Fibonacci; the unmemoized tree-recursion might be pushing it a bit, though.
20:02:57 <oerjan> koeffisienter is perfectly valid in norwegian ;D
20:03:09 <AnMaster> fizzie, that was the case here that was given to design a better one than
20:03:35 <oerjan> AnMaster: another efficient way of doing fibonacci uses matrix multiplication
20:03:43 <AnMaster> oerjan, koefficienter? Well I know about that, but not in the context of recursive definitions
20:04:10 <AnMaster> oerjan, got a link or something?
20:04:17 <AnMaster> oh wait it is there on wikipedia
20:04:41 <oerjan> f(n) = [[1 1] [0 1]]^n [1 1] modulo probably serious off-by-one errors
20:04:53 <AnMaster> um... we haven't got to "eigenvalues" yet, they are later in the course
20:05:26 <AnMaster> and I haven't been able to successfully figure it out myself from wikipedia
20:05:32 <oerjan> you don't eigenvalues to multiply matrices or take their powers
20:05:52 <oerjan> the eigenvalue thing is just another path to that x^n method, i think
20:06:17 <AnMaster> still, so far all we have done with matrixes has been the gauss/whatever-the-guy-was method for solving linear equation systems
20:07:04 <AnMaster> oh "Augmented matrix" is the English term
20:07:11 <AnMaster> it's "utökad matris" in Swedish
20:07:35 <oerjan> AnMaster: see the place in "Matrix form" section where it says closed form
20:07:54 <fizzie> Wikipedia flat-out gives you the F_{2n-1} and F_{2n} formulas, which I guess are what you'd get from the matrix power and a square-and-multiply exponentiation. (This is really a guess, I don't feel like thinking.)
20:10:28 <oerjan> you can essentially calculate F_n from the binary expansion of n (might be a bit off-by-one there since it's F_{2n-1})
20:10:43 <oerjan> mind you, F_n still grows exponentially
20:12:21 <oerjan> i know it can work because we implemented it in lambdabot in #haskell a couple of times
20:12:50 <AnMaster> oerjan, what about a non-exponential-to-calculate one?
20:12:51 <fizzie> Or to say the same thing in another way, but since I already had this thing mostly written: with those formulas, given F_n and F_{n-1} (okay, and F_{n+1}, but that's just their sum), you should be able to calculate F_{2n} and F_{2n-1} and use that to get F_{4n} and F_{4n-1} and so on; much like you can utilize x, x^2, x^4, x^8 you get from squaring things to compute any x^n.
20:13:23 <oerjan> AnMaster: er the _size_ of F_n is exponential in n, you're not going to get away from that is what i am saying
20:13:49 <AnMaster> oerjan, ah right. Where F is the sequence itself? (forgot if it was was upper or lower case f that was fib)
20:14:12 <oerjan> i haven't even bothered to get the notation right ;)
20:14:33 <oerjan> well, wikipedia uses F_n
20:17:21 <AnMaster> how comes you never see notation like: a < b > c ? (to say that b is bigger than both a and c, but not saying if a or c is largest)
20:17:31 <oerjan> http://en.wikipedia.org/wiki/Fibonacci_number#Another_identity seems like the most relevant section
20:17:34 <AnMaster> hm that would be same as a =< b < c
20:18:55 <Ilari> BTW: Eigenvalues are closely associated with groups of linear diffrential equations.
20:19:37 <AnMaster> Ilari, I don't think we got to the differential ones yet
20:20:01 <oerjan> Ilari: this fibonacci stuff is essentially linear _difference_ equations, and there are many analogies between the theories iirc
20:20:43 <AnMaster> oerjan, well it is in the chapter on "recursion and induction"
20:21:02 <Ilari> Basically, x is eigenvalue of matrix X if (xI - X)Y = 0 has nontrivial solutions (Y is not 0). Which is equivalent to det(xI - X) = 0.
20:21:05 <oerjan> hm or wait am i confusing it
20:21:36 -!- adam_d_ has joined.
20:21:42 <AnMaster> Ilari, where did the I in xI come from?
20:21:51 <oerjan> I is the identity matrix
20:21:55 <Ilari> AnMaster: I is identity matrix of apporiate size.
20:22:24 <ais523> why is #esoteric talking about eigenstructure?
20:22:26 <AnMaster> I think that is listed in the course description for November
20:24:03 <oerjan> <AnMaster> how comes you never see notation like: a < b > c <-- i think it's because you usually try to keep chains of relational operators somewhat transitive
20:24:16 <oerjan> you would say a, c < b instead
20:24:36 <oerjan> or b > a, c perhaps more likely
20:24:54 <fizzie> oerjan: "<oerjan> AnMaster: er the _size_ of F_n is exponential in n --"; the value of F_n grows exponentially, but does the *size* of it, really? I mean, F_n is something like φ^n, and the size of the decimal representation of 10^n is obviously linear in n, and yet 10^n > φ^n. Well, okay, maybe that was just a different definition of "size".
20:24:59 <AnMaster> oerjan, well, the first one looks nicer
20:25:09 <Ilari> If one uses multiply row and add to other to reduce the matrix to upper triangular form, the determinant is product of numbers in main diagonal in triangular form.
20:25:24 <oerjan> fizzie: i didn't want to mention that :D
20:25:50 <AnMaster> it made more sense than most other stuff so far
20:25:51 <fizzie> You could put "a" and "c" on top of each other, and next to a very large "< b".
20:25:53 <oerjan> however, f(n) is also exponential in n, and the size of f(n) is exponential in the size of n
20:26:16 <AnMaster> fizzie, that would look like a/c < b but someone forgot the _
20:26:19 <oerjan> the latter would be the relevant thing in complexity theory
20:27:23 <fizzie> In the very happy case your variables are a, b, e instead, you could use " < b"; but maybe that's a bit limiting.
20:27:39 <AnMaster> fizzie, ooh that would be another nice notation: for a < b < c: <small>a</small> <medium>b</medium> <large>c</large>
20:28:52 <AnMaster> also not sure how to differentiate =< and ==
20:29:34 <ais523> INTERCAL has had this idea already
20:29:39 <ais523> you can combine '. into !
20:30:11 <ais523> which is great, as ' is a paren character (nearly always opening paren in that context, although could theoretically be closing), and . is a sigil
20:30:47 <AnMaster> ais523, and what is the sigil for ?
20:31:08 <ais523> AnMaster: it means "16-bit scalar variable"
20:31:17 <fizzie> LaTeX gives you ten official sizes; \tiny, \scriptsize, \footnotesize, \small, \normalsize, \large, \Large, \LARGE, \huge and \Huge. (For some reason \HUGE is not listed; maybe it would be overkill.) Unfortunately those don't work in math mode.
20:31:22 <AnMaster> ais523, ah, wasn't the name one-spot?
20:31:24 <ais523> and there's no statement but comments in which .' isn't a syntax error
20:31:36 <ais523> those are the names of the sigil
20:31:44 <ais523> and by extension-of the types
20:31:51 <ais523> "sigil"'s a general programming language concept
20:31:55 <ais523> just like "variable", or "identifier"
20:32:22 <fizzie> Ooh, Wordnet doesn't know about "sigil". Shameful.
20:32:34 <ais523> strange, as it has a non-programming meaning too
20:32:38 <fizzie> Perl's $s, @s and %s are sigils, anyway.
20:32:56 <AnMaster> so why would wordnet NOT know it?
20:32:58 <ais523> interestingly, http://en.wiktionary.org/wiki/sigil doesn't have the programming meaning
20:33:04 <AnMaster> I mean, it isn't very uncommon
20:33:14 <AnMaster> ais523, what is the programming meaning?
20:33:21 <AnMaster> I know the non-programming meaning
20:33:25 <ais523> AnMaster: a character used to introduce a variable name
20:33:33 <ais523> often in a hungarian-notation style, but enforced by the language
20:33:46 <fizzie> AnMaster: <fizzie> Perl's $s, @s and %s are sigils, anyway.
20:33:49 <AnMaster> well, it is generally irritating unless shell
20:33:56 <AnMaster> fizzie, yes but I don't KNOW perl
20:34:02 <oerjan> yay spam on the esolang forum
20:34:12 <ais523> I keep forgetting the esolang forum exists
20:34:17 <fizzie> I assume everyone's seen *that* much of Perl.
20:34:20 <ais523> (and I'm not an admin there, so reporting it to me doesn't really help)
20:34:27 -!- MigoMipo has quit ("mi co'a sipna").
20:34:38 <ais523> AnMaster: how good are you at reading programs in languages you don't understand?
20:34:38 <oerjan> ais523: i got better after i subscribed to the rss
20:34:50 <ais523> I mean, not counting esolangs
20:34:51 <oerjan> although it's still months between posts
20:34:53 <AnMaster> ais523, depends on what language it is. if it is APL or K, or J: Not at all
20:35:05 <oerjan> ais523: i don't know who is an admin there, perhaps only graue
20:35:16 <ais523> there's a continuum, probably
20:35:27 <ais523> more or less anyone could read correctly-indented unobfuscated Lua, for instance
20:35:36 <ais523> but Perl's rather harder, due to its huge density
20:35:36 <AnMaster> ais523, C like? probably quite ok. Same for lisp-like. Haskell? Nop, perl neither.
20:35:42 <ais523> as in, you can say a lot in a few characters
20:35:46 -!- adam_d has quit (Connection timed out).
20:35:48 <AnMaster> but that seems uncommon for one-liners
20:35:59 <AnMaster> as for python, I know it a bit
20:36:11 <fizzie> There's a lot of Perl written that's pretty much like "shell script, except with regexps".
20:36:20 <fizzie> One-liners are a special breed though.
20:36:25 <ais523> AnMaster: http://ais523.sartak.org/TAEB-AI-Planar/lib/TAEB/AI/Planar.pm is a Perl file I've written that aims to be clear rather than obfuscated
20:36:29 <ais523> well, mostly me, sorear has a fwe patches there
20:37:10 <ais523> it's designed for downloading via darcs, rather than viewing online
20:37:20 <ais523> AnMaster: TAEB's lead maintainer
20:37:27 <ais523> who came up with the idea in the first place, and wrote quite a bit of the code
20:37:50 <AnMaster> ais523, what does "extends" do? kate's perl mode doesn't think it is a key word for example
20:38:02 <ais523> it's sugar for object-orientation
20:38:14 <ais523> it isn't part of the core lang at all
20:38:15 <AnMaster> ais523, brb phone, will take a few minutes
20:56:34 -!- SimonRC has quit (Read error: 113 (No route to host)).
21:18:46 -!- jix has joined.
21:23:00 -!- poiuy_qwert has joined.
21:38:46 <AnMaster> ais523, I just saw something funny:
21:38:56 <AnMaster> someone quoting "click" in the context of "link"
21:39:58 <AnMaster> (to be exact, the site with info for the local busses)
21:42:25 <AnMaster> 'Here you can see our time tables. "Click" on the time table on the left that you are interested in. You need Acrobat Reader to view the time tables.'
21:43:25 <AnMaster> I would say it is definitely a Swedish word that isn't uncommon, and hasn't been during the last 10 years or so
21:44:00 <AnMaster> FireFly, http://www.lanstrafiken.se/template/ResePlanerareUnder____2565.aspx
21:44:39 <FireFly> I read it as "someone writing 'click' instead of 'link'"
21:44:54 <AnMaster> FireFly, no I just find it funny to use quotes around that word
21:45:06 <FireFly> heh, yeah, well I wouldn't do that at least
21:46:36 <oerjan> i suppose you "wouldn't"
21:46:57 <AnMaster> oerjan, I fail to see anything wrong with "wouldn't"
21:47:21 <AnMaster> rather than pointing out some error
21:48:50 <AnMaster> oerjan, you can only take a pun so far before it becomes "lame"
21:49:51 <AnMaster> though, that maybe should have been "false"
21:50:02 -!- Gracenotes has joined.
21:50:23 <AnMaster> FireFly, it seems they go out of their way to make it accessible, except those features are broken without javascript
21:50:24 -!- adam_d_ has changed nick to adam_d.
21:50:54 <oerjan> AnMaster: you do realize that after the first couple of lines it's your reaction that makes it funny to continue, right?
21:51:22 <AnMaster> oh and why the hell would you provide a sign language version of a web page?
21:51:29 <AnMaster> audio version yes. but sign language?
21:51:46 <ais523> AnMaster: what does a sign language webpage look like?
21:51:56 <AnMaster> ais523, I haven't clicked on the link yet
21:52:18 <AnMaster> ais523, there is a link to a *.mov and *.wmv
21:52:49 <AnMaster> ok it is just someone talking the stuff in sign language it seems. Don't understand that of course
22:00:54 <AnMaster> hm, ok I have a theory: 1) they also provide the travel service for disabled people (anyone from US will have no clue what this is about) 2) This city is some sort of national center for higher education for deaf people
22:01:01 <AnMaster> perhaps there is some logic behind all this
22:04:24 <AnMaster> ais523, that Planar.pm, what does it do?
22:04:39 <AnMaster> I mean, it seems to need to fit into a much larger context to make sense
22:04:45 <AnMaster> it doesn't make sense stand-alone
22:06:15 <AnMaster> use constant veto_travel => 0; <-- huh?
22:06:24 <ais523> what's the huh for there?
22:06:29 <ais523> it simply creates a constant
22:06:36 <ais523> it's sugar for sub veto_travel { 0; }
22:06:38 <AnMaster> ais523, is this travel with level teleport?
22:06:51 -!- kar8nga has quit (Remote closed the connection).
22:06:58 <AnMaster> ais523, plus, what does the "use" mean
22:07:42 <ais523> and use constant imports constant.pm
22:07:46 <AnMaster> TAEB->config->get_ai_config->{'overall_plan'} // 'SlowDescent'; <-- that fucked syntax highlighting in kate, at the second / everything is green
22:07:48 <ais523> which is responsible for doing the sugar
22:07:53 <ais523> AnMaster: it's a 5.10ism
22:08:08 <ais523> // is a new operator in Perl 5.10
22:08:27 <ais523> replaces undef with the given value, has no effect on anything else
22:08:42 -!- poiuy_qwert has quit.
22:08:46 <ais523> i.e. if the left argument doesn't have a value, return the right argument
22:08:50 <ais523> otherwise return the left argument
22:08:54 <AnMaster> ais523, who thought such an operator was a good idea?
22:09:06 <ais523> I don't know; but it is a useful operator
22:09:14 <ais523> for doing things like default values
22:09:26 <ais523> and not quite sugar, it avoids having to use a temporary var
22:09:30 * oerjan thinks he wanted that // recently (ironically for the /// interpreter) but didn't know about it
22:09:32 <ais523> or write the left argument out twice
22:10:00 <AnMaster> traits => [qw/TAEB::AI::Planar::Meta::Trait::DontFreeze/],
22:10:04 <oerjan> i worked around it though :)
22:10:39 -!- Gracenotes has quit (Remote closed the connection).
22:10:40 <AnMaster> ais523, also what does the whole module thing do? implement some sort of strategy for this nethack bot?
22:10:48 <ais523> AnMaster: when freezing state, don't freeze that
22:10:56 <ais523> and yes, that's one of the modules doing the strategy for TAEB
22:11:16 <ais523> AnMaster: as in, continuations but on disk
22:11:36 <ais523> so you can exit the program, load it again, and it's in the same state again
22:11:48 <AnMaster> on tape or other long term storage media I guess it would be "freeze dried" even
22:13:12 <AnMaster> ais523, why is most stuff not stored?
22:13:24 <AnMaster> plus there seems to be an awful lot of repetition here
22:13:27 <ais523> it's stuff like caches that can be reconstructed anyway
22:14:15 <AnMaster> there is lots of complete copies, or very similiar entries to:
22:14:17 <AnMaster> isa => 'HashRef[ArrayRef[TAEB::AI::Planar::Plan]]',
22:14:17 <AnMaster> traits => [qw/TAEB::AI::Planar::Meta::Trait::DontFreeze/],
22:14:34 <AnMaster> surely this can be simplified in some way?
22:15:12 <ais523> yes, it could be, but it would be less clear
22:15:21 <ais523> you could generate the code at runtime, for instance
22:15:35 <ais523> several bits of TAEB actually do that, but it grates against my typical programming style
22:17:46 <AnMaster> "this can lead to inaccurate results sometimes when our resource levels change" <-- eh?
22:19:41 <AnMaster> ais523, "Phasers can walk anywhere." <-- is that true for all levels? Such as sokoban?
22:19:52 <ais523> no, it's an approximation for monster routing
22:20:01 <ais523> which has to be simple
22:20:09 <ais523> for efficiency reasons
22:20:17 <ais523> as in, we're trying to predict where enemy monsters will go
22:20:36 <ais523> and it's better to guess wrong in a way that gives the monsters more places to move to, than one that doesn't
22:20:41 <AnMaster> ais523, how much slowdown is caused by, say, perl?
22:20:53 <AnMaster> also why does it have to be fast? nethack is turn based
22:21:06 <ais523> so it can be tested in a reasonable amount of time
22:25:55 <AnMaster> ais523, where are the tactical plans listed?
22:26:23 <ais523> AnMaster: in a load of other modules
22:26:25 <ais523> that's just one of them
22:26:39 <AnMaster> ais523, wait, is this a tactical plan itself?
22:26:59 <ais523> it's a Perl module, the plans are something entirely different
22:27:05 <ais523> they're objects of classes defined in other modules
22:27:10 <AnMaster> <ais523> that's just one of them <-- was the bit that confused me
22:27:16 <ais523> that's just one of the modules
22:27:32 <AnMaster> ais523, so how many tactical plans are there?
22:28:11 <ais523> there are both tactical and strategic plans
22:28:52 <AnMaster> ais523, what sort of things are the strategic ones?
22:28:55 <ais523> including a few files that are not plans in themselves, but libraries used by plans
22:29:20 <ais523> see http://ais523.sartak.org/TAEB-AI-Planar/lib/TAEB/AI/Planar/Plan/
22:29:24 <ais523> for a list of both sorts of plans
22:29:31 <AnMaster> ais523, also why dislike levitation?
22:29:44 <ais523> things like SlowDescent are strategic; things like Walk are tactical
22:29:50 <ais523> and because lots of things can't be done while levitating
22:30:03 <ais523> and you need to recognise that fact in order to avoid getting stuck in an infinite loop
22:30:05 <AnMaster> ais523, levitation is still very useful
22:30:24 <AnMaster> say, getting up from below castle
22:30:29 <AnMaster> you don't want to fall down again
22:30:32 <ais523> AnMaster: yes... but, say you want to pick an item up
22:30:39 <ais523> you /need to know/ that you can't pick up items while levitating
22:30:45 <ais523> to a bot, you have to tell it
22:31:08 <ais523> and the talking about levitating is to tell it that
22:31:53 <ais523> AnMaster: read http://taeb-blog.sartak.org/2009/08/planar-taeb-ai.html for the explanation of how the code works
22:42:23 <fizzie> Nethack AI has long been on the list of possible topics on the Prolog course, I think; I'm not aware of anyone attempting it, though.
22:43:36 <ais523> fizzie: I'm attempting it
22:44:29 <fizzie> Yes, but on the course, I mean.
22:53:22 -!- KingOfKarlsruhe has quit (Remote closed the connection).
22:56:06 -!- adam_d has quit ("Leaving").
23:27:15 <ais523> hmm... it seems Novell ported Mono to the iPhone
23:34:55 -!- FireFly has quit ("Later").
23:35:48 <GregorR> Does that imply that one could write software for iShit without the iShit SDK?
23:37:00 * oerjan recalls reading that Apple won't accept any app which can run another program
23:38:13 -!- BeholdMyGlory has quit (Read error: 54 (Connection reset by peer)).
23:38:33 <coppro> like the C64 emulator THAT RUNS BASIC OMG
23:44:58 -!- ehird has joined.
23:47:45 -!- Sgeo has joined.
23:48:03 <GregorR> Yeah, I recall that too, but I thought you were implying the contrary.
23:48:10 <GregorR> But I guess Apple's douchebaggery knows no bounds.
23:48:22 <oerjan> ehird: how can you sleep in a well? isn't that awfully wet?
23:48:40 <ehird> aWell, well, well. What have we here. Some questions?
23:48:50 <ehird> GregorR: hurf durf
23:50:43 <ehird> 13:49:42 <AnMaster> 1> true == "true".
23:50:43 <ehird> 13:49:42 <AnMaster> false
23:50:43 <ehird> otoh, [116,114,117,101]=="true", and lists stringify as strings depending on if they're ascii or not
23:51:11 <ehird> must be nice changing [116, 114, 117, 3]'s last element to 101 and getting back "true"
23:52:26 <ehird> 14:08:54 <AnMaster> ais523, who thought such an operator was a good idea?
23:52:36 <ehird> it's "default the former to the latter"
23:52:39 <ehird> which is a good operation
23:52:47 <ehird> defaults when unavailable are a Good Thing
23:52:52 <ehird> and // encapsulates that
23:53:34 <coppro> ehird: you can call a function that prints it as a numerical list I think
23:53:47 <ehird> who cares, that default behaviour is fucked up :)
23:54:07 <ais523> ehird: which lang is that list example in?
23:54:17 <ehird> strings are lists of numbers
23:54:28 <ehird> but all lists with ascii printable contents stringify as strings
23:54:32 <coppro> and the interpreter prints them in strings if it can, in numbers otherwise
23:54:33 <ais523> "lists stringify as strings"
23:54:40 <ais523> depending on if they're ASCII?
23:54:43 <ehird> ais523: [97,98] shows as "ab"
23:54:44 <coppro> that's just the default printing
23:54:47 <ehird> I haven't seen anything like Erlang's lists-are-magically-strings in another language
23:54:52 <ehird> I think I'll call it magic-typing
23:55:02 <ehird> this list MAGICALLY becomes a string!
23:55:08 <coppro> they aren't; lists and strings are the same consturct
23:55:08 <ehird> oops, it's back again... magically
23:55:21 <coppro> it just changes how it's shown
23:56:53 <ehird> 13:44:00 <AnMaster> FireFly, http://www.lanstrafiken.se/template/ResePlanerareUnder____2565.aspx
23:57:05 <ehird> my eyes are really soothed by the colour scheme there...
23:57:09 <ehird> they just really like it
23:57:17 <ehird> well, mainly the header
23:58:00 <ehird> yeah don't they know it's just Adobe Reader now??????????????????? :P
23:58:43 <ehird> 15:35:48 <GregorR> Does that imply that one could write software for iShit without the iShit SDK?
23:58:43 <ehird> the SDK isn't shit... although what you said didn't imply that anyway
23:59:04 <ehird> i mean objective-c/cocoa is pretty okayrad
23:59:57 <ehird> i don't want to say rad because i haven't used it for anything myself, but the general consensus is that it's rad, and what i've read seems awesome