00:05:56 -!- Phantom_Hoover has quit (Ping timeout: 245 seconds).
00:21:23 -!- hailtothethief has quit (Ping timeout: 240 seconds).
00:22:09 -!- clochette has joined.
00:23:01 -!- clochette has quit (Client Quit).
00:40:46 <alise> <nooga> i like stoner rock
00:40:48 <alise> SO HORRIBLE, TO LIKE
00:46:37 <Ilari> Wasn't that address (192.88.99.1) the global IPv4 anycast of 6to4 gateways?
00:56:49 -!- jcp has quit (Ping timeout: 264 seconds).
01:10:12 <alise> heh, BytePusher runs at 3.93 MHz
01:12:08 <alise> cool, bytepusher uses C as an assembler XD
01:17:16 <alise> Gregor: It has a little header of C code that defines some stuff like jmp() and lbl().
01:17:25 <alise> Then it has loops to automate the generation of tedious code.
01:17:31 <alise> And it all gets spat out to a .BytePusher file.
01:17:35 <Gregor> I'm confused by what the 4MHz number has to do with the C-as-asm statement.
01:17:43 <alise> I was just looking at BytePusher.
01:19:27 <alise> it's common in minimalist riscs to have a skip-if-condition and a single unconditional jump instruction rather than conditional jumps, right?
01:20:44 <nooga> i wonder if one could mimic try ... catch ... finally in C
01:21:16 <oerjan> try ... fail ... give up
01:21:19 <alise> Gregor: you mean have an extra conditional in each instruction?
01:21:20 <nooga> using clever macros
01:21:26 <alise> but means every instruction could branch
01:21:26 <Gregor> nooga: A few wonky macros, no problem.
01:21:28 <alise> which is sort of ~not fun~
01:21:34 <alise> nooga: yes, it's been done
01:21:42 <alise> Gregor: hmm, does it do anything fancy?
01:22:02 <Gregor> alise: Almost always the condition is just used on branch instructions, but ANY instruction can be conditional.
01:22:10 <alise> "try and catch in C", fourth freaking result
01:22:18 <alise> Gregor: That must fuck everything up.
01:22:34 <alise> Gregor: Also, that's a pretty CISC thing to do, isn't it?
01:22:45 <alise> MIPS is cooler anyway :P
01:22:45 <Gregor> alise: ARM isn't really RISC :P
01:22:59 <Gregor> MIPS just has conditional branches.
01:23:41 <alise> Gregor: what's the point of condition flags and the like? are they faster than registers? i've forgotten
01:23:53 <alise> or is it just to avoid legislating a format for booleans :P
01:23:56 <Gregor> I'm not architectury enough to answer that.
01:25:52 * alise wonders if "eq d,a,b" should leave d alone if a =/= b, or set it to zero
01:25:54 <alise> the former, I think
01:26:02 <alise> you can always zero it beforehand
01:26:50 <alise> Gregor: all these registers are so bloated compared to OISCs >_>
01:26:59 <alise> The PDP-8 only had 8 instructions, and it was CISC! :P
01:30:21 <Ilari> Sgeo: BTW, how much you get from test-IPv6 tests? :-)
01:38:27 <alise> LD r1, r1; LD r2, r2; EQ r3, r1, r2; SNZ r3; JMP end; ... end:
01:38:31 <Sgeo> I just found out about it from Wikipedia
01:38:34 <alise> That's an awfully long series of instructions for if (*x==*y)
01:39:54 <Ilari> I get 10/10 for both dual stack and v6 only tests... :-)
01:39:59 <alise> Modern top-down programming is tested rigorously at every step even before the full design has completed; unfinished components are filled in by stubs so that the program can be tested. ...So top-down programming has turned into a freaky version of bottom-up programming. Lovely!
01:40:16 <alise> Ilari: If only anyone else did so you could put it to use >:)
01:44:22 <alise> 07:15:05 <fizzie> And dbc's ascii-art.
01:44:30 <alise> [suddenly, dbc talks]
01:44:40 <alise> 07:15:32 <fizzie> I think there was in... not more than five years ago.
01:47:25 <dbc> I don't talk much.
01:50:07 -!- Kordalien has joined.
01:55:41 <alise> dbc: You're so predictable, talking when someone says you're not going to talk.
01:55:52 <alise> dbc: In fact, I bet you don't say something very soon.
01:57:07 -!- alise has quit (Quit: Leaving).
02:00:42 -!- cal153 has quit (Ping timeout: 272 seconds).
02:15:01 <Gregor> Somebody needs to write an OISC backend for GCC.
02:15:06 -!- Mathnerd314 has quit (Ping timeout: 245 seconds).
02:21:24 -!- MB_MB has joined.
02:22:58 -!- MB_MB has left (?).
02:27:38 -!- cal153 has joined.
02:34:45 <Gregor> And binutils I spose ..
03:50:16 -!- augur has joined.
04:14:16 -!- oerjan has quit (Quit: Good night).
04:34:32 -!- lament has joined.
04:47:13 -!- augur has quit (Ping timeout: 264 seconds).
05:02:44 -!- sshc_ has joined.
05:05:03 -!- augur has joined.
05:05:55 -!- sshc has quit (Ping timeout: 240 seconds).
05:17:26 <coppro> what's wrong with this picture (hint, 4 things): 00:16 < Oleg_> I have a question. What C++ code for unicode would correspond to ANSCII value 129?
05:18:42 -!- GreaseMonkey has joined.
05:19:25 <Sgeo> Wait, n.. oh, still
05:19:54 <Sgeo> was thinking the 129, but then thought it was acceptable, but then that's some extended junk, right?
05:19:56 <lament> 129, "code for unicode", and Oleg
05:20:05 <Sgeo> "C++" has nothing to do with anything
05:21:19 <coppro> my 4 were: ANSCII, 129, "C++ code for unicode" twice; once for assuming C++ has any unicode support, twice for asking for a "code for unicode"
05:21:59 <Sgeo> As far as the C++ thing goes, I just saw it as "Well, the question should fundamentally be language-independent"
05:22:01 <lament> also you shouldn't use the nick Oleg unless you're really smart
05:24:20 <lament> coppro: because the real Oleg is the person who owns this site http://okmij.org/ftp/
05:25:37 <coppro> that's quite impressive
05:25:45 <coppro> incidentally, I think I'm insane
05:26:07 <coppro> I'm considering trying to gun for a triple major and a minor
05:40:09 <Gregor> With those majors in some things borderline-useful?
05:40:17 <Gregor> Or are we talking philosophy, linguistics and psychology?
05:43:36 <pikhq> Gregor: Linguistics is borderline-useful. Keep in mind that you put it into application rather often. ;)
05:44:26 <Gregor> I guess by "borderline-useful" I mean "puts food on the table"
05:44:46 <Gregor> Linguistics can only do that paired with something actually useful (e.g. CS)
05:45:00 <pikhq> Unless you're in academics.
05:45:12 <pikhq> In which case it doesn't matter you could be studying basket weaving.
05:46:01 <pikhq> And psychology is something like 3/4ths bullshit, due to lack of knowledge.
05:46:11 <pikhq> Philosophy? Formal study of bullshit.
05:46:26 -!- Harpyon has joined.
05:46:30 <pikhq> Which makes it both useless and astoundingly fascinating at the same time.
05:47:48 -!- Harpyon has quit (Client Quit).
05:51:41 -!- sftp has quit (Remote host closed the connection).
05:55:39 <Gregor> ALSO pikhq: Go read the Fythe spec dammit!
05:56:45 <Gregor> (The first bit I actually added some text to that is)
06:01:59 -!- Rugxulo has joined.
06:02:24 <Rugxulo> welp, haven't been here in a while
06:04:15 <Rugxulo> probably missed something interesting ... oh well (not going to read hundreds of old logs, even if it has only been like two or three months, heh)
06:07:42 <Rugxulo> okay, I'll just say it, I know it's lame to most of you, but I did port my Befunge-93 interpreter to Modula-2, and I (optionally) crammed its source into less than 80x25 (w/ example too)
06:08:14 <Rugxulo> crappy, crazy, useless but whatever ... tada! ;-)
06:10:16 -!- Harpyon has joined.
06:18:14 -!- Harpyon has quit (Quit: Harpyon).
06:19:55 <coppro> things I want: candies that taste like vitamin C pills but do not give indigestion
06:23:57 <coppro> I mean like exactly the same
06:24:14 <Rugxulo> I just meant why complicate it? just eat an orange ;-)
06:25:37 <coppro> because it does not taste the same
06:26:05 <Rugxulo> close enough, plus it's easily available and has vitamin C
06:26:19 <Rugxulo> and probably cheaper too (hopefully)
06:27:29 <coppro> b) I don't want vitamin C. The problem is the damned pills give you indigestion if you have too many
06:28:13 <coppro> because they are yummy
06:28:34 <Rugxulo> well your stomach doesn't, obviously, so you have to compromise
06:29:51 <coppro> Sgeo: you're an expert at not working. help me stop
06:47:42 <lament> i think philosophy is pretty cool
07:00:17 -!- tombom has joined.
07:45:31 -!- Rugxulo has left (?).
07:50:13 -!- tombom has quit (Quit: Leaving).
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:32:12 -!- dbc has quit (Quit: Seeeeeya).
08:34:22 -!- Kordalien has quit (Ping timeout: 276 seconds).
08:50:49 -!- cheater99 has quit (Quit: Leaving).
09:10:41 -!- dbc has joined.
09:25:35 -!- lament has quit (Ping timeout: 276 seconds).
11:07:42 -!- GreaseMonkey has quit (Quit: Welcome honored guest. I got the key you want! would you like onderves. of Yourself).
11:31:42 -!- sebbu has quit (Read error: Connection reset by peer).
11:32:22 -!- sebbu has joined.
11:57:08 -!- Sgeo has quit (Read error: Connection reset by peer).
12:06:17 -!- Sgeo has joined.
12:16:42 -!- oerjan has joined.
12:40:25 -!- rodgort has quit (Quit: Coyote finally caught me).
12:40:37 -!- rodgort has joined.
12:52:04 -!- Phantom_Hoover has joined.
13:04:07 -!- BeholdMyGlory has joined.
13:29:25 -!- oerjan has quit (Quit: Later).
13:32:34 -!- Harpyon has joined.
14:00:36 -!- Phantom_Hoover has quit (Quit: Leaving).
14:00:52 -!- Phantom_Hoover has joined.
14:01:41 <Slereah> Is it about a middle aged woman who discovers the internet
14:01:46 <Phantom_Hoover> Fortunately, I saw the trailer for it immediately before watching Inception, which erased the horror from my mind for at least a month.
14:01:46 <Slereah> Because that is basically facebook
14:04:41 -!- augur has quit (Ping timeout: 245 seconds).
14:04:58 -!- BeholdMyGlory has quit (Remote host closed the connection).
14:17:45 -!- sshc_ has changed nick to sshc.
14:39:27 -!- augur has joined.
14:56:10 -!- ineiros has quit (Quit: x).
15:14:57 -!- ineiros has joined.
15:33:40 -!- sftp has joined.
15:49:35 -!- sebbu2 has joined.
15:52:36 -!- sebbu has quit (Ping timeout: 245 seconds).
16:00:21 -!- alise has joined.
16:01:52 <alise> 21:17:26 <coppro> what's wrong with this picture (hint, 4 things): 00:16 < Oleg_> I have a question. What C++ code for unicode would correspond to ANSCII value 129?
16:02:02 <alise> "code for unicode", esp. supposing unicode is a charset
16:02:10 <alise> 129 is outside of ASCII
16:02:31 <alise> prolly "C++ code for [meant to be charset]" and thinking "unicode" is a charset is two things
16:03:20 <Phantom_Hoover> An immensely annoying number of people think the ASCII is 8-bit.
16:03:30 <alise> 21:46:11 <pikhq> Philosophy? Formal study of bullshit.
16:03:34 <alise> my philosophy-studying friend can confirm this :)
16:03:43 <alise> Phantom_Hoover: lawl "the ASCII" :)
16:03:58 <alise> 22:19:55 <coppro> things I want: candies that taste like vitamin C pills but do not give indigestion
16:04:07 <alise> multivitamins: stupid
16:04:09 <alise> their taste: AWESOME
16:08:20 <Gregor> You can just chug citric acid.
16:09:31 <Gregor> The American Standard Code for I[something] Interchange?
16:09:43 <Gregor> Hahaha you said "The"!
16:10:01 <Phantom_Hoover> I was waiting for someone to notice that I say that, actually.
16:10:56 <Gregor> Well, people on Internet tend to add or remove the word "the" indiscriminately.
16:10:59 <Phantom_Hoover> It occurred to me one day and I haven't been able to stop since.
16:11:36 <Phantom_Hoover> I suspect that I shall now end up talking about the TCP and the UTF-8.
16:12:03 <Gregor> The transfer control "TCP" protocol.
16:12:21 <Gregor> Dern, /transmissions/ control.
16:12:36 <Gregor> Dern, /transmission/ control (where'd that plural come from :P )
16:13:06 <alise> If I get a Ph.D. and change my name to Iliad, I'll be Dr. I. Hird, or, backwards, Dr. I. Hird.
16:13:10 <alise> I have a new life goal.
16:13:46 <alise> Phantom_Hoover: You do realise that acronyms aren't really their expansions?
16:13:48 <Phantom_Hoover> Why not change your last name to whatever ais' middle one is.
16:13:52 -!- Wamanuz has joined.
16:13:52 <alise> That's why RAS syndrome is a stupid complaint/.
16:13:59 <alise> "HURR HE SAID ATM MACHINE"
16:14:01 <alise> Phantom_Hoover: Ian?
16:14:06 <alise> Because Ian is a crappy name!
16:14:11 <alise> Maybe Imhotep. I would become invisible.
16:14:34 <Phantom_Hoover> Scottish names are good atmaking the English mispronounce them.
16:16:36 -!- Wamanuz has quit (Remote host closed the connection).
16:16:57 -!- Wamanuz has joined.
16:19:27 -!- BeholdMyGlory has joined.
16:19:32 -!- BeholdMyGlory has quit (Changing host).
16:19:32 -!- BeholdMyGlory has joined.
17:00:31 -!- Wamanuz has quit (Ping timeout: 245 seconds).
17:01:53 <Gregor> Well, nobody likes the English.
17:03:05 -!- BeholdMyGlory has quit (Read error: Connection reset by peer).
17:03:34 -!- BeholdMyGlory has joined.
17:05:39 -!- Wamanuz has joined.
17:05:52 <Gregor> The English LEAST of all.
17:06:00 <Gregor> Nobody hates the English more than the English.
17:06:59 -!- tombom has joined.
17:08:33 <Phantom_Hoover> In the same vein, HP Lovecraft was a dedicated Anglophile.
17:11:19 <alise> "I know that astrology isn't a science," said Gail. "Of course it isn't. It's just an arbitrary set of rules like chess or tennis or, what's that strange thing you British play?"
17:11:19 <alise> "Er, cricket? Self-loathing?"
17:13:01 <alise> (Next line: ""Parliamentary democracy. The rules just kind of got there. They don't make any kind of sense except in terms of themselves. But when you start to exercise those rules, all sorts of processes start to happen and you start to find out all sorts of stuff about people. [...]")
17:15:15 <alise> Phantom_Hoover: And it left us with "...And Another Thing".
17:15:33 <alise> A work of fanfiction so unsatisfactory as an entry in the H2G2 trilogy, it makes me want to write a better one.
17:16:13 <alise> Phantom_Hoover: Oh, it is an excellent imitation of Douglas' style. But it's ... frozen, somehow, a stereotype rather than a homage.
17:16:29 <alise> And while it's a decent book, it's a bad H2G2 book.
17:16:35 <alise> Rename the characters, republish, good book.
17:17:51 -!- augur has quit (Remote host closed the connection).
17:25:17 <Sgeo> I think the only reason my singing sounded bad was because it wasn't in-sync with the music underneath
17:25:24 <Sgeo> I think raw singing by itself should be ok
17:29:52 <Phantom_Hoover> The tone was all wrong, the voice was annoying and the timing wasn't just desynced, it was completely wrong.
17:31:29 <Sgeo> But those things, in and of itself, shouldn't be grating
17:32:54 <alise> Sgeo: *You* *cannot* *sing*
17:33:01 <alise> It is as simple as that, my friend.
17:33:28 <Sgeo> So when I'm bored at a bus stop, I should try to avoid vocalizing whatever song's floating through my head?
17:33:39 <Phantom_Hoover> I cannot sing either. I realise this, and I avoid singing at all costs.
17:34:29 <Phantom_Hoover> http://www.conservapedia.com/Obamageddon THIS WORKS EVEN BETTER WITH AN ATHEIST NON-RHOTIC ACCENT
17:34:42 <alise> <Sgeo> So when I'm bored at a bus stop, I should try to avoid vocalizing whatever song's floating through my head? ;; oh god yes
17:35:08 <Sgeo> Can I at least hum?
17:35:44 * alise listens to your thing again juts to make sure
17:35:46 <Phantom_Hoover> DO NOT ATTEMPT TO MAKE ANY FORM OF TUNE WITH YOUR LUNGS
17:36:12 <alise> http://sgeo.diagonalfish.net/unknownsongs.wav ;; yeah don't hum dude :{
17:36:15 -!- FireFly has joined.
17:36:22 <Sgeo> alise, that's not humming
17:37:22 <alise> Phantom_Hoover: i need a link :{
17:37:41 <Phantom_Hoover> alise, http://sgeo.diagonalfish.net/paint_it_black_karaoke.ogg
17:38:06 <Phantom_Hoover> I'm actually fairly desensitised to Sgeo's voice by now.
17:38:29 <alise> Phantom_Hoover: Postcode, weaknesses?
17:38:54 <alise> Phantom_Hoover: Is their weakness GUNS?
17:38:57 <Phantom_Hoover> I seem to infuriate him, though, so that's one attack strategy.
17:52:59 -!- hailtothethief has joined.
18:01:00 -!- Wamanuz has quit (Read error: Connection reset by peer).
18:01:12 -!- Wamanuz has joined.
18:01:45 -!- Wamanuz has quit (Read error: Connection reset by peer).
18:02:02 -!- Wamanuz has joined.
18:12:08 <Gregor> Phantom_Hoover: DO NOT ATTEMPT TO MAKE ANY FORM OF TUNE WITH YOUR LUNGS // /me tries to figure out a way to make a tune with his lungs that isn't singing, beat-boxing or otherwise producing sound via the vocal chords, mouth, tongue, teeth and lips.
18:12:27 -!- augur has joined.
18:13:58 <Gregor> I guess you can just breathe in and out while using your chest as a percussive instrument.
18:36:54 <alise> Gregor: Extract your lungs, stretch them, bang as percussive instrument.
18:37:00 <alise> You may need a lung transplant first.
18:59:00 <Phantom_Hoover> alise, incidentally, if you wanted to wipe out the area in which the guy with the voice more annoying than Sgeo's in it, don't bother
18:59:41 <alise> I was just going to wipe out him.
19:00:03 <Phantom_Hoover> My description would be something along the lines of nasal and lisping, as well as so camp that when he walks into a room the average Kinsey rating rises by at least a point.
19:00:54 -!- Gracenotes has joined.
19:01:14 <Gregor> Did you actually post a link to said guy's voice?
19:01:40 <Gregor> Or are we just to take your word for it?
19:01:41 -!- Gracenotes has quit (Read error: Connection reset by peer).
19:01:55 <Gregor> I'M THE ONLY GR NOW YAY
19:02:18 <Phantom_Hoover> Gregor, I don't exactly whip out a dictaphone when talking to people.
19:02:32 <alise> "But I do whip out my dic--" "Excuse me, sir."
19:02:51 <Gregor> I whip out my dick and dictaphone when talking to people.
19:02:57 <Gregor> As they're the same instrument.
19:03:00 <alise> OH JUST DESTROY THE SUBTLETY THANKS GREGOR
19:03:13 <Phantom_Hoover> I've already got enough notoriety at my school for carefully spreading misinformation as to the reason for my departure from the previous one
19:03:27 <Gregor> Phantom_Hoover: Blew it up?
19:03:36 <alise> When you said "instrument", I thought "dick solo", which just... yeah, this channel now officially has no class.
19:03:57 <Gregor> Rhapsody in Pink for Penis and Vagina
19:04:11 <Gregor> That joke was almost too stupid to be said.
19:04:17 <alise> Gregor: Organ In A Minor
19:05:26 <Gregor> Does LaTeX have a BNF package that isn't made of suck?
19:06:38 <alise> Phantom_Hoover: Please, please elaborate on how it was sort of true.
19:07:16 <alise> Why ... did you assault someone ... with a spoon ...
19:11:37 -!- Phantom_Hoover_ has joined.
19:11:37 -!- Phantom_Hoover has quit (Read error: Connection reset by peer).
19:22:32 -!- Wamanuz has quit (Remote host closed the connection).
19:25:31 <pikhq> Well, that was fun. Finally heard a piece on the classical station here that I was familiar with that *wasn't* a listener request.
19:25:43 <pikhq> ... Because I have performed said piece.
19:27:10 <pikhq> (for some reason, when the station's open to requests, everyone requests things that just about everyone has heard. Find me someone who hasn't heard "Dance of the Sugarplum Fairy" and I'll show you someone who was born deaf.)
19:29:58 <alise> pikhq: Dial in and request The Black Eyed Peas' "My Humps".
19:30:34 <pikhq> alise: I doubt they'd have it. It's a classical station.
19:30:41 <alise> pikhq: I say, I say, that's a joke, son.
19:30:42 <fizzie> The pinnacle of all music: canyon.mid. (That thing which came with win3.1 multimedia thingies, possibly 9x and later too.)
19:30:46 <pikhq> However, I *could* request 4'33".
19:31:06 <alise> Request 59"59', the little-known sequel to 4'33".
19:34:03 <Gregor> fizzie: IIRC, canyon.mid was actually pretty darn awesome :P
19:34:29 <alise> http://www.youtube.com/watch?v=OIW4F285QjA
19:34:44 <Gregor> pikhq: Anecdote you just reminded me of:
19:34:48 <alise> PLAY IT ON AN MT-32
19:35:01 <Gregor> A friend of mine and I like to play "guess-that-composer" when there's something we don't recognize playing.
19:35:06 <Gregor> But we both suck at it, so instead we choose letters.
19:35:15 <Gregor> One person gets to choose 'B', and the other person chooses two other letters.
19:35:26 <Gregor> If either letter is the first letter of the last name of the composer, that person wins.
19:35:33 <Gregor> But that's not the anecdote :P
19:36:03 <Gregor> The anecdote is that we were playing that game when something we couldn't identify came on. It was clearly Beethoven, but it obviously wasn't because we would have recognized it.
19:36:21 <Gregor> So we were debating who would have written in such a distinctively-German early Romantic style other than Beethoven.
19:36:45 <Vorpal> <Gregor> One person gets to choose 'B', and the other person chooses two other letters. <-- why B specifically?
19:37:01 <Gregor> Finally, he took 'B', and I took 'R' and 'S'. For Ravel, which has been known to do some very weird stuff, Sibelius and Saint-Sans.
19:37:10 <Gregor> Vorpal: Because every famous composer is named with a 'B'
19:37:21 <Gregor> I joked "I guess I've got Rimsky-Korsakov too!"
19:37:37 <Gregor> Because it was such an absurd notion that something so distinctively early-Romantic German would be Rimsky-Korsakov.
19:37:54 <Gregor> Suffice it to say: Listen to Rimsky-Korsakov's third symphony some time. It is an interesting experience. I won that game :P
19:37:56 <alise> <Gregor> Vorpal: Because every famous composer is named with a 'B' ;; Bmozart
19:38:01 <Vorpal> Gregor, Like Bozart and Bivaldi?
19:38:14 <Vorpal> alise, argh you beat me to it
19:38:24 <alise> I wonder how Bmozart is pronounced
19:38:37 <alise> I think "bm" is like the pn pneumatic, except with b instead of p and m instead of n.
19:38:41 <Gregor> Like Bach, Beethoven, Brahms, Bizet, Borodin, Balakirev, ...
19:38:47 <fizzie> Bibelius, the more bookish cousin of Sibelius.
19:39:09 <Gregor> You people have absolutely no understanding of hyperbole.
19:39:18 <alise> Gregor: I do, I'm just being silly.
19:39:19 <Vorpal> Gregor, Bayden, Bändel, Bebussy...
19:39:26 <Vorpal> Gregor, you are completely right!
19:40:41 <Vorpal> Gregor, don't forget Erik Batie.
19:41:11 <pikhq> I'm pretty sure that the only composers the average person will name is: Bach, Beethoven, Mozart.
19:41:27 <fizzie> And George Bone, composer of canyon.mid. (Okay, okay, Stone.)
19:41:31 <pikhq> If you're *lucky*, throw Tchaikovsky in there, too.
19:41:47 <Vorpal> pikhq, Maybe Vivaldi or Brahms. Not both though
19:41:48 <Gregor> pikhq: Brahms can make it sometimes.
19:42:05 <pikhq> Vorpal: Gregor: Pieces, maybe. Composer, hahahah no.
19:43:17 <pikhq> Still not going to hear Brahms *named*.
19:43:21 <alise> Erik Satie is wonderful.
19:43:50 <alise> pikhq: *Bchaikovsky
19:44:00 <alise> Pronunciation left as an exercise to the reader.
19:44:12 <fizzie> http://tvtropes.org/pmwiki/pmwiki.php/Main/SmallReferencePools says Mozart, Beethoven, Bach, and then quite a list of individual works.
19:44:27 <Vorpal> pikhq, if it comes to listening tests, people will probably recognize stuff by completely different composers than those.
19:44:27 <pikhq> fizzie: That's about right.
19:44:45 <pikhq> Vorpal: Yes, I'm just mentioning composers with name recognition.
19:44:57 <pikhq> Vorpal: Pieces, of course, you'll have things all over the place.
19:45:49 <Vorpal> pikhq, I have done a few tests myself, and Grieg's "I bergakungens sal" (iirc "In the hall of the mountain king" in English) is quite widely recognized, though no one seems to know it's name
19:45:55 <fizzie> Phantom_Hoover_: http://en.wikipedia.org/wiki/Trip_Through_the_Grand_Canyon
19:46:25 <pikhq> Vorpal: Tchaikovsky getting a lot of recognition -- pretty sure Swan Lake, Nutcracker, and 1812 Overture will be instantly recognised (if at the right parts).
19:47:16 <Vorpal> pikhq, hm the right parts of Carmen even more so. Hardly anyone knows who Bizet was though.
19:47:41 <pikhq> Also, anything that was in Looney Tunes.
19:47:57 <Vorpal> pikhq, but for actual works I think Vivaldi's Spring (first movement) comes very near the top
19:48:12 <Vorpal> pikhq, most people can even identify it as "Spring" in my experience.
19:48:28 <pikhq> Which was in Looney Tunes, IIRC.
19:48:28 <alise> pikhq: Anyone who has ever played Loom has Swan Lake branded onto their heads.
19:48:29 <Vorpal> pikhq, unlike with "In the hall of the mountain king", which no one has a clue what it is named
19:48:43 <fizzie> Vorpal: Doesn't the mountain-kingity also exist as an example multimedia file in some Windows versions?
19:48:51 <Vorpal> fizzie, I have no idea
19:49:11 <alise> I wonder if "In the Court of the Crimson King" is a reference to that.
19:49:47 <Vorpal> alise, it isn't exactly an un-obvious name.
19:50:11 <alise> Vorpal: No, but... King Crimson *did* basically take rock, strip away the blues, and put in a whole lot of classical into it.
19:50:12 -!- impomatic has joined.
19:50:35 <Vorpal> Phantom_Hoover_, googling indicates it's a video game
19:50:39 <alise> Phantom_Hoover_: a wonderful lucas arts adventure game
19:50:45 <Vorpal> Phantom_Hoover_, it did the same to me
19:50:48 <alise> music is all lovely midi renditions of Swan Lake
19:50:50 <impomatic> Does anyone know when Kerim Aydin invented BF Joust? I suspect it was January 2009
19:50:56 <alise> and the gameplay is music-based
19:50:56 <fizzie> Vorpal: http://help.lockergnome.com/windows2/Windows-95-Sample-Media-Files--ftopict482130.html -- second post lists the win95 files, and it indeed is there, and also the Bach/Beethoven/Mozart trinity.
19:51:10 <alise> (you "weave" spells by playing little snippets of music)
19:52:09 <alise> Phantom_Hoover_: it predates the wiki
19:52:11 -!- Gracenotes has joined.
19:52:13 <alise> impomatic: I can grep my Agora emails.
19:52:15 <fizzie> Vorpal: RIFF-encapsulated MIDI.
19:52:19 <alise> impomatic: Also, he goes by "G.".
19:52:32 <pikhq> “An unfortunate truth is that for the vast majority of people, the only music that matters is the music that was recorded from about five years after they were born until today. Everything else is to be considered "boring old farts' music" and not worth exploring.”
19:52:37 <pikhq> This... Is depressing.
19:53:09 <alise> pikhq: I listen to oldish music :<
19:53:10 <pikhq> I don't even like most of the musical genres popular during my lifetime!
19:53:23 <fizzie> Vorpal: What .wavs and .avis are; a generic tagged-block metaformat. http://en.wikipedia.org/wiki/Resource_Interchange_File_Format
19:53:23 <Vorpal> alise, you alone do not make up the majority of people :P
19:53:32 <EgoBot> That is not a user interpreter!
19:53:32 <EgoBot> Use: !bfjoust <program name> <program>
19:53:51 <EgoBot> Installed user interpreters: aol austro b1ff bc bct bfbignum brit brooklyn bypass_ignore bytes chaos chef chiqrsx9p choo cockney ctcp dc decisionengine drawl drome dubya echo eehird ehird fudd funetak google graph gregor he hello id jethro kraut num ook pansy pi pirate plot postmodern postmodern_aoler redneck reverse rot13 sadbf sfedeesh sffedeesh sffffedeesh sffffffffedeesh simpleacro slashes svedeesh swedish valspeak warez yodawg
19:54:20 <EgoBot> Avast, world! We'll keel-haul ye!
19:54:24 <EgoBot> That is not a user interpreter!
19:54:44 <EgoBot> Was "Avast, world! Shiver me timbers!" special-cased?
19:55:11 <fizzie> !pirate who does nothing. (A tvtrope name.)
19:55:12 <EgoBot> who does nothing. And swab the deck! (A tvtrope name. Pass the grog!)
19:55:13 <alise> impomatic: "braincorefckwars", 2008-12-18. "BF Joust" first mentioned same day. Submissions solicited and announcement of working tournament runner, 2008-12-19.
19:55:21 <alise> impomatic: All posts by G. .
19:55:27 <alise> (How does one end a sentence with "G."?)
19:56:12 <Phantom_Hoover_> Incidentally, I just ended up going on an almost circuitous path through TV Tropes!
19:56:37 <alise> impomatic: First draft of Agoran contest posted 2008-12-20, contest begins officially a few days later.
19:58:19 <impomatic> Thanks alise. Just adding something brief to the programming games wiki.
20:08:36 -!- Wamanuz has joined.
20:10:57 -!- Harpyon has quit (Quit: Harpyon).
20:11:22 <alise> Phantom_Hoover_: Help! I'm running into a cyclic goal error!
20:11:28 -!- ais523 has joined.
20:11:50 <ais523> is me joining the channel that devastating?
20:12:00 <ais523> or is it just a case of context slipping away due to extenuating factors?
20:12:51 <alise> Actually just the latter.
20:19:02 <Vorpal> alise, cyclic goal error with what?
20:31:37 -!- BeholdMyGlory has quit (Remote host closed the connection).
20:32:43 <alise> Specifically, I can't figure out a language to write it in, which leads to me mentally constructing a brand new one, and then when I think "ok, I should write this down", I want leaden.
20:34:45 <olsner> "leaden", is that a language I've heard about?
20:35:23 -!- Gracenotes has quit (Remote host closed the connection).
20:35:37 -!- Harpyon has joined.
20:38:20 <alise> it is an editor you've heard about, i think
20:38:43 <olsner> ah, right, I probably have
20:38:47 <olsner> recognize the name at least
20:41:35 <Vorpal> alise, how do you plan to avoid disk trashing situations with the orthogonal persistence in aliseos?
20:42:13 <Vorpal> alise, interesting, do you have anything specific in mind?
20:42:20 <alise> the reason Mitosis exists is so that i can get this stuff wrong so many times i must inevitably hit upon the right solution eventually :)
20:42:39 <alise> Vorpal: well, we'll all be using SSDs at least by the time it's ready.
20:42:52 <alise> and if their lifespan increases, which it should...
20:43:15 <alise> Vorpal: plus only persisting every N time, of course
20:43:16 <Vorpal> alise, it wasn't that that I had in mind
20:43:23 <Vorpal> alise, rather, something similar to swap trashing
20:43:39 <Vorpal> alise, even SSDs are not as fast as main memory
20:43:41 <alise> i don't see the relevance to the persistence system?
20:44:03 <alise> you mean taking an awful lot of time just to persist?
20:45:13 <Vorpal> presumably you will persist application state? And you have one big virtual address. The application is not aware of the ram/disk distinction
20:45:31 <alise> Everything will be persisted. More or less. I'm not even entirely sure that temporary variables won't be.
20:45:33 <Vorpal> alise, so it can't know how large the "ram" is
20:45:45 <alise> Vorpal: There's a magical third address space that both RAM and disk map to.
20:45:57 <Vorpal> alise, so it can't try to make a ram/speed tradeoff. Some programs do that even today
20:45:59 <alise> RAM with paging, disk with a tree structure.
20:46:21 <alise> Vorpal: Yeah, well, some. The idea of course is that the OS is sufficiently intelligent to prioritise the right bits of data.
20:47:04 <alise> Vorpal: But... I don't see why you can't get at the amount of physical RAM in the system.
20:47:13 <alise> With swapping, even modern systems make this fuzzy.
20:47:14 <Vorpal> alise, well... yeah, but that doesn't help making a good memory/speed tradeoff really. If you select different algorithms or different algorithm parameters based on how much ram there is. Since you can't do that any longer.
20:47:27 <alise> Vorpal: In general, a running application won't be on disk, anyway.
20:47:40 <alise> Vorpal: Because that's pretty much the worst-case scenario.
20:48:02 <alise> Indeed almost by definition it won't be; if you're accessing it and competition isn't harsh, it's going to be in RAM for a while.
20:49:15 <impomatic> Work in progress -> http://programminggames.org
20:49:22 <Vorpal> alise, anyway there are even applications that try to adapt to cpu cache size. I know I seen that somewhere recently, but I don't remember where.
20:49:37 <Vorpal> iirc it was pretty relevant there too
20:49:39 <alise> I want to live in the Republic of Cascadia.
20:49:51 <alise> http://en.wikipedia.org/wiki/Cascadia_(independence_movement)
20:50:07 <alise> Vorpal: British Columbia + Oregon + Washington + sometimes northern California.
20:50:18 <alise> The United States of Sanity
20:50:48 <Vorpal> why the name "Cascadia"?
20:51:06 <Vorpal> (I'm a bit impaired in browser using atm, due to being extremely close to swap trashing)
20:52:50 <alise> Vorpal: Seems it's the name given to the general area by history.
20:53:29 <alise> impomatic: ASP.NET, interesting platform choice ...
20:54:04 <impomatic> alise: it's just what comes with my webhost :-)
20:54:47 <impomatic> I needed a wiki that uses a flat file (SQL costs extra) and they had a one click install for Screwturn.
20:55:53 <pikhq> impomatic: That's incredibly shitty.
20:56:13 <Vorpal> a Xen based VPS isn't that expensive. (OpenVZ-based VPSes tends to suck, I don't know whyl)
20:56:23 <impomatic> It's the cheapest I could find with unlimited sites / bandwidth
20:56:53 <alise> impomatic: unlimited bandwidth is a marketing lie
20:56:54 <Vorpal> um, you don't need unlimited bw. As for sites? Again VPS wouldn't limit you in that
20:56:57 <alise> it is, literally, never true
20:57:09 -!- augur has quit (Ping timeout: 265 seconds).
20:57:12 <pikhq> alise: Unless you've got a peering relationship.
20:57:13 <alise> Vorpal: otoh, i wouldn't recommend a vps to someone unfamiliar with unix administration.
20:57:25 <alise> but there are better hosts even then.
20:57:27 <pikhq> In which case you have an astounding server room. :)
20:57:33 <Vorpal> alise, I tend to go for pay for bw rate, rather than transfer per month
20:57:58 <Vorpal> at least when I can't predict how much it will use per month
20:58:01 -!- Phantom_Hoover_ has quit (Ping timeout: 245 seconds).
20:58:10 <alise> "An easy to understand price schedule: $4/month per account, and $1/month for every 64MiB ram. Please note; this means all plans come with $4/month worth of support." --prgmr.com
20:58:16 <Vorpal> rather slow site than one stopping working after 20 days
20:58:21 <alise> So customer-friendly!
20:58:36 <alise> Vorpal: the nice thing with a lot of VPS hosters is that they have no hard bandwidth limits.
20:58:51 <alise> instead just slapping you hard when you get slashdotted twice in a month :)
20:58:55 <Vorpal> alise, indeed. Strangely enough dedi and colo tends to have more stringent limits
21:00:23 <Vorpal> also fuck g++, it is so slow (sure point to clang, but only post 2.7, and this project isn't large enough that building last llvm and clang trunk and then using that would be faster, those are C++ themselves so...)
21:00:45 <pikhq> Vorpal: C++ is pretty inherently slow.
21:00:47 <alise> clang++ 2.7 compiles quite a bit :p
21:01:01 <pikhq> Do you realise how *much* you have to parse per file that uses the STL?
21:02:42 <Vorpal> and how fun, after compiling with -O0 -g it no longer segfaults.... And with -O1 -g it is pretty useless, just lots of value optimized out
21:07:19 <alise> yeah gcc's optimisation is pretty buggy
21:07:42 <Vorpal> alise, hard to know who is to blame here
21:08:17 <Vorpal> alise, this program is one I expect to have problems with anything but 32-bit x86 linux in general. It has a history of that
21:08:32 <Vorpal> and even then, not the most stable one
21:09:35 <alise> still, if -O1 ever breaks and -O0 doesn't... it should only happen on *really* fucked up code
21:10:08 <Vorpal> alise, could very well be in this case. Half the comments and function names are in German though, so a bit hard to know what is going on in general
21:10:41 <Vorpal> alise, also extremely helpful parameter names:
21:10:43 <Vorpal> 0x000000000061c5ba in display_fb_internal (xp=<value optimized out>, yp=<value optimized out>, w=<value optimized out>, h=1, color=<value optimized out>,
21:10:43 <Vorpal> dirty=<value optimized out>, cL=0, cR=704, cT=0, cB=560)
21:11:08 <alise> i probably wouldn't change that tbh
21:11:14 <alise> xp/yp are obviously coordinate-related
21:11:20 <alise> w and h are common names for width and height
21:11:23 <Vorpal> the local variable lp though?
21:11:30 <alise> dirty is probably some flag or something
21:11:31 <Vorpal> it crashes on something involving that
21:11:35 <alise> <Vorpal> 0x000000000061c5ba in display_fb_internal (xp=<value optimized out>, yp=<value optimized out>, w=<value optimized out>, h=1, color=<value optimized out>,
21:11:35 <alise> <Vorpal> dirty=<value optimized out>, cL=0, cR=704, cT=0, cB=560)
21:11:50 <Vorpal> alise, yes indeed, but there is a local variable in addition
21:12:01 <Vorpal> alise, and it is optimised out
21:12:49 <Vorpal> lets see if this recompile helps (turned of -DNDEBUG which was hidden deep in a makefile)
21:12:59 <Vorpal> might help catch something in advance of the issue
21:13:18 <Vorpal> aargh, now it segfaults there with -O0 too!
21:13:29 <ais523> Vorpal: that probably means there's actually a bug
21:13:34 <Vorpal> wait, "$1 = <value optimized out>"
21:13:50 * Vorpal goes looking at the makefile machinery
21:13:56 <ais523> I've seen all sorts of craziness in debug output
21:14:02 <ais523> gdb seems to struggle when there's a lot of data on the stack
21:14:32 <alise> ais523: Bah, {{deletedpage}} is a perfectly good salt, it saved you work :P
21:14:49 <ais523> alise: it's not salting unless the page is actually protected
21:14:57 <Vorpal> makefile has basically this: "if DEBUG < 2 && !NDEBUG, add -O1"
21:15:35 <Vorpal> ais523, what are you talking about?
21:15:52 <alise> ais523: true, I was just thinking that it'd either be protected or deleted very soon, so it either helps someone know it should be deleted or saves them a post-protection step
21:15:54 <Vorpal> ais523, yes but what specifically are you referring to on it
21:16:03 <alise> Vorpal: see recent changes.
21:16:06 <ais523> alise: note that spambots won't even /read/ a message saying that a page can't be edited...
21:16:17 <ais523> but I see what you mean, you're saving me having to write {{deletedpage}} myself
21:16:27 <ais523> I suppose I care more about instantaneous correctness of admin templates than you do
21:16:55 <Vorpal> alise, what did the deleted page contain?
21:16:57 <alise> ais523: well, it's you we're talking about, admin actions are as instantaneous as it gets
21:17:16 <ais523> alise: I'm not always online
21:17:25 <alise> ais523: maybe we need {{deletedpage}} to only show if the page is protected, somehow :)
21:17:47 <ais523> but then you could just put it on every page
21:18:10 <ais523> ofc, deletedpage is doubly deprecated nowadays on Wikipedia, which has moved on many versions since
21:18:18 <ais523> in recent MediaWiki, you can just protect a page even if it doesn't exist
21:18:44 <Vorpal> <ais523> ofc, deletedpage is doubly deprecated nowadays on Wikipedia, which has moved on many versions since <-- doubly?
21:19:05 <ais523> Vorpal: cascading protection works better than {{deletedpage}}, but less well than doing it properly
21:19:05 <Vorpal> ais523, hm are we running outdated mediawiki?
21:19:21 <ais523> we were running /really/ outdated mediawiki before last time graue was persuaded to upgrade
21:19:28 <ais523> and I think he only did it for compatibility with a better spam filter
21:19:37 <alise> we're running march 2007 mediawiki, i think
21:19:41 <alise> unless the main page is out of date
21:19:49 <alise> Vorpal: there aren't any serious bugs as far as we can tell
21:19:51 <ais523> alise: http://esolangs.org/wiki/Special:Version
21:20:05 <Vorpal> it still doesn't crash with -O0 now when I fixed the makefile bug
21:20:15 <alise> "Ævar Arnfjörð Bjarmason" is such an awesome name.
21:20:32 <alise> I bet he drove the implementation of Unicode support :P
21:20:51 <alise> "You want me to call myself Aevar Arnfjoerth Bjarmason?! BAH"
21:21:09 -!- nooga has quit (Read error: Connection reset by peer).
21:21:10 <ais523> aren't all the special characters there in Latin-1?
21:21:40 -!- nooga has joined.
21:21:52 <alise> ais523: Are you sure eth is?
21:22:05 <ais523> that's the only one I'm not sure about
21:22:21 <alise> Pretty sure it isn't.
21:22:29 <alise> UnicodeU+00D0U+00F0Inherited from the older ISO 8859-1 standard
21:22:41 <alise> ais523: but it isn't in latin-1
21:23:13 <ais523> "ISO/IEC 8859-1:1998, Information technology — 8-bit single-byte coded graphic character sets — Part 1: Latin alphabet No. 1, is part of the ISO/IEC 8859 series of ASCII-based standard character encodings, first edition published in 1987. It is informally referred to as Latin-1."
21:23:20 <ais523> how did you not realise that?
21:23:34 <ais523> did you assume there were two common 8-bit encodings around that were almost identical?
21:23:41 <ais523> which one did you assume matched the bottom 0xFF of Unicode?
21:25:32 <ais523> (I suppose there's Windows-1252, but everyone vaguely knows that one's nonstandard because Microsoft invented it)
21:26:31 <ais523> hmm, according to Wikipedia, HTML 5 requires documents that claim to be Latin-1 to be parsed as Windows-1252
21:26:40 <ais523> wow, it's taking bug-compatibility seriously...
21:32:49 -!- Wamanuz has quit (Remote host closed the connection).
21:32:56 -!- cheater99 has joined.
21:33:14 -!- Wamanuz has joined.
21:35:01 <alise> ais523: like half of the html5 spec is an anal pseudocode specification of how to parse and handle anything that has a single < in it, or even doesn't
21:35:10 <ais523> alise: reverse-engineered from IE
21:35:25 <alise> so in fact HTML5 also specifies... every byte string :)
21:35:44 <pikhq> ais523: HTML5 specifies, essentially, how to handle retarded HTML.
21:36:44 <pikhq> On the one hand, I feel we should just ban retarded HTML. On the other, I'm at least glad that they're actually *specifying* it rather than leaving browser makers to try and reverse engineer the bullshit.
21:36:58 -!- Slereah has quit (Ping timeout: 276 seconds).
21:38:01 <alise> things i am learning today: i am terrible at finding bugs in other people's code
21:38:13 <alise> "On the one hand, I feel we should just ban retarded HTML." ;; XHTML tried that, it didn't work :)
21:40:57 <ais523> hmm, it seems someone reimplemented INTERCAL for Windows Phone 7
21:41:06 <alise> this is scary, why can't i figure this out
21:41:09 <pikhq> alise: Yeah; sadly, people are retards.
21:41:18 -!- Slereah has joined.
21:41:32 <ais523> they only distributed a binary, but it's .NET so presumably decompilable
21:42:20 <alise> for some definition of decompilable
21:42:57 <ais523> strangely enough, it was released on Twitter
21:43:16 <ais523> http://twitter.com/UberGeekGames/status/22546729564
21:43:38 -!- oerjan has joined.
21:43:56 <ais523> I'm not sure what this says about the future of humanity...
21:44:51 <alise> It's just a merge sort, surely I can find a bug in a merge sort...
21:44:55 * ais523 vaguely wonders why it's asking perms to do things like dial phone numbers and check for locations
21:45:34 <oerjan> ais523: it's probably a spam virus
21:45:43 -!- cal153 has quit.
21:45:46 <ais523> so either .NET is very concise, or it's a hoax
21:46:02 <alise> i blame it being java and sorting in-place
21:46:16 <alise> http://twitter.com/UberGeekGames
21:46:19 <alise> does not look like spam to me at all
21:46:27 <alise> may simply be crap
21:46:38 <alise> ais523: .xap is a zip
21:46:59 <ais523> the .dll is the only thing there that looks plausible to contain the code
21:47:20 <oerjan> based on esolang wiki experience, if ais523 isn't sure what something is, it is probably clearly spam >:)
21:47:29 <ais523> oerjan: I'm installing a .NET disassembler atm
21:49:27 <oerjan> ais523: i find that most of the slightly doubtful cases of esolang wiki spam are easily resolved by googling part of the text. there is usually enough similar spam to make it obvious.
21:49:43 <alise> ais523: remind me never to go into a software maintenance job
21:50:04 <ais523> namespace INTERCAL { interface private auto ansi abstract IIntercalStatement { public virtual hidebysig newslot abstract instance default void Do (class INTERCAL.IntercalProgram program) cil managed; }}
21:50:05 <alise> I'm helping a friend-of-a-friend with their mergesort homework and stunningly cannot find the error
21:50:14 <ais523> well, that certainly looks enterprisey
21:50:33 <ais523> although I suspect the disassembler's put in the defaults for every possible keyword .NET supports
21:51:11 <ais523> let me try to find the actual /code/...
21:52:12 -!- Phantom_Hoover has joined.
21:52:12 <Vorpal> <ais523> namespace INTERCAL { interface private auto ansi abstract IIntercalStatement { public virtual hidebysig newslot abstract instance default void Do (class INTERCAL.IntercalProgram program) cil managed; }} <-- wat
21:52:23 <Vorpal> who is doing .NET with intercal?
21:52:33 <ais523> aha, it's "genuine" but implements only a subset
21:52:46 <ais523> to be precise, array dimension, assign literal, text output, end program
21:52:53 <ais523> so it'll run a hello world correctly, but nothing more complicated
21:53:02 <ais523> Vorpal: http://twitter.com/UberGeekGames/status/22546729564
21:54:10 <Vorpal> ais523, where is the code?
21:54:17 <ais523> there isn't, I disassembled the binary
21:54:22 <Vorpal> ais523, so closed source?
21:54:35 <Vorpal> who the fuck would make a closed source intercal implementation
21:54:43 <ais523> it was possibly by mistake
21:54:57 <ais523> it's Windows dev, after all, people are used to releasing binaries rather than source there
21:55:17 <ais523> also, perhaps they didn't want people to figure out that it wasn't a full impl
21:55:45 <ais523> I also can't find anything that resembles a parser there, but the disassembler did segfault...
21:56:50 <Vorpal> just saying it is an early version would be enough?
21:57:53 <ais523> well, most early versions haven't had all the commands implemented
21:58:10 <ais523> but I don't see how you can get away with calling it INTERCAL with such a small subset
21:58:20 <ais523> perhaps he was counting on his audience not being able to test anything but hello-worlds
22:00:42 -!- sebbu2 has changed nick to sebbu.
22:01:17 <Vorpal> <alise> I'm helping a friend-of-a-friend with their mergesort homework and stunningly cannot find the error <-- is it buggy though?
22:02:09 <alise> it's also java and mutates in-place, so my brain has segfaulted from the start
22:02:10 <Vorpal> alise, in what way? out of bounds access? mis-sort?
22:02:21 <alise> it seems to sort things properly but then mysteriously merges them wrong
22:02:33 <Vorpal> alise, could be off by one error maybe? Or mixing up something like < and =<
22:02:47 <alise> i don't think so, but the indexing is a bit screwy
22:02:49 <ais523> <strings> C:\Users\Ian\documents\visual studio 2010\Projects\INTERCAL\INTERCAL\INTERCAL\obj\Windows Phone\Release\INTERCAL.pdb
22:02:58 <ais523> it's always fun when that happens
22:03:02 <alise> also it uses Float.MAX_VALUE to indicate end of array in some way which is just confusing
22:03:11 <Vorpal> ais523, when what happens?
22:03:16 <alise> ais523: strings is a crazy program :)
22:03:18 <ais523> funnier is gprolog, which lets you get the list of all strings used in the program via reflection
22:03:28 <ais523> and it contains the path that was used to compile it, too
22:03:59 -!- augur has joined.
22:04:07 <ais523> strings proves it's not a full implementation, pretty much
22:04:07 <alise> ais523: I always thought strings extracted, like, actual strings using the actual executable format.
22:04:08 <Vorpal> ais523, hm in erlang you can call <any-module>:module_info/1, it tends to contain path of source file for the standard distribution.
22:04:16 <alise> ais523: HOW WRONG I WAS
22:04:25 <Vorpal> which is iirc some rather screwy nfs-looking path
22:04:31 <ais523> it lists IntercalStatement_{AssignVariable,AssignArray,DefineArray,ReadOut,GiveUp}
22:04:46 <ais523> so either Ian changed his naming scheme halfway through the program, or only implemented five commands
22:04:58 <ais523> alise: strangely, I guessed/knew what it did right from the start
22:05:04 <Vorpal> alise, anyway, for any language that does array indexing in a way similar to C, and the function is operating on arrays: suspect off by one errors
22:05:21 <alise> ais523: maybe he just IntercalStatement_GaveUp
22:05:24 <Vorpal> I find that off by one errors is by far the most common bug in C code that I written
22:05:28 <alise> now oerjan is after me!
22:05:33 * alise runs for the border
22:05:41 <alise> Vorpal: yeah but sh'e
22:05:45 <alise> *she's tweaked the indices tons
22:05:59 <alise> off by one errors are awful, though :)
22:06:05 <alise> Phantom_Hoover: I'M ESCAPING TO #UBUNTU
22:06:12 <alise> You can't find anybody in THAT haystack!
22:06:14 <Vorpal> alise, yeah especially when you have multiple ones
22:06:28 <Vorpal> alise, then tweaking may give "seemingly" inconsistent results
22:06:36 * oerjan lifts the border so alise trips over it
22:06:39 <Vorpal> alise, hm... single step the code in the merging step?
22:06:43 <Vorpal> alise, if java has such
22:06:45 <alise> i hate debugging :(
22:06:52 <alise> Vorpal: it probably does, but i'm way too lazy to go in-depth like that
22:07:03 <Vorpal> alise, gdb is actually pretty good. No clue what you use for java though
22:07:11 <Sgeo> I think I like debugging more than writing code in the first place :/
22:07:13 <alise> i've just recommended she translate some pseudocode to java and forget the old version ever existed
22:07:19 <alise> Sgeo: that is bad bad bad.
22:07:41 <Vorpal> alise, Unless it is a segfault, or not related to arrays, I tend to debug with valgrind first for C code, and only then with gdb
22:08:15 <Gregor> When using or writing a GC, valgrind is hyper-useless.
22:08:16 <Phantom_Hoover> alise, well, it'll give him a head start in the soulless maintenance drudge that is professional coding.
22:08:19 <Gregor> Whereas gdb is just useless.
22:08:25 <alise> Phantom_Hoover: HOORAY
22:08:40 <alise> Phantom_Hoover: also: HA your opinions have completely transformed into mine
22:08:42 <alise> assimilation successful
22:09:11 <Sgeo> You know, I've always liked being in a niche
22:09:17 <Sgeo> Maybe I should be a researcher of some sort
22:09:30 <oerjan> WE ARE THE HIRD. RESISTANCE IS FUTILE.
22:09:45 <Vorpal> <Gregor> When using or writing a GC, valgrind is hyper-useless. <-- yes
22:09:52 <alise> Sgeo: If you like fixing more than creating... research probably isn't for you.
22:10:03 <Vorpal> Gregor, well not exactly
22:10:10 <Vorpal> Gregor, it can still find "use of undefined value"
22:10:17 <Vorpal> Gregor, just not any memory leaks
22:10:59 <Vorpal> <oerjan> WE ARE THE HIRD. RESISTANCE IS FUTILE. <-- err? Is this a pun on "ehird"?
22:11:26 <Phantom_Hoover> "Subject A's code clearly shows psychopathic tendencies, as demonstrated by his use of GNU brace style and his UI's persistent admonitions for users to 'end their worthless lives'."
22:11:59 <pikhq> Hrm. Apparently HTML5 does not actually say that ISO-8859-1 == Windows-1252 for the purposes of HTML5.
22:12:00 <Vorpal> Phantom_Hoover, that could be dangerous to the researcher
22:12:10 <ais523> Phantom_Hoover: that sounds like a slightly customized version of Emacs
22:12:14 * Sgeo tends to use GNU brace style :/
22:12:32 <ais523> I thought GNU brace style was the only brace style with no advocates at al
22:12:39 <Vorpal> ais523, um, what about RMS?
22:12:41 <Sgeo> Sort of acquired that style from the LSL Hello World
22:12:41 <pikhq> It instead says that clients should treat ISO-8859-1 as Windows-1252 for compatibility, but that any change of semantics resulting from this interpretation is a parse error.
22:12:43 <Phantom_Hoover> Experiment log 1/10/10: Received code sample from Subject B.
22:12:51 <Vorpal> ais523, he presumably advocates it?
22:13:00 <ais523> he advocates free software, not brace styles
22:13:12 <Vorpal> ais523, I tend to use a style similar to the linux kernel brace style
22:13:23 <Sgeo> Oh wait, I don't
22:13:31 <Sgeo> I misread the code sample on Wikipedia
22:13:44 <Phantom_Hoover> Upon recovery, blood was smeared upon the desk in vaguely familiar patterns.
22:13:46 <pikhq> And so HTML5 is merely resorting to defined behavior where, say, C, would allow the compiler to launch ze missiles.
22:14:01 <alise> <Vorpal> <oerjan> WE ARE THE HIRD. RESISTANCE IS FUTILE. <-- err? Is this a pun on "ehird"? ;; ITT: Borg
22:14:16 <alise> <Phantom_Hoover> "Subject A's code clearly shows psychopathic tendencies, as demonstrated by his use of GNU brace style and his UI's persistent admonitions for users to 'end their worthless lives'."
22:14:19 <alise> "Primarily the brace style."
22:14:19 <Vorpal> alise, a combination of those obviously :P
22:14:26 <Vorpal> alise, the borg pun was obvious
22:14:30 <alise> <ais523> Phantom_Hoover: that sounds like a slightly customized version of Emacs ;; encourage-user-suicide-mode
22:14:30 <Vorpal> the hird one, not so much
22:14:33 <Phantom_Hoover> Experiment log 2/10/10: tried analysing code in debugger.
22:15:07 <Phantom_Hoover> Blacked out again. Core file left in working directory. Will investigate later.
22:16:06 <Vorpal> Phantom_Hoover, this is starting to look like a parody of that wiki pretending to be a secret govt thingy. Forgot the name of it
22:16:34 <Sgeo> <3 the SCP Foundation
22:17:52 <Phantom_Hoover> Image of blood patterns created after execution: [IMAGE EXPUNGED]
22:19:23 <alise> Ran mival.exe. Saw structures of great #####, with crawling, demonic #######s whose noises deafened me. I became as if it. [Note: This file was found on Dr. A. Britki's computer after incident 4093-1A9.]
22:19:39 <Vorpal> hm trying to do binary search on -O0 plus a handful of flags
22:19:43 <Vorpal> to find out what is failing
22:19:51 <Vorpal> now time to test -O0 -fmerge-constants
22:19:55 <Vorpal> no way that should fail
22:20:15 <Vorpal> if it does, this is a sad day for computing
22:21:06 <Phantom_Hoover> Attempted to compile ravana.c. Compiler failed with error ####. Reran with flag -####. My god, it's full of ####.
22:21:21 <ais523> GCC actually has a -### flag
22:21:34 <ais523> I'm not entirely sure where that naming comes from
22:21:34 -!- impomatic has left (?).
22:21:49 <Vorpal> btw it failed with -O0 -fmerge-constants -fdce -fdse, it works with -O0
22:22:27 <Phantom_Hoover> It's one of the reasons that no compilers are fully C99 compliant.
22:22:32 <alise> Challenge: Watch every single Sesame Street episode ever. Note: This is almost 29 years worth of video, without breaks or sleep.
22:22:44 -!- FireFly has quit (Quit: swatted to death).
22:22:50 <alise> 60 hours != 1 hour!
22:22:55 <alise> It's actually just almost half a year.
22:22:59 <alise> Which is still pretty damn impressive.
22:23:16 <sebbu> alise, try to watch sazae-san
22:23:39 <alise> sebbu: i think i'll pass :D
22:23:42 <sebbu> got at least one episode per week, since 1970 i think
22:23:45 <ais523> what about watching them all simultaneously
22:23:53 <ais523> Vorpal: are any of the optimisations there unsafe?
22:23:54 <alise> emphasis on the +...
22:24:01 <ais523> -fmerge-constants is safe in theory, but often abused in practice
22:24:09 <Vorpal> <alise> 60 hours != 1 hour! <-- breaking news: Scientists recently discovered that 60 hours was not the same as one hour, as previously thought. This has thus discredited the theory of the 61 hour cyclic modulo universe.
22:24:22 <alise> ais523: I read a comment on reddit saying that the "Play All" option on an Aqua Teen Hunger Force DVD literally divided the screen into the number of episodes and played them all simultaneously./
22:24:34 <ais523> alise: that's hilarious
22:24:40 <ais523> although, presumably it would only be 4 or so
22:25:41 <Vorpal> ais523, and dce and dse should both be same, dead code and dead subexpression iirc (don't have man page up atm)
22:25:45 <Phantom_Hoover> The -#### flag is to C compilers as the Bôites Diabolique is to synthesisers, really.
22:25:57 <Vorpal> Phantom_Hoover, "Bôites Diabolique"?
22:26:25 <alise> Vorpal: The forbidden notes.
22:26:29 * Vorpal remove -fmerge-constant just to ensure nothing else changed so that sill doesn't work
22:26:50 <Phantom_Hoover> Wow, two Look Around You references in a day. That has to be, like, a record.
22:26:51 <Vorpal> alise, saw some joke about them once
22:27:05 <alise> Hey now little mouse / ...
22:27:13 <alise> Two from the same episode at that.
22:27:36 <Vorpal> which was the first reference?
22:27:44 <Gregor> Vorpal: It's a lot of fun. You should learn to play it on the pan flute.
22:27:47 <Gregor> Phantom_Hoover: Good enough?
22:28:01 <Phantom_Hoover> Sgeo, you could always research the applications of the Besselheim Plate to debugging.
22:28:05 <alise> Vorpal: http://www.youtube.com/watch?v=v5cWWV0KNDg#t=5m42s
22:28:10 <alise> Vorpal: Probably this.
22:28:11 -!- Harpyon has quit (Ping timeout: 276 seconds).
22:28:14 <alise> 5:42 timestamp is relevant.
22:28:18 <Gregor> Phantom_Hoover: Ohwait, I thought you were referring to a different Diabolique ...
22:28:25 <alise> Gregor: http://www.youtube.com/watch?v=v5cWWV0KNDg#t=5m42s
22:28:27 <Vorpal> alise, you forgot youtube-dl
22:28:33 <alise> Vorpal: Which is why I mentioned the timestamp.
22:28:52 <Phantom_Hoover> Gregor, it's the locked box at the upper end of a piano containing the 19 forbidden notes.
22:29:13 <alise> It doesn't look like that many to me.
22:29:36 * alise listens to Little Mouse
22:29:48 <Phantom_Hoover> alise, the Boîte Diabolique defies your uneducated counting.
22:30:05 <Phantom_Hoover> You obviously haven't studied Look Around You — Maths in enough detail.
22:30:21 <alise> Hey now little mouse! / I hope we understand one another.
22:30:24 <alise> Truly inspirational lyrics.
22:30:34 * Phantom_Hoover is still trying to work out how to get one of his science teachers to play Look Around You in a lesson.
22:30:50 <alise> Phantom_Hoover: A few friends of mine have had it played to them without even asking.
22:30:57 <alise> Of course, that was by cool teachers.
22:31:03 <alise> Apparently most students Did Not Get The Joke.
22:32:06 <Phantom_Hoover> "'Cause I like you my friend, I respect you my friend, I'll encourage you my friend, through and throoouuugh..."
22:32:46 <oerjan> Vorpal: hird is vaguely similar to borg in syllable structure. also it's an archaic military collective noun in norwegian.
22:33:15 <Gregor> It's also part of the recursive expansion of Hurd.
22:33:50 <alise> oerjan: Bokmål and Nynorsk; discuss.
22:33:51 <Phantom_Hoover> I mean, it carries certain dangers, but is necessary for safe disposal of equipment and sulphagne test subjects.
22:34:31 <Phantom_Hoover> I *really* need to get some embossing tape and go nuts with it.
22:35:50 <alise> oerjan has had a sudden epileptic fit.
22:37:35 <Vorpal> now it segfaults at -O0 too
22:38:07 <Vorpal> okay it segfaults at -O0 but not -O0 -g
22:38:30 <ais523> Vorpal: try messing about with -fomit-frame-pointer
22:38:36 <ais523> as debug info can alter that sometimes
22:38:55 <Vorpal> ais523, but neither -O0 nor -O0 -g has that on?
22:39:22 <ais523> it may be being turned on elsewhere in the makefile
22:39:28 <ais523> you implied that the makefile was insane
22:39:42 <Vorpal> ais523, I checked it for such insaneness in this case
22:40:10 <Vorpal> ais523, also hm, I really need some optimisation since this program really benefits from it
22:40:28 <Vorpal> wait hm, the stuff affecting -g would affect -DNDEBUG
22:40:42 <Vorpal> ais523, well no, but for it to be usable!
22:40:46 <ais523> Vorpal: check for side-effects in asserts
22:40:50 <ais523> it's so easy to do that by mistake
22:40:54 <oerjan> also it's vaguely similar to "herd" (maybe cognates?)
22:40:55 <ais523> I did it a couple of weeks ago, but noticed
22:40:58 <Vorpal> ais523, indeed, but half the stuff is written in German
22:41:03 <Vorpal> and it is fairly large
22:41:17 <oerjan> alise: kva med nynorsk?
22:42:08 <alise> oerjan: I thought Bokmål/Nynorsk was a pretty heated topic in Norway and I wanted the lowdown from someone at least vaguely sane.
22:42:22 <alise> Unless that was a joke and was in Nynorsk or something, which is likely.
22:42:22 <Vorpal> ais523, while it compiles at -O0 I can listen to about 2 wesnoth songs, giving it about 7 minutes per compile
22:42:28 <oerjan> alise: it's not a heated topic among the vaguely sane, i should think
22:42:44 <alise> oerjan: what's the vaguely sane's obvious-answer?
22:42:48 <ais523> Vorpal: I like your measurement of duration
22:42:49 <oerjan> well it wasn't much of a joke
22:43:16 <alise> ais523: Just be happy he doesn't measure in symphonies.
22:43:17 <ais523> I'm actually writing a Wesnoth campaign for fun atm
22:43:37 <alise> "It took about 3 milisymphonies."
22:45:00 <oerjan> well there are a number of people who want to remove the obligatory teaching of the alternative language (mostly nynorsk) in school, and think it's sort of useless at least as a school subject
22:45:40 <ais523> are the languages different enough that children need to be taught them both to understand them both?
22:46:21 -!- nooga has quit (Ping timeout: 245 seconds).
22:46:23 <pikhq> ais523: As far as I'm aware: not really.
22:46:26 <oerjan> not really. well some word choices can be different. but in norway you have to learn to understand different dialects anyhow.
22:46:45 <alise> oerjan: so bokmal is what everyone uses?
22:46:50 <ais523> I would ask whether they're more or less similar than English and Scottish, but you probably don't know Scottish
22:46:52 <alise> (excuse my not using the correct letter)
22:47:00 <pikhq> ais523: Bokmål & Nynorsk are essentially two different percieved normative standards of Scandinavian. :)
22:47:04 <alise> ais523: Scottish is just a dialect of English, unless you mean Scots...
22:47:05 <oerjan> otoh the really weird word choices are gradually dying out, i think.
22:47:17 <Sgeo> Grrrrrrrrrrrrrrrr Firefox
22:47:22 <oerjan> alise: not everyone, but 90% or so.
22:47:25 <ais523> alise: the language "Auld Lang Syne" is written in
22:47:34 <pikhq> ais523: That's Scots.
22:47:40 <Sgeo> When I click a tab, I do NOT mean to close it
22:47:47 <ais523> it's pretty close to English, but some of the words are different, and some of the grammar
22:48:12 <alise> ais523: mutually intelligible, is the word
22:48:15 <oerjan> there is _one_ whole county which is nynorsk, the rest are either bokmål or split by municipality (which are mostly bokmål)
22:48:17 <alise> pikhq: what's that almost-English language?
22:48:19 <alise> that i mentioned once
22:48:21 -!- nooga has joined.
22:48:24 <alise> and you said it wouldn't make learning it harder?
22:48:32 <pikhq> alise: West Frysian?
22:48:36 <alise> oerjan: that sounds confusing :)
22:48:44 <pikhq> alise: The one that is essentially Middle English without the Latin?
22:48:55 <alise> i recall it being more similar
22:48:58 <pikhq> ais523: Scots has more to do with Middle English than Modern English, BTW.
22:49:12 <alise> also what's that thing explaining atoms written in english without germanic stuff or something like that?
22:49:17 <alise> i've forgotten the name
22:49:18 <ais523> Middle English is somewhat harder to read than Scots
22:49:22 <ais523> although still possible, more or less
22:49:30 <pikhq> alise: Middle English is very very close to English.
22:49:35 <ais523> it uses a leading y for past participles, which is really confusing if you don't know German
22:49:47 <pikhq> ais523: Yeah, that's because Scots has followed a few of the orthographic changes in English.
22:49:49 <alise> pikhq: yes, but the West Frysian Lord's Prayer isn't very intelligible to me
22:49:55 <alise> but it probably is that
22:50:02 <alise> http://en.wikipedia.org/wiki/West_Frisian_language#Sample_Text
22:50:12 <pikhq> alise: Yes, that's because of the mostly-Germanic vocabulary.
22:50:23 <Vorpal> <ais523> Vorpal: I like your measurement of duration <-- mhm
22:50:34 <alise> <alise> also what's that thing explaining atoms written in english without germanic stuff or something like that?
22:50:34 <alise> <alise> i've forgotten the name
22:51:17 <Vorpal> ais523, but I ran out of wesnoth songs now. I would have to start repeating them
22:51:24 <Vorpal> so I guess I need to move to something else
22:51:35 <ais523> oh, I often set them on a loop for an entire day or so
22:51:43 <Vorpal> ais523, I couldn't stand that :P
22:54:06 <pikhq> Also, Scots is in the midst of language attrition, and hence is starting to gain a lot of features of Standard English...
22:54:26 <oerjan> alise: http://groups.google.com/group/alt.language.artificial/msg/69250bac6c7cbaff
22:54:35 <Sgeo> WHY MUST ALL CORDS IN MY POSSESSION BREAK?
22:55:58 -!- cpressey has joined.
22:56:36 <cpressey> So I had this dream where I was evaluating a homebrew OS that was based on FreeDOS or something
22:56:45 -!- Harpyon has joined.
22:57:13 <cpressey> And it had something in it that emulated another OS, like a Commodore 64 or something (I know that's not an OS but this was a dream, right?)
22:57:31 <Phantom_Hoover> cpressey, had it been coded by a shoe with multiple personality disorder?
22:57:45 <cpressey> And when you quit the OS (again, this was a dream), it popped up a message like those old shareware nag boxes
22:58:07 <cpressey> And part of this message was written by alise
22:58:33 <cpressey> And it was encouraging us to consider what it would be like if operating systems could emulate each other through successive fractal refinement of their semantics
22:59:34 <cpressey> Well, other than after reading that, in the dream, it was clear this was some open-source project that alise contributed to, and that it worked something like that, but in an unrefined way
22:59:48 <ais523> you had a dream that was internally consistent?
22:59:53 <ais523> it's rare for mine to be like that
22:59:57 <ais523> but while asleep, I don't notice the inconsistencies
23:00:21 <cpressey> ais523: well, parts of it were.
23:00:28 <ais523> either that, or I notice one which is comparatively minor compared to all the rest
23:00:37 <ais523> and that instantly causes me to realise I'm asleep, and wake up
23:00:54 <ais523> I should really keep a set of maps near my bed, the most common side-effect is that I completely lose track of local geography
23:01:52 <Vorpal> <cpressey> And it was encouraging us to consider what it would be like if operating systems could emulate each other through successive fractal refinement of their semantics <-- wow, the mind boggles
23:02:29 <cpressey> Oh yeah, one more detail: this OS was called "Ancestor"
23:02:40 <ais523> hey, did Wikipedia just go down?
23:02:44 <Vorpal> cpressey, hey this would make a good sci fi story
23:03:13 <Vorpal> ais523, works for me, but slow
23:03:13 <ais523> hmm, working again now
23:03:58 <Phantom_Hoover> cpressey, how can one refine semantics with fractional dimension?
23:05:00 <cpressey> Phantom_Hoover: as Vorpal said, the mind boggles.
23:05:21 <Phantom_Hoover> Does fractional dimensions actually imply self-similartiy
23:06:19 <oerjan> Phantom_Hoover: almost certainly not
23:06:22 <ais523> reminds me of some thoughts I was having on data types shorter than a bit
23:07:25 <Vorpal> cpressey, "Ancestor" is a good name for an OS. the "successive fractal refinement" sounds like a perfect technobabel to introduce AI. Just two things left: decide if the AI is good or bad (no one writes stories about AIs that are somewhere in between those extremes), and actually write the story
23:07:50 <oerjan> you should easily be able to get a fractal dimension from a splitting up in which you make every part _different_ in a recursive but non-repeating way, i think
23:08:55 <oerjan> say you split up as a triangle on the top, then as a square in one part, a pentagon in another, etc. etc...
23:09:11 <oerjan> i think that should work
23:10:27 <cpressey> oerjan: it seems to me i've seen or thought of that before (possibly as a logo program)
23:10:47 <cpressey> but not from the angle of "hey this is a fractal but isn't self-similar"
23:10:54 <ais523> here, here's a possible encoding for a trit that contains exactly one trit of data: 0: 01 preceded by any even number of 0; 1: 10 preceded by any even number of 0; 2: 11 preceded by any even number of 0
23:11:10 -!- Wamanuz has quit (Remote host closed the connection).
23:11:17 <ais523> because there are multiple possible encodings for each value, you can use those as part of representing another value
23:11:52 <ais523> e.g. you can combine it with an encoding for a quint: 0 = 00, 1 = 01, 2 = 10, 3 = 11, 4 = use secondary encoding for previous trit
23:12:14 <Vorpal> ais523, isn't a trit just like base 3 iirc?
23:12:18 <ais523> and then the trit plus the quint together fit into 4 bits nicely, with a bit left over (which is managed using the tertiary encoding for the trit)
23:12:25 <ais523> Vorpal: yes, but this is representing a trit using bits
23:12:38 <ais523> now, the crazy part that alise will love: remove the first bit from that encoding of a trit
23:12:46 <ais523> you now have something that equals one trit minus one bit
23:12:56 <ais523> which is less than a bit of data
23:13:01 -!- Phantom_Hoover has quit (Ping timeout: 245 seconds).
23:13:15 <ais523> (you have two possible values, 0 and 1; but they may inject more data into other parts of the program, depending on the context)
23:13:32 <ais523> so, say, you can add an extra bit and an extra quint, and /still/ fit it into four bits
23:13:45 <ais523> which you couldn't do if you started with an actual bit
23:13:46 <Vorpal> ais523, this is quite absurd :D
23:13:49 -!- Wamanuz has joined.
23:14:12 <ais523> Huffman coding can be generalised based on a similar principle, and in the limit you get arithmetic coding
23:15:32 -!- BeholdMyGlory has joined.
23:19:02 <cpressey> ais523: ok, i'll be blunt. i have never understood why it always takes more than one bit to represent the information present in less than one bit
23:19:25 <ais523> cpressey: it takes more than one bit to /exploit/ that information
23:19:40 <ais523> if you have less than a bit of information, anything that encodes it must contain padding
23:19:47 <cpressey> well, "exploit", "observe", "make relevant"...
23:20:01 <ais523> so you encode that less-than-a-bit, and something else as well, and fit it into a smaller size than the bit + the something else
23:21:01 <Vorpal> to be blunt I have a trouble with "less than one bit of information"
23:21:54 <ais523> Vorpal: well, consider run-length encoding
23:21:55 <Vorpal> I guess imagining something with 0.5 logical states is just not possible
23:22:04 <Vorpal> ais523, yes I know how it works in compression
23:22:15 <Ilari> Vorpal: String of 0 or 1s where one can do better than coinflip to predict the next bit has less than 1 bit of information per bit.
23:22:22 <Vorpal> ais523, but I never imagine that as "half-bits" or such
23:22:27 <ais523> here's a simple one: count, value
23:22:34 <ais523> if you just get the count, and the count has a value of 1
23:22:39 <ais523> then the count contains less than a bit of info
23:22:43 <ais523> because you still need the count to know what the next bit is
23:23:22 <ais523> it contains only a very small amount of info, in that you know that the byte after is not the same as the next byte
23:23:34 <ais523> that's only a small fraction of a bit of info
23:23:53 <Vorpal> ais523, how large fraction?
23:23:58 <ais523> now, this is counterbalanced by the case that a count of, say, 200, would contain a lot more than a bit of info, if you assume the original to be randomly distributed
23:24:00 <Vorpal> 0.5? 0.2? something else?
23:24:06 <Vorpal> ais523, how do you decide that
23:24:19 <Vorpal> ais523, I can't think of a sensible way to put a number to that
23:24:21 <ais523> because it allows you to encode 256 possibilities in something that only has 255 possible states
23:24:32 <Ilari> Like if you have string of n bits, where probability of 0 is 75% and probabilty of 1 is 25% and bits are independent, then on average you need about 0.8113n bits to reprent it (as n grows without bound).
23:24:35 <ais523> so it's log_2(256/255) bits
23:25:06 <ais523> 0.005646564 bits according to my calculator
23:25:20 <ais523> obviously, if your compression scheme does that a lot, it's not a very good compression scheme :)
23:27:34 <Vorpal> ais523, I still don't get "<ais523> because it allows you to encode 256 possibilities in something that only has 255 possible states"
23:27:52 <ais523> Vorpal: because you know that the next byte is not the same as the current one
23:27:57 <ais523> as otherwise the run count would have been more than 1
23:29:16 <Vorpal> ais523, still it is only less than a bit in the encoding scheme
23:29:22 <Vorpal> it is actually one physical bit
23:29:55 <Vorpal> well probably more if stored on a harddrive, what with their crazy RLL stuff (or whatever they use nowdays)
23:31:21 <cpressey> I guess you can think of it this way: divide the entire future into two halves: one if you see 0, the other if you see 1. Based on what you've seen so far, you might be able to rule out parts of the future, even if you can't cleanly divide it in half like that.
23:31:43 <cpressey> I was going to say "world" instead of "future" initially; either works, kind of.
23:32:14 * Vorpal splits it in technobabel interpretation of quantum physics
23:32:48 <cpressey> still, this is not quite the same lines as: struct foo { int bar:8; int baz:0.33; }
23:33:05 <cpressey> or however C does bit-sizing of structurs
23:33:10 <cpressey> I haven't seen that in a long time
23:33:18 <Vorpal> cpressey, I think it does it that way
23:33:29 <Vorpal> the meaning of that int baz:0.33; escapes me
23:33:40 -!- augur has changed nick to Zoidberg.
23:33:40 <Vorpal> C sensibly define it as compile time error presumably
23:33:49 -!- Zoidberg has changed nick to augur.
23:34:09 <Vorpal> cpressey, however, now define an esolang where something like "struct foo { int bar:8; int baz:0.33; }" would have a sensible meaning
23:34:22 <cpressey> "I want to be able to store an integer between 0 and 0.33 here"
23:34:43 <Vorpal> cpressey, but "integer between 0 and 8" isn't what bar:8 means
23:34:53 <Vorpal> it means an integer of 8 bits
23:35:46 <Vorpal> cpressey, -1 presumably
23:36:04 <Vorpal> since 1 bit is not "between 0 and 2^1" (that would be 0,1,2)
23:36:13 <Vorpal> (assuming a closed range)
23:37:23 <Vorpal> the -1 is outside the exponent
23:37:39 <Vorpal> 8 bits: 2^8 = 256, 2^8-1 = 255
23:37:42 -!- GreaseMonkey has joined.
23:38:09 <Vorpal> cpressey, so .33 bits is a value between 0 and 0.25701337452182837 (approx)
23:38:34 <Vorpal> cpressey, have fun assigning a sensible meaning to that
23:38:45 <Vorpal> cpressey, I look forward to seeing the results, but now I must sleep
23:39:06 <Vorpal> actually it need not be sensible
23:39:16 <Vorpal> as long as it is fun and actually works in the esolang
23:39:25 <Vorpal> well, you know what I mean)
23:41:17 <Vorpal> cpressey, and um, more implementable than TURKEY BOMB
23:41:30 <Vorpal> (not saying it has to be easy, just not impossible)
23:46:17 <cpressey> to say that it is impossible is a half-truth
23:57:26 -!- tombom has quit (Quit: Leaving).