00:00:05 <ehird> you're all welcome.
00:00:10 <oerjan> Slereah: i have no idea what that was about
00:00:28 <psygnisfive> wasnt sure if thats what you were talking about or not :D
00:00:34 <oerjan> however you can avoid -> if you let forall have a premise
00:01:03 <Slereah> You could do the Schonfinkel thingy
00:01:13 <oerjan> ehird: P -> Q = forall a \in P : Q
00:01:27 <oerjan> that's what CoC does iirc
00:01:52 <oerjan> Slereah: i'm not unicode clean
00:02:48 -!- puzzlet has quit (Read error: 110 (Connection timed out)).
00:03:09 <ehird> Using the definition of xor:
00:03:09 <ehird> (A ⊕ B) = (A ∧ ¬B) ∨ (¬A ∧ ¬B)
00:03:13 <ehird> A ⊕ B = ∀C. ((∀C. (A ⇒ ((B ⇒ ⊥) ⇒ C)) ⇒ C) ⇒ C) ⇒ (((∀C. ((A ⇒ ⊥) ⇒ (B ⇒ C)) ⇒ C) ⇒ C) ⇒ C)
00:03:18 <ehird> Delightfully verbose.
00:03:42 <Slereah> f(x) |^x g(x) = forall x. f(x)|g(x)
00:04:21 <ehird> http://pastie.org/465557.txt?key=nevd9omyyjsa2nvbn4g
00:04:40 <ehird> oerjan: do you know why that xor is so verbose?
00:04:44 <ehird> i'd expect it to be shorter
00:04:53 <oerjan> ehird: CoC logic is intuitionistic, so A -> B is _not_ (¬A \/ B)
00:04:54 <ehird> i suppose it has some redundant clauses
00:05:54 <oerjan> i don't think xor is a particularly intuitive operation :D
00:06:29 <ehird> oerjan: now to translate that to haskell
00:09:24 <ehird> type Bottom = forall a. a
00:09:24 <ehird> type Not a = a -> Bottom
00:09:25 <ehird> type a `And` b = forall c. (a -> (b -> c)) -> c
00:09:27 <ehird> type a `Or` b = forall c. (a -> c) -> ((b -> c) -> c)
00:09:29 <ehird> type a `Xor` b = (a `And` Not b) `Or` (Not a `And` b)
00:09:31 <ehird> now to prove modus ponens
00:09:47 <oerjan> huh so that's what wolfram looks like http://news.bbc.co.uk/2/hi/technology/8026331.stm
00:10:09 <coppro> looks like my math teacher
00:10:27 <oerjan> i had imagined him rather more maniacal, with huge hair
00:10:47 <Slereah> That he finally solved his problem made him so horny
00:10:57 <Slereah> He gave him his special "wolfram prize"
00:11:38 <oerjan> no dice, wolfram only breeds with CAs
00:12:06 <ehird> dammit, modus ponens doesn't type yet
00:12:25 <ehird> oh, trivial mistake
00:13:23 <ehird> Slereah: translated my implies-and-bottom logics to haskell
00:13:59 <ehird> Slereah: oerjan: http://hpaste.org/fastcgi/hpaste.fcgi/view?id=4466
00:14:12 <Sgeo_> "How old are you now? You're 20?" "Yes" "You've hit the double-digits!"
00:16:46 <Sgeo_> ehird, as of today
00:17:00 <ehird> Sgeo_: The world would start making a lot more sense if I stopped viewing 20 as a bastion of maturity.
00:17:18 <oerjan> i'm 38, and definitely not mature
00:17:34 <Sgeo_> ...did you just say that I'm immature?
00:17:50 <ehird> Sgeo_: Mmmmmaaaaayyyyyyyyyyyynooooooooyyyyyyyyyybe.
00:18:34 <ehird> Ssssssssssssssssssseeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeintercourse.
00:19:10 * pikhq has another ~year to go before hitting 20.
00:19:12 <ehird> oerjan: say, how could I implement a boolean with this?
00:19:13 <oerjan> seintercourse? sorry i don't speak french that well
00:19:50 <ehird> I'm trying to define (type Set a = a -> Bool), 'cept without using Bool since that's against the rules
00:19:52 <oerjan> well the system F definition is something like
00:20:30 <ehird> oerjan: right but that's sort of tautological
00:20:43 <ehird> since there aren't really distinct a's in the type system, right?
00:21:33 <oerjan> that's too subtle for me at this time, i'm about to go to bed
00:23:45 <ehird> although that's not really conditional any more
00:25:55 -!- oerjan has quit ("Nude Gait").
00:55:39 <ehird> pikhq: what do you use for helvetica on linux btw? i could convert the os x font, but...
01:03:29 <pikhq> Argh. So *that's* why I don't use Helvetica.
01:03:37 <pikhq> Oh, Helvetica's installed alright...
01:03:42 <pikhq> The bitmap font. Ick.
01:04:15 <pikhq> You could probably just use the OS X font.
01:04:17 <ehird> pikhq: AIIIIIIIIIIIIIIIIIIIIIEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE
01:04:23 <ehird> And yeah, but OS X fonts generally lack hinting.
01:04:35 <ehird> Because OS X's text rendering doesn't need it :P
01:04:49 <ehird> Probably finding a Windows one would be best
01:05:18 -!- ehird_ has joined.
01:05:22 <ehird_> so I got debian all how I like it, hooray
01:05:56 <ehird_> but more or less does!
01:06:08 <ehird_> at least I think more or less does.
01:06:15 <ehird_> hmm, xchat-gnome could do with more line spacing
01:08:49 <ehird_> Let's try this "smuxi" client.
01:12:11 -!- ehird_ has quit ("Leaving").
01:12:26 -!- ehird_ has joined.
01:13:07 <pikhq> I, myself, am partial to irssi.
01:13:55 <ehird_> pikhq: I'd be all over irssi if it had a GUI frontend.
01:14:12 <ehird_> As it is, I have an objection to mimicking a restricted GUI on top of VT terminal codes...
01:14:51 <pikhq> What is this "GUI" of which you speak?
01:15:16 <ehird_> pikhq: An attitude typical of many unix users, who should be sent back to the 80s. :)
01:16:07 <ehird_> Netcraft confirms it: Smuxi is nice.
01:16:14 <pikhq> Okay, so I'd be using an actual terminal instead of a psuedoterminal.
01:16:22 <Slereah> They should be sent back to the 50's.
01:16:29 <Slereah> To program by rerouting wires.
01:16:31 <pikhq> Oh, and I'd be using the OS that inspired UNIX-HATERS.
01:16:40 <ehird_> pikhq: Or a lisp machine.
01:16:45 <ehird_> You could actually buy them those days.
01:17:00 <pikhq> Oh, *God*... UNIX-HATERS would be full of perfectly valid points!
01:17:04 <pikhq> ehird_: Oh, right.
01:17:31 <pikhq> Except that many of the points are becoming increasingly moot unless you want them to not be.
01:17:50 <pikhq> Eh, 'tis the Unix way.
01:17:58 <pikhq> But oooh. A Lisp machine.
01:18:02 <pikhq> Buckey bits and Emacs. :)
01:18:35 <ehird_> The 80s was great for computers; bad for music, hairstyles.
01:19:17 <pikhq> Oh, I dunno about it being bad for musi... Oh, wait. *80s*.
01:19:31 <pikhq> That decade that my music collection almost entirely skips.
01:19:57 <pikhq> (I have some Styx. That's about... It, as far as the 80s go)
01:20:14 <AnMaster> wolfram alpha says "Launching May 2009"
01:20:14 <ehird_> pikhq: On December 31st, 1979 at 23:59:59 -- specifically, one planck time before January 1st -- the world's collective music lobe was knocked out in a freak accident.
01:20:19 <AnMaster> is there any info on when in May
01:20:21 <ehird_> It is well known that it takes 10 years, precisely, to grow back.
01:20:28 <AnMaster> since it is the second may now
01:21:09 <pikhq> ehird_: And it still needs at least two decades to recuperate.
01:21:35 <ehird_> AnMaster: i do not know
01:25:08 <ehird_> http://imgur.com/Gcy.png Yum.
01:25:13 <ehird_> (Sans tiny resolution due to VMity.)
01:25:21 <ehird_> Modulo, not sans, that is.
01:26:08 <ehird_> bsmntbombdood: Pray tell, how is it shitty?
01:26:31 <ehird_> The future of unergonomical pseudo-leetness, yes.
01:26:40 <ehird_> I'm not one for inhumane interfaces
01:27:40 <AnMaster> (yes I was doing it on my old pentium3)
01:28:13 <Slereah> You are both an idiot and a glutton for punishment
01:29:01 <AnMaster> Slereah, I just didn't want it on my main desktop
01:31:41 <ehird_> http://isohunt.com/torrent_details/63227281/Helvetica.ttf?tab=summary
01:32:30 <ehird_> Windows Helvetica TTFs, useful for Linuxing.
01:32:31 <AnMaster> ehird, don't you have those on OS X
01:32:39 -!- WangZeDong has joined.
01:32:59 <ehird_> AnMaster: OS X fonts lack hinting, and often are tuned for just OS X's font rendering
01:33:03 <ehird_> Windows fonts are closer to Linux
01:33:33 <ehird_> Now how do I install these painlessly?
01:35:19 <ehird_> go to fonts://, drag in
01:36:23 <pikhq> Stick them in ~/.fonts
01:37:26 <ehird_> pikhq: that's what it does
01:37:29 <ehird_> I'd prefer system-wide though :(
01:38:02 <pikhq> As root, stick it in /usr/share/fonts/
01:38:37 -!- MizardX has quit ("What are you sinking about?").
01:38:49 <ehird_> pikhq: that contains subfolders truetype, type1 and X11
01:39:04 <ehird_> Does /usr/local/share/fonts work, I wonder?
01:39:11 <pikhq> It might. Worth trying.
01:39:52 <AnMaster> ehird_, why do OS X ones lack hinting
01:40:04 <ehird_> because OS X doesn't hint
01:40:15 <ehird_> its rendering is accurate enough not to
01:40:29 <AnMaster> ehird, becuase of high res screen
01:40:38 <ehird_> because the font rendering is better :P
01:41:26 <ehird_> pikhq: how do I add a directory to fontconfig's list?
01:41:34 <ehird_> it doesn't look at them :<
01:44:51 -!- ehird_ has quit (Remote closed the connection).
01:49:26 -!- Slereah has quit (Read error: 110 (Connection timed out)).
02:38:55 -!- rabideejit has joined.
02:47:29 <rabideejit> I thought of a new language. It's one of these cheesey thematic languages.
02:48:15 <rabideejit> It should be called JUSTICE. It should read like a court transcript.
02:48:17 <GregorR> We'll be expecting your submission to http://www.esolangs.org/wiki/ promptly :P
02:48:48 <GregorR> Have an implementation, or is it still in the design stage?
02:49:37 <rabideejit> Well, I'm trying to think of an interesting foundation for the language. While loops etc seem rather dull. Perhaps some sort of combinatory logic would be good.
02:49:48 <rabideejit> The basic idea is it should read like a court transcript.
02:49:57 <rabideejit> Actually, I'm a little pissed, it's coming back to me
02:50:08 <rabideejit> What it is is: it's the lambda calculus
02:50:19 <rabideejit> The evidence presented at the trial represents the bound variables
02:50:57 <rabideejit> Named functions can be called by someone in the court saying that there's precedence from another trial...
02:52:59 <rabideejit> Haha. I'm not sure -- they could be used for conditionals, but I think that would make the syntax weird -- My major problem is that in court transcript, I don't think you get to see the minutes of the juri
02:53:37 <pikhq> rabideejit: You do get the jury's results, however.
02:53:45 <pikhq> Obviously, the jury should be for input and output.
02:54:00 <GregorR> pikhq: Yeah, but you only get the results once.
02:54:34 <pikhq> Hrm. I guess this is a kangaroo court, then.
02:55:54 <rabideejit> perhaps kangaroo would be a good name for the language
02:56:17 <pikhq> A kangaroo court is show trial.
02:56:28 <pikhq> Or, rather, the court running the show trial.
02:58:10 <rabideejit> Anyway, I have to sleep. Your feedback is excellent. I shall return with tidings of the progress. Goodbye!
02:58:32 -!- rabideejit has left (?).
03:58:58 -!- pikhq has quit ("leaving").
07:15:57 -!- Sgeo_ has quit (Read error: 110 (Connection timed out)).
07:43:20 -!- Slereah has joined.
07:57:07 -!- WangZeDong has quit (Read error: 110 (Connection timed out)).
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:44:21 -!- MizardX has joined.
09:02:37 -!- tombom has joined.
09:21:06 -!- oerjan has joined.
09:22:27 -!- M0ny has joined.
09:31:51 -!- Dewio has changed nick to Dewi.
10:02:27 -!- oerjan has quit ("leaving").
10:03:49 -!- Dewi has quit ("off into upgrade land").
10:47:26 -!- M0ny has quit ("When you get sad stop being sad and be awesome instead.").
10:50:01 -!- FireFly has joined.
10:58:48 -!- BeholdMyGlory has joined.
11:38:22 -!- MigoMipo has joined.
11:51:40 -!- puzzlet_ has quit (Read error: 60 (Operation timed out)).
11:51:47 -!- puzzlet has joined.
12:23:19 -!- puzzlet_ has joined.
12:27:25 -!- FireFly has quit ("Later").
12:27:40 -!- FireFly has joined.
12:36:09 -!- puzzlet has quit (Read error: 110 (Connection timed out)).
13:02:01 -!- M0ny has joined.
13:09:56 -!- Dewi has joined.
13:37:29 -!- MizardX has quit (Read error: 60 (Operation timed out)).
13:37:34 -!- MizardX has joined.
13:45:38 -!- tombom has quit (Read error: 110 (Connection timed out)).
14:40:40 -!- puzzlet_ has quit (Read error: 104 (Connection reset by peer)).
14:40:45 -!- puzzlet has joined.
14:48:25 -!- oklopol has joined.
14:52:54 -!- ais523 has joined.
14:59:28 -!- oklopol597 has joined.
15:07:58 <ais523> oklopol597: link to the logs
15:10:43 -!- oklopol has quit (Read error: 110 (Connection timed out)).
15:23:39 -!- oklopol597 has quit (Read error: 110 (Connection timed out)).
15:50:17 -!- MizardX- has joined.
15:50:26 -!- tombom has joined.
15:50:51 -!- MizardX has quit (Read error: 54 (Connection reset by peer)).
15:50:55 -!- MizardX- has changed nick to MizardX.
15:55:28 -!- ehird has left (?).
15:56:57 -!- ehird has joined.
16:00:35 <AnMaster> ais523, any idea why gcc generates .eh_frame sections in the ELF binaries when compiling C programs. As far as I understood it, it is used for unwinding in case of exceptions but none of the -fno-exceptions -fno-unwind-tables and so on seems to remove them
16:01:02 <ais523> err, no, I ignored all the exception-handling stuff when doing gcc-bf
16:01:06 <ais523> so it almost certainly doesn't work for C++
16:01:22 <AnMaster> ais523, yes but why is it generating them for C!?
16:01:38 <ais523> are they just debug data, I wonder?
16:02:16 <ehird> i think that Debian's fontconfig has less illegal-in-US subpixel stuff than ubuntu's...
16:02:46 <ais523> Ubuntu's European, so doesn't care about the illegal-in-US stuff when it's legal everywhere else
16:02:50 <ehird> OTOH, debian have special non-US repositories
16:02:55 <ehird> for all mirrors other than the US ones
16:03:02 <ehird> so I wonder if they have the illegals?
16:05:08 <AnMaster> ah I finally made it reduce the the .eh_frame segment to a few bytes. Byt instead it grew the .text segment with 64 bytes?
16:05:50 <ais523> anyway, I spent this morning writing an Underload interpreter
16:06:04 <ais523> which is pretty fast, it's only about 20% slower than programs generated from the ehird/me Underload compiler
16:06:16 <ais523> and I intend to expand it to handle Underlambda some time later
16:06:54 <ehird> ais523: you know how you use noscript to get rid of annoyances and be more secure? it's adware, and uses obfuscated code to further this end: http://adblockplus.org/blog/attention-noscript-users
16:07:12 <ais523> ehird: yes, I'm reading the story atm
16:07:14 <AnMaster> ais523, this makes no sense: http://pastebin.ca/1410386
16:07:19 <ais523> but I never visit the noscript homepage
16:07:27 <ais523> I went and blocked that through about.config ages ago
16:07:37 <AnMaster> ais523, difference was adding -fno-asynchronous-unwind-tables as well
16:07:41 <ehird> ais523: you still have obfuscated code designed to subvert adblock plus running on your FF instance
16:09:39 <ehird> 16:08 ski expects `mod n 0 = n' .. :/
16:09:51 <GregorR> ais523: SO SHOW US THE GOODS
16:11:06 <ehird> " wget http://keithp.com/~keithp/truetype.tar.gz
16:11:06 <ehird> (and wait... it's about 3 megs ;)"
16:11:27 <ais523> http://pastebin.ca/1410392
16:12:11 <ehird> ais523: I like how the compiler is infinitely clearer.
16:12:27 <ais523> what makes you think that?
16:12:33 <GregorR> ais523: You realize that newlines and comments don't make the interpreter run slower, right? X-P
16:12:39 <ehird> err ... have you read that interpreter, ais523?
16:12:45 <ehird> the compiler is a trivial transformation for 90% of it
16:12:53 <ehird> the interpreter has bitshift rubbish
16:12:58 <ehird> and int_least32_bigger_t
16:13:01 <ais523> and the bitshift's implementing utf-8
16:13:04 <ehird> and big_least_32_bigger
16:13:12 <ehird> ais523: err ... why?
16:13:19 <ehird> Underload is ASCII...
16:13:25 <ais523> ehird: err, what makes you think that/
16:13:39 <ehird> ais523: you didn't disagree when my compiler used ascii...
16:13:49 <ehird> and because it uses no non-ASCII characters
16:14:29 <ais523> well, in Underload the meaning of codes above 128 is irrelevant
16:14:52 <ais523> in Underlambda it isn't, and I'm trying to get the compiler to do both
16:15:23 <ais523> anyway, anyone complaining about a lack of newlines in that obviously doesn't like Python, it's indented exactly the same way
16:16:06 <ehird> ais523: or maybe they think different languages deserve different styles
16:16:14 <ehird> also, python uses 4-space indents
16:16:20 <ehird> precisely because it'd be impossible to read with 2-space indents
16:16:24 <ehird> since everything would munge together
16:16:32 <ais523> just resize spaces in your editor
16:16:57 <Deewiant> ehird: Noscript was already updated to remove that Adblock Plus filter subscription.
16:17:44 <GregorR> !underload (Hello, world!)S
16:17:55 <GregorR> Again with the not compiling.
16:18:01 <GregorR> Again with the not compiling.
16:18:02 <GregorR> !underload (Hello, world!)S
16:18:03 <EgoBot> Usage: derl (-o|-a) [inputfile]
16:18:14 <ais523> you need to give -o for Underload
16:18:17 <ais523> or invoke with the name derlo
16:18:48 <GregorR> !underload (Hello, world!)S
16:18:50 <ehird> Deewiant: Still doesn't stop it being shady. I wouldn't struct it anyway.
16:19:56 <ais523> !underload (a(:^)*S):^
16:20:22 <ehird> Deewiant: Next stop: Missing the point
16:20:30 <ehird> !underload (:aSS):aSS
16:20:48 <ais523> it optimizes tailcalls, so that'll just run until it runs out of time
16:21:00 <ehird> my compiler optimized tailcalls too :P
16:22:22 -!- FireFly has quit ("Later").
16:23:32 -!- FireFly has joined.
16:23:34 <GregorR> "Click Here now http://embryogenesiswatches.cn"
16:25:17 <ais523> ehird: my interp's a lot more memory efficient than your compiler, though
16:25:32 <ais523> !underload ((+)S:*:^):^
16:25:32 <EgoBot> +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
16:25:34 <AnMaster> <ehird> Deewiant: Next stop: Missing the point <-- I clearly see what ehird meant there
16:25:49 <ais523> ehird: try that program on your compiler, I bet it won't get nearly that far
16:25:51 <AnMaster> because the idea is good. And it used to be good at least.
16:26:08 <ehird> ais523: er, what? It would run forever.
16:26:11 <ehird> I optimized tail calls.
16:26:16 <ais523> ehird: look more carefully at it
16:26:20 <ais523> it uses exponential memory
16:26:29 <ehird> Shrug. Then you need exponential memory.
16:26:31 <ais523> although constant callstack space, it uses exponential data-stack space
16:27:09 <fungot> +++++++++++ ...too much stack!
16:27:15 <ehird> http://www.oyhus.no/SubLCD.html this is pretty cool
16:27:25 <AnMaster> that was very quick to give that error in fungot
16:27:26 <fungot> AnMaster: ' of course they answer to their names? is it particular or universal? perhaps looking-glass milk isn't good to fnord oh, oh, hear!" cried sylvie, in a melancholy voice. ' it's very provoking,' humpty dumpty cried, breaking into a sudden passion. ' you've only a few inches high, and was gone in a moment.
16:27:52 <AnMaster> that is a very good summary of the book fungot
16:27:52 <fungot> AnMaster: " oh, don't go on like that!'
16:28:19 <ais523> hmm.... thutubot doesn't like exponential programs
16:28:25 <ais523> it should have hit the too much stack warning already...
16:28:31 <AnMaster> ais523, doesn't it have a limit
16:28:43 <ais523> AnMaster: yes, and it should have reached it by now
16:28:51 <thutubot> ++++++++++ ...too much memory used!
16:29:05 <ais523> nah, the limit's obviously just a bit too high
16:30:12 -!- oerjan has joined.
16:32:30 <oerjan> <AnMaster> ais523, any idea why gcc generates .eh_frame sections in the ELF binaries when compiling C programs. [...] <<< obviously it's for framing canadians.
16:32:56 <ehird> http://www.oyhus.no/Monitors6.html Holy. Fuck.
16:33:31 <ehird> AnMaster: " I recently threw them away, since they used so much current that they often blew the house fuses and aborted the server. "
16:34:08 <ehird> it's a horrible xinerama setup
16:34:15 <ehird> xinerama would work with monitors without any borders :P
16:34:55 <AnMaster> ehird, where on the page is it
16:35:29 <AnMaster> ehird, not the same page on it that you linked though. meh
16:36:46 -!- KingOfKarlsruhe has joined.
16:38:05 <AnMaster> there, disabled noscript. Disabled javascript.
16:46:15 -!- MigoMipo_ has joined.
16:46:15 -!- MigoMipo has quit (Nick collision from services.).
16:46:38 -!- MigoMipo_ has changed nick to MigoMipo.
16:46:58 -!- MigoMipo has left (?).
16:50:28 <oerjan> <ehird> 16:08 ski expects `mod n 0 = n' .. :/
16:50:47 <oerjan> that _is_ the most logical definition if you make it defined
16:51:06 <oerjan> since k + 0*m = n has only one solution k
16:51:35 <ehird> er, it has no solutions, no?
16:52:13 <oerjan> of course it has no solution fulfilling the expected inequality 0 <= k < n
16:52:24 * ais523 tries to figure out whether nobody being able to read my Underload interp is a good thing
16:52:45 <oerjan> ais523: i haven't tried yet
16:53:11 <ais523> I spent hours debugging the GC this morning
16:53:37 <oerjan> oh it's not written in a language with GC?
16:53:59 <oerjan> ais523: reference counting should suffice
16:54:24 <ais523> just I use the refcounting for other purposes too
16:54:29 <ehird> refcounting is so inefficient
16:54:50 <ais523> not in terms of memory used
16:55:13 <ais523> Underload programs that use memory exponentially tend to only use quadratic memory in derl
16:55:30 <ais523> due to the compression it uses to store thigns
16:59:01 -!- MizardX has quit (Read error: 131 (Connection reset by peer)).
16:59:03 -!- MizardX- has joined.
16:59:06 <oerjan> ais523: sharing subexpressions i assume
16:59:14 <ais523> oerjan: yes, that's how it's done
16:59:40 -!- MizardX- has changed nick to MizardX.
16:59:44 <ais523> also, if subexpressions are only used in one place, I modify them when doing ^ or *, rather than creating another data structure to hold the relationship
17:00:20 <ais523> that becomes CAT(a,b) in memory if a is shared with anything
17:00:33 <ais523> if it isn't, I just change the end pointer of a to point to the start of b
17:00:41 <ehird> constitutes about half the things I hear from AnMaster
17:00:51 <ais523> I use UCS-4 internally
17:01:07 <AnMaster> ais523, right. Makes sense. Have you integrated this in EgoBot yet
17:01:17 <ais523> AnMaster: no, GregorR's integrated it in EgoBot
17:01:28 <EgoBot> Supported commands: bf_txtgen help info 1l 2l adjust axo bch befunge befunge98 bf bf16 bf32 bf8 dimensifuck glass glypho kipple lambda lazyk linguine malbolge pbrain qbf rail rhotor sadol sceql trigger udage01 underload unlambda whirl
17:02:12 <ais523> AnMaster: yes with debug level set 2 or higher
17:02:19 <ais523> lower than that, it skips free at program end
17:02:59 <ais523> ehird: I was valgrinding it a /lot/ to test for bugs
17:03:06 <ais523> more or less every bug in there showed up on valgrind
17:03:12 <ehird> still, a rather pointless thing to ask someone who shows a propgram
17:03:12 -!- coppro has quit (Connection reset by peer).
17:03:16 <AnMaster> ais523, you indention is mixed space and tabs
17:03:45 <ais523> anyway, here's an expression it took over an hour to get working:
17:03:54 <ais523> !underload (((1)S)^):^^
17:04:21 <ais523> that's a minimal testcase for an obscure optimizer bug
17:04:38 -!- coppro has joined.
17:04:57 <AnMaster> ais523, not likely any other interpreter will hit it then
17:05:21 <ehird> ais523: now I have to make a new interpreter/compiler that's really fast
17:05:24 <ais523> well, unless they try to optimise the same way
17:05:25 <ehird> I will do TYPE INFERENCE.
17:05:32 <AnMaster> ais523, anyway, your code: tl;drmoi
17:05:33 <ehird> :::*** will be stored as {num,3}
17:05:44 <ais523> ehird: I'm planning to do that too, some time
17:05:55 <ais523> both my unfinished Overload interps optimise numbers already
17:05:55 <ehird> I have patented it
17:06:02 <ais523> I have prior art, so there
17:06:13 <ehird> ais523: Like the patent system ever listens to prior art
17:06:56 <ais523> derl's O((log n)^2) for numbers atm
17:06:57 <AnMaster> ehird, also it is the same as I already do in my brainfuck compiler, optimising +++ and such to set constants when they are known (like cell is known to be 0 before, from a [-] or whatever)
17:06:58 <ais523> which isn't all that bad
17:07:11 <ais523> that's in memory, it's O(n) in actual processing for the same reason as brainfuck
17:07:16 <ehird> that is not even close to the same
17:07:51 <AnMaster> ehird, yes. "Optimising sub-optimal esolang representations for numbers into efficient representations"
17:08:09 <AnMaster> yes it covers church numerals too
17:08:58 <AnMaster> in fact, it would include constant folding in any esolangs too
17:10:51 <ais523> AnMaster: if you want to concatenate two linked lists
17:10:59 <ais523> then you set the last pointer in the first one to point to the start of the second
17:11:08 <ais523> it only works if nothing else is using the first one, though
17:11:17 <ais523> that saves a bit of memory, I was more paranoid about memory usage than speed for some reason
17:11:40 <AnMaster> I mean, I have done that in programming before. And never knew what it was called.
17:12:04 <AnMaster> ais523, it saves speed too doesn't it
17:12:15 <ais523> probably, but I haven't profiled
17:12:21 <ais523> it avoids the need to malloc
17:12:30 <ais523> but it requires a while loop to find the end of the list
17:12:31 <AnMaster> I mean you only have to traverse the first list, so O(n) (unless you already have a pointer it's end, in which case that is O(1))
17:12:32 -!- pikhq has joined.
17:12:42 <ais523> yep, but the other way's O(n) too
17:12:47 <ais523> the question is which is faster on a typical list
17:13:10 <AnMaster> ais523, wouldn't the other way be copying the linked list, AND finding the end of one of them.
17:13:21 <AnMaster> Thus being O(n) too indeed (I think)
17:13:23 <ais523> the other way is to do a lazy cat
17:13:31 <ais523> which is what I do when the first list is shared and so can't be modified
17:13:46 <ais523> I create a cat element in memory, and evaluate it later when the first list has already been consumed
17:13:52 <ais523> so the first list doesn't need copying
17:13:57 <ais523> as it's already gone by then
17:14:21 <AnMaster> ais523, well compared to that, lazy cat should be faster right?
17:14:37 <AnMaster> where do you store this lazy cat
17:14:38 <ais523> I suspect lazy cat will be faster on large lists, certainly instantaneously
17:14:41 <ais523> but it produces more memory
17:14:45 <ais523> and it's stored on the heap
17:14:55 <ais523> I'm wondering if lazy-catting everything would make memory usage a higher computational order
17:14:59 <AnMaster> ais523, no, I meant, how do you check in the code for a lazy cat situation
17:15:02 <ais523> which could make speed a higher computational order
17:15:13 <ais523> AnMaster: it's in the big switch, look for case CAT:
17:15:22 <ais523> and then I have the zero stack
17:15:31 <ais523> which is a stack of pointers to use instead if you hit a null pointer
17:15:38 <ais523> lazy cat's implemented in terms of it
17:16:08 <AnMaster> ais523, how do you store it relative the first list
17:16:23 <AnMaster> "<ais523> I create a cat element in memory, and evaluate it later when the first list has already been consumed"
17:16:24 <pikhq> Obviously, this lazy cat is stored right next to Einstein's cat with a long tail.
17:16:33 <ais523> AnMaster: it has a pointer to the first list and the second list
17:16:33 <AnMaster> I may have misinterpreted that
17:16:44 <ais523> once it's evaluated, it returns the first list, and pushes a pointer to the second on the zero stack
17:16:58 <ais523> then when the end of the first list is reached, it hits a NULL, so it pops the second from the zero stack and keeps going
17:17:26 <oerjan> ais523: hm if you cat n lists together, you only get n-1 cat cells, so as long as you don't use it for empty lists i think you only get a constant multiplier at most
17:17:38 <ais523> oerjan: I don't use it for empty lists
17:17:52 <ais523> but you can end up catting the lazy cats themselves
17:17:55 <ais523> and you end up in a big lazy glob
17:17:59 <ais523> which is why I think it might blow up
17:18:28 <ais523> say (x):*:*:* becomes x lazy-catted to itself, that lazycat's lazy-catted to itself, and that lazycat's lazy-catted to itself
17:18:36 <ais523> so you have more lazy cats than actual elements there
17:18:47 <ais523> OTOH, doing that's much more efficient than actually trying to store the number
17:19:00 <oerjan> ais523: um but you cannot use mutation on self-catting, obviously
17:19:26 <ais523> actually, come to think of it, you could; if the second list was modifiable, you could modify that
17:19:43 <AnMaster> ais523, can you do that exponentially. I mean cause that interpreter to create an exponential number of such lazy cats
17:19:49 <ais523> and if the first list was modifiable then, you could force the lazycat at that point and then attach to the forced list
17:19:51 <ais523> AnMaster: I'm not sure
17:20:06 <ais523> OTOH, it's certainly better than the naive approach, which would be hyperexponential for the same program
17:20:34 <ais523> AnMaster: yes, just by considering what would happen if you forced all the cats immediately
17:20:50 <ais523> !underload ((+)S:*:^):^
17:20:51 <EgoBot> +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
17:21:03 <pikhq> Spiffy. Just spiffy.
17:21:03 <AnMaster> I thought you meant a program creating hyperexponential number of lazy cats in it
17:21:17 <fungot> +++++++++++ ...too much stack!
17:21:26 <ais523> AnMaster: yes, I haven't modified thutubot
17:21:34 <ais523> you can tell that by the fact it hasn't left and rejoined
17:21:40 <ais523> thutu's a compiled language, at least in all known implementations
17:21:58 <AnMaster> ais523, it will run out some time soon or what
17:22:01 <ais523> I don't think it's necessarily a bug, just a limit (in this case, memory use) set slightly too high
17:22:05 <ais523> it hits the memory limit first
17:22:12 <ais523> time's measured in commands
17:22:17 <ais523> and Thutu is O(n) slower than anything else
17:22:21 <thutubot> ++++++++++ ...too much memory used!
17:22:32 <AnMaster> ais523, I missed that message last time
17:23:01 <AnMaster> rather. just left 5 seconds before
17:24:33 <AnMaster> ais523, tested ick with gcc 4.4 yet
17:24:43 <ais523> I suspect it will work, though
17:24:59 <ais523> given that gcc will have been trying not to break compatibility, and ick's designed to run on anything as it is
17:25:22 <ais523> "If you're using IE, do not be alarmed. This page does not really look like crap, it is only your browser."
17:25:24 <pikhq> GCC tries not to break compatibility, but sometimes compatibility is broken.
17:25:33 <AnMaster> ais523, it does have parts that are gcc specific
17:25:43 <pikhq> (some projects depend on behavior of the optimising pass.)
17:26:00 <AnMaster> unless you changed that recently
17:26:03 <ais523> it depends on ({ }) and a command-line option to gld
17:26:07 <pikhq> For example, ffmpeg won't compile without -O.
17:26:26 <ehird> 17:25 ais523: "If you're using IE, do not be alarmed. This page does not really look like crap, it is only your browser."
17:26:31 <pikhq> Erm. It'll compile, it won't execute at all.
17:26:32 <ehird> that site IS ugly, though, ironcially
17:26:42 <AnMaster> pikhq, how do they debug it then
17:26:50 <ehird> you can use -g with -O
17:26:59 <AnMaster> yes right. But it doesn't work well
17:27:10 <oerjan> ehird: the graphical version of muphry's law? :D
17:27:11 <ais523> ehird: I'm actually surprised that string only gives one result
17:27:16 <AnMaster> -O0 works a lot better when you are working in gdb
17:27:17 <ehird> ais523: in quotes, ofc
17:27:21 <ehird> yep, ubuntu's subpixel rendering is different and better than Debian's
17:27:25 <pikhq> It is my professional opinion that the ffmpeg developers are freaking mad.
17:27:58 <pikhq> ehird: Probably because Ubuntu's a fork of sid. ;p
17:28:05 <ehird> AnMaster: soon two?
17:28:21 <ehird> pikhq: no -- because Ubuntu doesn't care about breaking US law, just european law
17:28:26 <ehird> it may be in a debian non-US repository
17:28:26 <AnMaster> ehird, yes. one from that site. One from google crawling the logs for this channel
17:28:38 <ehird> AnMaster: you confused me since tunes are making a new site, supposedly
17:28:46 <pikhq> And Ubuntu doesn't even care about breaking European law.
17:28:51 <pikhq> Just South African law. ;)
17:28:53 <AnMaster> ehird, with spiffy GUI for the logs
17:28:58 <ehird> Canonical aren't a south african country
17:29:01 <ais523> pikhq: their main servers are in Europe
17:29:03 <ehird> AnMaster: the logs are not a part of the tunes project...
17:29:08 <ais523> ehird: they aren't even a european country
17:29:10 <pikhq> Oh? Guess I'm wrong.
17:29:11 <ehird> AnMaster: the tunes project is mainly the OS project
17:29:13 <ais523> although that would be impressive
17:29:23 <AnMaster> ehird, you failed to detect the sarcasm...
17:29:43 <AnMaster> ais523, where are they a company then
17:29:43 <ehird> AnMaster: people, as a rule, can't detect sarcasm without any hints that it would be sarcasm -- i.e. it could easily be perfectly serious -- and that is not funny at all
17:29:53 <ehird> since, y'know, we're not psychic
17:30:09 <ais523> AnMaster: Europe, they just aren't a country
17:30:24 <ais523> the main server's in the UK, although that doesn't mean the company is of course
17:30:27 <ehird> ais523: heh, country XD
17:30:30 <AnMaster> ais523, oh I mentally corrected the typo
17:30:33 <ehird> AnMaster: yes, but my sarcasm is detectable
17:30:44 <ehird> AnMaster: i disagree, since others have detected it
17:30:53 <ehird> 17:28 AnMaster: ehird, with spiffy GUI for the logs
17:30:54 <ehird> 17:28 AnMaster: or what
17:30:59 <ehird> @anyone: does this smell even slightly of sarcasm to you?
17:31:43 <ais523> it is potentially sarcasti
17:31:48 <ais523> but you can't tell without more context
17:32:24 -!- Dewi has quit (Remote closed the connection).
17:32:30 -!- Dewi has joined.
17:34:07 * ehird concludes that Debian, if it has a non-US nice freetype, is his best choice.
17:35:36 <AnMaster> ais523, hyperexponential == "like exponential but with the hyper operator" right? I can only find "hyperexponential distribution" which seems to be a slightly different concept
17:35:49 <ais523> AnMaster: something that blows up faster than exponential
17:36:07 <ehird> ais523: hyperexponential reaches infinity in finite time, no?
17:36:09 <AnMaster> so not specific to the hyper operator then
17:36:24 <ais523> I may have used the wrong term
17:36:26 <ehird> I'm pretty sure that's true
17:36:34 <ehird> Maybe "superexponential"
17:36:45 <oerjan> AnMaster: greek:hyper = latin:super
17:36:52 <ehird> there should be a way to install some of the ubuntu polish on debian.
17:37:05 <ais523> I don't know what happens if you enable the ubuntu repos in debian
17:37:12 <ehird> like the cleaned up system→preferences/administration and the nice wireless/wired connection helper thingies
17:37:12 <ais523> I wonder if it's as disastrous as doing it the other way round?
17:37:15 <AnMaster> oerjan, why are you highlighting *me* about that
17:37:17 <ehird> ais523: almost certainly
17:37:36 <ais523> the connection helper thing is NetworkManager + nm-applet-gnome
17:37:42 <ais523> Debian probably has the packages
17:37:45 <ehird> but yeah, I'd like some of ubuntu's polis without its bloaty feeling and without the goddamn ubuntu logo :-D
17:37:49 <oerjan> AnMaster: <AnMaster> ais523, hyperexponential == "like exponential but with the hyper operator" right?
17:37:54 <pikhq> (don't know if they *still* have that...)
17:37:58 <ehird> ais523: system→preferences/administration cleanup is a godsend though
17:38:02 <ehird> that thing's a maze on debian
17:38:06 <oerjan> that sort of implied you didn't know the basic meaning of the prefix
17:38:28 <AnMaster> oerjan, well hyper == super == "more wow" ;P
17:38:37 <ehird> still, I can handle Debian, prolly
17:38:44 <ehird> esp. since its installation is more flexible
17:38:49 <ehird> so I can do my LVM rubbish for the ssd
17:39:24 <ais523> err, AnMaster used an interrobang?
17:39:27 <ehird> AnMaster: It's slang.
17:39:28 <fizzie> Debian does have the network-manager package, and a network-manager-gnome systray applet.
17:39:31 <ehird> It doesn't mean "crap".
17:39:37 <AnMaster> the correct term is "baroque" for lvm
17:39:50 <ais523> AnMaster: I didn't expect it from you
17:39:52 <AnMaster> ais523, it isn't like the first time...
17:39:53 <ehird> Oh, and to piss off AnMaster:
17:39:56 <ehird> LVM rubbish LVM rubbish LVM rubbish LVM rubbish LVM rubbish LVM rubbish LVM rubbish LVM rubbish LVM rubbish LVM rubbish LVM rubbish LVM rubbish
17:40:09 <ais523> AnMaster: I haven't seen you do an interrobang before
17:40:10 <ehird> fizzie: I wonder if it comes with its desktop-environment packageset
17:40:27 <AnMaster> ais523, I haven't seen you around more than at most half of the days
17:40:41 <ais523> AnMaster: I don't have Internet at home
17:41:03 <AnMaster> ais523, I know. It doesn't change the fact that you have rather limited coverage of the channel and don't read all the logs.
17:41:12 <AnMaster> (the latter I don't blame you for)
17:41:53 <fizzie> Well, the "gnome" package recommends network-manager-gnome, which would pull in the network-manager itself. But it's just a "recommends", not a "depends" or a "suggests".
17:42:08 <ehird> I'd like to say something CONTRAVARSIAL:
17:42:11 <ehird> I like Compizzzzzzzzz
17:42:14 <oerjan> AnMaster: 50% is more than enough for statistical evidence
17:42:54 <AnMaster> oerjan, sure. However it doesn't *always* work. As shown in this case.
17:43:36 <AnMaster> fizzie, hm what is the difference between "recommends" and "suggests" in *.deb based systesm
17:44:14 <ehird> pikhq: do you know if there's a way to tell Debian to use gksudo instead of gksu?
17:44:19 <ehird> Since I want to lock the root account.
17:45:51 <pikhq> No; I don't use a desktop environment.
17:46:04 <ehird> pikhq: What's that got to do with anything?
17:46:09 <fizzie> The "recommends" list is supposed to contain "packages that would be found together with this one in all but unusual installations" while "suggests" should have things that are potentially useful but not in any sense required. So I guess "suggests" is the milder version, in fact. Funny that aptitude orders them depends/recommends/suggests. Well, maybe it's just alphabetical in that sense.
17:46:10 <ehird> gksudo is just a gtk sudo thing :P
17:46:23 <ehird> All Gtk-but-you-know,-rooted apps in debian use it by default
17:46:43 <AnMaster> only GTK app I use on a regular basis is Gimp I think
17:46:50 <AnMaster> possibly there is some other that I forgot about
17:47:05 <pikhq> The only GTK app I use on a regular basis is Xulrunner.
17:47:05 <ehird> AnMaster: firefox?
17:47:20 <ehird> anyway, gtk's a fine graphical toolkit.
17:47:40 <pikhq> If you don't have to code for it.
17:47:48 <ehird> pikhq: Oh, no doubting, GObject is a mess.
17:47:48 <fizzie> Debian-installer asks whether you want to use a root account with a password, or whether just to set up sudo. I've no idea about graphical-but-root stuff, though.
17:47:53 <pikhq> Gobject is... Evil.
17:47:54 <ehird> Nothing a bit of abstraction couldn't fix.
17:48:01 <AnMaster> there is no good GUI toolkit if you have to code for it that also looks good
17:48:03 <ehird> fizzie: Does it? It didn't to me. Do you mean the advanced install?
17:48:20 <ehird> AnMaster: Yeah... Qt is OK to code for... but OTOH it doesn't really have a nice selection of themes.
17:48:29 <ehird> But, Qt 4.4 has QGtkStyle, which renders directly via gtk.
17:48:30 <fizzie> ehird: Well, I've always done the expert install-mode thing, yes. It asks a bit more questions that way, I guess.
17:48:32 <ehird> Which is kick-ass.
17:48:52 <ehird> So maybe Qt is a nice Gtk abstraction ;-)
17:48:59 <AnMaster> ehird, except that doesn't work well when I tested it
17:49:07 <ehird> Does it not? Darn.
17:49:18 <AnMaster> buggy for most GTK themes I tried.
17:49:22 <pikhq> And there's gtk-qt-engine. I wonder what happens if you use both of them?
17:49:36 <GregorR> Massive failure I'm thinkin' :P
17:49:36 <AnMaster> the default gtk theme worked ok with it
17:49:50 <pikhq> AnMaster: I thought that Tk had recently started rendering using GTK?
17:50:05 <AnMaster> I might not have hit that version yet
17:50:15 <ehird> as a separate theme
17:50:23 <ehird> Tk has some ways to make it look modern
17:50:23 * pikhq doesn't know; no Tk usage.
17:50:27 <ehird> But not by default, eurgh.
17:50:31 <AnMaster> because the one I have here looks somewhat like.... Motif + CDE + pure X
17:50:40 <ehird> you guys ever used nedit? :-)
17:50:50 <ehird> http://www.nedit.org/
17:50:53 <ehird> very old X11 editor
17:50:56 <pikhq> Guess it'll be in officially in 8.6?
17:51:07 <ehird> AnMaster: it reminded me:
17:51:09 <ehird> http://www.nedit.org/technotes/looks-1.php
17:51:44 <pikhq> AnMaster: At the time it was written, the main UI toolkit was Motif.
17:51:45 <AnMaster> it still looks bad in the latter one
17:51:45 <ehird> And the less-3Dness :P
17:51:56 <ehird> AnMaster: It looks acceptable
17:52:03 <ehird> I mean, there's not much there to look ugly
17:52:07 <AnMaster> pikhq, actually I like motif. For the retro feeling it gives to mosiac
17:53:18 <AnMaster> (yes I have a mosaic version which runs on modern Linux and even have some bugs fixed in it, oh and it is set to not display css or scripts inline, just to make it able to render google.com reasonably)
17:54:01 <fizzie> It certainly looks better than, say, xman.
17:54:07 <ehird> fizzie: oh man, xedit
17:54:10 <ehird> you ever used xedit?
17:54:14 <ehird> that thing did SEARCH!
17:54:24 <ehird> i wonder what toolkit it was
17:54:29 <ehird> it's the black and white one
17:54:33 <ehird> with just a two pixel border on buttons
17:54:57 <AnMaster> ehird, I think I used xedit once. Used defined as ^C
17:54:57 <pikhq> I used to have OpenMotif installed... I can't remember why.
17:55:20 <ehird> AnMaster: xedit is perfectly intuitive
17:55:22 <ehird> it's just very very X
17:55:40 <pikhq> Oh, I still do have it.
17:55:40 <pikhq> Why do I have xpdf?
17:55:54 <ehird> pikhq: Evince uses gtk?
17:56:05 -!- puzzlet has quit (Read error: 60 (Operation timed out)).
17:56:07 -!- puzzlet has joined.
17:56:14 <pikhq> ehird: xedit didn't use a toolkit. It used libX11.
17:56:24 <ehird> pikhq: well, those widgets are in xman too
17:56:30 <fizzie> The "scrollbar scrolls down on left-click, up on right-click, and you can drag it with middle-click" stuff is the horrible.
17:56:45 <AnMaster> fizzie, that is like in emacs iirc
17:56:50 <AnMaster> but you can change it there I think
17:56:54 <fizzie> xman's linked with libXaw.so.7 here, so I guess it's from Athena.
17:57:03 <AnMaster> never been bothered enough to figure it out
17:57:16 <ehird> AnMaster: xterm is the main offender
17:57:21 <ehird> that's why you should use urxvt.
17:57:26 <AnMaster> ehird, actually xterm isn't too bad
17:57:29 <ehird> or gnome-terminal if you're not gnome averse
17:57:32 <ehird> AnMaster: it is for the scrollbar
17:57:36 <ehird> that's exactly the behaviour it has
17:57:47 <ehird> xterm has a scrollbar.
17:57:52 <ehird> you can turn it off but it has one.
17:58:02 <AnMaster> so I guess it is off by default
17:58:08 <AnMaster> I never turned it either on or off
17:59:16 <fizzie> "scrollBar (class ScrollBar): Specifies whether or not the scrollbar should be displayed. The default is ‘‘false.’’"
17:59:22 <fizzie> Of course I don't know if that's the universal default.
18:00:30 <AnMaster> gentoo tends to not change defaults for stuff. And not install customised icons replacing KDE/Gnome/whatever logos with the distro ones
18:02:46 <pikhq> Yeah, it's kinda nice that Gentoo only tends to do patches to, y'know, make shit work right.
18:03:01 <ehird> Debian does that stuff a bit, but OTOH you get a nice desktop that always works together
18:03:07 <ehird> And you can probably replace the debian logoshits.
18:07:00 -!- M0ny has quit ("When you get sad stop being sad and be awesome instead.").
18:16:28 <AnMaster> I mean, a logo at the boot or such (though I prefer text bootup) is fine with me. And possibly default desktop background picture. (I'm going to replace it with a single solid colour anyway!)
18:16:36 <AnMaster> but when it gets to the menu...
18:16:54 <ehird> AnMaster: all debian does beyond that is make gnome's menu icon the debian logo instead of the gnome foot
18:17:01 <ehird> which is probably easily replacable
18:17:04 <ehird> i agree they should do less
18:17:08 <AnMaster> ehird, can you get the MP foot
18:17:11 <ehird> but it's easily reversible, nothing really permanent
18:17:27 <ehird> AnMaster: It's probably just a png or an svg
18:17:39 <ehird> The gnome foot is ugly anyway, Debian's logo is nicer :P
18:17:54 <ehird> I dunno, that spiral is pretty nice.
18:18:06 <AnMaster> ehird, but the MP one is pretty isn't it
18:18:16 <ehird> I have no idea what the fuck you're talking about.
18:19:29 <AnMaster> ehird, you know their foot right
18:19:49 <AnMaster> (I'm never sure when it comes to you)
18:22:20 -!- pikhq has quit ("leaving").
18:23:44 -!- MizardX- has joined.
18:23:53 -!- MizardX has quit (Read error: 54 (Connection reset by peer)).
18:24:19 -!- MizardX- has changed nick to MizardX.
18:24:42 <AnMaster> oerjan, IWC! (Or have I said that already today?)
18:25:55 <oerjan> i wonder if DMM is going to rewrite history for very long
18:26:25 <oerjan> (well, apart from the historical themes, for which the answer is obviously "yes")
18:26:27 <AnMaster> also mezzacotta was interesting today. I never heard the word "astrogator" before
18:26:51 <ehird> AnMaster: an agoran office for keeping the gate to the stars, obviously
18:27:30 <AnMaster> ehird, spacecraft navigator in fact
18:27:36 <ehird> ais523: do you know how that started btw? It's not the Speakor.
18:27:56 <ehird> Speaker is an agoran office.
18:28:00 <ehird> Agoran spells offices with or.
18:28:03 <ehird> so rulekeeper is Rulekeepor
18:28:03 <AnMaster> Did you mean: speaker Top 2 results shown
18:28:06 <ehird> but it's Speaker, not Speakor
18:28:10 <ehird> and I was wondering why
18:28:13 <ehird> speaker was there from the start
18:28:15 <ehird> so it must be newer than that
18:28:21 <ehird> and I was wondering how it started
18:28:50 <ehird> AnMaster: dude, it's from 1993 to present
18:28:56 <ehird> and a large number of years were lost
18:28:56 <oerjan> ehird: when i was around it was only for -keepor
18:29:02 <ehird> I think the current logs start around 2004, right ais523?
18:29:03 <AnMaster> ehird, surely there is a "search" feature
18:29:13 <oerjan> i don't think it was Bankor, say
18:29:13 <ehird> and as I said, pre-2004 logs are lost
18:29:20 <ehird> AnMaster: it's just mailman.
18:29:30 <AnMaster> ehird, right. Crappy in other words.
18:29:33 <ehird> it was logged on escribe.com
18:29:38 <ehird> which then promptly died and took all logs with it
18:29:54 <ehird> and you guys think i'm crazy to hate rafb.net :-)
18:30:16 <ais523> rafb.net always takes logs with it
18:30:21 <ais523> for the others, it's less predictable
18:30:37 <AnMaster> I always set expire on any other pastebins when pasting to this channel
18:30:52 <AnMaster> oh and ais523's program above was set to expire I noticed
18:30:58 <ais523> AnMaster: no it wasn't
18:31:23 <ehird> we'll see who's laughing when civilization collapses because we couldn't find the specs to a nuclear anti-destruction shield that were pasted on rafb.net 3 years ago
18:31:44 <ehird> the last action done by humanity? reading the logs where everyone praises its technical prowess.
18:32:24 <AnMaster> yay a 72 KB executable for cfunge *with* the floating point fingerprints.
18:32:34 <AnMaster> Deewiant, can you get such a small ccbi
18:33:09 <AnMaster> without floating point using fingerprints: 59 KB
18:33:12 <Deewiant> No, even hello world with a non-minimized D stdlib is over 100 KB
18:33:54 <Deewiant> And minimizing the stdlib would result in pretty much C :-P
18:34:51 <GregorR> ARGH, just how much memory does Java WANT X_X
18:35:02 <GregorR> ulimit -v $(( 64 * 1024 )) # not enough for java
18:35:10 <GregorR> ulimit -v $(( 256 * 1024 )) # not enough for java
18:35:37 <AnMaster> -march=native -Os -DNDEBUG -fno-unwind-tables -fno-async-unwind-tables -Wl,-O1,-s
18:35:58 <AnMaster> the key things there are "-fno-unwind-tables -fno-async-unwind-tables"
18:36:08 <AnMaster> even for C code gcc generates unwind stuff
18:36:27 <AnMaster> and even with those I still have .eh_frame and .eh_frame_hdr
18:36:31 <AnMaster> but a lot smaller such sections
18:39:06 <AnMaster> GregorR, err I think it over-allocates
18:39:17 <AnMaster> there is a command line option iirc
18:39:24 <GregorR> Doesn't seem happy still X_X
18:39:42 <fizzie> You can tweak the initial Java heap sizes and such with something like "-Xms16m" to make it start with a 16-megabyte heap; I have no idea what the default could be. Although I guess setting the limit might work just as well.
18:39:51 <AnMaster> -Xss<size> set java thread stack size
18:40:04 <fizzie> Still, it seems to use a horrible amount of virtual memory, not so much resident.
18:40:25 <fizzie> Virtual size of this one java process is 1413696k, resident size 55960k.
18:40:52 <AnMaster> GregorR, found out where the buffer came from btw
18:42:18 <Deewiant> Virtual usage doesn't really matter, does it?
18:42:39 <AnMaster> GregorR, for cfunge running on mycology
18:42:40 <ehird> Virtual memory usage is irrelevant
18:42:52 <AnMaster> after it quit and then still sent it after reconnect
18:44:01 <Deewiant> Running foobar2000 in Wine results in four processes with 3.6 gigs of virtual memory usage
18:44:10 <AnMaster> GregorR, ... question was: have you fixed that issue or not
18:44:58 <AnMaster> Deewiant, I can get a full featured cfunge (32-bit cells, -Os, stripped) in 96 K for x86_64
18:45:18 <Deewiant> Yes, C tends to result in small executables, especially with dynamic linking.
18:45:23 <ehird> Has anyone said they cared yet, AnMaster? :P
18:46:03 <AnMaster> I got a call from Nokia about using it on their phones today. They said something about needing a fingerprint GPRS though.
18:46:34 <AnMaster> Deewiant, I'll try static linking, sec
18:46:37 <fizzie> According to /proc/<pid>/maps the Java VM has one exactly 617.5M-sized anonymous mapping, and one 308.75M one. I'm guessing those are some sort of fraction-of-available-memory things.
18:47:18 <ais523> -rwxr-xr-x 1 ais523 ais523 13892 2009-05-02 18:46 derl
18:47:19 <AnMaster> linking errors for 32-bit ncurses hm
18:47:21 <ais523> that's after stripping
18:48:11 <Deewiant> On the subject of file sizes of unrelated things:
18:48:13 <Deewiant> -rwxr-xr-x 1 deewiant deewiant 4669 2009-04-12 22:48 dobelx64
18:48:27 <AnMaster> Deewiant, that's impressive yes
18:48:46 <Deewiant> Stripping increased its size to 4888
18:48:50 <ais523> that's writtten in asm, isn't it?
18:48:55 <ehird> ais523: with custom headers
18:48:56 <ais523> also, how does stripping make something /bigger/?
18:49:09 <Deewiant> strip -s turned 4669 into 4888.
18:49:44 <Deewiant> Start of section headers: 120 (bytes into file)
18:49:47 <Deewiant> Start of section headers: 4696 (bytes into file)
18:49:48 <Deewiant> Start of section headers: 4696 (bytes into file)
18:49:55 <Deewiant> Anyway, that changed, at least.
18:50:18 <ehird> ais523: Right, that's it, I'm writing an optimized underload compirer.
18:50:19 <Deewiant> It also renamed STRTAB to .shstrtab, which adds a few bytes.
18:50:27 <ais523> ehird: optimized for what?
18:50:35 <ehird> ais523: Speed. Memory. Ponies.
18:50:57 <ais523> ehird: are you going to base it on the existing Underload compiler?
18:51:04 <ehird> No. I lost the code to that.
18:51:14 <ehird> I wanna call it overload in reference to overclocking but dammit that's taken :-)
18:51:24 * AnMaster tries on a non-multilib system
18:51:36 <ais523> please make it handle ((+):*:^):^ well
18:51:51 <ais523> you may want to add a ulimit when running that, though
18:52:23 <EgoBot> Supported commands: bf_txtgen help info 1l 2l adjust axo bch befunge befunge98 bf bf16 bf32 bf8 dimensifuck glass glypho kipple lambda lazyk linguine malbolge pbrain qbf rail rhotor sadol sceql trigger udage01 underload unlambda whirl
18:52:28 <ehird> ais523: (+)(:*:^):^ grows longer, no?
18:52:38 <EgoBot> +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
18:53:05 <ais523> you can't manage more than about 16 +s that way
18:57:27 <AnMaster> Deewiant, it is larger, because I can't strip that unwind info from the statically linked libraries
18:58:25 <AnMaster> Deewiant, yes but a large chunk of it is "useless" unwind info!
18:58:41 <Deewiant> And a large chunk of CCBI is "useless" TypeInfo!
18:59:14 <AnMaster> anyway 865KB for this full featured cfunge, stripped.
18:59:30 <AnMaster> 22 KB is unwind info from linked libraries
18:59:35 <AnMaster> cfunge 0.4.0 [+con +trace +exact-bounds +ncurses p:32 c:32]
18:59:45 <AnMaster> ais523, were you here when I added that
19:00:10 <ais523> I saw you developing it
19:00:12 <ais523> but not the finished product
19:00:36 <AnMaster> http://rafb.net/p/yVnwkh20.html
19:01:52 -!- WangZeDong has joined.
19:13:19 -!- Slereah has quit (Read error: 110 (Connection timed out)).
19:14:19 <AnMaster> I wish C had more exact ways to tell the compiler about aliasing
19:14:51 <AnMaster> ais523, rather crude. You can't say "this one will alias exactly this one, might alias that one, and won't alias anything else"
19:15:14 <ais523> yep, splint annoys me about that too
19:15:20 <ais523> because you can't specify aliasing precisely enough
19:15:38 <AnMaster> ais523, and you can't say "this pointer in your parameter list points to a block that is guaranteed to be aligned on a 16 byte boundary"
19:16:04 <ehird> when will people learn that c is fundamentally flawed
19:16:35 <ehird> yes, fundamentally
19:16:40 <AnMaster> For a start, what would you call cobol then
19:17:07 <olsner> at any rate, cobol is even more flawed fundamentally, but that doesn't make C unflawed
19:17:10 <ais523> which cobol, the mainframe lang or the esolang?
19:17:18 <AnMaster> if it is a *programming language* might be harder to know
19:17:26 <AnMaster> ais523, they are the same iirc
19:17:38 <ais523> there's an esolang called COBOL just to cause confusion
19:17:41 <ais523> based on playing cards, IIRC
19:17:53 <ehird> does anyone know why a failed login to a unix system lags a lot before telling you?
19:17:57 <ehird> a security measure/
19:18:07 <olsner> C has like 4 decades of workarounds, libraries, and coder experience to work around the flaws though
19:18:07 <ais523> yep, to prevent brute-forcing via a plugged-in keyboard
19:18:13 <ais523> well, keyboard simulator
19:18:20 <ais523> I think it's exactly 1 second of lag
19:18:25 <ehird> it's grrrrrrrrrrrrrrrrrrrrrrrr
19:18:33 <olsner> and bruteforce by telnet, obviously
19:18:56 <ais523> a good security cage is enough to make most unixy systems secure against even physical access, if they can't get through the cage
19:19:07 <ais523> arguably that isn't physical access, though
19:20:15 -!- Sgeo has joined.
19:20:27 <Deewiant> There's a couple-second sleep when the system shuts down after sending SIGTERM, too, even though it could continue early if all processes responded to the SIGTERM
19:21:14 <ais523> of course, if they slowhashed the passwords, there could be another reason for the lag
19:21:24 <ais523> but I don't think most Linux distros use slowhashing by defualt
19:22:03 <ais523> Sgeo: using a hash function that's really computationally expensive to compute
19:22:14 <ais523> so that even if someone gets read access to /etc/shadw somehow, bruteforcing is impractical
19:22:15 <Sgeo> Why would that be done?
19:23:11 <Sgeo> Do any Linux distros encrypt the files with the user's password?
19:23:15 <Sgeo> That would make sense, right?
19:23:21 <ais523> no, because they don't know what it is
19:23:47 <Sgeo> I meant the user's files, not the system files
19:23:51 <ais523> Ubuntu lets you encrypt keyrings, though
19:24:05 <ais523> and if you give the same password for them as your login password, then they'll be unlocked at login
19:24:08 <ais523> if you check the box to let you do that
19:24:18 <ais523> and then the passwords the files are actually encrypted with are in there
19:24:35 <AnMaster> (sorry had to go afk due to an optical failure)
19:24:53 <AnMaster> ais523, both cobol are esolangs IMO ;P
19:25:01 <AnMaster> one also happens to be a mainframe one as well
19:26:50 <AnMaster> http://en.wikipedia.org/wiki/COBOL#COBOL_2002_and_object-oriented_COBOL
19:27:55 <AnMaster> <ehird> does anyone know why a failed login to a unix system lags a lot before telling you?
19:28:35 <AnMaster> ehird, but why did it fail in the first place
19:28:52 <AnMaster> you don't typo your password a lot in general
19:33:56 <oerjan> <AnMaster> (sorry had to go afk due to an optical failure) <<< sorry to hear you broke your glasses
19:36:50 -!- EgoBot has quit (Remote closed the connection).
19:36:57 -!- EgoBot has joined.
19:37:41 <GregorR> Think plash is secure enough to run arbitrary C code? X-P
19:38:05 <ehird> GregorR: just run them in user mode linux
19:38:18 <GregorR> Yeah, that would be super, boot a UML every time X-P
19:38:29 <ehird> Deewiant: but is massively complex
19:38:42 <ehird> GregorR: Just reboot it when it starts doing weird shit
19:38:43 <Deewiant> ehird: What did you expect, it's C++ after all ;-)
19:39:03 <ehird> Deewiant: technically it's written in haskell
19:39:13 <Deewiant> Yes, but it has to deal with C++
19:39:37 <Deewiant> FWIW it also provides some handy C++-related tools which have nothing to do with running C++, which add to the complexity
19:42:55 -!- ehird_ has joined.
19:43:03 <ehird_> Yep, I still like this client.
19:43:08 <GregorR> !c printf("Hello, world!\n");
19:43:16 <ehird_> !c malloc(4587348957345345)
19:43:25 <GregorR> It still has all the same limits :P
19:43:33 <ehird_> !c printf("%i\n",1/0);
19:44:05 <GregorR> Just for ehird_, I'm outputting stderr to stdout :P
19:44:19 <GregorR> Or, y'know, failing to, whatever.
19:44:21 <EgoBot> <stdin>:8: error: expected identifier or '(' before 'return'
19:44:31 <ehird_> GregorR: Use clang, foo
19:45:00 <GregorR> Argh, where's my stderr X-P
19:45:06 <GregorR> !c fprintf(stderr, "Hello, world!\n");
19:45:06 <EgoBot> <stdin>: In function 'main':
19:45:25 <ehird_> GregorR: Clang is awesome
19:45:32 <GregorR> ehird_: It compiles to llvm, no?
19:45:47 <ehird_> It compiles cfunge, doesn't it AnMaster?
19:45:51 <ehird_> Therefore it can compile everything.
19:46:04 <GregorR> As far as I'm concerned, llvm is just yet one more environment to support ... unless llvm has nifty sandboxing support.
19:46:26 <ehird_> GregorR: You do know that llvm compiles to machine code, right?
19:46:43 <ehird_> It's just an intermediate step in the compilation process.
19:46:46 <GregorR> ehird_: I thought it was a JIT ... >_>
19:46:59 <GregorR> !c system("ls /home/egobot/");
19:47:11 <ehird_> !c system("rm -rf /home/egobot")
19:47:26 <GregorR> If that breaks, it's totally my fault X-P
19:47:33 <ehird_> !c system("ls /home/egobot")
19:47:52 <ehird_> !c system("http://4chan.org/")
19:47:55 <AnMaster> at least svn as of a few weeks ago
19:47:56 <ehird_> !c system("curl http://4chan.org/")
19:48:12 <GregorR> ehird_: That would work if curl was installed X-P
19:48:44 <oerjan> !c system("mv /home/egobot/egobot.hg /home/egobot/boo")
19:48:45 <EgoBot> Supported commands: help info 1l 2l adjust axo bch befunge befunge98 bf bf16 bf32 bf8 c dimensifuck glass glypho kipple lambda lazyk linguine malbolge pbrain qbf rail rhotor sadol sceql trigger udage01 underload unlambda whirl
19:48:50 <ehird_> !c system("ls /home/egobot")
19:49:00 <ehird_> !c system("sudo rm -rf /")
19:49:12 <ehird_> oerjan: that's not a very oerjan smiley :o
19:49:15 <GregorR> ehird_: You must think that plash is the most useless system ever :P
19:49:22 <ehird_> Gracenotes: But of course.
19:49:39 <ehird_> !c system("telnet irc.freenode.net 6667")
19:49:41 <ais523> more to the point, that wouldn't work even on a completely unsecured system
19:49:56 <ais523> how would you enter the password?
19:50:10 <AnMaster> ais523, btw, so you can coordinate for ick release, plan is to release a new stable cfunge either this evening or tomorrow or soon after. I realised that that -DCFUN_IS_IFFI changed the API so I'm going to increment the API version and make IFFI handle it (so soon you will have to pull from my ick branch again)
19:50:18 <oerjan> !c system("more magic")
19:50:31 <ais523> I wonder how the sandboxing is done?
19:50:41 <GregorR> ais523: http://google.com/search?q=plash
19:51:52 <AnMaster> but I would have to read scrollback then
19:51:59 <AnMaster> instead I'm going to code stuff
19:53:29 <oerjan> !c int i = 42; printf("%d\n", i);
19:54:26 <ehird_> 19:51 AnMaster [n=AnMaster@unaffiliated/anmaster] requested CTCP VERSION from ehird_:
19:54:35 <ehird_> ais523 may know it; it's in the debian/ubuntu repos.
19:54:51 <ais523> memorising the whole of the debian/ubuntu repos would be ridiculous
19:55:00 <AnMaster> err, do you call a -DFOO=bar a "macro" or is it macro only if it takes parameters
19:55:04 <AnMaster> I mean, what is the official term
19:55:21 <ais523> !c printf("%d",(int)__LINE__);
19:55:31 <ehird_> AnMaster: Smuxi. It's an IRC client for Gnome.
19:55:36 <GregorR> Oh, right, each #include is on its own line
19:55:36 <AnMaster> ais523, well you are the C expert here.
19:55:45 <ais523> AnMaster: I thought you were
19:55:52 <AnMaster> ais523, not C standard expert!
19:55:56 <ais523> and it's a preprocessor definition, technically
19:56:27 <AnMaster> "[...] must now define the below listed preprocessor definitions to strings with the relevant values"
19:56:43 <ais523> just because a term's used in a standard doesn't mean you have to /use/ it
19:56:47 <ais523> especially when it's unwieldy
19:57:36 <AnMaster> ais523, what would you recommend instead
19:57:39 <ehird_> AnMaster: your sentence is messy
19:57:53 <ehird_> "must now define the below listed" *Brain turns off*
19:58:21 <ais523> AnMaster: put an example there
19:58:26 <ais523> and ask people to edit it to suit their purposes
19:58:53 <oerjan> !c printf("%d", fac(5)); return 0; } int fac(int n) {if (n) return n*fac(n-1); else return 0;
19:59:40 <ehird_> "The application has to be able to decrypt the password in order to authenticate users when they login." <-- >_<
19:59:44 <GregorR> I'm fekking around with it :P
19:59:59 <ais523> !c FILE*in=fopen(argv[0],"rb"); while(!feof(in)) printf("%x",(int)getc(in));
20:01:10 <ais523> !c FILE*in=fopen(argv[0],"rb"); while(!feof(in)) printf("%x",(int)getc(in));
20:01:21 <ais523> hmm, why isn't that working?
20:01:25 <AnMaster> ais523, it is a long list of defines with descriptions. I'll pastebin it so you can find a better way.
20:01:31 <ehird_> !c you lied to me when you told me this was a program
20:01:41 <EgoBot> 7f454c46211000000000203e01000d04400000040000000060120000000000400380804002502206000500040000000040040000004004000000c01000000c010000008000000030004000020000000240000000240000001c00000001c0000000100000001000500000000000004000000004000000bc7000000bc700000000200000010006000c07000000c076000000c07600000020200000030200000000200000020006000e87000000e876000000e876000000a01000000a0100000080000000400040001c20000001c240000001c240000002000000002000000004000000050
20:01:42 <AnMaster> ais523, http://rafb.net/p/gmZTe577.html
20:01:48 <ais523> oh, it did work, just slowly
20:01:56 <ais523> that's rather sparse, probably it's the header
20:01:57 <AnMaster> ais523, that is not spell checked at all
20:02:26 <ais523> "Please define the following to string constants:"
20:02:43 <ehird_> Please #define the following as string constants
20:02:52 <ehird_> mainly as instead of to, but #define is clearer
20:03:16 <GregorR> !c int i=1;FILE*in=fopen(argv[0],"rb");for(;!feof(in);i++)printf("%02x%s",getc(in),(i%20)==0?"\n":"");
20:03:17 <EgoBot> 7f454c4602010100000000000000000002003e00
20:03:24 <oerjan> !c printf("%d\n", fac(5)); return 0; } int fac(int n) {if (n) return n*fac(n-1); else return 0;
20:03:29 <GregorR> Now EgoBot is spamming me.
20:03:39 <ais523> GregorR: yep, it returns multilines in /msg, doesn't it?
20:03:40 <oerjan> !c printf("%d\n", fac(5)); return 0; } int fac(int n) {if (n) return n*fac(n-1); else return 1;
20:03:47 <ais523> and !kill and !ps doesn't seem implemented atm
20:03:50 <GregorR> It'll get d/c'd pretty soon :P
20:03:52 <AnMaster> ais523, that misses the point of "you only need to worry if you aren't using cmake"
20:04:19 <ehird_> AnMaster: "If you are not building with cmake, #define these as string constants:"
20:04:23 <AnMaster> "Also any code using another build system than cmake (such as IFFI) must now define [..]"
20:05:35 * GregorR is a little bit surprised that EgoBot hasn't been dropped yet :P
20:06:10 -!- EgoBot has quit (Excess Flood).
20:06:10 <GregorR> It's still spamming the binary at me :P
20:06:13 -!- EgoBot has joined.
20:07:41 <AnMaster> GregorR, will it completely send the previous reply before starting next
20:07:45 <EgoBot> Supported commands: bf_txtgen help info 1l 2l adjust axo bch befunge befunge98 bf bf16 bf32 bf8 c dimensifuck glass glypho kipple lambda lazyk linguine malbolge pbrain qbf rail rhotor sadol sceql trigger udage01 underload unlambda whirl
20:07:57 <AnMaster> then I guess I made an error there
20:08:05 <oerjan> AnMaster: it answered mine while still spamming GregorR
20:08:07 <GregorR> !c printf("Hello, world!")
20:08:16 <coppro> why is there malbolge? Is it even possible to fit a useful malbolge program in one message?
20:08:24 <ehird_> coppro: !malbolge http://foo
20:08:25 <GregorR> Either that or befunge98 is broken in there right now :P
20:08:42 <ehird_> !malbolge (=<`$9]7<5YXz7wT.3,+O/o'K%$H"'~D|#z@b=`{^Lx8%$Xmrkpohm-kNi;gsedcba`_^]\[ZYXWVUTSRQPONMLKJIHGFEDCBA@?>=<;:9876543s+O<oLm
20:09:08 <AnMaster> !befunge98 'A aaa**k: aaa**k,a, @
20:09:09 <EgoBot> AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
20:09:26 <AnMaster> GregorR, so you don't add linebreaks
20:09:47 <AnMaster> !befunge98 'A aaaa***k: aaaa***k,a, @
20:10:33 <GregorR> I can't see that output easily.
20:10:43 <EgoBot> AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
20:11:22 <GregorR> EgoBot does slow everything down by insane degrees :P
20:12:06 <AnMaster> why didn't it output twice before
20:12:13 -!- ehird_ has quit (Remote closed the connection).
20:12:18 <AnMaster> 21.10:42 <AnMaster> !befunge98 aaaa***. @ never got a reply
20:12:37 <AnMaster> 21.11:01 <AnMaster> !befunge98 aaaa*** . a, @ got a reply
20:12:44 <AnMaster> and so did <AnMaster> !befunge98 aaaa*** . @ later
20:12:58 <AnMaster> GregorR, not even on what slox does
20:13:35 <GregorR> slox keeps a process from using more than a requested percentage of the CPU.
20:13:49 <ehird> GregorR: set up a reverse dns for that codu
20:13:53 <ehird> it comes as codu.xen.prgmr.com
20:14:07 <GregorR> ehird: Yeah, I know, I need to. I assume it involves talking to humans though, and that sucks ;)
20:14:19 <AnMaster> GregorR, ok. How does it do it
20:15:15 <AnMaster> GregorR, um. cfunge has a compile time option to set a 3 second alarm() at startup.
20:15:39 <ais523> is SIGSTOP the right one? I would have assumed SIGKILL would be better
20:15:47 <GregorR> I don't /think/ that that should be affected by SIGSTOP.
20:15:53 <ais523> it's about CPU throttling
20:16:00 <ais523> so you have to be able to start again
20:16:14 <GregorR> It's super-lame CPU throttling ^^
20:16:24 <AnMaster> GregorR, why not use linux built in stuff for it
20:16:37 <Deewiant> You can't have precise control with nice
20:16:45 <ais523> !underload (()(*))(~:^:S*a~^a~!~*~:(/)S^):^
20:16:48 <AnMaster> Deewiant, do you need that precise control
20:16:56 <Deewiant> If you want to do your own scheduling SIGSTOP is an easy way to get it
20:17:08 <Deewiant> AnMaster: I don't know if he does, but there are cases in which you do.
20:17:20 <AnMaster> Deewiant, I don't think SIGSTOP is safe
20:17:34 <AnMaster> what if it happens in the middle of a poll() or such
20:17:35 <GregorR> There are situations where you don't want a program to use much CPU even if the CPU is "free"
20:17:48 <ais523> AnMaster: you are using sigselect(), right?
20:17:55 <ehird> fizzie: does Debian's advanced install let you do kooky lvm stuff?
20:18:00 <Deewiant> AnMaster: poll can be interrupted by signals just like any other system call?
20:18:01 <GregorR> AnMaster: It's defined, but plenty of programs don't handle it properly.
20:18:01 <ehird> like, obscure options to lvm
20:18:04 <AnMaster> ais523, um you can't ignore STOP
20:18:14 <AnMaster> and it is pselect not sigselect
20:18:31 <GregorR> (Where by "plenty" I mean "none that I've found", but anyway :P )
20:18:51 <AnMaster> GregorR, I think it depends on very precise timing
20:18:55 <AnMaster> to get it to show said issues.
20:19:33 <GregorR> Suffice it to say that I don't trust Linux's scheduler enough to allow these programs to take 100% CPU *shrugs*
20:20:14 <GregorR> I find that setting the UID of a process rarely changes its scheduling priority :P
20:20:15 <AnMaster> /sys/kernel/uids/<uid>/cpu_share
20:20:45 <GregorR> Doesn't look like I have that.
20:21:02 <AnMaster> GregorR, it can use more, but setting them for different UIDs mean that some UIDs will be guaranteed more under load.
20:21:10 <AnMaster> like per user priority kind of
20:21:33 <fizzie> ehird: I'm don't think it does "kooky" by default, but you can start a shell to mess up with things during the install.
20:21:43 <AnMaster> Gracenotes, 2.6.23 or something like that
20:21:51 <ehird> fizzie: I did that, then it wanted to reformat them :)
20:22:04 <ehird> fizzie: It seems Debian doesn't have new enough kernel for ext4
20:22:06 <AnMaster> GregorR, then I guess it was disabled in kernel config
20:22:42 <AnMaster> ehird, 2.6.28, some older ones have ext4dev
20:22:53 <ehird> AnMaster: debian is 2.6.27
20:23:11 <fizzie> ehird: sid has 2.6.29 currently.
20:23:15 <AnMaster> ehird, try another distro, or build a newer one
20:23:25 <ehird> it'll be in testing by the time I get my machine then
20:23:45 <AnMaster> ehird, so you didn't like arch :(
20:23:55 <ehird> AnMaster: ECOULDN'TINSTALL.
20:24:06 <ehird> EWASTURNEDOFFBYINSTALLERANYWAY.
20:24:18 <ehird> AnMaster: EPEBVIRTUALMACHINEANDARCH,LOSER.
20:25:05 <AnMaster> ehird, EUSERNOTBOOTABLEINSERTNEWUSERANDPRESSANYKEY
20:25:31 <ehird> So the Arch philosophy is "if you can't install it on a machine, it's your fault, always".
20:25:35 <ehird> Great, I'll stick to Debian.
20:25:56 <AnMaster> ehird, but I still think it was PEBKAC
20:26:04 <AnMaster> I never heard of anyone with similar issues
20:26:05 <ehird> Fuck that, I did exactly what it said.
20:26:09 -!- bsmntbombdood has quit (Read error: 104 (Connection reset by peer)).
20:26:15 <AnMaster> ehird, what exact error message did it give
20:26:21 <ehird> AnMaster: I don't remember, goddamn
20:26:30 <AnMaster> you never mentioned back then either
20:26:33 <ehird> But I can install slackware with my hands tied, I think I could get Arch right
20:26:49 -!- bsmntbombdood has joined.
20:26:55 <AnMaster> ehird, hands tied behind your back or in front of you
20:27:08 <ehird> I need access to a keyboard and a screen.
20:27:09 <AnMaster> and slackware isn't hard to install
20:27:25 <ehird> It's harder than Debian
20:27:57 <ehird> darn. grub can't boot ext4 systems that use extents
20:27:58 <AnMaster> ehird, err, I had more issues installing debian than slackware. Both were about a year and a half a ago
20:28:10 <ehird> AnMaster: PEBKAC PEBKAC PEBKAC
20:28:12 -!- bsmntbombdood has quit (Read error: 104 (Connection reset by peer)).
20:28:31 <AnMaster> ehird, and. Don't use ext4 for your /boot
20:28:37 <ehird> You're an idiot and so's your mother and you should never be allowed to install Linux again if an installer failed for you.
20:28:42 <AnMaster> why are you allergic to separate partitions
20:28:43 <ehird> Also, I need to use ext4 for my /boot.
20:28:51 <AnMaster> ehird, why do you need ext4 on it
20:28:56 <AnMaster> I mean you won't access it a lot
20:28:57 <ehird> SSD alignment concerns.
20:29:09 <AnMaster> ehird, you won't access it a lot still.
20:29:23 -!- sebbu2 has joined.
20:29:26 <AnMaster> and you need just 32 MB to have plenty of free space on it
20:29:26 <ehird> I don't feel like fitting it in with other, properly aligned, partitions
20:29:55 <Deewiant> ehird: Do you need extents for it to be aligned properly?
20:30:09 <AnMaster> my boot on gentoo is 9.9 MB, on my arch 20 MB
20:30:10 <ehird> Deewiant: Probably not, but I cba to turn them off. :)
20:30:19 <AnMaster> but I have a distro generic fallback kernel on arch
20:31:15 <Deewiant> AnMaster: Is that used space or the whole partition size
20:32:18 <AnMaster> Deewiant, whole space is 32 MB on gentoo and 64 MB on Arch. Which is funny since the gentoo one is 64-bit and the arch one 32-bit
20:32:28 <AnMaster> Deewiant, I have multiple kernels
20:32:41 <AnMaster> not just the arch generic ones. But also my own ones
20:33:09 <AnMaster> gentoo one only have two custom ones
20:33:18 <AnMaster> current version, and new version to boot to next
20:33:24 <AnMaster> sometimes it has three versions
20:33:49 <AnMaster> in a complex system, see logs of channel from 2009 for more info
20:34:00 <Deewiant> Hmm, system.map isn't required, is it?
20:34:14 <Deewiant> Isn't it just for kernel debugging
20:34:17 <ehird> 20:33 AnMaster: in a complex system, see logs of channel from 2009 for more info ← how useless
20:34:24 <AnMaster> Deewiant, not for booting no. But it is needed to interpret kernel panics iirc
20:34:32 <AnMaster> since they tend to just contain addresses
20:34:44 <Deewiant> Yeah, so if I don't care about interpreting kernel panics I don't need the map
20:35:02 <AnMaster> I keep 2.6.x and 2.6.y always, where x is last one, and y is current.
20:35:05 <Deewiant> Isn't it recreatable by recompiling anyway
20:35:23 <AnMaster> when I change from 2.6.y to 2.6.y.z I don't remove 2.6.x
20:35:31 <AnMaster> I always keep one from that level
20:36:21 <AnMaster> Deewiant, moving /boot/old-configs to /root/old-configs reduced it to 8.5 MB
20:36:33 <AnMaster> old-configs contains configs all the way back from 2.6.9
20:36:47 <AnMaster> 2.6M /boot/kernel-2.6.27-gentoo-r8-L1
20:36:47 <AnMaster> 2.7M /boot/kernel-2.6.28-gentoo-r5-L1
20:37:26 <AnMaster> Deewiant, I used to have it at 2.2 before, but I needed more options later
20:37:48 <AnMaster> like usb audio for connecting usb keyboard. Parts of that I couldn't build as module.
20:38:07 <Deewiant> Hmm, I wonder where all that space is going given that du reports only 5.6M
20:38:09 <AnMaster> (keyboard meaning electrical piano, not qwerty here)
20:38:43 <AnMaster> Deewiant, there is stuff I only use very seldom
20:38:47 <Deewiant> AnMaster: Doesn't du report block-sized anyway, except with --apparent-size
20:38:47 <AnMaster> why would I want it compiled in
20:38:52 -!- sebbu3 has joined.
20:38:59 <Deewiant> Why not, the kernel is only 2.4M anyway :-P
20:39:02 <AnMaster> Deewiant, err yes that is what I said
20:39:23 <Deewiant> AnMaster: Yeah, so why does df report 4M on top of the 5.6M of du
20:39:24 <AnMaster> Deewiant, 6.6M for /lib/modules/2.6.28-gentoo-r5-1
20:39:37 <AnMaster> a lot more for /lib/modules/2.6.27-gentoo-r8-1, since it contains the nvidia module too
20:39:47 <AnMaster> which I haven't built for new kernel yet
20:40:08 <Deewiant> Which is something that couldn't be disabled
20:40:18 <Deewiant> So I do have 2.8K's worth of modules ;-)
20:40:26 <AnMaster> Deewiant, err scsi_wait_scan.ko can be disabled
20:40:40 <Deewiant> Comes with the whole SATA system IME.
20:40:43 <AnMaster> 4M + 394K, you are a bit closer
20:40:47 <AnMaster> maybe there is some other file there too
20:40:56 <Deewiant> AnMaster: No, the 5.6M included every single file+directory.
20:41:12 <AnMaster> Deewiant, did the 4.4 + 394K then
20:41:30 <AnMaster> and 4.4 + 394K == kernel + grub
20:41:57 <AnMaster> Deewiant, I have SATA system. I don't have scsi_wait_scan.ko
20:42:23 <Deewiant> AnMaster: Total for multiple files.
20:42:53 <AnMaster> Deewiant, some distros install them
20:43:07 <AnMaster> Deewiant, to make you able to write /boot/foo in grub config instead of /foo
20:43:15 <AnMaster> thus less bug reports from users
20:44:24 -!- sebbu has quit (Read error: 110 (Connection timed out)).
20:44:26 <Deewiant> menuconfig says SCSI_WAIT_SCAN = m
20:44:27 <AnMaster> you don't have that one I guess
20:44:36 <Deewiant> No, I have my own backup which is also 2.4M.
20:44:55 <AnMaster> so you don't have arch fallback ones installed then
20:45:16 <AnMaster> Deewiant, disable CONFIG_SCSI_SCAN_ASYNC
20:45:53 <Deewiant> Although I think I'll enable it, that sounds good. :-P
20:45:55 <AnMaster> Deewiant, I don't even have the option for SCSI_WAIT_SCAN there...
20:46:05 <AnMaster> Deewiant, CONFIG_SCSI_SCAN_ASYNC is why you need SCSI_WAIT_SCAN
20:46:41 <AnMaster> │ Symbol: SCSI_WAIT_SCAN [=n] │
20:46:52 <GregorR> !cxx cout << "Hello, world!" << endl
20:47:04 <ehird> GregorR: using that Georgi thing?
20:47:17 <GregorR> ehird: No, just g++ wrapped in the same gunk as usual.
20:47:25 <ehird> !cxx cout << [](){"A PACKAGE FOR YOU!"}
20:47:27 <EgoBot> /tmp/source.17577.cc: In function 'int main(int, char**)':
20:47:44 <Deewiant> AnMaster: Disabled everything under SCSI device support and it's still =m. Couldn't disable CONFIG_SCSI, though.
20:48:02 <AnMaster> GregorR, why did you include the namespace std:: by default
20:48:04 <ehird> GregorR: Well, it will.
20:48:08 <ehird> AnMaster: IRC line shorter.
20:48:18 <ehird> your definition of funny sucks :P
20:48:27 <EgoBot> /tmp/source.17622.cc: In function 'int main(int, char**)':
20:48:36 <ehird> 20:48 EgoBot: /usr/include/c++/4.3/istream:123: note: candidates are: std::basic_istream<_CharT, _Traits>& std::basic_istream<_CharT, _Traits>::operator>>(std::basic_istream<_CharT, _Traits>& (*)(std::basic_istream<_CharT, _Traits>&)) [with _CharT = char, _Traits = std::char_traits<char>] <near match>
20:48:41 <ehird> 20:48 EgoBot: /usr/include/c++/4.3/istream:127: note: std::basic_istream<_CharT, _Traits>& std::basic_istream<_CharT, _Traits>::operator>>(std::basic_ios<_CharT, _Traits>& (*)(std::basic_ios<_CharT, _Traits>&)) [with _CharT = char, _Traits = std::char_traits<char>] <near match>
20:48:50 <ehird> 20:48 EgoBot: /usr/include/c++/4.3/istream:134: note: std::basic_istream<_CharT, _Traits>& std::basic_istream<_CharT, _Traits>::operator>>(std::ios_base& (*)(std::ios_base&)) [with _CharT = char, _Traits = std::char_traits<char>] <near match>
20:48:51 <AnMaster> more harmonic. More like the error messages
20:49:05 <ehird> 20:49 EgoBot: /usr/include/c++/4.3/istream:242: note: std::basic_istream<_CharT, _Traits>& std::basic_istream<_CharT, _Traits>::operator>>(std::basic_streambuf<_CharT, _Traits>*) [with _CharT = char, _Traits = std::char_traits<char>] <near match>
20:49:13 <AnMaster> Deewiant, what are the deps of the symbol
20:49:22 <AnMaster> Deewiant, xconfig allows you to see it easily
20:49:24 <Deewiant> AnMaster: SCSI_WAIT_SCAN? How would I know?
20:49:34 <AnMaster> Deewiant, Show debug info in make xconfig
20:49:36 <GregorR> !cxx cout << [](){"A PACKAGE FOR YOU!"}
20:49:37 <EgoBot> /tmp/source.17673.cc: In function 'int main(int, char**)':
20:49:44 <AnMaster> Deewiant, I don't think menuconfig supports that
20:49:45 -!- bsmntbombdood_ has joined.
20:49:49 <GregorR> Hrm, I added -std=gnu++0x ...
20:49:51 <Deewiant> AnMaster: I've only ever used menuconfig and oldconfig, didn't even know about this :-P
20:49:56 <ehird> GregorR: does gcc do it?
20:50:08 <AnMaster> Deewiant, it has "show all options" which let you see all greyed out options too
20:50:12 <GregorR> ehird: Idonno exactly what g++ does and doesn't support, but it certainly supports /something/
20:50:22 <AnMaster> Deewiant, those things are in the option menu iirc
20:50:32 <Deewiant> AnMaster: ctrl+f doesn't find scsi_wait_scan...
20:50:42 <AnMaster> Deewiant, maybe case sensitive
20:50:53 <ehird> GregorR: add machine code!
20:51:01 <AnMaster> Deewiant, show all options first
20:51:04 <Deewiant> Or maybe it is but it didn't find it
20:51:06 -!- bsmntbombdood_ has quit (Read error: 104 (Connection reset by peer)).
20:51:07 <Deewiant> AnMaster: I did, and debug info
20:51:27 <Deewiant> Enabled everything under option
20:51:43 -!- bsmntbombdood_ has joined.
20:52:05 * ehird puts http://mastodon.biz/ in a vm
20:52:08 <Deewiant> Finds plenty for SCSI but nothing for WAIT
20:52:39 <Deewiant> I set it to =n in the .config but it got overridden
20:52:44 <AnMaster> Deewiant, disable the modules option
20:52:59 <AnMaster> Deewiant, since you didn't use modules anyway
20:53:05 <GregorR> ehird: Give me a header for a .asm file :P
20:53:24 <Deewiant> I wonder if it can be built as =y though, for the async
20:53:40 <AnMaster> Deewiant, if you don't have modules you don't need it
20:53:47 <AnMaster> and it is a user visible option in older kernels
20:53:53 <Deewiant> │ scsi_wait_scan module to ensure that all scans have completed."
20:54:09 <Deewiant> Oh, maybe I should read the whole paragraph
20:54:29 -!- sebbu2 has quit (Connection timed out).
20:54:45 <AnMaster> Deewiant, also I don't know if it has any effect on sata
20:54:53 <ehird> Ancient Linux GOOOOOOOOOOOOOOOOOOOOOO
20:55:05 <ais523> !underload (:aS(:^S^:)Sa:):^S^:(:aS(:^S^:)Sa:)
20:55:06 <EgoBot> (:aS(:^S^:)Sa:):^S^:(:aS(:^S^:)Sa:)
20:55:07 <ehird> http://mastodon.biz/
20:55:12 <ehird> an old distro w/ bsd userland
20:55:18 <ehird> gcc 2, linux 2.0.x
20:55:25 <ehird> 7 years old at the time of that writing
20:55:28 <ehird> which I imagine was a few years ago
20:55:40 <ehird> it mentions ISDN lines in the download page
20:55:40 <AnMaster> "The most recent version is INST0066, which is no longer all that recent (~7 years old), but I'm trying to decide whether to roll to one of the super-bloated newer Linux kernels or write my own USB stack plus SATA and UDMA drivers for 2.0.28."
20:55:58 <ehird> he's not insane, he just likes minimal sw
20:56:03 <ehird> ion author is a raving lunatic
20:56:40 <ehird> AnMaster: it even has a netinst!
20:56:41 <AnMaster> gcc no longer supports it nowdays. Since 4.4 it is obsolete
20:56:47 <ehird> also, I don't think he cares
20:57:00 <ehird> nothing depends on gcc 4 afaik
20:57:10 <ehird> he probably wouldn't use any gpl software.
20:57:17 <ehird> 1MB netinst .img boot GO!
20:57:32 <ehird> aww, no bootable medium found
20:57:35 <ehird> guess .img isn't .iso
20:58:00 <AnMaster> ehird, so you can't use it on your new computer
20:58:01 <ehird> damn that brings some memories
20:58:08 <ehird> AnMaster: I'll probably still a floppy on there
20:58:23 <ehird> AnMaster: that's what it is
20:58:28 <ehird> a floppy "livecd" that netinstalls
20:58:34 <AnMaster> ehird, you said it brings memories
20:58:34 <ehird> pretty modern for 7+n years ago
20:58:45 <ehird> never, just floppy linux
20:59:15 <ehird> I attempted to run every linux distro I could find in like 2004-2005
20:59:19 <ehird> Mostly shitty floppy ones
20:59:24 <AnMaster> nowdays even the bz2 image of the kernel wouldn't fit on a floppy
20:59:33 <AnMaster> but I guess if you disabled some stuff
20:59:37 <Deewiant> AnMaster: Super-bloated, like he said
20:59:40 <ehird> The first OS I tried when I got my AWESOME NEW ETHERNET ROUTER no more winmodem I can use other OSes yay!! was QNX
21:00:09 <AnMaster> so what did you think about it
21:00:25 <ehird> AnMaster: it did not live up to my high expectations, I thought it was a desktop OS.
21:00:54 <AnMaster> ehird, actually it does have a desktop
21:00:55 <ehird> AnMaster: hey, it did look pretty
21:01:05 <ehird> the gui wasn't unmodern, it just had no apps
21:01:18 <ais523> ehird: could you recompile apps for QNX, I wonder/
21:01:24 <ehird> ais523: it's posix
21:01:40 <ehird> AnMaster: it's just too retro for virtualbox!
21:01:50 <ehird> here's the wonderful splash screen I got:
21:02:03 <ehird> (WARNING: YOUR EYES WILL NEVER FORGIVE YOU)
21:02:16 <AnMaster> Deewiant, worked under qemu yes., Decided not to try under vmware
21:02:29 <ehird> even imgur can't take it
21:02:37 <ehird> AnMaster: i don't put files in shock-site hosters
21:02:49 <AnMaster> ehird, but you said it was a shock pic
21:02:57 <AnMaster> and not shock site hoster, just shock pic
21:02:58 <ehird> it's awful but not shocking :P
21:03:26 <GregorR> !asm movl $72, %edi; call putchar; movl $105, %edi; call putchar; movl $10, %edi; call putchar
21:03:36 <ehird> AnMaster: http://omploader.org/vMW00OQ
21:03:54 <ehird> it faded that in before booting the kernel
21:03:58 <AnMaster> ehird, aiie emebdded color profile
21:04:06 <ehird> AnMaster: os x does that :)
21:04:22 <ehird> and thus it died again; let's try... um ...
21:04:45 <Deewiant> !asm .intel_syntax; mov edi, 72; call putchar; mov edi, 10; call putchar
21:04:45 <EgoBot> /tmp/source.17919.s: Assembler messages:
21:05:28 <AnMaster> GregorR, so you prefer AT&T syntax too then
21:05:28 <GregorR> Deewiant: I can fix that. What's the directive to switch back to AT&T syntax?
21:05:45 <Deewiant> !asm .intel_syntax; mov edi, byte 72; call putchar; mov edi, byte 10; call putchar; .att_syntax
21:05:46 <EgoBot> /tmp/source.17963.s: Assembler messages:
21:06:08 <ehird> LET'S GOOOOOOOOOOO
21:06:10 <AnMaster> ehird, bochs got an awesome debugger though.
21:06:13 <ehird> -OOOOOOOOOOOOOOOOOOOOOOOOOOOOO-
21:06:16 <ehird> -OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO-
21:06:17 <Deewiant> !asm .intel_syntax; movzx edi, 72; call putchar; movzx edi, 10; call putchar; .att_syntax
21:06:17 <EgoBot> /tmp/source.18042.s: Assembler messages:
21:06:19 <ehird> -LAAAAAAAAAAAAAAGOOOOO-
21:06:36 <AnMaster> ehird, and wasn't bochs originally IOCCC entry
21:06:37 <GregorR> Apparently Deewiant can't even remember Intel syntax :P
21:06:54 <Deewiant> GregorR: I don't move constants into registers too often :-P
21:07:16 <Deewiant> !asm .intel_syntax; mov byte edi, 72; call putchar; mov byte edi, 10; call putchar; .att_syntax
21:07:17 <EgoBot> /tmp/source.18084.s: Assembler messages:
21:07:18 <ehird> How much disk do you think mastodon needs?
21:07:38 <Deewiant> GregorR: And FWIW "mov edi, 72" would work in FASM
21:07:50 <ais523> !underload (-)S(O)(~:S:*(- -)S~:^):^
21:08:01 <Deewiant> But it complains about ambiguous operand size
21:08:16 <AnMaster> ehird, qemu was related. Just not directly
21:08:24 <ehird> 21:07 ehird: How much disk do you think mastodon needs?
21:08:33 <Deewiant> !asm .intel_syntax; mov byte ptr edi, 72; call putchar; mov byte ptr edi, 10; call putchar; .att_syntax
21:08:34 <AnMaster> Binary file 2004/fs.tar matches
21:08:34 <AnMaster> 2004/gavin.hint: see http://bellard.org/ for QEMU (Fabrice Bellard is an IOCCC 2001 winner),
21:08:34 <EgoBot> /tmp/ccclhLmE.o: In function `main':
21:08:49 <Deewiant> ais523: Nope, undefined reference to edi
21:08:51 <ais523> !underload (-)S(O)(~:S:*(- -)S~:^):^
21:08:53 <AnMaster> Deewiant, att syntax is a lot faster to get it right in
21:09:00 <Deewiant> GregorR: And the directive was .att_syntax, incase you didn't notice :-P
21:09:10 <Deewiant> AnMaster: "mov edi, 72" should work.
21:09:32 <AnMaster> Deewiant, um, you mean mov $72, %edi surely?
21:09:42 <Deewiant> AnMaster: No, I mean intel syntax. I don't speak $%!@#%!@#%!
21:09:50 <AnMaster> Deewiant, I don't speak intel syntax
21:09:56 <ais523> hmm... obviously it isn't capturing stderr
21:09:57 <AnMaster> ais523, do you prefer Intel or AT&T syntax
21:10:03 <Deewiant> AnMaster: Like AT&T, but remove the ^$#!%!@ and flip operands.
21:10:14 <ais523> AnMaster: I grew up on Intel, but I haven't used asm recently enough to express a preference
21:10:14 <Deewiant> And lose the suffixes on instructions.
21:10:18 <ehird> AnMaster: Holy shit, mastodon has an ncurses installer!
21:10:18 <ais523> although gcc-bf is AT&T syntax
21:10:21 <AnMaster> Deewiant, no you add ^$#!%!@ and flip operands to get Intel
21:10:31 <ehird> err... intel has no such (#*$%*(@#
21:10:33 <Deewiant> AnMaster: No, there's no ^$#@!!@#$
21:11:32 <Deewiant> Why can't that crap understand 'mov edi, 72'.
21:11:52 <ehird> Choose the web/ftp site to install Mastodon 0066 from
21:11:56 <ehird> Pell (Oregon, USA)
21:12:19 <ehird> Admittedly, it freezes then
21:13:44 <ehird> Better download the real OS beta
21:13:47 <ehird> Not the netinstall
21:14:21 <ehird> ha, it comes with adobe acrobat
21:14:48 <ais523> sudo acrobat # watch as my system gets infected
21:15:09 <ais523> there's a cross-platform zero-day bug in latest Acrobat
21:15:18 <ais523> that lets PDF Javascript take over the system
21:15:30 <ais523> obviously the payload's different on different OSes, but the bug's the same one
21:15:34 <ais523> Deewiant: well, that's the obvious solution, yes
21:15:46 <ehird> ais523: it's adobe 4
21:15:54 <ehird> and it's not acrobat any more
21:16:04 <ais523> oh, not "acrobat reader"
21:18:02 <GregorR> !c printf("Hello, world!\n");
21:18:20 <AnMaster> Deewiant, did you use Intel or AT&T first
21:18:36 <AnMaster> Deewiant, I meant. Which one did you learn first.
21:18:40 <AnMaster> Deewiant, that is the issue then
21:18:48 <AnMaster> Deewiant, it is force of habit.
21:18:54 <Deewiant> I have absolutely no intention of learning it.
21:18:55 <ehird> AnMaster: you're a troll.
21:19:13 <Deewiant> AnMaster: There's no habit involved. I've coded exactly one thing in asm and did it in Intel because I evaluated both and found the other crap.
21:19:26 <AnMaster> ehird, oh. So all those persons saying Haskell is horrible, because they don't understand it, aren't trolls then
21:19:45 <Deewiant> I didn't say it's because I don't understand it, I said it's crap. :-P
21:19:45 <ehird> you're seriously being a retard, it's embarrasing
21:20:00 <Sgeo> http://www.thedailyshow.com/video/index.jhtml?videoId=225919&title=snoutbreak-09-what-to-call
21:20:00 <AnMaster> ehird, I met plenty of programmers who thought functional programming was stupid, because it can't do while loops and similar
21:20:22 <Deewiant> What does this have to do with anything?
21:20:23 <ehird> once again you prove yourself to be the one person who can say things completely irrelevant and act as if they're profound
21:20:29 <AnMaster> I'm saying this is the same thing. You don't understand AT&T syntax.
21:20:47 <ehird> i know it, i've used it, i hate it BECAUSE IT IS SHIT.
21:20:52 <AnMaster> You hate what you aren't used to it.
21:20:57 <ehird> AnMaster: Gentoo is awful. You don't understand Debian.
21:21:07 <ehird> You hate what you aren't used to it.
21:21:19 <AnMaster> then I would have no words left for Red Hat and *shudder* CentOS
21:21:21 <ehird> AnMaster: C is awful. You don't understand Pascal.
21:21:27 <ehird> You hate what you aren't used to it.
21:21:35 <ehird> AnMaster: No you don't.
21:21:38 <ehird> Evidence: You don't understand it.
21:21:41 <ehird> Because you hate it.
21:21:45 <ehird> You hate what you aren't used to it.
21:21:56 <ehird> You hate what you aren't used to it. →
21:21:59 <AnMaster> ehird, Deewiant said he didn't understand AT&T
21:22:27 <AnMaster> ehird, and you said I hate perl only because I don't understand it and such before
21:22:33 <AnMaster> which is exactly the same thing
21:22:43 <Deewiant> When did I say I don't understand it?
21:24:11 <AnMaster> Deewiant, I have coded a lot in both intel and AT&T syntax. I strongly prefer AT&T one. But I wouldn't actually call the intel one horrible most of the time.
21:24:19 <AnMaster> Deewiant, why do you think the AT&T one is so bad.
21:24:30 <Deewiant> &^!@#$!@#$! and the operand suffixes
21:24:39 <AnMaster> Deewiant, what is this "&^!@#$!@#$!" exactly
21:24:56 <Deewiant> AnMaster: For instance, $ in front of numbers
21:25:02 <AnMaster> Deewiant, AT&T is what most other platforms than x86 use.
21:25:18 <AnMaster> so while instruction set differ
21:25:23 <AnMaster> you don't have to learn a new syntax
21:25:30 <AnMaster> and I coded for many platforms
21:25:36 <Deewiant> You have to learn the instruction set anyway
21:25:38 <AnMaster> not having to learn another syntax
21:25:52 <Deewiant> AnMaster: Understand, then reply.
21:25:54 <AnMaster> <AnMaster> so while instruction set differ
21:25:55 <AnMaster> <AnMaster> you don't have to learn a new syntax
21:26:12 <Deewiant> My point was that learning the syntax along with the instruction set is not a noteworthy additional burden.
21:26:29 <Deewiant> The problem is learning the instruction set, not the syntax.
21:27:09 <AnMaster> Deewiant, there is a difference between "hm... so how did you write a mov to a register in this one..." and "hm... so what was the move instruction and the register name in this one..."
21:27:25 <AnMaster> when you coded for enough platforms not having to remember such details help
21:27:33 <Deewiant> Yes, and it is a small difference.
21:27:38 <AnMaster> especially if you haven't coded in it for some time
21:28:01 <AnMaster> Deewiant, so you dislike those prefixes ok
21:28:18 <AnMaster> so that is all the "&^!@#$!@#$!"
21:28:37 <AnMaster> Deewiant, a lot of noise for a small issue IMO
21:28:50 <AnMaster> Deewiant, why should I have to remember to write -> in C. AAARGH
21:29:02 <AnMaster> Deewiant, I was being sarcastic
21:29:21 <Deewiant> It /is/ a pointless distinction.
21:29:26 <AnMaster> AARGH WHY SHOULD I NEED TO KNOW THE DIFFERENCE BETWEEN /**/ AND /++/
21:29:54 -!- calamous has joined.
21:30:10 <Deewiant> True, that is a pointless distinction as well.
21:30:29 <Deewiant> I don't like D's approach of going for C compatibility.
21:30:43 <Deewiant> I understand why it does and it's probably a good idea but I don't like it.
21:31:12 <AnMaster> Deewiant, if you want a perfect language with no odd bits go use R5RS or someting
21:31:23 <Deewiant> That's not what this is about.
21:31:54 <AnMaster> Deewiant, yes, it is still force of habit.
21:32:05 <AnMaster> because sure you looked at GAS
21:32:22 <Deewiant> I could barely deal with either.
21:32:37 <Deewiant> I don't remember, that's probably over 10 years ago.
21:33:09 <AnMaster> Deewiant, oh also, you shouldn't need to write 0x
21:33:22 <Deewiant> You don't, you can write h instead. :-P
21:33:46 <Deewiant> You need some way to distinguish between different number bases.
21:33:49 <AnMaster> so if it saw any letters in the range a-f
21:34:01 <Deewiant> f00 can be both an identifier and a number.
21:34:10 <AnMaster> or possibly always use minimal possible base
21:34:32 <Deewiant> That's a good idea for an esolang
21:34:53 <oerjan> ^ul (-)S(O)(~:S:*(- -)S~:^):^
21:34:53 <fungot> -O- -OO- -OOOO- -OOOOOOOO- -OOOOOOOOOOOOOOOO- -OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO- -OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO- -OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO- -OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO ...too much output!
21:35:18 <thutubot> -O- -OO- -OOOO- -OOOOOOOO- -OOOOOOOOOOOOOOOO- -OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO- -OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO- -OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO ...too much output!
21:35:24 <AnMaster> !underload (-)S(O)(~:S:*(- -)S~:^):^
21:35:36 <AnMaster> ais523, why is EgoBot so slow there
21:35:38 <EgoBot> Supported commands: bf_txtgen help info 1l 2l adjust asm axo bch befunge befunge98 bf bf16 bf32 bf8 c cxx dimensifuck glass glypho kipple lambda lazyk linguine malbolge pbrain qbf rail rhotor sadol sceql trigger udage01 underload unlambda whirl
21:35:49 <ais523> AnMaster: my guess is that it's because it doesn't have a cutoff for too much output
21:35:54 <oerjan> !underload (-)S(O)(~:S:*(- -)S~:^):^
21:36:07 <Deewiant> It just cuts off /after/ it's run the whole thing or timed out.
21:36:16 <Deewiant> Well, I guess that's what you meant anyway.
21:36:28 <AnMaster> since it spammed with me the whole mycology as you remember
21:36:33 <ais523> (-)S(O)(~:S:*(- -)S~:^):^
21:36:35 <ais523> -O- -OO- -OOOO- -OOOOOOOO- -OOOOOOOOOOOOOOOO- -OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO- -OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO- -OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO- -O
21:36:37 <ais523> that's cut off by hand
21:37:01 <AnMaster> from yesterday or whenever it was
21:37:03 <Deewiant> How did you manage to run Mycology over IRC? O_o
21:39:48 <AnMaster> Deewiant, do !befunge98 http://rage.kuonet.org/~anmaster/mycology.b98
21:40:14 <AnMaster> last I checked it excess flooded off every few minutes
21:40:35 <ais523> AnMaster: do you have sanity.b98 up?
21:40:38 <ais523> that would be a lot saner...
21:41:04 <Deewiant> It's fully Befunge-93 compliant!
21:46:19 <AnMaster> ais523, pull from my ick branch.
21:46:32 <AnMaster> ais523, and I see no changes to pull from you
21:46:52 <ais523> AnMaster: final exams for me atm
21:46:56 <ais523> so I have to be careful about what I do with my time
21:47:05 <AnMaster> ais523, like spending hours on irc
21:47:05 <Deewiant> Me too, but that hardly matters ;-)
21:47:15 <ais523> AnMaster: yes, IRC is one of my favourite uses of spare time
21:47:19 -!- ais523 has quit (Remote closed the connection).
21:47:55 * oerjan _assumes_ that wasn't sarcastic, but you never know...
21:49:13 <GregorR> !sh echo Gregor shouldn\'t trust plash this much :P
21:49:13 <EgoBot> Gregor shouldn't trust plash this much :P
21:50:25 <GregorR> Deewiant: It puts it in a file, that's just an invalid script :P
21:50:49 <EgoBot> /home/egobot/egobot.hg/multibot_cmds
21:52:22 <Deewiant> !sh rm Makefile && ls Makefile
21:52:39 <GregorR> Deewiant: It's the '&&' that failed there.
21:53:34 <EgoBot> Linux codu.org 2.6.26-1-xen-amd64 #1 SMP Sat Jan 10 20:39:26 UTC 2009 x86_64 GNU/Linux
21:53:50 <GregorR> Very little of the filesystem is exposed.
21:54:05 <GregorR> AnMaster: http://google.com/search?q=plash
21:54:09 <AnMaster> GregorR, it seems fairly broken
21:54:22 <GregorR> AnMaster: Broken in what way? It seems to work great to me.
21:54:39 <AnMaster> I would have expected to see /bin and such
21:54:53 <EgoBot> bash busybox cat chgrp chmod chown chvt cp cpio date dd df dir dmesg dnsdomainname dumpkeys echo ed egrep false fgconsole fgrep fuser grep gunzip gzexe gzip hostname ip kbd_mode kill less lessecho lessfile lesskey lesspipe ln loadkeys login ls lsmod mkdir mknod mktemp more mount mountpoint mt mt-gnu mv nano nc nc.traditional netcat netstat open openvt pidof ping ping6 ps pwd rbash readlink rm rmdir rnano run-parts sed setfont sh sleep stty su sync tai
21:55:16 <GregorR> AnMaster: Because rm failed
21:56:03 <AnMaster> !sh /usr/bin/file /bin/busybox
21:56:03 <GregorR> AnMaster: The spamming is not going to be useful for >5 messages
21:56:04 <EgoBot> /bin/busybox: ELF 64-bit LSB executable, x86-64, version 1 (SYSV), dynamically linked (uses shared libs), stripped
21:56:24 <EgoBot> LD_LIBRARY_PATH=/usr/lib/plash/lib
21:56:42 <GregorR> !sh unset LD_LIBRARY_PATH; echo hi
21:56:49 <EgoBot> LD_LIBRARY_PATH=/usr/lib/plash/lib
21:56:51 <GregorR> !sh unset LD_LIBRARY_PATH; cat Makefile
21:57:10 <GregorR> The tricky thing about plash is that the LD_LIBRARY that's loaded in /allows/ it to see anything.
21:57:13 <AnMaster> !sh unset LD_LIBRARY_PATH; env | grep ld
21:57:14 <EgoBot> I_ARG=unset LD_LIBRARY_PATH; env | grep ld
21:57:16 <GregorR> By default it /can't/ see anything.
21:57:35 <GregorR> AnMaster: It has a special libc
21:57:46 <AnMaster> GregorR, yes right. But what prevents me from using another one
21:57:56 <AnMaster> I mean a statically linked binary or such
21:57:59 <GregorR> AnMaster: It's in a chroot jail with nothing in it, running as a fake user.
21:58:15 <AnMaster> GregorR, a different chroot than the egobot one?
21:58:31 <GregorR> (plash dynamically makes chroot jails)
21:58:43 <AnMaster> GregorR, so egobot runs as root
21:58:54 <GregorR> AnMaster: plash is setuid root to make chroots
21:59:09 <EgoBot> accessdb activateCmosToken add-shell addgroup adduser arp arpd ascii2enUS_scancode aspell-autobuildhash assetTag avahi-daemon biosdecode chgpasswd chpasswd chroot ck-log-system-restart ck-log-system-start ck-log-system-stop cleanup-info console-kit-daemon cpgr cppw createUnitTestFiles cron cytune defoma-reconfigure delgroup dellBiosUpdate dellLEDCtl dellLcdBrightness dellMediaDirectCtl dellWirelessCtl deluser disable_console_redir dmidecode dpkg-diver
21:59:18 <GregorR> ... activateCmosToken? >_O
21:59:33 <AnMaster> !sh /usr/sbin/activateCmosToken --help
21:59:36 <GregorR> !sh dpkg-query -l | xargs echo
22:00:14 <GregorR> BTW, the ultimate purpose of all this silliness is that I'm making daemons work again, and I'd like it to be possible to write daemons in non-esolangs.
22:00:16 <AnMaster> <EgoBot> SSH_CLIENT=65.183.185.209 36199 22
22:00:28 <GregorR> AnMaster: Its messages have exponential backoff.
22:00:28 <AnMaster> I thought we weren't doing env any more
22:00:33 <GregorR> AnMaster: So that was from a much earlier env.
22:00:50 <AnMaster> GregorR, how long will it spam me
22:01:16 <EgoBot> X11 [ a2p addpart addr2line appletviewer apropos apt apt-cache apt-cdrom apt-config apt-extracttemplates apt-ftparchive apt-get apt-key apt-mark apt-sortpkgs aptitude aptitude-create-state-bundle aptitude-run-state-bundle ar as aspell aspell-import awk base64 basename bashbug bdftopcf bdftruncate bsd-write c++ c++filt c2ph c89 c89-gcc c99 c99-gcc c_rehash cal calendar captoinfo catchsegv catman cc chacl chage chattr chcon chfn chkdupexe chrt chsh ck-h
22:01:18 <GregorR> AnMaster: By my calculations, 15 minutes, then once more 15 minutes after that :P
22:01:36 <GregorR> Deewiant: Old EgoBot let users run programs in esoteric languages that would get their own !commands.
22:01:43 <GregorR> Deewiant: I want to get that working again, and nearly have.
22:02:06 <Deewiant> GregorR: So you could add them dynamically over IRC, or what?
22:02:15 <AnMaster> GregorR, just simulate the freenode model
22:02:35 <AnMaster> GregorR, simulate the freenode rate limiting
22:02:47 <GregorR> I don't know how FreeNode does rate limiting.
22:03:02 <Deewiant> 2-second sleep between messages is enough.
22:03:11 <AnMaster> each time a line is sent, substract one
22:03:23 <AnMaster> anyway I don't remember the exact valyes
22:03:30 <AnMaster> but the algorithm was like that
22:04:28 <GregorR> I think I'll just use exponential backoff but cut it off earlier *shrugs*
22:05:33 <GregorR> Eh, I'll poke around at that later.
22:05:40 <AnMaster> and cut it off after 5 or 10 lines
22:05:48 <GregorR> AnMaster: I have no prioritization between processes.
22:05:53 <AnMaster> that is of course 2 second delay globally
22:06:06 <AnMaster> GregorR, so do it in the seralizer process
22:06:22 <GregorR> AnMaster: Yeah, but you could end up blocked by somebody else's process, which is irritating :)
22:06:29 <AnMaster> GregorR, otherwise you could still spam it off.
22:06:51 <GregorR> I'm not trying to make it impossible, just that you have to be actively malicious rather than just make a mistake.
22:07:07 <AnMaster> "<GregorR> AnMaster: Yeah, but you could end up blocked by somebody else's process, which is irritating :)"
22:07:32 <GregorR> You seem to think that multibot is far more complicated than it is :P
22:07:45 <GregorR> I have a bin that all the messages go in to. And that's it. That's the whole setup.
22:07:52 <AnMaster> GregorR, no. I think it *should* be
22:08:14 <GregorR> Not worth the PITA right now *shrugs*
22:08:16 <AnMaster> "Plash modifies library calls by providing a modified version of glibc, PlashGlibc. Most executables are dynamically linked to glibc, so they do not need to be recompiled in order to run under Plash."
22:08:23 <AnMaster> what if I got a static executable in there
22:08:42 <AnMaster> GregorR, do you use plash for the !asm and !c ones
22:08:44 <GregorR> <GregorR> The tricky thing about plash is that the LD_LIBRARY that's loaded in /allows/ it to see anything. <GregorR> By default it /can't/ see anything.
22:08:54 <GregorR> AnMaster: plash is used for /everything/
22:09:02 <AnMaster> GregorR, even for cfunge there
22:09:11 -!- Slereah has joined.
22:09:26 <AnMaster> GregorR, does it still work correctly
22:09:30 <AnMaster> have you checked with mycology
22:09:40 <GregorR> I decided I wanted to worry a little bit less about bug checking the interps :P
22:09:50 <GregorR> Mycology is 12 trillion lines of output, right?
22:10:12 <AnMaster> GregorR, there may be certain features only that doesn't work
22:10:23 <GregorR> Bleh, URL for mycology X_X
22:10:28 <AnMaster> mycology would check that throughtly
22:10:35 <GregorR> !befunge98 http://pastebin.ca/raw/1410667
22:10:43 <GregorR> !befunge http://pastebin.ca/raw/1410667
22:10:50 <GregorR> (Of course, I don't even know if that's valid befunge98 >_> )
22:11:01 <AnMaster> GregorR, http://rage.kuonet.org/~anmaster/mycology.b98
22:11:30 <GregorR> Have I mentioned that you're annoying? :P
22:11:30 <Deewiant> GregorR: That looks like it should work.
22:12:47 * oerjan gives GregorR his "AnMaster is annoying" badge
22:12:52 <GregorR> BAD: 32x doesn't set delta to (3, 2)
22:13:09 <AnMaster> GregorR, um. That doesn't happen freestanding does it
22:13:12 <Deewiant> That's pretty fail-y for an interpreter right there.
22:13:16 <oerjan> i just distribute the badges
22:13:26 <AnMaster> it doesn't happen when I do it normall
22:13:46 <GregorR> It does appear to happen freestanding.
22:13:58 <AnMaster> GregorR, sure you are using cfunge not the other one then
22:14:17 * oerjan gives AnMaster an "Elephants are pink, tiny with whiskers and can fly" badge
22:14:27 <lifthrasiir> hmm, i wonder what is the most optimizing brainfuck compiler and how does it do.
22:14:45 <AnMaster> GregorR, and it worked yesterday
22:14:47 <Deewiant> AnMaster: No interpreter on http://iki.fi/matti.niemenmaa/befunge/mycology-comparison.html fails at that.
22:14:54 <AnMaster> Deewiant, nor does cfunge here
22:15:16 <AnMaster> $ build/cfunge mycology/mycology.b98 | grep BAD
22:15:16 <AnMaster> If the interpreter loads the fingerprint, subsequent lines will be GOOD or BAD depending on whether the interpreter behaves according to the fingerprint's specification.
22:15:34 <AnMaster> which is not a BAD but just an info line
22:15:58 <AnMaster> GregorR, maybe you managed to mess up download *wgets from that url to check*
22:16:30 <GregorR> What part of "S - T - F - U" don't you understand.
22:16:37 <GregorR> I'm looking in to it, if I have questions I'll ask you.
22:16:43 <GregorR> If not, your spamming me that it should work is not helpful.
22:17:10 <AnMaster> I'm going to bed soon though, early morning tomorrow. But I will be around for maybe half an hour
22:18:32 <lifthrasiir> i have written some brainfuck-to-c compiler just out of curiosity, but i want to see world-class optimization ;)
22:19:11 <lifthrasiir> libbf seems to have some optimization passes but i'm not sure.
22:19:41 <AnMaster> lifthrasiir, I wrote one too. It kind of worked not very well
22:19:54 <AnMaster> like it compiled all in a single function
22:20:11 -!- WangZeDong has quit (Connection timed out).
22:20:23 <lifthrasiir> many compilers claiming optimization feature, in reality, only fold <<<<s and >>>>s etc.
22:20:57 <lifthrasiir> some compilers optimize [-], but that's all. i have seen only three or four compilers which optimizes multiplication loop.
22:21:12 <AnMaster> lifthrasiir, I did more. 1) folded +-<> 2) I made [-] set 0 3) [-]+++ -> set 3 4) I folded some loops into polynoms
22:21:19 <AnMaster> oh I also tried to reorganise code
22:21:19 <GregorR> It actually works fine, something (now fixed) removed the \x00 in the input.
22:21:56 <AnMaster> which was then folded into 2+>-<
22:22:32 <AnMaster> and only adjusted pointer when I entered some unbalanced loop or such
22:22:55 <lifthrasiir> "4) I folded some loops into polynoms" seems interesting. any details?
22:23:01 <AnMaster> lifthrasiir, I didn't optimise multiplication loop. I got bored before that.
22:23:08 <AnMaster> lifthrasiir, it was based on some other compiler
22:23:31 <lifthrasiir> http://hg.mearie.org/esotope/bfc/file/tip/esotope-bfc.py my attempt is here.
22:23:52 <AnMaster> http://bzr.kuonet.org/before/trunk/changes
22:24:07 <AnMaster> lifthrasiir, it didn't do it's job very well
22:24:50 <AnMaster> I got bored before that version
22:25:11 <AnMaster> http://bzr.kuonet.org/before/trunk/annotate/head:/src/optimize.c
22:25:25 <AnMaster> http://bzr.kuonet.org/before/trunk/annotate/head:/src/node.h
22:25:26 <AnMaster> http://bzr.kuonet.org/before/trunk/annotate/head:/src/node.c
22:26:39 <AnMaster> lifthrasiir, tell me if you want any more info
22:27:53 <lifthrasiir> well, i think my attempt is same as yours;
22:28:16 <AnMaster> lifthrasiir, "bff4" might be worth checking out
22:28:44 <lifthrasiir> it combines every <+>-s into the list and tries to eliminate dead code and propagate.
22:29:26 <lifthrasiir> so tightloop pass unrolls one loop, primarily for multiplications
22:29:54 <GregorR> !befunge98 http://pastebin.ca/raw/1410667
22:30:07 <GregorR> My sandbox sandboxed networking away from the wget that downloads the requested file X-P
22:30:20 <lifthrasiir> but since i'm using expression classes for intermediate representation i should write some simplification codes for it
22:30:24 <AnMaster> GregorR, so it broke \0 that way
22:30:44 <GregorR> AnMaster: No, that was unrelated.
22:30:46 <AnMaster> lifthrasiir, I think I planned to run mine recursively until optimisation didn't change the code tree
22:30:50 <GregorR> AnMaster: That was just because of how I was testing it.
22:31:10 <AnMaster> GregorR, so mycology works inside the jail then
22:31:22 <lifthrasiir> i got hello world program (as seen in wikipedia pages FYI) optimized to this:
22:31:24 <lifthrasiir> mptr[1] += 9; mptr[0] += (8 * mptr[1]); mptr[1] = 0; putchar(*mptr); mptr[1] += 7; mptr[0] += ((4 * mptr[1]) + 1); mptr[1] = 0; putchar(*mptr); ...
22:32:30 <AnMaster> GregorR, and if so you possibly doesn't need -S from cfunge. It would allow you to do some interesting stuff like file io and socket io
22:32:44 <GregorR> I don't want to give it sockets :P
22:32:57 <GregorR> I'm fine with file I/O at this point if you'd like to make a more precise option.
22:33:02 -!- puzzlet has quit (Read error: 60 (Operation timed out)).
22:33:05 -!- puzzlet has joined.
22:33:25 <AnMaster> GregorR, maybe I'll put that on TODO
22:33:47 <AnMaster> it isn't something I'll add right now anyway
22:33:55 <AnMaster> GregorR, what prevents the !asm one from making direct syscalls
22:34:42 <GregorR> Those were just a silly experiment (which, btw, is now gone :P )
22:35:08 <AnMaster> GregorR, you removed them after I said syscall didn't you
22:35:24 <GregorR> No, but I was removing them as you said that.
22:35:29 <GregorR> The only syscall I'm afraid of is socket.
22:35:41 <GregorR> All the other stuff is fine because it's running in a chroot jail as a fake user.
22:36:07 <EgoBot> Supported commands: bf_txtgen help info 1l 2l adjust axo bch befunge befunge98 bf bf16 bf32 bf8 dimensifuck glass glypho kipple lambda lazyk linguine malbolge pbrain qbf rail rhotor sadol sceql trigger udage01 underload unlambda whirl
22:36:09 <GregorR> ... iptables can do that? >_O
22:36:21 <AnMaster> GregorR, iirc there is an owner match yes
22:36:37 <GregorR> If so, I'll set that up and put them back.
22:36:38 <AnMaster> GregorR, not sure if it is part of standard kernel sources or just part of gentoo hardened
22:37:18 <ehird> 22:11 GregorR: Have I mentioned that you're annoying? :P ← You don't say.
22:37:19 <ehird> 22:34 GregorR: Those were just a silly experiment (which, btw, is now gone :P )
22:37:26 <ehird> !c printf("NOOO!!\n")
22:37:34 <ehird> RIP EgoBot-being-fun 2009-2009
22:38:06 <AnMaster> │ CONFIG_NETFILTER_XT_MATCH_OWNER: │
22:38:14 <AnMaster> │ Socket owner matching allows you to match locally-generated packets │
22:38:14 <AnMaster> │ based on who created the socket: the user or group. It is also │
22:38:14 <AnMaster> │ possible to check whether a socket actually exists. │
22:39:05 <AnMaster> [!] --uid-owner userid[-userid] Match local UID
22:39:05 <AnMaster> [!] --gid-owner groupid[-groupid] Match local GID
22:39:05 <AnMaster> [!] --socket-exists Match if socket exists
22:39:07 -!- tombom has quit ("Peace and Protection 4.22.2").
22:39:30 <GregorR> I'm trying to see if there's a way to make it range-based though.
22:39:32 <AnMaster> GregorR, you use a standard kernel
22:39:53 <ehird> Bah, I'll have to make my own arbitrary-code-execution bot. WITH BLACKJACK AND HOOKERS.
22:40:21 <ehird> AnMaster: PLASH IS FASCIST SYSTEM RESTRICTEr
22:40:32 <AnMaster> ehird, he used that all the time for C
22:40:38 <ehird> EXECUTION LIBÉRATIONE!
22:40:59 <ehird> Yay Mastodon downloaded.
22:41:00 <AnMaster> ehird, what is the dirt over the first E in the second word
22:41:42 -!- oklopol has joined.
22:42:03 * lifthrasiir comparing http://pastie.org/466231 (before output) and http://pastie.org/466232 (esotope-bfc output)
22:42:39 <lifthrasiir> of course i didn't write template for expanding memory. :p
22:43:23 <AnMaster> ehird, didn't you have it before
22:43:29 <ehird> AnMaster: the netinstall didn't work
22:43:40 <ehird> so I downloaded the beta disk
22:44:00 <ehird> AnMaster: 300-so megabytes @ 50kb/sec.
22:44:16 <ehird> Screenshot of the installer: (it has COLOURS!)
22:44:45 <ehird> AnMaster: http://omploader.org/vMW00cA
22:44:48 <ehird> And it's not THAT big.,
22:44:53 <AnMaster> lifthrasiir, your optimise better right
22:45:02 <ehird> It contains all the binaries, probably unstripped, with libc 5 and gcc 2
22:45:13 <ehird> No networking at all, and a full system with things like adobe acrobat
22:45:16 <AnMaster> lifthrasiir, as well as produces nicer output
22:45:23 <lifthrasiir> AnMaster: not so good yet, but i want to optimize further so there is only putchar('H'); putchar('e'); ... and so on. :)
22:45:46 <AnMaster> lifthrasiir, would only be possible for a special case
22:45:49 <ehird> lifthrasiir: you can optimize any loop with the same amount of <s and >s to a polynomial
22:46:07 <lifthrasiir> ehird: yes. and what i have done is exactly that
22:46:19 <ehird> OK, AnMaster: Mastodon install time.
22:46:27 <AnMaster> <ehird> AnMaster: http://omploader.org/vMW00cA <-- simplistic installed
22:46:32 <ehird> wow, it even has a helpfile
22:46:38 <lifthrasiir> though it cannot handle non-constant increments/decrements for now.
22:46:38 <AnMaster> but why so much transparent padding around the window
22:46:38 <ehird> you can press F1 and get info on the current installer screen
22:46:46 <ehird> use a non-black BG
22:47:02 <AnMaster> ehird, err I use the checker pattern
22:47:05 -!- Gracenotes has quit (Remote closed the connection).
22:47:13 <AnMaster> and I see it a third of the way
22:47:39 <AnMaster> lifthrasiir, so is it better than mine or worse
22:48:02 <ehird> AnMaster: Here, have a screenshot of the help: http://omploader.org/vMW00cQ
22:48:19 <lifthrasiir> AnMaster: i'm not sure, i'll check for them.
22:48:33 <ehird> It's a pretty sleek installer for a >7 year old project by one guy
22:48:47 <ehird> Automatic installation GOOO!
22:48:50 <AnMaster> lifthrasiir, I didn't properly do nested loops
22:48:54 <ehird> lawl, it only does the first disk
22:48:56 <ehird> no other option :D
22:49:02 <AnMaster> lifthrasiir, and as I said I gave up on it
22:49:19 <ehird> AnMaster: http://omploader.org/vMW00cg
22:49:25 <AnMaster> lifthrasiir, is esotope-bfc written in C or some other language
22:49:25 <lifthrasiir> then yours and mine is perhaps at same level, i also didn't optimize nested loops
22:50:13 <AnMaster> it had to work on old computers
22:50:28 <ehird> AnMaster: I dunno, so far this is the easiest, quickest Linux distro install I've done :-)
22:50:46 <ehird> Kind of disillusioning. Yay, it's installing kernel version 2.0.28 :-P
22:50:56 <ehird> And Perl 5.6! And Tcl/tk 8!
22:51:01 <ehird> A veritable smorgasbord of languages.
22:51:07 <ehird> It's bloated, clearly ;-)
22:51:07 <AnMaster> ehird, what is the last tcl/tk
22:51:13 <ehird> AnMaster: it was 8.0.4 this one
22:51:17 <ehird> and I think 8.6.something
22:52:02 <ehird> AnMaster: it's installed, now it wants me to set up the identity!
22:52:16 <ehird> Domain name: yeolde
22:52:30 <ehird> AnMaster: it just puts it in /etc/hosts, I think
22:52:53 <AnMaster> GregorR, so now we get !asm and such back
22:52:59 <ehird> Configure this device by hand OR Use dhcp/bootp for automatic configuration OR Do not configure this device
22:53:07 <ehird> hurr, when's the last time anyone last manually configured a networking device :P
22:53:23 <ehird> GregorR: It wants me to set up "Yellow Pages (NIS) administration"
22:53:29 <ehird> I'll just leave NIS domain blank and NEXT :-P
22:53:36 <ehird> Use ntp to set the time?
22:53:45 <ehird> AnMaster: what's a good utc ntp server
22:54:04 <ehird> what's a good ntp server
22:54:24 <Deewiant> ehird: http://www.pool.ntp.org/zone/europe
22:54:48 <ehird> [0123].europe.pool.ntp.org, OK
22:55:12 <Deewiant> ehird: Well, unless you want to use a UK one or whatever.
22:55:24 <ehird> Deewiant: It's mastodon; I really don't care :-)
22:56:12 <ehird> AnMaster: http://omploader.org/vMW00dg
22:56:14 <ehird> This distro is nice :P
22:56:15 <GregorR> Oh, I've managed to entirely kill it, awesome :P
22:56:32 -!- EgoBot has quit (Remote closed the connection).
22:56:35 -!- EgoBot has joined.
22:56:38 <EgoBot> Supported commands: bf_txtgen help info 1l 2l adjust asm axo bch befunge befunge98 bf bf16 bf32 bf8 c cxx dimensifuck glass glypho kipple lambda lazyk linguine malbolge pbrain qbf rail rhotor sadol sceql sh trigger udage01 underload unlambda whirl
22:56:54 <ehird> AnMaster: Gooooooooo!
22:56:56 <AnMaster> ehird, shouldn't you be complaining about it not being GUI
22:57:05 <ehird> Well, it's simple enough that I don't care.
22:57:12 <ehird> Please set the root password for this machine!
22:57:27 <ehird> I shall set it to the same as what my user password will be BECAUSE I AM HARDCORE AND YOU ARE A FEEBLE VM
22:57:40 <ehird> Could not open new password file, left unset XD
22:57:43 <AnMaster> GregorR, I assume you have limits
22:57:55 <EgoBot> Supported commands: bf_txtgen help info 1l 2l adjust asm axo bch befunge befunge98 bf bf16 bf32 bf8 c cxx dimensifuck glass glypho kipple lambda lazyk linguine malbolge pbrain qbf rail rhotor sadol sceql sh trigger udage01 underload unlambda whirl
22:57:58 <ehird> AnMaster: Wow, it has a dialog-based X configuration system
22:58:02 <ehird> Most distros don't have that :P
22:58:42 <ehird> So does Debian etc but not many ones YOU like ;-)
22:58:53 <ehird> Configure the monitor, and it gives a long list of models.
22:59:14 <ehird> PROBING YOUR VIDEO HARDWARE
22:59:25 <AnMaster> Deewiant, that sounds interesting
22:59:43 <ehird> YOU MUST INDICATE HOW MUCH VIDEO MEMORY YOU HAVE!
22:59:44 <AnMaster> ehird, wow. I would have expected CGA
23:00:14 <ehird> Haha it wants to run at 640x480
23:00:28 <ehird> My VM imitates Cirrus Logic GD5446
23:00:30 <ehird> So I'll choose that
23:01:08 <ehird> Test current configuration!
23:01:12 <ehird> This is all on the installer CD still
23:01:14 <AnMaster> GregorR, do you kill all the processes in said config
23:01:26 <ehird> The x server seems to have died ;_;;;
23:01:30 <ehird> No keyboard config XD
23:01:40 <GregorR> AnMaster: I /think/ so X-D
23:01:42 <AnMaster> GregorR, what if I double forked. Would it still be killed after 30 seconds
23:02:03 <ehird> It's having keyboard issues
23:02:33 <ehird> Fatal server error:
23:02:37 <ehird> You must specify a keyboard in XF86Config
23:02:42 <GregorR> AnMaster: Your megaforks are all dead.
23:03:11 <ehird> I'll fix it once this is installed
23:03:11 <EgoBot> core file size (blocks, -c) 0 data seg size (kbytes, -d) unlimited scheduling priority (-e) 0 file size (blocks, -f) 10240 pending signals (-i) 16382 max locked memory (kbytes, -l) 64 max memory size (kbytes, -m) unlimited open files (-n) 1024 pipe size (512 bytes, -p) 8 POSIX message queues (bytes, -q) 819200 real-time priority (-r) 0 stack size (kbytes, -s) 8192 cpu time (seconds, -t) 30 max user processes (-u) 32 virtual memory (kbytes, -v) 131072
23:03:17 <AnMaster> !sh ulimit -a | xargs echo '| '
23:03:19 <EgoBot> | core file size (blocks, -c) 0 data seg size (kbytes, -d) unlimited scheduling priority (-e) 0 file size (blocks, -f) 10240 pending signals (-i) 16382 max locked memory (kbytes, -l) 64 max memory size (kbytes, -m) unlimited open files (-n) 1024 pipe size (512 bytes, -p) 8 POSIX message queues (bytes, -q) 819200 real-time priority (-r) 0 stack size (kbytes, -s) 8192 cpu time (seconds, -t) 30 max user processes (-u) 32 virtual memory (kbytes, -v) 131
23:03:26 <AnMaster> !sh ulimit -a | xargs -1 echo '| '
23:03:43 <ehird> Haha, no space left on device
23:04:00 <GregorR> AnMaster: I have no idea why that command didn't work :P
23:04:03 <ehird> AnMaster: 512MB; it seems bloat was everywhere even then
23:04:04 <AnMaster> <ehird> You must specify a keyboard in XF86Config
23:04:18 <GregorR> !sh ulimit -a | xargs -1 echo -n '| '
23:04:18 <ehird> AnMaster: Dude. Xorg started in 2004.
23:04:32 <Deewiant> ehird: Consider losing that tetex and whatnot.
23:04:50 <ehird> Welp, it's booting
23:05:03 <ehird> WORST CONSOLE FONT EVER
23:05:15 <ehird> http://omploader.org/vMW00eQ
23:05:27 <ehird> It's like COMIC SANS, CONSOLE EDITION
23:05:53 <ehird> yeolde login: root
23:06:09 <AnMaster> ehird, didn't it make you add an user
23:06:21 <AnMaster> ehird, maybe because it errored out when it ran out of space
23:06:25 <ehird> AnMaster: no, it continued
23:06:29 <ehird> Deewiant: We have root
23:06:34 <ehird> it just forbids logging in to it
23:06:42 <ehird> AnMaster: by skipping that step
23:06:49 <Deewiant> ehird: Keep guessing, i.e. try other users.
23:06:54 <ehird> Deewiant: There are none.
23:06:58 <ehird> AnMaster: The write-x-config step.
23:07:06 <GregorR> !sh wget http://localhost -O - | head -n 1
23:07:10 <AnMaster> ehird, try again with enough disk space
23:07:15 <GregorR> !sh wget http://127.0.0.1 -O - | head -n 1
23:07:16 <EgoBot> <?xml version="1.0" encoding="utf-8" ?><!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
23:07:27 <AnMaster> GregorR, ok that is localhost though
23:07:29 <ehird> AnMaster: how much disk do you think I need?
23:07:43 <GregorR> AnMaster: I was just pointing out the small oddity that opening sockets is OK, they just can't be remote :P
23:07:54 <AnMaster> ehird, whare are the specified system reqs?
23:08:06 <AnMaster> ehird, try whatever crysis needs then
23:08:22 <ehird> Lawl, creating a 2gb image freezes Q
23:08:45 <ehird> Qemu gui frontend for mac.
23:09:01 <AnMaster> ehird, try virtualbox or parralells
23:09:04 <ehird> BTW http://www.colorforth.com/GA.htm
23:09:15 <ehird> AnMaster: virtualbox's vm crashes on mastodon, my parallels trial is probably up
23:09:52 <AnMaster> GA32 is a 32-computer chip in an 88-pin package
23:10:39 <ehird> it's chuck moore's new company
23:13:35 <ehird> Heh, it comes with emacs 19
23:15:53 <ehird> awesome, x almost works ;)
23:16:27 <ehird> AnMaster: cursor is just a black square
23:16:35 <ehird> hokay, let's try this
23:16:36 <AnMaster> ehird, did it let you set up users
23:17:17 <ehird> AnMaster: Methinks, that I have some work to do regarding X configuration:
23:17:45 <ehird> Eh, it won't upload
23:17:49 <ehird> Basically, the text is mangled.
23:18:18 <ehird> that's because you didn't use Xfree86 in ~2000 in a virtual machine, AnMaster.
23:18:37 <AnMaster> but you said it "didn't upload"
23:18:53 <AnMaster> I never had issues with ompload
23:19:24 <AnMaster> ehird, <ehird> http://omploader.org/vMW00eQ <-- what is "Recovering jove files"
23:19:35 <ehird> jove is an old editor
23:19:56 <GregorR> !cxx cout << "jove sux" << endl
23:19:56 <ehird> let's try this again
23:20:11 <ehird> apt-cache show jove
23:20:21 <AnMaster> ehird, I'm not on such a distro
23:20:34 <ehird> you've done the same before
23:20:35 <GregorR> Debianers love to troll apt-* though :(
23:21:12 <ehird> GregorR: it's not our fault we have a comprehensive, stable package base at our fingertips
23:21:27 <AnMaster> Available versions: 4.16.0.70.3.1 {unix98}
23:21:27 <AnMaster> Homepage: ftp://ftp.cs.toronto.edu/cs/ftp/pub/hugh/jove-dev/
23:21:27 <AnMaster> Description: Jonathan's Own Version of Emacs - a light emacs-like editor without LISP bindings
23:21:56 <GregorR> I'm sure it looks more-or-less like emacs.
23:22:22 <GregorR> ............................................ ohhhhhhhhhhhhh kay.
23:22:34 <AnMaster> GregorR, or did I misunderstand you
23:22:45 <AnMaster> maybe you said emacs was great there
23:22:54 <GregorR> Description: ... a light emacs-like editor ...
23:23:00 <GregorR> Therefore I concluded that it looks more-or-less like emacs.
23:23:12 <AnMaster> <GregorR> !cxx cout << "jove sux" << endl
23:23:31 <GregorR> I knew nothing about it, I was just being an ass :P
23:23:45 <AnMaster> GregorR, so which editor do you use
23:24:12 <AnMaster> GregorR, you can't irc in vim as easily. Though there is an extension for it
23:24:23 <GregorR> My text editor is not my operating system.
23:24:27 <GregorR> It is, in fact, my text editor.
23:24:51 <GregorR> emacs would be a perfectly good operating system if it had a better text editor.
23:25:22 <AnMaster> emacs is a lisp system. elisp yes. Which is not as nice as clisp or scheme
23:26:11 -!- KingOfKarlsruhe has quit (Remote closed the connection).
23:31:39 <ehird> AnMaster: you use bochs right
23:31:43 <ehird> what card does its generic vga emulate?
23:32:14 <AnMaster> ehird, I more often use qemu and vmware-server
23:32:24 <AnMaster> haven't used vmware-server since January
23:32:33 <AnMaster> haven't used qemu for over a year
23:32:53 <GregorR> !forth ." Hello, world!" CR
23:32:54 <AnMaster> in fact I haven't used any kind of virtual stuff since January
23:33:04 <ehird> GregorR: what forth?
23:33:39 <ehird> GregorR: plz to be printing stderr
23:33:40 <GregorR> gforth is case-insensitive.
23:33:43 <ehird> AnMaster: CR isn't idiomatic
23:33:43 <Deewiant> "Most versions of MS-DOS 6.22 do not idle the CPU when they are idle." Damn, I didn't know that
23:34:34 <ehird> GregorR: Can you please make it talk to the channel?
23:34:41 <ehird> I haaaaaaaate having to switch :p
23:34:55 <AnMaster> <EgoBot> in file included from *the terminal*:0
23:34:55 <AnMaster> <EgoBot> /tmp/input.19662:1: Invalid memory address
23:34:58 <ehird> !forth VARIABLE: a a @ . cr
23:35:06 <GregorR> ehird: The problem is that stdout and stderr aren't serialized WRT each other.
23:35:11 <GregorR> ehird: So your cr came first.
23:35:13 <ehird> !forth VARIABLE: a a . cr
23:35:31 <GregorR> ehird: The problem is that stdout and stderr aren't serialized WRT each other.
23:35:40 <ehird> !forth VARIABLE a a . cr
23:35:44 <GregorR> stdout and stderr, being the C files.
23:35:48 <ehird> !forth VARIABLE a 2 a ! a . cr
23:35:51 <ehird> !forth VARIABLE a 2 a ! a @ . cr
23:36:07 <ehird> !forth VARIABLE a : get-a a @ ; : set-a a ! ; 2 set-a get-a . cr
23:36:21 <ehird> AnMaster: who cares
23:36:21 <AnMaster> (same word in Swedish, easy to mix up
23:36:31 <ehird> I just want egobot to dump shit to the channel
23:36:43 <AnMaster> I prefer first line in channel only
23:36:51 <ehird> that's a rubbish back-off style
23:37:00 <ehird> and i'll paste it in anyway
23:37:01 <AnMaster> just make your irc client put it in your channel window
23:37:05 <ehird> since I do things in here to show people, duh
23:37:27 <ehird> AnMaster: then they should make the request via /msg
23:37:41 <ehird> If not, it should.
23:37:42 <AnMaster> ehird, and I think you are wrong anyway
23:37:59 <ehird> i've given good arguments, you haven't.
23:38:04 <ehird> i presume you'll refuse to give them
23:38:33 <AnMaster> <AnMaster> I prefer first line in channel only
23:38:35 <AnMaster> <AnMaster> just make your irc client put it in your channel window
23:38:40 <ehird> AnMaster: yes, but,
23:38:49 <ehird> the line in channel only, less spammy: If you want it public you'll paste it anyway
23:38:52 <ehird> If you don't use /msg
23:39:00 <AnMaster> ehird, you could have made your client put it in this channel virtually locally
23:39:07 <ehird> i'd still have to paste it
23:39:10 <AnMaster> in less time than this discussion
23:40:17 <EgoBot> /tmp/source.19948.c:7:6: warning: unknown escape sequence: '302'
23:40:20 <EgoBot> Supported commands: bf_txtgen help info 1l 2l adjust asm axo bch befunge befunge98 bf bf16 bf32 bf8 c cxx dimensifuck forth glass glypho kipple lambda lazyk linguine malbolge pbrain qbf rail rhotor sadol sceql sh test trigger udage01 underload unlambda whirl
23:40:29 <ehird> GregorR: so are you gonna make it print to the channel or not :P
23:40:46 <kerlo> !c puts("one\ntwo");
23:40:49 <ehird> That is not the same thing. Do not speak for GregorR.
23:40:58 <kerlo> !c puts("one\ntwo")
23:41:09 <EgoBot> /tmp/source.20145.c: In function 'main':
23:41:13 <AnMaster> ehird, well he did the Right Thing for asm before
23:41:18 <AnMaster> so I trust him to do that now too
23:41:19 <kerlo> What a descriptive error message.
23:41:37 <ehird> Hey, a UI you have to explain to each and every new user because it's so shit.
23:41:40 <ehird> That's so the Right Thing.
23:41:52 <AnMaster> because their clients are shit
23:41:58 <AnMaster> mine show what happened clearly
23:42:07 <ehird> Hear that everyone? If you don't use ERC, your client sucks.
23:42:31 <kerlo> !c puts((int)puts)
23:42:32 <EgoBot> /tmp/source.20187.c: In function 'main':
23:42:33 <ehird> Thus we prove that only the two clients that AnMaster likes are acceptable.
23:43:21 <EgoBot> /tmp/source.20230.c: In function 'main':
23:43:34 <kerlo> It appears to be a function.
23:43:57 <AnMaster> !c printf("%d\n", (int)fputs);
23:43:58 <EgoBot> /tmp/source.20280.c: In function 'main':
23:44:25 <EgoBot> /tmp/source.20320.c: In function 'main':
23:44:31 <kerlo> !c printf("%d", (long int) puts)
23:44:33 <Deewiant> AnMaster: He did no Right Thing for asm, he just used GAS, probably just because he had it installed already.
23:44:50 <AnMaster> Deewiant, no, he said he liked it
23:44:56 <kerlo> !c printf("%d", (long int) printf)
23:44:59 <kerlo> !c printf("%d", (long int) printf)
23:45:05 <ehird> Deewiant: gawd, just let him exist in his own GregorR-fellating world.
23:45:21 <kerlo> !c 4195272("%d", (long int) printf)
23:45:22 <EgoBot> /tmp/source.20459.c: In function 'main':
23:45:26 <Deewiant> In any case the Right Thing is more about the fact that GAS's support for Intel syntax is obviously crap :-P
23:45:42 <AnMaster> Deewiant, remember he coded a lot for MIPS
23:45:49 <AnMaster> which uses AT&T syntax most of the time
23:45:57 <kerlo> !c itfu("%d", (long int) printf)
23:45:58 <EgoBot> /tmp/ccX9BYHg.o: In function `main':
23:46:12 <Deewiant> Point being that if you're going to use GAS you're stuck with AT&T for all practical purposes.
23:46:28 <EgoBot> /tmp/source.20540.c: In function 'main':
23:46:34 <AnMaster> Deewiant, and with nasm you are stuck the other way
23:46:42 <ehird> I like how I can't follow what the fuck you're all doing.
23:46:49 -!- oerjan has quit ("Good night").
23:46:56 <EgoBot> /tmp/source.20582.c: In function 'main':
23:47:29 <EgoBot> Supported commands: bf_txtgen help info 1l 2l adjust asm axo bch befunge befunge98 bf bf16 bf32 bf8 c cxx dimensifuck forth glass glypho kipple lambda lazyk linguine malbolge pbrain qbf rail rhotor sadol sceql sh test trigger udage01 underload unlambda whirl
23:47:34 <AnMaster> GregorR, request for language: shakespear
23:47:58 <ehird> shakespeare is le shit
23:47:58 <AnMaster> http://shakespearelang.sourceforge.net/report/shakespeare/shakespeare.html
23:48:11 <ehird> it's not actually interesting
23:48:26 <ehird> a lot of rubbish things are
23:48:45 <AnMaster> The Infamous Hello World Program.
23:48:45 <AnMaster> Romeo, a young man with a remarkable patience.
23:48:45 <AnMaster> Juliet, a likewise young woman of remarkable grace.
23:48:45 <AnMaster> Ophelia, a remarkable woman much in dispute with Hamlet.
23:48:45 <AnMaster> Hamlet, the flatterer of Andersen Insulting A/S.
23:50:20 -!- oklopol has quit (Success).
23:50:30 -!- BeholdMyGlory has quit (Remote closed the connection).
23:50:39 <kerlo> !c ((int) printf)(itfu)
23:50:40 <EgoBot> /tmp/source.20661.c: In function 'main':
23:51:57 <ehird> Anyone know how to find out the RPM of my disk?
23:53:31 <ehird> That's very helpful of you
23:53:49 <AnMaster> ehird, answer is I don't know for OS X
23:55:15 <kerlo> !c ((int*) printf)(itfu)
23:55:16 <EgoBot> /tmp/source.20712.c: In function 'main':
23:55:37 <kerlo> !c ((int*) printf)("%n", 42)
23:55:38 <EgoBot> /tmp/source.20754.c: In function 'main':
23:57:32 <kerlo> !c ((int()) printf)("%n", 42)
23:57:32 <EgoBot> /tmp/source.20794.c: In function 'main':
23:57:33 <kerlo> 18:56 <EgoBot> /tmp/source.20794.c:7: error: cast specifies function type
23:57:33 <ehird> AnMaster: by the way, the annoying resize thing on OS X has one purpose
23:57:33 <ehird> if you hold down shift to make it minimize slowly, then killall Dock
23:57:33 <ehird> it stays in its half-warped state
23:57:33 <kerlo> !c ((int()*) printf)("%n", 42)
23:57:33 <EgoBot> /tmp/source.20831.c: In function 'main':
23:57:33 <ehird> and yes, you can interact with it
23:58:26 -!- puzzlet has quit (Remote closed the connection).
23:58:32 -!- puzzlet has joined.
23:58:39 <AnMaster> ehird, screenshot of interacting with it I mean
23:59:37 <ehird> http://omploader.org/vMW01NQ