00:06:30 -!- poiuy_qwert has quit (Quit: Leaving).
00:09:14 <AnMaster> <olsner> hmm, perhaps 'virrigt', but that doesn't imply the same kind of agitation
00:09:28 <AnMaster> I just don't know how it applies to a piano
00:09:40 <ais523> hmm, I just started using flymake-mode in Emacs, and it's amazing
00:09:59 <ais523> it repeatedly syntax-checks the code you're working on
00:10:06 <Gregor> Do you have something against my personifying things :P
00:10:10 <ais523> and syntax errors get highlighted almost immediately
00:10:18 <ais523> much faster than in any other IDE I've used
00:10:37 <ais523> I was expecting it to take several seconds, like NetBeans, etc., does
00:11:41 <oerjan> Gregor: no, but the things don't like it
00:12:31 <Gregor> oerjan: YOU MEAN THE PEOPLE
00:13:10 <oerjan> the vitally challenged people
00:14:07 <Gregor> "vitally challenged" is a pretty awesome phrase :P
00:25:16 <ais523> AnMaster: I was using it for Perl
00:25:20 <ais523> although I'm not sure what langs it works for
00:25:29 <ais523> I suspect, most of the popular ones
00:25:58 <ais523> besides, it doesn't normally have compile-time syntax errors
00:26:06 <ais523> and sure, read the source to intercal-mode
00:26:27 <ais523> AnMaster: it complains on a few things at compile-time, but that's arguably a bug
00:26:41 <AnMaster> ais523, what about the wrong dimensionality and such?
00:27:10 <ais523> given that dimensionality can change at runtime, how could you do it otherwise?
00:27:13 <AnMaster> ais523, it also complains if I feed it /dev/random
00:27:41 <ais523> maybe it only works on regular files
00:27:42 <AnMaster> ais523, something about it not looking like intercal iirc?
00:27:52 <ais523> /dev/random doesn't end ".i"
00:28:09 <AnMaster> head --bytes 5000 /dev/urandom > foo.i
00:28:46 <AnMaster> also it complains about missing files
00:29:40 <ais523> but flymake, you wouldn't expect to complain about missing files
00:29:47 <ais523> given that it, you know, checks the file you're working on atm
00:29:50 <ais523> which presumably exists
00:29:58 <AnMaster> anyway what about head --bytes 5000 /dev/urandom > foo.i
01:07:27 -!- FireFly has quit (Quit: Leaving).
01:08:08 -!- ais523 has quit (Remote host closed the connection).
01:33:09 -!- kwertii has joined.
01:33:57 -!- Gracenotes has joined.
02:22:37 -!- BeholdMyGlory has quit (Remote host closed the connection).
03:51:52 -!- augur has joined.
03:54:33 -!- jcp has quit (Quit: I will do anything (almost) for a new router.).
04:16:27 -!- jcp has joined.
05:26:16 -!- oerjan has quit (Quit: leaving).
06:53:05 -!- lament has joined.
07:04:50 -!- lament has quit (Remote host closed the connection).
07:04:55 -!- lament has joined.
07:05:39 -!- Slereah has quit (Ping timeout: 260 seconds).
07:31:35 -!- kwertii has quit (Quit: bye).
07:45:39 -!- Slereah has joined.
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:00:29 -!- oerjan has joined.
08:36:07 -!- lament has quit (Ping timeout: 246 seconds).
08:42:34 <AnMaster> pikhq, in #irp: <pikhq> AnMaster: Congrats, you are now uberop. <-- what?
08:54:43 -!- oerjan has quit (Quit: leaving).
09:19:47 -!- FireFly has joined.
09:22:28 -!- jcp has quit (Quit: I will do anything (almost) for a new router.).
09:28:10 -!- MigoMipo has joined.
09:43:34 <fizzie> AnMaster: Just a guess, but probably referring what "/msg chanserv access #irp list" should reveal to you.
09:46:12 -!- adam_d has joined.
09:58:39 -!- tombom has joined.
10:03:44 -!- MigoMipo has quit (Remote host closed the connection).
10:19:19 <AnMaster> Deewiant, other change: yn-rep 1000000 reduced from 8.169 to 5.370 seconds on my computer
10:23:30 <AnMaster> y-rep-n saw a similar, but not quite as large, speed increase. Didn't test it for quite that large values due to "limited" ram
10:27:10 -!- MizardX has quit (Read error: Connection reset by peer).
10:27:39 -!- MizardX has joined.
11:18:55 -!- chickenzilla has quit (Read error: Connection reset by peer).
11:36:46 -!- chickenzilla has joined.
12:06:56 -!- BeholdMyGlory has joined.
13:12:43 -!- cheater2 has quit (Read error: Connection reset by peer).
13:13:15 -!- cheater2 has joined.
13:39:32 -!- Alex3012_ has joined.
13:40:55 -!- Alex3012 has quit (Ping timeout: 245 seconds).
13:41:06 -!- Alex3012_ has changed nick to Alex3012.
13:45:32 -!- Gracenotes has quit (Quit: Leaving).
14:02:09 -!- MizardX has quit (Quit: reboot).
14:07:57 -!- MizardX has joined.
14:27:52 -!- BeholdMyGlory has quit (Read error: Connection reset by peer).
15:06:20 -!- SimonRC has quit (Ping timeout: 246 seconds).
15:09:23 -!- coppro has quit (Quit: I am leaving. You are about to explode.).
15:19:12 -!- tombom_ has joined.
15:22:01 -!- tombom has quit (Ping timeout: 264 seconds).
15:40:14 -!- SimonRC has joined.
16:05:34 -!- cheater2 has quit (Ping timeout: 264 seconds).
16:10:23 -!- Lawlabee-L has joined.
16:10:32 -!- Lawlabee-L has changed nick to Gregor-L.
16:21:37 -!- Slereah has left (?).
17:29:25 <Deewiant> AnMaster: Timeouting lines are now visible in the memory-vs-time plots
17:32:22 -!- BeholdMyGlory has joined.
17:37:31 -!- augur has quit (Ping timeout: 260 seconds).
17:38:54 <AnMaster> Deewiant, busy today and tomorrow. I hate deadlines
17:39:34 <Deewiant> I hate gnuplot with gigabytes of memory usage
17:40:39 <Deewiant> Well, not really; I think I mostly hate awk and/or my disk being so slow that it takes half an hour to generate the plots
17:59:20 -!- MizardX has quit (Quit: brb).
18:06:12 -!- cheater2 has joined.
18:09:35 -!- augur has joined.
18:11:20 -!- Slereah has joined.
18:24:33 -!- augur has quit (Read error: Connection reset by peer).
18:25:10 -!- augur has joined.
18:32:35 -!- augur has quit (Ping timeout: 245 seconds).
18:39:48 -!- augur has joined.
18:43:47 -!- MizardX has joined.
18:48:11 -!- ehirdiphone has joined.
18:56:20 -!- oerjan has joined.
19:04:31 <ehirdiphone> anyway Deewiant, how would you like another contender for Fungicide, written by someone who doesn't demand to know every detail of every benchmark?
19:07:08 <ehirdiphone> Deewiant: Sufficient motivation levels reached; funge time.
19:07:38 -!- tombom has joined.
19:07:52 <Deewiant> It should get through Mycology, to start with :-P
19:07:53 -!- tombom_ has quit (Ping timeout: 258 seconds).
19:08:13 <ehirdiphone> Hey, I can do n-dimensional funge /and/ speed: specialise the code based on a co
19:08:33 <Deewiant> That's what CCBI already does :-)
19:08:41 <ehirdiphone> Now I need a language good enough to do that.
19:09:23 <Deewiant> Compile-time specialization for an infinite number of dimensions == infinitely big binary :-)
19:09:44 <ehirdiphone> No. You compile it for a certain dimension.
19:10:41 <oerjan> today is a momentous day.
19:10:53 <ehirdiphone> Maybe I'll write it in some wonderful Scheme macros that spit out C.
19:11:07 <oerjan> i know, it's unheard of
19:11:30 <Deewiant> Am I in the wrong year or something?
19:11:45 <oerjan> oh wait, i'm being an idiot
19:18:03 -!- Gregor-L has quit (Ping timeout: 268 seconds).
19:22:04 <Deewiant> ehirdiphone: FWIW I don't mind demands except when they've already been met on the site
19:23:34 <ehirdiphone> The naming, of course, is the most important thing. Ascus, shiro, or hyphae?
19:24:47 <Deewiant> Dictionary knew the other two but not that one
19:26:37 <Deewiant> It's shroom-related and distinctive: what could be better?
19:27:48 <Deewiant> If you want that angle, "Jah"?
19:28:11 <oerjan> yeah infinite dimensional funge should definitely use a psychoactive mushroom
19:28:54 -!- ehirdiphone has quit (Remote host closed the connection).
19:29:14 -!- ehirdiphone has joined.
19:32:05 -!- adam_d has quit (Ping timeout: 265 seconds).
19:39:56 -!- ais523 has joined.
19:47:19 -!- impomatic has joined.
19:49:01 <pikhq> Hmm. I wonder how small of a Brainfuck interpreter I could get.
19:49:24 <pikhq> (note: architecture is i386-linux-unknown, with the following programs: shish)
19:50:43 <impomatic> http://impomatic.blogspot.com/2009/01/brainf-interpreter-in-redcode.html
19:51:20 <impomatic> There was a contest about 10 years ago to write the smallest Brainfuck interpreter in x86
19:54:53 <ais523> didn't someone fit one entirely into ELF headers?
19:55:10 <pikhq> Compiler, and not quite.
19:55:28 <ais523> still, there's a fun article somewhere about someone trying to write the shortest possible ELF Linux binary, that just did return 42;
19:55:39 <ais523> and they even managed to fit many of the ELF headers into the ELF headers
19:55:43 <Deewiant> http://www.muppetlabs.com/~breadbox/software/tiny/teensy.html http://www.phreedom.org/solar/code/tinype/
19:56:05 <ehirdiphone> ais523: Also the guy who wrote the cgi interval game
19:56:18 <ais523> and Google's INTERCAL style guide
19:56:24 <ais523> which implies, indirectly, that he works for Google atm
19:57:18 <pikhq> Aaaaw. The Brainfuck compiler segfaults for LostKng.b
20:05:12 -!- ehirdiphone has quit (Quit: Get Colloquy for iPhone! http://mobile.colloquy.info).
20:09:23 -!- ehirdiphone has joined.
20:10:50 -!- ehirdiphone has quit (Client Quit).
20:25:46 -!- oerjan has quit (Quit: Now with even more afk).
20:33:59 -!- Oranjer has joined.
20:42:55 -!- adam_d has joined.
20:54:11 -!- Gregor-L has joined.
20:55:38 -!- augur has quit (Ping timeout: 246 seconds).
21:01:28 <Gregor> Why'd you rename shsh -> shish? Or did I misread the first time?
21:02:43 <pikhq> Gregor: Thought that was the name.
21:02:55 <pikhq> I've been reducing the binary size of shish even further.
21:03:04 <pikhq> Would you like the source to it ATM?
21:03:16 <Gregor> Not even a little bit.
21:03:19 <pikhq> http://sprunge.us/HcSN
21:03:35 <pikhq> Notable features include not depending on libc.
21:03:50 <Gregor> I hope those small* features get inlined.
21:04:02 <Gregor> Your f-word overwrote my f-word mid-sentence.
21:04:58 <Gregor> http://codu.org/aforteforpiano/19e.ogg I can (nearly) play the first page of Nocturne 3 :)
21:05:28 <pikhq> I could probably minimise things further by going into assembly, but, well. The shell is already sitting at 1107 bytes.
21:06:33 <Gregor> The fact that you have an int main suggests strongly that your assertion that it doesn't depend on libc is false.
21:07:02 <Gregor> The glibc-provided entry function does garbage you don't want or need.
21:07:12 <pikhq> dietlibc, actually. But yeah.
21:07:21 <pikhq> I'm going to go copy in a _start and _exit.
21:08:09 <Gregor> Just replace _start entirely.
21:08:11 <pikhq> Actually. Bah. I'm just going to rename main _start and add _exit.
21:08:36 <pikhq> Gregor: Copy it in from my "small-libc" directory. Which is where all those *other* functions come from. :P
21:11:21 <Gregor> You can replace ELF with a.out, can't you?
21:11:24 <pikhq> -nostdlib is such an awesome library.
21:12:10 <Deewiant> pikhq: If it's that small, I'd go ahead and take the -Os and hand-optimize it
21:12:44 <pikhq> Gregor: Yes, but I'd need to build a new compiler for that.
21:12:58 <pikhq> Or just manually munge the assembly into a.out.
21:13:05 <pikhq> Speaking of, does nasm still support a.out?
21:13:07 <Gregor> pikhq: Alternatively, you could use objcopy like somebody who isn't a moron.
21:13:20 <pikhq> I didn't know about objcopy.
21:13:34 <Gregor> And I didn't no how to phrase that not as an insult ^^
21:15:36 <pikhq> Hrm. Well, it's 32 bytes, but I get an exec format error.
21:15:41 <oklopol> so you *did* how to phrase it as such
21:16:01 <Gregor> pikhq: Most linuxes don't support a.out
21:16:06 <pikhq> Let me go attempt to compile in a.out support.
21:16:14 <pikhq> Gregor: Yeah, Imma build the module.
21:16:23 <Gregor> And remove ELF support too, who needs it :P
21:16:51 <pikhq> Gregor: No, I mean build the module for my *own* system. So I can test it without building a new kernel. ;)
21:17:30 <pikhq> Let's see if this 32-byte program is, in fact, a full shell.
21:18:42 <pikhq> zsh: exec format error: ./shish.1.1
21:19:20 <Gregor> Lesse your objcopy line
21:19:40 <pikhq> objcopy --output-target=a.out-i386-linux shish shish.1
21:19:50 <Gregor> Simple enough *shrugs*
21:19:56 <Gregor> So what's shish.1.1 then?
21:20:21 <pikhq> Erm. Different file.
21:20:58 <pikhq> Seems I have to use sstrip to get out the comment section from the ELF file so objcopy will actually... Copy.
21:21:08 <pikhq> And I still get an exec format error.
21:22:00 <Gregor> file doesn't give wonky results? Maybe your a.out support is borkleborked :P
21:22:04 <pikhq> Okay, then. objdump hates the a.out file.
21:22:15 <pikhq> Contents of section .text:
21:22:15 <pikhq> objdump: out of memory allocating 18446744073709551584 bytes after a total of 0 bytes
21:22:35 <pikhq> And the ELF file, it refuses to disassemble.
21:22:44 <Deewiant> Looks like the size of your .text section is -32 or something
21:22:45 -!- rapido has joined.
21:23:08 <pikhq> Without sstrip, it handles the ELF file correctly.
21:23:18 <pikhq> I'm going to guess sstrip does weird stuff.
21:23:51 <pikhq> Well, good to know.
21:25:21 <Gregor> Yuh, maybe just strip ... objcopy can also strip out sections if you use enough objcopy-foo.
21:25:36 <pikhq> objcopy --output-target=a.out-i386-linux -S -R .comment shish shish.1
21:25:49 <pikhq> That produces a file that makes objdump report a file of 0s.
21:25:59 <pikhq> shish.1: Linux/i386 demand-paged executable (ZMAGIC), stripped
21:27:31 -!- rapido has left (?).
21:33:18 -!- augur has joined.
21:35:23 <AnMaster> does openoffice actually not support svg???
21:36:07 <AnMaster> I guess I'll have to do with EPS. Lol
21:37:31 <Gregor> I wonder if you could make Linux support a DOS-COM-like format :P
21:37:44 <AnMaster> hm there is svg on my laptop in openoffice
21:38:18 <pikhq> That's nearly a.out.
21:39:53 <Gregor> pikhq: Minus the header ^^
21:40:30 <AnMaster> Gregor, sure. If you a) patch kernel b) write a custom user space loader
21:40:47 -!- augur has quit (Ping timeout: 246 seconds).
21:41:04 <AnMaster> or c) use some header + user space loader + the binfmt that allows selecting interpreter from header magic
21:44:37 <pikhq> BTW, it is possible to compile Linux without an executable format.
21:56:29 <pikhq> Hmm. Near as I can tell, the Linux support for a.out is a lie.
21:57:11 <pikhq> Given that I've generated an a.out file as close to the "proper" way as I can.
21:57:44 <pikhq> I generated a .o file, then stripped the .comment section, then had ld link and output an a.out file.
21:58:01 <AnMaster> why does openoffice drop text when importing svg
21:58:09 <AnMaster> and eps it ended up using the preview at all time
21:58:45 <pikhq> Ah well. a.out format appears to have overhead for this program, compared with ELF.
21:59:20 <pikhq> This program statically allocates a lot of stuff. a.out does the static allocation via a series of 0s in the file.
21:59:31 <pikhq> ELF notes the size that's been statically allocated.
22:00:39 <Gregor> ... why do you have any BSS at all?
22:00:46 -!- BeholdMyGlory has quit (Read error: Connection reset by peer).
22:01:23 <AnMaster> http://qa.openoffice.org/issues/show_bug.cgi?id=49991 <-- aaaaaargh
22:02:21 <pikhq> Though, I could just stick that on the stack and see what size I get.
22:09:44 -!- cheater2 has quit (Read error: Connection reset by peer).
22:09:59 -!- cheater2 has joined.
22:12:20 -!- augur has joined.
22:12:35 -!- adam_d has quit (Ping timeout: 245 seconds).
22:13:25 -!- oklofok has joined.
22:14:19 <pikhq> This is smaller than many of the tiny ELF utilities.
22:14:52 <pikhq> Not exactly normal C, though.
22:14:52 -!- oklopol has quit (Ping timeout: 252 seconds).
22:24:06 -!- tombom has quit (Read error: Connection reset by peer).
22:24:47 -!- tombom has joined.
22:27:50 <Gregor> It's in C ... for a ridiculous definition of C :P
22:29:33 <pikhq> http://sprunge.us/MEcJ
22:29:59 <pikhq> I'm not thinking of better ways to shrink that, short of removing the environment handling.
22:30:26 -!- ais523_ has joined.
22:30:30 -!- ais523 has quit (Ping timeout: 245 seconds).
22:31:10 <ais523_> hmm, I can reboot my computer and reinstall my wireless drivers in the time it takes me to pingout from Freenode? impressive
22:31:11 -!- ais523_ has changed nick to ais523.
22:31:21 <ais523> well, rebuild and reinstall
22:31:38 <ais523> I compile my wireless drivers from source as there isn't a binary package for them yet
22:32:04 <Deewiant> pikhq: if (*(unsigned*)tokened[0] == 'set\0')
22:35:32 -!- Alex3012_ has joined.
22:37:12 -!- Alex3012 has quit (Ping timeout: 265 seconds).
22:37:21 -!- Alex3012_ has changed nick to Alex3012.
22:44:23 <Deewiant> Smaller check for set than four separate tests
22:45:05 <Deewiant> And similarly for cd, although you'll need a shift
22:45:22 <pikhq> I... Don't think that works, though. The 'set\0' literal there, that is.
22:45:28 <pikhq> I'll futz with it in a bit.
22:45:59 -!- augur has quit (Ping timeout: 260 seconds).
22:46:20 -!- augur has joined.
22:46:27 <Deewiant> You can always write it out as an integer if you're worried :-P
22:48:08 -!- BeholdMyGlory has joined.
22:52:14 <pikhq> Doesn't seem to work right.
22:52:46 <Gregor> You're gonna run into endianness issues.
22:53:03 <pikhq> Yeah, that'd do it.
22:53:37 <Deewiant> Since you're assuming x86 anyway it's '\0tes', neh?
22:53:42 -!- tombom has quit (Quit: Leaving).
22:54:17 <pikhq> That said, making the if statements a *single* expression gets it down to 488 bytes.
22:54:33 <Deewiant> You compiling with optimization?
22:54:40 <Deewiant> Also, & is presumably smaller than &&
22:54:53 <Deewiant> Remember to compare -Os and -O3 too
22:55:15 <pikhq> I have been. -O3 is the second-smallest optimization ATM.
22:55:45 <pikhq> "&" is *larger* than &&.
22:55:47 <Deewiant> And if you just care about the binary size, I reiterate that I'd be working on the asm level :-P
22:56:08 -!- coppro has joined.
22:56:15 <pikhq> But I don't wanna!
23:00:32 -!- oerjan has joined.
23:11:17 <pikhq> http://sprunge.us/XUcK
23:12:02 <pikhq> ... Unsigned short? WTF?
23:12:39 <pikhq> Okay, it's shorter that way.
23:12:52 <pikhq> I guess accidents made my code shorter. *shrug*
23:13:42 <pikhq> Ah. But that's incorrect. Doesn't check the null.
23:28:50 -!- ais523 has quit (Remote host closed the connection).
23:31:16 -!- FireFly has quit (Quit: Leaving).
23:54:23 * Sgeo is writing Haskell in his [C++] Data Structures class
23:54:30 <Sgeo> Because I'd rather not be doing this
23:54:45 -!- jcp1 has joined.
23:56:05 <coppro> I write code in class that isn't related all the time
23:57:23 -!- augur has quit (Ping timeout: 260 seconds).
23:59:37 <Sgeo> This code is related, just in the wrong language