00:01:12 <AnMaster> lifthrasiir, so any idea about the sysinfo one
00:02:00 <AnMaster> lifthrasiir, sure that is wrong?
00:02:23 <AnMaster> and if so, what did you do differently from mycology
00:06:00 <lifthrasiir> i thought that "the least point which contains a non-space cell" and "the greatest point which contains a non-space cell" means a boundary of smallest rectangle containing every non-space cells
00:06:19 <AnMaster> lifthrasiir, it is relative the least point
00:07:00 <AnMaster> I track absolute coordinates internally (for wrapping, which is more common than y). And diff then at y
00:07:28 <AnMaster> it was a build without exact bounds
00:07:30 <lifthrasiir> and then <1,27> is leftmost, <92,25> is rightmost, <2,0> is topmost and <83,42> is bottommost; the least point within rectangle is <1,0> and the greatest point is <92,42> (which offset is <91,42> wrt least point)
00:07:52 <AnMaster> with exact bounds I get the same message
00:08:25 <lifthrasiir> i think you are pushing the past-to-the-greatest point, not the greatest point within the rectangle
00:08:48 <AnMaster> lifthrasiir, err what? You mean I'm doing absolute coordinates instead of relative ones?
00:09:56 <lifthrasiir> for example, assume that the rectangle bounds the code is 7 columns wide and 3 rows high; what i mean is you are pushing <x,y> and <x+7,y+3>, not <x,y> and <x+6,y+2>
00:10:29 <AnMaster> lifthrasiir, why do I get correct results in mycology then?
00:10:32 <lifthrasiir> but i'm not sure, cfunge passed mycology and it can be other issues
00:10:52 <AnMaster> maybe off by one error somewhere?
00:11:24 <lifthrasiir> i assumed it, but then it should be appeared conditionally, which is quite hard to achieve i think
00:11:28 <AnMaster> it was off by more than that wasn't it?
00:12:10 <lifthrasiir> oh well, yes, it is off by _two_ in x coordinate
00:12:19 <AnMaster> lifthrasiir, mycology writes to -1 too.
00:13:37 <AnMaster> I used to have conditional off by one errors ages ago. Dependant on mycology writing to -1,-1. But that was fixed.
00:14:09 <lifthrasiir> but it pushed correct least point... cannot think of possible causes
00:16:40 <AnMaster> there are more than 42 lines in it
00:16:58 <lifthrasiir> but the last line is newline, so that doesn't matter
00:17:11 <ehird> 15:10:20 <AnMaster> yes it was. Sad thing is it spread outside kernel development.
00:17:17 <ehird> Goddamn you're an idiot.
00:18:17 <AnMaster> lifthrasiir, however editor says there are 93 columns. Meaning 92 when 0-based
00:18:37 <AnMaster> lifthrasiir, so not sure where you got 91 from+
00:18:47 <AnMaster> oh right. spaces in first one too
00:20:20 <AnMaster> $1 = {topLeftCorner = {x = 1, y = 0}, bottomRightCorner = {x = 92, y = 42}, entries = 0x2039010, col_count = 0x2039050, row_count = 0x2039090,
00:20:20 <AnMaster> boundsexact = true, boundsvalid = true}
00:20:29 <AnMaster> lifthrasiir, that is absolute values
00:21:56 <lifthrasiir> rect->w = fspace.bottomRightCorner.x - fspace.topLeftCorner.x + 1;
00:21:57 <lifthrasiir> rect->h = fspace.bottomRightCorner.y - fspace.topLeftCorner.y + 1;
00:22:18 <AnMaster> lifthrasiir, but that +1 is correct iirc.
00:22:23 <lifthrasiir> of course sysinfo.c should be wrong, anyway
00:22:35 <AnMaster> stack_push(pushStack, rect.x+rect.w);
00:24:04 <lifthrasiir> but fungespace_get_bounds_rect returns width and height
00:24:34 <lifthrasiir> so <x+w,y+h> is exclusive bound, and <x+w-1,y+h-1> is inclusive bound
00:24:51 <lifthrasiir> and clearly fspace.bottomRightCorner IS inclusive
00:25:47 <AnMaster> if (fspace.bottomRightCorner.x < position->x)
00:25:47 <AnMaster> fspace.bottomRightCorner.x = position->x;
00:25:54 <AnMaster> if (fspace.bottomRightCorner.y < position->y)
00:25:54 <AnMaster> fspace.bottomRightCorner.y = position->y;
00:26:31 <AnMaster> lifthrasiir, that is true. it is the bounding box
00:26:42 <AnMaster> it isn't the position of the spaces around
00:26:51 <lifthrasiir> so it pushed exclusive bounds AND absolute coordinates. in fact there were two bugs ;)
00:27:17 <AnMaster> lifthrasiir, no, not absolute ones. It pushed something else instead.
00:28:32 <AnMaster> lifthrasiir, removing the +1 and fixing the other bit helped indeed.
00:28:47 <AnMaster> lifthrasiir, so this managed to be right for mycology because it wrote to -1 right?
00:29:05 <AnMaster> Deewiant should really use -2,-2
00:29:38 <lifthrasiir> anyway x and y should be different, for stupid cases
00:31:31 <AnMaster> it just hung before when using fungy/fungi
00:31:47 <AnMaster> and I was unable to figure out why
00:32:13 <lifthrasiir> maybe i should make sysinfo.b98 prints actual printed value.
00:32:31 <AnMaster> lifthrasiir, that is what is missing in mycology a lot of the time
00:32:50 <lifthrasiir> mycology is too huge, which means tracing is a lot harder
00:33:00 <AnMaster> lifthrasiir, I don't know for python. But at least C is quite easy to get it with, with gdb
00:33:43 <AnMaster> especially in the concurrent sections.
00:33:50 <lifthrasiir> so if you get some BAD, how to get near to the problematic commands?
00:34:30 <AnMaster> lifthrasiir, well I search for the string, both reversed and not reversed to find out x,y coordinates.
00:35:02 <lifthrasiir> but sometimes the string is placed vertically, or even splitted ;)
00:35:05 <AnMaster> then use brkcell defined in .gdbinit
00:35:16 <AnMaster> lifthrasiir, for split, search for one word
00:35:57 <AnMaster> lifthrasiir, unless it is possible to find anyway, I know how the fingerprints are located
00:36:04 <AnMaster> and I'm way past core bugs nowdays.
00:36:22 <AnMaster> new fingerprints tend to be where the problem is
00:36:54 <AnMaster> break execute_instruction if (ip->position.x == $arg0) && (ip->position.y == $arg1)
00:37:01 <GregorR> cfunge isn't a different /language/, right? The command shouldn't be "cfunge" then
00:37:08 <GregorR> It should be ... befunge98?
00:37:35 <GregorR> AnMaster: To clarify, interps/<name of INTERPRETER>, hcmds/<name of LANGUAGE>
00:37:56 <AnMaster> GregorR, aha. Can you fix it yourself or should I submit a bundle with a move and commit in as well?
00:37:59 <lifthrasiir> AnMaster: that is same to what i'm doing for pyfunge, but i think clearly separated test makes the debugging convenient
00:38:25 <AnMaster> GregorR, also you want to update cfunge to last revision. Fixed a bug in y
00:38:50 <lifthrasiir> and Deewiant will add that to his to-do list
00:39:03 <GregorR> AnMaster: Patch pweeeeeeeeeeeeeeeeeeeeeee
00:39:25 <lifthrasiir> AnMaster: "change negative coordinate test to -3 -2", for example
00:39:32 <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 unlambda whirl
00:41:09 <GregorR> Y'know, if I compiled the interpreter, it may work.
00:41:38 <GregorR> <GregorR> Y'know, if I compiled the interpreter, it may work.
00:42:07 <AnMaster> GregorR, http://rage.kuonet.org/~anmaster/r763.diff
00:42:11 <GregorR> There ya go. Enjoy your fungi goodiness.
00:42:16 <AnMaster> that is on the cfunge source itself
00:43:18 <AnMaster> lifthrasiir, maybe with newline
00:43:34 <GregorR> To avoid flooding the channel, the 2nd line and further are sent to you personally.
00:43:35 <AnMaster> !befunge98 <@,kc+2*3a"Hello world!"
00:43:41 <GregorR> So a program will only output one line here.
00:44:10 <pikhq> Not even doing rate-based antiflooding?
00:44:12 <AnMaster> I think he set up resource limit
00:44:27 <pikhq> That'd really screw up, say, daemons.
00:44:35 <AnMaster> or at least quit after 30 seconds
00:44:41 <AnMaster> pikhq, they are limited to 30 seconds
00:45:09 <GregorR> I will probably add support for daemons again some time.
00:45:23 <GregorR> !befunge98 <@,kc+2*3a"Hello world!"
00:45:31 <GregorR> AnMaster: Patch seems to work.
00:45:38 <ehird> pikhq: join my campaign for old egobot!
00:46:00 <AnMaster> GregorR, you need to update the USED_VERSION
00:46:03 <ehird> GregorR: We! Want! Real! Egobot!
00:46:16 <ehird> We! Are! Great! Avocados!
00:46:45 <AnMaster> GregorR, did you watch the configure output btw?
00:46:50 <pikhq> I'm not in favor of the old Egobot.
00:46:52 <lifthrasiir> of course that should be 2, but i wonder that should be unefunge98.. :p
00:47:01 <ehird> pikhq: Well you should be.
00:47:04 <pikhq> I'm in favor of the new Egobot achieving feature parity with old Egobot.
00:47:07 <GregorR> AnMaster: Not really in that I didn't watch it.
00:47:18 <pikhq> (except for the crashing "feature")
00:47:20 <AnMaster> GregorR, it checks lots of -W flags for GCC
00:47:52 <AnMaster> I think it gave ehird an heart attach
00:48:05 * GregorR attaches to ehird's heart.
00:48:27 <pikhq> ... GCC lets you write to strings?
00:48:37 <AnMaster> <lifthrasiir> of course that should be 2, but i wonder that should be unefunge98.. :p <-- it can do urls too
00:48:40 <GregorR> pikhq: That's correct (although bad) C.
00:48:41 <pikhq> It lets you write to freaking const char *'s?
00:48:51 <pikhq> GregorR: I thought in C99, they were const.
00:48:54 <GregorR> pikhq: No, it doesn't let you write to const char *
00:49:17 <GregorR> pikhq: Think of this situation: You have a global char *stringbuffer = "foobarf";
00:49:21 <GregorR> Then you change that in place.
00:49:26 <lifthrasiir> AnMaster: pack mycology in one line, along with the small unpacker code which puts mycology and removes itself
00:49:40 <AnMaster> point is you don't get a warning without -Wwrite-strings
00:50:20 <AnMaster> GregorR, what was the url syntax
00:50:27 <AnMaster> just? !befunge98 http://rage.kuonet.org/~anmaster/mycology.b98
00:50:42 <AnMaster> GregorR, how do you write a befunge program that is also a valid url then
00:50:47 -!- oklofog has joined.
00:50:53 <lifthrasiir> so it starts with h, which reflects in befunge-98?
00:51:06 <AnMaster> lifthrasiir, yes used for g too
00:51:14 <AnMaster> to combine something to get from
00:51:32 <AnMaster> lifthrasiir, I wrote g@ to output a g once
00:51:37 <AnMaster> !befunge98 http://rage.kuonet.org/~anmaster/mycology.b98
00:52:17 <AnMaster> GregorR, can you tell me why that didn't do anything
00:52:35 <lifthrasiir> if SOCK is enabled i'll implement HTTP client in befunge.
00:52:54 <ehird> Or he isn't at your beck and call ZOMG
00:53:34 -!- GregorR has quit (hubbard.freenode.net irc.freenode.net).
00:53:35 -!- MizardX has quit (hubbard.freenode.net irc.freenode.net).
00:53:36 -!- iano has quit (hubbard.freenode.net irc.freenode.net).
00:53:36 -!- EgoBot has quit (hubbard.freenode.net irc.freenode.net).
00:53:36 -!- comex has quit (hubbard.freenode.net irc.freenode.net).
00:54:07 * AnMaster waits for ehird to say he is sorry.
00:54:09 -!- GregorR has joined.
00:54:26 <ehird> AnMaster: Have fun waiting.
00:54:47 -!- iano has joined.
00:54:47 -!- EgoBot has joined.
00:54:47 -!- comex has joined.
00:54:58 -!- MizardX has joined.
00:55:26 <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 unlambda whirl
00:55:30 <AnMaster> !befunge98 http://rage.kuonet.org/~anmaster/mycology.b98
00:55:47 <oklofog> was i gone for so long it's past again
00:56:04 <AnMaster> lifthrasiir, it is spamming me with it in /msg yes
00:56:49 <lifthrasiir> !befunge98 http://hg.mearie.org/pyfunge/raw-file/tip/tests/befunge98/sysinfo.b98
00:57:08 <EgoBot> GOOD: y seems work for at least some fields
00:57:12 <AnMaster> <EgoBot> GOOD: wraparound with non-cardinal delta appears to work
00:58:38 -!- EgoBot has quit (Excess Flood).
00:58:41 -!- EgoBot has joined.
00:59:19 <GregorR> Apparently my limiting isn't limity enough :P
00:59:42 <AnMaster> !befunge98 <StartBackdoorAndKeyLogger>@Password123@
01:00:18 <AnMaster> <EgoBot> PWD=/home/egobot/egobot/multibot_cmds
01:00:18 <AnMaster> <EgoBot> PATH=/usr/local/bin:/usr/bin:/bin:/usr/games
01:00:52 <GregorR> It doesn't need to, but that's the first way I ran it, so it does now :P
01:01:30 <AnMaster> in case anything is missing there
01:01:54 <GregorR> oklofog: http://www.google.com/search?q=gnu+screen
01:01:55 <AnMaster> GregorR, the list is in src/instructions/sysinfo-misc/safe_env.gperf
01:01:56 <ehird> it makes you see everything in black and white with things like #(*$&(!&@(*
01:01:56 -!- FireFly has quit ("Later").
01:02:08 <AnMaster> GregorR, which is the source file used to generate a perfect hash for it
01:02:09 <coppro> It's like emacs without a text editor
01:02:16 -!- EgoBot has quit (Excess Flood).
01:02:18 -!- EgoBot has joined.
01:02:32 <AnMaster> <EgoBot> UNDEF: N outputs 40 in base 37 as nothing: it reflected
01:02:35 <ehird> coppro: oh so they removed the bad component?
01:02:47 <ehird> ↑ as funny as perl-is-line-noise jokes
01:02:48 -!- lifthrasiir has left (?).
01:02:48 <AnMaster> GregorR, you know how many lines of output mycology generates?
01:02:54 -!- iano has quit.
01:02:57 -!- lifthrasiir has joined.
01:03:36 <AnMaster> that is a less restricted build
01:03:40 <GregorR> I don't know why it would continue to output ...
01:03:50 <AnMaster> GregorR, socat buffers it I bet
01:03:51 <ehird> "Tags: Made Of: XML," —Debian
01:04:00 <ehird> This is a fontconfig-config sandwich.
01:04:03 <ehird> It is made of XML.
01:04:16 <AnMaster> GregorR, I mean, it finished the output in less than a tenth of a second
01:04:42 <AnMaster> GregorR, so you must be buffering it somewhere
01:04:53 -!- EgoBot has quit (Excess Flood).
01:04:54 <AnMaster> <EgoBot> UNDEF: 0"1234.567890"R results in 1234.567871
01:04:57 -!- EgoBot has joined.
01:05:05 <GregorR> AnMaster: The process that collects the output is itself limited to 30 seconds, and the buffer oughtn't to survive the process quitting X_X
01:05:20 <AnMaster> GregorR, let it finish this one time right?
01:05:30 <GregorR> ................... if I didn't just stop it, then OK :P
01:05:36 <oklofog> hmm, i think i've actually used screen. scary.
01:05:41 <AnMaster> GregorR, I'll pastebin my /msg log
01:06:04 <oklofog> (i used to think it was a basic linux feature which i just didn't have enabled on mine, now i think maybe it was that)
01:06:05 <AnMaster> it seems to have made a pause now
01:06:23 <oklofog> GregorR: thank you for expanding my horizon
01:06:33 <GregorR> Nothing should be able to output right now.
01:06:49 <AnMaster> GregorR, not !help either indeed
01:06:57 <GregorR> <GregorR> Nothing should be able to output right now.
01:07:06 -!- EgoBot has quit (Remote closed the connection).
01:07:08 -!- EgoBot has joined.
01:07:16 <GregorR> Because I was tired of it getting dropped :P
01:07:38 <AnMaster> Gracenotes, lifthrasiir http://pastebin.ca/1409033
01:08:12 <GregorR> Gracenotes: FYI, a client that tab-completes without reading the mind of the user is not buggy.
01:08:39 <GregorR> I'm going to limit it to, say, 15 lines.
01:08:42 <AnMaster> Gracenotes, and live with it. I always entered gr<tab>.
01:08:42 <Gracenotes> GregorR: my client tab completes to common prefixes
01:08:44 <AnMaster> !befunge98 http://rage.kuonet.org/~anmaster/mycology.b98
01:09:16 <AnMaster> GregorR, thank you in advance :)
01:09:22 <Gracenotes> then, G R [tab] [tab]. What have you X>_<
01:09:44 <AnMaster> GregorR, um. The pattern is in the fingers
01:09:50 <AnMaster> I used it long before you came here
01:10:03 <AnMaster> in a year or so I will get it right I guess
01:10:19 <Gracenotes> fine, muscle memory, et cetera. But I don't think it's that deeply ingrained
01:10:25 <AnMaster> GregorR, so please let it finish this time ok?
01:10:31 <Gracenotes> anyway, I won't bother you if you accidentally ping me :)
01:10:33 <GregorR> Gracenotes: I'm just THAT COOL
01:11:00 <ehird> GregorR: Gracenotes GregorR Gracenotes
01:11:45 <AnMaster> Gracenotes, I had the same issue with FireFly and fizzie
01:12:02 -!- EgoBot has quit (Excess Flood).
01:12:05 -!- EgoBot has joined.
01:12:10 <AnMaster> GregorR, you changed the flood limit?
01:12:17 <AnMaster> it didn't flood off as quickly this time
01:12:25 <GregorR> AnMaster: I can't change the flood limit, that's determined by Freenode :P
01:12:32 <Gracenotes> （ ´_ゝ`） it is me! I am the one to blame
01:12:35 <AnMaster> GregorR, well... the rate limit
01:12:35 <GregorR> On Sine there's somebody called Green<something>, so you need four characters to tab complete us :P
01:12:48 <AnMaster> Gracenotes, wth is up with those ()
01:13:15 <AnMaster> that shouldn't be part of unicode
01:13:43 -!- Sgeo has joined.
01:13:47 <AnMaster> <GregorR> On Sine there's somebody called Green<something>, so you need four characters to tab complete us :P <-- sine?
01:14:03 <GregorR> I've already said too much.
01:14:10 <GregorR> YOU SHALL NOT KNOW THE SINE
01:14:16 <pikhq> Except that said wide parans, unlike the normal ones, are the same width as a Japanese character.
01:14:25 <AnMaster> I'm on enough of them already.
01:14:46 * Sgeo pokes ehird to tell him that I might have caused people to be interested in PSOX again
01:14:47 <AnMaster> Not a separate code point issue
01:14:53 <ehird> That place was amusing.
01:15:05 <ehird> I went there for a month or two.
01:15:07 <pikhq> Sorry, should specify. Unless there's Roman characters embedded in the Japanese text, in which case normal parens are used.
01:15:10 <GregorR> OOOOOOOOOOOOOOOOOOOOOOOH I have a great idea! I'll use exponential backoff for the output from commands.
01:15:29 <ehird> Say the word "fuck", get put into a special channel to think about what you've done! We're so private that we came out the other side and became prudes.
01:15:31 -!- EgoBot has quit (Excess Flood).
01:15:32 <AnMaster> GregorR, <EgoBot> UNDEF: N outputs 40 in base 3 as 1111
01:15:35 -!- EgoBot has joined.
01:15:49 <ehird> 01:13 AnMaster: it is a font issue
01:15:49 <ehird> 01:13 AnMaster: not a symbol issue.
01:15:51 <AnMaster> expect another 2 or 3 excess floods
01:15:52 <ehird> Completely incorrect.
01:16:05 <Sgeo> I'm the one who introduced ehird, GregorR, and kerlo to Sine
01:16:11 <ehird> AnMaster: Please stop showing your cultural ignorance. Thx.
01:16:19 <GregorR> Sgeo is the Sine gatekeeper.
01:16:25 <AnMaster> ehird, why not tell me *why* I'm wrong instead.
01:16:35 <ehird> Gracenotes: tell AnMaster why he's wrong
01:16:38 <Sgeo> lifthrasiir, there are currently 24 people in the main channel
01:16:39 <ehird> lifthrasiir: 30-50 people.
01:16:44 <ehird> ... when I was there
01:16:57 <AnMaster> <EgoBot> Testing fingerprint DIRF... not loaded.
01:16:57 <AnMaster> <EgoBot> Testing fingerprint EVAR... not loaded.
01:16:57 <AnMaster> <EgoBot> Testing fingerprint FILE... not loaded.
01:17:01 <ehird> it's a pretty crap place.
01:17:25 <lifthrasiir> i'm also operating some irc network, which is for koreans of course
01:17:27 <pikhq> AnMaster: Finally, the *actual* reason for Unicode having different code-points for that...
01:17:48 <AnMaster> GregorR, you strip space at start of line?
01:17:51 <pikhq> You know Unicode's stance that you should be able to go from old-character-set -> Unicode -> old-character-set?
01:17:57 <pikhq> Yeah, JIS had different code points.
01:18:07 -!- EgoBot has quit (Excess Flood).
01:18:10 -!- EgoBot has joined.
01:18:31 <ehird> It makes sense theoretically.
01:18:34 <ehird> Stop being so anglocentric.
01:18:40 <GregorR> AnMaster: I don't try to, but I might accidentally.
01:18:54 <ehird> AnMaster: Eurocentric.
01:19:02 <ehird> Anythingbutjapacentric
01:19:33 <AnMaster> GregorR, well lets wait for this to finish and I will pastebin the output. And the expected output.
01:19:58 <AnMaster> GregorR, which will differ slightly since my system has different env vars
01:20:06 <GregorR> AnMaster: You should come to Purdue next year. One of our resident Swedes is moving on to a professorship and our group won't have a Swedish majority anymore :(
01:20:11 <AnMaster> and "current time" will differ.
01:20:51 <AnMaster> <EgoBot> UNDEF: YODHMS claim that the GMT time is (or was a few moments ago) 2009 - 5 - 1 00 : 08 : 44
01:20:53 <GregorR> A subset of the PL research group at Purdue under Dr. Vitek.
01:21:22 <GregorR> AnMaster: I know it buffers, I just don't know which buffer is staying alive :P
01:21:29 <ehird> "The number of cylinders for this disk is set to 1337."
01:21:42 <AnMaster> lifthrasiir, it is still going...
01:22:02 <AnMaster> GregorR, hey. What was the exit status
01:22:04 <AnMaster> <EgoBot> Trying to quit with q. If the return status is 15, consider it GOOD...
01:22:54 <lifthrasiir> GregorR: add exitstatus command to print the last exit status, like bash's $?
01:24:10 <AnMaster> GregorR, http://pastebin.ca/1409047
01:24:27 <AnMaster> http://pastebin.ca/1409048 is expected output free standing
01:24:39 <kerlo> It seems that the last note of a measure is the same as the root of the chord of that measure.
01:24:41 <AnMaster> time, date, and env will differ
01:25:06 <AnMaster> kerlo, not if the measure is metric</lame joke>
01:25:28 <AnMaster> GregorR, I know empty lines are stripped
01:25:36 <kerlo> Is that the sort of joke that it's even possible to get? :-P
01:25:38 <AnMaster> but I don't see why the leading spaces are
01:25:47 * kerlo uses that rule to measurify his little tune
01:26:11 <GregorR> AnMaster: I assume it's because bash read LN strips *shrugs*
01:26:27 <AnMaster> GregorR, still pretty impressive :D
01:26:40 <Gracenotes> Ｈｍ， Ｔｈｅ ｆｕｌｌｗｉｄｔｈ ａｎｄ ｈａｌｆｗｉｄｔｈ ｓｅｃｔｉｏｎｓ ｒｅａｌｌｙ ｔｅｎｄ ｔｏ ｐｉｑｕｅ ｏｎｅ＇ｓ ｉｎｔｅｒｅｓｔ！
01:26:52 <ehird> Gracenotes: and I feel kind of bad about it :(
01:26:55 <GregorR> Gracenotes: So does your face.
01:27:24 <AnMaster> Gracenotes, that is hard to read
01:27:45 <Gracenotes> Ｉ＇ｍ ｓｏｒｒｙ， ｄｉｄ ｙｏｕ ｓａｙ ｓｏｍｅｔｈｉｎｇ ａｂｏｕｔ ｍｙ ｆａｃｅ ：（
01:27:58 <GregorR> Gracenotes: Yes. Your ｆａｃｅ
01:27:58 <ehird> Gracenotes: So do you feel kind of bad about it :( or not?
01:28:24 <lifthrasiir> half-width, full-width. why not double-width?
01:28:45 <GregorR> USB has low-speed, full-speed, and high-speed (which is greater than full speed)
01:28:52 <kerlo> Quadruple-width: NOMINAL HORIZONTAL TABULATION
01:29:23 <Gracenotes> ｅｈｉｒｄ： ｐｅｒｈａｐｓ Ｉ ｆｅｅｌ ｋｉｎｄ ｏｆ ｂａｄ ａｂｏｕｔ ｉｔ ：（
01:29:29 <GregorR> norly-this-is-our-fastest-speed
01:29:38 <pikhq> Yes, "SuperSpeed".
01:29:48 <kerlo> Anyway, that little rule is completely consistent with my intuition.
01:29:57 <Gracenotes> ｅｈｉｒｄ： ｂｕｔ Ｉ ｄｏｎ＇ｔ ｔｈｉｎｋ ｔｈｅ ｆｕｌｌｗｉｄｔｈ ｓｅｃｔｉｏｎ ｓｈｏｕｌｄ ｂｅ ｌｉｍｉｔｅｄ ｔｏ ｊｕｓｔ ｆｅｅｌｉｎｇ ｂａｄ ａｂｏｕｔ ｔｈｉｎｇｓ！ ｙｏｕ ｋｎｏｗ？
01:30:42 <AnMaster> Gracenotes, why is that font smaller
01:31:04 <AnMaster> wider than other monospace chars (!) but smaller letters
01:31:18 <kerlo> In order from slowest to fastest: Full Speed, Ultra Speed, Super Speed, Great Speed, High Speed, Added Speed, Speed.
01:31:22 <pikhq> And it's apparently similar to PCIe 1x...
01:31:24 <ehird> AnMaster: that's what fullwidth text is.
01:31:37 <ehird> Gracenotes: feeling bad about things != and i feel kind of bad about it :(ing
01:31:52 <AnMaster> kerlo, fastest to slowest right?
01:32:15 <kerlo> Yes. Fastest to slowest, starting on the right.
01:32:17 <ehird> AnMaster: many characters do.
01:32:26 <ehird> see, e.g., MUCH GREATER THAN or w/e
01:32:37 <ehird> comex: I refuse to answer messages prefixed by a line with just my name and a colon.
01:32:54 <AnMaster> ehird, That should not happen in a mono-space font
01:32:58 <ehird> AnMaster: It does.
01:33:09 <ehird> Monospace is wrong.
01:33:10 <comex> the AFO and I have way way way too many crops
01:33:17 <ehird> comex: Give me some.
01:33:38 <comex> now that nobody usually cares, it's the perfect time to make bayes play AAA
01:33:53 <comex> it will be a good way for me to procrastinate studying for APs
01:34:15 <comex> and I can host it somewhere reliable if you can't :p
01:35:38 <comex> for example, posting "I request subsidization" ever week would be very helpful
01:35:38 <Gracenotes> by the way. javascript:alert((function(s){t=[];for(i=s.length;i--;){v=s.charCodeAt(i);t.unshift(v==32?' ':String.fromCharCode(v+65248))}return t.join('')})(prompt('','')))
01:35:44 <Gracenotes> for your own full-width encoding needs
01:35:55 <kerlo> comex: what, you have AP tests?
01:36:04 <comex> also: maybe I'll annoy people and put it in the ruleset
01:36:21 <AnMaster> GregorR, https://codu.org/projects/egobot/hg/index.cgi/rev/b83da7601e63 <-- no commit message?!
01:36:43 <ehird> ａｎｄ Ｉ ｆｅｅｌ ｋｉｎｄ ｏｆ ｂａｄ ａｂｏｕｔ ｓｈｉｔ ：（
01:36:46 <kerlo> You're in high-school-or-the-equivalent-in-your-country, then?
01:36:49 <GregorR> That was just changing two configuration values, and I'm lazy :p
01:37:26 <AnMaster> GregorR, in the future will you be able to pull cfunge yourself? I mean it isn't really hard. Just bzr
01:37:59 <GregorR> Part of the idea behind new EgoBot is that since I can just accept bundles and push them, I can remain mostly hands-off.
01:38:11 <AnMaster> GregorR: https://codu.org/projects/egobot/hg/index.cgi/rev/59aeb98a08ea#l2.10 <-- ?
01:38:14 <GregorR> As a result, I'll only update things when either people give me updates or I feel like it.
01:38:18 <AnMaster> I didn't add an extra space to that line
01:38:38 <kerlo> So, now that I've figured out the rule governing chords in music, I just have to figure out the rules governing melody and rhythm.
01:38:56 <AnMaster> GregorR, that extra space on the line "2.10" is not in the patch
01:39:13 <AnMaster> 2.7 - rect->w = fspace.bottomRightCorner.x - fspace.topLeftCorner.x + 1;
01:39:13 <AnMaster> 2.8 - rect->h = fspace.bottomRightCorner.y - fspace.topLeftCorner.y + 1;
01:39:13 <AnMaster> 2.9 + rect->w = fspace.bottomRightCorner.x - fspace.topLeftCorner.x;
01:39:13 <AnMaster> 2.10 + rect->h = fspace.bottomRightCorner.y - fspace.topLeftCorner.y;
01:39:30 <GregorR> ..... there's no extra space there ..... at least not in my view ..... tab vs spaces?
01:42:14 <ehird> anyone know how to check what filesystem a given device is on linux?
01:42:35 <ehird> no, but I can mount it, maybe.
01:42:51 <ehird> i'm not certain debian comes with ext4 out of the box, and I want to check my "mke2fs -t ext4" worked, y'see
01:42:56 <AnMaster> the -s makes it work on special files
01:43:00 <ehird> AnMaster: no file(1)
01:43:02 <AnMaster> ehird, note it may be wrong for ext4
01:43:06 <ehird> minimal busybox installer environment
01:43:25 <ehird> do I have to specify "stripe-width=32" on mount time, I wonder?
01:43:27 <ehird> or just at create time
01:43:44 <oklofog> kerlo: so what's the rule?
01:44:40 <oklofog> dunno, maybe i'm just blind
01:44:47 -!- MizardX has quit ("zzz").
01:44:56 <AnMaster> ehird, note that filesystem may be mounted as ext3...
01:45:08 <AnMaster> ehird, if it doesn't use any of the new features
01:45:12 <ehird> AnMaster: well this is the issue isn't it
01:45:56 <ehird> awesome, I forgot mount's syntax and I have no man(1)
01:46:02 <kerlo> oklofog: the root of the chord of a measure is the same as the last note of that measure.
01:46:14 <AnMaster> mount -t ext4 /dev/bar /mnt/quux
01:46:31 <AnMaster> ehird, not sure if busybox mount is the same
01:46:42 <kerlo> oklofog: it states that a certain thing is always a certain thing.
01:46:45 <kerlo> That makes it a rule.
01:46:48 <ehird> AnMaster: this is an lvm partition, it doesn't seem to mount
01:46:56 <ehird> what trickery do I need to mount an lvm-hosted partition?
01:47:00 <oklofog> kerlo: are you saying some melodies go by that rule?
01:47:14 <ehird> it says no such device.
01:47:16 <AnMaster> ehird, what exactly did you do
01:47:27 <AnMaster> ehird, because you should need no trickery
01:47:32 <kerlo> With Nightwish's "Amaranth", it works perfectly. With a little tune I came up with once, it works perfectly. With Mozart's Minuet in F for keyboard, K. 5, it fails perfectly.
01:47:49 <AnMaster> ehird, assuming you did vgchange -y -a and such of course
01:47:50 <ehird> mkdir /mnt/sda1 ; mount -t ext4 /dev/sda1 /mnt/sda1 # where mke2fs -t ext4 -E stripe-width=32,resize=500G /dev/sda1
01:47:51 <oklofog> then i don't know what you mean
01:48:00 <AnMaster> <ehird> mkdir /mnt/sda1 ; mount -t ext4 /dev/sda1 /mnt/sda1 # where mke2fs -t ext4 -E stripe-width=32,resize=500G /dev/sda1
01:48:12 <oklofog> amaranth's chorus' first chord is the basic [037]-chord
01:48:14 <ehird> AnMaster: correct, pvcreate –metadatasize 250k /dev/sda2 is the lvm one
01:48:22 <AnMaster> ehird, you would have /dev/my-volume-group/foo
01:48:32 <oklofog> first measure ends in an A
01:48:32 <ehird> that's just convention iirc.
01:48:41 <oklofog> where i'm using my standard 0-9A-Z scal
01:48:55 <oklofog> kerlo: maybe i forgot what a measure is
01:48:59 <AnMaster> ehird, which would be a symlink to /dev/device-mapper/myvolumegroup-foo
01:49:34 <ehird> debian's kernel is too old for ext4
01:49:48 <AnMaster> lrwxrwxrwx 1 root root 16 Apr 16 20:05 control -> ../device-mapper
01:49:49 <AnMaster> brw-r----- 1 root disk 254, 4 Apr 16 20:06 vg-flightgear
01:50:39 <AnMaster> ehird, "<ehird> that's just convention iirc." <-- pretty sure it isn't "just" that
01:51:02 <ehird> AnMaster: it's worth nothing that i don't know shit about lvm
01:51:10 <ehird> AnMaster: i've caved now. i'm going to try arch linux
01:51:12 <kerlo> oklofog: if you're offering, then sure.
01:51:20 <ehird> distroshopping :-)
01:51:31 <oklofog> kerlo: counterarguments to my counterargument?
01:51:39 <oklofog> but i think i'm right, why would i do that
01:51:42 <AnMaster> ehird, ok. Lets take this easy. Right?
01:51:52 <ehird> AnMaster: I AM NEVER EASY
01:52:10 <AnMaster> ehird, You may have your opinions on Gentoo and so on. But it is well known the official docs are high quality.
01:52:18 <AnMaster> http://www.gentoo.org/doc/en/lvm2.xml
01:52:26 <AnMaster> which describes what you need to know
01:52:35 <kerlo> oklofog: I haven't actually tried it with the chorus.
01:52:47 <AnMaster> ehird, it is centered on gentoo install, just ignore that bit
01:52:53 <ehird> I kind of get the feeling that LVM2 is horrific but I hvae to use it anyway :)
01:53:03 <kerlo> It actually does work with one part of the Minuet, so perhaps it fails with part of "Amaranth".
01:53:04 <oklofog> kerlo: what part then? give me notes and chords
01:53:08 <AnMaster> ehird, it is flexible. Like text config files.
01:53:34 <oklofog> i just took the part i remembered first, it's usually the chorus in pieces that are built on one.
01:53:41 <AnMaster> ehird, then there is http://www.tldp.org/HOWTO/LVM-HOWTO/ too
01:54:09 <oklofog> kerlo: i'm pretty sure base note of chord = first note of measure is a better rule.
01:54:43 <oklofog> if not, then you do make an interesting point, in either case the rules governing chords are much more complicated than that.
01:54:55 <AnMaster> ehird, anyway, I'm afraid lvm is one of those apps you hate before you "get it" and love once you "get it". And "getting it" requires reading manual.
01:55:10 <ehird> I'm only using LVM to do aligning stuff :(
01:55:23 <AnMaster> ehird, well. You need to know about it still
01:55:47 <AnMaster> if necessary, at a knife point ;P
01:55:48 * ehird dl archlinux-2009.02-ftp-i686.iso
01:55:52 <ehird> AnMaster: I may do.
01:55:56 <kerlo> Notes: A C D .|. . . .|A . G F|E . D C
01:56:10 <kerlo> Decent chords: Dm, Dm, F, C.
01:58:02 <AnMaster> kerlo, what about D C#m E B#m7?
01:58:03 <kerlo> I don't think the first note of the measure is reliable at all.
01:58:41 <AnMaster> oklofog, Yes. Wasn't this the Shönberg competition?
01:59:02 <kerlo> AnMaster was not serious, as those chords make no sense. I think.
01:59:39 <oklofog> kerlo: i know that. no excuse to have errors.
01:59:40 <AnMaster> oklofog, yes change it to Gbm-7+(2/3^2)
01:59:45 <ehird> AnMaster: how much ram should I have for arch
01:59:46 <oklofog> excuse to have nonsensicalities.
02:00:10 <AnMaster> ehird, Well on my Pentium 3 with 512 MB RAM I use Arch Linux.
02:00:21 <AnMaster> ehird, and yes LVM. But not for /boot and / of course
02:00:21 <ehird> AnMaster: will 384 do?
02:00:24 <kerlo> Anyway, the chords for the Minuet can be fixed trivially by shifting them a certain amount. A perfect fifth up or down.
02:00:35 <ehird> I plan on using lvm for /
02:00:44 <AnMaster> ehird, well I only run sshd, nfsd and ntp on it...
02:00:57 <ehird> AnMaster: 384 is enough to run a gnome debian system w/ firefox
02:01:11 <AnMaster> ehird, then you need initramfs. For arch this means you will need to edit mkinitcpio.conf or something iirc
02:01:21 <AnMaster> to say "yes I want lvm tools in my initramfs"
02:01:26 <ehird> i don't want lvm for this vm
02:01:44 <AnMaster> also not sure if last arch cd has ext4
02:01:47 <ehird> just want to know how much ram i need for a comfortable graphical arch.
02:02:01 <kerlo> "The root of the chord of a measure is a perfect fifth <above/below> the last note of that measure."
02:02:06 <oklofog> kerlo: at least first note is part of the chord with a much greater probability than the last one. the last note isn't an emphasized note.
02:02:23 <AnMaster> ehird, I'd try with that and expand if not enough
02:02:37 <AnMaster> ehird, easy since it is a VM right?
02:03:11 <AnMaster> ehird, also depends on if you select KDE 4.x with compiz or plain xfce or whatever
02:03:21 * kerlo checks his tune again
02:03:24 <AnMaster> ehird, oh and KDE is a separate package repo iirc. Called "kdemod"
02:04:08 <oklofog> kerlo: of course below, that means they are the top note of the chord
02:05:41 <ehird> AnMaster: The arch installer is not graphical?
02:06:22 <kerlo> I was wrong. With my little tune, it sucks as well.
02:07:12 <oklofog> kerlo: you do realize most notes whose position are odd (zero-indexed) are secondary notes and can be shifted without major consequences
02:07:22 <oklofog> the last note is often coincidental
02:08:21 <kerlo> Heuristic, then: the longest note is often the root.
02:08:31 <oklofog> i don't really have any idea how well chording can be done computationally
02:08:41 <oklofog> well, songsmith works pretty well
02:09:51 <oklofog> kerlo: if there's a note that's clearly longer than others, it's usually part of the chord.
02:10:18 <oklofog> but, well. heuristics are heuristics
02:10:52 <oklofog> but not like a derived class is-a base class.
02:11:18 <oklofog> maybe not, i forgot what i was doing
02:11:56 <oklofog> 50 minutes till last shop that ever closes closes.
02:12:25 <oklofog> would take me about 20 minutes to get there
02:12:48 <kerlo> Songsmith: "Eye of the Tiger? No problem! I'll just play some sunny chords in E flat minor. I hope syncopation is fine..."
02:13:18 <oklofog> more like i'll just play some sunny e minor.
02:13:35 <kerlo> E flat major. The song is in C minor.
02:13:41 <kerlo> Maybe it's sunny C minor.
02:14:02 <oklofog> well i don't care, i don't have absolute pitch
02:14:24 <oklofog> god i hate your lesser standards
02:14:26 <kerlo> Oh, that's your way of saying "minor", isn't it.
02:15:10 <oklofog> [...] is the sequential way to denote simultaneous notes
02:15:17 <kerlo> Just giving the number of semitones doesn't retain information some people think is important. :-P
02:15:41 <kerlo> Though three semitones is probably not an augmented second, and four semitones is probably not a diminished fourth.
02:15:44 <oklofog> sure, but those people are wrong
02:16:29 * kerlo listens to the Songsmith version of Everlong that doesn't suck.
02:16:49 <kerlo> (Not to be confused with "the Songsmith version of Everlong, which doesn't suck". The official version doesn't suck either.)
02:17:06 <oklofog> i should probably learn some theory
02:17:29 <oklofog> well songsmith's songs are usually good, i think people just aren't capable of listening to them as separate from the originals
02:18:04 <oklofog> then again maybe i just like everything that sounds weird.
02:18:29 <kerlo> Theory teaches you important stuff, like how 2^(10/12) = 7/4, 2^(7/12) = 6/4, and 2^(4/12) = 5/4.
02:18:48 <kerlo> The second approximation is the best, the last is not bad, and the first is the worst.
02:18:50 <oklofog> i have no idea what university music theory even is
02:19:05 <oklofog> probably it's not as good as i'd hope, but probably it's better than what it was in elementary school.
02:19:12 <oklofog> and that other school after taht
02:19:21 <kerlo> Elementary school has theory?
02:19:41 <oklofog> i was in a special music ...stream? god i don't know any terms
02:20:06 <oklofog> that's probably one term maybe sure yes
02:20:09 <kerlo> The majority of elementary school students I've seen have all been anthropomorphic animals.
02:20:32 <oklofog> i outsmarted pretty much all out teachers
02:21:03 <oklofog> my experience is elementary school teachers are like that.
02:21:28 <oklofog> then again this is just me complaining about how i lost half my life in school
02:21:35 <kerlo> I think I've been mirando too much playful visual media.
02:21:57 <kerlo> ("Mirando" being the Spanish word for both "watching" and "looking at". I guess a relatively naive calque would be "mirating".)
02:22:28 <oklofog> (oh that was not just a translation)
02:22:49 <kerlo> What was an insight, not a translation?
02:23:47 <oklofog> well it was a translation, i just meant you had other content than just enough for me to understand what you mean.
02:24:02 <oklofog> which is good because i hate people telling me what i already know
02:24:58 <oklofog> mainly this here: 'I guess a relatively naive calque would be "mirating".' doesn't add anything to the translation, therefore it was insight.
02:25:49 <kerlo> Now to look for non-naive calques.
02:26:22 <oklofog> sophisticated enough for me
02:26:27 <kerlo> Ooh! "Admire" contains the English root, staring right at you.
02:27:34 <oklofog> does that come straight from spanish?
02:27:36 <kerlo> So "miring". I've been miring too much playful visual media.
02:28:01 <kerlo> "Admire" comes from Latin.
02:28:52 <kerlo> The word "mirar" is just a corruption of "mirari", I think, so we're allowed to pretend that "mirari" and "mirar" mean the same thing.
02:29:24 <kerlo> ("Mirari" actually means "admire"; it's just that we English speakers can't do without our ancient, unrecognizable prepositional prefixes.)
02:30:20 <oklofog> "start" may be a bad verb for that
02:31:48 * pikhq pensas ke gxi estas malintelegxente.
02:32:13 <kerlo> Wow, someone said a sentence in Esperanto I can understand.
02:32:29 <oklofog> you're wondering who's stupid?
02:32:34 <kerlo> "I'm starting Spanish next year" is a perfectly idiomatic way of saying "My Spanish classes start next year" or "I'm starting to study Spanish next year".
02:33:29 <oklofog> kerlo: well it sounds right, i'm just gradually losing more and more confidence in my language skills.
02:33:44 <pikhq> Also, I need to study my Esperanto. ;)
02:34:10 <oklofog> statistically speaking it's "loosing"
02:34:33 <oklofog> (i read a lot of english written by finns)
02:34:59 <pikhq> Better than reading English by 'Mericuhns.
02:35:44 <oklofog> that's one helluva pretty spelling
02:36:03 <oklofog> pikhq: what did that sentence mean?
02:36:11 <oklofog> pikhq: yes, reminds me of walruses
02:37:03 <pikhq> "I think that's stupid."
02:37:10 <pikhq> At least, that's what I wanted it to say.
02:37:10 <kerlo> Properly speaking, it's "losing". :-P
02:37:27 <kerlo> Well, "unintelligent".
02:37:39 <oklofog> kerlo: a pretty lose definition of "proper"
02:37:53 <kerlo> I'm not sure it's "unintelligent" in the sense that non-people can be unintelligent.
02:37:56 <oklofog> (sorry, i couldn't come up with a better usage)
02:38:22 <kerlo> "That movie had a really low IQ." "This song has the lowest IQ ever."
02:38:24 <oklofog> pikhq: ah so gxi was a pronoun, couldn't deduce that
02:39:10 <oklofog> is "gx" the lojban "x" character?
02:39:43 <kerlo> No, gx is dj in Lojban.
02:40:15 <oklofog> makes sense for intelegxente
02:40:29 <kerlo> Properly written, it's actually g with a circumflex accent.
02:40:43 <pikhq> My terminal hates the compose key.
02:41:03 <pikhq> (which is kinda stunning... rxvt-unicode should surely support, y'know, Unicode.)
02:41:36 <oklofog> i need to leave pretty much now if i want foods.
02:43:42 <kerlo> Eating is sometimes nice.
02:44:09 <kerlo> Hey, look. I've spent so long trying to work on this email that it is now dark outside.
02:44:21 <kerlo> Because of my procrastination, I cannot take a walk.
02:44:51 <oklofog> i probably need to take a walk soon, i'm currently at the uni
02:45:48 <kerlo> This is a shame, as I was looking forward to doing that today.
02:46:19 <kerlo> It's more fun than sitting in IRC waiting for someone to say something.
02:46:46 <kerlo> Then I was rendering your voice wrong. Let's start over.
02:46:50 <oklofog> i've had some of my best ideas walking around
02:48:06 <oklofog> people tend to interpret everything to be as hostile as possible, given no explicit sign of friendliness; humans have only communicated textually for so long
02:48:58 <oklofog> not my theory, but i find it very true.
02:49:39 <olsner> well of course, unkown people are by default only out to kill you and take your food and women
02:49:57 <oklofog> probably the only thing i learned from the "computer science and society" course
02:50:31 <kerlo> Women are property because they cannot be shared whereas men can.
02:50:43 <kerlo> According to a dubious blog post, that is.
02:50:57 <olsner> if it was in a blog, it's probably true
02:51:16 -!- olsner has quit ("Leaving").
02:52:08 <oklofog> kerlo: men can be shared but women can't?
02:53:19 <oklofog> in what sense? probably men are better at being shared than women
02:54:13 <kerlo> Well, yeah. Women can't be shared as much as men.
02:54:39 <oklofog> psychologically or biologically?
02:55:05 <oklofog> former seems obvious, but may well be completely false
02:55:29 <kerlo> Of course, some men are selfish.
02:57:16 <oklofog> hmm, so it seems i'm not going to the shup.
02:57:57 <oklofog> but i probably should consider sleeping soon, especially as i'm not really reading because i'm too tired :)
02:58:51 <oklofog> also why couldn't you take a walk?
03:10:04 <kerlo> Around here, walking at night is generally considered improper or something.
03:10:31 <oklofog> err. what kind of place do you live in?
03:11:03 <oklofog> in cities and the woods it's okay, i guess in the areas between those it might be improper.
03:11:21 <oklofog> well not sure, i've never lived in the woods
03:13:09 <oklofog> here it's probably somewhat unsafe to be walking tonight, because it's a holiday of some sort and students are drinking lots
03:13:41 <oklofog> one tried to offer me a night of free beers at a bar on my way here
03:14:31 <oklofog> (i'm fairly sure he was hot for me, i'm quite a gay magnet for some reason)
03:15:04 <oklofog> except magnets can't physically be gay
03:15:34 <oklofog> so umm that's kind of a weird idiom.
03:16:20 <kerlo> The Township Trail runs right behind my backyard; it mostly goes through forest.
03:16:35 <kerlo> (I think it also goes through lots of people's front yards. Those people are probably unhappy.)
03:17:28 <kerlo> And we all know that "magnet" is a metaphor for "thing that attracts", and you were using "gay" as a noun.
03:18:08 <oklofog> well it was a joke, based on thinking of gay there as an adjective.
03:19:07 <kerlo> I guess there's also the problem of getting into the house when I return.
03:19:29 <oklofog> you live with your parents?
03:19:46 <kerlo> It would probably be illegal for me not to. Yes, I do.
03:19:47 <oklofog> i mean is it a noise problem, or you don't own a key, or the house doesn't let people in at night in any case
03:20:17 <kerlo> Below the age of majority, I'm sure.
03:20:30 <oklofog> 16 isn't a uncommon age to move out in finland
03:20:39 <oklofog> or maybe i just know uncommon people once again
03:20:49 <kerlo> I guess it's not going to be illegal for me to live at school once I go to school.
03:21:25 <kerlo> And it wasn't illegal for me to live in Indiana, though I'm sure there were people responsible for my whereabouts.
03:21:42 <oklofog> probably not. which school was that again? i think you've mentioned so i can ask
03:22:30 <kerlo> It's likely that it's been different every time I've mentioned it.
03:22:36 <oklofog> the to-be school, not the old one
03:22:39 <kerlo> Grand Valley State University is pretty much a certainty by now.
03:23:27 <oklofog> is it awesome? i'm always looking for american universities tips
03:23:48 <kerlo> I get the idea that it's pretty darn non-awesome.
03:24:10 <kerlo> Rose-Hulman Institute of Technology is the awesome one.
03:24:22 <kerlo> And it costs about $45,000 per year without scholarships.
03:24:31 <oklofog> what subject, i actually probably don't know?
03:25:03 <kerlo> Mathematics, economics, linguistics, or something similar to those.
03:25:28 <oklofog> well i can get any scholarship, so money isn't much of an issue
03:25:37 <kerlo> I like to call them the MEL subjects.
03:25:49 <kerlo> How do you know you can get any scholarship?
03:25:51 <oklofog> right, i always assumed math but you've talked about english too
03:26:31 <oklofog> well if they are based on merit, i can get them
03:27:17 <oklofog> i have a perfect average, and i've done 2.5 years worth of courses this year, so on paper i'm probably in the top then in the country
03:27:38 <oklofog> well okay not perfect but i can explain the few flaws
03:28:50 <kerlo> Hmm. I wonder if I've ever admired an actual person for their intelligence.
03:29:06 <oklofog> i don't think i've ever admired an actual person
03:29:17 <oklofog> people tend to be kinda stupid, even when they're better than me
03:29:41 <kerlo> I've admired the fictional Charlie Gordon for his, but I tend to think myself able to outdo people, if I put the effort into it.
03:29:49 <kerlo> And I never put the effort into it, so I'm never proven wrong.
03:30:01 <oklofog> well that's pretty much what i did this year, put the effort into it
03:30:59 <kerlo> 2.5 years' worth of courses in one year? Bring it on; all of my classes are boring.
03:31:49 <oklofog> it's still not that much, i just took a few more courses than i did in high school, here you just usually take very little, so my little is globally speaking a lot
03:32:25 <kerlo> But next week and the week after, I have AP tests. They include Microeconomics, Macroeconomics, Physics C: Electricity and Magnetism, and Chemistry.
03:32:49 <oklofog> well those all sound interesting enough.
03:33:11 <kerlo> The bulk of my knowledge about economics comes from a one-semester economics course I took two years ago. A lot of my knowledge about physics comes from a one-year physics course I took six years ago.
03:33:27 <kerlo> Advanced Placement; the idea is getting college credit for classes taken in high school.
03:33:33 <oklofog> i don't have any knowledge about either :<
03:33:50 <kerlo> Anyway, I'm hoping that the economics tests will be pretty intuitive.
03:34:23 <oklofog> i got "college credit" for the courses i took at the actual university during high school
03:34:45 <kerlo> I think the amount of stuff there is to know about physics is relatively small.
03:34:48 <kerlo> Yeah, not the same.
03:35:14 <oklofog> and from the cisco ccna stuff i did, which was 100% surface learning, i don't remember any of it
03:35:48 <kerlo> Chemistry is something there is a lot to know about, but I got an 800 on the SAT II for it, and I took Honors Chemistry last year.
03:36:00 <kerlo> Let's see, my other AP tests are...
03:36:04 <oklofog> i didn't actually know how to study at high school, i did well enough by just using my brain in the exams
03:36:38 <oklofog> there's not a lot to know about physics?
03:36:48 <oklofog> where's Slereah when you need him
03:36:51 <kerlo> English, Biology, and Calculus BC, I'm taking the classes for, so there's little question that I'll get 5s on those.
03:37:09 <oklofog> oh same grade system, fail-5?
03:37:30 <kerlo> These differential equations govern gravity. These differential equations govern electricity and magnetism. That's it.
03:38:36 <kerlo> 1 is "no recommendation", 2 is "possibly qualified", 3 is "qualified", 4 is "well qualified", and 5 is "extremely well qualified".
03:38:59 <kerlo> These refer specifically to "college credit or advanced placement".
03:40:25 <kerlo> And the other ones I'm taking are Computer Science AB and Physics C: Mechanics.
03:40:32 <oklofog> speaking of studies, i have four exams next week, so i should probably start considering leaving soon
03:41:14 <kerlo> Each test takes four hours; I think I'll finish each day by 5 PM at the latest. This leaves a nice four hours to study for the next test, eh?
03:41:38 <oklofog> how many you have in one week?
03:41:52 <kerlo> They're spread out relatively evenly over two weeks.
03:42:10 <oklofog> i usually do exams in about 3 hours, therefore i have about 14 hours a day for reading
03:42:28 <kerlo> What, four hours of sleep?
03:42:29 <oklofog> i have 6 exams in the next 2 weeks
03:43:02 <oklofog> but yes, that's how much i sleep when i have exams, or less
03:43:33 <oklofog> "OMG WHAT IF I FORGET THE DETAILS OF THIS HEADER"
03:44:16 <kerlo> So, it looks like I have a test on Tuesday, May 5; one on Wednesday, May 6; three on Monday, May 11; one on Tuesday, May 12; one on Wednesday, May 13; and two on Thursday, May 14.
03:44:45 <kerlo> But the three on May 11 are really more like one and two halves, since the Physics C ones are half as long as the others.
03:45:23 <coppro> also, /me should scroll up
03:45:28 <oklofog> that's a lot of exams; then again if it's about aptitude i guess it doesn't matter that much.
03:45:38 <kerlo> Also, I have a load of Spanish homework due on May something.
03:45:56 <kerlo> My Spanish teacher said that I need to be doing about five assignments a day to catch up.
03:46:08 <oklofog> i like to surface learn all the details before the exam, which isn't the point for aptitude stuff (and not for exams either, it's just how i roll).
03:46:13 <coppro> Due to the fact that I'm not in AP and the fact that AP exams won't get me anything at my target university, I'm not writing any AP exams.
03:46:54 <kerlo> What is "aptitude"?
03:47:19 <oklofog> the thing i thought ap was about, and still think it is
03:47:41 <oklofog> and not about learning some specific new thing
03:48:04 <kerlo> So practical ability, neither knowledge nor intelligence exclusively but a combination.
03:48:48 <oklofog> something like that, i just meant it's probably the kind of exam that's more about how well you know the general subject than it is about being you know studied it lots.
03:49:03 <oklofog> like you know understanding. forget it :P
03:50:18 <kerlo> That's why I don't plan on studying lots. :-P
03:50:22 <oklofog> i just assume it's not the kind of thing you revise that much before the exam, but more the kind of thing that tries to measure what you learned in high school or whatever.
03:51:35 <oklofog> hmm, i'll probably leave in 10 minutes
03:52:06 <kerlo> I can study for one on Monday, one on Tuesday, one on Wednesday, one on Thursday, one on Friday, one on Sunday, one on Monday, one on Tueday, one on Wednesday.
03:52:39 <oklofog> next time i don't have to study is probably 19.5.
03:52:49 <kerlo> Or I could be responsible and begin studying the day after tomorrow.
03:53:05 <oklofog> excluding these random 3 hour irc breaks
03:53:08 <kerlo> Yeah, right. I have things due Monday.
03:53:26 <kerlo> And I have an email to finish. Be back right before you leave or something.
03:54:04 <oklofog> alrighty, i'll try to consume a few more pages of this trivial networking crap
03:54:08 <coppro> but what I /do/ need to do is my calculus work
04:00:18 * kerlo finishes the email.
04:00:46 * kerlo wins, non-exclusively.
04:01:00 <oklofog> except i'll just leave irc and read the rest of this thing
04:01:16 <oklofog> so yes, we see like the swedes say.
04:01:21 -!- oklofog has quit ("PJIRC @ http://webirk.dy.fi").
04:06:06 <kerlo> And I guess I have to go now as well.
04:06:33 <kerlo> I have two options: be tired, or get addicted to caffeine.
04:06:58 <kerlo> The obvious explanation for my being tired is that I slept too much last night. :-P
04:07:56 <kerlo> So, plan: read no more than 50 pages of Flowers for Algernon, then work on Spanish. I'm currently on page 185.
04:10:59 <Gracenotes> I think I'll include a lolcat in my cryptography presentation
04:11:12 <Gracenotes> namely this specimen, http://geekfriendly.org/blog/wp-content/uploads/2007/06/schrodinger_s-lolcat.jpg
04:23:07 <GregorR> We were talking about PSPACE-complete games in Theory today, and one that he mentioned I think has the potential to be used as the basis for a real game. Given a graph, a starting vertex, and two players, each player selects a vertex pointed to by the current vertex, and crosses the current vertex out. You're not allowed to move to a vertex which is crossed out, and if you can't move, you lose.
04:45:38 -!- pikhq has quit (Read error: 110 (Connection timed out)).
04:57:41 -!- rodgort has quit ("Coyote finally caught me").
04:58:00 -!- rodgort has joined.
04:58:10 -!- lifthrasiir has quit ("Changing server").
04:58:45 -!- lifthrasiir has joined.
05:00:39 -!- lifthrasiir has left (?).
05:02:01 <kerlo> That seems really simple.
05:02:31 <kerlo> It makes me wonder if there's a way to estimate the maximum nimber of a group given a random sample.
05:03:20 <kerlo> The obvious answer is no, because it's PSPACE-complete.
05:04:54 -!- lifthrasiir has joined.
05:09:50 -!- psygnisfive has joined.
06:23:48 <Gracenotes> I can hear the sound of a thousand geeks scrambling to make such a script and be Internet Famous
06:24:32 <bsmntbombdood> there's got to be some ebay-scraping libraries already
06:27:08 <psygnisfive> it'd be hilarious if someone set up a beartrap in the pentagon
06:29:39 <Gracenotes> yeah, and they'd probably get a thousand life sentences, if not death by disembowelment and a thousand advertisers trying to use the recent "terrorist attack" to sell their security
06:32:38 <Gracenotes> my clock is telling me it's too late for aimless rants
06:33:47 <Gracenotes> oh no, your clever wordplay has rendered every bit of logical reasoning I've picked up over the years utterly useless! :X
07:12:27 -!- Slereah has joined.
07:13:02 <Gracenotes> psygnisfive: you clearly don't know how to speak Engrish
07:14:13 <Gracenotes> yeas, "set us up" sounds more natural and more menacing
07:14:49 <Gracenotes> but your base are not have property natural
07:26:16 -!- WangZeDong has quit (Read error: 110 (Connection timed out)).
07:46:25 -!- comex has quit (Read error: 104 (Connection reset by peer)).
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:11:46 -!- Ilari has quit ("Shutdown...").
08:42:43 -!- kerlo has quit (hubbard.freenode.net irc.freenode.net).
08:43:25 -!- kerlo has joined.
09:03:26 -!- oerjan has joined.
09:15:05 -!- Ilari has joined.
09:24:01 <AnMaster> <ehird> AnMaster: The arch installer is not graphical? <-- Yes. Ncurses has line art... :P
09:24:30 <AnMaster> I mean, you won't ever need to use it again after.
09:24:36 <AnMaster> since Arch is pure rolling release.
09:25:40 -!- Sgeo has quit (Read error: 110 (Connection timed out)).
09:39:46 -!- lifthrasiir has quit (Remote closed the connection).
09:41:12 -!- lifthrasiir has joined.
09:54:46 -!- BeholdMyGlory has joined.
10:11:42 -!- oerjan has quit ("leaving").
10:11:58 -!- M0ny has joined.
10:29:47 -!- KingOfKarlsruhe has joined.
10:42:54 -!- MizardX has joined.
10:57:16 -!- tombom has joined.
11:26:53 -!- BeholdMyGlory has quit (Remote closed the connection).
11:40:37 -!- BeholdMyGlory has joined.
11:42:06 -!- FireFly has joined.
13:01:20 -!- Judofyr has quit (Remote closed the connection).
13:43:28 -!- KingOfKarlsruhe has quit (Remote closed the connection).
14:33:46 -!- kar8nga has joined.
14:48:00 -!- iano has joined.
15:04:10 -!- iano has quit.
15:08:01 -!- M0ny has quit ("PEW PEW").
15:10:11 -!- M0ny has joined.
15:11:13 -!- Hiato has joined.
15:35:00 -!- olsner has joined.
15:46:08 -!- Gracenotes has quit ("Leaving").
15:50:19 -!- coppro has quit (Read error: 110 (Connection timed out)).
16:06:12 -!- Hiato1 has joined.
16:06:24 -!- Hiato has quit (Read error: 104 (Connection reset by peer)).
16:33:31 -!- kar8nga has quit (Read error: 60 (Operation timed out)).
16:35:09 -!- kar8nga has joined.
16:37:35 -!- ais523 has joined.
16:48:19 -!- louzer has joined.
16:50:03 <louzer> Is there a mechanical way to make a quine of a combinator?
16:50:23 <louzer> something like a fixed point combinator?
16:52:42 <Slereah> What do you call a quine from a combinator, exactly?
16:54:50 <louzer> A quine from an expression made of combinators is something that evaluates producing the same expression that was evaluated.
16:55:48 -!- Gracenotes has joined.
16:56:19 <Slereah> That would be... a non-normal form?
16:56:26 <ais523> louzer: ```sii``sii is the most famous combinator quine, in that sense
16:56:33 <ais523> although any such quine is necessarily an infinite loop
16:57:03 <Slereah> And I'm not sure you can actually compute to check for it.
16:57:54 <Slereah> Unless you can, since it's not actually every non-normal forms
16:58:02 <ais523> Slereah: that's an interesting question. "Does a combinator expression halt" is obviously unsolvable. "Does a combinator expression ever return to its starting point when evaluated" is I suspect unsolvable, but it's not obvious from the halting problem that it is
17:01:56 <AnMaster> it has some new dependencies. Which has more dependencies.
17:02:38 <AnMaster> two deps left ot compile before it's time to compile gcc itself.
17:03:13 <ais523> I love the Ubuntu/Debian "install all build dependencies of this package" command
17:03:33 <ais523> lets you build what parts of the system you like by hand, without chasing dependencies
17:05:00 <AnMaster> ais523, they are not in stable arch linux at least yet.
17:05:02 <louzer> ais523, Perhaps there are ways for a combinator expression to loop forever without ever reaching its starting point
17:05:13 <ais523> louzer: there definitely are
17:05:21 <AnMaster> ais523, plus I had to recompile one dep (gmp) to enable some more options for it
17:05:45 <ais523> ```s``skki``sii becomes ```sii``sii after once round the loop, IIRC
17:05:56 <ais523> and from then on is ```sii``sii forever
17:06:00 <louzer> ais523, that was quick
17:06:18 <ais523> heh, ```s.a.b``s.c.d is the usual expression I use to test Unlambda interps
17:06:22 <AnMaster> ais523, what about one not ever reaching the same state again. As in, every state is unique
17:06:25 <ais523> so I'm rather familiar with how it loops
17:06:34 <ais523> AnMaster: I think that's possible too, but harde
17:07:10 <ais523> imagine something like (:*:^):^, translated to Unlambda
17:08:38 <Slereah> Around stacks, never relax.
17:10:11 <louzer> what if use Zobrist hashing to test whether some evaluation step during evaluation gets repeated? We will be able to do impossibility-space tradeoff attack on halting problem lol.. (People use Zobrist hashing to find whether chess board states get repeat when searching down the game tree.)
17:11:33 <louzer> wait ordinary hashing would do
17:12:07 <louzer> because expressions can be evaluated unambiguosly
17:12:51 <louzer> zobrist would help to check whether parts of the expression are getting repeated
17:13:39 <AnMaster> !befunge98 http://www.quote-egnufeb-quote-greaterthan-colon-hash-comma-underscore-at.info/befunge/ytest.bf
17:14:03 -!- MizardX has quit (Read error: 104 (Connection reset by peer)).
17:14:14 <AnMaster> ais523, did you miss that yesterday about egobot?
17:14:20 -!- MizardX has joined.
17:14:53 <louzer> I bet Turing would say there exists an expression that does not have any detectable pattern that signifies non-haltability
17:16:06 <AnMaster> oh and another thing... I made a patch to add befunge-98 to it. Using cfunge. Which made me add some new compile time options for cfunge. To enable/disable some things.
17:16:32 <AnMaster> (TURT is disabled in cfunge in EgoBot, so is NCRS and TERM. And it runs with -S so no unsafe IO)
17:18:40 <louzer> Guys, so there cannot be a mechanical way to generate a quine of a combinator expression because that would mean there exists halting predictor which is impossible?
17:19:36 <ais523> oh, you can certainly generate some infinite combinator loops
17:20:52 <ais523> ```s``ska``skb``s``skf``skg is an infinite loop no matter what a, b, f, g
17:21:07 <ais523> (Unlambda experts will know why I skipped c, d, and e)
17:21:21 <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 unlambda whirl
17:21:31 <EgoBot> EgoBot is a bot for running programs in esoteric programming languages. If you'd like to add support for your language to EgoBot, check out the source via mercurial at https://codu.org/projects/egobot/hg/
17:21:45 <ais523> it's nice to see the exclamation mark prefix come back
17:22:26 <AnMaster> there are instructions for how to add new languages.
17:23:22 <ais523> but that would been I'd need an Underload interp written in something sane
17:23:53 <louzer> Life would have been so much easier if I had an Oracle machine.. fuck reality! it doesn't want me to have one.
17:25:22 -!- thutubot has joined.
17:25:22 -!- thutubot has quit (Remote closed the connection).
17:25:25 <AnMaster> ais523, C, java, haskell and several other ones are supported.
17:25:25 -!- MizardX has quit (Read error: 60 (Operation timed out)).
17:25:32 <AnMaster> it executes the interpreter binaries.
17:25:39 -!- thutubot has joined.
17:26:18 <ais523> now 2 out of 5 bots here can do Underload
17:26:22 <ais523> rather than 1 out of 4
17:26:25 <ais523> Slereah: that wasn't deliberate!
17:26:38 <ais523> if it were, I'd have capitalised it properly
17:27:11 <ais523> !unlambda ```s.a.b``s.c.d
17:27:29 <ais523> I probably shouldn't have fed it an infinite loop
17:27:41 <ais523> Slereah: that produces no output, obviously
17:28:06 <ais523> I can only conclude I accidentally crashed it with that loop
17:28:14 <GregorR> Did you guys kill EgoBot >_<
17:28:21 <ais523> GregorR: I think quite possibly
17:28:27 <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 unlambda whirl
17:28:37 <ais523> ok, it's listening to you
17:28:41 <EgoBot> EgoBot is a bot for running programs in esoteric programming languages. If you'd like to add support for your language to EgoBot, check out the source via mercurial at https://codu.org/projects/egobot/hg/
17:29:03 <GregorR> It's possible, and even likely, that unlambda doesn't work.
17:29:17 <ais523> I know that bf works, though
17:29:20 <ais523> and I'm not getting output from it
17:29:30 <ais523> !bf ++++++++++++++++++++++++++++++++++++++++++++++++++++++++.
17:29:37 <ais523> it's just not accepting input via !
17:30:16 <GregorR> What should ````kiss.x output?
17:30:17 <ais523> we certainly need to add Underload, but the problem is that the only two known sandboxed Underload interps are written in esolangs
17:30:38 <ais523> and `s.x has no output
17:30:43 <ais523> because it's an incomplete expression
17:31:08 <ais523> how does EgoBot handle infinite loops?
17:31:12 -!- MizardX has joined.
17:31:14 <Slereah> !unlambda ````````kiss.m.y.a.s.s
17:31:41 <GregorR> It nices down, kills long-running scripts, and refuses to allow anything to take more than 25% CPU
17:32:13 <GregorR> Does underload have the ability to run I/O or whatnot?
17:32:19 <ais523> underload does O but not I
17:32:50 <AnMaster> one of the deps fail configuring
17:32:52 <ais523> +ul (()(*))(~:^:S*a~^a~!~*~:(/)S^):^
17:32:54 <thutubot> */*/**/***/*****/********/*************/*********************/**********************************/******************************************************* ...too much output!
17:32:58 <ais523> ^ul (()(*))(~:^:S*a~^a~!~*~:(/)S^):^
17:32:58 <fungot> */*/**/***/*****/********/*************/*********************/**********************************/*******************************************************/*****************************************************************************************/********************************************************************************* ...too much output!
17:33:00 <AnMaster> it claims a header that does exist doesn't...
17:33:00 <GregorR> So there's no complelling reason not to use a normal underload interp?
17:33:11 <ais523> the only compelling reason is that there are no normal underload interps
17:33:28 <GregorR> Well, one that's in an esolang is fine, really.
17:33:33 <ais523> it's simple enough to implement that everyone just wrote esolang versions
17:33:36 <GregorR> So long as it's an esolang that EgoBot supports.
17:33:36 <ais523> Slereah: no, fibonacci
17:33:42 <ais523> primes don't go up exponentially
17:34:07 <ais523> really, the best thing would just be to write an interp in something sane
17:34:21 <ais523> there's an efficient C version, but it's a compiler not an interp
17:35:01 <ais523> Underload interps have been written in: JS, Thutu, Befunge, Redcode, BF
17:35:18 <ais523> the other problem is it's very easy to use up memory really quickly in an Underload program
17:35:19 <lifthrasiir> AnMaster: i found some example that rcfunge is faster than cfunge... :p
17:35:22 <ais523> do you have a limit on memory used?
17:35:48 <AnMaster> lifthrasiir, the actual program?
17:36:02 <ais523> also, the only way to output a newline in Underload's to put a literal newline in the program
17:36:06 <ais523> but that could just be done via URL
17:36:06 <lifthrasiir> it puts p at <9^6,0> so it tests pure performance of the fungespace get
17:36:10 <AnMaster> lifthrasiir, where does it put it
17:36:29 <AnMaster> lifthrasiir, is those value average over several runs btw?
17:36:32 <GregorR> ais523: That sort of issue I just dont care about ;)
17:36:48 <ais523> <lifthrasiir's link> 'q99*:*:*0p3
17:37:22 <ais523> I don't see how the ' before the q is ever overwritten
17:37:35 <AnMaster> lifthrasiir, it runs over lots of empty funge-space indeed,
17:37:49 <lifthrasiir> ais523: no, it puts q at <9^8,0>, so after ...p3 there is a generated q and exits
17:38:09 <AnMaster> lifthrasiir, what about ccbi and pyfunge on it
17:38:41 <ais523> I suppose an interp really optimised for that program would skip all the spaces in one go, and handle that in O(1) not O(n) time
17:38:43 <lifthrasiir> AnMaster: i don't have a working ccbi build yet, due to out of time, and pyfunge... is too slow.
17:38:55 <AnMaster> ais523, you need to find all those spaces.
17:39:07 <ais523> you could know there was nothing in the columns inbetween
17:39:10 <AnMaster> which means lots of "check if value exist in hash table" for cfunge.
17:39:39 * ais523 tests Language::Befunge
17:39:45 <lifthrasiir> ~/hg/pyfunge$ echo '88*99*::**0p' | time ./pyfunge -v98 -
17:39:50 <AnMaster> ais523, yes I could. But then I would need to check the hash table for column count
17:39:58 <AnMaster> which would for most programs be a waste of time.
17:40:22 <ais523> that's why I said "optimized for that program"
17:40:52 <lifthrasiir> but still i cannot think why rcfunge is faster than cfunge then. maybe because of mac strangeness? :S
17:40:57 <AnMaster> lifthrasiir, feel free to tune the size of the static array to include that. #defines are near line 100 of src/funge-space/funge-space.c. Note the comment about
17:41:13 <AnMaster> #define FUNGESPACE_STATIC_X 512
17:41:13 <AnMaster> #define FUNGESPACE_STATIC_Y 1024
17:42:05 -!- puzzlet has joined.
17:42:27 <AnMaster> anyway: each must be a power of two. And (FUNGESPACE_STATIC_X * FUNGESPACE_STATIC_Y * sizeof(funge_cell)) % 128 == 0
17:42:57 <AnMaster> the comment *was* correct after all
17:43:30 <ais523> jqbef98 is much slower than either cfunge or RC/funge
17:43:37 <ais523> not at all surprising, it's one of the slowest on Mycology
17:43:48 <AnMaster> lifthrasiir, I'll try profile it later today. Doing some other stuff atm. And food is soon ready.
17:44:11 * ais523 kills the process after 90 seconds
17:45:16 <AnMaster> lifthrasiir, not that I know rcfunge code very well.
17:45:41 <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 unlambda whirl
17:45:58 <AnMaster> ais523, were you around when Deewiant tested on 32 MB large text file from project gutenberg?
17:46:03 <lifthrasiir> FYI, i have #defined FUNGESPACE_STATIC_X to 0x4000000 and FUNGESPACE_STATIC_Y to 1, and cfunge became very slow (i.e. 60s and still running)
17:46:06 <AnMaster> With a @ to just check parsing.
17:46:18 <lifthrasiir> for smaller FUNGESPACE_STATIC_X there is no difference.
17:46:23 <ais523> I like that idea, give a @ followed by 32 MB of junk?
17:46:47 <AnMaster> ais523, ah. cfunge was fastest. 27 seconds or such and 2.2 GB RAM. CCBI was way more. And more than half an hour (forgot how long exactly)
17:46:55 <AnMaster> don't remember what other ones he tested.
17:47:06 <ais523> also, my guess for the 0x4000000 define is that cfunge caused the computer to swap by trying to access that much memory
17:47:38 <ais523> that's a 1 followed by 26 zeroes, isn't it?
17:47:54 <ais523> wait, *4 or 8 due to width of an int
17:48:07 <ais523> 256 MB isn't an insane amount for some modern computers
17:48:14 <ais523> presumably yours can't handle it, though
17:48:51 <ais523> 256 MB should fit in 2GB without swapping, though
17:53:50 <ais523> anyone here have ideas of how to write a super-optimised sane Underload interp?
17:53:54 <ais523> as opposed to a compiler?
17:54:53 <Deewiant> AnMaster: 3.x or 4.x RAM and 4x minutes
17:56:39 <GregorR> ais523: Super-optimiZed is not a requirement. (But American spelling is lawl ;) )
17:56:56 <ais523> yes, especially as I'm British
17:57:08 <ais523> I write in a mix of British and American English unless I concentrate
17:57:12 <ais523> due to being on the Internet so much
17:57:28 <ais523> but if I'm going to write an interp in a sane language, I want it to run quickly
17:57:29 <GregorR> AMERICANS HAVE TAKEN OVER THE INTARWEBS
17:58:33 <Deewiant> AnMaster: " CCBI was way more. And more than half an hour (forgot how long exactly)"
17:58:36 <AnMaster> <ais523> 256 MB should fit in 2GB without swapping, though
17:59:08 <ais523> but cache misses shouldn't slow cfunge down that dramatically
17:59:29 <Deewiant> Life was simpler when all memory was the same speed
17:59:47 <ais523> even back in the 6502, memory access to the bottom 256 bytes of memory was faster
18:00:03 <ais523> although only 16 bytes of it was available to user programs
18:00:52 <ais523> who's Lawrence Woodman, I wonder? He's been writing about Redcode and SUBLEQ
18:03:14 <AnMaster> ais523, 256 MB is quite a bit to fill with spaces initially though
18:03:33 <ais523> although again it shouldn't take that long
18:03:34 <AnMaster> even though I use streaming non-temporal stores.
18:03:46 <ais523> memtest86 can fill my entire memory with anything in less than a second
18:03:50 <ais523> and I have more than 256 MB
18:04:21 <AnMaster> actually, not the entire memory
18:04:30 <AnMaster> it needs to reserve some for itself
18:04:51 <ais523> yes, the entire memory
18:04:59 <AnMaster> ais523, I mean. For it's own code.
18:05:00 <ais523> it fills most of it, then moves itself to a different location, then tests the bit where it was
18:05:11 <AnMaster> ais523, what about BIOS reserved bits.
18:05:47 <ais523> I'm not sure how it handles those
18:06:01 <ais523> I imagine it doesn't test memory-mapped registers with random data, though!
18:07:00 <AnMaster> lifthrasiir, did that one you set ever finish?
18:07:27 <AnMaster> if not I suspect it just doesn't handle 1 for the dimension very well.
18:07:43 <lifthrasiir> AnMaster: i terminated it before it finishes.
18:18:28 <AnMaster> what does "ulp" mean in the context of floating point
18:19:06 <AnMaster> ok. And what does that actually _mean_
18:19:23 <Deewiant> ulp(x) is the difference between the two floating-point numbers closest to x
18:19:52 <lifthrasiir> a gap between it and closest representable number greater than it
18:24:33 <AnMaster> ais523, does one need to do the bootstrap thing when building a new gcc version using the previous gcc version.
18:24:35 <ais523> possibly in this channel, but only an older version, and I bet there are people who know more about gcc than me elsewhere
18:24:53 <ais523> and you never "need" to do the bootstrap thing, not even for building with a different compiler
18:25:02 <AnMaster> ais523, why does it default to being used then
18:25:15 <ais523> because it gives a more efficient binary in most cases, and because you can't test unless you bootstrap
18:25:20 <GregorR> To verify trhat the generated compiler is correct.
18:27:07 <GregorR> And because the compiler doesn't taste like chicken until it's compiled itself.
18:27:45 <lifthrasiir> well, anyone can connect to eso-std.org domains? my DNS didn't resolve them at all.
18:28:03 <ais523> err, eso-std.org no longer exists
18:28:07 <ais523> ehird dropped the domain, and it's been parked
18:28:22 <ais523> if there's something for particular you're looking for that was there, I might know where it is now
18:29:34 <AnMaster> odd, why does trying to open a new ssh connection only work when the target computer isn't loaded.
18:29:43 <AnMaster> existing ones work fine in all cases.
18:30:59 <AnMaster> and even if I ctrl-z whatever is putting that load on the system... It doesn't help.
18:31:22 <AnMaster> as in, any existing ssh that is currently connecting doesn't finish connecting.
18:33:32 <AnMaster> especially configure running prevents it working. Compiles do it sometimes.
18:35:56 <ais523> heh, Microsoft's servers crashed when they tried to release Windows 7 RC to the MSDN and TechNet subscribers
18:36:27 <Deewiant> I was under the impression they just got slashdotted
18:36:45 <AnMaster> someone should slashdot slashdot
18:36:59 <GregorR> Why was eso-std.org taken down? (<-- ehird?)
18:37:08 <ais523> Deewiant: I consider being slashdotted as a form of crashing
18:37:29 <ais523> GregorR: I'm not entirely sure, I'm no good at following ehird's logic with server maintenance
18:37:39 <Deewiant> Slashdottedness doesn't require manual intervention to resolve
18:37:42 -!- oerjan has joined.
18:37:52 <ais523> well, just slashdotted then
18:38:19 <GregorR> ais523: I don't suppose there was anything hugely relevant there anyway? :)
18:38:28 <ehird> GregorR: It was dormant.
18:38:42 <ehird> Will not be soon due to moving to prgmr
18:38:54 <ehird> 03:23 GregorR: We were talking about PSPACE-complete games in Theory today, and one that he mentioned I think has the potential to be used as the basis for a real game. Given a graph, a starting vertex, and two players, each player selects a vertex pointed to by the current vertex, and crosses the current vertex out. You're not allowed to move to a vertex which is crossed out, and if you can't move, you lose.
18:39:09 <ehird> AnMaster: well, I can try Arch.
18:39:11 <ehird> downloading Slackware atm
18:39:41 <ehird> Started to, then had to go sleep.
18:39:45 <ais523> ehird: that game GregorR mentioned, if you did it with a directed graph you could use it to model any finite game, with an appropriate graph
18:39:54 <ehird> ais523: Even hawtter.
18:40:12 <oerjan> well nxn Go is supposedly PSPACE-complete iirc
18:40:13 <ehird> mayhaps I will try a BSD afater slack
18:40:39 <ais523> ehird: are you messing with lots of different Linux distros?
18:41:14 <AnMaster> ais523, could you model chess with it
18:41:28 <ehird> AnMaster: arch using grub on a cd freaked me out
18:41:34 <ais523> AnMaster: yes, in theory, but the graph would be insanely massive
18:41:45 <ehird> it worked just fine
18:41:51 <GregorR> ais523: How massive? If exponential, that's not a valid reduction :P
18:41:54 <AnMaster> ais523, how would you do it. I mean how do you represent a game in it.
18:42:02 <ais523> GregorR: yes, exponential
18:42:11 <GregorR> Right, so that doesn't prove anything about Chess.
18:42:13 <ais523> AnMaster: with one node for every position, and directed arcs that don't let you go back
18:42:17 <ais523> GregorR: no, it wasn't meant to
18:42:21 <oerjan> AnMaster: chess has some timeout rules that force it to be finite
18:42:35 <GregorR> So in short, every game with a finite number of states can be represented as a graph of states.
18:42:50 <oerjan> (not just no. states but also length)
18:43:01 <AnMaster> ais523, you could only represent turn-based games with moves on a graph or board I think. And only deterministic ones.
18:43:22 <AnMaster> for example, you couldn't represent any game with dice right?
18:43:28 <ehird> AnMaster: anyway, I don't mind a textual installer, it just tends to imply unpolishedness in other areas
18:43:36 <ais523> you're right, it would have to be deterministic
18:43:49 <ehird> ais523: just add a token
18:43:54 <ais523> but you don't need moves on a graph/board, anything with a finite number of defined states will do
18:43:55 <ehird> meaning a number from 0-N inclusive
18:43:58 <AnMaster> ais523, and turn based. Not deterministic and continuous.
18:44:04 <GregorR> Is there a theory of nondeterministic graphs? :P (Quantum graphs?)
18:44:09 <ais523> AnMaster: continuous implies infinite number of states
18:44:17 <AnMaster> ais523, well maybe wrong word.
18:44:20 <ehird> AnMaster: I assume I want an ftp/http installation source?
18:44:27 <ehird> I'm using the ftp disk
18:45:10 <AnMaster> ehird, now it was almost a year ago I installed. So I don't remember all details
18:45:14 -!- tombom has quit ("Peace and Protection 4.22.2").
18:45:22 <oerjan> GregorR: strictly speaking a reduction from something bounded finite is not really exponential no matter how much it blows up
18:45:39 <oerjan> (you just stick it in the constant multipliers)
18:45:41 <ehird> AnMaster: oh god, I have to use the awful partitioning program
18:45:47 <AnMaster> ais523, make a graph to represent D&D first edition.
18:45:55 <AnMaster> I thought there was fdisk on it
18:46:10 <ais523> D&D first edition is nondeterministic and has an infinite number of states
18:46:13 <Deewiant> cfdisk crashed on my disk IIRC
18:46:14 <GregorR> oerjan: True. But I was assuming "generalized" Chess with nxn boards *shrugs*
18:47:24 <ehird> Pacman preparation failed.
18:47:31 <ehird> Even slackware's install is better
18:48:44 <GregorR> Slackware has an installer? :P
18:48:52 <oerjan> GregorR: i (very vaguely) recall that loop quantum gravity theory may use something that's almost quantum graphs, although the spin or was it area annotations may be on tuples of lines so it's a bit higher-dimensional
18:49:16 <ehird> GregorR: It has a command which tells you how to run the disk partitioner, and a program that lets you select what packages to install which then copies them to the disk.
18:49:24 <ehird> The package installer even uses ncurses.
18:49:40 <GregorR> They've given up on their heritage! :(
18:49:42 <ehird> Slackware isn't unusable... just minimalist.
18:50:12 <GregorR> LFS is how people who aren't pansies install their distro.
18:50:15 -!- thutubot has quit (Remote closed the connection).
18:50:22 <ehird> GregorR: I am going to kill you now.
18:50:28 <GregorR> LFS is so awesome it scared thutubot away.
18:50:35 <ehird> GregorR: You will then die.
18:50:46 <ehird> "They're just sitting there taking up (precious) disk space." — LFS
18:50:53 <ehird> Precious disk space.
18:51:05 <ehird> I think the last time I thought disk was precious was in the nineties.
18:51:31 <GregorR> That's why "precious" is in parens X-P
18:51:45 <GregorR> You could always install Gentoo (wimpmode LFS)
18:51:59 <ehird> I hate Gentoo with the fiery passion of a thousand suns.
18:52:10 <GregorR> Because it's wimpmode LFS.
18:54:21 <oerjan> <ais523> Slereah: that's an interesting question. "Does a combinator expression halt" is obviously unsolvable. "Does a combinator expression ever return to its starting point when evaluated" is I suspect unsolvable, but it's not obvious from the halting problem that it is
18:55:07 <oerjan> i think that's nearly obvious though, because you can take an integer function with unsolvable halting and turn it into a quine
18:56:08 -!- Leonidas_ has joined.
18:56:57 -!- thutubot has joined.
18:57:06 <ehird> GregorR: WOW, you can use cfdisk OR fdisk!
18:57:16 <ehird> Slackware is amazing.
18:57:21 <oerjan> also, if something returns to its starting point that can obviously be proved, just as it can be proved _if_ something halts
18:57:33 -!- Leonidas_ has changed nick to Leonidas.
18:58:00 <oerjan> (i guess this is really "obvious if you've thought about such problems before")
18:58:11 <AnMaster> <ehird> Even slackware's install is better <-- I never had any issues.
18:58:26 <AnMaster> but cd version I used was from august last year.
18:58:41 <ehird> Cool, slackware calls ext2 "standard".
18:59:13 <AnMaster> <GregorR> LFS is how people who aren't pansies install their distro. <-- I done it. And HLFS. You learn a lot, but you don't want to use it.
18:59:24 <AnMaster> ehird, why are you not going with ubuntu or debian
18:59:33 <ehird> AnMaster: Boredom. I have cheap VM creation.
18:59:37 <ehird> And a fast internet connection.
18:59:40 <oerjan> ehird: old standards are also standards?
18:59:42 <ehird> Distros are like pokemon,.
18:59:58 <ehird> Emacs is a separate package series in the slackware installer :DDDDDDDDDDD
19:00:11 <ehird> lawl, it comes with tex by default
19:00:22 <ehird> and games! BLOATWARE MORE LIEK
19:00:28 <ehird> I bet it comes with hunt(1)
19:00:36 <ehird> That bastion of funnity
19:00:37 <AnMaster> ehird, no idea about arch in vm
19:00:51 <ehird> "full Install everything (4.8+ GB of software, RECOMMENDED!)"
19:00:51 <AnMaster> I only tried it on a non-virtual computer.
19:01:00 <AnMaster> where it worked (and works) perfectly.
19:01:17 <AnMaster> <ehird> I bet it comes with hunt(1)
19:01:25 <ehird> AnMaster: a bsdgame
19:01:37 <ehird> goes over the network!
19:01:44 <ehird> although you'd best use localhost tbh
19:01:56 <ehird> you walk around nethack-style uncovering tunnels
19:01:57 <ehird> and shoot each other
19:02:22 <AnMaster> ehird, ASCII art with look down like nethack or adventure game style
19:02:44 <AnMaster> there are bsdgames that are text adventure
19:02:46 <ehird> nobody did adventure game style graphics then
19:02:58 <ehird> 19:01 ehird: you walk around nethack-style uncovering tunnels
19:03:02 <ehird> that implies graphical
19:03:07 <AnMaster> ehird, adventure style == adventure(1)
19:03:19 <ehird> AnMaster: fail. it's ADVENT(1)
19:03:37 <AnMaster> ADVENTURE(6) BSD Games Manual ADVENTURE(6)
19:06:09 <ehird> why does lfs need a guide?
19:06:13 <ehird> it seems pretty simple
19:06:37 <AnMaster> you need to create a temp system in /tools first
19:06:48 <ehird> AnMaster: err, I meant installing from an existing linux
19:06:55 <AnMaster> ehird, yes even then that is recommended.
19:08:30 <ehird> "Debra Lynn Murdock and Ian Ashley Murdock"
19:08:37 <ehird> — List of divorces for aug 10-16 07
19:09:03 <GregorR> Not that either of them worked on Debian by then anyway :P
19:11:46 <ehird> Maybe I should make Hirdux ;-)
19:13:50 <ehird> AnMaster: Ian Murdock + Debra Murdock = Debian.
19:13:55 <oerjan> !bf ++++++++[>++++++++<]>[.]
19:14:05 <oerjan> !bf ++++++++[>++++++++<]>.
19:14:06 <ehird> AnMaster: Ian Murdock founded it...
19:14:08 <ehird> Debra was his wife.
19:14:12 <ais523> ehird: you have a girlfriend called Ux?
19:14:36 <oerjan> !bf ++++++++[>++++++++<-]>[.]
19:14:45 <fungot> @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ ...
19:14:51 <oerjan> !bf ++++++++[>++++++++<-]>.
19:14:54 <ehird> Hirdux would deviate from all standards whenever possible
19:15:05 <AnMaster> oerjan, it runs for 30 seconds before ending program
19:15:10 <ais523> ehird: ReactOS already exists...
19:15:20 <ehird> but I mean, while still being linux
19:15:22 <ehird> I'd probably have no /usr
19:15:24 <oerjan> i see EgoBot doesn't handle infinite output
19:15:32 <ehird> Well, I'd have /usr, but it'd be /home.
19:15:39 <AnMaster> oerjan, process is killed after 30 seconds.
19:15:53 <AnMaster> I guess it is line buffer output
19:16:01 -!- Mony has joined.
19:16:05 <ehird> I'd also get rid of /boot
19:16:13 <ehird> After all, it's not anything to do with booting, it's just the kernel!
19:16:39 -!- Mony has changed nick to Guest38386.
19:16:51 <oerjan> AnMaster: i am not saying it's surprising
19:17:10 <ehird> I wonder if the kernel is executable?
19:17:12 <oerjan> but it's much more fun when you can do infinite lists and stuff
19:17:13 <ehird> ais523: do you know?
19:17:41 <Deewiant> AnMaster: Consider adding an @
19:17:44 <ais523> ehird: I don't, but I doubt it is
19:17:50 <ais523> you'd have to wrap it in a container
19:17:56 <ais523> and even then, it wouldn't work running in user space
19:18:03 <ehird> ais523: but if you +x'd it, and put it in a container, and was in kernelspace, would it be executable?
19:18:06 <AnMaster> ehird, how far do your funge get in mycology now btw
19:18:09 <ehird> Or is it not "directly" executable?
19:18:12 <EgoBot> kkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkk
19:18:28 <ais523> ehird: well, in order to load a kernel you copy the image into memory and give it the instruction pointer
19:18:34 <ais523> that's the same process you need to run an executable
19:18:36 <ehird> AnMaster: older program
19:18:49 <ehird> ais523: right then, my kernel would go in /bin!
19:18:58 <ehird> /bin/linux or something
19:19:14 <AnMaster> because that should output lots of F not lots of k
19:19:14 <ehird> Deewiant: Why? It's a binary.
19:19:34 <ehird> Deewiant: Well phooey to that!
19:19:43 <Deewiant> ehird: Binaries are processes and belong in /proc
19:19:56 <ehird> Maybe I'd unify /bin and /lib
19:20:02 <ehird> Since the distinction is often shaky.
19:20:08 <ehird> Especially for scripts
19:20:11 <Deewiant> /bin is for all non-textual data
19:20:23 <oerjan> 09:53:50 <ais523> anyone here have ideas of how to write a super-optimised sane Underload interp?
19:20:35 <ehird> Deewiant: Probably I'd have /bin be all binaries & libraries.
19:20:36 <ais523> ehird: just have executable shared libraries
19:20:44 <oerjan> as for memory, you could always do some sharing of structures
19:21:03 <oerjan> would be easy in haskell
19:22:00 <oerjan> (just make * a constructor, essentially)
19:22:49 <AnMaster> ehird, better idea: /sys/* /proc/* /user/*, first one is system files. All of them. Config files, program, libraries, kernel, ..., second one is for those pseudo file things needed. You should combine /proc (classical) /sys and /dev in it
19:23:34 <AnMaster> no files may be written outside those dirs (that is, /foo is invalid)
19:23:36 <ais523> AnMaster: why use directories at all, if you're banning subdirs
19:23:47 <ais523> actually, the existence of dirs but not subdirs is nicely ridiculous
19:24:21 <AnMaster> ais523, for upgrades obviously. If user creates a file "X" and then you add X in new release you are into issues
19:24:30 <AnMaster> with /sys/X and /user/X you have no issues.
19:24:36 <ehird> http://pastie.org/465240.txt?key=suxdvgfmuazte7rug8rvw ← The Hirdux Hierarchy
19:24:45 <ehird> Simple, effective!
19:25:14 <ehird> AnMaster: /bin/grub.bin
19:25:20 <ehird> .bin means "raw executable data, yo, no headers"
19:25:51 <AnMaster> ehird, you need to patch grub then.
19:25:59 <ais523> ehird: you should have a SIGUSR3 signal
19:26:01 <ais523> you know, just in case
19:26:10 <AnMaster> ehird, because it expects a certain layout.
19:26:10 <ehird> ais523: that's up to the kernel
19:26:17 <ehird> AnMaster: grub can boot non-linux OSes
19:26:20 <ais523> you can patch the kernel!
19:26:22 <AnMaster> ehird, and linux systems doesn't work without /proc.
19:26:23 <ehird> just do the multiboot stuff
19:26:27 <ehird> and fine, I'll have /proc too
19:26:35 <AnMaster> ehird, oh and /sys too iirc nowdays
19:26:42 <ehird> then I'll use an older kernel
19:26:46 <AnMaster> you might get that to work without udev
19:27:01 -!- Judofyr has joined.
19:27:09 <ehird> i'd also need /mnt
19:27:12 <ehird> but that file has the basic structure
19:27:20 <AnMaster> ehird, what is the thing with /usr/root
19:27:33 -!- Judofyr has quit (Remote closed the connection).
19:27:40 <ehird> AnMaster: Basically, /usr/foo is the same hierarchy as /, except that there's no /usr/foo/usr (duh) and /usr/foo/home is their home directory.
19:27:48 <ehird> So you have /usr/anmaster/bin, etc.
19:27:52 <ehird> And /usr/anmaster/home/music
19:28:00 <ehird> ~ = /usr/$USER/home
19:28:18 <AnMaster> why is there a bin outside their home but inside their user
19:28:27 <ehird> ais523: tell him why it makes sense
19:28:37 <ehird> (/usr/foo is not a home directory, it's just that user's personal system files)
19:28:40 <AnMaster> <ais523> actually, the existence of dirs but not subdirs is nicely ridiculous <-- IIRC Mac OS 1 (!) had something like that
19:28:44 -!- Sgeo has joined.
19:28:50 <ehird> AnMaster: what would be /home/anmaster/local/bin is now /usr/anmaster/bin
19:28:58 <ais523> AnMaster: DOS 1 didn't have dirs at all
19:29:02 <ehird> AnMaster: what would be /home/anmaster/music is not /usr/anmaster/home/music
19:29:08 <ehird> AnMaster: see? it separates the two distinct concepts.
19:29:17 <AnMaster> ehird, hm. So what should be in home then
19:29:21 <ais523> ehird: err, Windows Vista works like that
19:29:23 <ehird> AnMaster: /usr/anmaster/home/music
19:29:25 <ais523> so that's hardly original
19:29:29 <ehird> ais523: i never said it was
19:29:34 <ehird> AnMaster doesn't understand it, though.
19:29:44 <ehird> AnMaster: also /usr/anmaster/code/foo
19:29:52 <AnMaster> ehird, that should be home/code?
19:30:03 <ehird> /usr/anmaster is just things like a user's local binaries, logfiles, databases, blah blah blah
19:30:11 <ais523> what if I write a program called .
19:30:11 <ehird> /usr/anmaster/home is their personal scratchspace
19:30:17 <ehird> ais523: it exists.
19:30:18 <ais523> can I use .. as a dotfile to store its data?
19:30:19 <AnMaster> ehird, anyway /lib needs subdirs on Linux. Where do you place kernel modules now
19:30:38 <AnMaster> since you can have more than one kernel version installed
19:30:41 <ais523> AnMaster: clearly, you distinguish kernel modules by the file extension
19:30:45 <ehird> AnMaster: /bin/linux/{kernel.bin,moduleblah.so,...}
19:30:57 <AnMaster> doesn't solve the version problem
19:31:02 * Hiato1 just wants to hear suggestions for an interesting, non-eso programming language from the chatroom. Is that so much to ask!?!?! :P
19:31:12 <AnMaster> kernel expects a structured layout iirc
19:31:12 <ehird> AnMaster: don't care about multiple versions
19:31:16 <ehird> and fuck the kernel :)
19:31:19 <ehird> i'll use bsdddddddddddd
19:31:32 <oerjan> Hiato1: i am obliged to mention haskell, despite not doing much in it... :D
19:32:06 <AnMaster> ehird, you'd still need to patch a lot of libraries to make this work
19:32:08 <oerjan> also, factor was recently mentioned
19:32:21 <ehird> don't caaaaaaare they can suck my diiiiiiiiiiiiiiiiiiiiiiiiiiick
19:32:24 <ehird> they suuuuuuuuuuuuuuuck
19:32:29 <ehird> just tell it that /bin is /lib.
19:32:39 <Hiato1> oerjan: right, though I've looked at it. I was thinking imperative, having just discovered Io, but it seems to have slowed to nothing in terms of dev/activity
19:32:48 -!- M0ny has quit (Connection timed out).
19:32:52 <Hiato1> hrmm... will hcek it out
19:32:52 <AnMaster> ehird, at the very least you need to patch the build system and some files to do that.
19:32:58 <ehird> Hiato1: factor is quite imperative, but stack based
19:33:03 <ehird> haskell is fun though.
19:33:16 <Deewiant> D is traditional-style imperative
19:33:17 <ais523> yay, concatenative langs
19:33:22 <ehird> erlang is not nice >:(
19:33:38 <ais523> also, you should learn Prolog if you never have done before
19:33:38 <AnMaster> Hiato1, concurrent and easily distributed. Functional language.
19:33:48 <AnMaster> ehird, you just hate it because I like it.
19:33:51 <ais523> I think every programmer deserves to know Prolog, even if they never use it or don't like it
19:34:01 <ehird> AnMaster: no, I've disliked erlang for a while
19:34:08 <ehird> its concurrency model isn't great imo.
19:34:10 <AnMaster> ehird, so I should learn haskell just to watch you hate it.
19:34:17 <AnMaster> ehird, better than shared memory at least.
19:34:20 <ehird> AnMaster: are you listening to me?
19:34:21 <Hiato1> hrm.. well.. not exactly writing a complex system, just looking for romething fun to play with
19:34:27 <ehird> i don't give a damn about what languages you use
19:34:32 <ehird> and yes, I do, several
19:34:41 <ais523> Hiato1: definitely I'd recommend Prolog, then, making a complex system in it probably wouldn't work well
19:34:56 <ehird> evidently you're -not- listening
19:34:57 <ais523> but just like Perl's good for text-processing one-liners, Prolog's good for puzzle-solving ten-liners
19:35:09 <AnMaster> I guess stm is short for statement.
19:35:23 <Deewiant> Software Transactional Memory?
19:35:30 <AnMaster> ehird, I googled stm. Gives "STM - STMicroelectronics N.V. (ADR) (NYSE)"
19:35:38 <AnMaster> and "Société de transport de Montréal - [ Translate this page ]"
19:35:38 <Hiato1> ais523: ok, great. But it clumsy and over complex for normal tasks, right? I mean, that would be nice
19:35:51 <ehird> AnMaster: you've told me to JFGI for things that don't give a relevant first result and complained when I complained
19:35:55 <ehird> so there it stands
19:36:04 <ais523> Hiato1: it's a bit wordy when trying to do normal tasks, and the standard library isn't all that good, but it's not terrible
19:36:25 <AnMaster> ehird, you did that first though.
19:36:30 <AnMaster> So was just following your style
19:36:47 <AnMaster> ehird, at least I never linked non-existent wikipedia articles.
19:36:47 <Hiato1> ais523: ok, sounds promising. Know any good resources or should I ask "the g"?
19:37:22 <ais523> Hiato1: it helps to have someone who knows the language to learn it from, IME
19:37:38 <ais523> you can grab a compiler+interpreter+REPL like gprolog without much trouble, though
19:37:52 <Hiato1> ais523: right, so IRC channel's the way to go then
19:38:27 <ais523> you'll want to grab the docs for your interp, just to remember the names of standard library stuff that you use
19:38:39 <ais523> or just write them by hand, almost the whole of Prolog apart from I/O can be written without using the library at all
19:38:45 -!- louzer has quit (Read error: 104 (Connection reset by peer)).
19:38:51 * Hiato1 checks out wiki article
19:38:59 <Hiato1> now that's a nice feature
19:39:09 <AnMaster> lrwxrwxrwx 1 root root 15 Jan 6 11:54 /usr/lib/libavcall.la -> ../libavcall.la
19:39:21 <oerjan> Hiato1: also, Scala or Ocaml for languages that are combined object-oriented and (impure) functional, with advanced static types
19:40:00 <Hiato1> hrmm.. righty'o but I'm currently intreuged with prolog, Io, and factor
19:40:10 <Hiato1> they all look hideously complex OD
19:40:23 <ehird> Io is very simple.
19:40:26 <ehird> It's also, unfortunately, shit.
19:40:38 <oerjan> Hiato1: well then do scheme
19:40:42 <ais523> I learnt Ocaml for a University project, it's an interesting imperative/functional mix; I'd suggest it isn't a good language to learn, but it's a good language to use once you've learnt imperative and functional concepts on a different lang
19:41:12 <oerjan> ehird: i guess R6RS _did_ get hideously complex?
19:41:17 <Gracenotes> ais523: interesting. some people say it's a good link from imperative languages to more functional ones
19:41:36 <ais523> Gracenotes: I hate to look at the programming style of people who tried to learn OCaml like that
19:41:48 <Hiato1> ais523: right. Well I'm learning functional witn Haskell
19:42:03 <Gracenotes> ais523: heh. Programming style in what language? OCaml?
19:42:04 <ais523> I personally program it almost entirely functionally, just using things like imperative globals or exceptions or whatever when they're more useful
19:42:48 <ehird> I should probably make hirdux.
19:43:02 <AnMaster> strings -a /usr/lib/opengl/nvidia/*/*.so | grep -F 'GCC: (GNU)' | sort -n | uniq is interesting
19:43:10 <oerjan> ais523: well, Ocaml was my first static functional language
19:43:16 <AnMaster> they used lots of different versions:
19:43:25 <ais523> oerjan: did you know a dynamic functional language before that, though?
19:43:27 <ais523> if so, it's not so bad
19:43:30 <oerjan> (i guess i mostly knew scheme beforehand)
19:43:31 <ehird> The installer consisting of making a package manager configuration file that points to /mnt/newsystem as the root and running it with a bunch of packages to install, hopefully :P
19:43:45 <AnMaster> GCC 3.2, GCC 3.3.3 (SuSE), GCC 3.4.6 (altstack), GCC 4.1.1 (altstack)
19:43:53 <AnMaster> no idea what those "altstack" are
19:46:17 <oerjan> AnMaster: oh, and IWC :D
19:47:58 <oerjan> if i wanted to spoil it for you, don't you think i would have included some actual _content_?
19:49:43 -!- Ilari has quit (hubbard.freenode.net irc.freenode.net).
19:49:43 -!- Slereah has quit (hubbard.freenode.net irc.freenode.net).
19:49:43 -!- EgoBot has quit (hubbard.freenode.net irc.freenode.net).
19:49:43 -!- ineiros has quit (hubbard.freenode.net irc.freenode.net).
19:49:43 -!- ehird has quit (hubbard.freenode.net irc.freenode.net).
19:49:43 -!- olsner has quit (hubbard.freenode.net irc.freenode.net).
19:49:44 -!- Dewio has quit (hubbard.freenode.net irc.freenode.net).
19:49:51 <AnMaster> oerjan, anyway, I usually read it within an hour after it is updated.
19:50:55 -!- olsner has joined.
19:50:55 -!- Dewio has joined.
19:51:04 -!- ehird has joined.
19:51:53 -!- Ilari has joined.
19:51:53 -!- Slereah has joined.
19:51:53 -!- EgoBot has joined.
19:51:53 -!- ineiros has joined.
19:52:01 <oerjan> except today i got derailed
19:54:55 <ehird> I suppose making Hirdux wouldn't actually be hard if I compromise with the hierarchy :P
19:55:52 * oerjan interpreted that as "Hurdux", and thought that was a suitably insane idea
19:56:10 <AnMaster> ehird, you must base it on Hurd.
19:56:48 <oerjan> where the -ux is from linux, not generic unix
19:57:09 <AnMaster> ehird, or at least do something mad like non-matching kernel and user land (exceptions allowed are tools to load/unload/list kernel modules, and some glue code in libc as needed)
19:57:25 <ehird> BSD userland w/ linux kernel is mundane and done before.
19:57:45 <AnMaster> bsd userland, darwin libc, hurd kernel,
19:57:56 <ehird> Hurd is an unacceptable monostrosity
19:57:58 <AnMaster> libc is of course part of userlang
19:58:23 <oerjan> ehird: what are the acceptable monstrosities?
19:58:25 <Deewiant> Patch the binary if you run into problems
19:58:26 <AnMaster> solaris userland, glibc, Plan9 kernel
19:58:41 <ehird> AnMaster: Fun fact: fun thing+fun thing not always = fun thing
19:59:07 * ehird 's slackware greets him with:
19:59:13 <oerjan> cthulhux, with non-euclidean file system
19:59:13 <ehird> Crucifixes are sexy because there's a naked man on them.
19:59:17 <AnMaster> Deewiant, oh, so it wasn't french then
19:59:19 <ehird> Thanks, Slackware!
19:59:27 <ehird> oerjan: speaking of which
19:59:35 <ais523> ehird: that's quite a fortune quote
19:59:43 <ehird> $ # wow, this new oerjanix is modern
19:59:58 <ehird> $ gnuverse-simulator
20:00:03 <ehird> > ;; Oh god it uses Guile now.
20:00:11 <ehird> > (set! fine-structure "a pound of pennies")
20:00:36 <AnMaster> ehird, that was like months ago
20:00:37 <ehird> > (set! message-style 'old-uppercase)
20:00:40 <ehird> > ;; can't loose that
20:01:39 <ehird> $ cthulhu-scanner # I think we're infected by the sheesh-cthulhu virus
20:02:01 <ehird> AnMaster: you don't care? what bad sysadministration!
20:02:19 <AnMaster> ehird, of course I do. But I would do a clean reinstall in case it includes a rootkit
20:02:34 <ehird> AnMaster: Oh, I have a far more fun way to fix cthulhus.
20:02:36 <ehird> oerjan: scan results?
20:02:45 -!- WangZeDong has joined.
20:02:46 <oerjan> THERE ARE NO GREAT OLD ONES IN THE SYSTEM
20:02:53 <ehird> oerjan: Well, make some then.
20:03:19 <AnMaster> ehird, /usr/share/openssh/Ssh.bin
20:03:39 <AnMaster> no clue what it is. But ssh always installs such a file
20:04:11 <AnMaster> Deewiant, /usr/share/Ssh.bin on arch
20:04:19 <AnMaster> # file /usr/share/openssh/Ssh.bin
20:04:20 <AnMaster> /usr/share/openssh/Ssh.bin: DBase 3 data file (507582464 records)
20:04:31 <ehird> that's a big database!
20:04:43 <ehird> oerjan: Well, I'll just assume that the old ones lied and said they weren't there.
20:05:09 <ehird> > (forever (thread (create-universe :old-ones #f :ehird-has-a-million-pounds #t)))
20:05:11 <oerjan> BUT THERE _ARE_ NO GREAT OLD ONES, THIS IS A _NEW_ SYSTEM
20:05:20 <ehird> > ;; By the omega point hypothesis, we are in one of those universes.
20:05:43 <AnMaster> ehird, check your account balance
20:05:47 <Deewiant> AnMaster: Google suggests that it's for smart cards.
20:05:52 <ehird> AnMaster: I'M RICH!!!!!!!!!!!!!!!
20:06:00 <ehird> oerjan: Well, okay, we're clean now.
20:06:13 -!- Guest38386 has changed nick to Mony.
20:06:16 <ehird> > (set! pi (dialog "What's it to you?" pi))
20:06:17 -!- Mony has changed nick to M0ny.
20:06:28 <ehird> ;; that'll give the mathematicians a fright
20:06:44 <AnMaster> Why not go work on your befunge implementation
20:07:01 <AnMaster> Deewiant, I'm planning for cfunge 0.4.1 tomorrow. btw
20:07:13 <AnMaster> Deewiant, just thought I'd let you know.
20:07:25 <oerjan> POUND OF PENNIES ARE MISSING
20:07:42 <AnMaster> ehird, hey. Give them back to him.
20:07:54 <AnMaster> that is where the million pounds were taken from.
20:08:02 <ehird> > (set! fine-structure "million poundsworth of pennies")
20:08:20 <ehird> what can't science do?
20:08:31 <AnMaster> ehird, ^Z would suspend guile not the program right
20:08:40 <oerjan> RAMPANT PENNY INFLATION
20:08:44 <ehird> AnMaster: DO NOT ARGUE WITH GNUVERSE
20:09:50 <AnMaster> oerjan, that was a really bad pun
20:10:11 <ehird> oerjan: i'm scared that the "pi-god" will kill mathematics
20:10:14 <ehird> please provide an update
20:10:50 <ehird> oerjan: (note that i did > (set! pi (dialog "What's it to you?" pi)))
20:10:53 <AnMaster> oh my, we are set for economic jokes
20:11:18 <AnMaster> I'll go code on cfunge or something...
20:11:39 <oerjan> GIANT BULL FOUND IN SPACE
20:12:04 <oerjan> BULLSHIT FORMS PLANETS
20:13:27 <ehird> AnMaster: no more economics
20:14:27 -!- Slereah has quit (Connection timed out).
20:14:54 -!- pikhq has joined.
20:15:34 <oerjan> UNIVERSE ENTERS RECESSION. BULL IMPLODES DUE TO GRAVITY OF SITUATION.
20:16:17 <fizzie> OpenSSH README.smartcard: "To enable -- load the Java Cardlet to the Cyberflex card -- sectok> jload /usr/libdata/ssh/Ssh.bin"
20:16:19 <AnMaster> Deewiant, actually, it might take a day or two more
20:16:37 <AnMaster> depending on how much work I have to do for gcc 4.4
20:16:47 <AnMaster> in case of new warnings or whatever
20:16:53 <AnMaster> gcc 4.4 is still compiling atm
20:17:26 <oerjan> PLANET EVOLVES SAPIENT BEARS
20:18:13 -!- WangZeDong has changed nick to Slereah.
20:18:50 <oerjan> CIVILIZATION FORMS BASED ON BEAR MARKETS
20:20:51 <oerjan> MAINLY BASED ON TRADING STAG MEAT
20:22:10 <oerjan> GENETIC ENGINEERING GREATLY INCREASES MEAT PRODUCTION
20:22:44 <oerjan> ALSO KNOWN AS STAGFLATION
20:22:49 <AnMaster> <oerjan> CIVILIZATION FORMS BASED ON BEAR MARKETS <ehird> groan <-- IDGI
20:23:16 <ehird> AnMaster: The terms bull market and bear market describe upward and downward movements respectively and can be used to describe either the market as a whole or specific sectors and securities (stocks).
20:23:21 <ehird> http://en.wikipedia.org/wiki/Market_trends#Bear_market
20:24:41 <oerjan> FTL TRAVEL INVENTED. USED TO INVESTIGATE NEARBY BLACK HOLE.
20:24:47 <ehird> For the lose travel
20:25:39 <oerjan> GIANT BULL RETRIEVED, REVOLUTIONIZING MEAT PRODUCTION
20:26:18 <oerjan> UNIVERSE LEAVES RECESSION
20:26:25 * pikhq listens to "Dark Side of the Moon"
20:26:39 <AnMaster> pikhq, sounds like a reference to a TP novel
20:26:50 <ehird> AnMaster: sounds like a pink floyd album.
20:26:55 <pikhq> It's a Pink Floyd album.
20:27:11 <AnMaster> I didn't know about that though :P
20:27:22 <ehird> AnMaster: pratchett book: 1976. dsotm: 1973.
20:27:35 <AnMaster> ehird, so it is the other way around then
20:27:36 <ehird> and "dark side of the moon" as a phrase is older.
20:27:41 <ehird> AnMaster: no, it's not any way around
20:27:43 <AnMaster> or possibly not connected at all
20:29:09 <oerjan> UNIVERSE ENDS IN THE BIG RIP-OFF
20:29:11 <AnMaster> ehird, did you know that the city of "Sto Lat" in the Discoworld happens to be the name of a Polish traditional song. I read some interview with TP where he said that completely unitentional.
20:29:13 <ehird> AnMaster: I'm surprised you didn't know about the album; it's generally considered one of the most known/significant albums of all time (maybe on that scale for "modern cultural icon" too)
20:29:24 <AnMaster> http://en.wikipedia.org/wiki/Sto_lat
20:29:37 <Deewiant> I don't think I've ever heard of the album
20:29:52 <ehird> Deewiant: Seriously?
20:30:10 <Deewiant> I'm somewhat known for not knowing things
20:30:33 <AnMaster> http://wiki.lspace.org/wiki/Sto_Lat mentions it too. But not the "unintentional" bit.
20:30:46 <pikhq> It's one of the best-known albums.
20:31:05 <pikhq> It stayed in Billboard's top 200 albums list for about 13 years...
20:31:05 <oerjan> now if sto helit had a hidden meaning we might start doubting him
20:31:19 <ehird> pikhq: it's one of the best known items of modern culture...
20:31:29 * ehird runs xdm. /me's eyes burn out
20:31:32 <pikhq> ehird: Yeah, definitely.
20:31:48 <ehird> CRISS CROSS BLACK/WHITE AS FORM OF: GRAY, TORTURE
20:32:58 <oerjan> ah, so "sto lat" means a hundred years
20:33:31 <ehird> "in which the subject is wished a life one hundred years ("sto lat") long."
20:33:37 <ehird> i see the polish don't have very high ambitions
20:34:05 <ehird> He wasn't talking about xdm :P
20:34:10 <ehird> also, xdm IS terrible
20:34:17 <ehird> http://upload.wikimedia.org/wikipedia/commons/e/ec/Xdm_Screenshot.png
20:34:19 <AnMaster> ehird, yes. Worse than twm even
20:34:20 <ehird> zoom in on that gray
20:34:26 <ehird> it seriously kills your eyes
20:34:29 <ehird> it hurts to look at
20:34:47 <AnMaster> ehird, that is the default bg colour of X with no window manager stared
20:35:01 <Deewiant> Zoom out and it looks prettier
20:35:21 <AnMaster> I don't know what format <color> is in
20:36:41 <AnMaster> Deewiant, what if you want #E6E6F8 for example
20:36:55 <AnMaster> which is what I prefer for bg colour
20:37:42 -!- kar8nga has quit (Remote closed the connection).
20:38:58 <pikhq> Either that or X11 color names.
20:39:40 <AnMaster> lifthrasiir, in my tests cfunge is about twice as fast on that example file
20:39:55 <AnMaster> lifthrasiir, Release build of cfunge with 32 bit cells
20:40:02 <fizzie> "For backward compatiblity" also "#e6e6f8" color specification should work.
20:41:04 <AnMaster> cfunge: 32bit, exact-bounds, concurrency
20:41:13 <AnMaster> hm I should make a build string thingy or such
20:41:24 <AnMaster> so you can get all that info from cfunge -f in one line
20:42:07 * ehird replaces xdm with wdm
20:42:28 <ehird> anyone used icewm?
20:42:42 <ehird> Deewiant: how can I make its menus not require clicking to open?
20:43:27 <Deewiant> Can't remember how to config that
20:44:06 <ehird> IceWM is less Cool-For-The-Sake-Of-Unusability-I'm-So-Minimalist-Yo
20:44:21 <Deewiant> Which might be why I didn't like it ;-)
20:44:40 <Deewiant> I can't actually remember, I had some small issue with each of Ice and Flux
20:45:04 <fizzie> Changing window managers every now and then is good for you; keeps you from getting into a rut.
20:45:30 <ehird> fizzie: why, which do you use?
20:46:07 <fizzie> ehird: Oh, I've been using awesome lately. But I've been through so many I couldn't reconstruct the complete list anyhow.
20:46:39 <fizzie> The "lavender" X11 color name (#e6e6fa) is probably closest to that requested color.
20:48:15 <fizzie> I did consider trying out Ion, but, well, this is directly quoted from Ion's FAQ re Xinerama support: "Xinerama this, Xinerama that, baawaawaa!! Too bad. I'm not going to waste my time rubbing your unecological penis enlargement."
20:48:39 <ehird> fizzie: Tuomov is a retard; he's switching to Windows.
20:48:44 <ehird> Because he had some problems with an old Ubuntu.
20:48:51 <ehird> His blag is quite lollerific
20:49:01 <Deewiant> I wonder what he's going to do with Ion on Windows
20:49:11 <ehird> Deewiant: Run Linux in a VM, I think he said.
20:49:13 <ais523> blag = short for "weblag", the amount of time it takes to access a website due to internet latency?
20:49:24 <ehird> ais523: Let's just say yes.
20:49:58 <Deewiant> ehird: I wonder how running Linux in a WM is supposed to be an improvement over running it natively
20:50:06 <ehird> Deewiant: Why are you applying logic
20:50:11 <AnMaster> <fizzie> The "lavender" X11 color name (#e6e6fa) is probably closest to that requested color. <-- ?
20:50:50 <AnMaster> well that isn't very close IMO
20:51:52 <fizzie> "-- it [the Ion license] is targeted against distributions – the middle-men, the equivalents of the big record companies in FOSSphere – that have become far too powerful, wielding de facto central control over easily installable software --" such vitriol.
20:52:04 <AnMaster> lifthrasiir, so how is your compiled then.
20:52:24 <lifthrasiir> AnMaster: with default options. no change applied.
20:52:42 <AnMaster> lifthrasiir, set the build type to RELEASE and try again
20:52:56 <AnMaster> maybe I should try to figure out how to make that default
20:54:03 <ehird> fizzie: he's also threatened to license it under a license which requires packagers to upgrade it 30 or so days after every release or remove it
20:54:07 <ehird> because debian had a version a month or two old
20:54:27 <fizzie> Yes, well: "The Linux/FOSS distributions are such powerful entities that will abuse authors (and users!) as long as they remain in power."
20:54:52 -!- bsmntbombdood has quit (Read error: 104 (Connection reset by peer)).
20:55:16 <AnMaster> lifthrasiir, by changing to "Release"?
20:55:17 <ais523> ehird: I'm sure that wouldn't be DFSG-compliant
20:55:26 <ehird> ais523: it wouldn't eb.
20:55:36 <AnMaster> lifthrasiir, I tend to use ccmake though.
20:55:42 -!- bsmntbombdood has joined.
20:55:43 <AnMaster> should figure out how to make default
20:55:46 <ehird> he said debian would have to move it into the non-free repository, but i'd just leave the old version or remove it altogether
20:56:04 <ais523> even funnier, fork the old version
20:56:13 -!- bsmntbombdood has quit (Read error: 104 (Connection reset by peer)).
20:56:20 <ehird> "(They might not bother: the trend seems to be to force everyone to use Xft shite and blurred fonts.)"
20:56:34 <ehird> this just in: fonts that have actual curvature that you can read are BLURRY!
20:57:00 <ehird> Mm, IceWM+gtk theme mist+Debian is quite comfortable.
20:57:03 -!- bsmntbombdood has joined.
20:57:10 <fizzie> Ion3 is in debian non-free now.
20:57:21 <ehird> If only I could hover on IceWM menus to make them view.
20:58:04 <ehird> Also maybe smooth out the dis— you know, I could rewrite ionwm with gtk in a few hours, probably :P
21:00:26 <ehird> anyone know what that pretty package installer thing is?
21:00:30 <ehird> the more lightweight one
21:00:44 <ehird> ais523: ubuntu's "add/remove"
21:00:55 <ais523> oh, I've known it at least twice
21:00:59 <ais523> but forgotten each time
21:02:05 <ehird> Is the problem Gnome-related?
21:02:07 <ehird> — Ion bug reporting page
21:02:15 <ehird> Did you run any Gnome software in your session (especially gdm or gnome-settings-daemon)?
21:02:16 <ehird> Yes. Remove all Gnome software from your system. Then go to next step.
21:02:19 <AnMaster> <ehird> this just in: fonts that have actual curvature that you can read are BLURRY!
21:02:27 <AnMaster> that is opposite of what you think
21:02:28 <ehird> damn tuomov is funny
21:02:44 <ais523> ehird: gnome-app-install
21:04:58 -!- MizardX- has joined.
21:05:10 <ehird> Maybe Hirdux will end up as Debian + a shell script to install some stuff :P
21:06:01 -!- MizardX has quit (Nick collision from services.).
21:06:25 -!- MizardX- has changed nick to MizardX.
21:07:14 <AnMaster> ais523, I'm about to break IFFI again...
21:07:31 <AnMaster> ais523, to be able to provide useful -v option somewhat like gcc -v to check compile flags.
21:07:42 <ais523> why would that break IFFI? I don't use your main
21:07:52 <AnMaster> ais523, /home/arvid/src/cfunge/trunk/src/main.c:196: error: expected ‘)’ before ‘CFUN_USER_CFLAGS’
21:08:04 <AnMaster> well I could add some #ifndef IFFI
21:08:21 <AnMaster> so you could skip that completely
21:08:24 <ais523> there aren't any useful command-line options with IFFI as-is
21:08:54 <AnMaster> ais523, true. But you still need parts of main.c. Since some variables are defined there.
21:09:32 <AnMaster> but yes I could ifdef them out when ICK_IS_IFFI or whatever is defined (either ICK_* or CFUN_IS_IFFI should be the name of the define)
21:10:24 <ehird> Deewiant: you know how you said a desktop environment is needless?
21:10:42 <ehird> if you don't have one you get to use 5 billion incompatible setting programs
21:10:47 <ais523> AnMaster: CFUN_ would be fine, define names don't come up in user code
21:10:58 <Deewiant> I don't use setting programs, I use setting files :-P
21:11:10 <ehird> Deewiant: Gee, that sure sounds fun.
21:11:20 <AnMaster> ais523, CFUN_* FUNGE_* and CF_* are all used in cfunge. Or no prefix at all. Depends on how old that code is.
21:11:27 <Deewiant> ehird: What do you config so much that it matters?
21:11:39 <ehird> Nothing much, but it is irritatin
21:11:44 <fizzie> I find it hard to take this stuff seriously when terms like "the AA fascist movement" are used.
21:11:59 <fizzie> (When speaking of people who like anti-aliasing in their fonts.)
21:12:24 <AnMaster> fizzie, I assume he has one of those 300 dpi monitors then
21:12:33 <AnMaster> where AA actually would be bad.
21:12:35 <ehird> AnMaster: No, he just things edges are edgy.
21:12:44 <ehird> Grr, gksudo sux since you lose theme settings
21:12:55 <fizzie> AnMaster: No, no. "200dpi is not enough for blurring not to show. I can see the dots in 300dpi prints. -- You have to embrace the pixels, not act as if they are not there."
21:12:58 <AnMaster> ehird, why are you running graphical programs as root!?
21:13:03 <ehird> AnMaster: synaptic, etc.
21:13:14 <AnMaster> lots of not well reviewed code
21:13:15 <ehird> only as insecure as running a cli program as root.
21:13:41 <AnMaster> (but it is likely to change in the future, thanks to GEM and kernel mode switching)
21:14:00 <AnMaster> <fizzie> AnMaster: No, no. "200dpi is not enough for blurring not to show. I can see the dots in 300dpi prints. -- You have to embrace the pixels, not act as if they are not there." <-- what
21:14:12 <ehird> AnMaster: He's mad.
21:14:21 <ehird> What's debian's nonfree repository called again?
21:15:01 <fizzie> "You're expected to sacrifice the present on the altar of the future, and to sacrifice personal ergonomics on the altar of fads and techno-high priests."
21:15:12 <fizzie> It is very fontcubeish.
21:15:33 <fizzie> When you antialias, you're sacrificing the present on the altar of the future!1
21:16:56 <ehird> Why is it so ugly?
21:16:56 <ehird> a) It is not. Are you sure you're not on a bad trip from the aqua in teletubbyland?
21:17:52 <ehird> &°·̂‡†̂̀‡̂%7ˆÞt675
21:17:55 <AnMaster> checking for suffix of object files... configure: error: in `/home/anmaster/gcc/build/i686-pc-linux-gnu/libgcc':
21:17:55 <AnMaster> configure: error: cannot compute suffix of object files: cannot compile
21:17:55 <AnMaster> See `config.log' for more details.
21:18:09 <ehird> fizzie: man, footnote 1
21:18:19 <ehird> he actually ties antialiasing into being un eco friendly
21:18:48 <fizzie> Yes, they're building separate power plants for all the CPU cycles consumed by "the blurring".
21:18:51 <pikhq> ehird: That's quite impressive.
21:19:40 <pikhq> Also, I suspect that these people have only seen very, very bad anti-aliasing.
21:19:53 <ehird> pikhq: No, he complains about Xft.
21:19:56 <ehird> Tuomov is just insane, man.
21:20:04 <ehird> OS X antialiasing would probably drive him insane
21:20:11 <ehird> OMG IT'S AS CLEAR AS PRINT!! ANTI-PIXEL BLURRING EVIL!
21:20:16 <ehird> ALL MY BOOKS USE BITMAP FONTS
21:20:16 <pikhq> Huh. I was under the impression that Xft actually did decent anti-aliasing.
21:20:26 <ehird> pikhq: Eh, it's alright. Not as good as OS X's.
21:20:32 <AnMaster> ais523, I believe it ignores LDFLAGS somewhere
21:20:42 <ehird> pikhq: Its subpixel rendering in particular is quite horrific
21:21:03 <pikhq> Though OS X handles fonts not designed with it in mind much better.
21:21:21 <ais523> AnMaster: gcc's configure is terrible, you'll go mad if you look at it
21:21:29 <ehird> Freetype handles Helvetica nicely
21:21:34 <ehird> That's about it :-P
21:21:36 <ais523> so much so that even though it did everything C-INTERCAL needed, I wrote my own method of doing it rather than using gcc's
21:21:40 <pikhq> ehird, that's because of patent law.
21:21:57 <pikhq> It can't use *good* auto-hinting, because that's patented.
21:21:58 <ehird> pikhq: I've ran the HIGHLY ILLEGAL IN THE US patched freetype.
21:22:17 <ehird> The autohinter, in particular, produces consistently worse results than the interpreter.
21:22:53 <pikhq> There's a reason I run Helvetica, Bitstream, and Lucida. ;)
21:23:34 -!- Hiato1 has quit ("Leaving.").
21:23:53 -!- Hiato has joined.
21:25:08 <ehird> I'd better write my nice WM.
21:25:16 <AnMaster> ais523, seems I can set BOOT_LDFLAGS
21:25:28 <AnMaster> and then wait a few more hours
21:25:30 <ais523> there are so many steps in compiling gcc that there are a lot of options
21:25:38 <AnMaster> http://gcc.gnu.org/install/build.html
21:25:50 <AnMaster> a grep shows there is a corresponding LDFLAGS one
21:26:00 <ehird> Shouldn't be exceedingly hard. Draw a simple border around a window -- mimicking Metacity's Mist theme -- with some simple buttons, double-click-to-maximize behavior and drag-to-move.
21:26:29 <ehird> Then, write a panel application that displays a bog standard menu with blah blah launchers, some application shortcuts and a window list, still with gtk.
21:27:47 <fizzie> Unfortunately, in practice, X.
21:28:02 <ehird> Especially since I want to do it non-reparenting.
21:28:04 <ehird> That could turn out hard.
21:30:30 <GregorR> I'm sorry, but I own patent #54372895472543: "IRC Nicks Beginning with the letters 'G' then 'R'". You will have to change your nick or pay me royalties.
21:31:01 <ais523> IBM took out a patent on not using Lotus Notes in meetings
21:31:29 <ais523> yes, although luckily it's more specific than that
21:32:37 <fizzie> "METHODOLOGY AND PROCESS FOR SUPPRESSING DE-FOCUSING ACTIVITIES DURING SELECTIVE SCHEDULED MEETINGS"; quite a name.
21:32:44 <ehird> Anyone else have an appreciation for the gtk/metacity theme Mist? It really does make gtk quite the nice.
21:34:53 <ais523> ehird: it's included in Ubuntu Jaunty, but I'm not sure about it
21:35:09 <ehird> ais523: you just can't handle the minimalism!
21:35:24 <ehird> You're, um, sacrificing usability at the altar of the future. I think
21:35:41 <fizzie> Anything with a title bar is not minimalist.
21:35:51 <pikhq> GregorR: I already own patent 0xDEADBEEF: "Nicknames representable only using ASCII", though. You violate that. ;)
21:36:01 <ehird> fizzie: The gtk theme is separate from the metacity one.
21:36:06 <pikhq> Using only, rather.
21:36:10 <ehird> Although my mega wm will imitate the window decoration style; it is rather tiny
21:36:22 <GregorR> I was having weird problems with my credit card, so I went to my bank and they "fixed" it, then I charged something to it, and now they've canceled it X_X
21:36:36 <fizzie> That's one definition for "fixed".
21:37:16 -!- tombom has joined.
21:37:59 <ehird> fizzie: have you seen the Mist window decoration? It's just a tiny blue line with the window title and the three ubiquitous buttons as boring white symbols :-P
21:38:06 <AnMaster> <ais523> IBM took out a patent on not using Lotus Notes in meetings <fizzie> "METHODOLOGY AND PROCESS FOR SUPPRESSING DE-FOCUSING ACTIVITIES DURING SELECTIVE SCHEDULED MEETINGS"; quite a name.
21:38:08 <ehird> It's not tinywm minimalist, but it's minimalist
21:38:30 <fizzie> Using Lotus Notes during the meeting is the "de-focusing activity" they're suppressing.
21:38:30 <AnMaster> and doesn't IBM develop Lotus *
21:38:31 <ais523> AnMaster: the patent was about Lotus Notes, and how to prevent people using it during a meeting
21:38:47 <ais523> the joke, of course, is that if IBM patent not using Lotus Notes, everyone else has to use it
21:39:22 <fizzie> http://appft1.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PG01&p=1&u=%2Fnetahtml%2FPTO%2Fsrchnum.html&r=1&f=G&l=50&s1="20090063996".PGNR.&OS=DN/20090063996&RS=DN/20090063996
21:39:28 <ais523> it's an office suite, I think
21:39:33 <ais523> is it based on OpenOffice? I can't remember
21:39:51 <fizzie> Lotus Notes is e-mail and calendaring and such.
21:39:53 <ehird> Lotus notes is ancient
21:40:00 -!- puzzlet has quit (Remote closed the connection).
21:40:05 <ehird> ais523: 1989 ancient
21:40:07 -!- puzzlet has joined.
21:40:28 <AnMaster> ais523, did they do it as a joke
21:40:42 <ais523> I don't see anything particularly joky about it
21:40:52 <ais523> just because it's possible to joke /about/ it doesn't make it in itself a joke
21:41:12 <ehird> ibm filing a patent as a joke?
21:41:16 <fizzie> But really, what they patented is just a system that lets you automagically (and enforcedly) disable some "distracting" features during meetings.
21:41:24 <ehird> stupidest thing I've ever heard
21:41:28 <fizzie> It's only jokeable if you deliberately misinterpret it; like most things are, I guess.
21:43:16 <fizzie> Patent language is quite a joke on its own, though. "A method for the suppressing computing system activities during a scheduled event, the method comprising: creating an invitation for an event at an event scheduling application executing at a primary computing system, the event invitation comprising an exclusive attendance event acceptance category and a non-exclusive attendance event acceptance category; .." continuing ad nauseam.
21:43:54 <fizzie> "-- accepting the event invitation at the event scheduling application executing at the secondary computing system, where in the instance that the event is accepted as an exclusive attendance event then communication and non-event scheduling application executing activities within the secondary computing system will be suspended upon the initiation of the event, and in the instance that the event is accepted as a non-exclusive attendance event a message is ge
21:43:54 <fizzie> nerated and delivered to the primary event scheduling application reporting that the event has been accepted as a non-exclusive event --"
21:44:05 <fizzie> All that without any punctuation in it.
21:44:17 <fizzie> Wait, there's a couple of commas.
21:45:53 <fizzie> The description parts usually seem human-readable, but the patent-claims-part is like in some sort of crude mockery of English.
21:47:37 <pikhq> Sadly, our patent system has completely and utterly failed at its purpose.
21:47:54 <ehird> pikhq: It's about as misguided as copyright law...
21:48:08 <pikhq> You know, publishing how an invention works so that after some time, it's part of the body of human knowledge?
21:48:30 <pikhq> Yeah, now we just publish some bullshit that could possibly describe the invention and then sue everyone.
21:49:53 <pikhq> ehird: Except that copyright reform would be quite a bit more radical. ;)
22:17:41 <AnMaster> I plan to add some way to get revision too
22:20:25 <AnMaster> Compiled on: Linux 2.6.27-gentoo-r8-1 (x86_64)
22:20:34 <AnMaster> now to find how to get compiler version from cmake too
22:31:25 -!- M0ny has quit ("When you get sad stop being sad and be awesome instead.").
22:35:41 -!- MizardX has quit (Read error: 104 (Connection reset by peer)).
22:36:45 -!- MizardX has joined.
22:46:07 -!- Sgeo_ has joined.
22:46:44 -!- tombom has quit ("Peace and Protection 4.22.2").
22:47:34 <AnMaster> GCC's preprocessor have computed includes
22:47:55 <AnMaster> http://gcc.gnu.org/onlinedocs/gcc-4.4.0/cpp/Computed-Includes.html#Computed-Includes
22:48:15 <AnMaster> I was looking for an official list of GCC predefined macros
22:48:24 <AnMaster> it seems to define __VERSION__ to what I need.
22:48:40 <AnMaster> but I'm not sure it is official or not
22:54:30 -!- Sgeo has quit (Read error: 110 (Connection timed out)).
22:57:07 <psygnisfive> but, ironically, the implementation would be inspired by something from linguistics ;p
22:58:03 <psygnisfive> an expression like the following pseudohaskell: _ ++ "foo" ++ _
22:58:31 <psygnisfive> would become the lambda \a.\b.(a ++ "foo" ++ b)
22:59:06 <psygnisfive> would become the lambda \a.(a ++ "foo" ++ a)
22:59:20 <Deewiant> I think Scala has approximately-if-not-exactly that
22:59:31 <Deewiant> Of course not Haskell syntax, but the _s.
22:59:53 <psygnisfive> the _'s arent supposed to be haskell's "these dont matter" variables
23:00:11 <psygnisfive> rather, theyre a sort of "i dont care about the name of this variable" variable
23:01:22 -!- puzzlet has quit (Remote closed the connection).
23:01:26 -!- puzzlet has joined.
23:01:36 <psygnisfive> have some sort of type-shifting operations
23:02:02 <psygnisfive> meaning, whatever type the thing that you give the lambda
23:02:57 <psygnisfive> but _ ++ "foo" has a type-gloss of t (String/String)\String String
23:03:43 <oerjan> erm, not String -> String?
23:04:12 <oerjan> oh some linguist thing...
23:04:34 <psygnisfive> a gloss is just an item-by-item "Translation" of sorts
23:04:59 <psygnisfive> so _ is type t, ++ is type (String/String)\String, and "foo" is type String
23:06:59 <psygnisfive> so the type-shifter consumes the items left-to-right like
23:08:02 <oerjan> well, that (String/String)\String i seem to recall you doing with grammar previously
23:08:37 <psygnisfive> yeah, its CCG directional lambda application notation
23:09:11 <psygnisfive> essentially (A\B)/C is a left-first lambda type
23:09:46 -!- FireFly has quit ("Later").
23:09:46 -!- MizardX has quit (Read error: 60 (Operation timed out)).
23:09:59 <AnMaster> "The compiler values a multi-character character constant a character at a time, shifting the previous value left by the number of bits per target character, and then or-ing in the bit-pattern of the new character truncated to the width of a target character. The final bit-pattern is given type int, and is therefore signed, regardless of whether single characters are signed or not (a slight change fro
23:09:59 <AnMaster> m versions 3.1 and earlier of GCC). If there are more characters in the constant than would fit in the target int the compiler issues a warning, and the excess leading characters are ignored."
23:10:31 <AnMaster> http://gcc.gnu.org/onlinedocs/gcc-4.4.0/cpp/Implementation_002ddefined-behavior.html
23:10:52 <ais523> AnMaster: multicharacter character constants in C exist, but are implementatino-defined
23:11:02 <AnMaster> ais523, yes. But why that behaviour
23:11:12 -!- Hiato has quit (Read error: 104 (Connection reset by peer)).
23:11:12 <AnMaster> why is that specific behaviour useful
23:11:26 <ais523> it's the most common behaviour for a multichar constant
23:11:47 <AnMaster> why is that an useful behaviour though
23:12:19 <lifthrasiir> for example, pushing handprint in funge-98. e.g. ip_push(ip, 'WTF!');
23:12:58 <psygnisfive> so the type-shifter gets to the ++ and says; ok, _ is type t, and ++ is type (Str\Str)/Str, so I have to lift _ to type Str/(Str\Str)
23:13:18 -!- BeholdMyGlory has quit (Remote closed the connection).
23:13:44 <AnMaster> lifthrasiir, I don't do it that way, I have a defined hexdecimal constant
23:13:48 <lifthrasiir> okay, i'm just kidding but such constants are hardcoded to several file formats, for example.
23:14:14 <psygnisfive> once _ is lifted to Str/(Str\Str), it can compose with ++ (Str\Str)/Str to produce something like \x.(_ ++ x)
23:14:35 -!- MizardX has joined.
23:15:33 <psygnisfive> the _ itself is probably some sort of magic lambda that stays on the outermost periphery of this whole thing
23:15:51 -!- coppro has joined.
23:16:26 <psygnisfive> so that "_ ++" actually produces something of the type \M.\x.(M ++ x)
23:16:46 <psygnisfive> but since the _ objects are parsing magic, not application magic, its all cool.
23:17:22 <ehird> psygnisfive: scala & arc do that, but with less ridiculous "type gloss" rubbish.
23:17:28 <ehird> it's a trivial syntactic translation
23:18:00 <psygnisfive> the type gloss stuff is a) not rubbish, and b) just a way to represent the types for describing whats going
23:18:18 <ehird> that's "nice" but also useless, it's a trivial syntactic translation
23:18:55 <psygnisfive> maybe! but im thinking about it in combinatory terms. :P
23:20:58 <psygnisfive> mainly because it has no lambda delimiters to denote the whole thing as a lambda
23:21:12 <psygnisfive> so im not sure how i'd code something to recognize the extent of the lambda
23:23:07 <psygnisfive> i dont even remember at this point why i had this idea. XD
23:23:16 <psygnisfive> but it seemed like something i could use at the time
23:23:35 <oerjan> i recall there were some discussions in #haskell about this
23:24:08 <oerjan> mostly, it wasn't considered worth complicating the language for, i think
23:25:16 <oerjan> (haskell already has sections, which is essentially an implicit such _ at one end of a parenthesized (x ++) or (++ x)
23:25:47 <psygnisfive> the intention was to have a sort of... auto-lambda-izer for arbitrary expressions
23:26:42 <oerjan> it gets really messy if nested, with itself or with ordinary lambdas
23:29:43 <AnMaster> ais523, can you update IFFI build process to define CFUN_IS_IFFI
23:31:43 <oerjan> like, if you have both _ and _a, and no specific indication which is innermost
23:31:43 <ais523> AnMaster: pretty easily, but not now
23:32:09 <ais523> AnMaster: because it's late and I'm about to go home
23:32:11 <psygnisfive> oh. well, my intention was it'd just be in-order
23:32:13 <oerjan> so _ _a could mean either \_ _a -> _ _a or \_a _ -> _a _
23:32:15 <ais523> and in the middle of conversations with other people
23:32:20 <AnMaster> ais523, and do it tomorrow then?
23:33:16 <oerjan> or wait, even \_ -> _ \_a -> _a
23:34:00 <psygnisfive> the magic lambdas just go in order of appearance of the _'s
23:34:30 <oerjan> well i guess the question is where the whole thing starts
23:34:47 <psygnisfive> the intention is that its the whole expression that the _ appears in
23:35:44 <oerjan> ah yes that was another problem for haskell, it would mess with referential transparency
23:35:45 <AnMaster> ais523, I'm going to commit that then
23:36:09 <psygnisfive> {...} could delimit the magic lambda, i guess.
23:36:10 <oerjan> since you couldn't just substitute in an expression containing _'s
23:37:30 <psygnisfive> a language that takes full advantage of type-shifting and function composition
23:37:30 <AnMaster> ais523, pull from me please :)
23:38:25 <psygnisfive> so that expressions like sqrt + negate 5 == (\x -> sqrt x + negate x) 5
23:39:03 <ehird> psygnisfive: you just invented forks
23:39:14 <ehird> in J, (sqrt+negate)x = sqrt x + negate x
23:39:25 <ehird> well all commands in j are one-char
23:39:27 <AnMaster> ais523, btw since a few days cfunge has an option to disable all floating point calculations. (That is disable all fingerprints using them, core doesn't use it)
23:39:33 <ehird> and psygnisfive, that's EXACTLY your concep.
23:39:36 <psygnisfive> in J you need to use special forked functions, i think, ehird
23:39:53 <AnMaster> ais523, this is of course a step in my evil plan to take over the world of embedded funge. ;P
23:39:55 <ehird> ((verb1 verb2 verb3) noun) = ((verb1 noun) verb2 (verb3 noun))
23:40:46 <psygnisfive> because it should be able to do more complex things
23:40:56 <AnMaster> ais523, it is possible (with advanced options) to get cfunge only to depend on libc now. Dropping need of libm, librt (optional, but previously always used if found), ncurses (likewise before)
23:41:05 <ais523> time to go home, anyway
23:41:09 -!- ais523 has quit (Remote closed the connection).
23:41:23 <AnMaster> since clock_gettime() is in libc on freebsd
23:41:30 <ehird> psygnisfive: you can do that with just a fork
23:41:30 <AnMaster> and all other platforms I know of
23:41:41 <oerjan> psygnisfive: something like this has been done partially in haskell by creating numerical type class instances for functions
23:41:53 <AnMaster> GregorR, limited output more yet
23:42:00 <ehird> psygnisfive: if you have an apply function you can do it, probably
23:42:04 <ehird> the fork will be ugly but meh
23:42:22 <oerjan> although that's not very generic
23:42:27 <psygnisfive> the interpreter can recognize type-mismatches, and lift around them
23:42:50 <ehird> psygnisfive: the problem is that your ideas, while sounding nice, often come down to trivial transformations :)
23:42:55 <ehird> i attribute this to you being a linguist ;-)
23:43:18 <AnMaster> psygnisfive, use some complex ones then
23:44:24 <GregorR> <AnMaster> GregorR, limited output more yet // no
23:44:30 <psygnisfive> its not really a simple transformation tho
23:44:38 <oerjan> psygnisfive: well AnMaster is hardly the person to ask about convincing ehird ;D
23:44:45 <psygnisfive> its not a syntactic thing, atleast not at the level of the syntax of the language
23:45:53 <oerjan> psygnisfive: but then you are instead getting around to needing a "sufficiently advanced type system"
23:46:30 <ehird> have I mentioned that dependent types solve every woe
23:47:10 <ehird> permutations of (values|types) indexing on (values|types), truly une beautifule!
23:47:53 <oerjan> psygnisfive: i know that haskell's type system very much _avoids_ trying to deduce alternatives to type mismatches, presumably because it would be unsolvable and/or unbearably messy
23:47:54 <psygnisfive> oerjan, on the offchance that you're interested in the sort of ideas that im basing this off of, check out Combinatory Categorial Grammar
23:48:08 <ehird> http://adblockplus.org/blog/attention-noscript-users
23:48:30 <oerjan> (well with type classes that is)
23:48:31 <psygnisfive> its a minor extension on plain old categorial grammar, which uses a simple directional-application function type
23:48:56 <psygnisfive> see, type-lifting and such are very well defined operations in CCG
23:49:20 <psygnisfive> theres basically only two novel operations in CCG that make it do what it does
23:49:59 <psygnisfive> given a function of type a, you can live to type (a -> b) -> b
23:50:13 -!- puzzlet_ has joined.
23:50:57 <psygnisfive> the other novel operation in CCG isnt novel to us, its just function composition, but its novel for CG
23:50:59 <ehird> psygnisfive: if you will speak of linguistics without any context, please learn logic too.
23:51:26 <oerjan> curry-howard isomorphism
23:51:28 <ehird> oerjan: yep, and (a,b=>c) == (a=>b=>c)
23:51:44 <ehird> you can do all logic with just forall and →, I think.
23:51:50 <ehird> with _|_ = forall a.a
23:51:57 <psygnisfive> oerjan, yes, sure, i guess. i mean, afaik these ideas themselves arent magically knew, you know, but
23:52:00 <ehird> (and ~a = a→_|_ ofc)
23:52:24 <ehird> a^b = forall c.(a→b→c)→c
23:52:25 <oerjan> ehird: i think calculus of constructions does that
23:52:27 <psygnisfive> i mean, CCG has been around since the early 80s, i guess? and its an extension off of traditional logic as applied to natural language
23:52:40 <ehird> oerjan: what about or?
23:52:54 <ehird> aVb = forall c. ((a→c)V(b→c))→c
23:52:57 <ehird> but that's self-referential
23:54:01 <ehird> aVb = forall c. (a→c)→(b→c)→c, maybe?
23:54:14 <ehird> if you can't satisfy one, do (a→_|_)?
23:54:27 <ehird> yeah, that seems right
23:56:28 <ehird> oerjan: do you have the unicode _|_?
23:56:54 <psygnisfive> oerjan: ftp://ftp.cogsci.ed.ac.uk/pub/steedman/ccg/manifesto.pdf CCG manifesto. lol
23:57:26 <psygnisfive> because theres no fun in using someone elses language
23:58:07 <oerjan> Slereah: ~a is defined using _|_
23:58:46 <Slereah> oerjan : Only if you chose it such :3
23:58:58 <psygnisfive> plus, i dont know if J does what im asking for, so nya. :P
23:59:11 <Slereah> Defining symbols with _|_ means you have to include at least two
23:59:12 <psygnisfive> not like im /asking/ for this, as such, but you know
23:59:35 <oerjan> Slereah: _|_ = forall a.a, said ehird
23:59:41 <psygnisfive> wtf is all this -> and _|_? _|_ looks like the symbol for bottom
23:59:55 <ehird> A ∧ B = ∀C. (A ⇒ (B ⇒ C)) ⇒ C
23:59:57 <ehird> A ∨ B = ∀C. (A ⇒ C) ⇒ ((B ⇒ C) ⇒ C)