00:48:24 -!- tusho has quit.
01:14:42 -!- puzzlet_ has quit (Read error: 104 (Connection reset by peer)).
01:15:42 -!- puzzlet has joined.
01:21:15 -!- puzzlet_ has joined.
01:31:58 -!- psygnisfive has joined.
01:34:31 -!- puzzlet has quit (Connection timed out).
02:31:54 -!- boily has joined.
02:39:21 -!- calamari has joined.
03:14:15 -!- boily has quit ("leaving").
03:45:39 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | The talk page is made of gold.
03:55:54 -!- calamari has quit ("Leaving").
05:06:32 -!- psygnisfive has quit ("http://www.mibbit.com ajax IRC Client").
05:23:19 -!- GreaseMonkey has joined.
05:25:12 * oerjan spots a kiwi and runs after it
05:49:29 -!- oerjan has quit ("leaving").
05:52:31 -!- bsmntbombdood has joined.
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:54:32 -!- fungot has joined.
08:55:01 <fizzie> I'm not completely certain that bot works correctly. This is what I got when I tried to test it via privmsg:
08:55:12 <fizzie> 10:54:16 <fizzie> Hey, fungot, does your nonsense generation work?
08:55:12 <fizzie> 10:54:16 <fungot> fizzie: i think not.
08:55:12 <fungot> fizzie: heh. i don't think
08:55:13 <fungot> fizzie: 30min sounds like a plural
08:55:41 <fizzie> fungot: Oh! So you said "i think not" because you actually do not think at all. Now I get it.
08:55:42 <fungot> fizzie: i was wishing we could use only 1-ary functions and currying really mutually exclusive?
09:04:18 <oklopol> where was the source again?
09:04:45 <fizzie> http://zem.fi/~fis/fungot.b98.txt is most of it.
09:04:47 <fungot> oklopol: take the last three days or so
09:05:18 <oklopol> i like the love triangle there
09:05:30 <fizzie> It's not the most current version, I think I got rid of NULL fingerprint usage at least.
09:05:38 <oklopol> seems it's some kind of tokenizer?
09:06:03 <fizzie> The triangle is the token-to-punctuation code of the nonsense generator.
09:06:34 <oklopol> you should add some more triangles
09:06:38 <fizzie> The first 16 or so tokens are handled by going to that 'j' and then letting that pick one of the 'v's there.
09:07:53 <fizzie> There's another jump table -like structure around line 264, for the brainfuck bytecode thing, but it's significantly less triangley.
09:07:54 -!- olsner has quit ("Leaving").
09:11:16 <fizzie> I don't really know why I bothered writing 200000 as "2aaaaa*****" when I could've gotten the reasonably close 2^18 with something like "8:*::**", a lot shorter.
09:29:49 -!- oklobol has joined.
09:31:04 -!- oklopol has quit (Read error: 113 (No route to host)).
09:45:39 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | PSOX is centered around brainfuck. Stating anything to the contrary is just a lie.
09:48:32 -!- sebbu2 has joined.
09:58:07 <oklobol> so, how about that mips in wireworld
10:05:33 -!- oklobol has changed nick to oklopol.
10:07:11 -!- sebbu has quit (Read error: 110 (Connection timed out)).
10:31:02 <AnMaster> <fizzie> The first 16 or so tokens are handled by going to that 'j' and then letting that pick one of the 'v's there.
10:32:12 <fizzie> There's another (a vertically oriented one) in the brainfuck bytecode interpreter.
10:32:59 <AnMaster> fizzie, why do you use a triangle btw?
10:33:20 <AnMaster> you could just write the code paths vertically
10:33:40 <AnMaster> and you seem to go right up again
10:33:43 <fizzie> It's pretty that way, and there was empty space. :p
10:34:03 <AnMaster> fizzie, you know that you are probably the only person who can read the code?
10:34:22 <AnMaster> fizzie, and you got a lot of trailing whitespaces on the line below "looptest"
10:34:23 <fizzie> I'm not sure I can read it all that well, since I haven't worked with it in a week or so.
10:34:47 <fizzie> Yes, there was probably something on that line and I erased only the visible parts.
10:34:56 <AnMaster> fizzie, you should document the code structure, I mean a diagram showing "brainfuck region" and so on
10:37:13 <fizzie> I probably would've documented it better if I had an editor which could attach comments to arbitrary regions of a text file.
10:38:15 <AnMaster> fizzie, I meant somewhat like the image on http://www.quinapalus.com/wires11.html
10:39:27 <fizzie> I guess I could do something like that, although it really isn't large enough to create a really impressive picture, and it's just tall and thin, not very square-like at all.
10:40:42 <fizzie> The "no more than 80 columns" rule I've followed has maybe made it easier to edit, but it's disappointingly linear code now.
10:51:39 <AnMaster> fizzie, want to see one of my own boringly linear test suites?
10:51:49 <AnMaster> http://rafb.net/p/VQGDpf66.html
10:53:09 <oklopol> "detroppus ton" i first thought this was latin or something
10:54:04 <AnMaster> oklopol, you clearly aren't seasoned in reading funge
10:54:04 <oklopol> i want to project that wireworld program on my wall
10:54:29 <oklopol> took like a second to realize it's a gnirts
10:54:47 <oklopol> but no i'm not seasoned to reading it
10:55:06 <AnMaster> oklopol, ever looked at mycology source?
10:55:28 <oklopol> that's really the only funge program i've read parts of
11:00:39 -!- GreaseMonkey has quit ("Jesus loves you").
11:27:56 <AnMaster> Deewiant, ah cause of infinite loop was u popping from TOSS to SOSS when positive instead of the other way
11:28:08 <AnMaster> still weird loop it ended up in
11:28:20 <AnMaster> BAD: 101-{} doesn't leave stack top as 0 and next as 1
11:28:20 <AnMaster> BAD: fedcba0{05-} doesn't leave 15 on stack
11:28:27 -!- tusho has joined.
11:28:27 <AnMaster> and I still don't understand that
11:28:43 <AnMaster> would be useful to print what happened instead
11:39:44 -!- oklopol has quit ("( www.nnscript.com :: NoNameScript 4.2 :: www.regroup-esports.com )").
11:42:46 -!- kar8nga has joined.
11:48:43 -!- sebbu2 has changed nick to sebbu.
12:02:04 -!- ais523 has joined.
12:06:09 <Slereah_> "# Any "if" statement requires at least 14,000 subconditions. "
12:06:22 <Slereah_> Silly Emplosions looks a little too silly.
12:06:38 <ais523> Silly Emplosions seems to have the same attitude as INTERCAL
12:06:44 <ais523> just without the need to be practical too
12:08:06 <tusho> ais523: I am going to write that autogreeter now, because I want to, and you have inspired me.
12:08:06 * ais523 is reading about the whole Firefox 3 EULA thing
12:08:27 <ais523> it seems the EULA is there to protect their trademarks, among other things
12:08:34 <ais523> and Canonical and Mozilla have got into a row about it
12:09:18 <tusho> ais523: i've decided that i really don't like mozilla
12:09:23 <tusho> their attitude really sucks
12:09:38 <ais523> yes, it's annoying me quickly too
12:09:48 <tusho> ais523: i have maintained for a while that they're no more a free software group than Sun or IBM or Apple
12:09:56 <tusho> but people just tended to ignore those kind of remarks
12:10:00 <tusho> now it seems more are listening...
12:10:02 <Slereah_> http://nukees.com/comics/nukees20080915.gif
12:12:07 <tusho> ais523: MWAHAHA! I HAVE CONQUERED APPLESCRIPT AND "hi ais523.scpt" VERILY EXISTS
12:12:08 <Slereah_> We've got CS class again this year I think D:
12:12:23 <tusho> here's the cringeworthy code: [[using terms from application "Colloquy"
12:12:23 <tusho> on member joined m in room
12:12:23 <tusho> if m's name is "ais523" then
12:12:23 <tusho> tell room to send message "hi ais523"
12:12:26 <tusho> end using terms from]]
12:12:42 <ais523> that is so AppleScript
12:12:53 <ais523> it's one of those languages which is instantly recognisable
12:14:59 -!- tusho has quit.
12:15:14 -!- tusho has joined.
12:15:25 <ais523> tusho: do you want me to /cycle to test?
12:16:00 -!- ais523 has left (?).
12:16:00 -!- ais523 has joined.
12:16:42 <tusho> ugh, where IS the scripts menu
12:18:00 -!- tusho has quit (Client Quit).
12:18:12 -!- tusho has joined.
12:18:32 -!- ais523 has left (?).
12:18:32 -!- ais523 has joined.
12:19:01 <tusho> ais523: any gcc-bf news?
12:19:04 <tusho> pikhq was asking yesterday
12:19:15 <ais523> tusho: I've done all the easy parts, just a few hard parts left
12:19:18 <ais523> and haven't really tested anything
12:19:30 <tusho> (pretty eventful day actually: Figs from circa 2007 was here and I got a submission to #1 place on reddit :P)
12:19:50 <ais523> also, what submission?
12:19:57 <tusho> http://www.reddit.com/r/programming/comments/71e5c/sketchpad_an_advanced_painting_program_from_1963/
12:20:07 <tusho> the CAD-ish program that invented OOP,
12:20:09 <tusho> non-functional languages,
12:20:18 <tusho> and a metric fuckton of other stuff
12:20:29 <tusho> the first comment there (a quote from the video) is nice: ""How could you do [all this] in one year?" "I didn't know it was hard.""
12:38:45 -!- Sgeo has joined.
12:50:38 -!- kar8nga has quit (Read error: 110 (Connection timed out)).
12:55:52 <tusho> http://www.webfoundation.org/news/2008/09/welcome-to-the-world-wide-web-foundation.html Neat. What does it do.
12:57:46 <ais523> tusho: site times out for me
12:57:51 <tusho> "The World Wide Web Foundation seeks to advance One Web that is free and open, to expand the Web's capability and robustness, and to extend the Web's benefits to all people on the planet. The Web Foundation brings together business leaders, technology innovators, academia, government, NGOs, and experts in many fields to tackle challenges that, like the Web, are global in scale."
12:57:58 <tusho> I guess that's good. A bit vague, but sure, whatever.
12:58:07 <tusho> (Tim Berners Lee is a co-foundery thingy)
12:58:23 <ais523> sorry, it's just the name reminded me of the OpenDocument Foundation
12:58:32 <ais523> who were a few guys who came up out of nowhere
12:58:41 <ais523> and then decided they weren't supporting ODF any more
12:59:27 <tusho> 'Steve became CEO of the World Wide Web Foundation in September 2008.'
12:59:29 <tusho> it was just founded you idiots
12:59:35 <tusho> you can't 'become' it
13:00:17 <tusho> ais523: From the site for Helvetica the film
13:00:19 <tusho> p.body { font-family: Verdana, Helvetica, Arial, sans; }
13:01:50 -!- kar8nga has joined.
13:03:22 -!- kar8nga has left (?).
13:14:13 <AnMaster> yay efunge almost pass mycology now, just a few details about y left
13:14:19 <AnMaster> basically, command line and env
13:16:07 <AnMaster> http://rafb.net/p/Qj6DvB92.html
13:16:53 <AnMaster> the indention got some issue with y, it is different if I run efunge freestanding instead of under the erlang shell
13:23:21 <ais523> AnMaster: all the easy bits are finished, there are several hard bits left
13:23:33 <ais523> and it isn't really called anything stable atm
13:24:11 <tusho> http://www.djangopony.com/ Django (python web framework) pony.
13:24:54 <tusho> (Follow up: http://hackety.org/2008/09/15/documentsRevealDjangoPonyTailOfLies.html)
13:26:32 <tusho> obviously sketchpadded though
13:27:27 <tusho> AnMaster: no. sketchpadded
13:27:37 <tusho> you can tell from the parallel lines
13:36:21 <tusho> AnMaster: why is known for his bizzare hand-drawn stuff, though, yes
13:36:43 <tusho> his old site had "code in images" like:
13:36:53 <tusho> http://redhanded.hobix.com/bits/gem_mirror_only.html, http://redhanded.hobix.com/cult/ahaNoticeTheExpandoWhichPrecludes.html, http://redhanded.hobix.com/bits/theHeartOfTryRuby.html
13:46:01 -!- cmeme has quit (zelazny.freenode.net irc.freenode.net).
13:49:38 -!- cmeme has joined.
13:57:25 -!- SirDayBat has joined.
13:57:37 -!- oklobol has joined.
14:03:21 <ais523> tusho: someone /msged me in NetHack, thinking it was about hacking
14:03:35 <ais523> so they asked me if I knew anything about hacking
14:03:39 <ais523> I said not that I'd tell anyone over IRC
14:03:47 <ais523> so they said they had Messenger too...
14:04:05 <tusho> ais523: aw, you should have pretended you thought he meant hacking as in the non-cracking sense
14:04:46 <AnMaster> ais523, how do you make plural of gnirts?
14:04:57 <ais523> AnMaster: gnirtses is about the best you can get
14:05:06 <ais523> within normal English pluralising rules
14:10:47 <ais523> fizzie: that would be the plural of gnirtsy
14:10:59 <ais523> on the other hand, it's not the sort of word which pluralises very well at all
14:14:18 <tusho> aoidajsiojasiodhaiusdhuadjs
14:21:20 -!- Sgeo has quit (Read error: 110 (Connection timed out)).
14:23:04 <oklobol> actually didn't work that well because no third s was added
14:23:20 <oklobol> funny how i thought there would be one but didn't put it there
14:35:03 <AnMaster> well I just needed a name for a function to push a list of gnirts in efunge
14:39:11 <AnMaster> bzr branch http://rage.kuonet.org/~anmaster/bzr/efunge/b98 efunge98
14:39:25 <Deewiant> AnMaster: for 101-{} and the others, the whole point of printing the problematic code is that you can do "echo '101-{}@' > foo.b98 && run-interpreter foo.b98"
14:40:04 <AnMaster> Deewiant, and the u got direction messed up -> infinite loop issue?
14:40:12 <Deewiant> and before you ask me about it again, yes there is no corresponding GOOD for that and the next test :-P
14:40:22 <Deewiant> AnMaster: probably fixed in the update I did this morning?
14:40:33 <Deewiant> AnMaster: a 'v' going to nowhere, right?
14:40:51 <AnMaster> not going to mess with u code again
14:41:08 <AnMaster> stackstack and y are the ugliest bits of code I have
14:41:28 <Deewiant> you haven't done fingerprints then yet, I suppose? ;-)
14:41:42 <AnMaster> nor am I sure how I would do it
14:42:00 <ais523> hmm... I'm going to have to write asmfunge at this rate
14:42:15 <AnMaster> but see the bzr link I posted a few lines above if you are interested!
14:42:26 <AnMaster> Deewiant, it doesn't implement tio= either
14:42:48 <AnMaster> and I'm not going to work more on it today
14:43:14 <AnMaster> Deewiant, an odd bug is that indention for mycology's y output differ between freestanding and running inside the erlang shell
14:43:18 <AnMaster> can't figure out the cause of that
14:43:32 <Deewiant> AnMaster: s/indention/indentation/ btw
14:43:52 <Deewiant> I've seen you make that mistake a couple of times so I figured I'd correct you :-)
14:43:57 <ais523> Deewiant: does Mycology indent with tabs?
14:44:04 <ais523> if so that might be the reason
14:44:30 <AnMaster> ais523, I use io:format in both cases, so I don't see why it would mangle tabs with erl -no_shell and not mangle them when starting from inside shell
14:44:58 <ais523> AnMaster: it's possible that the tabstops are different in the two situations
14:45:27 <Deewiant> of course it tabstops differ the indentation doesn't, it's still 1 or 2 tabs :-P
14:47:33 <AnMaster> ais523, Deewiant the two ways of running below each other: http://rafb.net/p/GSynYV92.html
14:48:26 <Deewiant> rafb eats tabs so that's not very informational
14:48:33 <ais523> it seems that you have a tabstop at 1 the second way of running
14:48:42 <AnMaster> Deewiant, well so does my terminal when I copy paste with select
14:48:54 <AnMaster> I don't think the erlang interactive shell even works if you redirect it
14:49:19 <ais523> AnMaster: does it work over ssh?
14:49:30 <ais523> if so, ssh into your own computer and redirect that via tee
14:49:54 <ais523> ssh localhost | tee filename
14:49:55 <AnMaster> but ssh simulates a terminal afaik
14:49:59 <ais523> (sorry, corrected version)
14:50:08 <ais523> would mean that you'd get the output in filename as well as onscreen
14:50:09 <AnMaster> well I used erlang over ssh even
14:50:18 <AnMaster> ais523, doesn't tee buffer output?
14:50:19 <ais523> so that's one way to capture the output
14:50:28 <ais523> AnMaster: only to the file, not to stdout I think
14:50:42 <ais523> at least not if stdout's a terminal
14:50:53 <ais523> OTOH the /input/ to tee might be buffered, which could be a problem
14:50:57 <AnMaster> when I redirected the freestanding output to a file, it contained tab chars
14:51:14 <AnMaster> though cfunge in same terminal give normal tabs
14:51:54 <AnMaster> I guess my terminal settings have got fucked up somehow
14:52:32 <AnMaster> well yes was just that terminal
14:52:49 <ais523> although I still have no idea how you managed to set a tabstop in column 1
14:52:58 <AnMaster> Deewiant, btw apart from getting env vars the program is fully portable I think
14:53:14 <AnMaster> I need to use os:get_env() for environment vars
14:53:20 <AnMaster> so that may or may not work on other OS
14:53:31 <AnMaster> ais523, catting a binary file or something like that I bet
14:53:47 <ais523> do you often cat binary files?
14:53:55 <ais523> if so, the "reset" command is very useful to know
14:54:07 <AnMaster> but as I said, reset fixed the tab stop
14:54:13 <AnMaster> but I probably did cat one by mistake
14:54:21 <ais523> before I learnt about reset I used to cat /dev/random to reset the terminal
14:54:31 <ais523> if you did it often enough the terminal normally ended up sane by chance eventually
14:54:34 <AnMaster> ais523, that is likely to mess it up even more
14:54:44 <ais523> but I didn't know that at the time
14:55:00 <ais523> and mostly it was about getting the terminal out of high-bit-set mode
14:55:09 <ais523> the resulting terminals often gave curses a field day, though
14:55:14 <ais523> and messed up ls output from time to time
14:55:28 <AnMaster> Deewiant, anyway I think efunge qualifies for mycology results page, though a release will have to wait a bit
14:55:39 <AnMaster> Deewiant, and it's the first bignum one
14:57:20 <AnMaster> wait I will change fingerprint to make use of bignum, I think that is valid
14:57:49 <ais523> heh, there should be a bignum fingerprint with a name like "BIGNUM"
14:57:56 <ais523> so you need 48-bit or better ints to use it...
14:58:07 <AnMaster> ais523, no need, since all cells are bignum here
14:58:19 <ais523> AnMaster: I was speaking in general
14:58:30 <AnMaster> "efunge - A bignum funge interpreter in Erlang" for fingerprint
14:58:52 <ais523> some ridiculously large number
14:59:23 <AnMaster> "efunge - A BIGNUM Befunge-98 interpreter in Erlang"
14:59:44 -!- jix has joined.
15:03:28 <AnMaster> 13> io:format("~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~.16B~n", "efunge - A BIGNUM Befunge-98 interpreter in Erlang").
15:03:28 <AnMaster> 6566756E6765202D2041204249474E554D20426566756E67652D393820696E74657270726574657220696E2045726C616E67
15:03:57 <AnMaster> so 16#6566756E6765202D2041204249474E554D20426566756E67652D393820696E74657270726574657220696E2045726C616E67
15:04:05 <AnMaster> Deewiant, what do you think about that for handprint?
15:04:57 <AnMaster> That the interpreter's handprint is 1022815351171939080750602515080902596655856329413174779991860562785344354620645109770120275823379669394014875628110704231
15:05:38 <ais523> gah, you beat me to it
15:05:53 <ais523> calculating that handprint
15:06:11 <ais523> I was just looking up how to get a character code in Haskell...
15:06:34 <AnMaster> ais523, I generated the format
15:06:54 <AnMaster> using lists:seq() and then lists:map()
15:07:07 <AnMaster> since in erlang a string is just a list of integers
15:07:47 <AnMaster> it makes a lot of funge stuff a lot simpler
15:07:47 <Deewiant> AnMaster: I sort of think that handprints should be as small as possible so that you can test for them portably
15:07:51 <ais523> AnMaster: a string is a list of char in Haskell
15:07:57 <ais523> and I can't remember the function to convert a char to an integer
15:08:02 <AnMaster> Deewiant, hehe ok, but it would be fun
15:08:29 <AnMaster> Deewiant, still the long handprint is much much more fun!
15:08:31 <Deewiant> http://haskell.org/hoogle/?hoogle=Char+-%3E+Int
15:08:52 <ais523> Deewiant: how do you cast an Int to an Integer in Haskell, then?
15:09:08 <AnMaster> Deewiant, and you could test for it, if you overflow you know it can't be it!
15:09:24 <AnMaster> anyway is funge cell integer overflow well defined?
15:09:41 <Deewiant> ais523: http://haskell.org/hoogle/?q=Int%20-%3E%20Integer again :-)
15:09:48 <ais523> Deewiant: thanks, I know much of the structure of Haskell but am useless with its stdlib
15:09:52 <ais523> and I was looking there too
15:10:34 <Deewiant> AnMaster: that handprint overflows to some 32-bit value which could theoretically be in use by another interpreter
15:10:49 <Deewiant> and no, overflow is not discussed in the specs
15:10:53 <AnMaster> Deewiant, hm true, so EFUN it is
15:11:20 -!- oerjan has joined.
15:11:51 <AnMaster> hi oerjan, do you think 0x6566756E6765202D2041204249474E554D20426566756E67652D393820696E74657270726574657220696E2045726C616E67 is a sane number to use for a unique ID of an interpreter?
15:12:20 <AnMaster> really? Deewiant doesn't agree
15:12:35 <oerjan> assuming it was picked at random
15:12:53 <AnMaster> "efunge - A BIGNUM Befunge-98 interpreter in Erlang"
15:13:15 <AnMaster> normally they are like CFUN, CCBI, RCSU (iirc) and so on
15:13:57 <AnMaster> anyway I guess I shall go with EFUN instead
15:15:14 <oerjan> ok assuming you wanted to call that from another interpreter, it would seem rather useful if that didn't have to be a bignum one
15:15:39 <AnMaster> oerjan, but agree the long one is more fun :D
15:16:00 <AnMaster> That the interpreter's handprint is 1022815351171939080750602515080902596655856329413174779991860562785344354620645109770120275823379669394014875628110704231
15:16:13 <AnMaster> That the interpreter's handprint is 1162237262
15:16:23 <AnMaster> the latter is EFUN as a 32-bit integer
15:16:51 <AnMaster> hm really I should make it a prime number :/
15:16:55 <oerjan> clearly there needs to be a handprint for requesting unlimited length handprints
15:17:14 <AnMaster> one char difference and it is a prime
15:18:55 <Deewiant> AnMaster: that wraps around to "lang" btw, heh
15:19:19 <oerjan> i mean are any of those primes
15:19:20 <AnMaster> Deewiant, oh you mean in 32-bit
15:19:22 <ais523> well my Haskell program makes it 1022815351171939080750602515080902596655856329413174779991860562785344354620645109770120275823379669394014875628110704231
15:19:23 <Deewiant> AnMaster: your too-long handprint, in 32-bit it's "lang"
15:19:27 <ais523> which is the same result you got
15:19:59 <ais523> it's a three-liner the way I wrote it, I was just a little rusty at Haskell so it took me a while
15:20:11 <AnMaster> Deewiant, anyway lang is a cool 32-bit handprint
15:20:20 <Deewiant> concatMap (flip showHex "" . ord) $ string
15:20:36 <ais523> what's special about that?
15:20:52 <Deewiant> but in retrospect I guess that's obvious
15:20:57 <Deewiant> since it's the last 4 chars of the string >_<
15:21:00 <ais523> no, I created lang from that
15:21:13 <Deewiant> ais523: it's that integer when truncated to 32 bits
15:21:19 <AnMaster> Deewiant, that depends on how signed wrapping works
15:21:26 <ais523> I thought there was something interesting about the number
15:21:49 <ais523> AnMaster: gcc's definition is interesting, it assumes signed wraps like unsigned except sometimes in the control expression of a loop
15:22:06 <ais523> or other comparison-like situations
15:22:15 <ais523> AnMaster: gcc's taking advantage of that fact, obviously, as its definition is weird
15:22:42 <AnMaster> find me a prime fingerprint in 32-bit that works for the name "efunge"
15:22:49 <AnMaster> otherwise I may just use the long one
15:23:05 <AnMaster> it shall be either prime and the name fitting
15:23:36 <AnMaster> but EFUM (which is prime) makes no sense
15:24:29 <AnMaster> but I totally fail to see how it is related
15:24:38 <ais523> AnMaster: fungus = mushroom
15:25:29 * ais523 likes the concept of e-mushrooms anyway
15:25:32 <tusho> AnMaster: what's so good about primality?
15:25:46 <ais523> have a fingerprint that's the sum of two cubes in two different ways
15:25:54 <ais523> it doesn't have to be 1729, make it much bigger
15:25:58 <ais523> then you can annoy people with ti
15:26:33 <AnMaster> ERFN is even and factors into "1163019854: 2 581509927"
15:26:34 <oerjan> well the first cannot be since it's even
15:27:39 * oerjan wonders what the perfect number fingerprints are
15:27:40 <AnMaster> ais523, huh what do you mean with cube here?
15:27:59 <ais523> AnMaster: a number that's some integer to the power 3
15:28:15 <AnMaster> two different cubes in two different ways?
15:28:25 <Deewiant> AnMaster: a^3 + b^3 and c^3 + d^3
15:28:40 <Deewiant> AnMaster: two cubes, in two ways.
15:28:45 <ais523> a number is the sum of two cubes if it's a^3+b^3
15:29:00 <ais523> and the sum of two cubes in two different ways if there are two possibilities for (a,b) that work
15:29:07 <AnMaster> anyway I got no idea if 1022815351171939080750602515080902596655856329413174779991860562785344354620645109770120275823379669394014875628110704231 is prime or not
15:29:25 <AnMaster> how many bits is that number hm?
15:29:43 <oerjan> could do the 3 and 11 tests at least
15:29:54 <oerjan> actually haskell should handle those fine
15:30:13 <ais523> AnMaster: it's divisible by 3
15:30:45 <oerjan> 6, 28, 496, 8128, 33550336, 8589869056, 137438691328
15:31:11 <ais523> oerjan: perfect numbers?
15:31:15 <AnMaster> oerjan, err hm some of those are not primes, so what are the prime factors of 1022815351171939080750602515080902596655856329413174779991860562785344354620645109770120275823379669394014875628110704231
15:31:26 <ais523> wow, that 33550336 struck me in the face
15:31:32 <ais523> I've spent all week staring at numbers that start 3355
15:31:36 <Deewiant> that's a tough one, 45 seconds already :-P
15:32:08 <ais523> 0x02000000 is 33554432
15:32:08 <oerjan> well that's a perfect synchronicity
15:32:15 <AnMaster> 10> io:format("~.16B~n", [33550336]).
15:32:19 <ais523> and it's the lowest memory address on the heap in gccbf
15:32:33 <AnMaster> 11> io:format("~.16B~n", [8589869056]).
15:32:35 <ais523> AnMaster: not really, all perfect numbers are the product of a power of 2 and 1 less than a power of 2
15:32:45 <ais523> all even perfect numbers anyway
15:32:46 <oerjan> i thought we were doing byte encoding not hex
15:33:06 <ais523> that's bad as it means that any even perfect handprint will end with lots of NULs
15:33:11 <AnMaster> is it so that all are 0x1F...0... with equal count of F and 0?
15:33:20 <AnMaster> or is that just true for those two?
15:33:26 <ais523> AnMaster: yes, I think so, past the first few
15:33:29 <AnMaster> 12> io:format("~.16B~n", [137438691328]).
15:33:44 <ais523> AnMaster: that C is half an F and half a 0
15:34:14 <Deewiant> AnMaster: {{3, 1}, {233, 1}, {463, 1}, {7309, 1}, {79843, 1}, {1196959691, 1}, {11426407403, 1}, {1516849785751, 1}, {77863659174921299, 1}, {3508614362114968649, 1}, {955524906494690973199465859798407059913, 1}}
15:34:21 <AnMaster> 13> io:format("~.2B~n", [137438691328]).
15:34:21 <AnMaster> 1111111111111111111000000000000000000
15:34:40 <AnMaster> Deewiant, how long did it take?
15:35:13 <AnMaster> I should select an 4096 bit number as fingerprint as a tribute to gpg key sizes
15:35:25 <tusho> AnMaster: why not... EFUN
15:35:35 <tusho> AnMaster: so fucking what?!
15:35:44 <AnMaster> either prime or larger than 32 bits
15:35:54 <AnMaster> tusho, because it is esoteric!
15:36:12 <tusho> AnMaster: it's pointless and not very funny either
15:38:11 <oerjan> i'd say it's a prime example of esotericness
15:39:41 <oerjan> i was trying not to say that
15:39:47 <Deewiant> evidently there's no noun for it
15:40:01 <ais523> heh, I noticed an Unlambda interp on hackage, checked the .cabal file and it seems that Oerjan here is at least partially responsible
15:40:05 <ais523> I should have guessed, really
15:40:22 <tusho> we should call esotericy peoples perpetrators
15:40:36 <ais523> tusho: that term is reserved for people who perpetrate INTERCAL compilers
15:40:43 <tusho> and they're "responsible" for languages and interps
15:40:44 <oerjan> there are several unlambda interpreters in haskell i think
15:40:51 <tusho> ais523: but going by my system, what would esolangs and implementations be?
15:41:57 <AnMaster> Deewiant, hm would it be possible to implement MVRS without implementing concurrency I wonder?
15:42:21 <tusho> so perpetrators are responsible for plans and crimes
15:42:33 <oerjan> hm esoteri(ci)sm means something slightly different
15:43:31 <AnMaster> Deewiant, anyway are you going to test efunge at some point? That project will go dormant now until Friday at least
15:43:44 <Deewiant> and possibly at some point, yes
15:44:20 <AnMaster> Deewiant, adding it would be simple, since it don't implement anything like TERM or NCRS or any other fingerprint that may need manual checking
15:44:22 <oerjan> esotericity and esotericness both give about 500 google hits
15:45:39 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | Hi..
15:45:50 <optbot> oerjan: because the mind isn't barraged by things like - say, physical constants?
15:46:18 <oerjan> oh i'd say G lays heavily on me
15:47:25 <oerjan> optbot: except i'm not sure that's the right idiom
15:51:51 <oklobol> what what coming from all them directions graaaa how can i find it in that small a heap
15:52:43 <oklobol> i mean, absolutely nothing
15:52:46 -!- oklobol has changed nick to oklopol.
15:59:32 -!- Ilari has quit ("Won't be back for a while...").
16:12:32 <oerjan> http://en.wiktionary.org/wiki/ilari
16:12:34 -!- Hiato has joined.
16:12:56 <oerjan> http://en.wiktionary.org/wiki/hiato
16:13:03 <Slereah_> Once again, they're making us program on fucking paper.
16:13:04 <ais523> oerjan: bet it doesn't have an entry for ais523
16:13:58 <oerjan> http://en.wiktionary.org/wiki/User:Ais523
16:14:01 <oklopol> what's wrong with programming in c on paper
16:14:04 <ais523> oerjan: that's cheating
16:14:20 <Slereah_> You can't really correct yourself
16:14:26 <tusho> does the search suggest tush for tusho?
16:14:36 <ais523> tusho: Tusho is a placename
16:14:37 <Slereah_> I'd rather program in pseudo code on paper
16:14:46 <tusho> ais523: is it? cool. where
16:15:00 <Hiato> heh, oerjan, where'd you pull this out from?
16:15:17 <ais523> tusho: Ethiopia, apparently
16:15:27 <ais523> http://www.maplandia.com/ethiopia/oromiya/east-harerghe/tusho/
16:15:53 <oerjan> Hiato: google and wiktionary
16:19:08 <oerjan> Hiato: does your nick mean something else?
16:19:30 <AnMaster> Deewiant, hm why haven't I seen Mike for some time
16:20:32 <oerjan> AnMaster: mike was actually another secret identity of Radovan Karadzic
16:20:40 <Hiato> oerjan: nope, I actually stole it from a similar one in worms (the original, it was Hialto)
16:21:14 <tusho> Hiato: the game worms? :P
16:22:13 <oerjan> Hialto seems to be a figure from Norse legends
16:22:55 <tusho> 1, dc, 2, wp, wa, 3d, 4, (haven't kept up with the ones after)
16:23:02 <Hiato> one of the random names "generated"
16:23:28 <tusho> i mean, which game?
16:23:37 <tusho> i think wa was the first with random names
16:24:00 <Hiato> http://www.urbandictionary.com/define.php?term=Hiato
16:24:05 <oklopol> i've only really played wp, but it was awesome
16:24:19 <tusho> oklopol: W:A is a lot better than WWP with the recent patch
16:24:34 <tusho> it's WWP-crap-bugs+smoother+features
16:26:14 <AnMaster> Deewiant, I sent off a question to Mike asking about tagged tuples for FPDP
16:26:46 <ais523> AnMaster: FPDP seems reasonably simple to me, clearly it should just use 128-bit float on a 64-bit system
16:26:50 <AnMaster> Deewiant, it could even be written using g/p, but not printed out using o
16:26:59 <AnMaster> ais523, I can't do a union of int and double in erlang!
16:27:11 <AnMaster> so I need to use something like {double,1.345}
16:27:17 <AnMaster> on the stack to represent them
16:27:18 <ais523> AnMaster: you could copy the bits by hand, I suppose?
16:27:27 <oerjan> AnMaster: erlang is dynamically typed isn't it
16:27:44 <AnMaster> but point is following FPDP specs is an issue here
16:27:50 <oerjan> so can't you type check?
16:28:10 <AnMaster> oerjan, well yes I could use is_float()
16:28:39 <AnMaster> ais523, yes erlang can do that kind of, but it is messy and I'd rather avoid it
16:28:50 <AnMaster> and also what if a float was a trap value
16:29:07 <AnMaster> efunge is aimed to be 99.9% percent portable
16:29:11 <ais523> AnMaster: what if someone runs FPDP, then stores the values in the playfield, then runs into them as a command?
16:29:19 <AnMaster> only non-portable erlang function I use so far is os:get_env()
16:29:35 <AnMaster> and as far as I can see in the docs os:get_env() works on both windows and linux
16:30:20 <Deewiant> AnMaster: you can store them however you like as long as they transparently behave like normal funge-space cells
16:31:25 <AnMaster> Deewiant, btw for efunge / is correct path separator on windows I think, since erlang use forward slash on windows and convert it internally
16:31:32 <Deewiant> AnMaster: and Mike is very active in e-mail, he just isn't here anymore
16:31:41 <AnMaster> Deewiant, I did email him about it
16:31:51 <AnMaster> Deewiant, and no they won't be able to be written out using o
16:31:58 <Deewiant> AnMaster: / is not correct on Windows IMO, no matter what.
16:32:08 <AnMaster> Deewiant, well using \ in efunge would fail on windows
16:33:06 <AnMaster> Deewiant, no but the docs seem to indicate it
16:33:23 <AnMaster> I don't have a windows install around
16:33:37 <AnMaster> Deewiant, do you have erlang on it?
16:33:40 <Deewiant> what do I have to type, I have Eshell V5.5.5 open
16:34:30 <AnMaster> file:open("C:\\SomeFileOfYours", [read])
16:34:34 <AnMaster> file:open("C:/SomeFileOfYours", [read])
16:34:39 <AnMaster> Deewiant, err add a . after each
16:35:01 -!- Slereah_ has quit (Read error: 54 (Connection reset by peer)).
16:35:04 -!- Slereah2 has joined.
16:35:14 <Deewiant> AnMaster: both give {ok,<0.number.0>} where number was 32 with / and 34 with \\
16:35:26 -!- jix has quit ("This computer has gone to sleep").
16:35:28 <AnMaster> Deewiant, q(). to quit erlang shell
16:36:16 <AnMaster> Deewiant, well there is the filename module to *convert* to native paths, meant for interacting with external tools
16:36:24 <AnMaster> but I can't find path separator in there
16:36:31 <AnMaster> only more higher level functions
16:36:41 -!- jix has joined.
16:37:05 <Deewiant> AnMaster: can't you do something manually to detect OS?
16:37:16 <AnMaster> there is the os module but it is unportable
16:37:17 <ais523> system("uname"), if it fails you're on Windows
16:37:24 <AnMaster> and erlang will always translate paths
16:37:36 <AnMaster> Deewiant, I prefer it to work on other OS like OS9 iirc
16:37:41 <Deewiant> ais523: unless it prints CYGWIN_NT-<number> like it does here :-)
16:38:09 <AnMaster> Think twice before using this function. Use the filename module if you want to inspect or build file names in a porta-
16:38:09 <AnMaster> ble way. Avoid matching on the Osname atom.
16:38:27 <AnMaster> so well I say that / is native in *erlang*
16:38:41 <AnMaster> and I define ERTS (Erlang RunTime System) to be my OS
16:38:50 <ais523> I'm beginning to be a bit inspired to write perfectly portable float -> int and int -> float convertors
16:38:54 <ais523> which copy the bit pattern using arithmetic
16:39:00 <AnMaster> ais523, err erlang only got double
16:39:16 <ais523> but the same idea I mean
16:39:35 <AnMaster> and I *think* the format of it is IEEE 7xx (whatever, can't remember the number) but not sure if implementation/OS defined
16:40:00 -!- oerjan has quit ("leaving").
16:40:10 <ais523> AnMaster: everything uses IEEE-format floating point in practice nowadays
16:40:22 <ais523> it was the easiest format to get hold of soft-float libraries for
16:40:25 <AnMaster> ais523, well something could use x87 ones
16:41:09 <AnMaster> Deewiant, no I can't find out native path separator in any way it seems
16:41:48 <Deewiant> AnMaster: how would I convert a path to native
16:42:00 <AnMaster> not without building a fake filename, translating to native format, and then looking at some specific index in the string, and that would be error-prone since some OS use a totally different format
16:42:24 <AnMaster> Deewiant, filename:nativename("string").
16:42:37 <AnMaster> Converts Path to a form accepted by the command shell and native applications on the current platform. On Win-
16:42:37 <AnMaster> dows, forward slashes is converted to backward slashes. On all platforms, the name is normalized as done by
16:42:58 <Deewiant> AnMaster: can you push / on OS 9?
16:43:04 <AnMaster> Macintosh HD:Folder:SubFolder:file
16:43:18 <AnMaster> Deewiant, point is in all cases erlang will handle it transparently from my point of view
16:43:35 <Deewiant> AnMaster: slashes would work on OS 9? O_o
16:44:12 <AnMaster> now I don't know if there is still any erlang implementation for OS 9
16:44:20 <AnMaster> I'm pretty sure one existed at some point
16:44:27 <AnMaster> Deewiant, no but erlang would convert it
16:44:38 <AnMaster> http://erlang.org/doc/man/filename.html
16:44:42 <Deewiant> AnMaster: yes, so "slashes would work on OS_9? O_o" is true
16:44:51 <AnMaster> http://erlang.org/doc/man/file.html
16:45:05 <Deewiant> AnMaster: what about if you have a file called / in OS 9
16:45:20 <Deewiant> or a file containing / in its name
16:45:44 <AnMaster> Deewiant, hm don't remember, iirc it wasn't legal
16:46:45 <fizzie> "Mac allows all except a colon (though a forward slash, /, can cause issues for POSIX paths)."
16:47:02 <AnMaster> hm didn't they change it in OS 9.1 or so
16:47:07 <fizzie> And the Internet also says that with some C compilers/runtimes on OS 9, /s work in fopen/friends.
16:47:11 <AnMaster> for forward compatiblity with OS X
16:47:36 <AnMaster> wow it was years ago I used macs
16:48:07 <AnMaster> Deewiant, anyway as far as I can tell, letting erlang handle what path separator is actually used is more portable
16:48:17 <AnMaster> that is the point of erlang runtime
16:49:12 <AnMaster> Deewiant, but I guess efunge works under windows or?
16:50:15 <AnMaster> Deewiant, to give it arguments from the shell you could use efunge:start("foo.b98", ["a", "b"]).
16:50:21 <AnMaster> Deewiant, oh you had erlang installed anyway?
16:51:58 <AnMaster> "In Windows, all functions return file names with forward slashes only, even if the arguments contain back slashes. Use join/1 to normalize a file name by removing redundant directory separators."
16:52:05 <AnMaster> from the filename module man page
16:52:55 <AnMaster> Deewiant, or what did filename:nativename() return for you?
16:53:15 <Deewiant> AnMaster: nativename("/") was "\\"
16:53:45 <AnMaster> Deewiant, and it is meant for when you want to send off the path to an external tool, as far as I can tell
16:54:14 <AnMaster> as the other functions return with forward slashes even on windows I assume that is what erlang prefer internally
16:55:27 <Deewiant> AnMaster: what I'd do is nativename(join("foo", "bar")) and get what's between foo and bar :-P
16:55:59 <AnMaster> Deewiant, I define my OS to be Erlang, I will make = evaluate erlang, won't use os:system()
16:56:14 <AnMaster> just because I don't think you could claim it was invalid :P
16:56:30 <Deewiant> unless I can manage to break it ;-)
16:56:53 <AnMaster> Deewiant, well of course you use any = for breaking commands
16:57:14 <Deewiant> what I meant was finding a case where \ would work and / doesn't
16:57:47 <ais523> AnMaster: and why would you want to do that?
16:58:09 <AnMaster> ais523, "<Deewiant> unless I can manage to break it ;-)" "<AnMaster> Deewiant, well of course you use any = for breaking commands"
16:58:13 <ais523> surely ccbi shouldn't need the -9 to kill?
16:58:19 <Deewiant> AnMaster: 2008-09-15 18:57:14 ( Deewiant) what I meant was finding a case where \ would work and / doesn't
16:58:44 <Deewiant> I don't really care enough to bother trying :-P
16:58:47 <AnMaster> anyway as efunge doesn't currently implement i or o
17:03:23 <AnMaster> hm eval in erlang looks tricky
17:04:35 <AnMaster> seems you have to call several functions on the string: erl_scan functions to tokenise it, then erl_parse functions to get a parse tree, finally erl_eval functions
17:07:14 -!- ais523_ has joined.
17:07:44 -!- ais523 has quit (Read error: 104 (Connection reset by peer)).
17:08:01 -!- ais523_ has changed nick to ais523.
17:39:36 <oklopol> i can't say i believe that anyone can actually create computers.
17:39:42 <oklopol> they most likely always were
17:40:47 <oklopol> just think about it, no one can make anything that small
17:41:13 <AnMaster> and the original computers were huge
17:41:25 <oklopol> tools must be made as well.
17:41:57 <oklopol> i believe you can make huge computers, yes
17:42:28 <oklopol> anyway, i wish there was a book about how a computer is actually made, from scratch, but not for dummies, a book that actually explains each and every detail
17:43:02 <oklopol> currently i have a massive body of knowledge about circuit design and digital logic, but i have no idea how the fuck these things are actually made
17:43:29 <ais523> oklopol: are you wondering about how things like ICs are made?
17:43:33 <oklopol> not that i care, now that i think about it
17:43:39 <ais523> mostly it's done using chemistry and light
18:16:31 <fizzie> I have a feeling the details are going to be somewhat complicated when the line width goes below 100 nm. Lithography itself is reasonably understandable, though.
18:17:57 <ais523> basically there are various light-sensitive chemicals
18:18:11 <ais523> which can be washed away in a particular acid if they've been exposed to light, but not otherwise
18:18:33 <ais523> in order to make a chip all you need is to be able to place p-type and n-type silicon in arbitrary positions in 3D
18:18:41 -!- Hiato has quit ("Leaving.").
18:19:22 <ais523> so you start out with a block of p-type (for instance), then protect it with a light-sensitive chemical, use optics to expose it to light in a certain pattern, wash away the exposed chemical, then fill the atmosphere with something that turns p-type into n-type
18:19:54 <ais523> to get a 3d pattern rather than just 2d, you continue this alternating between p-type and n-type, and doing it faster each time so the p-type/n-type flip doesn't go as deep each time
18:21:30 <ais523> Slereah2: I was replying, very late, to oklopol
18:21:35 <ais523> oh, and btw hi optbot, hi fungot
18:21:36 <fungot> ais523: it was evoli whom you introduced to begin, i think
18:21:36 <optbot> ais523: i'm blaming the world
18:47:19 -!- olsner has joined.
19:22:48 <oklopol> i guess i could believe it if it's done with light
19:23:03 <oklopol> since i can imagine ways to make light patterns smaller
19:24:02 <AnMaster> probably not visible wavelengths though
19:24:16 <ais523> I think they normally use ultraviolet, not too sure about that though
19:25:27 <ais523> I imagine they have something like a reverse microscope, to scale the pattern down to chip size
19:26:53 <oklopol> now i'd like to know details out of interest, not out of doubting the existance of computers
19:27:17 <oklopol> would be fun to have built a computer from scratch
19:27:48 <oklopol> wonder how many manhours that would take if you know exactly what to do
19:31:11 <AnMaster> ais523, is wikipedia down? en.wikipedia.org times out for me
19:31:21 <ais523> AnMaster: let me check
19:31:33 <AnMaster> ah well google cache will work for now
19:31:58 <ais523> yes, it's up, just slow
19:47:22 <ais523> tusho: anyone else who cares: I found a brilliantly appropriate format in which to distribute C-INTERCAL
19:47:49 <tusho> so it's probably perfect
19:47:51 <ais523> they removed tar from the POSIX standard
19:47:54 <ais523> and put in pax in its place
19:47:58 <tusho> ais523: what? when?
19:48:06 <Deewiant> http://en.wikipedia.org/wiki/Pax_(Unix)
19:48:12 <ais523> to address some of the deficiencies with tar
19:48:21 <ais523> it's forward-compatible so it can be read with a tar reader
19:48:36 <ais523> so nobody will actually have any trouble decoding the paxball once they know that method works
19:48:55 <ais523> the only practical difference is that the files' ctimes are recorded, in the case of C-INTERCAL
19:49:26 <ais523> although pax is a lot more versatile than that
19:49:29 <tusho> what is wrong with tar
19:49:32 <ais523> also it allows unlimited-length filenames
19:50:11 <ais523> tusho: filenames limited to 99 characters, some information about the file not recorded, GNU tar not actually conforming to the standard?
19:50:24 <ais523> at least, not by default
19:50:31 <Deewiant> AnMaster: I guess you should be using pax too, right?
19:50:32 <ais523> it can generate pax though, nowadays
19:50:42 <tusho> Deewiant: ooh, i don't think he can respond to that one
19:50:47 <ais523> Deewiant: that's the great part, everyone /ought/ to be using pax, nobody does though
19:50:47 <tusho> it fits him perfectly
19:51:05 <ais523> thus it's totally unusual, and fits the spirit of INTERCAL very nicelt
19:51:21 <Deewiant> AnMaster: see above discussion
19:51:31 <AnMaster> Deewiant, is it important? I'm kind of busy
19:51:33 <ais523> Deewiant: had you heard of pax before today?
19:51:38 <Deewiant> ais523: everybody /ought/ to be using INTERCAL, nobody does though? ;-)
19:51:50 <tusho> i think it's very important
19:51:57 <ais523> Deewiant: INTERCAL isn't a POSIX standard, neither is tar, pax is
19:52:04 <Deewiant> ais523: I think I had, hadn't paid much attention to it though if I had
19:52:17 <Deewiant> I'm fairly sure I had but I couldn't remember without a wikipedia lookup
19:52:17 <ais523> probably most people don't go around reading man 5 tar in their spare time
19:52:40 <Deewiant> ais523: just poking fun at the way it fits in the spirit of INTERCAL :-)
19:53:04 <ais523> Automake have heard of it, although it isn't the default; unfortunately they put .tar.gz/.tar.bz2/tar.lzma extensions on so as not to confuse people, but I'll fix that
19:53:15 <ais523> Deewiant: by being unlike all other programming languages is the original spirit
19:53:29 <tusho> ais523: loads of languages use automake
19:53:33 <tusho> you should write your own build system in c
19:53:34 <ais523> the modern spirit is "by being unlike all other programming languages in a way that the other programming languages can reasonably be claimed to be incorrect"
19:53:47 <ais523> tusho: it's a custom tweaked automake
19:53:58 <tusho> Deewiant: that would be quite thoroughly difficult
19:54:02 <tusho> ais523: yes, but that's more intercally
19:54:03 <ais523> Deewiant: INTERCAL needs special extensions to be able to do file handling and invoke other programs
19:54:19 <ais523> basically the whole thing would be written in extensions, which is not the point
19:54:22 <olsner> longjmp without setjmp <3
19:54:32 <tusho> LOVE AT FIRST SIGHT
19:54:59 <olsner> love at first "The program has executed a very large jump." popup
19:55:30 <ais523> the worrying thing is that under gcc it's possible to fake a jmp_buf in a way that will work to some extent on a lot of systems
19:56:18 <ais523> it goes something like {auto char c; jb[0]=__builtin_frame_pointer(0); jb[1]=(&c)-1; jb[2]=&&label; label: ;}
19:56:28 <ais523> of course you'd be mad to try anything of the sort in production code
19:56:34 <tusho> ais523: that is brilliant
19:56:38 <tusho> ais523: put that into c intercal
19:56:54 <ais523> tusho: this is the sort of knowledge you get after playing around with the internals of gcc for several weeks
19:56:58 <tusho> ais523: actually, set that as your quit message
19:57:06 <ais523> really, though, I wouldn't give it more than a 50/50 chance of working
19:57:33 <ais523> especially as I've probably messed up somewhere
19:57:39 <ais523> hmm... that -1 should be +1 on x86, I think
19:57:48 <ais523> or possibly +4, depending on alignment
19:57:57 <ais523> I should probably make c a long long to avoid problems with that
20:04:17 <olsner> hmm, segfaults for me :P
20:04:37 <olsner> and the function is apparently called __builtin_frame_address rather than pointer
20:04:48 <ais523> and I'm not at all surprised at that segfault
20:05:06 <ais523> for all I know I got the elements of that in the wrong order
20:05:14 <ais523> try swapping 1 and 2, I'm pretty sure about 0
20:05:41 <olsner> also, I'm on x86_64 :P
20:05:50 <olsner> but it does do the same when compiled with -m32
20:06:15 <ais523> hmm... maybe I'll research a corrected version that at least works on x86
20:10:15 <AnMaster> and it is horribly implementation defined, or undefined
20:10:18 <ais523> well that was meant to be portable to a whole range of processors, most likely I messed it up somewhere
20:10:24 <ais523> that was kind-of the point
20:10:36 <ais523> normally you'd just use setjmp
20:11:03 <AnMaster> ais523, normally I wouldn't mess with either setjmp or longjmp
20:11:06 <ais523> I think every single statement of that, apart from the auto char c and the null statement at the end, is undefined
20:11:17 <tusho> i love it when AnMaster nitpicks about portability, readable & maintainable code, etc. when we're messing around with stuff in #esoteric
20:11:36 <AnMaster> tusho, maybe it is some sort of meta-esotericness?
20:11:36 <ais523> well, I think it's useful to nitpick about that normally, but my code was intentionally awful
20:11:47 <tusho> ais523: of course - normally
20:11:50 <tusho> but we're in #esoteric
20:11:58 <AnMaster> and yes I see it is something you shouldn't do
20:12:04 <AnMaster> not even with lots of doxygen comments.
20:12:05 <tusho> unreadable, unmaintainable and unportable code is a holy grail
20:12:16 <tusho> see! now he's ranting about decent documentation
20:12:29 <AnMaster> tusho, you need a new sarcasm detector
20:13:10 <tusho> AnMaster: your breed of sarcasm is neither funny, has any hint of sarcasm being in it at all (seeing as most of the time it's your actual opinion backed up from previous times) and is merely annoying
20:14:44 <tusho> 'AnMaster:' <- this is a magical little thing that lets you know who I'm talking to
20:14:57 <ais523> please you two, calm down
20:15:01 <AnMaster> tusho, yes I assumed it was some weird mistab
20:15:08 <ais523> I may have to research a fixed setjmp-by-hand
20:15:15 <ais523> to quell this argument somewhat
20:15:25 <tusho> ais523: no, i've dealt with anmaster's annoyance for many months now and it's just pissing me off
20:15:33 <AnMaster> also if you think I like readable code, how do you explain this then:
20:15:35 <AnMaster> set(Fungespace, #fip{offX = OffX, offY = OffY}, {X,Y}, V) ->
20:15:35 <AnMaster> set(Fungespace, {X+OffX, Y+OffY}, V).
20:15:46 <AnMaster> I wouldn't call that readable, yet I wrote it and I'm happy with it
20:15:47 <ais523> AnMaster: that is readable compared to half the stuff I write
20:15:50 <tusho> AnMaster: except i have seen people actually find my sarcasm funny, always is exaggerated so you can tell it is sarcasm
20:15:55 <tusho> and is not generally annoying
20:16:00 <tusho> also, that code is perfectly readable
20:16:04 <tusho> it's just not imperative code.
20:17:23 <tusho> i wouldn't have a problem with this if it wasn't completely impossible to /ignore AnMaster properly seeing as there is regularly discussion with him of which at least one side is interesting
20:17:54 <ais523> you two seem to get into different sorts of discussions really
20:18:02 <ais523> maybe we need to split #esoteric into two
20:18:12 <tusho> it's hilarious that I was actually the one who invited AnMaster to #esoteic
20:18:30 <ais523> tusho: you're both credits to #esoteric; at least, /I/ enjoy both sorts of discussion
20:18:37 <tusho> AnMaster: is definitely
20:18:44 <tusho> you were in #bash asking about your brainfuck impl
20:18:58 <ais523> but tusho gets annoyed when I have long interesting discussions with AnMaster and considers my half interesting but infuriating due to the lack of context and AnMaster's half annoying
20:18:58 <tusho> I asked some stuff about it and helped a bit in /msg, invited you to #esoteric, and then a few months later you just started being incredibly annoying
20:19:02 <AnMaster> tusho, what nick were you using then?
20:20:05 <ais523> AnMaster: I'm trying to remember when it was
20:20:17 <ais523> probably it was about some Befunge implementation, most likely cfunge
20:21:17 <AnMaster> ais523, which you like right? it seems tusho hates it though
20:21:26 <ais523> AnMaster: yes, I like it
20:22:23 <AnMaster> and I actually timed it on life.bf (befunge93 game of life) and cfunge is way faster than any other funge98 I found so far, I guess a dedicated befunge93 with a static 80*25 array would be faster
20:22:51 <AnMaster> all built with -march=k8 -msse3 -O2
20:23:04 <AnMaster> ccbi I had to use Deewiant's build for though
20:29:11 <AnMaster> ais523, no responded with "trolls live in Norway"
20:29:52 <AnMaster> ais523, no timestamps but see "potar" in http://rafb.net/p/RJkQJY27.html
20:32:04 <ais523> http://hasthelargehadroncolliderdestroyedtheworldyet.com/ <-- why?
20:32:24 <oklopol> *me is xxxx. -> why are you xxxx? is actually not that improbable
20:32:41 <ais523> oklopol: ELIZA, IRC version?
20:32:41 <AnMaster> ais523, that got some java script...
20:33:02 <ais523> you can look at the source if you really want to
20:33:09 <ais523> it checks a variable to see if the world's ended
20:33:35 <ais523> but I don't see why anyone would go to the trouble of registering the domain in the first place
20:34:31 <ais523> what exactly is K-lining?
20:34:42 <oklopol> i have a friend who got a k-line on freenode by explicitly asking for one for about a week
20:34:57 <ais523> being killed by the server admins as opposed to being kickbanned from a channel?
20:35:25 <AnMaster> ais523, like a ban from the whole network
20:35:27 <ais523> AnMaster: that isn't very helpful as I don't know what either is
20:35:37 <AnMaster> k-line on freenode is whole network
20:35:46 <AnMaster> on traditional networks k-line was server-local
20:35:57 <ais523> must do something pretty bad to annoy the opers like that
20:36:22 <oklopol> in quakenet, you just have to connect 5 times from the same ip
20:36:31 <AnMaster> ais523, a staffer joined just before, I guess he (yes I know that staffer is a he) banned after he saw the spamming
20:36:52 <ais523> I thought channel kickbans were more common than server k-lines though
20:37:05 <oklopol> probably on multiple channels spamming the channel name
20:37:41 * ais523 wonders how long an ELIZAbot would last in a high-traffic channel if it tried to reply to everyone
20:38:07 <oklopol> heh, i just realized the domains page says "volimo's homepage", i smell copy paste
20:38:15 <AnMaster> ais523, sendq or recvq exceeded? or if it rate limited itself, would lag several hours
20:38:52 <ais523> AnMaster: I seriously doubt it would overflow the sendq, if it did then the same would happen to everyone else in the channel
20:38:57 <ais523> as they're all receiving the same things
20:39:24 <AnMaster> ais523, err I just didn't remember if it was send/recv from client or server point of view they were named after
20:39:58 <ais523> I know, I did /who on Freenode once by mistake
20:40:14 <ais523> it managed several screenfuls of channels before killing me, though
20:41:23 <oklopol> talking about excess floods?
20:41:43 <ais523> in particular, what would happen if you put an ELIZAbot in, say, #ubuntu, which replied to every message
20:42:08 <oklopol> with the current traffic, probably nothing
20:42:21 <ais523> ok, say it's the release day of Intrepid
20:42:26 <ais523> to guarantee high traffic
20:42:33 <olsner> my stack layout seems quite insane here, &c isn't even *near* the stack pointer
20:42:35 <oklopol> in quakenet, a full message can be sent every 7 seconds without excess flood
20:42:49 <ais523> olsner: that could explain why it isn't working, to some extent
20:43:00 <oklopol> 5 second delay gives may get you killed.
20:43:21 <oklopol> freenode lets you send a bit faster, since this is a relatively small network
20:43:28 <oklopol> well, that's no reason, more like a heuristic
20:44:23 <oklopol> you could send at least 3 responses to everything on #ubuntu without getting excess flood killed
20:44:36 <ais523> ok, so not a floodkill
20:44:49 <ais523> how long do you think it would take for the people there to get annoyed and find a chanop and kill the bot?
20:45:39 <ais523> "Are you a bot?" "Would you like me to be a bot?2
20:45:51 <ais523> that would make it obvious pretty quickly, I imagine
20:46:03 <oklopol> when i asked for a mips expert, where you here, ais523
20:46:14 <ais523> I don't think so, but I'm not a mips expert
20:46:22 <ais523> if I was I would have said GregorR was and I wasn't
20:46:34 <ais523> at least GregorR reimplemented mips in JS, which has got to be worth something
20:46:35 <oklopol> yeah that's the redirection i was given
20:46:55 <oklopol> but GregorR is not active, so i had to recheck.
20:48:02 <oklopol> i do know mips quite well now, read the most of computer organization and design last week, and it introduces mips pretty thoroughly
20:48:07 <olsner> hmm, even with the use of the proper register offsets into the jmp_buf (which are 3, 4 and 5 for bp, sp, ip), and adding an arbitrary offset to &c, it still seems to go haywire
20:48:30 <oklopol> so i don't actually need any help, i guess i'm just asking out of interest
20:48:44 <tusho> it's a parody of isitchristmas.com
20:48:50 <ais523> ah, I didn't know that
20:48:51 <tusho> which has parodies such as isxkcdshittytoday.com
20:49:00 <tusho> and isbarackobamaamuslim.com
20:49:04 <ais523> but why is a website needed to tell people if it's christmas?
20:49:08 <tusho> ais523: it's funny
20:49:24 <tusho> ais523: it's absurd, yet funny because it's so overblown
20:49:27 <tusho> ais523: all these have rss feeds
20:49:44 <ais523> tusho: how often do they get new entries? Once a year?
20:50:00 <tusho> for isitchristmas, NO or YES
20:50:02 <oklopol> isxkcdshittytoday says yes :o
20:50:02 <ais523> well, twice a year for isitchristmas
20:50:05 <tusho> for isxkcdshittytoday.com always YES
20:50:14 <tusho> http://hasthelargehadroncolliderdestroyedtheworldyet.com/ - well you can guess
20:50:25 <tusho> http://hasthelhcdestroyedtheworld.com/ <-- someone else got the idea
20:50:40 <tusho> http://www.hasthelhcdestroyedtheearth.com/
20:51:03 * ais523 almost faints in horror at the idiocy of the Internet
20:51:09 <ais523> actually, I genuinely feel a bit sick now
20:51:26 <tusho> what's wrong with you, generally you love pointless ideas carried out to a huge maximum
20:51:29 <tusho> that is, er, intercal
20:51:41 <ais523> tusho: they have to be interesting pointless ideas carried out to a huge maximum
20:51:50 <tusho> ais523: an rss feed is a pretty huge maximum
20:52:03 <tusho> I also like http://notalwaysblue.com/ (javascript required).
20:52:07 <tusho> It is not always blue.
20:52:08 <ais523> also, doing it more than once is not funny
20:52:14 <tusho> ais523: it's not one person
20:52:16 <ais523> the original is possibly a good idea
20:52:17 <AnMaster> ais523, why exactly is INTERCAL not LR(N)?
20:52:20 <tusho> they're all different people
20:52:25 <ais523> AnMaster: I wrote an essay about it
20:53:16 <ais523> http://code.eso-std.org/c-intercal/pit/tests/arrtest.doc
20:53:20 <ais523> that's actually a text file
20:53:26 <ais523> the extension is misleading the web server for some reason
20:53:27 <tusho> i wanna register http://alwaysblue.com/
20:53:37 <tusho> ais523: you know exactly the reason
20:54:18 <ais523> tusho: that site is not even usually blue
20:54:27 <tusho> ais523: no, that's the joke
20:54:30 <tusho> it focuses on the blue for no reason
20:54:33 <tusho> everything has a chance
20:54:40 <tusho> but it is 'not always blue'
20:54:58 <ais523> I didn't check the list of colours quickly enough
20:55:28 <ais523> AnMaster: anyway, quick summary:
20:55:35 <ais523> DO .1 <- ,3SUB",2SUB.1".2 (reduce required)
20:55:41 <ais523> DO .1 <- ,3SUB",2SUB.1".2~.3"".4 (shift required))
20:55:59 <ais523> and it's possible to put an arbitrary expression in place of .3
20:56:13 <ais523> however, doing this sort of thing is actually against the spec
20:56:24 <ais523> which basically says "you mustn't do anything which makes the language not LR(1)"
20:56:36 <ais523> that amused me when I read it...
20:56:59 <AnMaster> DO .1 <- ,3SUB",2SUB.1".2~.3"".4 (shift required))
20:57:15 <ais523> AnMaster: bad copy-and-paste, sorry
20:57:33 <ais523> yes, but I didn't copy the line above
20:57:34 <AnMaster> (It could be the start of this statement:
20:57:34 <AnMaster> DO .1 <- ,3SUB",2SUB.1".2~.3"".4 (shift required))
20:57:52 <ais523> it matches in the orginal but not my paste here, which is why I said "bad copy and paste"
20:58:41 <AnMaster> ais523, apart from this, what prevents it from being LALR(1)?
20:58:53 <ais523> nothing, it's just the array syntax
20:58:56 <ais523> everything else is fin
20:59:13 <AnMaster> <ais523> which basically says "you mustn't do anything which makes the language not LR(1)" <-- didn't that mean LALR then?
20:59:24 <ais523> could mean either really
20:59:32 <ais523> they express the restriction in very vague terms
20:59:43 <olsner> hmm, LALR is a subset of LR, right?
20:59:48 <ais523> it's pretty much impossible to know what it means until you try to implement INTERCAL yourself, or have someone explain it to you...
21:00:30 <AnMaster> ais523, how can " be both "sparks" and "ears"?
21:01:05 <ais523> there is no requirement to use one or the other except in the case of array indexing
21:01:18 <AnMaster> so how is there a conflict if ' is spark
21:01:22 <ais523> when you have to alternate to avoid that little bit of parsing trickiness
21:01:24 <AnMaster> then you could just see if it was ' or "
21:01:38 <ais523> AnMaster: the point is that you have to choose either ' or " to avoid the conflict
21:01:46 <ais523> whereas in most expressions you can just use " everywhere
21:02:00 <ais523> ".1~".2~".3~.4""" is a perfectly legal INTERCAL expression
21:02:44 <ais523> INTERCAL parens are nondirectional, thus causing the current notational confusion
21:02:55 <ais523> . is a sigil to denote a 16-bit variable, like $ in Perl
21:03:02 <ais523> and ~ is the select operator, which is a pain to explain
21:03:17 <Deewiant> nondirectional? ( and ) are interchangeable?
21:03:28 <ais523> Deewiant: they're both " in INTERCAL, or alternatively '
21:03:30 <AnMaster> ".1~".2~".3~.4""" <-- if they are like ( and ), there is no way you can handle that with one symbol lookahead
21:03:33 <ais523> " can only match " an ' can only match '
21:03:46 <ais523> AnMaster: there is, the trick is to see whether there's an operator or number immediately before them
21:03:47 <AnMaster> because you wouldn't know if it was a closing on a opening one on a new level
21:04:02 <ais523> that doesn't work in arrays though because they can have multiple expressions in a row
21:04:11 <Deewiant> AnMaster: 3( is invalid and +) is invalid, I guess
21:04:21 <ais523> that's what I was referring to
21:04:31 <AnMaster> Deewiant, except intercal doesn't have addition iirc
21:04:42 -!- Corun has joined.
21:04:44 <ais523> none of which are particularly standard
21:04:47 <Deewiant> whatever, you know what I meant
21:04:53 <AnMaster> ais523, select being the worst one?
21:05:01 <ais523> AnMaster: select is probably the hardest to explain
21:05:04 <ais523> it's one of the most useful though
21:05:11 <ais523> and the only one known to have been invented independently
21:05:20 <ais523> well, $ is mingle, that alternates bits in its arguments
21:05:44 <AnMaster> ais523, iirc I read that select matched some operator on some other existing system though it was unknown by the intercal inventors
21:06:02 <ais523> e.g. #5$#2 is 101$10 which is 100110 which is 38
21:06:10 <ais523> AnMaster: yes, that's from the Wikipedia article
21:06:33 <ais523> AnMaster: no, it takes two arguments
21:06:50 <ais523> it takes the first bit of the first, then the first bit of the second, then the second bit of the first, then the second bit of the second, and so on
21:06:58 <ais523> until all the bits have been copied into a new number
21:07:07 <tusho> ais523: is that all?
21:07:20 <ais523> tusho: that's what $ does, and it isn't all that powerful
21:07:36 <ais523> then there are unary AND, OR, and XOR, which work on alternate bits of their arguments
21:07:41 <ais523> s/alternate/consecutive/
21:07:58 <ais523> so it's the first bit AND the second, the second AND the third, and so on
21:08:13 <ais523> except that you have the last AND the first filling in the very top bit
21:08:26 <ais523> they're written & V ? in C-INTERCAL
21:08:37 <ais523> I'll try to explain select anyway, it's probably the most useful
21:08:45 <tusho> ais523: wait, isn't $ select
21:08:57 <tusho> ah, so select is the complex useful one
21:09:02 <tusho> well, I'll implement $
21:09:23 <AnMaster> ais523, well I seen a diagram which wasn't useful at all
21:09:33 <ais523> AnMaster: that diagram's a joke AFAICT
21:10:02 <ais523> anyway, to calculate .1 ~ .2, you work out the bitwise-and of .1 and .2, then stable-sort the bits of the result by the bits of .2
21:10:02 <AnMaster> "The program arrtest.i will not actually run, but it will compile as far as degenerating C code" <-- "degenerating"?
21:10:24 <ais523> AnMaster: INTERCALese for "generate", they tend to mess around with words like that if it doesn't cost too many extra characters
21:10:48 <AnMaster> stable-sort the bits of the result by the bits of the second operand.... um wait what?
21:10:58 <AnMaster> you mean all 0 first, then all 1?
21:11:06 <tusho> ais523: hmm, you can't implement mingle without knowing how many bits in an int, right?
21:11:16 <ais523> tusho: correct, but it's always 16 in INTERCAL
21:11:23 <AnMaster> tusho, sizeof(int) * CHAT_BIT in C
21:11:24 <ais523> well, 16 before the mingle, 32 afterwards
21:11:25 <tusho> alright, i'll just require that then
21:11:38 <AnMaster> tusho, what are you writing this for?
21:11:48 <ais523> INTERCAL has both 16- and 32-bit datatypes but you aren't allowed to mingle 32-bit values
21:12:43 <AnMaster> ais523, what sorting algorithm does ick use for select?
21:13:02 <AnMaster> I assume some custom one would actually be fastest here
21:13:13 <ais523> AnMaster: I think it's custom
21:13:26 <ais523> I can look it up, the implementation of the forward operators is one of the bits I never touched though
21:13:38 <AnMaster> could be O(n) with O(2n) for memory
21:14:02 <AnMaster> could be if one was reversed stable sort
21:14:10 <AnMaster> so you filled in from each end
21:14:24 <ais523> AnMaster: well the only possible digits in binary are 0 and 1
21:14:28 <ais523> that makes the sorting a lot faster
21:14:32 <ais523> also, 0 AND anything is 0
21:14:37 <ais523> so that can also be used to optimise the result
21:15:33 <AnMaster> is the source for the original intercal compiler around any more?
21:15:50 <ais523> AnMaster: OK, I checked, basically it just scans the second argument looking for 1s, and copies corresponding values from the first argument to the result
21:16:10 <ais523> oh, except it saves time by bitshifting the arguments around on occasion
21:16:19 <ais523> AnMaster: not as far as I know
21:16:23 <AnMaster> ais523, ah so O(n) and O(2n) for memory?
21:16:32 <ais523> however, there was an interview with the original creator
21:16:52 <ais523> anyway it turned out that they didn't implement integers as integers at all
21:17:02 <ais523> they implemented them as strings holding the binary representation of the number
21:17:06 <ais523> and manipulated them with regexen
21:17:13 <tusho> AnMaster: it was in SNOBOL, iirc
21:17:16 <tusho> so that would have been natural
21:17:24 <ais523> tusho: SPITBOL, a dialect of SNOBOL, but yes
21:17:36 <AnMaster> still that would have been slow, and very very slow on those old computers
21:18:01 <tusho> why does it matter
21:19:00 <tusho> AnMaster: because i really don't get why you're bothered by stuff like that
21:19:06 <tusho> who cares if intercal programs run fast?
21:19:09 <tusho> why is it an issue?
21:19:24 <ais523> tusho: actually part of the charm of C-INTERCAL is trying to get INTERCAL to run fast
21:19:44 <tusho> ais523: but nobody would fault it if it failed
21:19:52 <ais523> well, it does fail to some extent
21:20:00 <ais523> but it's a lot faster than CLC-INTERCAL
21:20:04 <ais523> and probably a bit faster than J-INTERCAL
21:21:06 <AnMaster> ais523, does ick use SSA internally or?
21:21:19 <ais523> AnMaster: no, it maps each INTERCAL variable to a C variable
21:28:17 <AnMaster> ais523, what about other stuff, common sub-expression elimination and such?
21:28:23 <ais523> AnMaster: when you find out how bad INTERCAL control flow is normally, you'll see why
21:28:44 <ais523> anyway, the sort of optimisations C-INTERCAL do are mostly manipulating the INTERCAL to make it more C-like
21:28:57 <ais523> there's some CSE involved, but normally that can be left to the C compiler when the code's C-like enough
21:29:40 <ais523> common subexpression elimination
21:29:55 -!- comexk has joined.
21:30:06 <AnMaster> ais523, ever thought about compiling to native code?
21:30:19 <ais523> AnMaster: a bit, if I did I'd do it as a gcc frontend
21:30:26 <ais523> but it works well compiling to C
21:31:06 <ais523> AnMaster: the intermediate formats used by C-INTERCAL and by GCC are pretty similar
21:31:50 <oklopol> ais523: i may have killed you
21:31:55 <tusho> [[ You should have received a copy of the GNU General Public License
21:31:55 <tusho> along with this program; if not, write to the Free Software
21:31:55 <tusho> Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
21:32:01 <tusho> What if the FSF move offices?
21:32:07 <tusho> Won't they be violating the GPL or something?
21:32:16 <oklopol> ais523: was watching death note, and decided to try out whether notepad has the same power
21:32:25 <AnMaster> tusho, didn't they at some point?
21:32:30 <ais523> tusho: it would certainly make for a good Slashdot story
21:32:36 <oklopol> you were the only one on #esoteric the name and face of whose i could immediately recall
21:32:41 <ais523> the GPL is no-derivs so we couldn't even change it
21:32:49 <ais523> oklopol: I'm still alive I think
21:32:59 <tusho> i believe death note has a time limit thing
21:33:02 <tusho> 24 hours or something?
21:33:04 <tusho> haven't actually watched it
21:33:07 <oklopol> you should've died already
21:33:17 <Deewiant> I haven't watched it either, read it though
21:33:17 <oklopol> i guess the microsoft dudes didn't add tat feature
21:33:19 <tusho> creepy that oklopol knows ais523's face by heart, though
21:33:26 <oklopol> 40 seconds in the tv series
21:33:26 <tusho> AnMaster: manga/anime thing.
21:33:28 <Deewiant> AnMaster: manga/anime/live-action movie
21:33:32 <AnMaster> ais523, tusho: I seen another address in older copies of GPL with different address
21:33:50 <Deewiant> oklopol: I wonder why they always change stuff like that
21:33:57 <oklopol> tusho: i know yours too, but ais523's nick was the one i saw first
21:34:04 <tusho> Deewiant: it's OVER 8000
21:34:15 <oklopol> Deewiant: in an anime the time actually *matters*
21:34:18 <Deewiant> in the manga it's not even said
21:34:24 <comexk> in the full version of even the english, he says "it's over 8000... over nine" to himself
21:34:25 <tusho> hmm. i bet apple added death note functionality to TextEdit, since they rock so much
21:34:30 <Deewiant> oklopol: it's not like it's accurate anyway
21:34:33 <AnMaster> anyway you are making no sense
21:34:43 <oklopol> there are a lot of inconsistencies
21:34:44 <AnMaster> I guess that is due to not watching manga or anime
21:34:50 <tusho> oklopol: you may die in 30 seconds of choking on your computer
21:34:57 <ais523> tusho: TextEdit is one of the few applications that can actually open OOXML
21:35:05 <ais523> however it does so by just extracting the text, I think
21:35:16 <comexk> don't you just have to unzip it like odt
21:35:30 <oklopol> death note is the only anime i've watched
21:35:31 <tusho> comexk: i use textmate :p
21:35:41 <comexk> I cracked it once though
21:35:46 <oklopol> i've tried to watch a few other, but it's impossible
21:35:48 <tusho> but, i assume the death notey code kind of has high licensing fees.
21:35:56 <tusho> so i doubt the textmate dev would have it
21:35:57 <comexk> but it's nice to know how to use a disassembler and be able to crack your own software
21:36:03 <tusho> apple probably do which is why i tried textedit
21:36:11 <tusho> oklopol is actually dead now
21:36:16 <tusho> if he says anything its just what he said before death
21:36:19 <tusho> he has like a 5 year lag
21:36:37 <Deewiant> that's fine as long as the responses are topical
21:36:42 <tusho> just so that when he is talking about his death
21:36:45 <tusho> and pleading for help
21:36:54 <tusho> and we all call up the emergency services
21:36:58 <tusho> we learn that he has been dead for 5 years
21:37:02 <tusho> DUN DUN DUNNNNNNNNNNNN
21:37:22 <oklopol> Deewiant: impossible, they simply tend to suck
21:37:41 <oklopol> well okay, lucky star was awesome :)
21:37:47 <oklopol> but i haven't dl'd that yet
21:37:57 <tusho> oklopol: how are you talking about things from the future?
21:38:14 <tusho> it is late 2003 right now for your messages that we are getting
21:38:26 <tusho> oklopol: from __future__ import anime?
21:38:30 <oklopol> import("__future__").socket
21:39:03 <Deewiant> oklopol: meh, you're just watching the wrong ones
21:39:50 <oklopol> they both had __'s, so i naturally got confused
21:42:53 <comexk> I joined so I could see your mingle thingy
21:43:04 <ais523> comexk: e pasted it to me in a PM
21:43:10 <ais523> tusho: paste it in-channel!
21:43:23 <tusho> http://pastebin.ca/1203677
21:43:28 <tusho> is bugged, the last two expressions are messed
21:43:33 <tusho> need moar nummers in the *
21:43:36 <tusho> but yeah that's the basic idea
21:43:43 <tusho> slower than c-intercal's, but easy to understand
21:43:54 <AnMaster> tusho, nice one, and unreadable
21:44:06 <tusho> you just have to know what it's trying to do
21:44:12 <AnMaster> tusho, not on a 80 char wide terminal no
21:44:21 <tusho> i could wrap it but i generated that code
21:44:29 <tusho> and that's not the point
21:44:36 <ais523> tusho: C-INTERCAL's mingle is beautifully readable
21:44:41 <ais523> if a bit non-obvious how it works
21:45:08 <tusho> well, that's the thing
21:45:14 <AnMaster> ais523, do you agree tusho' unrolled one is unreadable?
21:45:14 <tusho> mine is less readable structurally wise
21:45:17 <tusho> but is more obvious how it works
21:45:24 <tusho> AnMaster: c-intercal's is more unrolled
21:45:24 <ais523> AnMaster: yes, I agree with you
21:45:34 <tusho> if you just read the first few expressions
21:45:34 <AnMaster> he is obviously speed crazy since he unrolled the loop
21:45:38 <tusho> you get the idea of how it works
21:45:39 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | more ideas are swimming through my head than I can clearly keep track of; that generally means its night time for me. I'll have some more glypho fluff tomorrow though, to be sure!.
21:45:44 <tusho> AnMaster: no, that's stupid
21:45:58 <tusho> the point is you can look at the first few expressions and see the algorithm
21:46:04 <ais523> and C-INTERCAL's is a common-subexpression-eliminated version of mingle, which stores parts of the expressions in parts of the word
21:47:13 <AnMaster> tusho, if I had done that you would have said it was speed crazy and unreadable
21:47:18 <tusho> AnMaster: no i wouldn't have
21:47:35 <tusho> i do not judge code based on who wrote it
21:47:42 <tusho> it just so happens that yours is pretty consistent in its ugh-ness
21:48:02 <AnMaster> <tusho> no, readable <AnMaster> ais523, do you agree tusho' unrolled one is unreadable? <ais523> AnMaster: yes, I agree with you
21:48:15 <tusho> *it's not a fucking battle
21:48:25 <AnMaster> you seem to treat it that way normally
21:48:26 <tusho> i don't give a shit how many people think my code is readable, because it is and I am prepared to justify that
21:48:53 <tusho> and no, i don't know where you are getting the delusions but I have never cared what people think of your code, i attack it purely by myself and if other people agree that is irrelevant
21:48:55 <ais523> tusho: it's too wide for the screen and therefore can't be justified without word-wrapping, left-aligned is more useful for code anyway
21:49:09 <tusho> ais523: when I was hand-writing I aligned the two b and a expressions
21:49:15 <tusho> by putting the + at the end of the return on the first line
21:49:30 <ais523> tusho: I was avoiding the argument for the sake of a bad pun, sorry
21:49:41 <tusho> didn't see that pun
21:57:40 <AnMaster> http://public.web.cern.ch/public/en/Science/Antimatter-en.html <-- *groan* at the pun at the end of that
22:23:47 -!- jix has quit (Nick collision from services.).
22:23:57 -!- jix has joined.
23:37:08 -!- ais523 has quit ("9").