00:08:31 -!- megatron has changed nick to moozilla.
00:46:44 -!- tusho has quit.
01:45:22 -!- olsner has joined.
02:14:25 -!- olsner has quit ("Leaving").
03:04:44 -!- Tritonio_ has quit (Remote closed the connection).
03:04:54 -!- CO2Games has joined.
03:15:38 <CO2Games> I want to make my own brainfuck language. How would I go about getting a compiler for it?
03:20:48 <CO2Games> Or should I go for an interpretted language
03:21:18 <pikhq> Making your own Brainfuck language?
03:21:23 <pikhq> If it's Brainfuck, it's not your own.
03:22:25 <CO2Games> Well I want to modify it and make it my own
03:24:19 <CO2Games> And I want a seperate data storage area, not an array, more of a single variable I can cache things into
03:25:12 <lament> it's probably better for everyone if you write your own compiler or interpreter
03:26:01 <CO2Games> Like so I can ^<<<V and that would copy the value to the cache, then move back three, and copy the value from cache
03:26:40 <CO2Games> And I'm looking for a way to use a pointer to a specific piece on the tape
03:26:53 <lament> and the single-value cache is the only vailable storage?
03:28:00 <CO2Games> Like in assembly, you have the stack, but you also have areas you can place data, and get it later
03:28:58 <CO2Games> It just would be nice if we got that.
03:30:05 <CO2Games> I mean a tape is great but there's still a small piece of ram in the back
03:30:42 <lament> my guess is that it would make brainfuck too easy to be interesting
03:32:03 <CO2Games> like you do a bunch of stuff, no idea where you are after some loops. 0@, there I'm back at the start
03:32:37 <CO2Games> Since the left edge is not infinite on this 'tape', there would be a device to rewind to the start of the tape
03:33:14 <CO2Games> Also, my first brainfuck program >+[++++++++++>,----------]<[<]>>[.>]
03:33:51 <CO2Games> Takes input until a 10, then echos all of the data
03:35:15 <lament> have you tried running it?
03:35:29 <CO2Games> That's why this one actually works now
03:36:12 <CO2Games> Before I just had a program that went until a null pointer but I'm not exactly going to hit a NULL button on my keyboard to stop the program
03:36:52 <CO2Games> What I'm looking for is a way to force a loop to happen at least once, a do...while, rather than a while
03:37:43 <CO2Games> Yeah I know but then you use up another space
03:38:12 <CO2Games> I would much rather have >{++++++++++>,----------}<[<]>[.>]
03:40:08 <CO2Games> I guess I could work on a parser for uhh whatever I'm going to call my little brainfuck derivative
03:45:34 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | Look at it..
04:24:07 <CO2Games> Ok I've decided on having two tapes, the program code cells, and the data storage cells
04:40:34 <lament> you don't need much of a parser when every instruction is 1 char
04:41:24 <lament> brainfuck was actually designed to make it easy to compile/interpret
05:09:44 <CO2Games> Yeah but I've got a lot of commands planned
05:32:16 -!- calamari has joined.
05:35:02 -!- GreaseMonkey has joined.
06:21:41 -!- calamari has quit ("Leaving").
06:29:14 -!- oklopol has quit (Read error: 113 (No route to host)).
07:41:34 -!- oklopol has joined.
07:57:25 -!- kar8nga has joined.
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:15:41 -!- oerjan has joined.
09:26:01 -!- Judofyr has joined.
09:27:07 -!- oklopol has quit (Read error: 113 (No route to host)).
09:27:33 -!- oerjan has quit ("leaving").
09:33:38 -!- poiuy_qwert has quit.
09:45:34 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | and for location of stacks and direction of growing, I'll make that command line arguments.
09:49:07 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | how do you tell which ones are bad?.
09:56:17 -!- kar8nga has quit ("Leaving.").
10:00:22 -!- GregorR has quit (zelazny.freenode.net irc.freenode.net).
10:00:22 -!- puzzlet has quit (zelazny.freenode.net irc.freenode.net).
10:01:48 -!- puzzlet has joined.
10:11:50 -!- lifthrasiir has quit (zelazny.freenode.net irc.freenode.net).
10:11:50 -!- dbc has quit (zelazny.freenode.net irc.freenode.net).
10:11:52 -!- moozilla has quit (zelazny.freenode.net irc.freenode.net).
10:11:52 -!- rodgort has quit (zelazny.freenode.net irc.freenode.net).
10:12:06 -!- moozilla has joined.
10:12:06 -!- dbc has joined.
10:12:06 -!- lifthrasiir has joined.
10:12:06 -!- rodgort has joined.
10:13:56 -!- GregorR has joined.
10:27:04 <AnMaster> What exactly does "If an attempt is made to modify the result of a comma operator or to access it after the next sequence point, the behavior is undefined." mean?
10:27:22 <AnMaster> it is from C99 standard 6.5.17 paragraph 3
10:27:27 <AnMaster> it is from C99 standard 6.5.17 paragraph 2
10:32:26 <AnMaster> however that seems like it should be undefined to me
10:32:32 <Deewiant> probably nothing you'd ever run into in practice
10:32:40 <Deewiant> sounds awfully obscure, anyway
10:33:07 <AnMaster> Deewiant, well IOCCC probably...
10:34:36 <AnMaster> Deewiant, anyway I suspect obscure use of the comma operator is an excellent way to obfuscate C code.
10:34:55 <Deewiant> I think anything nontrivial is UB
10:35:25 <AnMaster> Deewiant, however the specs give "f(a, (t=3, t+2), c)" as a valid example
10:35:42 <AnMaster> would mean f(a, 5, c) according to the example's description
10:35:45 <Deewiant> yeah but that's fairly nontrivial ;-)
10:36:24 <AnMaster> Deewiant, you added a "non" to much I think
10:36:56 <AnMaster> Deewiant, still if I ever saw something like like that in real code I would have to read it twice then exclaim wtf :P
10:57:24 -!- Tritonio_ has joined.
11:03:38 -!- GreaseMonkey has quit ("Jesus loves you").
11:34:15 -!- AnMaster_ has joined.
11:46:47 -!- AnMaster has quit (Connection timed out).
12:04:42 -!- oklopol has joined.
12:05:13 -!- puzzlet has quit (Read error: 104 (Connection reset by peer)).
12:05:50 -!- puzzlet has joined.
12:09:17 -!- tusho has joined.
12:12:52 -!- Tritonio_ has quit ("Leaving.").
13:22:55 -!- AnMaster_ has changed nick to AnMaster.
13:24:01 -!- oklofok has joined.
13:24:21 -!- oklopol has quit (Read error: 113 (No route to host)).
13:53:36 -!- Tritonio_ has joined.
14:02:24 -!- kar8nga has joined.
14:09:05 -!- Judofyr has quit.
14:43:14 <AnMaster> can you code a compiler (in a non-TC language) for a TC language?
14:46:25 <tusho> for a lang which goes through 'a=b' in order and applies to stdin and finally outputs to stdout:
14:46:40 <tusho> well, you get my meaning
14:46:44 <tusho> brainfuck->c is just trivial string replacements
14:47:23 <AnMaster> bf->simple asm in Befunge-93 should also be possible
15:04:25 -!- Tritonio_ has quit ("Leaving.").
15:28:02 -!- kar8nga has quit (Read error: 110 (Connection timed out)).
15:37:35 <CO2Games> Alright I just finished my cache and tape classes, not sure if they work or not.
15:45:34 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | cya.
16:12:33 -!- Ilari has joined.
16:27:05 -!- CO2Games has quit ("And I invented doors, no joke!").
16:37:43 -!- Hiato has joined.
16:40:38 -!- Hiato has quit (Client Quit).
16:45:02 -!- Hiato has joined.
16:45:52 -!- oklofok has quit (Read error: 113 (No route to host)).
16:51:42 -!- Hiato has quit ("Leaving.").
16:56:24 -!- Judofyr has joined.
16:59:44 <tusho> i thought ais would be here today.
17:03:25 -!- oklopol has joined.
17:20:49 -!- Tritonio_ has joined.
17:28:35 <AnMaster> tusho, btw cfunge compiles successfully on GCC and ICC. TCC lacks a certain C99 feature currently cfunge uses but they plan to add that. PCC I couldn't even get to compile a simple hello world on my system
17:28:51 <AnMaster> compiling llvm/clang currently, though clang is experimental
17:28:55 <tusho> i'm not sure why you decided to tell me all that
17:29:05 <AnMaster> so will be fun to see if it works
17:29:09 <tusho> also, can't wait for clang to replace gcc
17:29:15 <tusho> from what i've heard it is completely awesome
17:29:29 <AnMaster> however it is still far from complete
17:29:46 <tusho> perhaps ais523 should be writing an LLVM backend instead...
17:29:47 <AnMaster> tusho, also I suspect trying to compile a kernel using llvm could cause serious issues
17:30:04 <tusho> AnMaster: well, presumably they're targeting regular apps first
17:30:21 <AnMaster> anyway GCC won't be replaced for a long time
17:30:34 <tusho> bit of a shame, it is _quite_ warty
17:30:37 <AnMaster> maybe on OS X, since Apple seems to be LLVM fans
17:30:48 <tusho> nah, os x is still very BSDy
17:31:06 <AnMaster> tusho, you ever been to #llvm on oftc?
17:31:12 <AnMaster> a lot seem to match *.apple.com
17:31:23 <tusho> yea, well, apple have done a lot with llvm
17:31:49 -!- Mony has joined.
17:31:57 <AnMaster> tusho, however, llvm is rather large
17:32:05 <AnMaster> I guess the cause is that it is written in C++
17:32:08 <tusho> there's quite a difference
17:32:13 <AnMaster> not something I would like on an embedded system
17:32:26 <AnMaster> on an embedded system I would probably try something like PCC
17:32:29 <tusho> but you can't just tweak a regular compiler for embeddedness
17:32:34 <tusho> so that's unsurprising
17:32:39 <AnMaster> tusho, PCC and/or TCC should work fine there
17:33:45 <tusho> AnMaster: does this look ok in konqueror to you? http://eso-std.org/
17:33:52 <tusho> it's the style i'm going to use for my blahhhg
17:33:58 <tusho> a screenshot would be appreciated
17:34:04 <tusho> although tbh i don't give a crap about konqueror, nobody uses it
17:34:08 <AnMaster> tusho, so you broke the real site?
17:34:19 <AnMaster> tusho, it looks better than in firefox
17:34:28 <tusho> AnMaster: the real site had nothing there
17:34:33 <tusho> except a test post
17:34:43 <tusho> with the default skin
17:34:48 <AnMaster> tusho, why is there an empty white box above the title and below the red bar?
17:35:00 <tusho> AnMaster: works in ff3, it's just that ff2 is broken
17:35:03 <tusho> that's the same bug IE has
17:35:10 <tusho> the white box is meant to stretch the whole page
17:35:13 <tusho> i.e. the posts are meant to be in the white box
17:35:19 <tusho> does konqueror not even show a white box at all?
17:35:25 <tusho> lol, i knew it was broken but dayum :)
17:35:31 <AnMaster> tusho, in ff2 all links are underlined, in konq they are underline on hover
17:35:38 <tusho> AnMaster: that's a konqueror default
17:36:01 <tusho> a screenshot in ff2&konqueror would be much appreciated, i don't have them to test with
17:36:32 <AnMaster> tusho, the red box around ESO have 90 deg corners in konq
17:36:42 <tusho> AnMaster: again, screenshot would you :P
17:36:51 <AnMaster> tusho, just will take a few minutes
17:42:28 <AnMaster> tusho, yes I said a few minutes...
17:42:53 <tusho> AnMaster: you were 6 minutes.
17:47:42 <AnMaster> ######################################################################## 100.0%
17:47:50 <tusho> AnMaster: so use another image host
17:47:52 <tusho> it's not rocket science
17:48:04 <tusho> i generally use xs.to
17:48:07 <tusho> or you could use bayimg.com
17:48:31 <tusho> AnMaster: are you going to die to upload two f*cking images via a browser
17:49:10 <tusho> AnMaster: why, exactly? What's the big deal?
17:49:32 <tusho> AnMaster: http://bayimg.com/ just upload it, sheesh
17:49:51 <tusho> what the fuck is the problem with clicking 'browse', selecting a file and clicking upload?
17:50:00 <tusho> why do you make such a goddamn fuss over having to use a GUI program?
17:50:06 <AnMaster> tusho, give me a page that doesn't time out...
17:50:17 <tusho> Both xs.to and bayimg.com fail?
17:50:26 <tusho> How big is that file?!
17:50:52 <AnMaster> tusho, hm I think my ISP got issues
17:51:54 <tusho> totally broken, both of em
17:52:12 <tusho> AnMaster: can't be arsed, it's for my personal blog
17:52:21 <tusho> works in firefox 3 & ie 6/7
17:52:28 <tusho> that'll make up, uh, all of my traffic.
17:52:47 <tusho> no reason why it shouldn't, it's based on webkit like safari
17:53:00 <tusho> the markup is very lean
17:53:03 <tusho> it should be very nice in lynx
17:53:14 <AnMaster> tusho, well it works in ff2 and konq too, just not the CSS bit
17:53:19 <AnMaster> and that doesn't work in lynx either
17:53:48 <AnMaster> tusho, doesn't work well in lynx
17:53:54 <tusho> it works fine for me
17:54:03 <AnMaster> is what it looks like at the top
17:54:16 <tusho> what is the problem...
17:54:23 <tusho> page title, the <link> elements, then the h1
17:54:26 <tusho> that's how lynx displays things,
17:54:42 <tusho> no, i have lynx here
17:54:45 <tusho> the point is it's correct.
17:54:56 <tusho> #Feed is from the <link> element to the atom feed in my <head>
17:55:01 <tusho> the second ESO is from the <h1>
17:55:05 <tusho> the first one is from the <title>
17:55:27 <tusho> AnMaster: anyway, firefox's problem there is that that white thing should stretch down to the bottom of the page
17:55:30 <tusho> and the posts should be contained within it
17:55:55 <tusho> AnMaster: exactly the same except it lacks the white box altogether...
17:56:07 <AnMaster> tusho, and got non-rounded corners for the red thing
17:56:16 <tusho> AnMaster: yea, but only firefox and webkit support those
17:56:20 <tusho> well, gecko and webkit
17:56:32 <AnMaster> tusho, so it is a non-standard CSS bit?
17:56:32 <tusho> -moz-border-radius-bottomleft: 5px;
17:56:32 <tusho> -webkit-border-bottom-left-radius: 5px; -moz-border-radius-bottomright: 5px;
17:56:32 <tusho> -webkit-border-bottom-right-radius: 5px;
17:56:36 <tusho> AnMaster: it's in CSS3
17:56:45 <tusho> but nothing implements it apart from gecko and webkit
17:56:48 <tusho> and since it isn't standard yet
17:56:53 <tusho> they use the standard CSS method for unstandard things
17:57:04 <tusho> they disagree a bit on how to do things like 'bottom left', though
17:57:20 <tusho> but yes, it's not-yet-standard CSS because css3 isn't final yet
17:57:41 <tusho> anyway, the problem with firefox 2 and konqueror is that they don't let you apply styling to unrecognized elements
17:57:44 <tusho> which is a bug bug bug bug bug bug
17:57:46 <tusho> and entirely their fault
17:57:56 <tusho> (i use HTML5 elements like <article> and such)
17:58:14 <tusho> IE6/7 has the same bug but a tiny piece of JS fixes it
17:58:15 <AnMaster> tusho, they don't claim to support HTML5 I guess
17:58:22 <tusho> AnMaster: it's not to do with that
17:58:27 <tusho> they should let you style <asdasdad>
17:58:30 <tusho> it's how css works
17:58:36 <tusho> it is an actual bug on their part
17:58:37 <AnMaster> if you used XML + your own XML Schema I guess it would work?
17:58:43 <tusho> but it's still a bug
17:58:58 <AnMaster> tusho, well add a custom XML schema for it?
17:59:11 <tusho> i don't want my pages to be xhtml, can you remember the huge rant in #ESO?
17:59:16 <tusho> i believe I have stated my cause very well
17:59:22 <tusho> and the people working on html5 100% agree with me too
17:59:41 <AnMaster> tusho, so where in the html5 file does it say anything about what version of HTML it is?
17:59:53 <tusho> that doctype is html5-specific
17:59:57 <tusho> all the pre-html5 ones look like this
18:00:04 <AnMaster> yes I know what they look like
18:00:04 <tusho> <!DOCTYPE html "-//W3C//dicks dicks butts">
18:00:11 <tusho> AnMaster: however, you're wrong
18:00:13 <AnMaster> how can you know what html version
18:00:15 <tusho> as browsers ignore the doctype
18:00:19 <tusho> apart from quirks mode vs standards mode
18:00:25 <tusho> they don't actually listen to what version you put in there
18:00:26 <AnMaster> tusho, what about future html6 version?
18:00:31 <tusho> AnMaster: <!DOCTYPE html>
18:00:43 <AnMaster> tusho, so if there were any breaking changes stuff would go pretty bad?
18:00:48 <tusho> AnMaster: it does anyway
18:00:52 <tusho> browsers ignore whatever html version you specify
18:00:56 <tusho> they use the doctype to determine:
18:01:01 <AnMaster> tusho, you mean the major ones do?
18:01:13 <tusho> 1. whether you actually have one (quirks mode vs standards mode)
18:01:19 <AnMaster> but that doesn't prevent some potential browser from caring about it
18:01:25 <tusho> AnMaster: such a browser would be useless
18:01:33 <tusho> as the number of broken webpages would be sky-high with it
18:01:47 <AnMaster> tusho, it could have a quirks mode as well
18:01:53 <tusho> the web is shit, people put shit that contradicts other shit in their shit, and as a browser maker you have to bend over, take it, and add a metric fuckton of hacks to make them work
18:01:59 <tusho> it's lame but it's true
18:02:15 <tusho> AnMaster: btw - http://annevankesteren.nl/2005/07/html5-doctype
18:24:14 <AnMaster> apart from that, which I suspect is a LLVM or clang bug, it works
18:35:38 <AnMaster> tusho, found the issue, it clang claims to be GCC by defining __GNUC__, so that exposes some inline asm in a system header... which then clang fails horribly at
18:35:53 <tusho> well, I can see why they claim to be gcc
18:35:54 <AnMaster> not even going to bother with reporting a bug
18:35:56 <tusho> tons of stuff dies out without it
18:36:20 <tusho> AnMaster: your browser sends a user agent that lies
18:36:25 <tusho> and covers up its lie with (compatible;
18:36:28 <tusho> gonna stop using that?
18:36:30 <AnMaster> tusho, yes I hate that as much
18:36:32 <tusho> it does it SO THINGS WORK
18:36:40 <tusho> AnMaster: you'd prefer it if everything was broken, huh
18:37:14 <Deewiant> really, the issue is crap that does #ifdef __GNUC__
18:37:25 <tusho> Deewiant: but he's not going to report a bug because they're DIRTY LIARS
18:37:31 <tusho> that is ridiculous
18:37:55 <tusho> evil, malicious, dirty, conniving LIARS
18:37:58 <tusho> they probably lie to everyone
18:38:09 <tusho> why would he deal with such sleazy folk
18:44:18 <oklopol> i wouldn't report a bug either
18:44:30 <tusho> oklopol: no but you wouldn't report it anyway
18:44:37 <tusho> whereas AnMaster has stated that the reason is because "they're liars"
18:44:58 <oklopol> i don't mind weird reasons
18:51:16 <AnMaster> <Deewiant> really, the issue is crap that does #ifdef __GNUC__
18:52:02 <AnMaster> Deewiant, well yes, ICC does it too, but at least ICC 1) doesn't fail at it 2) got an option to turn that off (-no-gcc). Though I would prefer if it defaulted to off rather than on
18:52:05 <Deewiant> yes, so file a bug with glibc :-)
18:52:18 <Deewiant> AnMaster: well, as for 1) clang is beta
18:52:53 <AnMaster> Deewiant, indeed. but they want me to try to create a minimal test case not including any system headers.. well as far as I can see it is a preprocessor issue
18:52:58 <Deewiant> and as for preference, again, too much stupid stuff does #ifdef __GNUC__
18:53:22 <pikhq> It's really irritating to define __GNUC__ if you don't implement GNU C... Stuff that dies without __GNUC__ defined doesn't *deserve* to compile.
18:54:19 <pikhq> I thought glibc was fairly explicitly designed to not need __GNUC__.
18:54:34 <Deewiant> "look at Tiny C Compiler development list for workaround they had to put up with to use glibc headers. (TCC refuses to define __GNUC__, but what a pain it causes.)"
18:54:55 <pikhq> As far as the Linux kernel goes, yeah; why, exactly, *do* they decide to only work with GCC?
18:55:27 <Deewiant> probably because they decided to use C for speed, which makes sense, but C is such a crap language that they want to use the GNU extensions, which also makes sense. :-P
18:55:59 <pikhq> Glibc has absolutely no excuse.
18:56:23 <pikhq> The GNU coding standards fairly explicitly require cross-UNIX compatibility.
18:56:24 <tusho> pikhq: exactly so it's better to support it
18:56:33 <tusho> except that it makes them DIRTY SLEAZY LIARS
18:57:47 <Deewiant> hmm, no compilers I have can optimize "switch (foo) { case bar: stuff; break; default; assert (0); }" to "stuff;" :-(
18:59:42 -!- KingOfKarlsruhe has joined.
18:59:54 <AnMaster> <pikhq> It's really irritating to define __GNUC__ if you don't implement GNU C... Stuff that dies without __GNUC__ defined doesn't *deserve* to compile.
19:00:24 <AnMaster> however it is a lot worse that compilers die when they define __GNUC__ and then get some weird stuff from system headers
19:00:45 <AnMaster> the compiler authors only got themselves to blame
19:01:10 <Deewiant> of course they're going to do something about it given that it's /glibc/
19:01:11 <AnMaster> Deewiant, as for TCC, it works almost fine on cfunge
19:01:23 <tusho> Deewiant: and they'd get it done faster
19:01:23 <AnMaster> there seems to be one single C99 construct that cfunge uses that TCC can't handle
19:01:25 <tusho> if AnMaster reported a bug.
19:01:29 <tusho> because they're LIARS
19:01:46 <tusho> AnMaster: ah, so you like sleazy liars now
19:01:48 <AnMaster> since they ask you to use their mailing list
19:02:04 <tusho> i was talking to Deewiant
19:02:15 <tusho> you win the reading comprehension prize
19:02:19 <AnMaster> <tusho> AnMaster: ah, so you like sleazy liars now
19:02:30 <AnMaster> or why did you mention my nick
19:02:35 <tusho> AnMaster: please read properly
19:02:39 <tusho> but I was talking to Deewiant
19:03:03 <tusho> and how they'd fix the bug quicker if it were to be, you know, reported.
19:03:11 <AnMaster> tusho, and well it is not because they are liars, but well they are too. but because they only got themselves to blame
19:03:11 <tusho> but you won't, because they're "liars"
19:03:15 <AnMaster> if they claim to be another compiler
19:03:26 <tusho> AnMaster: so tell them and let them fucking fix it
19:03:37 <tusho> or is it COMPILER EAT COMPILER EVERY COMPILER FOR HIMSELF
19:06:34 -!- Tritonio_ has quit ("Leaving.").
19:11:06 -!- Tritonio_ has joined.
19:16:45 -!- GreyKnight has joined.
19:17:53 <GreyKnight> I got a memo from some randomer asking me to come back here, I guess the addition of ancient logs to the /topic has something to do with it :-)
19:32:19 -!- CO2Games has joined.
19:32:45 <pikhq> Now, if I knew who you were, that'd be great. :p
19:35:59 <optbot> Mony: !"BF!"100[1-$][['<]['>]['+]['-]['.][',][91][93]8CHOOSE;!,]#
19:40:56 <tusho> he's a guy from a september 06 log i read a while back!
19:41:15 <tusho> unless someone else is memoing people demanding their return
19:41:17 <tusho> which would be strange
19:41:30 <tusho> and pikhq, i seem to remember you talking to him in the log :P
19:42:11 <tusho> pikhq: here http://tunes.org/~nef/logs/esoteric/06.09.02
19:42:15 <GreyKnight> <tusho> [GreyKnight] is cool. <-- wooo critical acclaim
19:42:33 <GreyKnight> how is the wacky world of esoterica these days, anyway
19:42:34 <tusho> GreyKnight: rated 5 out of 5 by irc magazine!
19:42:41 <tusho> and ... pretty much the same, i think.
19:43:44 <pikhq> Odd; I think I was around in '06.
19:43:57 <pikhq> Did I join in '05 or '06?
19:44:02 <tusho> pikhq: http://tunes.org/~nef/logs/esoteric/06.09.02 has you in it
19:44:11 <tusho> and is the log i read with GreyKnight in it :P
19:44:18 <tusho> GreyKnight: oh, as for why I saw that particular log
19:44:20 <pikhq> Guess I just forgot Sir Grey.
19:44:28 <tusho> our bot here (say hi optbot)
19:44:28 <optbot> tusho: I might add <ii (reminiscent of html's "ISINDEX") for a form with just a text field, a submit button and optionally a prompt :P
19:44:39 <tusho> - spews out random quotes from the entire backlog of esoteric
19:44:41 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | Sure it does..
19:44:45 <tusho> and keeps our topic interesting.
19:44:53 <tusho> I saw a quote that interested me, so I grepped for it and found that log.
19:44:58 <tusho> and ended up reading it. ~fin~
19:46:06 <AnMaster> I think I seen the nick before somewhere
19:46:31 <tusho> AnMaster: you really are good at asking a question that we've just discussed and answered...
19:46:35 <tusho> he is from this log http://tunes.org/~nef/logs/esoteric/06.09.02
19:46:39 <tusho> well http://tunes.org/~nef/logs/esoteric/06.09.01 and http://tunes.org/~nef/logs/esoteric/06.09.02
19:46:55 <AnMaster> tusho, I'm on a PDA atm, connecting to the bnc, I don't have a lot of scrollback...
19:47:04 <tusho> AnMaster: you only need like 10 lines of scrollbar
19:47:30 <AnMaster> tusho, issue is I just connected at "<tusho> GreyKnight: oh, as for why I saw that particular log"
19:49:19 <AnMaster> ok off again, should be back to normal computer soon
19:51:58 <CO2Games> Agh I can't get my brainfuck interpreter to work
19:53:06 <AnMaster> by the way I recently came up with a new (I think) way of optimising bf
19:53:26 <AnMaster> after a [-] you can know you hit 0. Right?
19:53:56 <GreyKnight> I see bf is still popular these days, then
19:53:59 <AnMaster> now if that isn't directly followed by any unbalanced loops or such
19:54:20 <CO2Games> Actually I'm not sure [-] would do that in all cases
19:54:21 <AnMaster> you could optimise away [-]>>++++<<+ to set *one* instead
19:54:34 <tusho> GreyKnight: it shall always be...
19:54:46 <tusho> AnMaster: trivial optimization
19:54:47 <tusho> everything does it :P
19:55:01 <AnMaster> in fact apart from in unbalanced loops (as in different > and < count) you could optimise about everything to set constants or add fixed numbers
19:55:12 <CO2Games> I have a problem with the brainfuck interpreter I'm writing
19:55:20 <AnMaster> and you could do loop renumbering
19:55:26 <AnMaster> even partly in unbalanced loops
19:56:37 <CO2Games> Oh I wrote my first brainfuck program
19:56:58 <AnMaster> I seen it done for balanced loops
19:57:14 <tusho> tons of stuff does that
19:57:19 <tusho> see bf4.c or whatever
19:57:21 <tusho> by that mazonka guy
19:57:42 <AnMaster> tusho, my other idea: Befunge: you could probably JIT it
19:58:02 <AnMaster> at least not by any reasonable way
19:58:08 <tusho> see befunge wiki page
19:58:34 <AnMaster> tusho, can't find the word JIT there?
19:58:43 <tusho> there is more to reading than ctrl-f
19:59:02 <AnMaster> "The bf2c compiler included with the standard Befunge-93 distribution uses threaded code: each instruction is compiled to a snippet of C code, and control flows through the snippets just as it does in a Befunge interpreter (that is, conditionally on the value of some 'direction' register.) This does not result in a significant advantage over a good interpreter."
19:59:27 <AnMaster> also "Note that the bf2c compiler is not correct since it does not handle p correctly, but it would not be impossible to make it do so (although the C language might not be well-suited for this.)"
19:59:50 <tusho> AnMaster: how about actually reading
19:59:57 <tusho> and who cares about the link
20:00:00 <tusho> it's still been done
20:00:24 <AnMaster> and it is not JIT really for B98
20:00:24 <tusho> AnMaster: yes, evidently they are lying to suppress your ideas.
20:00:34 <tusho> the hard bit is g and p.
20:00:36 <tusho> and the directions
20:00:41 <tusho> both 93 and 98 have that
20:00:50 <AnMaster> tusho, betty only handles cardinal directions
20:00:59 <tusho> so? it's not a hard extension
20:01:11 <AnMaster> because in befunge 98 you can set delta to, say, 5,3 or whatever
20:02:37 <AnMaster> well pre-compiling lines may not be that useful then
20:02:52 <tusho> so pre-compile instructions
20:02:55 <AnMaster> tusho, anyway the real challenge would be actually compiling it
20:03:13 <tusho> just compile the instruction and a switch based on the direction
20:03:15 <AnMaster> yes, to self modifying asm I guess
20:03:31 <AnMaster> tusho, and if the instruction changes? using p
20:03:37 <AnMaster> then you need to self modify it
20:03:41 <tusho> then recompile it and mov it into place
20:03:45 <tusho> betty did this, obviously
20:03:48 <tusho> since its the only way
20:04:00 <AnMaster> tusho, it could use some JIT framework to do it...
20:05:03 <AnMaster> anyway it should be possible to make some patological thing like <bytes for longest instruction><jump handler>, to actually use a 25*80*sizeof(celloverhead)
20:06:15 <GreyKnight> I think what we *really* need is some novel languages :-p
20:06:18 <AnMaster> which would basically be place for longest instructions (fill rest with nops) + a direction handler
20:06:30 <AnMaster> GreyKnight, like ais523's Feather?
20:07:08 <AnMaster> GreyKnight, considering the *grammar* is self modifiable using time jumps
20:07:16 <AnMaster> so you can rewrite it to be a different language
20:07:30 <AnMaster> which would cause a time jump and reinterpreting the whole thing iirc
20:08:02 <AnMaster> GreyKnight, well no links, since ais523 have been unable to write up specs or implementation so far
20:08:09 <AnMaster> haven't yet manage to figure out all the details
20:08:22 <AnMaster> oh btw he is the C-INTERCAL maintainer so I'm sure he will manage it :P
20:09:31 <AnMaster> GreyKnight, the fact that it involve time jumps make it extra interesting
20:09:53 <GreyKnight> rewinding the execution timeline and so on?
20:10:15 <AnMaster> Only two implementations have it
20:10:36 <Deewiant> and they implement it differently and neither implementer knows who's wrong and about what ;-)
20:10:41 <AnMaster> RC/Funge (author invented it), CCBI (author implemented it and found RC/Funge's one didn't work)
20:11:07 <Deewiant> and since then, RC/Funge fixed its, but differently to CCBI, and now they're different in an unknown way :-P
20:11:12 <AnMaster> and that is why I won't implement it in cfunge
20:11:15 <GreyKnight> good, "complex" keeps us in business ;-)
20:11:34 <AnMaster> GreyKnight, http://www.rcfunge98.com/rcfunge_manual.html#TRDS
20:12:12 <AnMaster> GreyKnight, hm.. worked a second ago...
20:12:28 <Deewiant> welp, google cache probably has it
20:12:34 <tusho> oh and Deewiant develops ccbi
20:12:39 <tusho> since nobody seems to have mentioned that :P
20:13:06 <tusho> and dbc is daniel b cristofani, but i think he was here in '06 too
20:13:19 <AnMaster> http://64.233.183.104/search?q=cache:gFspwyCbPjkJ:www.elf-emulation.com/funge/rcfunge_manual.html+rc/funge&hl=en&ct=clnk&cd=2
20:13:59 <AnMaster> Deewiant, lets just hope it doesn't last long now
20:14:26 <Deewiant> yay, I just got NULL to work in CCBI 2
20:14:41 <Deewiant> I guess that's a good point to stop hacking for today
20:14:58 <AnMaster> Deewiant, you know NULL and ROMA are easy :P
20:15:26 <Deewiant> of course the actual work there was writing the fingerprints framework, not NULL :-P
20:15:46 <tusho> wait. Deewiant is doing work?
20:15:56 <AnMaster> tusho, his posix_fadvice version ;)
20:16:08 <tusho> Deewiant: why are you doing work man
20:16:10 <Deewiant> AnMaster: no, only NULL now :-P
20:16:21 <AnMaster> Deewiant, so loading more than one fingerprint doesn't work yet?
20:16:32 <Deewiant> AnMaster: no, I've just only implemented NULL
20:16:41 <Deewiant> and figured I'd stop for today
20:16:48 <Deewiant> not like it'd take long to get ROMA, MODU, etc
20:17:17 <AnMaster> Deewiant, cathedral style development? Or finally going open bazaar style? ;)
20:18:11 <Deewiant> can't be bothered to generate dev snapshots and I'm not sure it makes sense to run a static HTML mercurial repo
20:18:32 <Deewiant> (since I don't have the option of running a proper mercurial server)
20:18:36 <tusho> every piece of bugging zealotism becomes a joke if you add /me runs after it
20:18:43 <tusho> AnMaster: KILL JEWS
20:18:43 <AnMaster> Deewiant, what about going to, say, darcs?
20:18:49 <Deewiant> (and since I don't want to run on hgweb or some other site which ain't mine)
20:18:59 <Deewiant> AnMaster: say, because darcs sucks :-P
20:19:26 <Deewiant> maybe if I looked into it properly I'd like it too
20:19:30 <AnMaster> I like some of it's features certainly
20:20:05 <AnMaster> very useful when you want to backport some changes to a stable branch for example
20:20:23 <Deewiant> it's just a matter of your workflow
20:20:45 <AnMaster> Deewiant, hm? I need to back/forward port bugfixes randomly
20:20:46 <Deewiant> with git, I guess the idea is that you make a new local branch (or what are they called) for each individual change
20:20:59 -!- tusho has quit ("And then-").
20:21:04 <AnMaster> Deewiant, don't think that will help for that
20:21:07 <Deewiant> and then if you need to pull a change to a stable branch you just pull from the change's branch
20:21:09 <AnMaster> since they would have diverged
20:21:16 -!- tusho has joined.
20:21:24 <AnMaster> and you would need to pull in other changes from said branch that came from the main branch
20:21:24 <Ilari> Deewiant: Yup. Topic branches. One branch per change (where change can span multiple commits)...
20:21:27 <Deewiant> well, of course in the cherry-pick case you also have to merge them :-P
20:21:53 <AnMaster> Ilari, where would you branch it from?
20:21:55 <Deewiant> I don't see the advantage of cherry-pick in that case, you need to merge the changes in any case
20:22:08 <AnMaster> for a bug fix that needs to go into both
20:22:15 <Deewiant> I guess it's smarter to branch from stable in that case
20:22:16 <Ilari> Nitpiking, merge the branch, pulling branch impiles also fetching it...
20:22:59 <AnMaster> well wouldn't merging the changes from that topic branch require any changes missing from the target of the merge, and that came from previous branches
20:23:33 <AnMaster> rev 70 (trunk): branch of bug fix branch xxx
20:23:40 <AnMaster> find you need to merge it into stable too
20:23:53 <Ilari> Deewiant: At least git.git itself commits more critical bugfixes to mainintance branch and then merges mainintance branch into development (and then merges development branch into less stable development branch).
20:23:55 <AnMaster> wouldn't that require merging any changes between 50 and 70? If not
20:24:03 <AnMaster> If not then it is cherry picking
20:24:42 <AnMaster> maybe you didn't know the bug fix was needed for stable
20:25:14 <AnMaster> or maybe some other developer made it against trunk, and didn't know about stable needed it, he not being a stable maintainer
20:25:36 <AnMaster> Ilari, so tell me how that works out
20:25:37 <Deewiant> AnMaster: what about cherry-picking a bugfix change from trunk, how is it different
20:25:48 <AnMaster> <AnMaster> wouldn't that require merging any changes between 50 and 70? If not
20:25:49 <AnMaster> <AnMaster> If not then it is cherry picking
20:25:49 <AnMaster> <AnMaster> if it needs that: useless
20:26:04 <Deewiant> it's not cherry picking in that case
20:26:04 <AnMaster> thing is, the version control system should be able to keep track of it
20:26:40 <Deewiant> cherry picking is, take these X changes from branch foo and put them in branch bar
20:26:59 <Deewiant> with topic branches, you've just manually put some of the X changes in another branch
20:27:11 <Deewiant> and then it's merge from branch baz to bar
20:27:31 <AnMaster> yes... but merge would require merging any other changes from that branch
20:27:32 <Deewiant> or that's my understanding of it anyway, Ilari might be able to correct if I get something wrong
20:27:41 <AnMaster> <AnMaster> rev 50: branch off stable
20:27:41 <AnMaster> <AnMaster> rev 70 (trunk): branch of bug fix branch xxx
20:27:43 <Deewiant> AnMaster: and of course, merge does that, just like a normal merge
20:27:57 <AnMaster> Deewiant, so that means changes from *trunk* between 50 and 70
20:28:10 <AnMaster> if you can skip them, then it is cherry picking
20:28:21 <AnMaster> because they are part of the feature branch's history
20:29:19 <Deewiant> I'm not sure if you can skip them, I haven't done much branching myself let alone messing with git enough to know if it has an option for this :-)
20:29:31 <Deewiant> but even if you can, I wouldn't call it cherry picking
20:29:35 <Deewiant> a subset of cherry picking, maybe
20:29:50 <Ilari> If you notice that the bugfix is also needed for stable, maybe rebase the bugfix branch?
20:30:23 <AnMaster> that is why I prefer cherry picking
20:32:19 <Deewiant> I guess you're screwed in that case if you've already published the bugfix branch
20:32:29 <Deewiant> but I'm not sure if that's something you should be doing in any case
20:34:22 <AnMaster> well if you another developer did it, and you, the maintainer notice it
20:34:41 <AnMaster> but the original developer (who doesn't have commit access to the official trunk)
20:35:01 <AnMaster> Deewiant, "shouldn't be doing" still?
20:35:31 <Deewiant> like said, I don't know if topic branches should be published
20:36:10 <Deewiant> if not, the original developer shouldn't have and thus hasn't done it, thus he can rebase it and it can be applied to stable
20:36:19 <AnMaster> how else should a non-core developer, say Mr. Random Programmer using your project who wants to help contribute a fix. And you notice it applies to stable too, which Mr Random Programmer doesn't use
20:36:46 <Deewiant> he makes a topic branch for the change
20:36:52 <AnMaster> Deewiant, yes, and don't care about stable
20:36:56 <Deewiant> he pulls the changes from the topic branch into his main branch
20:37:04 <AnMaster> I certainly would only look at the part I use
20:37:07 <Deewiant> he pushes or somebody pulls into the devel branch
20:37:16 <Deewiant> you ask him to rebase his topic branch
20:37:31 <Deewiant> you pull from his topic branch into the stable branch
20:37:51 <AnMaster> well, I think cherrypicking is a cleaner approach
20:38:00 <AnMaster> and that doesn't prevent the invariant of commits
20:38:09 <Deewiant> like said it's a matter of your workflow
20:38:11 <AnMaster> IMO commits can not be changed once done
20:38:18 <Deewiant> in some sense I think cherrypicking is uglier
20:38:26 <Deewiant> since you don't get any separation of changesets
20:38:32 <AnMaster> rebase implies that your commits are mutable
20:38:34 <Deewiant> with topic branches, each branch is its own unit
20:38:47 <AnMaster> Deewiant, topic branches and cherry picking then
20:39:22 <AnMaster> it goes against the core principle
20:39:23 <Deewiant> I think making private history mutable is fine
20:39:35 <GreyKnight> wow, according to the wiki, the 2006 Esolang Contest is *still* going :-D
20:39:43 <GreyKnight> that shouldn't amuse me as much as it does
20:41:38 -!- vendetta has joined.
20:41:40 <AnMaster> anyway my workflow isn't topic branches really
20:42:00 <AnMaster> not unless they are big changes
20:42:03 <GreyKnight> I remember it from the last time I was here :-)
20:42:18 <AnMaster> I guess everyone just forget it?
20:42:19 <Deewiant> AnMaster: yeah, I think it's a git thing
20:42:22 <pikhq> GreyKnight: That's because judging has yet to finish. ;p
20:42:36 <pikhq> One of the judges hasn't been seen for months.
20:42:47 <pikhq> When did Sukoshi last show up? February?
20:43:02 <AnMaster> Deewiant, I make topic branches when I change major core stuff. Like when I last rewrote the entire funge-space code
20:43:21 <AnMaster> had to change the whole idea behind it
20:43:28 <GreyKnight> pikhq: if he can make it through until January then it will be a three-year contest, maybe? :-)
20:43:37 <AnMaster> actually probably May or even April
20:43:45 -!- vendetta has quit (Remote closed the connection).
20:45:20 -!- olsner has joined.
20:53:01 <oklopol> i think i should make an entry for the competition.
20:54:57 <CO2Games> My brainfuck interpreter works
20:55:08 <tusho> GreyKnight: you made that mistake in 2006
20:55:15 <tusho> when you discussed the shorthand english
20:56:01 <GreyKnight> well, most IRC clients don't have CTCP GENDER, so I just use "he" as a generic pronoun like normal
20:56:26 <GreyKnight> it's rarely important in discussions about esoteric programming what the gender of your co-conversationalist is anyway :-P
21:02:01 <AnMaster> CO2Games, also have you triest lostking under it?
21:02:08 <AnMaster> because if that doesn't work it is probably buggy
21:03:01 <CO2Games> And how many commands does it contain
21:03:32 <tusho> CO2Games: it's 2mb big
21:04:05 <CO2Games> How many command characters does it contain?
21:04:38 <CO2Games> so how many characters is that
21:06:37 <AnMaster> CO2Games, does it use several levels of nested loops?
21:07:11 <CO2Games> Yeah it doesn't support nested loops yet...
21:07:23 <AnMaster> http://jonripley.com/i-fiction/games/LostKingdomBF.html
21:07:43 <AnMaster> CO2Games, and you say you don't know how much 2 MB's of chars are?
21:07:50 <pikhq> Since it doesn't support nested loops, it's not turing complete.
21:08:01 <AnMaster> if you did, you would have known
21:08:12 <tusho> AnMaster: You gain +5 Holier Than Thou
21:08:18 <tusho> Ding-ding-ding-ding! Level up.
21:08:31 <pikhq> Most PEBBLE-compiled code wouldn't work in it.
21:08:42 <pikhq> CO2Games: I think it just *barely* works in the 30,000 cell limit.
21:09:10 <AnMaster> pikhq, in fact everything except trivial programs will need nested loops
21:09:43 <AnMaster> pikhq, oh btw how goes Def-BF?
21:10:20 <CO2Games> 1,024,000 code pieces is too much for my interpreter
21:10:32 <pikhq> AnMaster: No work on Def-BF. Busy.
21:10:59 <CO2Games> I use a static array to ensure that I can always reach a place via pointers
21:11:35 <pikhq> CO2Games: What language are you writing this in?
21:12:03 <AnMaster> I mean you got to support large programs
21:12:12 <CO2Games> I have a system for jumps and absolute pointer-movements
21:12:19 <pikhq> It's not hard in C++; just store the large program in a vector...
21:12:36 <CO2Games> It works as an array and I don't want to break that
21:12:50 <AnMaster> CO2Games, any interpreter or compiler that can't handle lostkingdom is not worth the files it was written in ;)
21:12:56 <AnMaster> really, it is the ultimate test
21:13:10 <AnMaster> everyone have to succeed at it
21:13:21 <AnMaster> CO2Games, my first was in *bash* think of that
21:13:48 <AnMaster> admittedly it took about 5 minutes to *parse* lostkingdom
21:13:49 <CO2Games> I can add support by changing the shorts to longs
21:13:57 <AnMaster> and that was even using awk to optimise it at string leve
21:14:09 <GreyKnight> CO2Games: good for you, now you can design your second: the rewrite of the first :-D
21:14:22 <AnMaster> CO2Games, um all cells are 8 bits
21:14:41 <CO2Games> Mine is signed shorts that don't wrap
21:14:45 <tusho> AnMaster: not defined
21:14:47 <AnMaster> CO2Games, if the memory cells aren't then it won't run lostking
21:15:04 <AnMaster> well lostkingdom won't work then
21:15:09 <pikhq> CO2Games, make them unsigned chars.
21:16:03 <CO2Games> It has more than just a pointer thinh
21:16:16 <AnMaster> CO2Games, it Def-BF is meant to be able to write a kernel in
21:16:27 <AnMaster> the high level code got functions
21:16:39 <AnMaster> CO2Games, I wrote a low level implementation
21:16:44 <AnMaster> to handle the pre-processed code
21:16:45 <GreyKnight> the real question is why "lostkingdom" relies on unspecified behaviour ;-o
21:16:50 <AnMaster> no one ever wrote the high level part yet
21:17:03 <AnMaster> GreyKnight, because everyone does it that way?
21:17:09 <GreyKnight> CO2Games: You should invent a nice new one! BF clones are ten-a-penny.
21:17:25 <AnMaster> GreyKnight, oh yes there are lots
21:17:29 <CO2Games> Nah I want something like brainfuck
21:17:35 <pikhq> GreyKnight: It's unspecified but de-facto behavior.
21:17:39 <CO2Games> I'm keeping the standard functions, mostly.
21:17:51 <CO2Games> I'm adding a static variable you can access
21:18:05 <AnMaster> GreyKnight, and apart from Def-BF and Boolfuck, they are just really stuff like "map ook to >, eek to >" or whatever (oh that variant is called "Ook!" iirc)
21:18:23 <CO2Games> Like, to copy a cell, ^ then move over to the destination cell and V
21:18:28 <pikhq> AnMaster: And Dimensifuck?
21:18:37 <AnMaster> pikhq, haven't heard of that one I think?
21:18:43 <AnMaster> or wait is that the several tapes one?
21:18:43 <pikhq> It's one I devised.
21:18:53 <pikhq> Not the one with several tapes.
21:18:55 <AnMaster> then it is one of the actually interesting ones
21:19:00 <pikhq> The code itself is in N dimensions.
21:19:02 <AnMaster> pikhq, oh how does it differ then?
21:19:18 <AnMaster> pikhq, that's called befunge/trefunge/n-funge :P
21:19:29 <CO2Games> In fact lost kingdom won't run on drainfuck because it has to be prepended with drainfuck:: at the start of the source
21:19:37 -!- kar8nga has joined.
21:19:37 <pikhq> No, though it's vaguely inspired by the funges.
21:19:49 <pikhq> 5 dimensional "Hello, world!".
21:19:55 <AnMaster> drainfuck:: <-- huh. why double : ?
21:20:35 <AnMaster> CO2Games, well try it with unsigned char to make sure lostkingdom works, so you get all the details like nested loops and so on
21:20:52 <CO2Games> No no, I'm going to stay with this
21:21:19 <AnMaster> CO2Games, k, so the nested loops probably are broken
21:21:19 <CO2Games> You see, the cell count and the maximum value per cell is only one off
21:21:34 <AnMaster> lostkingdom is the best test suite for bf there is
21:21:58 <CO2Games> It doesn't follow all of brainfuck's standards
21:21:59 <AnMaster> CO2Games, doesn't mean the same bugs couldn't show up
21:22:10 <pikhq> http://esolangs.org/wiki/Dimensifuck
21:22:32 <AnMaster> pikhq, you should make an opengl based editor with sound effects for it ;)
21:22:33 <CO2Games> http://esolangs.org/wiki/Drainfuck
21:23:15 <CO2Games> So far I have E, ^, and V coded, but none of them are tested
21:23:31 <AnMaster> pikhq, oh yes there is one for some semi-funge one btw. BeQunge or something like that, B93 + extra dimensions + not limited to 25x80. And buggy
21:23:46 <tusho> Slereah will love this
21:23:50 <tusho> Melab is one of the esco crew
21:23:55 * tusho goes and removes all their links again
21:24:14 <tusho> AnMaster: an esowiki user who added tons of unfinished languages and weird categories, etc
21:24:20 <tusho> and has now added back all the esco links into pages
21:25:28 <CO2Games> So lost kingdom uses wrapping?
21:25:57 <AnMaster> tusho, esco, sounds familiar, iirc silly idea + crappy implementation
21:26:13 <tusho> AnMaster: it tries to implement all esoteric languages in one shitty c++ program
21:26:24 <AnMaster> CO2Games, as in integer overflow
21:26:26 <tusho> they are unintentionally hilarious
21:26:32 <tusho> They say: "Byter is a language for training your brain"
21:26:37 <tusho> "Byter is a language for training brains"
21:27:24 <tusho> but yea they added links to themselves to every esolang article they support
21:27:29 <tusho> so i get to have fun with REMOVING THEM ALL!
21:27:33 <tusho> already done, actually
21:28:32 <dbc> CO2 you probably don't mean "next" and "previous".
21:28:48 <CO2Games> Yeah that's what the interpreter does atm.
21:29:07 <CO2Games> Wasn't thinking about nested stuff
21:29:35 <CO2Games> But now that I think about it...ooops.
21:30:40 <CO2Games> I need to check for the other markers and increment a counter if it encounters one of itself, decrementing when it meets an opposite and only when the counter is 0 should it accept an opposite as its own
21:31:01 <CO2Games> Also, lost kingdom would probably be shit if I got it to work on drainfuck
21:31:18 <CO2Games> drainfuck doesn't 'stream' the code right out of the file
21:31:26 <CO2Games> It parses it into the code tape
21:31:47 <CO2Games> The interpreter uses three tapes: code, data, and inactive
21:32:01 <GreyKnight> and the "fsck" joke got old decades ago anyway :-P
21:32:27 <tusho> CO2Games: most good interps parse
21:32:31 <CO2Games> X swaps the data and inactive tapes, L copies the data over the inactive
21:33:11 <CO2Games> So lost kingdom would take 4 megs of ram just for the code
21:33:46 <CO2Games> And that's more than my program allocates
21:34:25 <CO2Games> my program allocates a short for each positive value in a signed short
21:34:32 <dbc> The matching method you describe is comparatively slow and complicated.
21:34:42 <GreyKnight> maybe you could swap sections of tape in and out of memory to tempfiles depending on your current position? Would probably take a rewrite though.
21:35:01 <GreyKnight> listen to me, trying to give advice, don't even know this language :-D
21:35:26 <CO2Games> well, the tapes are all classes so I could essentially just give them a save and load command
21:35:33 <CO2Games> Thus you could have a save state for the programs
21:35:44 <CO2Games> Although they would all be of equal size
21:36:12 <CO2Games> Since the memory allocated is always the same assuming an error didn't happen that I didn't catch
21:36:29 <dbc> Anyway, arrays are quicker than tapes :)
21:36:44 <GreyKnight> I was talking about swapping out *sections* of tape though, to solve your "program's tape too big to have all at once" problem. Maybe I am talking nonsense, check with an expert :-)
21:37:07 <pikhq> You could just have a dynamic array...
21:37:17 <pikhq> vector<char> is not all that hard to write.
21:37:55 <CO2Games> also they are shorts, not single bytes
21:38:11 <AnMaster> <GreyKnight> "brainfsck" <-- that's just silly
21:38:20 <AnMaster> static code analyser for brainfuck!
21:38:44 <AnMaster> that is obviously what brainfsck would be
21:38:59 <CO2Games> Where did you guys here brainfsck
21:39:02 <GreyKnight> just what bf-land needs, more punnery :-P
21:39:18 <AnMaster> CO2Games, try #include <humor>
21:39:19 <dbc> Why signed?
21:39:38 <CO2Games> It doesn't wrap around anyways
21:39:41 <GreyKnight> AnMaster, try #include <relaxation> // :-)
21:40:00 <AnMaster> GregorR, well I'm not a C++ coder like CO2Games so I leave the .h in :P
21:40:08 <dbc> (doesn't wrap why?)
21:40:11 <AnMaster> I was just trying to talk to CO2Games in a way he knew
21:40:45 <dbc> in C++ it would take extra work to make a short NOT wrap, right?
21:41:09 <dbc> Like, you have to actually check right before you increment it whether it's at the top of its range, and then make a problem if it is?
21:41:37 <CO2Games> if it is it just lands at the top and doesn't bounce off or anything
21:41:42 <dbc> Lot of work to achieve a loss of functionality, unless the point of the program is to check whether a brainfuck program uses wrapping.
21:41:47 <CO2Games> or bottom if going the other way
21:41:58 <AnMaster> dbc, I think CO2Games is thinking of wrapping space rather than wrapping numbers
21:42:01 <CO2Games> Didn't think shorts already wrapped around
21:42:07 <AnMaster> it is the only sense his comments would make sense
21:42:16 <AnMaster> CO2Games, wrapping *signed* numbers is undefined
21:42:27 <AnMaster> wrapping unsigned numbers happen automatically
21:42:39 <dbc> AnMaster nah...because that wouldn't have anything to do with signed vs unsigned.
21:42:43 <AnMaster> if you overflow a unsigned number it will wrap of course
21:42:56 <AnMaster> but if you overflow a signed one... anything could happen
21:43:19 <AnMaster> for example you could hit a trap representation
21:43:21 <pikhq> AnMaster: That's mostly because the representation of signed numbers is undefined; 2's complement will, of course, wrap a bit differently from sign bit.
21:43:44 <AnMaster> pikhq, um ever heard of trap representation?
21:43:52 <dbc> Is there a computer in common use where overflowing or underflowing a signed number does something other than wrap (and set an overflow flag for other things to look at and react to as they like)?
21:44:06 <AnMaster> pikhq, a certain value may be reserved for "can't be used"
21:44:35 <AnMaster> pikhq, normally 0x8..... (that is all the rest are zeros) for int iirc
21:44:41 <pikhq> I actually think x86 sets an overflow flag. (correct me if I'm wrong)
21:44:57 <AnMaster> pikhq, x86 doesn't have a trap representation
21:45:08 <AnMaster> nor does any *common* platforms as far as I know
21:45:08 <pikhq> Overflow flag, not trap representation.
21:45:22 <AnMaster> pikhq, and yes it sets a carry flag iirc
21:45:31 <AnMaster> but you can't reach that from inside C
21:45:35 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | >.<.
21:45:42 <dbc> pikhq yeah, that's what I thought.
21:45:48 <pikhq> Unless one uses inline assembly. :p
21:46:13 <GreyKnight> basically the moral of the story is "don't let signeds overflow" :-)
21:46:48 <AnMaster> GreyKnight, potentially really bad, but mostly it works on normal computers
21:46:53 <pikhq> Especially not if you have to write code on something I may have designed.
21:46:55 <GreyKnight> AnMaster: I literally just got through fixing a bunch of someone else's code of that exact problem
21:47:01 <AnMaster> but yes you invoke undefined behaviour if you let signed overflow
21:47:12 * pikhq has been really tempted to use 0x0 for the syscall gate in a kernel
21:47:22 -!- KingOfKarlsruhe has quit (Remote closed the connection).
21:47:56 <AnMaster> pikhq, I was rather thinking for that C -> Lisp CPU compiler
21:48:14 <AnMaster> you know, Symbolics Lisp computers or something like that
21:48:29 <AnMaster> oh and their signed had bad issues when overflowing
21:48:58 <dbc> Anyway. In this "Drainfuck" there doesn't appear to be any need for negative numbers unless it's to address cells below cell 0, which I'm guessing don't exist...they haven't been mentioned, anyway.
21:49:33 <dbc> And therefore there's no obvious benefit to using signed cells.
21:49:46 <AnMaster> http://en.wikipedia.org/wiki/Lisp_machine
21:52:42 <dbc> Good number.
21:53:01 <CO2Games> Because it's the maximum a short can point to?
21:53:32 <dbc> It's part of the series 0, 1, 2, 4, 16, 65536...
21:53:56 <dbc> which are my favorite numbers.
21:54:24 <tusho> 65536 is a pretty bitchin' number
21:54:46 <dbc> (the one after that is too long to memorize)
21:55:45 <dbc> (and the one after THAT is too long to store anywhere)
21:57:29 <CO2Games> 1 2 4 8 16 32 64 128 256 512 1024 2048 4096 8192 16384 32768 65536
21:57:51 <oklopol> 1 5 5 7 7 3 8 23 98 124 8 8 345 2 2 947
21:58:30 <CO2Games> Now my long series of numbers simplified
21:58:31 <GreyKnight> is it Random Integer Sequence Day already?
21:58:43 <AnMaster> <dbc> (and the one after THAT is too long to store anywhere)
21:59:38 <dbc> f(n)=2^(f(n-1)).
22:00:01 <tusho> also known as "the powers of two"
22:00:05 <oklopol> 1 5 5 7 7 3 8 23 98 124 8 8 345 2 2 947
22:01:01 <CO2Games> 1 10 100 1000 10000 100000 1000000 10000000 100000000 1000000000 10000000000 100000000000 1000000000000
22:01:51 <AnMaster> dbc, you mean: f(1) -> 1; f(N) -> math:pow(2,f(N-1)).
22:02:39 <dbc> Yeah, but start with f(0) -> 0 instead.
22:02:53 <oklopol> 1 5 5 7 7 3 8 23 98 124 8 8 345 2 2 947
22:04:10 <AnMaster> 4> X = fun(0) -> 0; (N) -> math:pow(2,f(N-1)) end.
22:05:06 <dbc> Figured. fun vs f.
22:05:25 <AnMaster> dbc, need to ask in #erlang now
22:06:20 <AnMaster> dbc, the normal syntax would work in a compiled erlang module htough
22:07:20 <AnMaster> <ttyerl> Y = fun(0, X) -> 0; (N, X) -> X(N-1) end.
22:08:08 <AnMaster> 11> X = fun(0, Y) -> 0; (N, Y) -> math:pow(2, Y(N-1, Y)) end.
22:08:20 <GreyKnight> CO2Games: the typical life of a program of this sort is that eventually the author rewrites it from scratch at some point. ;-) It is actually a good thing because, having written one once, he has now figured out how to do it even better!
22:09:20 <AnMaster> GreyKnight, unless he or she wrote it in a functional language, in a modular way, and with good design
22:09:27 <AnMaster> then he or she should easily fix it
22:09:39 <tusho> always throw one away.
22:09:45 <tusho> even in functional languages
22:09:49 <tusho> it is a basic principle of software design
22:10:01 <tusho> write one knowing that you will throw it away, then throw it away and write another
22:10:03 <AnMaster> tusho, well yes, you throw the first interpreter away
22:10:12 <tusho> otherwise you get bloated, overengineered programs
22:10:15 <GreyKnight> AnMaster: actually, in any language, you can often get some improvements. Great modularity means you can throw bits of it away at a time, is all ;-)
22:10:18 <tusho> not that AnMaster would know anything about *those*...
22:10:26 <AnMaster> tusho, I already done it that way :)
22:10:36 <dbc> I've been trying to think what's the best way to match brackets in self-modifying variants of brainfuck. It can be done for constant cost added to increment and decrement (for the code "tape")...
22:10:45 <AnMaster> I replaced large parts of cfunge several times
22:10:47 <tusho> actually, writing a program with a kafka book as guidance would be lulzy
22:10:52 <AnMaster> since it is coded in modular C
22:11:02 <GreyKnight> (of course, in order to learn how to *do* good design in the first place usually involves a lot of throwing away on prior projects)
22:12:35 <AnMaster> dbc, only issue with that erlang code uses math:pow which returns floating point values
22:12:40 <AnMaster> so it can't compute the next value
22:12:48 <AnMaster> normal erlang numbers are bignums
22:13:08 <AnMaster> and I don't feel like writing my own pow
22:13:45 <AnMaster> dbc, actually math:pow/1 to be specific above ;)
22:14:08 <AnMaster> (oh yes, erlang got strange notation for referring to functions, it is name/<count of parameters>)
22:14:12 <dbc> (I used gnu bignum library to get the next number a long time ago, so I don't need such a program myself)
22:14:22 <AnMaster> (if parameter count differs then it is different functions)
22:15:35 -!- puzzlet has quit (Remote closed the connection).
22:15:43 -!- puzzlet has joined.
22:16:09 <dbc> CO2, how do you modify the code tape anyway? Or do you?
22:16:26 <AnMaster> Deewiant, I got a question about conforming funges in exceptional circumstances!
22:17:04 <AnMaster> Deewiant, for example, what if the programming language the funge is written in *can't* return a specific exit code to the shell. Or if the language will always print the return value of the function before exiting?
22:17:27 <AnMaster> Deewiant, does that prevent implementing Befunge-98 in such a language? or does it count as a special circumstance
22:17:40 <AnMaster> after all it is actually the shell of the interpreter of the interpreter that prints it out
22:17:47 <AnMaster> not the befunge interpreter itself
22:17:53 <AnMaster> so it is really not a part of it
22:18:03 <AnMaster> but a part of the shell (erlang shell or whatever) it runs in
22:18:52 <fizzie> That's not so strange: Prolog uses "predicate/arity" notation too.
22:19:12 <AnMaster> fizzie, anyway what do you think about the return value and stuff issue?
22:19:29 <AnMaster> the befunge interpreter won't run from command line either
22:19:45 <AnMaster> Erlang (BEAM) emulator version 5.6.3 [source] [64-bit] [async-threads:0] [hipe] [kernel-poll:false]
22:20:00 <AnMaster> then enter efunge:run("path/to/program.b98").
22:20:01 <oklopol> http://www.vjn.fi/pb/p554424121.txt
22:20:14 <fizzie> I haven't been really following all this, only just arrived back here.
22:20:27 <AnMaster> fizzie, question was just about 5 lines above
22:21:44 <AnMaster> oklopol, anyway you can't call that on 7
22:23:28 <fizzie> Personally I wouldn't really care about the "q" exit code missing if there's a reasonable reason.
22:24:00 -!- GreyKnight has quit ("zzzzz").
22:24:11 <oklopol> xrange(10)=[0,1,2,3,4,5,6,7,8,9], yes
22:24:46 <AnMaster> fizzie, well it wouldn't be missing. since everything has a return value in erlang, and the shell prints that out when the function returns
22:25:49 <fizzie> Well, that's certainly good enough.
22:26:09 <AnMaster> I got a befunge93 version here
22:27:24 <fizzie> No, really, I got the point; if your interpreter is a function under some sort of "eval"-type situation, I think you can and should substitute the "exit code" of the spec with "return value".
22:27:31 -!- Mony has quit ("À vaincre sans péril on triomphe sans gloire...").
22:27:41 <AnMaster> fizzie, http://rafb.net/p/NJr9IR26.html
22:27:50 <AnMaster> fizzie, the code is compiled, to bytecode
22:27:58 <AnMaster> which is then run under the erlang vm
22:28:08 <AnMaster> also I figured out how to do FPDP here
22:28:23 <AnMaster> you can't do the union style trick
22:28:38 <AnMaster> you can make the stack, a list, contain other stuff than integers
22:29:22 <AnMaster> now if your normal routines pop that, then you just let to code see some dummy value
22:29:33 <AnMaster> apart from of course keeping track of the metadata still
22:29:39 <AnMaster> so it is handled if pushed back
22:29:46 <AnMaster> fizzie, think that would work? :D
22:30:50 <AnMaster> fizzie, I plan to make it befunge 98 and then implement MVRS
22:31:05 <fizzie> With a non-homogenous container like a list that can store anything, sure, I don't see why you couldn't stick type-tagged tuples in there for the uncommon case.
22:31:31 <AnMaster> fizzie, yes, just need to handle it gracefully if the befunge program tries to do something strange with it
22:31:49 <fizzie> Does FPDP even define what happens then?
22:32:04 <AnMaster> I will probably make it reflect
22:32:43 <AnMaster> erlang doesn't have single/double
22:33:31 <AnMaster> I got no idea what type of float
22:34:15 <AnMaster> fizzie, anyway this befunge-93 is already turing complete, it got bignums
22:34:42 <fizzie> Google says Erlang uses IEEE 754 64-bit floats, i.e. doubles.
22:34:43 <AnMaster> shows [] as a prompt all the time
22:36:06 <fizzie> I have a piece of Perl for writing MATLAB's binary ".mat" file format which stores the desired MATLAB type of the value by "bless"ing (Perl's curious class system here) it to a class name indicating the type.
22:37:11 <fizzie> If I were inclined to write a Perl-based Funge thing, which I don't, I guess I would do FPDP and friends by sticking in the stack a reference to a scalar blessed to Funge98::Float or something. It would be appropriately perverse.
22:38:13 -!- kar8nga has quit (Read error: 110 (Connection timed out)).
22:42:58 <tusho> yay, http://eso-std.org/ has been all improved (note to people seeing it the first time: not the design that will be used for ESO's site but just for my personal blog)
22:43:01 <tusho> lots of little tweaks
22:44:28 <tusho> AnMaster: screenshot? :P
22:44:31 <fizzie> Incidentally.. is FPDP usally done by using two consecutive stack cells or something? I think I've seen talk like that here, although I haven't really been reading very carefully. If so, does it actually say so anywhere in the spec? Is there any spec other than the RC/Funge manual?
22:44:49 <AnMaster> tusho, sure, once I recover from the brain damage
22:45:03 <tusho> brain damage causing site
22:45:45 <tusho> AnMaster: firefox 3 doesn't crash
22:46:00 <AnMaster> tusho, just by having 3 lines of tabs in general
22:46:13 <tusho> psht i have like 300 tabs open at once
22:47:38 -!- Judofyr has quit.
22:48:13 <fizzie> At work approximately every ~100th time I try to plot something -- especially if it's a "pcolor" style spectrogram plot -- it goes and kills my X server. It is very very very very annoying, and I haven't really figured out a really good workaround for it.
22:48:32 <fizzie> The previous sentence was supposed to have the word MATLAB in there somewhere.
22:48:37 <fizzie> I'm not quite sure where.
22:48:45 <fizzie> Anyway, MATLAB's the "it" there.
22:48:48 <tusho> https://browsershots.org/screenshots/0269e9f697ee09b438699e1bb2ed4501/
22:48:53 <tusho> but totally not my fault
22:49:04 <tusho> just the same "elements it don't know can't be styled"
22:49:24 <fizzie> (I don't usually use FF3 for plotting.)
22:50:33 <AnMaster> tusho, http://omploader.org/vcTJ2
22:50:48 <tusho> AnMaster: gimme a sec
22:50:56 <AnMaster> tusho, well, just consider that design bad
22:51:05 <tusho> AnMaster: no, just consider that browser broken
22:51:23 <tusho> the people I know use firefox 3 and IE 7
22:51:29 <tusho> i don't really give a shit about anything els
22:51:45 <tusho> my code is totally valid & 100% semantic
22:51:50 <AnMaster> with that attitude, no more testing from me
22:51:53 <tusho> it works in the two browsers that will actually visit my blog
22:51:57 <tusho> and that's the end ofi t
22:52:17 <tusho> AnMaster: i didn't actually needed your testing
22:52:22 <tusho> but you were nearby so i told you to
22:52:27 <tusho> i could have asked someone else
22:54:42 <tusho> anyway, yay, all i need now is tusho.net
22:55:16 <fizzie> Valid and valid; the doctype doesn't even say what sort of HTML it is.
22:56:39 <tusho> currently a draft.
22:56:44 <tusho> http://validator.nu/ is an html 5 validator
22:56:57 <tusho> and <!DOCTYPE html> is the actual html 5 doctype
22:57:25 <tusho> its not actually valid html5
23:00:43 <tusho> http://validator.nu/?doc=http%3A%2F%2Feso-std.org%2F
23:04:55 <fizzie> For some reason that whole "doctype is just trickery to trigger a rendering mode in browsers" thing makes me feel vaguely uneasy. I do understand the rationale, it's just an emotional thing.
23:05:08 <tusho> fizzie: html 5 is pragmatic
23:05:23 <tusho> doctypes today ARE just for triggering rendering modes, and remembering them is hell anyway
23:05:40 <tusho> so the html 5 guys discovered a doctype that makes stuff go into standards mode and has no version and is easy to remember
23:05:43 <tusho> <!DOCTYPE html>, voila.
23:06:44 <fizzie> Yes, I did say I understood the rationale.
23:07:15 <AnMaster> I still think <!DOCTYPE html 5> or so would have been saner
23:07:24 <tusho> because that's not a valid doctype
23:07:25 <fizzie> I think I'm just being contradictionary for no reason; I like the Perl pragmatism, but for some reason I don't like it in HTML.
23:07:30 <tusho> and also, versioning is dead
23:07:46 <AnMaster> tusho, what will happen to xthml in the future
23:07:54 <tusho> AnMaster: HTML 5 can be expressed as XML
23:07:56 <tusho> and is called XHTML 5
23:08:02 <AnMaster> tusho, that was not my question
23:08:59 <tusho> AnMaster: xhtml 2.0 will probably be finalized in 10 years and a grand total of 3 people will use it
23:09:09 <tusho> xhtml 1.1 will continue to be used by a fringe
23:09:22 <tusho> AnMaster: xhtml 2.0 has absolutely no steam behind it
23:09:30 <tusho> and is losing pace compared to a snail
23:09:33 <AnMaster> tusho, ok. xhtml 1.1 still does
23:09:37 <tusho> AnMaster: not much.
23:09:43 <AnMaster> and that was the [citation needed]
23:09:46 <tusho> a fringe of geeks use it ... that's it
23:09:58 <tusho> AnMaster: you go out and find me some xhtml 1.1 websites
23:10:22 <AnMaster> sure, http://rage.kuonet.org/~anmaster/cfunge/
23:10:31 <tusho> AnMaster: ok, that's one, and it's fringe
23:10:38 <tusho> it's a nerd's site for an esoteric language interpreter
23:10:53 <tusho> fizzie: nerd's tiny site with fringe info.
23:11:27 <AnMaster> tusho, what about xhtml 1.0 then?
23:11:41 <tusho> AnMaster: it's used pretty widely and will probably continue to
23:11:53 <tusho> prediction: html 5 will slowly, but steadily gain steam
23:11:57 <AnMaster> tusho, http://www.w3.org/TR/xhtml11/
23:12:05 <tusho> AnMaster: zomg!11121212
23:12:24 <tusho> but yeah, xhtml 1.1 isn't really very exciting and i doubt it'll see a huge influx of sites
23:12:31 <tusho> but html 5 is getting quite the buzz out there
23:12:44 <tusho> obviously growth will be sloooooow but i don't think it'll stop for a while
23:12:53 <tusho> AnMaster: geeks are the ones who make webpages
23:13:01 <fizzie> http://xkcd.net/ is also XHTML 1.1, and it's at least popular, even if it's a "nerd" site.
23:13:14 <tusho> I haven't seen xhtml 1.1 on a site targeted at non-geeks, though.
23:13:39 <tusho> AnMaster: you're misinterpreting my point and turning it around so you look like an idiot again
23:13:48 <AnMaster> tusho, show me *one* single site using html5, apart from your own and one that is actually advocating html5
23:13:53 <AnMaster> so blog about html5 doesn't count
23:14:06 <tusho> http://intertwingly.net/, http://diveintomark.org/
23:14:13 <tusho> of course all of them are nerd sites now
23:14:28 <tusho> but the more it's specified and talked about, the more it's supported, the more non-geek sites will switch over
23:14:53 <tusho> intertwingly.net is XHTML 5
23:16:22 <CO2Games> dbc: I intend to add code-tape modification commands
23:16:42 <CO2Games> Also I can't get my thingy to work now
23:18:19 <AnMaster> tusho, all lighttpd's auto generated directory indexes are XHTML 1.1
23:18:29 <AnMaster> so that means a lot of sites probably
23:18:32 <tusho> AnMaster: that hardly counts as a site ;)
23:18:41 <AnMaster> tusho, counts once for every site :P
23:19:04 <tusho> and <UPPERCASE TAGS>
23:19:18 <tusho> AnMaster: that code probably hasn't been touched since 1995
23:19:24 <AnMaster> tusho, I hope html 5 forbids uppercase tags!
23:19:51 <CO2Games> why would it be forbidden anyways
23:20:52 <CO2Games> reguardless, I can't get my interpreter to work
23:21:18 <CO2Games> drainfuck::>+[++++++++++>,----------]<[<]>>[.>]E
23:21:43 <CO2Games> It gets lost after the first loop ends
23:21:54 <CO2Games> Both with and without nested loop support
23:22:09 <CO2Games> So it wasn't the addition of the nested loop support that broke it
23:28:36 <fizzie> www.cpan.org is XHTML 1.1 and it's also sort-of for non-"geeks", since it's pretty much targeted to all Perl programmers. I'm having trouble thinking of "non-geek" sites given that the places I visit...
23:39:13 <tusho> fizzie: is it served application/xml+xhtml
23:39:17 <tusho> if not, it's not xhtml 1.1
23:39:24 <tusho> it specifically disallows sending it as text/html
23:42:45 -!- oklopol has quit (Read error: 104 (Connection reset by peer)).
23:43:00 -!- oklopol has joined.
23:43:10 <fizzie> tusho: Could I have a citation on that? The XHTML 1.1 spec I have says "XHTML 1.1 documents SHOULD be labeled with -- application/xhtml+xml"
23:43:22 <tusho> fizzie: it's somewhere in the spec
23:44:03 <fizzie> I don't think it can be, if they already say SHOULD there.
23:44:14 <tusho> it's in some appendix of the xhtml 1.0 spec, then
23:44:18 <tusho> but it definitely is somewhere
23:44:27 <fizzie> That's not the XHTML 1.1 spec.
23:44:48 <tusho> http://www.w3.org/TR/xhtml-media-types/#summary
23:45:14 <tusho> even if it is valid, browsers treat it as invalid html 4
23:45:15 <tusho> http://hixie.ch/advocacy/xhtml
23:45:35 <fizzie> SHOULD NOT does not equal "specifically disallows".
23:46:22 <fizzie> On the other hand, cpan.org has their doctype public ID wrong, I think.
23:47:14 <tusho> AnMaster: because it's a fucking war
23:47:20 <tusho> jesus christ you are so childish, everything is about beating or winning
23:47:29 <tusho> i'm having what is referred to as a DISCUSSION
23:47:31 <AnMaster> tusho, hope you learn the difference between should and must in standards from now on
23:47:47 <tusho> AnMaster: go fuck yourself, i misremembered the standard, so fucking what
23:47:47 <AnMaster> tusho, and I'm just adapting to your style
23:47:55 <tusho> AnMaster: you're just trolling actually
23:48:13 <tusho> "adapting to someone's supposed style to irritate them" = "trolling"
23:48:30 <fizzie> I have no idea how to ask my Firefox "what do you think about this page?", but at least page info says "Render Mode: Standards compliance mode". I'm a bit disappointed that it doesn't say what standards those are.
23:48:31 <AnMaster> tusho, I mean adapt as in do the same as you do
23:48:47 <tusho> AnMaster: see, that is plainly trolling
23:49:00 <tusho> "Aww! I'm doing the same thing you do! Doesn't it piss you off? Haha!"
23:49:13 <tusho> 1. you are not doing what i do, you are exaggerating what you think I do
23:49:14 <AnMaster> tusho, iirc "When in Rome do like the Romans" is an English proverb?
23:49:25 <tusho> AnMaster: and also a bullshit one
23:49:55 <tusho> AnMaster: you're just trying to piss me off by acting like an exaggerated version of what you see me as, which is completely unrelated to how i am
23:50:01 <tusho> that is pretty much the definition of trolling.
23:50:06 <AnMaster> adapting to the local culture is better than trying some imperialist style and imposing on other
23:50:15 <AnMaster> as the English did with their colonies
23:50:24 <AnMaster> of course that is unrelated to modern time
23:50:56 <tusho> AnMaster: you're portraying yourself as some wonderful person who adapts to people to fit in
23:51:06 <tusho> you're just acting annoying to piss me off, and don't try and fucking sugar-coat it
23:52:03 <AnMaster> tusho, I'd say it was rather successful if your interpretation was correct. and sadly rather a failure if you respect the truth that I just said about it
23:52:18 <tusho> yes, yes it has pissed me off. congratu-fucking-lations, you are able to piss people off.
23:52:20 <tusho> what a wonderful powre.
23:52:25 <tusho> pissing people off is _easy_.
23:52:48 <tusho> AnMaster: now you may _NEVER_ accuse me of being a troll again, lest I accuse you of being an utter hypocrite
23:52:58 <AnMaster> tusho, ooh a super power? Captain piss off? The piss-off man?
23:53:02 <tusho> i hope that it was worth giving that up, since you seem to do it very often.
23:53:18 <AnMaster> tusho, of course I am a hypocrite, like you are
23:53:40 <AnMaster> tusho, and you still troll more than I do
23:54:00 <tusho> AnMaster: just shut the fuck up. you're being immature, trolling, being hypocritical, and shouting random accusations without bias. i am going now, i dearly hope you are less of an idiot/asshole in the morning
23:54:01 -!- tusho has quit.