00:12:10 back desu yo 00:13:08 someone should make a programming language that weeaboos could use 00:13:16 every statement has to end in "desu yo" 00:13:34 even better if it has to be in hiragana 00:14:32 x = 2 + 2 ですよ。 00:16:32 using a function as a predict necessarily requires the syntax ですか。 00:17:11 foo = bar が quux ですか。 00:17:40 i think it'll catch on 00:24:07 oh, no no 00:24:13 instead of x = 2 + 2 desu yo. 00:24:15 it has to be 00:24:20 x wa 2 + 2 desu yo 00:24:50 abd foo wa bar ga quux desu ka 00:24:56 or something like that 00:25:27 foo wa bar ga quux da ka n' desu yo 00:27:28 isn't japanese stack-based? 00:27:29 or whatever 00:28:10 most languages (the ones with parentheses at least (that I'm aware of)) are stack-based 00:28:11 no 00:28:20 but it is head-final 00:28:21 maybe 00:28:21 in a very vague sense 00:29:05 i mean, most formal models of japanese syntax make it almost completely, if not entirely, head final 00:29:30 but i wouldnt say that head-final-ness = stack-based 00:30:17 RodgerTheGreat: you could have been more subtl 00:30:17 e 00:30:18 ;) 00:30:50 * RodgerTheGreat shrugs 00:32:40 tusho, why do you ask about japanese being stack-based? 00:32:53 psygnisfive: just that it wouldn't be very hard to make it into a programming language 00:32:54 I guess 00:32:59 i mean 00:33:01 it would, obviously 00:33:03 but. I dunno 00:33:11 heh. it would it more ways than you know :) 00:33:35 yeah I figured that like 0.1 seconds after saying it 00:34:00 tho it would be easier than english i guess 00:38:48 in japanese, conditionals can be done with conjugations on verbs or similar 00:38:50 looking like 00:39:00 tara 00:39:21 or nara 00:39:29 we could use nara. 00:39:41 blah blah blah nara blah blah blah. 00:39:54 or as the japanese would say... 00:39:59 nani nnani nara nani nani 00:50:44 -!- tusho has quit. 00:51:11 -!- Corun has quit ("This computer has gone to sleep"). 00:55:50 -!- Corun has joined. 01:21:52 -!- Sgeo has joined. 01:40:40 hrrr 01:56:26 -!- lilja has quit ("KVIrc 3.2.0 'Realia'"). 03:13:58 -!- puzzlet_ has joined. 03:26:52 -!- puzzlet has quit (Read error: 110 (Connection timed out)). 03:27:39 -!- puzzlet_ has changed nick to puzzlet. 04:06:07 -!- Corun has quit ("This computer has gone to sleep"). 04:25:00 Hmm, who was it who had us all talking about sex in here? 04:25:43 Was it you, psygnisfive? I have no idea, really, apart from that it wasn't someone I recognize easily, and their nick was a bit longer than mine. 04:25:44 ihope: I'm going to go with "any female" for the $500, regis 04:26:00 ive always got you talking about sex 04:27:21 Yep, it was you. 04:27:31 Turns out that wasn't a sexual desire at all. Sorry. 04:28:02 ihope, how old are you? 04:28:38 -!- pikhq has quit ("leaving"). 04:29:03 Now is not the time to tell you that. :-P 04:29:44 Older than ehird, who I imagine will have his oldth birthday this year, or maybe next. 04:29:59 every year everyone has their oldth birthday. 04:31:01 Actually, old = 13. 04:31:19 well in that case yes, this is true about tusho. 04:31:56 Though then you're only mostly old, until you turn 18, in which case you're completely old, and then you turn 70 and become thoroughly old. 04:32:06 s/in which case/at which time/ 04:32:20 lol 04:32:23 18-70 04:32:25 thats a big gap 04:32:26 lol 04:32:47 how old are you ihope 04:32:49 tell me. 04:32:49 Nothing happens in that gap, apart from the Shive Transition, which happens at 25, and has the effect of allowing you to rent a car. 04:32:58 Mostly old. 04:33:10 so you're under 18? 04:33:14 Yep. 04:33:17 damnit, theres goes any hope of sexing you. 04:33:23 guess rape'll have to to. ::rape:: 04:33:36 Oh, by no means! Unless you're male or something. 04:33:45 oh i am. :p 04:35:17 ::pet:: dont worry, it only hurts the first few times ;O 04:35:49 oh, speaking of rape, i have 29 books on theory of computation. do you want? 04:36:04 If you can send them to me, yes. 04:36:22 http://www.wellnowwhat.net/transfers/Theory%20of%20Computation.zip 04:36:31 Is that 29 books? 04:36:45 yep 05:06:02 -!- cmeme has quit (Read error: 104 (Connection reset by peer)). 05:09:54 -!- cmeme has joined. 05:16:27 -!- CakeProphet has joined. 05:16:47 awwww yeah 05:21:19 -!- moozilla has quit (Read error: 104 (Connection reset by peer)). 05:22:15 -!- GreaseMonkey has joined. 05:24:40 -!- moozilla has joined. 05:27:05 sup cakeprophet 05:27:51 word 06:13:08 -!- Sgeo has quit (Read error: 110 (Connection timed out)). 06:16:57 -!- puzzlet_ has joined. 06:17:40 -!- puzzlet has quit (Remote closed the connection). 06:34:26 -!- CakeProphet has quit (Remote closed the connection). 06:59:32 -!- CakeProphet has joined. 07:11:34 -!- psygnisf_ has joined. 07:11:34 -!- psygnisfive has quit (Read error: 104 (Connection reset by peer)). 07:12:38 -!- olsner has joined. 07:59:59 -!- clog has quit (ended). 08:00:00 -!- clog has joined. 08:17:24 -!- Polar has quit (clarke.freenode.net irc.freenode.net). 08:17:25 -!- Quendus has quit (clarke.freenode.net irc.freenode.net). 08:17:50 -!- Quendus has joined. 08:17:50 -!- Polar has joined. 08:32:01 -!- olsner has quit ("Leaving"). 09:19:01 -!- moozilla has quit (Read error: 104 (Connection reset by peer)). 09:21:22 -!- moozilla has joined. 10:31:17 -!- GreaseMonkey has quit ("HydraIRC -> http://google.com <- Go find something better"). 10:31:44 -!- tusho has joined. 11:51:48 hi tusho 11:51:55 hello 11:52:09 bbiab 12:13:30 -!- oklopol has joined. 12:16:25 -!- lilja has joined. 12:55:26 -!- moozilla has quit (clarke.freenode.net irc.freenode.net). 12:57:23 -!- moozilla has joined. 12:58:40 -!- Sgeo has joined. 13:32:47 -!- lilja has quit ("KVIrc 3.2.0 'Realia'"). 13:54:23 -!- Hiato has joined. 13:55:48 -!- Corun has joined. 13:56:27 -!- Corun has quit (Client Quit). 13:57:07 -!- Corun has joined. 13:57:50 -!- Corun has quit (Read error: 54 (Connection reset by peer)). 14:01:01 -!- Corun has joined. 14:37:03 -!- pikhq has joined. 14:37:42 -!- Corun_ has joined. 14:42:46 -!- Corun has quit (Nick collision from services.). 14:42:54 -!- Corun_ has changed nick to Corun. 14:56:48 -!- Corun_ has joined. 15:02:25 -!- Hiato has quit ("Leaving."). 15:04:59 -!- Corun has quit (Read error: 110 (Connection timed out)). 15:10:41 does anyone know if there is some C escape sequence for "Form Feed", like \n for Line Feed and \r for Carriage Return? 15:11:02 -!- Corun_ has quit ("This computer has gone to sleep"). 15:12:30 AnMaster: yes, \f 15:12:35 thanks 15:14:24 Next time try the first link Google gives 15:14:38 http://gd.tuwien.ac.at/languages/c/cref-mleslie/FUNCTIONS/escape.html 15:16:18 btw anyone know what the form feed was actually used for? 15:16:34 I mean, it is one of those control chars like vertical tab that simply isn't used today 15:24:44 emacs source has vertical tabs 15:24:47 so do rfc' 15:24:48 s 15:24:52 *rfcs 15:28:08 tusho, sure it isn't formfeeds? 15:29:11 I'm well aware of Emacs' usage of form feeds to seperate chunks of code into logical pages. . . 15:29:36 (Emacs, of course, supports the use of this convention) 15:30:22 AnMaster: Yes. 15:30:31 * tusho shrugs 15:31:13 and trefunge uses it 15:31:18 formfeeds I mean 15:34:29 Hmm. 15:35:02 For my first programming assignment in college, should I use digraphs just to give the grader a headache? :p 15:35:20 pikhq, you mean trigraphs or? 15:35:28 I never heard of digraphs 15:35:39 An amendment to the C standard in '94 added digraphs. 15:35:48 An example: 15:36:02 does gcc support it without the -trigraphs switch? 15:36:12 pikhq, also what about the bison code? 15:36:16 did you get anywhere with it? 15:37:03 %:include 15:37:06 int main()<%char foo<::> = "Hello, world!\n";printf(foo);%> 15:37:13 Didn't get anywhere with it. 15:37:25 pikhq, but at least you pastebinned it somewhere? 15:37:34 Fuck. 15:37:39 I guess no 15:37:48 or what? 15:37:50 I make no guarantees about such things happening when it's not the weekend. 15:37:58 pikhq, ah right 15:38:00 sorry 15:46:38 -!- Corun_ has joined. 15:54:04 -!- MikeRiley has joined. 15:57:54 pikhq: what is <::> 15:58:17 [] 15:58:50 -!- eth01 has joined. 15:58:54 -!- eth01 has left (?). 16:08:31 tusho: that bit was evidentally invented by someone who didn't like lexer-writers 16:08:49 hee 16:09:24 And by someone who liked obfuscated C. 16:13:12 can you mix mono-, di-, and tri-graphs? 16:14:13 Yes. 16:14:30 You want even more insanity? 16:14:39 Digraphs, like monographs, are considered tokens. 16:14:40 monographs? 16:14:44 Trigraphs are preprocessed. 16:14:48 tusho: The ordinary characters. 16:14:58 a 16:19:14 In format English, if I want to put a footnote after a sentence, should the number be after or before the period? 16:20:14 I suggest 'whatever LaTeX does'. 16:20:39 pikhq, it depends on where you put the footnote symbol iirc 16:20:56 Bleck. 16:21:01 Well, I firmly don't know. 16:21:10 anyone else? 16:26:28 well, you could just do the logical thing 16:26:43 if it refers to the whole sentance, put it outside 16:36:56 -!- Corun_ has quit ("This computer has gone to sleep"). 17:02:30 -!- Hiato has joined. 17:03:51 pikhq: that obfuscated bubble sort has some nasty undefined constructs in it 17:04:05 SimonRC: It assumes GNU C, obviously. 17:04:20 more than that too 17:04:20 What other undefined constructs do you refer to, though? 17:04:30 it uses and increments a variable in one expression 17:05:00 "putchar(b[1[i]]) + b++;" 17:05:34 Ah yes. 17:05:53 I'm evil. 17:05:55 you do the same thing elsewhere 17:06:06 I think that's well-defined in GNU C, however. 17:06:14 pikhq: and non-pedantic compilers can make your code break 17:06:42 Non-GCC compilers choke on it already. 17:06:48 AnMaster: simple 17:06:52 if it's a footnote about the whole sentence 17:06:53 Expression statements are a GCC-only feature. 17:06:53 after the . 17:06:58 if it's about the last part of the sentence 17:06:59 before 17:07:11 tusho, thanks 17:07:16 the first code block is gratuitous; you can use comma-expressions 17:07:18 This was covered in J's sucky article[1]. 17:07:18 [1]: http://j.com/sucky 17:07:35 On another note, I hate J.[1] 17:07:35 [1]: http://cool.com/why_i_hate_j 17:07:47 -!- Tritonio__ has joined. 17:07:47 pikhq, what bubble sort? 17:07:49 tusho: looks like how I'd do it 17:07:53 hello 17:08:15 SimonRC: I also follow Djikstra's parenthesis stuff, mostly 17:08:30 pikhq, got a link to your implementation that you were discussing? 17:08:32 tusho: where is that? 17:08:34 or does SimonRC have a link? 17:08:38 Foo. (If a whole sentence is a parenthesized note, the period goes within the parens.) Bar. 17:08:45 and starts with a caps 17:08:51 http://pikhq.nonlogic.org/bubble.c 17:08:56 tusho: well duh; that's obviously the correct way to do it 17:09:00 SimonRC: no 17:09:03 i was buyy typing 17:09:05 that was just the first part 17:09:12 anyway, I now realise I don't quite follow djik's conventions 17:09:13 I do this 17:09:13 pikhq, oh god 17:09:18 Foo (if, however, it's within a sentence, start with a lowercase letter and end it outside). 17:09:18 (the "well duh" was partly ironic, BTW) 17:09:20 pikhq, golfing? 17:09:21 as for quoting 17:09:29 "If a whole sentence is a quote, within." 17:09:35 tusho: yep, that's how I do it too 17:09:37 If, however, I am just quoting "you suck", outsid. 17:09:38 e 17:09:41 No; my golfed version uses the preprocessor, and counts as a one-liner. :p 17:09:42 (that is non-standard) 17:09:50 standard is putting the comma in the quote 17:09:52 which is braindead 17:09:58 (don't have it handy; it's on frodo, which I won't have access to for a couple more weeks yet) 17:10:05 pikhq, hah, well afk for a bit, food 17:10:12 pikhq, tell me how it works though when I get back 17:10:15 pikhq: I suspect that the other two code blocks could be removed too if you turned the whole program into a state machine. 17:10:26 It's just bubble sort. 17:10:31 SimonRC: Meh; too much effort. 17:10:38 :p 17:10:41 (for fun, do it with recursion on main and ?: as the only control structures) 17:10:59 QUESCOL 17:11:10 (reference to http://rghosh.free.fr/essays/quescol.html, some idiot's article) 17:11:16 (he advocates writing JS with mainly recursion and ?:) 17:11:33 skip to /* parse a CSV line recusively, in quescol style */ 17:11:34 and laugh 17:11:42 also note JS' lack of tail recursion 17:12:06 tusho: ouch 17:12:18 otherwise, that would be a nice functional style 17:12:23 -!- Tritonio__ has quit (Client Quit). 17:12:27 SimonRC: apart from the syntax 17:12:33 so, apart from the whole point of it, it's nice ;) 17:12:43 also the fact that its totally non-idiomatic 17:12:50 so you're basically making it hell for anyone else to read or modify it 17:12:54 * SimonRC reads it... 17:13:42 my quoting style would put the punctuation in the quoted text if it was part of the quoted text. Note that I would distinguish quoting the whole contents of a sentance and quoting the whole sentance. 17:16:22 -!- MikeRiley has quit ("Leaving"). 17:17:52 -!- Hiato has quit (Read error: 104 (Connection reset by peer)). 17:18:08 needs slightly better indenting and more pares, but otherwise quite readable 17:18:15 -!- Corun has joined. 17:18:37 maybe splitting it into seperate functions for states 0 and 1 would help 17:20:50 and some of those conditions need to o the other way out 17:20:54 *go 17:21:49 wow, his indentation just goes down the plughole at the end of that function 17:23:17 in fact, escaping should be a third state 17:28:16 -!- MikeRiley has joined. 17:29:14 SimonRC, I find the "procedural" style easiest to follow 17:29:22 but I guess it is just due to being a C programmer 17:29:27 ah, waitamo, he continues to improve it... 17:29:49 SimonRC, you disagree I guess? 17:30:00 MikeRiley, hi! 17:30:01 SimonRC: even if it is kinda nice, it'll be horrid to write and its hideously non-idiomatic 17:30:03 hello 17:30:04 and the whole stack thing 17:31:28 multiple return values would be nice; dunno if js can do those 17:32:32 * SimonRC would like a comment box on that blog 17:34:59 MikeRiley, what exactly is the point of 3DSP? 17:35:37 it provides vector and some matrix operations... 17:35:47 on 3d vectors... 17:35:48 are they hard to implement in Funge? 17:35:53 not hard at all... 17:35:58 hm ok 17:36:29 i thought once about doing something really crazy and writing a simple raytracer in funge,,,and needed the 3d vector/matrix operations to do it... 17:36:34 MikeRiley, I do see the point of providing floating point (could be hard to implement) and sqrt, sin and so on (could also be hard to implement) 17:36:57 funges do not do function calls nicel 17:36:57 the floating point was not hard,,,since single precision is 32 bits, same as the cell size,,, 17:36:58 y 17:37:12 SimonRC, depends, try SUBR 17:37:19 if you use SUBR then function calls work out ok... 17:37:43 -!- lilja has joined. 17:37:53 MikeRiley, anyway how would you handle denormals and such in pure funge version of floating point? 17:37:59 it would be nice to have a "side call" instruction 17:38:01 I hope you didn't implement it using mini-funge! 17:38:06 do not need to... 17:38:11 no? 17:38:23 i use a union with a 32-bit integer and a floating point value... 17:38:33 as long as you do not use the fp cells for integer math, then it works fine... 17:38:39 MikeRiley, yes right, I thought you implemented it in mini-funge :P 17:38:43 which would be quite a feat 17:38:43 it pushes the "next" location on the return stack, then instead turns left/right and goes that way... 17:38:48 while doing it in C is easy 17:38:49 i read the cells using the union, and do the math from the union... 17:38:53 I do it using a union 17:39:02 nice and fungey rather than that crap with calculated offsets 17:39:04 MikeRiley, same 17:39:04 no, did not do it in minifunge, would have been too hard!!! eheheheheheeh 17:39:22 (a return would simply pop+jump) 17:39:37 MikeRiley, my interpreter will support mini-funge the day someone write TRDS in pure mini-funge (I don't think that is possible) 17:39:40 :P 17:39:48 i doubt that is possible!!! eheheheheheeheh 17:39:56 well, befunge is turing-complete 17:39:59 too tied into the interpreter... 17:40:05 you'd have to write a befunge interpreter in the mini-funge 17:40:09 which implements TRDS 17:40:10 yes... 17:40:13 that is what you would have to do.... 17:40:15 which would be crazy 17:40:23 and dang slow!!!!!!! 17:40:26 anyway mini-funge couldn't implement FILE either 17:40:33 not without expanding it.... 17:40:36 indeed 17:40:40 I done some extending 17:40:51 based on !Befunge's extended mini-funge 17:40:59 called dynafing 17:41:08 i have seen a page on that somewhere... 17:41:27 brb phone 17:41:29 phlamethrower.co.uk IIRC 17:41:47 phlamethrower = same site as the one with the befunge domain name 17:41:56 quite 17:41:57 save for the logo and link to the other domain 17:42:05 err, nt logo 17:42:07 just the index page 17:42:14 on the other one it adds '# Wondering what the deal is with the domain name? Head over to my befunge pages to find out.' 17:42:27 who runs that site? 17:42:29 seems it's still updated 17:42:39 My email addresses are phlamethrower@quote-egnufeb-quote-greaterthan-colon-hash-comma-underscore-at.info, or alternatively me at phlamethrower dot co dot uk. 17:42:40 his name is there somewhere 17:42:45 I think 17:42:51 i think it is... 17:43:28 hey Deewiant, did you see my email about the FIXP errors??? 17:46:05 I did now :-) 17:48:16 hmm, I wonder if the _ is just flipped 17:48:20 -!- Corun has quit ("This computer has gone to sleep"). 17:48:29 MikeRiley: what does the stack look like at (37,523) 17:49:26 hold on. let me see if i can find out... 17:49:41 RC/Funge-98 has a debugger, no? :-) 17:51:34 yes it does... 17:51:58 so just do "bp 37 523" or whatever the command was 17:54:19 does not look like 37 523 is an instruction it will stop at... 17:54:48 meh 17:54:51 there should be a _ there 17:54:51 tested break points at other addresses and works fine.... 17:54:56 hmmmmmm 17:55:04 maybe you've got a different origin than I do 17:55:09 line 524, column 38 17:56:05 or hmm 17:56:07 let me try again... 17:56:10 if it doesn't stop there 17:56:22 does it go to (45,520) (a w)? 17:56:34 let me try... 17:56:44 maybe the w is flipped actually 17:56:50 stopped at that 17:56:55 what's on the stack 17:56:59 hold on 17:57:10 0 350000 8191 8191 8192 17:57:22 alright 17:57:33 and next it probably goes to (45,519) right? 17:57:38 let me see... 17:57:49 yes... 17:58:07 alright, so I've just got the w wrong 17:58:38 hopefully a simple fix... 17:58:47 yeah, it is 17:59:10 or it should be 17:59:17 ;-) 17:59:20 eheheheh 17:59:47 meh, stupid code results but oh well 17:59:57 since there isn't much room there so I have to do a bit of jumping around :-P 18:00:12 or there is quite a bit of room actually, I just don't want to move code around 18:00:52 yeah,,,always fun editing funge code!!!! 18:02:15 your asin is wrong 18:02:22 17.45760312 is the exact result 18:02:30 17.4575 is wrong then 18:03:10 depends on rounding and precision.... 18:03:18 not rounding 18:03:26 since it's over 17.4576 you can't round it to 17.4575 18:03:32 and since this is fixed point, the precision is fixed 18:03:32 true,,,,, 18:03:53 in my interpreter it is doing all the math in floating point and converting back... 18:03:56 so you can't say it's due to floating point inaccuracy :-P 18:03:59 sure 18:04:03 maybe you should doubles instead :-) 18:04:06 alright, so I've just got the w wrong 18:04:07 +use 18:04:08 Deewiant, what? 18:04:08 probalby should!!! 18:04:14 Deewiant, what is going on with w? 18:04:26 w turns left or right or goes forward 18:04:30 I had the left and right cases the wrong way around 18:04:31 Deewiant, yes I know 18:04:37 Deewiant, in all places? 18:04:40 in two places 18:04:47 in mycology that is? 18:04:53 yes, in FIXP 18:05:01 Deewiant, anyway, does this affect my FIXP implementation? 18:05:04 you're probably not getting any BADs from FIXP so it doesn't affect you 18:05:07 ah 18:05:08 right 18:05:10 no BAD 18:05:16 it's the places where it checks for two possible answers 18:05:25 whereas i am getting a couple bads with numbers that are correct... 18:05:25 Deewiant, however if I change k behaviour I get BAD from random places in mycology 18:05:35 AnMaster: yes, that's to be expected 18:05:41 and mycology is to be updated in that regard. 18:05:56 Deewiant, well when will it be updated like that? 18:06:07 weekend, as I've said often before 18:06:12 ah right 18:06:14 or hopefully the weekend, anyway 18:06:24 yeah,,,always fun editing funge code!!!! 18:06:27 I got a trick 18:06:32 to make maintainable funge code 18:06:42 it makes it almost like a normal programming language 18:06:42 and that trick is??? keep space!!!! eheheheheeh 18:06:48 just end every line with: 18:06:49 11x> 18:06:53 11x> 18:06:58 aligned like that 18:07:09 i see.....interesting.... 18:07:18 " si gnidaeH :FEDNU">:#,_A.a"seerged">:#,_ 11x> ; UNDEF: Initial heading ; 18:07:18 "( era sdnuoB :FEDNU">:#,_U.."( )",,,..a'),, 11x> ; UNDEF: Bounds ; 18:07:18 a7+3*5*N 0C 11x> ; Clear with blue. Set pen to black ; 18:07:18 0H 1P 11x> ; Set direction, pen down. ; 18:07:19 MikeRiley: there's a new version of mycology up, just redownload it, should work hopefully... 18:07:20 is an example 18:07:24 MikeRiley, that works really well 18:07:28 ok,,,,will go and grab it.... 18:07:42 brb 18:09:42 yep,,,,that looks better.... 18:10:07 Hey, the lambda calculus functions form a ring, don't they? Makes me wonder if there are any prime functions. 18:10:13 a couple more bads scratched of my list.... 18:10:36 $ ./cfunge -S catseye/diagt.b98 18:10:36 thread 2 started 18:10:36 thread 1 started 18:10:36 OK 18:10:47 Deewiant, the thread error went away when I fixed the k issue 18:10:54 so I guess it was just an effect of that 18:10:59 alright 18:11:46 i guess that clarifies the k issue a bit more.... 18:12:06 MikeRiley, btw with the k as catseye there will be some issues, I can paste them if you want (as my implementation passes except for that 18:12:19 BAD: 3k4 leaves more than 3 fours on stack 18:12:19 BAD: "a b" takes more than 5 ticks 18:12:19 BAD: 0y pushes wrong stack stack size 18:12:21 those ones 18:12:30 yeah, i was just ignoring those one for now... 18:12:37 Deewiant, but tell me how the "BAD: "a b" takes more than 5 ticks" depends on k 18:12:41 I would like to hear 18:12:53 as if I would know offhand 18:12:54 the "a b" one works fine for me.... 18:12:59 that one shows good.... 18:13:13 the other 2 show bad tho... 18:13:50 oh,,,so the k issue effects the 0y one???? well,,,i guess i can quite trying to fix that one!!! eheheheeheheheheh 18:14:19 there are some places in mycology where I've just used a value that happened to be left on the stack 18:14:33 so if something previous is BAD, the value may be different 18:14:39 which can cascade really deeply 18:14:40 makes sense... 18:14:59 something like that might only show up 100 lines later 18:15:08 MikeRiley, I get "good" for the "a b" one when k works as in CCBI 18:15:45 which probably means you got another error hiding it hehe 18:15:53 that one has always been good for me,,,,after i fixed the multi-space problem... 18:16:19 well do you get all GOOD up to that point? 18:16:24 hmmmmmm, i should break k in mine and see what it does... 18:16:32 hold on....let me look... 18:16:36 of course he doesn't, since 3k4 is before that 18:16:38 yes not very hard to break 18:16:47 Deewiant, ah right duh 18:17:06 the 3k< shows up bad right above it 18:17:22 and the 0y one and the 3k4 one... 18:17:26 everything else shows good.... 18:17:29 or undef 18:17:57 actually, the 0y is below that point,,,so no effect from that... 18:18:32 so only bads above the "a b" are the 3k4 and the 3k< 18:19:10 MikeRiley, could be two bugs hiding each other 18:19:13 MikeRiley, I had that once 18:19:14 not sure what is the deal with the 3k< since way i see it,,,it is going to take 3 ticks,,,,depending on how you test.....1 tick for the 3, 1 tick for the k,,which the ip will now go the other way, and then another tick for the 3 again... 18:19:23 certainly possible... 18:19:23 MikeRiley, when I fixed one the other part broke 18:19:29 I had *two* off by one errors 18:19:32 i have had that happen to me as well... 18:19:33 that masked each other 18:19:44 and showed up when it tested negative funge space 18:20:11 I also had some off by two errors iirc 18:20:17 those are the aggravating kinds of problems... 18:20:50 -!- Corun has joined. 18:20:51 oh yes I also found a few bugs in mycology 18:21:04 MikeRiley, and one (indirectly) in CCBI 18:21:14 oh and quite a few crash bugs in ccbi 18:21:24 "quite a few"? 18:21:24 MikeRiley, does all your fingerprints handle division by zero? 18:21:28 do remind me :-P 18:21:32 they should.... 18:21:39 CPLI didn't in mine, IIRC 18:21:44 but not certain,,,i need to check it... 18:21:48 Deewiant, division by zero in FIXP (iirc), CPLI, and some more 18:21:52 TOYS too(?) 18:21:56 could be some division by zero bugs still there.... 18:21:57 meh, that's all one bug :-) 18:22:02 MikeRiley, I used fuzz testing on cfunge 18:22:11 I got a script for that, but it is cfunge specific 18:22:35 but issues found with it can be adapted (in fact some tests in tests/ in cfunge are that) 18:23:01 anyway it is cfunge specific because it expects that the program will pass valgrind completely in normal operation with no memory leaks 18:23:02 breaking the k,,,,allows both the "a b" and the 3k< to show good 18:23:04 and errors 18:23:11 MikeRiley, ok that's odd... 18:23:25 and fixing the k the "a b" shows good and the 3k< does not 18:23:52 breaking the k also shows the 0y one correctly... 18:23:54 well here "a b" breaks only with broken k 18:23:56 which is odd 18:24:01 interesting.... 18:24:55 MikeRiley, does RC/Funge search past spaces when looking for instruction to repeat with k? 18:25:05 I got a mail from C. Pressy that said that was what to do 18:25:15 also seek past any matching ; .... ; 18:26:03 MikeRiley, if you give me your mail I can forward it 18:26:27 it does look past spaces,,, 18:26:31 and past ; pairs... 18:26:35 ah :) 18:26:38 ccbi doesn't iirc 18:26:49 MikeRiley, does it also handle wrapping? 18:26:58 not sure there,,,,would have to test for that.... 18:27:04 a sec 18:27:05 I got one 18:27:36 3k ; ; ;; ; f 18:27:44 add some code at the end to print out to check 18:29:34 will try it.... 18:29:50 -!- CakeProphet has quit ("lolwhut?"). 18:30:18 that works just fine... 18:30:18 * SimonRC goes 18:30:24 MikeRiley, my funge-108 draft can be found at http://rage.kuonet.org/~anmaster/funge-108/ 18:30:29 in pdf and lyx formats 18:30:31 going to grab it... 18:30:38 lyx is just a GUI frontend to LaTeX really 18:31:02 got it... 18:31:21 www.lyx.org (but you can view the pdf version easily enough) 18:31:53 thanks, will take a look at that... 18:32:19 by the way,,,on the 3k ; ; ;; ; f i get 4 15s on the stack... 18:32:38 well yes that would match Funge-98 behaviour 18:32:49 yep... 18:33:18 (not Funge-108, you would get 3 15s, as it *always* skips go forward then then) 18:33:21 wait... 18:33:23 hm 18:33:47 Deewiant, when it skips over after k in 108... Should it skip over next *instruction* or next char 18:33:54 as this may differ in cases like: 18:34:00 3k 4 18:34:08 C. Pressy said you should seek past spaces 18:34:24 hmm 18:34:32 the only case in which it matters is if the k instruction modifies funge-space 18:34:42 Deewiant, not really, if matters if you turn 18:35:04 well it does in the way I implement "seek past spaces" 18:35:05 i understand that spaces are ethereal, and for the most part they do not exist as far as the ip is concerned 18:35:13 yep, same with ; 18:35:13 MikeRiley, yep... 18:35:24 same thing with the ; pairs.... 18:35:27 aye 18:35:38 so seek past next instruction? 18:36:00 AnMaster: well, do you think that it makes sense that if you turn in "2k ]" that you continue from under the ]? 18:36:15 Deewiant, um no 18:36:23 then it has to be the char, no? 18:36:30 otherwise it's special cases again 18:36:30 well but: 18:36:44 1kv 18:36:46 18:36:47 18:36:51 a 18:36:57 should that jump over the spaces? 18:37:01 not that it matters in this case 18:37:09 but it matters when you have: 18:37:12 1k a 18:37:15 again 18:37:20 where it would be logical to jump over the a 18:37:26 if you want "1k ]" 18:37:28 as otherwise you execute it once to many 18:37:34 s/1/2/ 18:37:51 to not continue from under the ], then it has to be the char, or you get special cases, which we would like to avoid 18:37:54 Deewiant, question is. should k resume execution 2 chars from itself. or 2 instructions from itself 18:38:00 assuming the position wasn't changed 18:38:07 (but delta may have been) 18:38:23 and my answer is the same as I have already said twice 18:38:29 my take on k, for all cases except 0k,,,,is that the ip stays at k until the end of the instrution, and then the ip moves normally along its delta... 18:38:33 but I never said "continue under ]" 18:38:42 I said should it just jump over one char 18:38:48 or should it jump over one instruction 18:38:52 wherever that instruction may be 18:38:58 1k ] 18:39:00 a 18:39:01 or 18:39:02 1k ] 18:39:05 18:39:06 18:39:07 a 18:39:09 err 18:39:11 a 18:39:13 is what I mean 18:39:15 if you have it jump and instruciton,,,,then lkv should jump the instruciton underneath the k... 18:39:17 after some blank lines 18:39:28 why should their be exceptions of the delta is changed?? 18:39:33 specs do not call for that.... 18:39:37 no there shouldn't be 18:39:39 I didn't say that 18:39:45 gah must try to say it clearer 18:39:52 will bbiab with some examples 18:39:58 ok,,, 18:40:10 I don't know, it seems silly to me that "1k 4" continues after the 4 but "1k ]" after whatever's under the k 18:40:16 but if you're fine with it, fine 18:41:00 all my FIXP errors are now fixed!! :) 18:41:21 i agree.,,,which is why 1k 4 continues from the k.....and not the 4.... 18:41:22 Deewiant, MikeRiley: http://rafb.net/p/5Dpvny52.html 18:41:33 the delta will move it past the spaces to execute the 4 again... 18:42:06 MikeRiley, that will make it impossible to do 0 iterations: 0ka 18:42:10 will still just 10 once 18:42:12 that's the way I see it as well 18:42:23 except the 0 iterations causes a jump,,,,but that is an exception.... 18:42:37 MikeRiley, so it is impossible to do one iteration then? 18:42:40 i wonder if the 0 just means the k does not do anything,,,the instruction follwoing would execute as normal... 18:42:46 that's even more odd IMO 18:42:54 from what i understand of the specs,,,yep,,,not possible for 1 iteration... 18:43:06 which I think should be possible in Funge-108 18:43:17 again,,,the specs do not say that the IP moves from the k while k is being executed.... 18:43:36 i agree,,,,and would like to see the iteration of 1.... 18:43:44 but i think in Funge-98, it is not possible with k... 18:43:48 right 18:43:53 but in 108 it will be 18:43:56 good!!!! 18:43:57 the issue here is spaces 18:44:07 as you can see in http://rafb.net/p/5Dpvny52.html 18:44:11 spaces are ethereal, as if they do not exist... 18:44:26 MikeRiley, so k should jump past the next instruction always? 18:44:31 or next char always? 18:44:36 in my mind it does not matter if you have 1kab or 1k ab, you will get the same result... 18:44:38 anyway next char doesn't work 18:44:49 2k; this is a comment;a 18:44:57 now that doesn't make any sense 18:45:01 i agree,,,, 18:45:04 AnMaster: that could be a design criterion for you: inserting any ; pairs or spaces between k and instruction shouldn't matter 18:45:04 if you don't go past next char 18:45:07 ,,,,,, 18:45:20 spaces and ; pairs are ethereal, and should not be taken into account when k moves the ip... 18:45:31 Deewiant, indeed it shouldn't matter, k will seek past spaces and ; 18:45:40 as the mail from Pressy that I forwarded said 18:45:57 MikeRiley, agreed. So this means that: 18:45:57 1kv 18:45:57 18:45:57 a 18:45:57 b 18:45:58 AnMaster: yeah, but now you're bringing in all these ; cases 18:46:01 will push the b? 18:46:11 that is the only logical conclusion then 18:46:14 AnMaster: 2k; this is a comment;a should be identical to 2ka should be identical to 2k ;foo bar; a 18:46:15 will push the a and the b,,,,in 98,,,,and i guess in 108, the b.... 18:46:26 Deewiant, indeed it should 18:46:29 -!- Corun has quit ("This computer has gone to sleep"). 18:46:34 i agree there deewiant... 18:46:36 which means it should *seek past next instruction* 18:46:39 not *next char* 18:46:40 AnMaster: so what's the problem then :-) 18:46:42 which is what # does 18:46:44 if you are seeking yes.... 18:46:53 AnMaster: well, the way I see it is not that it seeks 18:47:01 1kv 18:47:01 18:47:01 a 18:47:01 b 18:47:06 that pushes a or b? 18:47:12 both 18:47:30 1kv 18:47:30 ; 18:47:30 Really? 18:47:30 ; 18:47:30 a 18:47:31 b 18:47:34 what about that? 18:47:36 yes, really. 18:47:45 because, as I just said, it shouldn't matter no matter what ; you put in there. 18:48:09 after k executes, it moves to the next instruction, as normal 18:48:15 there's no special seek or move involved 18:48:22 right? 18:49:07 hm 18:49:20 right the one after it just executed 18:49:24 this is 108 after all 18:49:28 not 98 18:50:13 with 108,,,you can just say,,,,that k skips the instruction it iterated,,,,if the ip moves somewhere other than the iterated instruction,,,,then it is not skipped... 18:50:36 MikeRiley, that would be CCBI behaviour 18:50:42 which causes extra special cases 18:50:53 why extra special cases??? 18:50:57 only 1 special case... 18:50:58 I hoped to avoid that by simply saying "always jumps forward" 18:51:12 hrrm 18:51:17 I guess that is the only sane way 18:51:30 only special case is,,,,if the next instruction to be executed by the ip is the iterated instrution, skip it.... 18:51:45 otherwise execute it... 18:52:08 that deals with changes in deltas without skipping anything and then would work properly for things like 0k or 1k.... 18:52:18 with a non delta changing instruction after it... 18:53:23 yes 18:53:36 what about next instruction being a p? so it overwrites itself 18:53:46 then you could argue it is not the same instruction ;P 18:53:56 ok,,,,maybe better to say,,,the ip skips the cell of the iterated instrution.... 18:54:07 no matter what is there... 18:54:09 right :) 18:54:11 and only if it's going to hit it next 18:54:18 I guess that works 18:54:19 yes,,,only if it hits it next... 18:54:24 that'd be what CCBI does currently? 18:54:32 "If the next cell to be executed after k would be the one it iterated on, then it moves ip forward to just past said instruction. The result is that k skips over the next instruction." 18:54:35 Deewiant, it is 18:56:17 all my FIXP problems are now solved.... 18:56:38 did you just switch to doubles? ;-P 18:57:06 that i did.... 18:57:16 solved all the remaining bads... 18:57:20 in FIXP... 18:57:32 it's probably still not a robust solution, if I just tested big enough numbers 18:57:37 but meh, whatever 18:57:50 what is the issue with the FIXP Deewiant? 18:57:59 no issue 18:57:59 not using epsilon to test? :P 18:58:13 well, it's fixed point 18:58:19 so it should be fully accurate to the fixed precision 18:58:26 I bet I would get other results if I used 128-bit SSE floating point numbers 18:58:29 but since I guess everybody does stuff via floating point 18:58:41 Deewiant, you do it by floating point too? 18:58:44 I suppose it'll break down at some point and you'll get an error in the last digit 18:58:55 because fixed point sin() would be a pain in C at least 18:58:55 yeah 18:59:02 yes it would... 18:59:04 and no, not really, I don't think so 18:59:25 just some fixed-point iteration or whatever 18:59:44 but I still can't be bothered to do it :-P 18:59:52 certainly could be done,,,just iterate the sin formula instead of using the c library to do it for you... 19:00:27 yep 19:00:35 It's not like it's impossible to compute the Taylor series... 19:00:39 yep 19:01:07 nope,,,in fact my trig functions on my elf stuff uses fixed point arithimetic on the taylor series.... 19:01:50 which was done in 1802 assembly,,,would actually not be that hard to do it in c... 19:02:22 -!- CakeProphet has joined. 19:03:02 s/,+ ?/, /g 19:03:05 oh....by the way....i have change FILE a little bit,,,,added another command to it... 19:03:06 ;) 19:03:57 -!- olsner has joined. 19:04:34 -!- Hiato has joined. 19:06:29 MikeRiley, about the issues with "a b" I suspect it may be related to some UNDEF before 19:06:35 messing up later on 19:06:40 with the changed k 19:06:42 possible... 19:06:48 that is all I think of as a reason for it 19:07:00 MikeRiley, you probably do other stuff at some undef than me 19:07:02 currently 15 bads remaining to be fixed, all but 3 are in the fingerprints... 19:07:07 that is possible.... 19:07:27 execution time of mycology is: 19:07:28 real 0m1.476s 19:07:28 user 0m1.400s 19:07:28 sys 0m0.052s 19:07:33 http://rafb.net/p/uVhUs586.html 19:07:50 MikeRiley, I guess some differ? 19:08:00 as for i and o 19:08:06 those are due to the -S switch 19:08:10 -S enables sandbox mode in cfunge 19:08:16 preventing any unsafe actions 19:08:24 and blocking some fingerprints 19:08:27 your undefs look like mine with the exception of i an o... 19:08:28 while -F disable all fingerprints 19:08:45 MikeRiley, I think my sandbox one is a good idea, allow safe fuzz testing too :) 19:09:05 yes, that does sound like a good idea... 19:09:05 anyway I get same result without -S and -F 19:09:38 to be exact, -S blocks: i, o, = and certain fingerprints 19:09:49 oh it also hides some environment variables 19:09:55 it just allow a safe set of them 19:10:09 might want to add that to mine as well... 19:10:48 http://rafb.net/p/sPhQzF38.html is the gperf file I use to generate the list of good environment variables 19:10:53 i assume the fingerprints you block are things like FILE and SOCK?? 19:10:55 gperf generates a perfect hash 19:11:05 MikeRiley, yes indeed PERL too 19:11:07 and some more 19:11:11 yes perl too... 19:11:16 I haven't yet implemented SOCK btw 19:11:22 but it would definitely be one to block 19:11:57 yes,,,it would... 19:12:06 MikeRiley, DIRF too btw 19:12:15 DIRF as well... 19:12:42 DIRF, FILE and PERL are the only ones that I block so far, but I haven't yet implemented some fingerprints 19:12:48 others look like they would be safe... 19:13:01 only system interaction ones would need to be blocked... 19:13:02 FPDP, FPSP TOYS yeah safe I think 19:13:21 technically STDIO is system interaction... ;P 19:13:30 but no I don't block &~,. 19:13:44 true,,,,,but a bit harder to do something destructive with stdio... 19:13:59 well I could really mess up your terminal 19:14:01 using control codes 19:14:07 yep..... 19:14:13 say, put it into some don't echo and use line graphics mode 19:14:36 (of course just type "reset" and hit enter to restore it) 19:14:37 could always block the character in sandbox mode... 19:15:12 yep.... 19:15:13 nah I wouldn't block esc 19:15:14 as well as control characters besides lf,cr,nl 19:15:29 messing up the terminal is really not that big a deal...as you said....reset would fix anything that happened... 19:15:49 yes assuming it wasn't hardwired to a printer 19:15:56 not likely these days though :P 19:16:07 yep...not too likely.... 19:16:29 and if it was it would the users own fault heh 19:17:01 anyway I need to push an updated funge108 one up now 19:17:32 got to run for a bit,,,,be back in a little while... 19:17:44 just put up a new one that reflects k better 19:17:49 upload done 19:17:51 and cya 19:18:03 will look when i get back.... 19:18:06 cya.... 19:18:08 -!- MikeRiley has quit ("Leaving"). 19:19:20 -!- Corun has joined. 19:23:15 Deewiant, 432kk5 19:23:20 this got 2 issues 19:23:25 at which k is the 5 executed 19:23:38 when it jumps past next instruction after first k 19:23:41 where does it end up? 19:24:03 and then what about 432kkv 19:24:08 from what k does that go down? 19:24:09 it jumps to the 5, no? 19:24:19 and the 5 is executed on the inner k 19:24:22 Deewiant, so it will push a final 5 after the first k right? 19:24:30 that's how I see it 19:24:41 inner k? 19:24:44 you mean second one? 19:24:45 so kkk is even trickier :-) 19:24:48 yep 19:24:57 Deewiant, so what about: 19:24:59 432kkv 19:25:01 ab 19:25:03 what about it 19:25:12 hmm 19:25:12 will that execute a or b after? 19:25:19 a 19:25:26 I'd say 19:25:56 because the outer k completes last 19:26:03 right 19:34:01 Deewiant, http://rafb.net/p/35jr7b24.html (fi is the ligature fi, blame LaTeX output in combination with said ligature existing in unicode...) 19:34:40 no, I blame you for putting an HTML entity instead of UTF-8 :-P 19:34:52 Deewiant, blame that on firefox + pastebin 19:34:59 of of them anyway 19:35:08 pastebin, specifically 19:35:09 I just copied it from pdf output 19:35:17 yeah one of them did it 19:35:34 "execute the 4 four times" s/4/5/ 19:35:43 ah thanks 19:36:09 fixed that typo 19:36:49 so in result, nested k isn't really useful anyway ;P 19:37:09 anyway this means some major headache in cfunge, due to the way it implements stuff 19:37:13 no, but it's not useless either 19:37:24 and yeah, might be a pain to implement :-) 19:37:43 basically yet more special casing 19:37:47 and refactoring 19:38:00 Deewiant, anyway what does ccbi do on nested k currently? 19:38:10 stuff 19:38:18 "stuff"? 19:38:31 as in you want me to read the code and try to figure out? 19:38:34 I think 432kk12345 would execute first 1 3 times, then 2 4 times, before jumping to the 3 19:38:37 but I'm not sure 19:38:41 no, you can just test it :-P 19:39:13 anyway that's an interesting interpretation you had :P 19:39:21 no, it's just no handling of the k case 19:39:28 so it moves after executing the inner k 19:39:31 as well as the outer k 19:39:36 well mine wound probably execute the second k on the second k 19:39:37 currently 19:39:56 but I can see how to handle multiple k nicely 19:40:02 it is called recursion 19:40:03 sure, it wouldn't be hard 19:40:05 with special casing 19:40:05 er, no 19:40:10 you don't need anything special 19:40:28 Deewiant, I do, as the instruction doesn't know *where* it is, only where the ip is 19:40:28 the only change to CCBI now would be to not move after executing an inner k 19:40:36 so it will look one ahead from current ip 19:40:55 result: I need to go forward once again and actually execute the second k at the second k 19:41:33 that, and restoring ip after each iteration on top k is all that is needed 19:41:46 this will work on 3 k too 19:41:54 brb 19:48:30 -!- Sgeo_ has joined. 19:50:02 Deewiant, what if the second k changes delta when it first executes? 19:50:13 wait yeah 19:50:25 no it is consistent still 19:50:28 then nothing 19:50:29 yes 19:50:49 however I need to execute the second k at itself each time. 19:53:10 Deewiant, um what if second k changes *position? 19:53:14 *position* 19:53:29 432kk#abcdef 19:53:30 say 19:53:38 I don't want to think about it >_< 19:53:48 4k# itself is messed up enough :-P 19:54:00 Deewiant, nor do I, but I need to think about it.... what should happen 19:54:11 or: 432kk# 19:54:12 err 19:54:13 so go ahead and think about it, I don't want to :-) 19:54:14 or: 432kkj 19:54:39 Deewiant, well what about something else that sets position without messing with other stuff? 19:54:49 say something from SUBR 19:54:51 I don't want to think about it >_< 19:54:52 I don't want to think about it >_< 19:54:57 ah right 19:55:15 Deewiant, well I can't come up with something sane, I thought about it for a while 19:55:54 Deewiant, I will have to ask MikeRiley when he gets back 19:56:08 I don't want to think about it >_< stop pinging me :-P 19:56:34 hrrm... 19:57:09 wow 19:57:12 just wow 19:57:16 internal compiler error 19:57:30 GCC failing on you? :-) 19:57:42 Deewiant, well this code is quite messed up now 19:57:46 heh 19:58:27 as the function got different prototypes with concurrency support and without it 19:59:03 #ifdef CONCURRENT_FUNGE 19:59:03 # define RunSelf() RunIterate(ip, IPList, threadindex) 19:59:03 #else 19:59:03 # define RunSelf() RunIterate(ip) 19:59:03 #endif 19:59:05 that works 19:59:13 no internal compiler error now 19:59:22 fk2kkkt 19:59:26 what!? 19:59:55 Deewiant, well I do handle 2kt and 2k@ correctly 20:00:02 that is why I need the different prototype 20:02:22 wait 20:02:56 In 432kk5 the second k would actually execute the 5 *five* times the second iterations 20:02:58 Deewiant, right? 20:03:09 as the first iteration pushed 5s 20:04:07 Deewiant, tell me, what should 32kk5 >:#._@ print? 20:04:12 *poke* 20:07:01 outer k pops two, executes twice: 20:07:10 inner k pops three, executed thrice: 20:07:14 5 20:07:23 inner k pops zero, nothing happens 20:07:39 oh, d'oh 20:07:46 first time: 5 5 5 20:07:47 inner k pops /five/, executed 5 times: 20:07:50 5 20:07:51 second time... 20:07:55 so 20:07:56 5 5 20:07:57 + 20:07:58 output: 5555555 20:08:00 5 5 5 5 5 20:08:14 Deewiant, nop, . adds a trailing space 20:08:16 according to spec 20:08:21 oh, it was a . not a , 20:08:31 well, anyhoo, 7 bytes 20:08:38 yes 20:08:40 are pushed on stack 20:08:47 and . wouldn't output a 5 20:08:53 it would output some control char 20:08:58 err 20:08:59 , 20:09:00 I mean 20:09:09 Deewiant, anyway then what about: 4k'a 20:09:12 that is a bit hard itself 20:09:19 I think cfunge may do the wrong thing here 20:09:25 that's like 4k# and I don't want to think about it :-) 20:11:24 heeeeey guys 20:11:55 Deewiant, yes... but I think it may push some ' 20:11:58 or a 20:12:01 not sue 20:12:02 sure* 20:13:34 Deewiant, help please http://rafb.net/p/Y2LKjU71.html 20:13:38 does that test seem right? 20:13:43 in cfunge it cause an infinite loop 20:13:49 and tracing goes bad 20:13:50 mad* 20:14:01 tix=0 tid=0 x=4 y=5: " (34) 20:14:01 tix=0 tid=0 x=4 y=4: B (66) 20:14:01 tix=0 tid=0 x=4 y=5: " (34) 20:14:01 tix=0 tid=0 x=4 y=6: (32) 20:14:01 tix=0 tid=0 x=4 y=7: > (62) 20:14:02 tix=0 tid=0 x=4 y=0: k (107) 20:14:05 and that makes no sense 20:14:40 seems right to me 20:14:50 (gdb) call FungeSpaceDump() 20:14:50 Fungespace follows: 20:14:50 432kkv 20:14:50 "" 20:14:50 DD 20:14:51 OA 20:14:53 OB 20:14:57 G" 20:14:59 " 20:15:01 >>:#,_a,@ 20:15:03 seem right there too 20:16:36 Deewiant, I see what the issue is 20:16:44 next time k will look at the " 20:16:45 not the v 20:16:49 as the delta changed 20:16:51 for the inner k 20:16:54 Deewiant, :P 20:17:16 but is that what we want 20:17:22 it is 20:17:23 -!- Sgeo has quit (Connection timed out). 20:17:30 seems weird 20:17:31 Deewiant, just my test case was fucked up then 20:17:38 but I guess it makes sense 20:17:48 Deewiant, nested k *doesn't* make sense 20:17:58 nor does k on stuff like ' or # 20:18:07 it makes a lot of sense, just these stupid delta/pos changers don't :-P 20:18:13 -!- KingOfKarlsruhe has joined. 20:18:14 this is just following the rule set defined and see where it leads 20:18:35 Deewiant, well even if you don't move after you have issues on that 20:19:43 432kkv 20:19:43 zz 20:19:43 z>0"DAB">:#,_a,@ 20:19:43 >0"DOOG">:#,_a,@ 20:19:45 ok that works 20:19:47 for testing it 20:24:31 yay that prints good now 20:33:32 -!- MikeRiley has joined. 20:35:46 MikeRiley, I got some questions for you 20:35:54 that Deewiant said " I don't want to think about it >_<" 20:35:59 432kk# 20:36:00 and 20:36:08 2k'a 20:36:24 MikeRiley, tell me what happens in those 20:36:24 ok,,, 20:36:43 well,,,i would think that 2k' would put two 's on the stack... 20:36:58 yes that is the what cfunge currently does 20:37:03 however. 20:37:17 and in the 432kk i would think this would end up popping a lot of the stack and do nothing else... 20:38:16 about ': "This pushes the Funge character value of the next encountered cell (position + delta) onto the stack" 20:38:23 so yes would push ' here I think 20:38:39 MikeRiley, no I didn't say 432kk 20:38:40 and since position is the k,,,,since the ip is not moved until after k is done,,,, 20:38:40 I said 20:38:44 should push the '.... 20:38:45 432kk# 20:38:48 as in jumping 20:38:53 to be exact 20:39:02 432kk#abcdef123456789 20:39:05 or whatever 20:39:07 i know....but i still think it would do no more than pop a lot off the stack,,,, 20:39:14 since the k executes at the first k... 20:39:23 well yes and no 20:39:26 the 2nd k,,,ip is still pointing at the first k.... 20:39:32 and so will just read the k again... 20:40:33 the ip did not move to the 2nd k.... 20:40:42 now when the k's exhaust the stack,,,, 20:40:51 the ip will move the # and jump to the b.... 20:40:53 http://rafb.net/p/pQcrKo80.html 20:40:55 MikeRiley, ^ 20:40:57 for funge108 20:41:02 looking 20:41:16 -!- pikhq has left (?). 20:41:46 your way of doing in 108 is reasonable...and makes it acuatlly usefull.... 20:42:00 yes and it been a headache really 20:42:13 that is just one apsect of k in funge108 ;P 20:42:17 it fills a page 20:42:27 yeah,,,,,k should have filled a page on the 98 spec.... 20:42:28 this code however: 4k#5 20:42:37 that will certainly jump over the 5 20:42:43 yes.... 20:42:54 -!- oerjan has joined. 20:43:03 so I'm not sure why it wouldn't cause the same on nested k 20:43:14 different situation... 20:43:30 from my understanding of this: 20:43:38 when 4k#5 executes,,,, 20:43:43 ip is pointing at k 20:43:47 anyway fixing the k going downward thing broke the 32kk5 >:#._a,@ test I had... 20:43:52 first iteration, ip is pointing at # 20:44:00 2nd iteration ip is pointing at 5 20:44:09 3rd iteration ip is pointing one cell past 5 20:44:19 4th interation ip is pointing one cell furter... 20:44:30 end of instrucion, k moves the ip and will be one cell further yet... 20:44:57 hrrm 20:45:04 since # moves the ip..... 20:45:19 yes... 20:45:21 it will keep moving with each iteration... 20:45:34 however the last iteration of the second k won't reset the ip 20:45:40 the way I see it 20:45:42 in something like kk...the ip remains pointing at the first k...since no instructi9on is changing ... 20:45:59 yes... and no... :/ 20:46:01 unless it's kk#... 20:46:16 but again,,,the ip is pointing at the first k.... 20:46:21 remember the CCBI style of skip next 20:46:31 *headache* 20:46:32 even when the 2nd k is executed, it is still pointing at the first k... 20:46:41 but the 2nd k then executes the # which moves the ip... 20:46:47 yes 20:47:00 but it is reset at next iteration, as the k executes *at* the k 20:47:00 only after the first k has finished fully iterated and falls through... 20:47:01 remember? 20:47:14 so you will have to reset it when *entering* 20:47:24 this is, the # must execute at the *second* k 20:47:35 even though the second k executes at the first k 20:47:36 :P 20:47:45 for 108, we can say that....the 98 spec does not work that way... 20:48:01 MikeRiley, yes this is figuring out how it would work in 108 20:48:05 ok,,,, 20:48:15 as well as 98 too kind of 20:48:20 in that respect,,,,i think something like 32kk# should execute the #... 20:48:24 that would be usefull.... 20:48:33 where as the 98 semantic is pretty much pointless... 20:48:48 because I think that part (apart from never skip next instruction after k) is undefined in 98 20:48:49 right? 20:49:13 well,,,,the 98 spec does not say that k modifies the ip from its standard way of moving... 20:49:23 indeed 20:49:32 while it makes both # and ' change the way they move 20:49:34 which would mean the instruciton must execute from the k... 20:49:40 yes... 20:50:48 in fact,,,,looking at the spec,,,,,4k'abcd should actuully put abcd onto the stack... 20:50:56 since the ' spec does say that it moves the ip.... 20:51:19 actually,,,it should put 'abc on the stack 20:51:54 good point 20:52:02 and then, by 98, execute the d, leaving the stack as ' a b c 13 20:52:03 actually cfunge may already work that way 20:52:20 since k also does not specify that the ip is reset on each iteartion...and since ' specefies a move,,,just like # does.... 20:52:32 let me try that in Rc/Funge-98....brb... 20:52:56 -!- Hiato has quit ("Leaving."). 20:53:14 $ ./cfunge tests/iterate-fetchchar.b98 20:53:14 13 cba 20:53:17 4k'abcd .,,,a,@ 20:53:20 that was the code 20:53:23 in Rc/Funge, i got cba' on the stack,,, 20:53:30 not what I had expected 20:53:40 wait 20:53:43 with c as top of stack... 20:53:50 $ ./cfunge tests/iterate-fetchchar.b98 20:53:51 13 cba' 20:53:51 from 20:53:56 4k'abcd .,,,,a,@ 20:54:07 one too few , in test program first time 20:54:29 MikeRiley, and no 13? 20:54:33 just use >:#,_ :-P 20:54:34 oh,,,,i have the 13 as well,,,,so 13 cba' 20:54:45 wonder what ccbi does :P 20:54:55 well,,,i guess we agree on that one!!!! 20:55:04 probably the same, apart from skipping over the d currently 20:55:51 MikeRiley, fix all BAD before the weekend (apart from it's incorrect k) ;P then CCBI will for the first time be less standard conforming than RC/Funge 20:55:52 hehe 20:56:05 eheheheeheheheheheheeh 20:56:08 unless you count fingerprints ;-) 20:56:21 well I mean fixing the BAD in fingerprints too 20:56:27 good luck with TRDS ;-) 20:56:31 eheheheheeheheheheh 20:56:32 ah true... 20:56:33 ouch 20:56:36 that will take some time 20:56:41 but then anmaster does not have that one... 20:56:47 indeed I don't 20:56:57 and probably never will if i guess correctly!!! eheheheheheheeheh 20:57:00 MikeRiley, btw when I looked at your fingerprints you seem to implement them as one function with a switch case 20:57:00 yeah, but we were comparing CCBI and RC/Funge-98 and not cfunge ;-) 20:57:13 yes, that is how they execute... 20:57:19 thought about changing it.... 20:57:21 MikeRiley, interesting way of doing it 20:57:25 I use function pointers 20:57:36 that is what i though of changing too...funtion pointers... 20:57:40 I also auto generate the list of available fingerprints from spec files 20:57:43 to generate an array 20:57:49 well, RC/Funge-98 still has the incorrect way of doing fingerprints, no? 20:57:56 yep... 20:58:01 the spec files can also be used to generate the initial template C files for the fingerprints 20:58:20 may make a compatability mode,,,kinda like the -Y for that,,,since i prefer the method that it currently uses... 20:58:38 I'd prefer it if you defaulted to spec compatibility, though 20:58:45 to me,,,the fingerprint mechanism in Rc/Funge-98 seems more usable... 20:58:45 MikeRiley, http://rafb.net/p/dHBaNm41.html 20:58:49 I have a bash script parsing that 20:59:05 to generate the array that fingerprint loading searches through 20:59:19 url need to be updated yes 20:59:41 interesting the way you do that... 21:00:03 sed 's|http://web.archive.org/web/20020816190021/http://homer.span.ch/~spaw1088/funge.html|http://www.elf-emulation.com/funge/rcfunge_manual.html|' -i *.spec 21:00:04 fixed :P 21:00:09 back when Rc/Funge was written,,,my programming skills were a bit more primitive...so there are lots of things that i would change now.... 21:00:16 then: 21:00:17 tools/gen_fprint_list.sh 21:00:25 to update the list 21:00:42 // FILE - File I/O functions 21:00:42 { .fprint = 0x46494c45, .uri = NULL, .loader = &FingerFILEload, .opcodes = "CGLOPRSW", 21:00:42 .url = "http://www.elf-emulation.com/funge/rcfunge_manual.html", .safe = false }, 21:00:44 there we have it :) 21:01:59 and now I uploaded yet a newer version of the funge-108 draft 21:02:18 will have to grab that... 21:02:56 got it.... 21:11:45 problems in FTH module are now all fixed... 21:11:51 FRTH module that is... 21:12:54 MikeRiley, you have put this up somewhere? 21:13:04 I mean version control system or so 21:13:34 not on the web, only within my own dev system... 21:13:37 it seems too few are keen on using public version control :) 21:13:44 well what one do you use? 21:13:46 I would, but I lack a server 21:13:47 cvs 21:13:51 oh my 21:13:55 why cvs? 21:14:10 that is the one that i learned years ago...and have never had a need to change it... 21:14:40 if you would like the current set of sources,,,i just need your email and i will send them to you... 21:21:24 well no need 21:21:32 I would, but I lack a server 21:21:39 Deewiant, what about some public hosting place? 21:21:45 too paranoid ;-) 21:21:50 there are such for bzr, svn, cvs and git at least 21:22:01 I'm sure there is one for mercurial too 21:22:06 not sure what one 21:22:11 there is 21:22:25 oh? 21:22:34 forget what it's called but there is :-P 21:22:39 I think there might even be two, actually 21:22:48 http://www.selenic.com/mercurial/wiki/index.cgi/MercurialHosting 21:23:14 -!- CakeProphet has quit (Remote closed the connection). 21:25:08 Deewiant, nothing as good as there is for bzr though 21:25:15 bzr got the very nice launchpad 21:25:24 svn and cvs have sourceforge 21:25:29 haven't really looked at any of them at all 21:26:30 Deewiant, most seem to have an commercial aspect 21:26:52 unlike launchpad (which offers bzr hosting). but launchpad got Ubuntu behind it 21:26:54 so not odd 21:26:56 bitbucket is the one I was thinking of 21:27:02 and freehg is the other 21:27:22 Deewiant, doesn't mercurial work on a plain http server? 21:27:23 IIPC fixed now... 21:27:23 bzr does 21:27:32 no, bzr and darcs only 21:27:40 Deewiant, ah. bad luck 21:27:45 or, well 21:27:52 if you have CGI support 21:27:54 then, sure 21:27:56 because that is one of the nicer features 21:28:06 but only bzr and darcs can serve with only static pages 21:28:07 right, but I mean without cgi 21:28:19 Deewiant, bzr is very nice btw IMO 21:28:26 it's kind of annoying that bzr needs the client to be able to see the source, though :-/ 21:28:29 even though I could set up more complex stuff on my server 21:28:34 I got root there (duh) 21:28:59 Deewiant, well yes I may fix the webui stuff tomorrow 21:29:21 try to get loggerhead working 21:29:22 anyway 21:29:28 it is mirrored on launchpad 21:29:33 launchpad has loggerhead 21:30:22 Deewiant, https://code.launchpad.net/~anmaster/cfunge/main 21:30:36 doesn't have the last changes though 21:30:43 Last mirrored: 5 hours ago 21:30:43 Next mirror: in 10 minutes 21:33:42 -!- oerjan has quit ("Good nizzzz"). 21:33:57 Deewiant, still it allows you to browse the code 21:34:25 Deewiant, MikeRiley: http://bazaar.launchpad.net/%7Eanmaster/cfunge/main/files 21:34:33 still not last version mirrored there 21:35:39 looking... 21:36:02 who are you, mike? 21:36:15 RC/Funge-98's author 21:36:19 nice 21:36:31 MikeRiley, next update of that mirror will happen in a few minutes it seems 21:36:32 nite 21:36:34 -> 21:36:39 oklopol, heh cya 21:37:46 oklopol! 21:37:52 its only 4:30! 21:37:55 you cant go to sleep! 21:37:57 .. 21:38:00 -!- psygnisf_ has changed nick to psygnisfive. 21:38:26 psygnisfive, I'm going to sleep soon too 21:38:35 wtf?! 21:38:36 still early afternoon for me... 21:38:45 how are you going to sleep so early?! 21:42:01 http://news.ycombinator.com/item?id=253693 21:42:02 lol 21:42:04 oh kids 21:43:04 Like it is small, so I should imagine it is fast? So like maybe for large problems involving lots of "number crunching"? 21:43:34 MikeRiley, 22:43 here 21:43:35 gah 21:43:36 So... The compiler itself is insanely efficient (I have heard that it is on the order of ~200 bytes). 21:43:46 psygnisfive, learn about timeszones 21:43:48 13:43 here... 21:43:56 anmaster: learn about humor 21:44:01 for (;;){} is pretty small too, but that doesn't mean it's efficient >_< 21:44:04 also, learn about formal languages 21:44:06 psygnisfive, error 21:44:17 humor is undefined 21:44:21 where is "here" anmaster??? 21:44:28 has anyone ever tried write a forth program backwards? 21:44:29 MikeRiley, Sweden 21:44:32 MikeRiley, and you? 21:44:41 las vegas,nevada,usa.... 21:44:44 ah 21:45:06 when Rc/Funge-98 was originally written, i was living in zurich,switzerland... 21:45:09 psygnisfive, eh? you mean as in piping one into rev? 21:45:19 er.. "into rev"? 21:45:23 $ echo "abc" | rev 21:45:24 cba 21:45:24 see? 21:45:29 yes, sure. 21:45:43 programming it like its lisp, but reading it backwards. 21:45:45 except: 21:45:46 $ echo -e "abc\ndef" | rev 21:45:46 cba 21:45:46 fed 21:47:08 tusho, you forgot to join #eso btw 21:47:16 :o 21:47:17 whats #eso? 21:47:30 Esoteric Standard Organisation 21:47:41 ask tusho 21:48:56 MikeRiley, anyway switzerland... Wouldn't that be same timezone as me? 21:49:23 GMT+1 21:49:30 currently CEST 21:49:34 due to summertime 21:49:41 (or as they say in US, daylight saving) 21:49:46 what does ESO do 21:49:48 yep,,,it was gmt+1 21:50:20 http://eso-std.org/ 21:50:28 have a homepage without info on it 21:50:32 that's what they do :P 21:50:33 theres no info there 21:52:06 psygnisfive, indeed not 21:52:18 psygnisfive, think ISO for esoteric languages 21:52:23 that is the plan anyway 21:52:33 how will THAT work? 21:52:44 psygnisfive, well I'm eager to find out too 21:52:52 bak 21:52:53 psygnisfive, helping with standards anyway 21:52:55 tusho 21:52:57 explain ESO 21:52:58 tusho, #eso 21:53:03 psygnisfive: no 21:53:04 AnMaster: no 21:53:05 tusho, you forgot to join 21:53:08 AnMaster is wrong 21:53:12 tusho, what? 21:53:24 everything you've said about eso in the past few minutes has been wrong 21:53:39 tusho, it was the way you described it to me 21:53:43 * AnMaster shrugs 21:53:46 anyway night 21:53:48 can 21:53:50 't recall that 21:53:52 but I was wrong, then 21:53:57 tusho, whats ESO do 21:54:52 psygnisfive: long story. 21:54:57 ive got time 21:55:15 i've not got the patience to type out all of it atm :) 21:55:20 :P 21:55:29 well when you do, maybe you should put it up on eso-std 21:55:37 that way you'll never have to type it again 21:56:40 -shrug- 21:59:32 -!- oklopol has quit (Connection timed out). 22:00:15 down to only 8 bads left... 22:02:09 -!- Corun has quit ("This computer has gone to sleep"). 22:02:47 MikeRiley: Funge-98 is dead! Long live Funge-98! 22:15:46 -!- CakeProphet has joined. 22:17:02 {Like it is small, so I should imagine it is fast? So like maybe for large problems involving lots of "number crunching"?} is the best thing I've heard all week, fwiw 22:17:03 -!- Corun has joined. 22:34:05 -!- timotiis has joined. 22:57:46 -!- MikeRiley has quit ("Leaving"). 23:20:15 -!- Corun has quit ("This computer has gone to sleep"). 23:25:44 -!- tusho has changed nick to universe. 23:26:07 -!- universe has changed nick to tusho. 23:29:35 -!- KingOfKarlsruhe has quit (Read error: 104 (Connection reset by peer)). 23:31:01 -!- Corun has joined. 23:39:21 -!- timotiis has quit ("leaving").