00:08:55 -!- tusho has quit. 01:12:33 -!- CO2Games has joined. 01:26:26 -!- CakeProphet has joined. 01:28:01 Oh great prophet, will I get any cake today? 03:05:57 -!- Corun has quit ("This computer has gone to sleep"). 03:07:55 Horray for brainstack 03:08:23 It has FUNCTIONS :D 03:08:32 And MACROS 03:08:58 #i10{++++++++++} 03:08:59 !i32{%i10%%i10%%i10%++} 03:09:00 (,[*i32**i32**i32**i32*.0,]) 03:09:47 # declares a macro, % inserts it 03:10:34 you should only be able to define functions and macros to symbols 03:10:43 or else it isnt esoteric enough :P 03:10:48 ! declares a function, * calls it 03:10:58 pfft 03:11:19 or i guess you could just compile into brainfuck 03:11:57 why D: 03:14:06 it turns into !i32{++++++++++++++++++++++++++++++++} 03:14:08 (,[*i32**i32**i32**i32*.0,]) 03:15:50 you could really just combine function and macro 03:15:58 if all a function does is textually inserts its parameters 03:16:15 which is what it appears to do 03:17:07 what? 03:17:30 Macro is just for cleaning up code 03:17:52 functions actually CALL and RET 03:18:11 BSFunci32: 03:18:11 MOV EAX,[pos] 03:18:11 INC [tape+EAX] 03:18:12 ... 03:18:13 MOV EAX,[pos] 03:18:14 INC [tape+EAX] 03:18:15 RET 03:18:16 start: 03:18:31 NOP 03:18:32 MOV EAX,[pos] 03:18:33 CMP [tape+EAX],0 03:18:34 JE BSWhile1E 03:18:35 BSWhile1B: 03:18:36 CALL BSFunci32 03:18:38 CALL BSFunci32 03:18:38 CALL BSFunci32 03:18:55 CALL BSFunci32 03:18:55 NOP 03:18:56 NOP 03:18:56 NOP 03:18:57 MOV EAX,[pos] 03:18:58 CMP [tape+EAX],0 03:18:59 JNE BSWhile1B 03:19:00 BSWhile1E: 03:19:15 end start 03:21:31 the question is whether a macro can do anything a function cannot 03:21:57 (functions could recurse, i hope) 03:31:47 If assembly can recurse functions, so can this 03:33:29 Unfortunately functions don't get their own tape. 03:34:09 so then any macro could be assembled as if it was a function instead, and the behavior of the program would still be the same 03:34:36 Wait, functions can't recurse? 03:34:48 sure, i think 03:34:59 How would a macro recurse? 03:35:14 it cannot. but it cannot do anything a function cannot, either 03:35:30 true... 03:35:48 macros are essentially #defines in c++ 03:36:06 Except they don't take parameters 03:36:09 so i thought 03:37:01 Although they can pop from the stack, and push back onto it. 03:37:04 you could let them take parameters, that would be a difference. easier to implement than for functions, too 03:37:23 hmm... 03:37:25 (i think) 03:37:52 I could give functions their own tape to work with. 03:38:26 Implementing parameters on macros sounds interesting 03:39:19 Although I'm not sure how they would work 03:40:52 I mean macros are exactly the same except they are inline instead of seperate code. 03:42:11 Here's an interesting ordeal... 03:42:56 ...how would parameters be passed to the macro? 03:43:09 Obviously not through the stack like functions. 03:43:26 they would be expanded inline of course 03:43:39 expanded inline? 03:43:48 like C's #define is one way 03:44:10 There isn't a parameter list for brainstack macros though 03:44:39 so it would require some added syntax 03:44:48 hmm... 03:45:13 ok, so how would these parameters be accessed? 03:45:39 inside the macro I mean 03:45:39 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | ...if I'm not identified.. 03:46:05 C's #define does it by name, you might access them as if they were macros themselves. or you could do it by argument position. 03:46:37 -!- CakeProphet has quit (Read error: 110 (Connection timed out)). 03:47:40 -!- oklobol has quit (Read error: 60 (Operation timed out)). 03:47:49 oh, so something like %macroname|9|+% for the call? 03:48:14 perhaps 03:48:37 yeah that would be the wiki style wouldn't it 03:48:54 (mediawiki templates) 03:49:11 (or resembling it) 03:54:53 and then maybe something like #macroname{@looper$u:@@MOV ECX,$0@+@loopc looper@} 03:55:32 $u for a unique id, incrementing every time the macro is used 03:55:42 So we don't get label conflicts 03:55:58 $0 for the 0th argument... 03:56:00 oh shit 03:56:11 #macroname{@looper$u:@@MOV ECX,$0@$1@loopc looper@} 03:56:12 there 03:56:51 shouldn't there be a $u at the end too 03:57:01 so %macroname|9|+% increments the ...oh right 03:57:09 #macroname{@looper$u:@@MOV ECX,$0@$1@loopc looper$u@} 03:57:20 so %macroname|9|+% increments the cell 9 times 03:57:53 %macroname|57|V% pops the stack 57 times 03:58:47 @ surrounds inline assembly 03:58:52 Which I should add right now 04:00:50 ok, added 04:00:51 case '@':fputs(gets_until('@'),asmout);break; 04:01:35 gets O_O 04:02:07 * oerjan runs away screaming 04:02:27 I wrote a function that uses getchar and keeps doing that until it gets the given character 04:02:53 And then dumps it out as a string 04:03:00 Err wait no it does fgetc 04:03:22 but where are you putting it? gets is just begging for buffer overrun 04:03:46 what? 04:04:01 It's used by the converter 04:04:22 how is the string allocated, i mean 04:04:35 char buffer [1024] 04:04:52 Yeah yeah I know I know 04:05:20 blah blah string std vector blah buffer overflow blah segfault win32 blah blah access violation, I GET IT. 04:05:39 I chose char arrays, and I stuck to it 04:06:10 But hey, I'm using vectors for the macros, and strings for the macro contents 04:06:29 uhh except the names 04:06:33 the names are 256 chars 04:10:16 * oerjan does a little segfault da 04:25:11 -!- calamari has joined. 05:06:55 Holy shit, can I eat you? 05:10:13 calamari: What's your involvement with Qemu? 05:11:13 hi 05:11:16 -!- megatron has joined. 05:11:32 calamari: Oh yeah ... blah blah blah pleasantries etc :P 05:12:05 GregorR: I use it.. and they needed someone to host the FAQ.. so I made it into a wiki, and it went from there 05:12:22 I haven't contributed much content.. mostly grammar and link fixing 05:12:25 Okidoke. Just had seen the URL and was wondering :P 05:13:13 although I had to use non-free virtualbox to communicate with my timex data link watch :( qemu usb kept giving data errors 05:13:26 Durn :( 05:13:40 I gave it a good try tho 05:13:51 qemu is faster 05:14:03 I'm using an array of Qemus to build things across multiple platforms automatically ^^ 05:14:18 it installed windows 98 at least 3 times as quickly as vbox 05:14:24 as->more 05:14:44 need to port bf to my watch 05:14:50 Whoot. 05:15:02 125 bytes of ram 05:15:28 plus whatever I can reclaim from my 900 bytes of program code 05:15:43 125 bytes D-8 05:16:35 well, bf interpreters aren't terribly large 05:16:46 so I'm hoping to get at least another 500 bytes 05:16:47 -!- oklobol has joined. 05:16:54 Is the SDK actually provided by Timex, or is it a user hack? 05:17:03 provided, amazingly 05:17:10 That is amazing. 05:17:14 My watch is nonprogrammable :( 05:17:29 and they even coded a build system.. windows only, but I guess the watch was codeveloped my microsoft so I can't blame them 05:17:33 my->by 05:17:56 I guess the earlier ones even had microsoft printed on the face 05:18:14 maybe that's why I didn't look into it.. hmm 05:27:16 My watch is ultrapowerful, but still no SDK *sigh* 05:27:26 which? 05:27:37 -!- moozilla has quit (Read error: 110 (Connection timed out)). 05:28:07 http://www.chinavasion.com/product_info.php/pName/super-cool-mobile-phone-wrist-watch/ 05:28:35 you got one of those? hahaha 05:29:08 does it work well as a phone? 05:29:13 Extremely. 05:29:20 I consider it to have been an excellent buy. 05:29:30 neat.. whats the typical battery life like? 05:31:11 One day with moderate phone use, two days on average. 05:31:30 It comes with two batteries, which is nice since you can just swap them out whenever and never have the phone off for long. 05:31:49 anyone working on hacking it? 05:32:02 Not to my knowledge. 05:34:29 well it might be possible.. does it allow firmware upgrades? 05:40:27 i need a hash function that incorporates something like levenshtein distance 05:40:40 strings with a levenshtein distance of less than n hash the same 05:42:25 or s/levenshten/similar algo/ 05:50:06 OOOOOH, generate the set of strings, sort it and hash the concatenation 05:50:08 of course 05:51:31 um this sounds like everything must hash the same, by transitivity 05:52:04 oh 05:52:05 duh 05:52:06 :( 05:52:43 except not 05:54:22 bbl, tired 05:54:25 -!- calamari has quit ("Leaving"). 05:55:31 -!- GreaseMonkey has joined. 06:05:56 -!- olsner has joined. 06:09:57 -!- oklobol has quit (Read error: 113 (No route to host)). 06:33:47 oerjan: Ok I think everything's working now 06:33:53 But I need to test the converter 06:34:31 aye 06:41:17 ok i was wrong 06:42:58 people often are 06:43:32 i wonder if this is even possible 06:44:18 depends what "this" is 06:45:00 i need to find the fuzzy-union of two sets in O(n) 06:46:58 i don't know what that is 06:49:21 ...i meant interesection 06:50:06 right, there was something intuitively trivial about that union 06:53:29 God damnit 06:53:38 I can't figure out a way to get nested loops working 06:55:26 O(m+n) then for two sets, or what? 06:55:40 My brainstack converter 06:56:29 Deewiant: yeah 06:58:48 unfuzzy intersection is easy, you just use a hash table 06:58:53 -!- GreaseMonkey has quit ("Jesus loves you"). 07:02:29 -!- oerjan has quit ("Good time-of-day"). 07:09:27 -!- megatron has changed nick to moozilla. 07:59:59 -!- clog has quit (ended). 08:00:00 -!- clog has joined. 08:15:32 -!- kar8nga has joined. 09:17:16 -!- GreaseMonkey has joined. 09:29:01 -!- olsner has quit ("Leaving"). 09:45:40 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | And this somehow has to do with /translating/ the specs?. 10:30:07 -!- jix has joined. 10:31:59 -!- GreaseMonkey has quit ("Jesus loves you"). 10:39:37 -!- kar8nga has quit (Read error: 60 (Operation timed out)). 11:52:09 Deewiant, I know this is picking at irrelevant details but... 11:52:11 GOOD: 1y and 5y do not disagree about = 11:52:11 No reliable cross-platform method of testing: assume = works 11:52:24 well you could test it in this case, check that it reflects :P 11:52:32 -!- tusho has joined. 11:52:33 since y indicates it is not implemented 11:52:37 bbl 12:09:25 -!- jix has quit ("CommandQ"). 12:16:26 -!- oklobol has joined. 12:24:13 -!- oklopol has joined. 12:24:13 -!- oklobol has quit (Read error: 104 (Connection reset by peer)). 13:24:55 -!- oklopol has quit (Read error: 104 (Connection reset by peer)). 13:25:03 -!- oklobol has joined. 14:04:41 -!- oklopol has joined. 14:04:49 -!- oklobol has quit (Read error: 113 (No route to host)). 14:29:37 -!- oklobol has joined. 14:30:42 -!- oklopol has quit (Read error: 113 (No route to host)). 14:52:41 -!- puzzlet has joined. 15:21:36 -!- CO2Games has quit ("And I invented doors, no joke!"). 15:45:40 -!- optbot has set topic: the entire backlog of #esoteric: http://tunes.org/~nef/logs/esoteric | what are the default namespace mappings?. 16:25:02 -!- oerjan has joined. 16:25:58 -!- puzzlet has quit (Remote closed the connection). 16:26:02 -!- puzzlet has joined. 16:40:04 -!- Hiato has joined. 16:40:35 -!- Hiato has quit (Read error: 60 (Operation timed out)). 16:42:29 -!- Hiato has joined. 17:00:04 -!- Slereah_ has joined. 17:00:04 -!- Slereah2 has quit (Read error: 104 (Connection reset by peer)). 17:20:29 bbl 17:54:56 heh, Mike is working hard getting Rc/Funge-98 to behave on MycoTRDS the same way CCBI does :-) 17:57:11 -!- Corun has joined. 18:02:16 -!- Corun has quit ("This computer has gone to sleep"). 18:09:35 -!- puzzlet_ has joined. 18:21:47 -!- psygnisfive has joined. 18:21:52 -!- puzzlet has quit (Read error: 110 (Connection timed out)). 18:21:54 hey tusho 18:21:57 hey oklobol 18:22:35 hey oerjan 18:22:38 hey anmaster 18:22:43 hey pikhq 18:22:44 hey psygnisfive 18:22:50 hey deewiant 18:22:52 :P 18:23:30 I guess I don't get a "hey." *sobs uncontrollably* 18:24:06 GregorR: hey. 18:24:18 hey all 18:24:23 done, problem solved. 18:24:52 X-P 18:24:59 ae5ir: who are you 18:25:12 I am the terror that flaps in the night 18:25:14 not really. 18:25:34 used to come in here maybe a couple years ago 18:25:36 as Robdgreat 18:25:38 no? you totally had me going there 18:27:27 No way that was on the order of years. 18:29:06 i dont know you much gregor D: 18:30:28 10/10/05 was when I registered the nick 18:31:53 and iirc this channel was the reason I first came to this network 18:33:44 I am an enigma. 18:35:02 better than an enema 18:35:06 I am an enigma wrapped in a mystery wrapped in questions wrapped in a warm flour tortilla served with chips and guacamole for only $4.95 for a limited time only. 18:35:43 Here's a five! 18:35:50 plus tax 18:36:12 Taxes? 18:36:16 RON PAUL! 18:36:19 PLON RAU! 18:36:32 No sales tax in OR :) 18:36:43 (I'm not in OR, but my point stands) 18:37:17 true? yes. relevant? perhaps not. 18:37:18 ;) 18:37:32 TOUCHÉ SIR 18:37:45 :D 18:50:33 i'll have that and a medium coke, thank you 18:51:04 oerjan: which kind of coke? 18:51:17 zero 18:52:13 it's either that or infinity 18:53:01 oerjan: that is not quite what i was referring to 18:53:19 ah. maybe. 18:53:25 the legal kind. 18:54:02 boring, i know. 18:56:26 also, i am sure i've seen robdgreat earlier this year. i kept confusing him with robert. 18:56:39 who isn't here much either? 18:57:17 -!- olsner has joined. 19:01:09 yeah I've been in here since back when, just not very frequently 19:05:49 I am confusing robdgreat with RodgerTheGreat 19:05:52 rodgort: Ok, who are YOU? 19:06:06 SirDayBat: Or you! 19:06:35 SirDayBat smells finnish 19:06:44 yeah 19:06:57 heh yeah I came back after a long absence and he accused me of stealing his nick -_- 19:07:18 OH YEAH 19:07:21 That's my confusion. 19:07:36 ae5ir: rodgerthegreat is an asshole, so i'm not surprised 19:07:39 s/robert/rodger/ 19:17:01 -!- jix has joined. 19:26:12 -!- oerjan has quit ("subinmo"). 19:40:22 -!- puzzlet_ has quit (Read error: 60 (Operation timed out)). 19:43:02 -!- puzzlet has joined. 19:43:22 -!- Hiato has quit ("Leaving."). 19:58:20 -!- Corun has joined. 20:01:14 -!- Corun has quit (Client Quit). 20:19:23 -!- Corun has joined. 20:21:05 -!- Corun has quit (Client Quit). 20:25:51 -!- atrapado has joined. 20:36:07 Here's what graphviz thinks of fungot's fingerprint-loading parts: http://www.cis.hut.fi/htkallas/bg.png 20:36:07 fizzie: thanks for your help 20:36:26 fungot: I'm not sure that graph will help you any, but sure, no problem. 20:36:26 fizzie: hold on a second 20:38:35 fizzie: did you do that yourself or generate it somehow 20:39:28 I guess there's some generation involved in any case, but you know what I meant 20:41:24 Wrote a bit of Java (since oklopol was talking about it the other day) to do a tiny amount of static Funge source code analysis. 20:42:28 neat - what can it do? checks for # _ or something or is it more clever? 20:42:56 Basically there's a "collect cells to a graph" function which, when coming across things like _| or "# followed by something that might reflect" calls itself recursively to handle the two branches, and then some more code to combine "uninteresting" (single in-edge, single out-edge type) vertices to longer strings. 20:43:39 cool :-) 20:43:49 Currently there's branching only for _| and the #-reflect case, but I guess I'll need to add some sort of heuristic for the "j" jump tables, as well as '?', if I intend to be able to pass whole fungot through it. 20:43:49 fizzie: it's n being shifted by k. dick fellow writes crap. at least i can 20:44:30 Currently it gets confused somewhere in fungot sources. Must fix later. Should go to sleep early today, so maybe tomorrow. 20:44:30 fizzie: don't worry about that stuff. 20:44:43 fungot: You're being surprisingly coherent today. 20:54:06 Wrote a bit of Java (since oklopol was talking about it the other day) to do a tiny amount of static Funge source code analysis. <-- wow? 20:54:39 fizzie, where can you see the source? 20:54:44 find* 20:55:34 fizzie, also you need to handle x and w, + some other 20:56:00 of course when the program is self modifying it will be harder 20:56:10 it's impossible 20:56:27 Deewiant, yes at least unless you can see no code cell get overwritten 20:56:45 and you can't see that, it's impossible. 20:56:51 you could tell the analyser "all stores in negative funge space are for data" 20:56:58 give it that hint 20:57:06 Deewiant: not impossible 20:57:07 superturing 20:57:12 and make the actual source code non-modifying 20:57:53 tusho: i.e. impossible. :-P 20:57:55 I agree there will always be cases it *can't* handle of course 20:58:03 Deewiant: :D 20:58:06 but you can make it more or less good 20:58:11 really want a goddamn oracle machine 20:58:12 also 20:58:14 wanna write stuff with language.c 20:58:15 like 20:58:19 graph of control flow in c 20:58:26 have functions be bubbles, and flowcharts inside 20:58:27 Deewiant, how come a human can work out the flow of such a program then? 20:58:31 are humans super turing? 20:58:33 with little arrows pointing to other bubbles and back again 20:58:33 -!- oklobol has changed nick to oklopimp. 20:58:35 into the flowchar 20:58:37 t 20:58:38 AnMaster: no 20:58:38 or do we just execute them in our head? 20:58:40 AnMaster: a human can't do it either. 20:58:42 AnMaster: no, we just have reasoning skills 20:58:45 they are not perfect 20:58:49 tusho, indeed 20:58:50 we cannot do it for all cases 20:58:54 very true 20:58:59 and most likely our reasoning could be simulated by a computer 20:59:00 and most likely humans do execute them in their heads. 20:59:01 we just don't know how 20:59:04 but yes 20:59:08 I know I do. :-P 20:59:09 we execute in our heads in a limited form 20:59:23 yes, follow the code flow to see what happens 20:59:57 what about "formally verified" programs. Obviously someone must have reasoned a lot about those? 21:00:17 no clue how you actually "formally verify" a program 21:00:23 i wonder if on any other media there's been as much leasurely chatter about turing completeness as in here 21:00:47 AnMaster: they are just theorem proving systems 21:00:51 the theorems could be written badly 21:00:58 or you can prove something yourself 21:00:59 -> same problem 21:01:05 Deewiant: that's even more flawfilled though 21:01:12 flawfilled? :-P 21:01:16 Yes, I'm just trying to make it good enough to handle fungot, not any sort of funge-98 program. I don't think I'll even do 'x' (even though in many cases a simple constant-folding-type of thing would work) since I don't use it. 21:01:35 those c graphs would be win 21:01:35 <3 21:02:21 AnMaster: formally verifying is proving things work as they're supposed to 21:02:32 proving is a vast subject with tons of techniques 21:02:45 some of which you most likely are aware of 21:02:52 induction, for instance 21:03:04 that's one of the techniques he doesn't understand 21:03:10 >_< 21:03:26 * oklopimp is a confident guesser 21:03:37 agreed with oklopimp on the not understanding 21:03:37 :P 21:04:42 Deewiant: induction isn't really something many people use in their thinking unless been taught to 21:05:15 induction, for instance <-- certainly, I used it in math in school 21:05:22 (i invented it for a math test once, but well i'm a pimp that's just how i roll) 21:05:28 damn 21:05:31 i was wrong 21:05:38 at least if it is what I think it is in Swedish 21:05:45 "bevis genom induktion" 21:05:57 it's what you think it is 21:06:17 some people do not understand induction 21:06:23 therefore AnMaster does not understand induction 21:06:59 like, f(n) = .... for any integer n always returns (n-1)/n, then you prove that it is true for n = 1, and then that if it is true for n = 1, then it have to be true for any integer n 21:07:17 err 21:07:30 err? 21:07:32 more like if it's true for all integers up to a certain number, it will be true for that number + 1 21:07:54 oklopimp, right, what I meant. Except I'm just home from Aikido, so kind of sleepy 21:08:12 yeah, i'm sure you get it, i just found your explanation confusing 21:08:28 aaaaanyway, induction is one of the ways you can verify algos 21:08:35 oklopimp, probably since I'm almost sleeping on the keyboard 21:08:50 so heading to bed shortly 21:08:57 have fun, i just slept like 4 hours 21:09:18 oklopimp, well did you just came home from a two hour session of Aikido? 21:09:35 no, i just woke up from a 4 hour sleep 21:09:40 ah well 21:09:40 i did aikido as a kid 21:09:42 it was fun 21:09:45 i kicked ass 21:09:47 then i quit 21:09:50 see you later -> 21:09:55 oklopimp, you quit? why? 21:10:10 (actually i guess you don't kick much ass in aikido) 21:10:18 oklopimp, indeed you don't 21:10:23 i quit because i was a kid and i got bored with things 21:10:27 shop -> 21:10:36 I suspected you meant it metaphorically(sp?) 21:10:45 Deewiant, can you explain why Finnish ppl seem to use "->" for "afk"? 21:10:51 at least all in here 21:11:19 AnMaster: stop ripping me 21:11:20 i asked that first 21:11:21 like 5 times 21:11:27 tusho, did you get any answer? 21:11:33 i think its just the culture, from what i gather 21:11:38 tusho, and yes I know you asked too 21:11:39 irc started in finland, right? 21:11:45 it did 21:11:48 so it just kind of grew, someone used it, it continued, but then we imported our chatspeak 21:11:53 and ... the divide just kinda grew 21:12:00 so finnish irc channels use it 21:12:01 anyway then I'm going to start using "<=" at the beginning of lines or something 21:12:02 but english ones don't 21:12:07 oooh no I got a better one 21:12:08 so when finns are on english ones 21:12:10 they use what's natural to them 21:12:12 so they use -> 21:12:13 brb 21:13:53 well I'm heading to bed now 21:13:54 Exeunt 21:15:14 Hey, the funge-graphery actually helped me notice a bug: there in ^save if 'O' fails, the code will go nowhere. 21:16:25 fizzie, ok, pastebin the code please 21:16:35 does it have some nice GUI btw? 21:17:34 There's a lot of files, and no GUI, and I won't be sharing the code just yet, still playing with it a bit. 21:17:52 fizzie, aww, would like to see it 21:18:20 -!- jix has quit ("This computer has gone to sleep"). 21:19:00 Are you sure? It's Java, of all things. 21:19:19 fizzie, well I got a jdk 1.6 around 21:19:24 for other apps I need 21:19:37 no I don't like coding java, but I accept using it 21:19:55 it is not my primary choice no, but if that app is good it is ok 21:20:21 (basically I need it to run a level editor on a game where I develop on the server code, which is in C) 21:21:13 Okay, I've got a 4920x15464-pixel image of fungot sources made with that thing; it might be the buggy. http://www.cis.hut.fi/htkallas/fungot.png 21:21:41 warning do not open in firefox 21:21:58 it crashed 21:21:58 I just opened it in firefox. 21:22:11 fizzie, firefox 2.x or 3.x? 21:22:25 3.something. 21:22:36 ah still 2.x due to stable in distro 21:22:44 3.0.2, OS X. 21:23:01 fizzie, there are a lot of empty bubbles? 21:23:27 Yes, I think it's because the node-combinatator removes <>^v from those nodes, but isn't intelligent enough to remove empty nodes. 21:23:28 also I would like to try it on some smaller program 21:23:30 to understand it 21:23:31 Must fixery that. 21:24:15 fizzie, what about http://catseye.tc/projects/befunge93/eg/life.bf 21:24:28 no clue if it is too self modifying 21:24:49 since I *can't* reason about that program 21:25:37 http://www.cis.hut.fi/htkallas/life.png says dot. 21:26:05 fizzie, where is the *start* node 21:26:27 There in the middle. 21:26:35 It's the one with a incoming "->" from nowhere. 21:26:39 fizzie, well that doesn't make it a lot clearer 21:27:11 fizzie, what about the program in http://bzr.kuonet.org/cfunge/trunk/annotate/418?file_id=count.b98-20080502072851-9xxna708q7kcdybc-1 21:27:14 Must remove those empty nodes, it'll be better then. And neato would produce nicer graphs except that it needs some serious adjustments to make nodes not overlap. 21:27:14 it is html that link 21:27:20 fizzie, ^ 21:27:33 very simple program 21:28:37 however it is golfed for space 21:28:49 more or less 21:28:53 count.png there. 21:28:55 could be smaller I bet 21:29:28 -!- Corun has joined. 21:29:38 fizzie, ah it fails at program wrapping around edge 21:29:43 which that one does for space reasons 21:30:05 see on line 3, .:< 21:30:09 that one will wrap 21:30:16 wait or does it fail? 21:30:18 hard to say 21:30:32 It should work. 21:30:40 Doesn't that ".:" go to the correct 1+... node? 21:30:49 count2.png is how neato layouts it. 21:30:50 ah yes it does 21:31:05 fizzie, you mean upside down compared to program 21:31:12 I find count.png easier to read 21:31:36 -!- atrapado has quit ("Abandonando"). 21:32:16 fizzie, well I think I can understand count.png since I wrote the original program 21:32:22 dot does hierarchic graphs better, since it gives a "rank" for each node. 21:32:46 fizzie, you mean count.png? yeah count.png is better 21:32:58 straight lines too 21:33:09 I mean vertically 21:33:17 fizzie, try antialias, it would help a lot 21:33:30 on the lines and the circles 21:34:18 If you insist on looking at the sources, they're at http://www.cis.hut.fi/htkallas/bg.tar.gz -- builds with ant, uses Runtime.exec to invoke Graphviz. And is very very incomplete and work-in-progress. 21:34:54 fizzie, I didn't insist at looking at source, but I insisted at running it 21:35:25 BUILD SUCCESSFUL 21:35:26 Total time: 6 seconds 21:35:31 for 6 files heh 21:35:34 err 21:35:35 8 files 21:35:49 fizzie, how do I run it? there is no jar file? 21:36:32 Haven't bothered to add a "dist" target to the build.xml. 21:36:48 Just do "java -cp build befgraph.BefGraph program.b98 output.png" in the top dir. 21:37:11 It is, after all, not a distribution yet. 21:37:35 wow 21:37:40 I just found a bug in cfunge... 21:37:45 wtf 21:37:46 * AnMaster fixes 21:37:58 it ignored - for integer input... 21:38:43 wait 21:38:48 so does ccbi? 21:39:00 Deewiant, are implementations supposed to ignore - in &? 21:39:08 Please input a number: -4 21:39:08 UNDEF: got 4 which is hopefully correct. 21:39:11 for both cfunge and ccbi 21:39:29 Please input a number: -4 21:39:29 UNDEF: got -4 which is hopefully correct. 21:39:31 for efunge 21:39:50 fizzie, which is right I wonder? 21:40:28 fizzie, lets see what happens on mycology 21:40:38 $ file mycology.png 21:40:38 mycology.png: PNG image data, 1859 x 1832, 8-bit/color RGBA, non-interlaced 21:40:48 That's surprisingly small. 21:40:58 lots of nodes on each other 21:41:10 Uh, I think I left "neato" there as the layout engine. 21:41:20 fizzie, where would I change it? 21:41:22 You can change it in BefGraph.java, just s/neato/dot/ 21:42:00 fizzie, it seems to be antialiased at least 21:42:02 - is a digit, so I guess "4" is more right: & specs say "reads and discards characters until it encounters decimal digit characters". 21:42:42 fizzie, well hm 21:42:50 I think the graphviz version at the SuSE box I was running it on might be a bit old and not support any antialiasationary stuff. Not sure. 21:42:53 $ file mycology.png 21:42:53 mycology.png: PNG image data, 1856 x 5019, 8-bit/color RGBA, non-interlaced 21:42:58 smaller? 21:43:02 wait 21:43:05 very very wide? 21:43:16 -!- oerjan has joined. 21:43:16 ah no tall 21:43:24 so that is Y x X 21:43:26 not X x Y 21:43:28 weird 21:43:46 fizzie, my dot seems to do antialiasing 21:43:47 weird 21:44:01 looks way nicer than the output you got 21:45:00 fizzie, http://omploader.org/vcjZl 21:46:50 fizzie, somehow I don't think that graph is complete 21:48:42 The dot I have at work is very very old, but since I started that program there, I've been running it there too. 21:48:58 And it might easily get confused by unexpected code flow. 21:49:10 fizzie, better check where it gets confused 21:49:16 I want to see everything up to y :D 21:49:53 oh wait it hits the "p does not modify space" I think 21:50:19 well you can't solve that 21:50:31 Yes, self-modification is no good. 21:50:46 Also it doesn't do x and ; yet; at least ';' is used in mycology. Although ; is easy. 21:51:00 yes it is 21:51:10 fizzie, what about []? 21:51:29 also the p check is way earlier 21:51:34 Those neither. Hey, fungot graph must be also semi-broken because of that. 21:51:42 There's at least one ] in there. 21:51:42 fizzie, w? 21:51:51 Nope. Don't use 'w' in fungot. :p 21:51:58 why not? 21:52:00 I like w 21:52:01 very useful 21:52:17 Sure, I just have my Befunge-93 roots. 21:52:31 fizzie, if you use [] then...? 21:56:37 That was just because I had to rearrange some code and got a >< pair that would've needed to overlap that was solvable with a single ]. 22:00:06 fizzie, err I mean those two instructions 22:00:12 not that specific construct 22:00:58 Yes, I had a "coming from up, do <; coming from down, do >" case which I did with a ]; I think that's the only place where fungot uses either [ or ]. Although there might be a second place somewhere. 22:01:50 I rather like the generated fungot.png scaled to 25% of the size. It sort-of looks very enterprisey. Now if I could just fix the empty-node-removal. I thought I got it right, but now it got confused because an empty node had two outgoing edges, which shouldn't really be possible. 22:02:43 http://rafb.net/p/xVQ6FN79.html <-- hm nice format 22:04:48 fizzie, well I'd like to see the static analysis as complete as possible 22:04:51 would be very nice 22:06:24 I'll do some improvements, but I doubt I'll care enough to start analyzing the state of the stack, for example. I did do that once for a Befunge-93 compiler which did constant-folding, though. 22:06:49 fizzie, nice! 22:07:07 fizzie, except well that would not work for self modifying programs 22:07:13 fizzie, but I'd like to see the code 22:07:30 I think you could JIT it though 22:08:14 I'm not sure where I have the code. 22:09:09 It was also written in Java, and it was supposed to target the Java VM so that it could recompile stuff when self-modification occurs. Got bored midway through and only did a proof-of-concept code-generation backend that generated C code. 22:10:00 ehehe? 22:11:05 fizzie, also most funge programs tend to use self modification to store variables at least 22:11:24 so lazy recompilation 22:12:02 fizzie, in fact I planned to rewrite that count program to use self modification and a ' 22:12:58 That's curious, I enabled the empty-node filtering and fungot.png grew in size. 22:14:17 I guess it's because I also added [ and ] instructions. 22:15:09 fizzie, where can I check out last version? 22:15:51 I got bzr, hg, darcs, svn, cvs, git and monotone installed 22:16:02 i still wanna make that c control flow graph it would pwnz0r 22:16:24 tusho, would certainly be interesting, but aren't there already such things? 22:16:30 probably 22:16:32 don't care 22:16:34 mine is better 22:16:43 tusho, go code it, maybe after botte? 22:17:18 i don't actually give a shit about botte right now since nobody is actually in a situation requiring it 22:17:22 i might code it if i can be bothered, that would be something 22:20:00 -!- ihope has joined. 22:23:04 AnMaster: I guess you can do a "git clone http://www.cis.hut.fi/htkallas/git/befgraph" but don't expect updates very often. 22:23:14 I guess I'll sleep now. 22:23:22 warning: AnMaster hates git 22:23:36 Well, git was listed there in the list of acceptable methods. 22:25:01 hm 22:25:08 tusho, I accept using git if I *have* to 22:25:10 so ok 22:25:24 * tusho points a gun at AnMaster's head 22:25:28 Use git for all your projects. 22:25:32 never 22:25:34 Or- bang. 22:25:34 and busy 22:25:50 I guess death doesn't fall under the umbrella of have to. 22:27:41 * AnMaster points a gun at tusho's head 22:27:49 Never use git again, use mercurial or bzr! 22:27:54 Or- bang. 22:27:58 Sure. 22:28:04 really? 22:28:13 AnMaster: No, because you didn't really point a gun at my head. 22:28:20 But I was acting in the context of the fictional gun-pointing 22:28:23 ah 22:28:24 as IRC is a virtual environment 22:28:29 If you pointed the gun at my head in real life, yes, of course I would. 22:28:33 that's what you _think_ 22:28:38 tusho, well I was breaking the fifth wall before ;P 22:28:40 oerjan: dun dun dun 22:28:44 AnMaster: Fourth wall. 22:28:44 oerjan, no no 22:28:49 and no tusho 22:28:51 maybe sixth 22:28:58 not not fourth 22:29:23 tusho, a channel like #esoteric got a non-trivial space time geometry ;P 22:29:27 * oerjan is starting to wonder if the room is 3d at all 22:29:31 (warning: probably technobable) 22:29:39 i don't think #esoteric obeys any kind of geometry 22:29:42 it's all sort of wishy-washy 22:29:46 also 22:29:59 there are things i wouldn't do even with a gun pointed at my head 22:30:05 -!- oklopimp has changed nick to oklopol. 22:30:06 oerjan, s/you _think_/she said/ 22:30:08 it's a Hausdorff compactum. we are still unclear on whether it is metrizable. 22:30:51 um wait 22:31:01 Hausdorff compactum? 22:31:03 s/compactum/compact space/ 22:31:22 ah well I can kind of guess now what thing it may be 22:31:25 i forgot the slight difference 22:31:37 oerjan, aspell doesn't like metrizable, please explain it 22:32:26 a topological space is metrizable if its topology can be got from some possible distance metric 22:33:02 (metric = distance function, essentially) 22:33:19 ah well that clarified one, no two, things at least: 1) we were talking about topology as I thought 2) You know way more than I do about it 22:33:29 oh and 3) I didn't get what you just said :P 22:33:32 * oerjan cackles evilly 22:34:05 i'm not entirely sure what he's saying either 22:34:31 but i'm hoping ihope is the only other guy here who understood that 22:34:50 Hum. 22:34:59 there are things i wouldn't do even with a gun pointed at my head 22:35:00 name some 22:35:01 well it was just an attempt at technobabble. the fact that the terms actually exist should not detract from that point. 22:35:52 oerjan: can I have some technobabble on the subject of oko 22:35:57 So, it's a Hausdorff compact space and we don't know whether it's a metric space thingy. 22:36:06 yeah 22:36:34 tusho: oko calls more for spiritual babbling, i feel 22:36:42 I think I've forgotten what Hausdorffness is. It's a separation axiom, I know. 22:36:48 oerjan: do that then 22:37:00 ihope: any two points are in disjoint open sets 22:37:09 Ah, disjoint. 22:37:27 I don't know if there's any particular Hausdorff compact space whose metrizability is unknown. 22:37:43 tusho: the essence of oko is in a transcendental logosphere of infinite reproduction 22:37:45 tusho: i would never kiss a guy in public, for instance 22:38:11 two guys kissing with a gun pointed at one of their heads 22:38:15 that would be ... an odd situation 22:38:21 avoiding death does not have an infinite priority for me. 22:38:41 Does asiekierka have an email address out there somewhere? 22:38:46 oklopol: I imagine it does, instinct-wise. 22:38:55 well okay i guess that's a bad example since i've kissed a guy while drunk once 22:38:58 but 22:39:06 if there was something i really didn't wanna do, i wouldn't do it 22:39:37 for instance, if someone pointed a gun at me and told me to tell them they're more intelligent than me, i probably wouldn't. 22:39:56 ihope: i'm sure there is something involving extra unprovable ZFC axioms or something. 22:40:55 I like adding true-but-unprovable axioms to things. 22:41:10 i like adding your mum to things 22:41:13 ihope, how do you know they are true then? 22:41:24 axioms aren't true 22:41:38 they're nothing 22:41:45 AnMaster: by the lack of a counterexample. 22:42:11 ihope: his nick at gmail.com popped up on the third google page 22:42:28 I'm not sure that makes it true, only a relative probability that it is in fact true 22:42:54 nothing can be proved, there can always be a mistake in the proof 22:43:28 but other than that, you seem to have misunderstood everything about math, AnMaster 22:43:49 oklopol, hah 22:44:01 if you had to guess (with a gun at your head), would you say that the Riemann hypothesis is 1) false 2) true and provable 2) true but unprovable 22:44:22 -!- CO2Games has joined. 22:44:26 yes I realize this is the wrong question to ask. But that makes it even more interesting (IMHO) 22:44:40 well (2) is the most interesting 22:44:48 oerjan, agreed 22:44:50 er, (3) 22:44:52 err, i don't know if it's true, because i've forgotten what that is 22:44:53 oh 22:44:57 hm I guess so 22:45:01 Ok, my brainstack compiler now supports macro parameters 22:45:05 your numbering was off 22:45:14 oerjan, off compared to? 22:45:27 Unfortunately my current problem involves std i/o 22:45:27 AnMaster: you used 2) twice 22:45:30 oh 22:45:31 right 22:45:32 sorry 22:45:34 meant 3 22:45:37 -!- sebbu2 has joined. 22:46:11 AnMaster: axioms are not "probably true", as in "things that cannot be proved but might be true", they are things we've decided to hold as true, to see what happens should we do so 22:46:36 what's riemann? explain it in an irc message 22:46:38 oklopol, yes I know what an axiom is 22:46:45 oklopol, http://en.wikipedia.org/wiki/Riemann_hypothesis 22:46:55 that's just a link with the word in it 22:46:59 hardly an explanation 22:47:02 *shrug* 22:47:06 %repeat|65|+%%repeat|25|.+% 22:47:19 Prints the alphabet in allcaps 22:47:29 But I still need to fix . and , 22:47:33 CO2Games, you know this is hardly a turing tarpit any more 22:47:36 And I need to implement [ and ] 22:47:37 night everyone 22:47:41 I know 22:48:29 * oklopol clicks link 22:52:03 btw AnMaster, it supports inline assembly via the @ command 22:52:37 CO2Games, ewwwwwww 22:52:42 and night really now 22:52:49 * AnMaster turns off screen 22:52:53 lol 22:53:03 AnMaster is gay 22:53:22 He has sex with pigs 22:53:44 litterally like oink oink 22:53:58 Yeah ok maybe he did turn off the screen 22:54:08 CO2Games is a bit confused about the word "gay" i think 22:54:25 No, he's gay *and* has sex with pigs 22:54:40 [citation needed] 22:54:41 not limited to pigs 22:54:43 lol 22:55:02 So anyway macro parameters work now 22:55:38 But I can't run the programs assembled and linked by masm compiled by BFToASM 22:55:41 :\ 22:56:38 shit I just can up with a really bad pun 22:56:57 it's infectious 22:57:05 "or-gasm", using or's in gasm 22:57:09 also, madness 22:57:13 madness? 22:57:21 THIS IS #ESOTERIC! 22:57:27 exactly 22:57:32 madness is mandatory here 22:57:38 heh yeah you got a point there 22:58:26 so is andgasm the dual concept? 22:59:37 of course it was. well, nearly. 23:01:07 -!- sebbu has quit (Connection timed out). 23:04:38 xorg-asm 23:04:43 xor-gasm 23:05:10 gorgasm 23:09:32 http://www.youtube.com/watch?v=aD4bn5pp32w&feature=related 23:11:11 Too bad that's actually a gasm-or. 23:12:04 "You know I'm molestin` ox's now." 23:20:16 it must be true. everyone knows people who misuse apostrophes are sexual deviants. 23:21:29 also, bathyspheres. 23:25:28 That's not a misuse of an apostrophe, it's a misuse as an apostrophe. 23:25:55 ? 23:27:22 because it wasn't actually an apostrophe character? 23:27:25 or what 23:27:37 err 23:27:43 actually i guess it was about the ox's 23:27:47 so i have no idea 23:27:52 misuse of accents *brr* 23:27:59 i'm just squintering my way through the world of scent and danger 23:29:23 scent-based UI _sound_ nice, until you get spam 23:30:47 -!- CO2Games has quit ("And I invented doors, no joke!"). 23:31:07 I wouldn't want a scent-based UI unless I were Florence Ambrose... 23:37:38 -!- puzzlet has quit (Remote closed the connection). 23:37:43 -!- puzzlet has joined. 23:42:23 -!- olsner has quit ("Leaving").