00:00:40 hmm... do i do a generic number-> bsmntbombdood-code function or english? 00:00:43 what did you imply? 00:01:06 dude, you are a genious 00:01:12 That loop code works perfect 00:01:24 well, thank you 00:02:16 now I need to think about the math 00:02:35 and write a spec 00:03:43 god i love metaprogramming :O 00:03:49 how come i never did it before? 00:03:50 >>> def printnumsfromntozero(n): 00:03:50 return makenumber(n)+"[.1-].]" 00:03:50 >>> printnumsfromntozero(7) 00:03:50 '11g11g11g11g11g11g1------------[.1-].]' 00:03:50 >>> runsome("11g11g11g11g11g11g1------------[.1-].]") 00:03:52 7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0 00:04:11 hmm 00:04:28 should open c++... i'll start liking that lang soon :< 00:04:54 hmmm 00:05:15 yeah, the ordo is not optimal 00:05:27 i mean O(n) to create a number 00:06:12 In the loop, you have to pop the stack 00:06:25 is that so? 00:06:47 my bot does that, believe, it's not that nice 00:06:51 *believe me 00:06:59 i'll get some coffee 8| 00:11:50 okay 00:11:53 added sqrt 00:11:55 as 'q' 00:12:04 now try 11g1--qg. 00:12:13 and wait 10 min for the generic one 00:15:06 hmm 00:15:27 does python have err... err.... err..... binary logical operators? 00:17:35 hmm i assumed being able to create 0.5 automatically makes it easy to create 0.25 etc :) 00:17:54 oh 00:17:56 it does... 00:19:12 argh 00:19:20 I can't figure out how to multiply 00:19:33 binary lowho? 00:19:35 loop 00:19:49 oklopol: no, with logs there's a better way 00:19:57 okay 00:20:01 what is it? 00:20:27 11g1--qqqqg = 0.0625 00:20:34 ">g true 00:20:55 #define ADD > ZERO <-- 00:20:59 then inverse would be nice 00:21:11 right, but I don't want to have to add another operation 00:21:26 trues 00:21:31 but, q is bad too? 00:21:44 yeah, I don't want any extra ops 00:21:44 makes fractions nice to do 00:21:48 okay 00:21:53 then it's loop 00:22:02 no 00:22:09 there has to be some way to do it with logs 00:22:20 yes 00:22:23 changing base 00:22:28 :\ 00:22:33 ? 00:22:34 i can't think of anything else 00:22:44 What do you mean changing base? 00:22:47 hmm, i might be wrong 00:23:00 i'll test before i say anything more stupid 00:23:44 maybe inv. log (exponent) is needed 00:24:59 yeah 00:25:05 i was wrong 00:25:14 haven't fiddled with logarithms in ages 00:25:14 that makes me sad :( 00:25:22 :( 00:25:27 -!- ihope has quit (Read error: 104 (Connection reset by peer)). 00:29:42 i thought there was a correlation between either logarithms with inverted base, but there seems not to be 00:29:53 (i'm not good at math, i just guess :)) 00:30:17 yeah, the coffee >> 00:30:56 -!- sebbu has quit ("@+"). 00:31:13 ok, for now i've added an operation "p", which inverts "g" 00:31:49 so now to multiply you can do ">g yeah ok 00:33:54 but, if you need decimal numbers, i'd say square root being a sister operation of a 2-base logarithm would make a nice addition to the lang... unless there is a nice way to do numbers with logs 00:34:07 decimal numbers -> floating point values 00:34:19 -> rational non-integer numbers 00:34:29 -> the right term fuck it 00:35:03 oklopol: back, sorry, kk, thanks 00:35:35 you need logs when calculating loans etc 00:35:35 what's a decimal number? 00:35:55 lament, i used the wrong term 00:36:13 a->v mean s/a/v 00:36:19 ... 00:36:22 s/v/b 00:36:39 ...someone kill me 00:36:59 * ryanakca gives oklopol a poison cookie :) 00:37:13 :) 00:40:07 argh, how to do this 00:40:08 so many people get the terminology wrong :( 00:40:10 they make me sad :( 00:41:31 aARARARaRARARAR 00:45:08 what, bsmntbombdood? 00:45:18 lament, i'm an idiot 00:45:27 inverting a logarithm 00:45:31 2**... 00:45:39 I don't want to have an extra command 00:45:44 ah okay 00:45:51 you said something peeish 00:46:02 yeah 00:47:30 my source: http://bsmntbombdood.mooo.com/logfuck.py.txt 00:49:45 read 00:49:53 nice 00:50:03 all my iffy parts fixed 00:50:05 nice work 00:50:19 but, i had to make it in 10 minutes to be first :) 00:50:27 heh 00:50:50 * oklopol likes q :'( 00:50:51 :DD 00:51:25 You can compute square roots withs logs 00:52:30 ah okay 00:52:34 shomme! 00:54:42 one sec 00:54:45 k 00:54:50 darn 00:54:57 i'm gonna hate myself when i see it :) 01:01:53 hrm 01:02:32 does it work? 01:02:34 well, sqrt(x) == 2**(.5 * log(x)) 01:02:43 :) 01:02:58 since sqrt is the only way to get 0.5, that's hard 01:03:03 no 01:03:03 you can get close tho 01:03:06 oh 01:03:15 #define DIV >g #define ZERO 1g 01:03:23 #define TWO 1 ZERO 1-- 01:03:25 well, with p, true :) 01:03:44 1 TWO DIV 01:03:47 is .5 01:03:49 but, you see, then you don't need sqrt 01:03:54 ye 01:04:07 right 01:04:08 so... 01:04:13 so, it's p or q 01:04:15 floating points numbers are a tad... limited. 01:04:23 CakeProphet, you are limited 01:04:29 :D 01:04:29 LOL NO 01:04:31 oklopol: p can do more than q 01:04:37 CakeProphet: What's your point? 01:04:38 yes 01:04:43 much more 01:05:02 a ha! 01:05:08 got it 01:05:11 so, you are a stupid head 01:05:11 #define SQRT >1 TWO DIV :O 01:05:27 yeah, okay 01:05:32 i can make a num conversion 01:05:52 and it'll make it to the spec and i'll be a hero -_____- 01:06:14 well... 01:06:16 What do you mean num conversion? 01:06:19 hrm.. 01:06:29 conversion of a number into lf-code 01:06:44 oh 01:06:58 but, it wouldn't be very optimal 01:07:01 easy tho 01:07:10 I'm going to change output from decimal to ascii now I think 01:07:25 yeah, for non-integers 01:07:28 *integers 01:07:45 >>> float("0." + "".join(["565656534" for x in xrange(100000)])) 01:07:46 0.56565653456565657 01:07:48 >>> 01:07:49 :( 01:07:57 I want giant floating point numbers!!!! 01:08:54 CakeProphet: gmp 01:09:01 i want 10-base to be illegalized 01:09:20 I want to use a base-5 numeral system! 01:09:24 THE LAW OF FIVES. 01:09:35 How do you implement dup in lf? 01:09:40 I can't figure it out 01:09:45 hmm... 01:10:02 you could consequitely move items from one stack to the next. 01:10:06 till you... get to the one you want? 01:10:12 dup 01:10:16 not lift 01:10:30 well... 01:10:32 it's close. :D 01:12:00 hmm 01:12:07 It's not turing complete unless I can implement dup I think 01:12:20 since [] is destructive 01:12:55 hmm, there is no way to use a number in a calculation without even then only having it in one place again... i think 01:13:08 you can only add a 1 01:13:21 Implement two stacks-of-queues. 01:13:24 like Stue. :D 01:13:29 except... two 01:13:31 TwoStueFuckLogs 01:13:36 you could call it. 01:13:41 no 01:13:42 and using that one IN ANY WAY with the number to duplicate will destroy the original number 01:13:43 too complicated 01:13:50 and you have the same prob with the new one 01:13:53 oklopol: yep :/ 01:14:10 i think i can prove it's impossible... but that's all i can do :\ 01:14:23 I could make > copy isntead of moving 01:14:26 ...if it's impossible, that's pretty obvious tho :) 01:14:30 yeah 01:14:43 but that causes other problems 01:14:58 such as.. not being able to remove items. 01:15:02 because it just copies. 01:15:11 you can. 01:15:15 oh 01:15:20 you could just have a copy command. 01:15:25 CakeProphet: no 01:15:28 that copies from stack to other stack 01:15:32 too many instructions 01:16:00 [ could copy the topmost, ] could pop :P 01:16:39 nah, wouldn't work... i think 01:16:42 hmmm 01:16:49 because you still couldn't dup for example 01:17:06 the dup would be ">0<[>]<" 01:17:11 s/the/then/ 01:17:18 s/0/1g/ 01:17:57 you're right 01:18:00 I think that works! 01:18:15 i didn't understand > and < right, yeah, it was good 01:18:26 s/understand/...grasp :) 01:20:51 crap, that doesn't work 01:21:09 :\ 01:22:11 mine says list index outta range 01:22:18 but i didn't allow that anyway 01:22:24 so it should 01:23:18 [>0]< 01:23:24 that would dup right? 01:23:32 eh 01:23:41 yeah 01:23:43 i thnk 01:23:47 no 01:23:58 wait 01:24:43 no 01:24:56 that does nothing 01:26:08 n, -> (start of loop) n n, -> n, n -> (end of loop) n 0, n -> n, n -> n n, 01:26:09 hmm.. 01:26:24 I wonder how you would use a tree as a languages data structure. 01:26:29 yes 01:26:33 my friend did that 01:26:49 it is nice 01:27:10 you can have tree data structures as vars then 01:27:14 very convenient 01:27:53 You could probably do something with genetics. 01:27:54 and trees. 01:27:59 and fractals. 01:28:01 and chaos 01:28:04 and evolution. 01:28:09 > goes left, < goes right, ^ goes up, so navigating is like bf but you can reserve infinite certain memory for vars 01:28:25 hmm... 01:28:27 that's cool. 01:28:36 you could take a lisp-ish approach and make the program a tree. 01:28:40 true 01:28:44 but he's not good 01:28:49 with the branches acting as... erm.. 01:28:51 you could 01:29:01 conditional pathways? 01:29:07 aaagh 01:29:18 you could just code it as a tree, no ascii shit 01:29:31 i don't understand why everything must be coded in ascii... 01:29:36 or unicode 01:29:55 except for cellular automatons 01:29:56 ok 01:29:57 of course 01:30:04 bsmntbombdood, what? 01:30:05 We make > copy instead of move 01:30:09 < acts as normal 01:30:12 okay 01:30:17 could do something like wireworld... but different. 01:30:18 ah 01:30:21 yeah 01:30:25 electrical circuitry type programming. 01:30:32 make it a GUI 01:30:34 it could be fun. :D 01:30:36 how do you drp? bsmntbombdood 01:30:41 oklopol: [] 01:30:46 yeah 01:30:58 what was wrong with my n, -> (start of loop) n n, -> n, n -> (end of loop) n 0, n -> n, n -> n n, tho? 01:31:05 what fails? 01:31:11 We should all COLLABORATE 01:31:11 the move 01:31:13 and make something sweet 01:31:22 but according to my vision.. 01:31:24 why does it fail? 01:31:27 because I said so. 01:31:31 CakeProphet, sure :) 01:31:39 now [ pops 01:31:46 FRACTALIC SELF-REPRODUCING EVOLUTIONARY GENETICS CODE 01:31:49 now 01:31:51 how 01:31:53 think 01:31:58 but but, what was wrong with the move? :'( 01:32:14 It was destructive 01:32:31 okay 01:32:35 how would it go then? 01:32:41 if not like that 01:32:53 i did it destructively in n, -> (start of loop) n n, -> n, n -> (end of loop) n 0, n -> n, n -> n n, i think 01:33:06 write that in the language 01:33:06 I'd like to make an esolang larger than a few commands and a tiny interpreter. 01:33:17 i did, but it fails 01:33:22 like... a major project... 01:33:39 oklopol: exactly 01:33:43 but, why? 01:33:56 how do the stacks evolve then? 01:34:07 crap, borked 01:34:19 because making > copy is not nice if it's not needed 01:34:27 because it's not very intuitive 01:34:40 * CakeProphet is thinking. 01:34:47 Yeah 01:34:48 and i think an esoteric language should always be intuitive... according to it's own rules 01:35:03 now we can't drop 01:35:13 ah 01:35:15 a genetic language isn't coming into my head intuitively. 01:35:15 yeah 01:35:35 but it seems like it would be simple. 01:36:00 you have some stuff... and some other stuff... and some stuff that tests that stuff... 01:36:08 and something to put it all in... an environment. 01:36:13 that really sums it up :) 01:36:19 and stuff starts dying.. 01:36:26 and stuff starts procreating. 01:36:28 yay... genetics. 01:36:37 which inevitably creates evolution. 01:37:04 AHA 01:37:04 i thought of making a language with the most basic component being flock movement... but if that's not done yet, i pity mankind 01:37:06 simplified DNA 01:37:18 a series of.... codes\ 01:37:22 that determine shit 01:37:33 and make a bunch of them... 01:37:42 or better yet... just make one starting code... the source code. 01:37:55 which spawns off offspring.. 01:38:05 two source codes? 01:38:07 yes. 01:38:09 two source codes. 01:38:11 that mate. 01:38:26 yeah 01:38:31 swap genes... reproduce... you have a bunch of source codes now. 01:38:39 first of all you need a lot of redundancy 01:38:48 and some start dying... the ones that you don't want. 01:38:52 nah... just two genetic strands. 01:39:16 but I don't want DNA.... the thought of coding in DNA-like language kinda makes me want to puke. 01:39:40 the codes you have first should be quines that are altered then executed until more new executable code is randomly created 01:40:00 maybe borrow some ideas from Thue-like-ness? 01:40:05 but, the language should aim at creating itself 01:40:05 but.. not direct string replacement. 01:40:09 yah. 01:40:26 I've been tossing through ideas like that. 01:40:37 a language that contains the instructions for creating itself 01:40:38 it need not have any other operation than cloning itself, changing thing to their "synonyms" etc 01:40:59 well... there needs to be some form of sexual interchange... otherwise you just have a bunch of clones. 01:41:19 well, yeah 01:41:26 we can leave out randomness then 01:41:29 which is good 01:41:36 could have it as an option. 01:41:40 non-determinism is cool sometimes. 01:42:16 because with one we'd need a seed, with two (that reproduce collaboratively) they are each others seeds, more sensible 01:42:21 yeah 01:42:29 but the seed could then be a random program 01:42:46 since all programs should of course be always-executing 01:43:01 but, redundant operations just wouldn't produce themselves 01:43:19 hmmm... the program instructions could correspond with each other... 01:43:23 like... 01:43:25 so there would always be a random amount of nops that always die 01:43:29 yes 01:43:35 do different things for two different pairings? 01:43:40 or... however we pair them up. 01:44:01 yes, these pair could be made so that best result i gotten 01:44:04 *pairs 01:44:07 *is 01:44:19 my s's are becoming extinct 01:44:25 haha... a self-optimixing language. 01:44:29 :) 01:44:36 fuck compilers. 01:44:39 did you coin that term just now? 01:44:42 just make the language do it itself. 01:44:57 ...well.. it was a typo. 01:44:59 but sure. 01:45:01 I coined it. 01:45:15 okay 01:45:22 i thought it was brilliant :) 01:45:37 since it mixes two progs while optimizing automatically 01:46:02 now, english again 01:46:06 by program.. do you mean hard file or... separate 'halves" 01:46:12 it could be either I guess. 01:46:19 i need more complicated english here than in my homework :\ 01:46:21 you could have two different files get it on as well. 01:46:34 file-sex. 01:46:52 you could do like dna 01:47:10 each instruction is two bytes, like +/ () %% etc 01:47:24 and you know how it works 01:47:38 eh... my biology is fuzzy. 01:47:40 okay 01:47:42 but I get the picture. 01:47:45 they sort of... interchange right? 01:47:52 random one of the pair is picked 01:48:12 hmmm... I don't think randomality could produce optimizations... but... who knows. 01:48:39 mixing program "�� ## 1%" with "aG oo /\" could be "�G #o %\" for example 01:48:44 oooh.. 01:48:48 the optimization wouldn't be there. 01:48:50 you could optimize it afterwards 01:48:54 it would be in the... massive amounts of this occuring. 01:48:59 and... which ones "live" 01:49:04 by being more efficient. 01:49:06