00:00:10 i could maybe do lojban minus gismu. 00:00:14 gismu 00:00:14 ? 00:00:26 i was thinking you'd teach real life concepts to it yourself 00:00:27 umm 00:00:30 gismu are root words 00:00:33 oh. 00:03:17 but probably i'd go for a simpler language. 00:03:35 and, you know, more mathy. 00:03:49 closer to Absolute Fundamental semantics 00:03:50 ... 00:04:06 * oklopol remembers a certain moment of madness 00:04:42 ^bf +++++++[>+++++++<-]>++.+. 00:04:42 34 00:04:58 Hm, is A XOR B = (A+B)(¬(AB)) ? 00:06:21 FireFly: yes 00:06:30 Good 00:06:42 I mananged to do the basics 00:06:44 : 00:06:45 :>* 00:07:08 also, ¬(AB) = ¬A + ¬B 00:11:25 Hm, yeah 00:11:33 Bleh, I'd better sleep 00:11:39 Until tomorrow 00:11:40 Or 00:11:45 Later today, actually 00:11:53 -!- FireFly has quit ("Later"). 00:12:11 buy. 00:12:29 sell! 00:18:47 -!- Corun has joined. 00:30:16 -!- yoR has left (?). 00:48:15 -!- Corun has quit ("Leaving"). 00:48:20 -!- CakeProphet has joined. 01:10:16 -!- Sgeo has joined. 01:26:07 heys 01:26:56 haze 01:29:50 -!- Corun has joined. 01:31:36 -!- Sgeo_ has joined. 01:31:36 -!- Sgeo_ has quit (Client Quit). 01:31:50 so 01:33:08 ive rewriting my interp for AntiGravity 01:33:31 -!- Corun has quit (Client Quit). 01:39:31 * AnMaster goes insane 01:39:53 I shall try to port it to python 3 while keeping it working under python 2 01:40:07 and the C API differs a lot 01:42:20 i think that's - not recommended. 01:47:51 -!- CakeProphet has quit (Read error: 60 (Operation timed out)). 01:48:33 -!- CakeProphet has joined. 01:51:43 oerjan, going insane? 01:52:17 oerjan, one of the devs use debian stable, if we didn't support both it would take like 5 years before we could upgrade! 01:52:41 no, trying to make something both Python 2 and 3 01:52:52 oerjan, well the C module, not the python code 01:53:03 oerjan, for the code you could just use 2to3 01:53:04 but 01:53:15 not so for the C code embedding python 01:53:27 http://docs.python.org/3.0/howto/cporting.html#cporting-howto 01:54:09 oh i was thinking you were going for a complete polyglot. at least that would be the properly insane thing. 01:54:25 oerjan, impossible for the C API 01:54:37 oerjan, as for that, it is just avoiding "print" and a few other things 01:54:53 2to3 shows that is basically one of the few changes 01:57:56 $ LD_LIBRARY_PATH=$HOME/local/python-3.0-dbg/lib PYTHON_HOME=/home/arvid/local/python-3.0-dbg ~/local/python-3.0-dbg/bin/2to3 *.py 01:57:56 RefactoringTool: Skipping implicit fixer: buffer 01:57:56 RefactoringTool: Skipping implicit fixer: idioms 01:57:56 RefactoringTool: Skipping implicit fixer: set_literal 01:57:56 RefactoringTool: Skipping implicit fixer: ws_comma 01:57:57 RefactoringTool: No files need to be modified. 01:58:00 for some files 01:58:07 of course it could still break 02:27:20 night 02:35:52 -!- psygnisfive has quit (Remote closed the connection). 02:50:33 There once was a man from Man-Fuck-Dick, 02:50:33 Whose horse enjoyed when he sucked it, 02:50:33 His limericks reversed, 02:50:33 The humor inversed, 02:50:33 'til the day that he moved to Nantucket. 02:52:16 Nantucket? 02:52:51 Meta-humor FAIL. 02:54:01 Meta-meta-humor success? 02:55:54 ^style 02:55:54 Available: agora alice darwin discworld europarl ff7 fisher ic irc* lovecraft pa speeches ss wp 02:55:59 fungot, hi 02:56:03 ... 02:56:08 someone talk to him? 02:56:17 fungot, how do you do 02:56:18 oerjan: i don't think exists? and all? 02:56:21 thanks 02:56:31 oerjan, I need some data to feed into an eliza 02:56:34 hm 02:56:37 which style 02:56:41 fungot is being descartian today 02:56:42 oerjan: hmn. first dead fnord family member in years. mother's father died when i was a big php advocate for awhile. not so much 02:56:52 eww 02:56:57 "mother's father died when i was a big php advocate for awhile" 02:56:59 :D 02:57:07 goes to show 02:57:53 PHP KILLS PEOPLE 02:57:53 yes 02:57:55 ^style lovecraft 02:57:56 Selected style: lovecraft (H. P. Lovecraft's writings) 02:58:00 fungot 02:58:02 AnMaster: the tiles on graves and rotting temples flash like ripples, as earth's gods dance against it; i shall see the dancing forms of the people always represented by the sacred reptiles appeared to be, while the boys were overseas. some fateful mood impelled me to ask for work. 02:59:54 fungot, oh? 02:59:55 AnMaster: after a time he became very numb and somnolent, moving more from automatic impulse than from reasoned fnord nor did he neglect a small store of gold for earthly use. 03:00:45 i'm not sure reasoned fnord is an improvement 03:32:10 -!- psygnisfive has joined. 03:32:20 ja 03:32:50 i need to add built ins to antigravity to allow for things like IO i think ya 03:32:58 what sorts of built ins should i have?? 03:35:11 something uplifting 03:35:14 * oerjan ducks 03:36:16 -!- CakeProphet has quit (Read error: 60 (Operation timed out)). 03:38:32 hmm 03:38:35 uplifting huh 03:38:37 sounds good! 03:39:06 also, "uplift" is occasionally used as a verb to mean "increase the intelligence of some animal to the point of sentience" 03:39:18 so an uplifting function would be quite interesting to have! 03:39:43 also, observe as i only now get the pun 03:40:21 it's a *whoosh* ... but but ... it comes back ... and hits! 03:40:39 :p 03:40:49 it left a dent in my head! D: 03:41:10 well dent do that then 03:41:18 :P 03:42:46 -!- CakeProphet has joined. 04:10:31 !! 04:10:41 i just had another idea for something to add to the AntiGravity system 04:10:51 besides types, which i will add eventually 04:11:29 you'll also be able to define attributes on the grammatical rules 04:11:54 so that you can not only define new parts of the grammar 04:12:20 but you'll also be able to perform attribute grammar types of computation while parsing 04:12:59 hm. 04:14:03 maybe. 04:14:06 :P 04:34:26 -!- psygnisfive has quit (Remote closed the connection). 04:45:40 -!- psygnisfive has joined. 05:21:31 -!- oerjan has quit ("leaving"). 06:22:21 TALK WON'T YOU 06:22:22 :| 07:11:19 07:11:51 .. oh attribute grammars, shiny ! 07:13:28 ski__: :P 07:14:44 well, the language itself is actually an unrestricted grammar engine, so it can do things attrgrams cant do, ofcourse. but im considering making the parsing engine capable of handling attrgrams for the potential odd kinds of solutions it might permit 07:59:59 -!- clog has quit (ended). 08:00:00 -!- clog has joined. 08:19:25 -!- jix has joined. 08:21:13 hmhm 08:21:40 * ski__ want to experiment with coupling AGs with LP 08:22:41 -!- jix has quit (Client Quit). 08:23:57 Antigravity with linear prediction. 08:24:40 why do you call it "AntiGravity" ? 08:29:16 just felt like it. 08:29:20 fizzie: linear prediction? 08:29:40 its a working name, i think. ill rename it something more nice eventually 08:30:22 * ski__ called a language `contagion', once 08:30:31 i like that name 08:30:52 the only abstraction primitive was anonymous continuations 08:31:01 huh. 08:31:35 I don't know why, but I've seen AG used as an abbreaviation for antigravity. And linear prediction is a rather fundamental signal processing thing. 08:32:10 oh .. i used `AG' as abbreviation of "Attribute Grammar" 08:32:18 AG is indeed a standard abbreviation for antigravity 08:32:31 so what exactly IS linear prediction 08:32:32 (maybe that was psygnisfive's point in calling it "AntiGravity", though) 08:32:51 no. the name preceded the potential use of attribute grammar qualities 08:33:20 (never heard of `AG' as abbreviation of "Anti-Gravity") 08:33:26 hm. 08:33:28 -!- Sgeo has quit ("Leaving"). 08:34:29 Linear prediction is just "predict next sample as a linear combination of p earlier samples"; you can use it for statistically modeling signals, and things like that. 08:35:44 ahh ok. 08:35:54 markov chains. 08:39:10 Er, not really, there are no state transition probabilities involved anywhere. 08:43:49 But sure, in the sense that both could be used for somewhat similar things. Although when modeling a random process, linear prediction is usually called autoregressive modeling. LP is used a lot in things like compressing speech to very low bitrates. The GSM codecs are linear-predictive. 08:45:10 hm. 08:47:57 so fizzie what /other/ built ins should i have? :P 08:48:10 since i dont know how i'd incorporate LP 08:48:30 -!- GreaseMonkey has quit ("Client Excited"). 08:59:34 no? 08:59:35 nothing? 09:03:42 I was: eating lunch. 09:04:48 Oh, antigravity was mentioned earlier too. I didn't bother to read the context. 09:05:12 I don't really know how linear prediction would be related to whatever-it-is-you-were-doing. 09:05:32 :p 09:05:40 what should i build in! D: 09:20:41 * ski__ just realizes fizzie might have interpreted `LP' as "Linear Prediction" 09:21:22 Yes, your "AGs with LP" is what made me say what I did. 09:21:34 i feared that :) 09:22:00 by `LP' i meant "Logic Programming" 09:23:59 ski__: regarding logic programming 09:24:16 AG does do some unification stuff, so you might be able to do some LP kind of stuff 09:24:36 yes 09:24:39 also 09:24:44 but 09:25:29 some AG systems statically stage tree traversal in several steps, to compute inherited and synthesized attributes 09:25:30 it wouldn't be elegant like prolog because you'd be building it on top of a non-logical system, and you could only prove truths, not discover truths, i think. 09:25:50 "some AG systems"? 09:25:52 i want to have a system that allows such static (maybe dynamic, too ..) staging of predicates 09:26:03 yes 09:26:32 i didn't realize there were things related to AG that _i_ wasn't making. :P 09:26:58 oh oh 09:27:01 Attribute Grammar 09:27:04 haha ok :) 09:27:14 dont abbreviate that :P 09:27:20 or atleast not like that 09:27:28 AG = AntiGravity XD 09:30:21 no :) 09:30:49 i believe that is a standard abbreviation, in relation to grammars, and programming languages 09:31:01 :P 09:32:24 so cmon! what should i add to my language as a primitive. 09:32:32 what is useful 09:32:47 continuations ? 09:32:55 variables ? 09:32:55 and what do i need to consider for laziness/forced evaluation 09:33:14 it cant have continuations, really. or it might be able to but im not sure how. 09:33:20 and it does have variables 09:33:21 (you might note that i have no idea what you already have in your language) 09:33:24 :p 09:33:32 it is right now literally just 09:34:29 a unrestricted tree rewriting engine with variables in the grammar's rules, and a unification system for pattern matching 09:35:17 thats really it. the core functionality is math rules, and a rule for substitution of an item in a tree 09:35:45 could you give some contrived samples to exemplify what you mean ? 09:35:58 no. it just ends up looking like haskell. 09:35:59 :P 09:36:28 how about giving a BNF, then ? 09:36:56 there is no real grammar for it either. 09:37:09 i mean, there are really only two things that define the language. maybe three. 09:37:15 = is used for definitions. 09:37:32 () define subtrees 09:37:55 and starting a symbol with a capital defines a syntactic variable 09:38:04 -!- impomatic has quit ("programming @ http://tr.im/xey"). 09:38:06 meaning ? 09:38:20 well, you know what an unrestricted grammar is, right? 09:39:06 i don't recall details .. but i imagine it's a couple or rules with no context-insensitive restriction .. 09:39:30 well, its basically just arbitrary rewrite rules 09:40:04 so that the kind of string you replace can look like anything 09:40:17 e.g. abc -> def 09:40:29 so you replace trees instead of strings 09:40:39 (i assume you have congruence) 09:40:46 congruence? 09:41:22 the important part is, if you know what an unrestricted grammar is, just imagine the rules that can have variables in them 09:41:37 if `blah' can be rewritten into `bleh', then any tree with `blah' inside it can be rewritten to the same tree with (some) occurance(s) of `blah' exchanged into `bleh' 09:41:57 e.g. abcX -> XXabc where X can be any symbol or tree part 09:42:05 so abcd -> ddabc, etc. 09:42:08 how about 09:42:25 aXbXc -> Xd 09:42:25 ? 09:42:27 oh i see what you mean 09:42:32 yeah thats fin too 09:42:34 fine* 09:43:15 * ski__ is btw irritated by the direction the `->' arrow in grammar rules point .. 09:43:35 in some sense, yes, it has congruence, but its not like a rule blah -> bleh will actually transform that tree, necessarily, because its lazy. 09:43:54 (when i add support for grammars to my LP language, i'll have it point the other direction ..) 09:43:59 why are you irritated?? 09:44:49 in `aXbXc -> Xd', the `->' is really a kind of implication .. but it is `Xd' which implies `aXbXc', and not the other way around 09:45:09 -> is not an implication :P 09:45:20 -> is an arrow telling you what gets rewritten as what :P 09:45:38 i'm pretty sure it can be interpreted as an implication in an ordered logic 09:45:56 it can be, and in THAT case you're right 09:46:16 (which, imo, is what grammars is all about) 09:46:19 but thats not how rewriting systems are termed 09:47:34 hm .. will you have confluence in your term-rewriting system ? 09:47:46 ^ 09:48:11 oh no that was congruence 09:48:13 whats confluence? 09:48:45 basically that which rewriting rules you choose in which order doesn't matter for the "end result" 09:49:10 there might be several possible rewriting rules you could use to rewrite a term 09:49:26 say `A' can be rewritten into `B', but also into `C' 09:50:10 then there should be a `D' such that both `B' and `C' will eventually be rewritten into `D' 09:50:30 (there might be some details regarding termination which i forget here, but that's the gist of it) 09:51:15 oh yeah no it is ordered because of pattern matching and stuff 09:51:51 but thats amongst definitions of "the same function", so to speak 09:52:13 09:52:24 yes 09:52:27 consider 09:52:46 foo (bar x (baz y z)) 09:53:02 you might rewrite according to `bar' or according to `baz' first 09:53:22 confluence says that it doesn't matter which you actually do 09:54:18 yah, im not entirely sure if thats going to happen. i mean, because its pattern matching, you could have a pattern that looks inside another "function application" 09:54:29 e.g. you could have a rule bar x y -> z 09:54:48 but you could also have another pattern foo (bar x y) -> w 09:55:00 or something like that 09:55:04 ok 09:55:18 where like.. if bar x y = 5 09:55:34 i suppose another example of confluence is 09:55:44 and foo (bar x y) = 10 09:55:49 foo 5 doesnt have to = 10 09:56:19 A + (B + C) --> (A + B) + C 09:56:29 yah 09:56:39 0 + A --> A 09:56:46 so if you have 09:57:00 0 + (x + y) 09:57:09 you could use either rule first 09:57:20 yah. 09:57:58 but anyway 09:58:01 thats what it is 09:58:10 and i need some ideas for what primitive rules should exist 09:58:19 e.g. IO and stuff 09:58:39 since its very purely functional, should i think about monad sorts of things, etc etc 09:58:52 (iirc, there's five ways to go from `a + (b + (c + d))' to `((a + b) + c) + d' here .. that they are all equal is called the pentagonal law) 09:59:06 :P 09:59:27 anyway, i need to run off to sleep. 09:59:45 maybe you could use some kind of unique state for I/O 09:59:48 (like Mercury) 09:59:51 private message me any ideas that you think would be important to include :p 10:00:05 well IO i can really just implement pretty naively 10:00:08 but whether i should or not 10:00:18 well, i don't know what you want, so it's hard to say what's important for you 10:00:50 well i just mean whats important in a usable programming language in general. 10:01:05 mhm 10:01:18 anyway, night. <3 10:01:23 m, night 12:46:24 -!- blAckEn3d has joined. 12:47:33 -!- blAckEn3d has quit (Client Quit). 12:54:19 -!- Hiato has joined. 13:04:45 where has ais gone 13:04:45 :\ 13:29:11 -!- Hiato has quit (Read error: 110 (Connection timed out)). 13:33:59 well do like ehird and just start coding in it... :D 13:34:00 god i hate that dood sometimes 13:34:02 (read: "envy") 13:34:07 --2008-04-09 13:35:55 [2008-04-10 04:33:07] < oklopol> ehird is a guy who lives here 13:36:03 yes 13:36:54 There is also a nice definition for "atom" by oklopol in that log: "a retarded string, somewhat" 13:37:05 verily 13:37:14 oklotalk-- is pretty cool. 13:37:26 it's very... orthogonal 13:38:37 i have the oklotalk-- source UNLIKE ANYONE ELSE :DD 13:39:06 i mean it even has tests. 13:39:07 holy fuck. 13:39:24 # Needs to be set quite high because the implementation has the massive 13:39:25 # conceptual defect of not supporting tail recursion in a 13:39:26 # language without any other type of iteration. 13:39:28 sys.setrecursionlimit(3000) 13:39:30 CONCEPTUAL DEFECT 13:39:56 Hang on to that source code, it'll surely be worth millions some day. 13:41:01 # TODO 13:41:01 # standard representation of oklotalk-- object, not implemented 13:41:02 def obj_to_str(a,cxt,depth=0): 13:41:04 return a.call([Atm('get')],"oo",cxt,depth=0).val 13:41:06 wut 13:41:57 # store all functions on stack, a function needs to evaluate (to a copy of itself) 13:41:58 # if it is already on stack when called 13:41:59 dyn_cls=set([]) 13:42:01 w h a t 13:42:08 # Verbose may be helpful when debugging, but prolly not. 13:42:09 verbose=False 13:42:41 # jsussiuuidfhsaudfh PRIVMSG #chan ::: 13:42:51 it's drunk code 13:56:02 -!- Hiato has joined. 14:11:07 (we actually talk about programming stuff once in a while as opposed to just talking about manga and social incompetence) 14:11:11 -- 2008-04-10 14:19:58 Andreou, the founder of #esoteric, was last here 2008-08-31 14:22:42 oklopol: 15:02:43 o 15:07:54 oklopol write my j for m 15:07:54 e 15:07:55 :'( 15:11:53 -!- ski__ has quit (brown.freenode.net irc.freenode.net). 15:12:39 http://love2d.org/screenshot_data/seaotters.png 15:12:40 ehird: yeah what about that stack thing :D 15:12:49 oklopol: wut stack thing 15:12:55 that one you linked 15:13:05 # store all functions on stack, a function needs to evaluate (to a copy of itself) 15:13:09 # if it is already on stack when called 15:13:11 dyn_cls=set([]) 15:13:15 its from oklotalk-- 15:13:19 i know. 15:13:20 I don't know it XD 15:13:23 it makes no sense 15:13:23 :DD 15:13:28 omfg hackety redesigned 15:13:52 it makes sense, just not... in a conventional way 15:14:16 -!- ski__ has joined. 15:14:45 oklopol: also I totally thought of the best way to make oklotalk 15:14:56 it's sort of like feather but not rly 15:15:19 it'd let you literally add anything. 15:15:24 even continuations, if it didn't have them. 15:15:27 litter ally 15:15:37 ya 15:15:39 cunt in your asians 15:16:06 wait did you actually say something 15:16:08 * oklopol reads 15:16:13 oh okay. 15:16:20 well good for you 15:16:21 lol, cunt in your asians 15:16:21 :) 15:16:35 also oklopol will you ever make oklotalk 15:16:36 :| 15:16:37 what do you want from j 15:17:03 well. i currently *do not code at all*. so nothing gets done really. 15:17:22 but i will probably be in a more productive state at some point again. 15:17:55 i want to do dis in J: 15:17:58 14:23 21:07 How would I get all length-two permutations? e.g., 1 2 3 -> 1 1, 1 2, 1 3, 2 1, 2 2, 2 3, 3 1, 3 2, 3 3? 15:17:59 14:23 21:07 Specifically, giving each of those permutations as the left-argument of another application 15:18:06 probably after my fourth period, if you know what i mean 15:18:45 umm 15:18:45 wait 15:25:46 (,/@:(([,])"0/~)) 15:26:29 but they aren't permutations :P 15:26:54 they are elements of the cartesian exponentiation 15:27:57 of course that probably isn't a word, but it should be. 15:30:15 oklopol: right but 15:30:25 surely there's a way to get that shorter if you're just giving them each to another func 15:30:31 I mean in APL it's liek one char, iirc 15:33:18 9:49:01 alternatively, you could put the jewnicode into auschwitz09:49:12 but then you'd be an nascii 15:35:29 ehird: maybe there is, i don't see it. 15:35:35 more sp : | 15:35:36 -> 15:35:40 waiiiit oklopol 15:35:44 <-' 15:35:44 when I have those permutations 15:35:47 mmmk 15:35:48 how do I give them to a funnnnnnnnnnction 15:35:51 <.< 15:36:11 hey, if you want to put them in a function, you can just do func/~ list :D 15:36:18 except 15:36:22 it has to be the left argu 15:36:23 ment 15:36:25 not the right wun 15:36:39 also, nope 15:36:40 +/~ 1 2 3 15:36:41 2 3 4 15:36:43 err. 15:36:43 3 4 5 15:36:45 4 5 6 15:36:54 that's not length 2 picks 15:36:55 yeah, then just ,/ it 15:37:00 ,/? 15:37:07 ,/ 1 2 3 15:37:07 1 2 3 15:37:12 yeah, do you know what / does as monadic 15:37:15 no 15:37:23