00:09:24 -!- goban has joined. 00:30:39 -!- goban has quit (Remote closed the connection). 00:34:51 -!- goban has joined. 00:36:34 -!- sebbu has quit ("@+"). 00:52:25 har har! 00:52:32 that map is public record and denying me access to it is punishable by 90 days in jail 00:52:37 according to the colorado revised statutes, title 24, section 72, part 2 00:52:45 i win :) 01:20:22 zzzzzzzzzzzzz 01:20:30 yay! 01:20:31 zzzzzzzzzzzzz 01:26:28 -!- jix__ has quit ("This computer has gone to sleep"). 02:07:04 -!- UnrelatedToQaz has joined. 02:07:20 hey 02:08:36 Baz-wha? 02:13:09 -!- UnrelatedToQaz has quit ("ChatZilla 0.9.78.1 [Firefox 2.0.0.6/2007072518]"). 02:18:47 baz-quux 02:20:04 bsmntbombdood: what map? 02:20:15 ihope: of storm sewers 02:25:33 Here, have a law: if an emergency vehicle is approaching, pull over and stop. 02:26:06 Apparently no exception is made for freeways. 02:26:29 so? 02:29:18 Oh, I was wondering and... apparently I felt like saying that. 02:33:50 i hate how in a torrent with many files, all the files are 50% done instead of half of them all done 02:34:21 s/, all/, often all/ 02:36:36 Seems that if you're on a freeway and there's an emergency vehicle in the thing connecting the two roads, you're supposed to move one lane "apart" from the vehicle if possible. 02:36:52 yes 02:37:14 What does "one lane apart" mean, exactly? 02:37:23 one lane away 02:37:29 With one lane between the vehicles, or one lane over? 02:37:39 what? 02:37:42 That is, not in the same lane? 02:37:50 one lane between 02:37:58 * ihope nods 02:38:05 learning to drive? 02:38:25 Yup. 02:38:45 Pollution-loving wimps :( 02:39:00 That describes most of the US. 02:39:16 Especially our fondness for urban sprawl and a lack of decent public transportation. 02:39:26 i have knowledge, so i'm a pollution-loving wimp? 02:43:47 If you have knowledge and don't use it, you're a hippy. 02:44:19 I guess the government's decided that 120 MPH is the fastest relative we want to have in a head-on collision. 02:44:28 s/relative/relative speed/ 02:44:55 wrong, if you're talking about the federal government 02:45:08 I said just "the government" for a reason :-P 02:45:28 Are there higher speed limits than 55 for two-way traffic elsewhere? 02:45:32 75mph is the fastest speed limit allowed if you want to receive highway funding 02:45:36 Heck, are there higher speed limits here? 02:46:03 Freeways... there's nothing for a while, and then suddenly there's an on-ramp and you find yourself between two cars when there are only two lanes. :-) 02:46:10 that is, a state must have a maximum speed limit <= 75mph to get highway funding 02:46:35 Can states have a speed limit of 75 on a two-way road? 02:47:03 i think states can have whatever speed limits they like 02:47:35 Well, yeah, but if they want highway funding. 02:47:39 "In 1865, the revised Locomotive Act reduced the speed limit to 4 mph in the country and 2 mph in towns." http://en.wikipedia.org/wiki/Speed_limit 02:48:18 that is of dubious constitutionality, in my opinion 02:57:52 http://www.isbc.com/business/mistakes.cfm 02:58:23 i can has 640k? 03:16:09 -!- ihope_ has joined. 03:16:46 There. Now that everything's upgraded and I have my wireless connection working again, I can actually start work on that Unlambda interpreter. 03:18:52 how are you going to do it? 03:19:21 Haskell, monads... 03:21:30 and the call/cc? 03:22:47 i wonder if you can implement call/cc with tree-rewriting 03:25:23 Continuation monads. 03:25:28 Rather, a continuation monad. 03:25:55 so you're compiling to haskell or what? 03:27:34 Parsing into a Haskell data structure and then running. 03:29:48 write an unlambda to brainfuck compiler in unlambda 03:32:52 Why not a Haskell-to-Unlambda compiler? 03:33:06 because that's too large 03:33:16 Pff. 03:33:34 although simplified haskell to unlambda might be interesting 03:33:36 * ihope_ decides that adding a useful combinator to Unlambda is worth removing its comment functionality 03:34:08 haskell is just too complicated to write a compiler for fun for 03:34:22 Mm, gotta go. 03:34:29 Bye. 03:34:33 -!- ihope_ has quit ("Lost terminal"). 03:36:34 i have no idea how to compile a functional language to bf 03:38:43 PEBBLE functions. :p 03:40:07 are they first order? 03:40:15 er, first class 03:42:40 -!- poiuy_qwert has joined. 03:43:57 No, I mean "functions" as in "they work". 03:44:02 Err. 03:44:04 "it works". 03:45:14 Hahaha 03:45:27 oh 05:29:47 g'night everyone 05:29:59 -!- RodgerTheGreat has quit. 06:54:44 -!- olsner has joined. 07:16:23 -!- cherez has quit (Read error: 110 (Connection timed out)). 07:41:09 (bsmntbombdood) i hate how in a torrent with many files, all the files are 50% done instead of half of them all done <<< get a good torrent program... 07:41:22 like? 07:41:28 or make one 07:44:01 µtorrent for one 07:44:18 you have to manually tell it to dl them one by one though 07:44:43 you can assign separate dl speeds for individual files 07:45:19 isn't that windows only? 07:46:40 i have no idea... but perhaps. 07:46:57 i have a feeling there are decent programs for unix as well... 07:47:49 actually, the reason i'm doing my torrenting on this machine is that i can use µtorrent 07:47:54 bittorrent sucks ass 07:59:59 -!- clog has quit (ended). 08:00:00 -!- clog has joined. 08:09:44 -!- olsner has quit. 08:28:52 -!- GreaseMonkey has joined. 08:38:06 Deluge is the equivalent for *nix, bsmntbombdood. 08:38:18 Once I found Deluge, I pretty much abandoned my own torrent client. 08:38:25 (Even though I had most of it done.) 08:38:42 hmm 08:39:06 I'm going to start hacking at Deluge though. First thing I want to do is add stuff from BitTyrant's new BT paper. 08:39:22 what stuff? 08:39:51 They have a new piece obtaining and queueing algorithm to speed up a torrent. 08:43:22 azureus' peer dht is pretty nice 08:53:43 -!- poiuy_qwert has quit. 09:03:49 -!- GreaseMonkey has quit (Success). 09:06:30 -!- GreaseMonkey has joined. 10:38:03 night all 10:38:23 -!- GreaseMonkey has quit ("guess what that button did"). 11:09:56 -!- jix__ has joined. 11:43:29 -!- ehird` has joined. 12:11:59 -!- jix__ has changed nick to jix. 12:32:16 -!- UnrelatedToQaz has joined. 12:33:52 !info ololobot 12:34:00 Wrong channel... 12:34:02 hang on. 12:34:13 There. 12:34:22 -!- UnrelatedToQaz has left (?). 12:37:07 :D 14:37:21 -!- wellons has joined. 15:17:55 -!- goban has quit (Remote closed the connection). 15:40:28 -!- puzzlet has quit (Remote closed the connection). 15:40:38 -!- puzzlet has joined. 15:44:49 Does anyone know about a brainfuck implementation that can multithread several programs working on the same memory array? 15:46:02 ? 15:46:12 like a boring brainfork? 15:46:25 seveninchbread did something like that 15:53:44 hmm... i see, thanks 15:54:58 i do not know where 15:55:00 you can get it. 15:55:21 (i had to split to multiple lines for obvious reasons) 15:56:19 if sib comes here, you can ask him, though it isn't exactly that hard making it yourselfidy helfity 15:56:21 skdjgf 15:57:21 i wrote something similar yesterday 16:01:49 and, i don't see the reason for splitting to multiple lines... :-( 16:09:27 whut? 16:09:35 can i lick your source? 16:10:50 -!- sebbu has joined. 16:14:58 you need subversion, use this command: svn checkout https://opensvn.csie.org/wbf2c/trunk wbf2c 16:15:36 i have a website that explains things a bit here: http://www.cse.psu.edu/~wellons/bf/ 16:43:17 !bf_gen oklopol 17:31:57 -!- jix has quit (Read error: 104 (Connection reset by peer)). 17:33:52 -!- jix has joined. 17:35:14 -!- cherez has joined. 17:57:12 -!- jix has quit (Nick collision from services.). 17:57:26 -!- jix has joined. 18:05:31 -!- Arrogant has joined. 18:14:41 -!- ihope has quit (Nick collision from services.). 18:15:09 -!- ihope_ has joined. 18:19:45 http://www.piepalace.ca/blog/asperger-test-aq-test/ 18:19:51 36 !! 18:24:34 "When I talk, it isn't always easy for others to get a word in edgeways. " like... it's hard to get to say something when i'm saying something? 18:24:56 i actually cannot translate "edgeways" there... 18:25:38 That's partially because the phrase is "can't get a word in edge-WISE" 18:25:51 i do know. 18:25:58 And it means nobody will shut up for long enough to let you talk :) 18:26:02 i still don't know what it means there :< 18:26:05 ah 18:26:12 so i totally failed what it mean 18:26:13 *means 18:26:19 So when this guy talks, he just babbles on and on and on and nobody else can talk. 18:26:28 so i totally got it right 18:26:33 Yeah. 18:26:47 i just wasn't sure 18:26:51 And it means nobody will shut up for long enough to let you talk :) << Was referring to the normal phrase, not his variation :) 18:26:52 now i am 18:26:59 ah 18:27:29 "I would rather go to the theatre than a museum. " you'd have to pay me a lot to get me to either... 18:29:32 -!- pikhq has quit ("leaving"). 18:30:57 I'd rather go to the moon than a theater 18:30:59 hmm... i can remember any phone number easily, for just about as long as i like... then again there's no way for me to remember a birthday... so i prolly get "average" on that just for having a weird memory 18:31:27 well, i can remember any number as long as it doesn't mean anything to me 18:31:43 when it gets a meaning, i remember the meaning 18:31:53 * SimonRC has dinner 18:32:23 SimonRC: you got 36? 18:32:30 i'm a mere 28 :\ 18:32:32 yes 18:32:42 * SimonRC is geekier than oklokok! 18:32:44 * SimonRC has dinner 18:32:44 i have a feeling i'd get a better score if i actually talked to a doctor :P 18:33:01 better == bigger 18:34:16 -!- poiuy_qwert has joined. 18:34:58 -!- ihope has joined. 18:44:28 You need javascript for it to work... 18:44:51 44, it says. 18:44:59 Not that I answered all the questions honestly. 18:47:16 oooh, aspies test 18:47:21 * sp3tt tries 18:47:33 I got like 120/140 on one once :( 18:48:48 well, i'm pretty good at getting friends, but people usually think of me as "the weird guy" 18:50:00 i really can't believe any of you is less geek than me 18:50:21 ... 18:50:22 morte 18:50:23 *more 18:50:24 38! 18:50:45 the test must be crooked! 18:52:07 I got it up to 49. 18:53:06 actually, i can easily think of any of those options suiting me 18:53:51 those are general big-picture questions, so i basically just pick a random situation i've been in and answer based on that 18:56:12 "I enjoy doing thing spontaneously." <<< thing is a character in addam's family 18:56:19 it's a hand 18:56:24 does this refer to masturbation? 18:56:43 like... hold on a min, i'll do some quick masturbation 18:59:17 i guess i'd have the same answer were it either 19:00:35 doing thing is a bit different than masturbation 19:01:52 well, i'm deriving it from the fact "masturbation" means "desecrating the hand" 19:03:29 Doing Thing. 19:04:14 "bbl guys, gotta do some thing ->" 19:04:29 "i hope you mean something" 19:04:33 "err... yes, sure" 19:04:45 'something' 19:04:58 i hate how i always fail quoting 19:26:00 -!- GregorR has changed nick to GreorR. 19:32:47 Actually I think it is "The Thing". 19:33:27 The Thing is a movie, Thing is the hand from The Adams Family 19:33:39 ah, ok 19:34:06 the thing is something by King? 19:34:53 no i dont think so 20:17:30 -!- GreorR has changed nick to GregorR. 20:18:42 random idea for some snippit of code: 20:19:03 accepts mouse-written drawings, then using some kind of OCR, converts it into mathematical notation in real time 20:19:30 so you could e.g. draw 231, a line beneath it, then 500 beneath that, it'd convert it correctly, then display a latexified (or whatever) version at the same size, replacing the drawing 20:19:44 once you've drawn the whole thing, it'll be converted to latex + displayed as an image 20:19:53 and, presumably, you could evaluate it or whatever 20:20:08 -!- oerjan has joined. 20:20:38 ehird`: like maple? 20:20:57 hmm.. i'm not sure if it does that though... i just know it has something like that 20:21:07 prolly something ridiculously simple 20:21:08 well, i'll show a sort-of-example 20:21:15 i write "1 +" using my mouse 20:21:45 within some very short space of time (hopefully <1 second), it'll disappear and be replaced by whatever latex looks like when rendering the equivilent equation, at roughly the same size 20:21:46 have you seen the video where the guy draws a cart on a blackboard then draws a play button and makes it move? 20:21:50 same as i carry on with the expression 20:22:03 and no 20:22:06 a touchpad that looks like a blackboard that was 20:22:08 cool vid 20:22:42 but anyway 20:22:43 you could draw strings and solid objects and make them move realistically in 2d 20:22:48 uh i wan tthat 20:22:52 it'd be pretty nice to be able to, e.g., get a tablet 20:22:56 start up an app 20:23:08 write some mathematical notation, look up at the app and see it in rendered latex form 20:23:09 also, your thing would be fun if you have a touchpad 20:23:15 and be able to evaluate it and stuff 20:23:54 like maybe if you wrote = and left it for a second it would evaluate :) 20:24:20 might be good for code too... 20:24:22 hmm... an ski parser that did that would be fun :P 20:24:33 * SimonRC thinks 20:24:35 i know there's a flash thingie, but you don't draw anything in that 20:24:47 IWBNI you could draw decision trees and tables and turn them into programs 20:25:06 I am sure this has actually been done before, but not, I suspect, with handwriting recognition. 20:25:24 Now there is a nice PhD topic 20:25:24 SimonRC: sounds good 20:25:41 -!- Arrogant has quit ("Leaving"). 20:25:41 SimonRC: but i think mathematical notation lends itself to being written more than, e.g. a decision tree 20:25:53 How about Haskell? 20:25:56 anything mathematical is simplest when you write it 20:25:59 thus.. 20:26:14 however, programming languages and decision trees and stuff - maybe not, they were designed for keyboard entry 20:26:28 maybe lisp. 20:26:40 Have you seen a Haskell paper? 20:26:46 nope 20:27:02 Haskell as programmed in is a mere shadow of the notation that the programmers have in their heads 20:27:22 in papers, the ASCII-art becomes nice math notations 20:28:01 when i program haskell i see haskell code... not mathematical notation 20:28:40 (<- => -> become arrows, | bceomes a proper bar, ++ join together, <+> <*> become symbol-in-diamond, etc) 20:28:56 i don't see that =p 20:29:48 they use a preprocessor called lhs2tex 20:30:11 I know 20:30:43 I feel that that notation is the "real" Haskell, and the programmed format is an approximation to it. 20:31:01 i think haskell is what the haskell report defines it as. 20:31:18 the programmed format may be older. 20:32:18 show me an example of the notation 20:34:09 <.< 20:34:22 heh 20:35:24 a big sample 20:35:26 where it's used lots 20:35:29 so i can see what it looks like 20:36:45 aaah, SimonRC beat me on the aspergers test 20:36:50 i got a 30 20:37:19 aspergers test? 20:37:39 would this be a scientific, based-in-reality test or a quiz test 20:37:46 quiz 20:38:14 link 20:38:44 http://www.piepalace.ca/blog/asperger-test-aq-test/ 20:39:22 it might as well be called a "social ability test" 20:40:00 except for the bits about dates and telephones and obsession 20:41:07 i'll probably do shite on it 20:41:09 and i have no idea why 20:41:25 quiz/test-type things just always give the opposite of what i really am :p 20:42:54 um 20:42:57 you are in denial! 20:42:57 WHERE is the submit button 20:43:07 you fill it in, then you get an answer at the bottom 20:43:11 THERE is none 20:43:15 hm wtf 20:43:17 49/50 20:43:17 it has some javascript fail 20:43:20 which did i miss 20:43:31 ah there 20:43:33 lots of scheme stuff replaces the word 'lambda' with the symbol lambda 20:43:39 i hate that 20:43:43 " Scores over 32 are generally taken to indicate Asperger's Syndrome or high-functioning autism, with more than 34 an "extreme" score." 20:43:45 i scored 35... 20:44:17 well; that fits considering i do have aspergers 20:44:23 the letter lambda looks terrible in a sexp 20:44:38 it only looks good in church's notation 20:45:36 at any rate, "texed" haskell certainly looks pretty 20:45:52 but ascii is not too shabby either 20:46:12 i think i got some of the questions wrong; anyway 20:46:15 yeah, the tex haskell does look nice 20:46:22 i protrayed myself to be a bit more sociable than i am :p 20:47:20 i think the whole point of that test is that it examines your opinion of yourself, rather than what you actually are 20:47:30 for example, the first question "I prefer to do things with others rather than on my own." 20:47:35 that doesn't even mean anything 20:47:39 what kind of "things"? 20:47:54 sex is the only thing i like doing with other people 20:48:14 i took it to mean any personal contact 20:48:52 -!- wellons has quit ("ERC Version 5.1.2 $Revision: 1.796.2.4 $ (IRC client for Emacs)"). 20:49:10 i certainly prefer to do foo, bar and baz with others, while qux and quux i will gladly do on my own. 20:50:39 "I am fascinated by numbers." - does this belong in a test on aspergers? 20:50:42 sure 20:50:59 i hate numbers 20:51:18 i love numbers 20:51:19 :p 20:51:23 sure to the first statement 20:51:49 if there's any numbers in it, it's not abstracted enough 20:52:29 helios24 has always been a bit primitive to me 20:53:04 bsmntbombdood: just think of numbers as the skeleton category of Set 20:54:44 it's still a number, whether it's 2 or {{}, {{}}} 20:55:27 wee, i got 26 on the test 20:55:44 barely beating a math contest winner 20:58:28 but of course it's useless without some sort of an indication of possible error size/probability 20:58:50 which i suspect to be around +- 10... 20:59:00 or at least +-5 21:00:59 Ā± <-- is that a +- sign? 21:01:32 -!- olsner has joined. 21:01:47 SimonRC: yes 21:03:47 irc needs some sort of unicode macro system 21:03:55 \+- transforms into Ā± or something 21:04:07 ...no 21:04:15 unicode is evil 21:05:09 unicode is not evil 21:05:33 -!- RodgerTheGreat has joined. 21:05:47 * ehird` finds a joined ++ in unicode so he can say "unicode is ++good" 21:11:04 āŒ  21:11:07 āŒ” 21:12:12 someone should modify nethack and make it abuse unicode and control characters to hell 21:12:15 i bet you could get it looking really good 21:13:59 & 21:14:24 heh 21:14:30 UNICODE SMILEY ā˜ŗ 21:14:41 ā˜¾ā˜½ 21:14:53 ā™² 21:15:42 ooh 21:15:45 what about a unicode befunge 21:15:50 ā˜ instead of ^, etc 21:16:58 INTRODUCING THE ā€½ā™Æ PROGRAMMING LANGUAGE 21:17:06 Designed to be the first name solely in unicode. 21:17:20 oh jesus 21:17:22 my friend made unicode print quite cool mazes 21:17:44 ...yes, he programmed a maze generator in unicode 21:17:59 is unicode turing complete? 21:18:16 a maze generator *generating* unicode is a different matter entirely 21:20:43 turing complete unicode... that makes my mouth water 21:20:45 that's just crazy 21:20:48 a turing complete CHARACTER SET 21:21:39 that would be rather fun 21:21:53 just add some control characters that can do branching.... oooh... 21:21:59 fun 21:22:18 ;-) 21:23:24 the algorithm for bidirectional rendering of text seems quite intricate though - a few carefully inserted extra rules could perhaps make it turing complete 21:24:13 i would love to see 99 bottles of beer in unicode using loops of some sort 21:24:30 this is probably rather like my theory that HTML by itself is *very nearly* TC 21:25:10 umm 21:25:11 no it isn't 21:25:19 it has nothing that would make it close to tc 21:25:28 it is in every way the archetypical markup language 21:25:36 you can do HTTP redirects (looping) 21:25:50 that doesn't really count 21:25:53 form elements and passed parameters via URLs also offer some possibilities 21:25:58 that's not HTML 21:26:01 that's HTML+HTTP 21:26:07 or HTML+some way to handle passed parameters 21:26:08 psh 21:27:11 javascript ;-) 21:27:52 I had an idea a while back for a network protocol in which the packet format represented a TC language that could be used for writing all kinds of networking utilities, games, file-transfer programs and viruses 21:28:17 sounds nicely esoteric 21:28:19 make a spec 21:28:40 RodgerTheGreat: make timing matter too ;-) 21:28:56 the entire language is one big race condition, hehe 21:29:00 and the fun part was that the system itself centered around being p2p- you'd "set up" a network by firing off a worm that would traverse the network and build routing databases on all the computers 21:29:44 that's neat 21:29:44 how do you stop exploits? 21:29:45 seriously 21:29:46 make it 21:30:02 bsmntbombdood: don't allow the interpreter to access HD or just about any unprotected external data, of course 21:30:11 you can still dos 21:30:22 you can ALWAYS do DoS 21:30:36 you can dos much easier if you're given a turing machine... 21:30:39 and yeah, scripts are sandboxed 21:30:54 RodgerTheGreat: "scripts" - it's just binary data, right? 21:30:55 not a HLL 21:31:04 kinda... 21:31:24 i think it should be low-level simple binary data... more like a protocol :) 21:31:33 make a simple assembler or something 21:31:43 yeah 21:31:45 it resembled Redcode a bit, actually 21:32:02 obviously you want the data actually transmitted as compact as humanly possible 21:32:11 self-modification and spawning new packets was a common practice, and the loopback device becomes really handy 21:32:13 otherwise the latency for something like, say the WWW, would be staggering 21:32:32 unicode could be tc if you added regexes natively and expanded them with a macro system 21:32:51 anyway, i'm gonna eat a food -> 21:35:18 I really love the idea of unicode macros 21:35:34 http://bash.org/?754254 21:35:34 RodgerTheGreat: how many opcodes do you think you could squeeze it down to? 21:35:39 8? 16? 21:35:43 16, surely 21:35:46 i wonder if there's an implementation alreday 21:35:49 16 would be good 21:35:53 yes 21:36:08 you could name them 0-F for obvious reasons :P 21:36:23 though for actually using the assembler maybe you could think of some better mnemonics, hehe 21:36:31 allow asm macros, though, that act just like an opcode 21:36:39 using only 16 opcodes all the time would be teeeedious 21:37:06 why 16? 21:37:14 bsmntbombdood: 4 bits 21:37:29 the more opcodes, the less you need, and it's easier to program 21:37:32 the spec was never finished, but I was also considering a high-level version that could be like "code"@AA.45.17.BB# to send a chunk of data off to the next machine 21:37:44 bsmntbombdood: this is a >network protocol< 21:37:52 compactness is more important than anything 21:38:04 you would have to make everything memory-mapped, and you use more space anyway for the addresses 21:38:19 there's plenty of tricks you could do 21:38:26 * SimonRC reads up... 21:38:45 using a full char would allow for a wide range of opcodes with the added benefit of being able to punch in code in a normal text editor 21:39:23 i'd assume you would use an assembler.. 21:39:26 sure, but it uses more space 21:39:33 and yeah, assembler is a must 21:39:36 otherwise, hex editor 21:39:37 you use less opcodes 21:39:45 how about 5 bits :D 21:39:53 32 opcodes should be enough for everyone! 21:39:58 lament: i was just thinking something like kinda like bnf for the macro system... and have tc regexes inbuilt in unicode format. just out of curiosity is that was you had in mind too? 21:40:04 like like like 21:40:10 1 is also a power of 2 21:40:21 0 bits ;-) 21:40:30 heh 21:40:36 "Your mom is so fat she sat on a binary tree and turned it into a linked list in constant time!" 21:40:44 Technically meaningless, but still funny :-) 21:40:45 oklokok: i was just thinking of something like tex markup... 21:40:47 ihope: lmao 21:40:58 (From SimonRC's link above) 21:41:06 hmm 21:41:15 making a useful machine with 16 opcodes would be a challenge indeed 21:41:18 a fun one, too 21:41:22 oklokok: otherwise it would be too smart, and do stuff when you don't expect it to 21:41:40 i'll have to look up tex markup 21:41:43 (just because you type +-, doesn't mean you always want it turned into a single symbol) 21:42:28 of course it can't be tex markup, because then you won't be able to discuss tex markup :) 21:42:39 MSN simply uses (foo) for its markup 21:42:42 (+-) 21:42:56 obviously that can't work in the context of programming discussions 21:43:09 people have so cute msn nicks 21:43:10 [c=3][c=38][c=27][c=23][c=29][c=46][/c]a[/c]n[/c]s[/c]k[/c]u[/c].krisu, joku teini<<33 21:43:31 oh god the eyes bleed 21:43:33 :P 21:43:39 (the eyes would bleed even more if i had msn plus and could see the colours) 21:43:43 heh 21:44:12 i could see the colors if i put that option on, but i hate colors and i love ascii line noise so that works better for me 21:46:41 * ehird` wonders what's the smallest a single instruction could be cut down to 21:46:48 4 bits for an opcode is reasonable.. 21:47:12 -!- poiuy_qwert has quit (Read error: 110 (Connection timed out)). 21:47:14 oh gods... http://bash.org/?8873 21:47:37 hello.jpg? 21:47:46 meta-goatse? 21:47:47 www.goatse.fr? 21:47:52 hello.jpg.jpg 21:47:53 .cz 21:47:59 a jpg containing a screenshot of hello.jpg 21:48:05 SimonRC: mirror 21:48:11 hello.(jpg^aleph_0).jpg 21:48:24 ehird`: indeed. The Tornado fighter used 4-bit opcodes 21:48:31 in its control computers 21:48:39 SimonRC: did it use 0-F as mnemonics? :p 21:49:24 ehird`: do you *really* think it had mnemonics? :P 21:50:01 hehehe 21:51:12 hm, opcodes don't need to be able to take immediate values 21:51:13 do they? 21:51:27 if you have a few registers.. 21:51:55 I guess you'd need load, store and some way to tell load and store where to load/store from/to 21:52:07 well yes 21:52:09 but apart from that 21:52:51 OP --> LD addr-to-data,reg; OP reg 21:53:06 but the address is a kind of immediate ;-) 21:53:17 i meant apart from the address :p 21:53:32 then no, you wouldn't need immediates 21:53:37 all i need is LD addr , ST addr and addresses 21:54:49 * ehird` makes decisions for the architecture 21:54:55 i think i'll use a sign bit it's simple 21:57:16 =) 21:57:42 -!- wellons has joined. 21:58:10 -!- Arrogant has joined. 21:59:12 hmm 21:59:13 -0 21:59:14 odd 22:03:22 :s 22:35:50 what are the implications of -0 22:41:39 YOUR MOM 22:41:42 heh 22:41:44 :/ 22:41:45 hmm 22:42:15 * ehird` thinks that he can squeeze a single opcode into 12 bits 22:42:24 err 22:42:26 s/opcode/instruction 22:43:12 no, wait, 8 bits 22:43:35 INS INS INS INS ARG1 ARG1 ARG1 ARG2 ARG2 ARG2 22:43:43 where each argument is a register name 22:43:47 of course, LOAD and STORE would be exceptions 22:44:10 err, what am i talking about 22:44:10 10 bits 22:44:41 maybe 9 bits 22:44:47 INS INS INS ARG1 ARG1 ARG1 ARG2 ARG2 ARG2 22:44:56 then all of the components are balanced 22:45:11 but... 8 opcodes... 22:46:05 ... good enough. 9 bits is insanely small! 22:46:51 8 is better than 9 22:47:11 no, i mean 22:47:12 7 is better than 8 22:47:19 7 is lucky 22:47:19 III111222 = 9 bits 22:47:25 and III gives me 8 opcodes 22:47:25 no, byte alligned is good 22:47:33 but 22:47:35 ok... 22:47:35 8 isn't lucky 22:47:38 so what should i make it? 22:47:50 add a parity bit 22:47:54 I I I 1 1 2 2 (SPECIAL QUANTUM CONTAINING BOTH A BIT OF 1 AND 2) 22:49:07 hmm, having LD and ST instructions longer than others could cause real problems 22:49:09 * ehird` wonders 22:51:11 someone suggest an extremely compact instruction format :p 22:51:52 gzip compressed 22:52:16 ...8 bit opcode, the following bits the operands, the number of which depends on the opcode 22:52:31 wait a minute, i should only need 2 bits for each argument 22:52:36 i mean, 3 registers is enough, right? 22:52:42 and might as well pipe it through gzip 22:52:43 no 22:52:52 you need memory access too 22:53:04 that'd all be done with some special instructions 22:53:07 like 22:53:26 LD REGISTER1 memaddr, MEM REGISTER1 22:53:29 ...some opcodes take 0 bytes of operand, some take 4 22:53:35 and then REGISTER1 would have the value at memaddr or something 22:53:43 and then you'd have MEMSET memaddr 22:53:44 or something. 22:53:45 keep it extensible 22:53:56 i'm going for "frickin' tiny", not "extensible" 22:54:19 being small is stupid if you can't use it 22:55:14 sure 22:55:16 i'm aiming for 22:55:34 1. relatively usable - if you're familiar with computer internals and asm, you should be able to program it with relative ease 22:55:37 2. as tiny as possible 22:56:23 -!- Sgeo has joined. 22:58:13 * ehird` wonders what bit-ness he should make the machine 22:58:14 4-bit? 8-bit? 22:58:30 ..8 bit 22:58:47 question: is nop actually essential? 22:58:54 i mean, i think you could get away without nop most of the time 23:00:24 damnit why is my computer making very high pitched and very short beeps randomly 23:00:32 not "beeeep" 23:00:34 more like "bip!" 23:00:37 you is been hacked 23:01:06 i is not been hacked 23:01:15 hmm 23:01:20 i really don't think NOP is needed 23:01:39 I think NOP is the quintessential non-essential operation :P 23:01:57 sure, but for asm... 23:01:58 *perhaps* useful to fill branch delay slots though 23:02:03 it seems to get an absurd amount of use 23:02:18 one very unneeded thing is ifneq 23:02:25 ifneq X Y Z can just be: 23:03:13 0 ifeq X Y 2 23:03:13 1 goto Z 23:03:13 2 stuff 23:05:21 i think i'll drop NOP 23:05:30 if anyone has a convincing argument for it, tell me 23:05:57 olsner: LD addr-to-data,reg 23:06:05 olsner: how would you put this data in the address? :) 23:06:50 ehird`: let's say addresses 0-200 is code and 200-250 is data... just LD 201,reg1 to load the contents of address 201 into register 1 23:07:10 and ST reg1,201 would put a recalculated value back into memory 23:07:16 ah, right 23:07:17 data sections 23:07:31 (modern assemblers have broken my soul :D) 23:07:46 not really - it could be loading code into registers to modify itself 23:07:48 only 200 bits for code though? lame :P 23:08:19 nah, just let bits 0-31 control the page number :P 23:08:29 good idea :p 23:09:21 i think i only need 2 bits for addresses 23:09:29 i mean, 4 addresses is enough to be comfortable with, right? 23:09:41 olsner: how would you personally define ST? 23:09:52 st reg,addr puts the value in reg in the memory addr? 23:10:23 4 addresses? that's just the same as having 4 registers though :P 23:11:24 err 23:11:26 i meant 4 registers 23:11:33 of course i shall use more for addresses :) 23:11:38 "i think i only need 2 bits for REGISTERS" 23:11:43 4 registers + memory 23:12:35 -!- poiuy_qwert has joined. 23:12:54 th 23:12:55 hmm 23:12:57 -!- sebbu has quit ("@+"). 23:13:01 i wonder how many bits i should use for addresses 23:13:41 7 23:14:04 77 23:14:05 only 128 places of addressable memory? 23:14:14 maybe just 8 bits. 23:14:35 now i have to decide what size to use for the values stored in memory and registers ;) 23:14:37 *:) 23:15:15 14 bits 23:16:01 12 23:16:48 people always ask for numbers, but then end up ignoring my nice numbers and having something boring like 16 23:16:54 heh 23:17:27 maybe just 8-bit values :) 23:17:33 well since he is refusing our good luck numbers, i suggest we compromise on 13. 23:17:52 maybe 8-bit values + 4 registers + 8-bit addresses isn't enough? 23:19:03 dunno 23:19:05 8-bit values + 4 registers + 8-bit addresses should be enough for anyone. 23:19:20 no seriously :p 23:19:28 do you think it would be enough to implement simple things? 23:19:55 no one is ever really gonna need more than 640k memory 23:20:40 i can't take the sarcasm any longer :( 23:20:56 * oerjan gives ehird` a lollipop. 23:21:09 <___< 23:21:19 * oklokok does /me because it's so cool 23:21:58 ehird`: i actually haven't read the logs so i don't really know what you are making :P 23:22:09 oklokok: i'm aiming for 23:22:09 1. relatively usable - if you're familiar with computer internals and asm, you should be able to program it with relative ease 23:22:09 2. as tiny as possible 23:22:09 i'm guessing cookies 23:22:20 a computer architecture, obviously =p 23:22:49 ah, so making a simulated asm for fun? 23:22:56 or emulated 23:23:20 or copulated 23:24:20 or you could just use brainfuck 23:24:26 -!- ihope_ has quit (Read error: 110 (Connection timed out)). 23:24:31 you only need 3 bits per opcode 23:25:18 brainfuck breaks rule 1 23:25:19 1. relatively usable - if you're familiar with computer internals and asm, you should be able to program it with relative ease 23:25:31 uh, brainfuck is usable 23:25:49 brainfuck can be coded with relative ease if you learn the basic trixxors 23:25:54 that is not the point 23:26:09 -!- Arrogant has quit ("Leaving"). 23:26:47 brainfuck plus arithmetic operators, 4 bits per operator and it's pretty simple to program in 23:26:47 :) 23:26:51 sigh 23:26:58 i am not looking to reinvent a brainfuck computer 23:27:16 i am looking to invent a very small architechture spec that is quite easy to program in if you know asm 23:27:45 you making like "the asm". 23:28:13 o 23:28:14 no 23:28:20 what the hell does that mean anyway 23:28:23 teh asm 23:28:28 -!- cherez has quit ("Leaving."). 23:29:54 Well, you'll want it to be very parallel, I think. 23:30:10 I think Game of Life is a decent platform. 23:30:25 i just want a simple low-powered computer that can still do 99 bottles of beer and some other stuff with at least mildly readable code 23:30:33 Though maybe BBM with quantum extensions would be better. 23:30:55 What, you want the assembly/machine code to be readable? 23:30:57 ihope: well done picking up where me and oerjan left 23:30:59 :P 23:31:10 lots of asm code is readable 23:31:16 if you know some asm for some platform, that is 23:32:43 Do you know what GCC does to multiply by 10? 23:32:50 ok, the only logic stuff i need is NOT, OR, AND, XOR i think 23:32:54 ihope: x86 asm is not readable 23:33:00 Oh. 23:33:03 gcc-generated asm is not readable anyway 23:33:22 Why do you want the assembly code to be readable? 23:33:34 yeah, who likes readability 23:33:48 Why not use a compiler from a high-level language instead? 23:33:57 you really don't get the point of this 23:34:04 No, I don't. 23:34:05 i mean readable as in not bloody brainfuck or unlambda! 23:34:11 i don't want some beautiful asm code 23:34:13 or anything 23:34:20 :P 23:34:36 ehird`: not trying to make you mad, just tired. 23:35:06 -!- olsner has quit. 23:35:10 What's wrong with BF and Unlambda, apart from the slowness? 23:35:22 they're hell to program in 23:35:24 absolute hell 23:35:40 You want to program in the assembly language? 23:35:41 gtfo of my #esoteric 23:35:44 * ehird` has store, load, move, xor, and, or, not ... hmm .. what next 23:35:49 bsmntbombdood: this is still esoteric :) 23:35:53 esoteric != hell to program in 23:36:00 ihope: At least 99bob, sure. 23:36:13 I see. 23:38:24 So you want readable and small. 23:39:36 Typed lambda calculus with data declarations/constructors? 23:39:55 I guess you also want it to look like an assembly language. 23:42:44 Wait, what do you mean by "small"? 23:43:11 Rather, what do you want that I stated as "small"? :-P 23:43:55 http://pastie.caboo.se/86457 here's my instruction set design 23:44:00 i think i might have screwed a part up 23:44:02 maybe it's not TC 23:44:49 Doesn't look infinite memory-y. 23:45:16 hmm 23:45:20 i'll rectify it tomorrow 23:45:27 Looks fine apart from that. 23:45:41 -!- ehird` has set topic: Esoteric programming language discussion | FORUM AND WIKI: esolangs.org | CHANNEL LOGS: http://ircbrowse.com/cdates.html?channel=esoteric | Rules: 1. Break at least one rule. 2: At least (10^10) bots must be on #esoteric | NOTE TO EHIRD DO NOT REMOVE: \instructionset. 23:45:47 i shall continue it tomorrow 23:45:51 Can't you send yourself a memo? 23:46:06 Won't clutter up our topic that way. 23:46:07 sure, put your shit in the topic... 23:46:11 -!- ehird` has quit (Read error: 104 (Connection reset by peer)). 23:49:16 bsmntbombdood: how often is bsmnt_bot reset? 23:49:31 what do you mean? 23:49:44 How often does it quit? 23:49:52 whenever i want it to 23:50:03 * ihope shrugs 23:50:56 ~exec self.raw("QUIT :Or we do") 23:50:57 -!- pikhq has joined. 23:50:57 -!- bsmnt_bot has quit ("Or we do"). 23:51:01 -!- bsmnt_bot has joined. 23:51:10 Indeed. 23:52:49 -!- ihope has set topic: Esoteric programming language discussion | FORUM AND WIKI: esolangs.org | CHANNEL LOGS: http://ircbrowse.com/cdates.html?channel=esoteric | Rules: 1. Break at least one rule. 2: At least (10^10) bots must be on #esoteric | NOTE TO EHIRD DO NOT REMOVE: \instructionset. 23:52:53 Er, oops. 23:53:15 At least that didn't do anything as far as I can tell :-)