00:02:44 what do you mean? 00:04:14 the kolmogorov complexity result. 00:04:23 can't find the optimal compression. 00:04:39 it doesn't need to be optimal 00:05:45 "8 repitions of 'a'" and "8 repitions of 'b'", while not any smaller than the original strings, still help to compare them 00:08:42 but it requires human interpretation. 00:10:33 no it doesn't 00:10:51 a computer just dumbly comparing programs would not be able to tell why that was essentially better than ["aaaaaaaa","bbbbbbbb"]!!0 and ["aaaaaaaa","bbbbbbbb"]!!1 00:11:21 compression algorithms... 00:11:40 8b1f 0008 de47 4612 0300 4c4b 0084 002e 66b6 add7 0009 0000 00:11:52 8b1f 0008 de52 4612 0300 4a4b 0082 002e 0d07 d1d7 0009 0000 00:12:01 -!- crathman has quit (Read error: 113 (No route to host)). 00:12:13 the first is 8 a's gziped, the second is 8 b's gziped 00:13:39 rather lousy compression :) 00:14:13 because it's so short 00:15:38 that method will not show similarity of strings, anyhow, unless you are very lucky with how the strings and gzip interact. 00:17:04 what you might do is to compare the compressed lengths of the two strings concatenated and each of them. 00:17:29 that might at least give some information 00:17:34 You could use a custom compression algorithm that is designed for this purpose 00:18:11 "aaaaaaaabbbbbbbb" gziped is 8b1f 0008 e00d 4612 0300 4c4b 8084 2824 02e0 1100 a19d 11b8 0000 0000 00:18:46 -!- CakeProphet has joined. 00:19:14 what if one of my strings is digits of pi, and the other is digits of 4-pi? 00:19:43 however gzip doesn't help with things that have no characters in common, does it? 00:20:05 happy_song: it can't consider _all_ similarities 00:20:13 oerjan: i dunno 00:21:54 bsmntbombdood: precisely 00:23:01 the similarity of A and B could be the length of a program to convert A into B 00:23:03 a la sed 00:23:33 a la Levenstein distance 00:26:36 -!- atrapado has quit ("multiverse?"). 00:27:33 "aaaaaaaa" -> "bbbbbbbb" = s/a/b/g 00:29:11 which reduces to kolmogorov complexity in the case A="" 00:29:32 right 00:49:18 -!- wooby has joined. 01:26:31 -!- sebbu has quit ("@+"). 01:27:25 -!- oerjan has quit ("Good night"). 01:51:48 I think levenshtein distance of compressed strings is a good similarity metric 01:54:57 using gzip, 01:55:05 dist("a"*8, "a"*8) => 0 01:55:15 dist("apple", "appel") => 4 01:55:28 dist("apple", "bannana") => 10 01:55:41 dist("a"*8, "b"*8) => 4 01:56:43 dist("a"*100, "b"*100) => 6 01:58:04 dist("abcde", "edcba") => 5 02:00:12 using bzip2 the results are ... different 02:08:24 -!- Pikhq has joined. 02:49:21 -!- jix__ has quit ("Bitte waehlen Sie eine Beerdigungnachricht"). 03:04:53 * Pikhq almost has an IME set up. . . 03:05:08 Just need to figure out how to get KDE to start with some environment variables exported. 03:05:10 Sukoshi? 03:05:42 . . . Found it. 03:07:07 -!- Pikhq has left (?). 03:11:39 -!- Pikhq has joined. 03:12:02 やった! 03:12:32 ARRRRRRRRR, TEH NOT ENGLISH 03:12:39 [You don't have #plof on autojoin :( :P] 03:16:30 少しがいるか? 03:16:42 ARRRRRRRRR, TEH NOT ENGLISH 03:16:53 It's for the other Japanese speaker in here. 03:54:04 ok, maybe min(levenshtein(a, b), levenshtein(compress(a), compress(b))) 03:57:24 polyglots with C and any language with #-style comments is so easy :) 03:58:07 because levenshtein(compress.compress()"the quick brown fox jumped over the fence"), compress.compress("the quic brown fox jumped over the fence")) => 6 03:58:18 where levenshtein(...) => 1 04:08:33 ちょっとね。 04:08:51 ....I want to write an operating system as a polyglot of a bunch of languages. 04:09:03 well... I say that. 04:09:05 but... 04:09:12 I'd probably go crazy 04:09:15 before taht happend. 04:10:12 -!- ShadowHntr has joined. 04:11:49 removing headers and shit from the gzipping should help 04:12:37 -!- happy_song has changed nick to lament. 04:13:00 most polyglots play around with comments to achieve polyglotness. 04:13:47 You almost have to. 04:13:51 C works well with bash-comments... you can make macros tha won't break the languages that treat it as comments. 04:13:53 yeah. 04:14:17 The ones that don't take advantage of comments take advantage of the intersection of language syntax :-P 04:14:47 heh... like... Python floor division and...... C comments???? 04:14:53 * CakeProphet attempts to think of a use. 04:15:16 well.. 04:15:20 that's still comment-based 04:15:28 You probably aren't going to find any between C and Python ... 04:15:33 I'm talking about things like C++ and Java. 04:15:35 function calls are nice and common. 04:16:03 yeah Python and C are tough... especially because of the differences in blocks. 04:16:09 -!- maverickbna has joined. 04:16:26 -!- ShadowHntr has quit (Nick collision from services.). 04:16:40 -!- maverickbna has changed nick to ShadowHntr. 04:17:25 -!- calamari has joined. 04:17:29 Perl and PHP are easy. 04:25:04 Perl is write-only, PHP is domain-specific (no matter what strange people want to do with it) 04:27:31 removing the checksum helped 04:27:38 python is good 04:28:43 helped A LOT 04:29:12 * CakeProphet ...likes Perl. :( 04:29:34 Bloody Perl. 04:29:38 Godawful language. 04:30:01 They miserably failed to draw "the line." 04:30:51 Pikhq: I'm here sort of. 04:31:59 * CakeProphet draws the line _________________________________________ 04:36:06 ok, CakeProphet, I think i've satisfactorily implemented an algorithm to compare strings 04:36:17 -!- ShadowHntr has quit (Connection timed out). 04:36:22 -!- lament has changed nick to happy_song. 04:38:33 and here she is: http://pastebin.ca/423400 04:41:11 note, that's the distance between strings, not a ratio of similarity 04:48:14 -!- Pikhq has quit (Read error: 110 (Connection timed out)). 04:55:24 hmmmm 04:55:31 what's it intended for 05:26:04 -!- ShadowHntr has joined. 05:36:20 CakeProphet: finding how similar two strings are 05:39:25 gives a value in the range [0..] 05:39:38 0 = the same string 05:46:16 -!- Sgeo has quit (Read error: 131 (Connection reset by peer)). 05:57:53 bsmnt, no I mean... do you want to find typos... or do you want "pure" difference in characters? 05:58:07 what do you mean? 05:59:26 it'll recognize typos, sure, but i wanted to create an algorithm that recognizes strings like "aaaaaaaaa" and "bbbbbbbbb" as similar 05:59:27 -!- FSi has joined. 05:59:55 -!- FSi has changed nick to Arrgh. 06:00:20 -!- Arrgh has changed nick to FSi. 06:01:14 -!- FSi has changed nick to Hardley. 06:04:23 -!- Hardley has quit. 06:06:50 data compression is interesting 06:07:10 oooh 06:07:32 for typos... addings a transposition operation does wonders. 06:10:01 transposition in compressed data won't mean much 06:14:46 yeah 06:15:09 data compression is the art of finding similarities 06:19:40 I wonder how effective "macro-like" compression would be. 06:19:42 like... 06:20:09 say you split data up and then check each of their similarities. 06:21:04 you could find parts that are similar to a large number of these chunks... and abstract that as a "macro"... which can take arguments corresponding to the differences. 06:21:21 that's just backreferences 06:21:37 lz77 06:21:52 hmmm? 06:21:55 like... 06:23:20 x5=Hello,\0! 06:24:18 you could compress anything that had that pattern into a macro like x(World) x(Goats) etc 06:25:11 for redundency that isn't always linear. 06:27:45 -!- ShadowHntr has quit (Read error: 145 (Connection timed out)). 06:37:28 or you could just make the decompressor an interpreter for a turing complete language... 06:40:27 -!- GreaseMonkey has joined. 06:58:31 -!- CakeProphet has quit (Read error: 113 (No route to host)). 06:58:37 I'd like to see a step by step view of bzip2 compressing something 07:26:42 -!- GreaseMonkey has changed nick to CCC^. 07:28:35 -!- CCC^ has changed nick to GreaseMonkey. 07:31:42 -!- nooga has joined. 07:31:51 GregorR: not universe 07:32:03 only a scene with bounds 07:32:38 Wow, that's so many conversations ago O_O 07:33:01 photons would disappear when they are absorber or they leave the boundaries or hit the virtual photographic film 07:33:08 absorbed* 07:33:32 I believe I mentioned something about an "arbitrary bounds" 07:33:56 oh 07:35:04 http://en.wikipedia.org/wiki/Photon_tracing 07:35:06 check that out 07:35:14 it was done in java 07:35:32 i would code in C with inline asm 07:37:26 http://www.cpjava.net/raytraces/DRUN.GIF 07:37:34 results r stunning 07:40:37 i think the best part of fun will be collecting computers for distributed rendering 07:48:45 g2g bye 07:49:25 i woludn't exactly call that picture stunning. 07:53:02 amusing, maybe. 07:59:59 -!- clog has quit (ended). 08:00:00 -!- clog has joined. 08:01:04 eww, java 08:05:25 afk, food 08:21:26 -!- calamari has quit ("Leaving"). 08:56:05 -!- nazgjunk has joined. 09:15:06 -!- sebbu has joined. 09:39:36 -!- sebbu2 has joined. 09:51:19 -!- sebbu has quit (Nick collision from services.). 09:51:22 -!- sebbu2 has changed nick to sebbu. 10:35:43 gonna go sleep, gnight 10:37:02 -!- GreaseMonkey has quit ("one day to go, then school's out"). 11:47:01 -!- happy_song has quit (Read error: 60 (Operation timed out)). 11:47:57 -!- happy_song has joined. 12:20:39 -!- jix__ has joined. 12:48:12 # I don't feel like dancing # 12:48:41 Sorry about the pink text, but that guy has the most irritating falsetto voice ever. 13:10:31 * SimonRC indicates that this might be of some interest to people here: http://jededag.blogspot.com/ 13:10:45 I know more than a little about the story behind it 13:48:08 aaa 14:27:15 -!- helios_ has joined. 14:27:21 -!- helios_ has quit (Client Quit). 15:08:27 a q q 15:17:17 -!- crathman has joined. 15:18:12 -!- crathman has quit (Remote closed the connection). 15:20:03 -!- crathman has joined. 15:21:09 -!- crathman has left (?). 16:44:32 -!- happy_song has quit (Read error: 110 (Connection timed out)). 18:08:42 -!- CakeProphet has joined. 18:59:29 hmmm... 18:59:39 I need a server... so I can host a Subversion repository. 19:02:59 or.. maybe find a free host... do such things exist? 19:15:42 -!- nooga has quit (Read error: 110 (Connection timed out)). 19:27:31 -!- nooga_ has quit (Connection timed out). 19:32:42 -!- atrapado has joined. 19:42:56 -!- oerjan has joined. 19:44:38 -!- ShadowHntr has joined. 19:45:15 -!- CakeProphet has quit ("haaaaaaaaaa"). 19:52:31 -!- CakeProphet has joined. 20:16:25 -!- nooga has joined. 20:40:56 -!- LookA has joined. 20:41:14 Howdy 20:41:41 -!- fax has joined. 20:41:53 cool 20:42:03 It does exist. 20:42:31 O_o 20:42:38 why are tunes logging this 20:43:01 ? 20:43:25 Some people wake up in the morning and read #esoteric logs instead of the newspaper. 20:43:38 haha 20:46:41 * oerjan points out that he reads #esoteric logs _and_ the newspaper. 20:47:47 although truth be told, i usually read #esoteric logs first. 21:10:33 EgoBot: !help 21:11:16 The ! must be first on line 21:13:51 Oh, thanks. 21:22:50 -!- LookA has left (?). 21:29:47 -!- sebbu2 has joined. 21:35:17 -!- jix__ has quit (Read error: 110 (Connection timed out)). 21:35:39 -!- jix__ has joined. 21:48:05 -!- sebbu has quit (Read error: 110 (Connection timed out)). 22:01:03 -!- dub_ has joined. 22:16:29 -!- lament has joined. 22:20:29 -!- atrapado has quit (Nick collision from services.). 22:20:39 -!- dub_ has changed nick to atrapado. 22:22:05 -!- Ryacko has joined. 22:24:41 Please say "Hello, World!" 22:25:04 -!- Ryacko has left (?). 22:29:48 dooo 22:30:07 oh hi bsmntbombdood 22:30:59 you are welcome to join #lispcafe as is everyone else here :D 22:31:23 #lispcafe? 22:32:00 bsmntbombdood: yet another lisp channel, but there is good discussion a lot of the time 22:45:16 I had an idea for an esolang recently 22:45:38 aha 22:45:40 it'd be an emo-person based programming language called "Theirspace" 22:45:56 the emo people (enodes) are arranged in a graph 22:46:35 they can exchange messages via comments, and internally store opinions of other enodes (an endex, or emo-index). 22:46:53 endices are naturally the product of what other people say about a specific person 22:47:25 so comments can include phrases with specific meanings like "you suck" or "this person sucks" that influence endices. 22:47:46 (the opposite of sucking would be being "totally HxC" 22:47:48 ) 22:48:18 * oerjan ponders the effect of the message "NN always lies about other people" 22:48:51 and enodes can also store rules imparted via comments from enodes they respect, like "don't listen to " or "tell people who write poetry to kill themselves" 22:49:20 heh, gonad 22:49:31 two data types, testicles and ovaries 22:49:33 -!- sebbu2 has changed nick to sebbu. 22:49:53 rules can also be obeyed or ignored based on the cliques enodes belong to 22:51:01 what do you guys think? Does it show promise? 22:52:18 it seems like it requires AI 22:52:24 it would be interesting if you could acutally execute myspace.com 22:53:11 oerjan: what do you mean? The enodes would just be looking for and generating sequences of canned phrases 22:53:23 it would only be designed to *resemble* online conversations. 22:53:47 ok 22:55:14 although, I guess you could use Markov chains or something to make it more spontaneous 22:55:27 then you could feed in actual emo-speak and the like. 22:55:53 computation through genetics! 22:55:53 and generate a garbled mishmash of words resembling random emo-speak 22:56:03 ooh 23:04:13 -!- wooby has quit (Read error: 104 (Connection reset by peer)). 23:04:28 -!- wooby has joined. 23:38:01 so like.. 23:38:11 is there anything better than Windows command prompt out there? 23:38:23 uh huh 23:38:26 ...and if you say bash I will FUCKING KILL YOU 23:38:27 bash 23:38:36 I meant... for Windows 23:38:59 wtf would you use windows? 23:39:23 ...because I can't get my network adapter to work for Linux. 23:39:36 ...and because I don't think Linux is too much better than Windows. 23:39:42 It's too extremes. 23:39:45 two 23:40:22 then don't use linux 23:40:25 use bsd 23:40:32 >.> 23:41:20 CakeProphet, try cygwin , a unix environment for windows 23:41:26 got it. 23:41:33 it's kind of clumsy 23:41:38 to use. 23:42:00 your mom is kinda clumsy 23:42:02 to use 23:42:11 cygwin is fun! 23:55:20 -!- calamari has joined.