00:02:50 -!- FireFly has quit (Quit: swatted to death). 00:04:05 -!- ralc has quit (Quit: Leaving). 00:12:31 -!- copumpkin has quit (Quit: Computer has gone to sleep.). 00:13:26 -!- pikhq has joined. 00:14:42 -!- pikhq_ has quit (Read error: Operation timed out). 00:32:45 -!- azaq23 has quit (Quit: Leaving.). 00:45:39 EVEN MORE HOMESTUCK 00:45:46 elliott et. al. 00:46:01 Yeah um I check often enough you don't have to remind me. :p 00:46:50 elliott, you check several times a day? 00:47:04 Who doesn't? 00:47:33 Actually this time it was that I loaded the SA thread and someone mentioned a new update. :p 00:48:12 In the past few minutes? (Just want to be sure we're on the same page here) 00:49:14 -!- elliott_ has joined. 00:49:23 00:48:12: In the past few minutes? (Just want to be sure we're on the same page here) 00:49:24 Yes. 00:49:32 -!- elliott has quit (Read error: Connection reset by peer). 00:49:34 Ok 00:53:24 huh, someone renamed the despotic language section in http://en.wikipedia.org/wiki/Esoteric_programming_language to turning tarpit 00:53:53 although i don't think that's quite right... 00:55:16 ??? 00:55:19 turing tarpit and...turning tarpit? 00:55:30 well of _course_ it's a pun 00:56:17 Maybe they should instead rename it to "Despotic languages (turning tarpit)" 00:56:25 (In Wikipedia, I mean) 00:57:54 -!- TeruFSX has quit (Remote host closed the connection). 00:58:13 is LOLCODE really an esoteric programming language? 00:58:17 it's C with different syntax 00:58:32 it's the black sheep of the family 00:59:00 it's clearly _intended_ to be esoteric, yet it's so only in a very shallow way 01:06:18 it's shitty 01:06:26 although i don't think that's quite right... 01:06:28 howso? 01:06:31 that's what we agreed is a better name 01:10:26 Try harder!!! 01:12:18 elliott_: i mean they are not actually _synonyms_ 01:12:25 a turning tarpit has to _turn_ 01:13:05 oerjan: well meharharharbaraba said that despotic was just another name for a turning tarpit 01:13:08 so.. 01:13:10 [asterisk]so... 01:13:18 maybe "turning tarpit" is a misnomer already :) 01:17:16 -!- micah|insane has changed nick to micahjohnston. 01:18:51 well the three articles in the category pretty much fits the definition of turning tarpit on our wiki. it is just not the same as the definition on wikipedia (which was the definition of despotic language) 01:20:44 -!- zzo38 has left. 01:23:05 Patashu: LOLCODE is composed principally of fecal matter. 01:24:05 -!- CakeProphet has quit (Ping timeout: 240 seconds). 01:27:13 -!- copumpkin has joined. 01:34:10 It's thundery outside, and my scaredy cat has shoved herself uncomfortably under my chest :P 01:35:51 Missed opportunity to say "scaredy cat cat". 01:36:09 That would be "scaredy-cat cat" X-P 01:36:16 Scaredy cat-cat. 01:43:42 (diff) (hist) . . N Free download of 7 Khoon Maaf‎; 01:26 . . (+5,110) . . Anicgibcamp (Talk | contribs) (New page:


'''[http://popmovz.com/415894?pnr_id=667 Download 7 Khoon Maaf movie, CLICK HERE!!!]'''


All formats: DVD, DivX, Xvid, Avi, Mkv, Mp4, HD, Blu-Ray
Downlaod video fo...) 01:43:43 finally 01:43:53 ive wanted to download 7 Khoon Maaf for my entire life 01:43:56 and now i can do it for fee 01:43:58 free 01:43:58 -!- lameNOT has changed nick to lament. 01:44:01 thank you Anicgibcamp 01:44:03 thank you so much 01:44:07 :) 01:54:25 (diff) (hist) . . N Yutham Sei Divx Hd‎; 01:51 . . (+2,934) . . Easmapohand (Talk | contribs) (New page:


'''[http://popmovz.com/422588?pnr_id=667 Download Yutham Sei movie, CLICK HERE!!!]'''


All formats: DVD, DivX, Xvid, Avi, Mkv, Mp4, HD, Blu-Ray
Downlaod video for ...) 01:54:26 OMG 01:54:27 EVEN MORE DOWNLOADS 02:04:10 -!- micahjohnston has changed nick to micahjohston. 02:04:13 -!- micahjohston has changed nick to micah|irssi. 02:04:17 -!- micah|irssi has changed nick to micah|||. 02:04:21 -!- micah||| has changed nick to micahjohnston. 02:04:28 -!- micahjohnston has changed nick to ec|ec. 02:05:51 -!- ec|ec has changed nick to micahjohnston. 02:08:58 -!- micahjohnston has changed nick to mj|irssi. 02:09:19 -!- mj|irssi has changed nick to micahjohnston. 02:10:35 ... 02:10:39 micahjohnston: stop 02:10:50 clearly micahjohnston is a secret agent communicating in nick changes 02:11:15 we must accept this as a matter of utmost importance to global security 02:11:17 oerjan: no, the period of nick changes convey the information. 02:11:27 periods* 02:11:29 steganickraphy 02:11:34 the nick itself does not have any meaning. 02:11:54 ah a clever ruse 02:12:20 except this needs some heavy error correction to beat lag 02:13:20 in reality some kind of modulation would have to be in use. 02:13:59 -!- micahjohnston has changed nick to not-quite-correc. 02:14:15 -!- not-quite-correc has changed nick to theyreontome. 02:14:24 -!- theyreontome has changed nick to micahjohnston. 02:15:07 "not-quite-correc"? clearly this scheme is using self-reference 02:15:36 -!- micahjohnston has changed nick to oerjanisoverthin. 02:15:45 -!- oerjanisoverthin has changed nick to overthinking. 02:15:50 -!- overthinking has changed nick to micahjohnston. 02:15:58 story of my life 02:17:20 oerjan you're overthin 02:17:23 eat a sandwich or something :/ 02:19:11 is esco still alive? 02:19:25 -!- micahjohnston has changed nick to heisnotthin. 02:19:28 -!- heisnotthin has changed nick to micahjohnston. 02:22:21 -!- oerjan has quit (Remote host closed the connection). 02:22:28 -!- oerjan has joined. 02:29:26 lifthrasiir: doubtful :P 02:30:35 elliott_: yeah, actually (as i mentioned hours ago) i'm working on something like it 02:31:04 though the main architecture differs. 02:31:55 Is it less terrible than Esco? :) 02:32:01 Is this esotope? 02:32:06 yes 02:32:11 http://hg.mearie.org/esotope/esotope 02:32:19 IMO such a project is interesting if it shares backends. 02:32:25 Or something like that anyway. 02:33:24 well, esotope attempts to implement N^2 combination from input language to output language, when there are N languages 02:33:33 combinations* 02:34:34 but since it is infeasible, i instead implement much less number of those combinations and fill the remaining cases transitively 02:35:12 for example if one has a processor from Text to Brainfuck, and one from Brainfuck to Ook!, then you don't have to write a processor from Text to Ook! 02:35:27 so an EsoInterpreters type thing 02:35:42 of course it's suboptimal when the language is not Ook!, but better than nothing. 02:35:45 elliott_: exactly. 02:36:05 to this end i had to implement some kind of meta-object system in ocalm 02:36:07 ocaml* 02:36:27 looks feasible? :p 02:37:29 I haven't looked yet :) 02:37:32 but, obligatory - 02:37:35 Haskell Haskell Haskell ;D 02:37:39 lol 02:50:30 a common backend thing for esoteric languages might be interesting 02:50:41 like, most of the bf tape optimisations aren't specific to bf, they work for any tape-with-head structure 02:52:41 whats a good word for rejecting-out-of-snobbery? 02:53:27 augur: Douchebaggery. 02:53:36 something i can put in a paper :P 02:53:43 lifthrasiir: Have you considered using not-Make? 02:53:50 pikhq: oh please :P 02:54:01 elliott_: :P 02:54:53 Slightly more serious note: any particular reason for Ocaml? 02:55:03 -!- Aune has quit (Quit: Hath Deprated). 02:59:36 pikhq: i) of course i dislike make generally, but for now i'm going to stick to it. ii) to use both imperative and functional features. 02:59:55 * lifthrasiir afk 03:06:06 Haskell is good at that too ;D 03:06:30 Indeed, only in Haskell are imperative actions first-class. 03:06:33 :P 03:21:08 Does it ever get to be a pain using monad transformers? 03:27:14 -!- pikhq_ has joined. 03:30:00 elliott_, TV Tropes on Homestuck comments that the wiki takes what trolls say at face value 03:30:10 -!- pikhq has quit (Ping timeout: 252 seconds). 03:30:17 Does the entire community consist of portions taking pot shots at each other? 03:30:50 I don't know since I'm not stupid enough to read the TV Tropes page, which consists of an alphabetical listing of every trope on the entire site. 03:31:17 I've never found any incorrect information on the wiki but I'm not exactly cross-referencing things. 03:31:25 Where does it say that anyway. 03:33:30 http://tvtropes.org/pmwiki/pmwiki.php/Homestuck/TropesS-U search for Unreliable Narrator 03:33:42 -!- copumpkin has quit (Ping timeout: 240 seconds). 03:33:56 -!- copumpkin has joined. 03:34:50 probably some random person added that :P 03:34:57 it's not like anyone would notice anyone adding anything to those pages 03:35:01 considering they're all fives miles long 03:35:42 anyway who cares, the tv tropes page is interminably boring 03:44:55 -!- zzo38 has joined. 03:45:13 Esolang wiki is not supposed to be the movie downloading wiki 03:45:59 IT IS NOW. 03:46:20 zzo38: blanking the pages is a waste of time, wait for an admin 03:47:14 well maybe it reduces their google hits in some cases 03:47:44 elliott_: I know, that is why I only put it on one of them. 03:48:57 hm i wonder if google detects links that are frequently removed... 03:49:31 do we use nofollow 03:49:31 dunno 03:50:19 oerjan: anyway the permalink is still in recent changes 03:50:42 not as a link 03:51:14 * lifthrasiir back 03:51:45 pikhq_, and it was quite painful for me to get used on it. ;) 03:51:52 unless you mean to the article version. that simply _has_ to be nofollow, otherwise we'd get swamped like haskellwiki once was 03:52:07 or robots.txt, even 03:52:22 There is a new esolang added called "Hatter". I don't know if "\apply" is allowed 03:52:46 and i'm also trying to learn ocaml in the course of the development (i do write code in Haskell as my daily job...) so it actually does not matter :p 03:52:58 Also, 03:53:12 lifthrasiir: psht, real haskellers don't get paid for it 03:53:18 and yes, links are nofollow 03:53:18 [THINLY DISGUISED JEALOUSY] 03:53:31 Also, "A numeric constant behaves as a hat that always yields the corresponding value. Data dropped into a numeric constant is lost." Is the "nop" hat similar? 03:54:02 why does tvtropes still exist 03:54:17 elliott_, as a part of my research, actually. *wink* 03:54:26 Patashu: because humanity is dead. RIP. 03:54:28 Patashu: Probably because they have a lot of things and informations on there. 03:54:32 lifthrasiir: >:E 03:54:43 tvtropes is the next wikipedia 03:54:47 soon we'll be feeding it into our AIs 03:54:50 so they can impersonate nerds 03:54:53 god 03:54:53 no 03:54:55 most obnoxious ais 03:54:57 No, it is a different set of information, it is a different kind of things. 03:55:00 every sentence would start 03:55:03 "This Troper" 03:55:16 Wikipedia is one thing, TV Tropes is other things. 03:55:20 i don't think ais would like tvtropes fed into him 03:55:26 Patashu, no, but the AI will automatically generate a work using its knowledge of tropes in order to amuse itself. 03:55:29 zzo38: whenever i need tautologies 03:55:32 i can always rely on you to deliver them 03:56:02 Okay good point 03:56:13 It will read tvtropes but ignore the 'examples' list 03:56:14 oerjan: Unless it was the AI designed specifically for that purpose, maybe? 03:56:25 zzo38: whoosh 03:57:47 hey zzo38 03:57:51 burgers are made of koala fish 03:57:54 also 03:57:57 all koalas are space 03:57:59 your opinions? 03:58:13 elliott_: I don't know. 03:59:20 Make up the 'unexamples' list, in case of Uncyclopedia. 03:59:32 no 04:02:37 New kind of exotic code golf: Make amount of alnum same as amount of symbol, and put (alnum=sym) after your name. 04:02:58 so exotic 04:03:10 elliott_: would you be willing to critique a draft of an introduction to a paper im writing? 04:03:21 maybe if it was tomorrow and not five am today 04:03:23 or er 04:03:25 maybe if it was later today 04:03:31 so what i am saying is no, but yes if you're asking future me 04:03:34 later today is fine! 04:03:37 or at least what i assume future me is going to be like 04:03:44 he might actually decide to hate you way too much to do it 04:03:44 idk 04:03:45 future D: 04:03:46 timey wimey 04:03:48 youll have to ask him when he starts existing 04:03:54 after i become him 04:03:56 transform into him 04:04:07 REGENERATE, IF YOU WILL 04:04:18 yes that is definitely what you think is taking place here. 04:05:35 -!- micahjohnston has changed nick to everyone. 04:05:43 -!- everyone has changed nick to micahjohnston. 04:07:19 -!- zzo38 has left. 04:07:21 elliott_: wellnowwhat.net/linguistics/cwatc_draft.pdf 04:07:58 -!- SimonRC has quit (Ping timeout: 260 seconds). 04:09:03 i think you should remove all references to "i", they make the paper impure and ugly. 04:09:12 if i didn't want to pretend people didn't exist i wouldn't be reading something set in computer modern. 04:09:31 elliott_: :X 04:09:41 if you want more coherent opinions WHOOPS YOU'RE SHIT OUT OF LUCK but slightly more amusing (but equally incoherent) ones will be available once i've slept 04:09:44 i hate using the impersonal style 04:09:47 it feels unnatural 04:10:12 and i shall stick with the fun DT literature in using personal pronouns 04:10:30 or 04:10:31 in other words 04:10:37 if its good enough for conor, its good enough for me 04:10:56 yeah but theres a difference 04:11:04 conor mcbride is fucking god incarnate and pretty much the best person to ever live 04:11:05 youre not 04:11:08 becuase youre not conor mcbride 04:11:14 if you want to do that 04:11:17 you will have to become conor mcbride 04:13:02 this is true 04:13:06 about mcbride being awesome 04:13:09 i fucking love his accent 04:13:34 wheres he from, btw 04:13:39 he sounds northern of some sort 04:13:52 scotland 04:14:00 -!- lifthrasiir has quit (*.net *.split). 04:14:01 unless i'm very much mistaken. 04:14:03 well, he actually sounds a bit like arthur darvill to me 04:14:09 and darvill is from birmingham 04:14:28 well aren't his two universities he's been at both in scotland... 04:14:31 that's fairly good evidence :P 04:14:51 hm no wait 04:15:05 strathclyde is glasgow 04:15:08 well dunno 04:15:09 but isnt he also at nottingham? 04:15:10 yeah but nottingham isnt :P 04:15:16 well he used to be 04:15:24 is epigram two out yet 04:15:30 his current email is at nottingham 04:15:46 at least if his site is up to date 04:15:53 http://www.e-pig.org/epilogue/?p=802 04:15:55 "Maybe ()"? 04:15:57 what a pointless type 04:16:07 oh i guess it might be to use the maybe monad 04:16:32 and nottingham is east midlands 04:16:42 last myth adventures comic :( 04:16:52 so not tooo far from birmingham 04:17:14 Maybe () ~= Bool! 04:17:26 not quite 04:17:30 well 04:17:31 augur: not strictly true in haskell 04:17:31 Maybe () has one more element 04:17:35 modulo bottom 04:17:42 that's a dangerous thing to modulo :) 04:17:46 ;) 04:17:47 i imagine it's being used for the maybe monad here anyway 04:17:53 can't have a bool monad after all 04:18:01 yep indeed it is 04:18:05 Bot, True, False vs. Bot, Nothing, Just (), Just Bot 04:18:10 _|_ 04:18:11 ftfy 04:18:14 \bot 04:18:16 kthxbai 04:18:55 i am a full unicode cultist 04:18:59 "Silent cumulativity is a kind of subtyping, and if you are not scared of subtyping, you have not properly understood it." 04:19:23 elliott_: i read conor's paper on ornamental algebras todayesterday 04:19:32 todayesterday 04:19:33 or was it yestertoday 04:19:34 what a nice word 04:19:52 so what will you do todaymorrow 04:20:00 and what a fucking paper man 04:20:02 sleep hopefully 04:20:15 hopefully indeed 04:20:41 http://www.rpscontest.com/ this is cool 04:20:49 the top programs are all pretty short, too 04:22:16 ok i will now be going to sleep, indeed 04:22:18 night 04:22:31 /msg me with comments on the intro, elliott_, oerjan 04:22:37 why /msg 04:22:41 theres a perfectly good channel right here 04:23:12 elliott_: scrollback! :( 04:23:15 ok fine 04:23:17 just ping me 04:23:19 im sleeping too 04:23:26 when you comment 04:25:16 You know what makes me sad? 04:25:24 K&R C and Smalltalk are contemporary languages. 04:25:35 (Smalltalk is the *older of the two*) 04:26:05 are you sure about that 04:26:09 hmm 04:26:22 Smalltalk. 1972. C. 1973. 04:27:04 isnt c older than that, thats just the first release 04:27:12 it took until nineteen eighty for smalltalk to go public 04:27:13 soo 04:27:18 you should find C's inception date instead 04:27:21 which is like late sixties i believe 04:27:29 1969 vs. 1969, then. 04:27:53 Also, "The C Programming Language" came out in 1978. 04:28:41 yes, but unix w/ c came out in seventy-three 04:28:49 and people didn't not use c until the book did they :P 04:28:52 True, true. 04:29:21 Still. *Smalltalk and C are contemporary languages*. 04:29:22 ok im going to 04:29:22 sleep 04:29:23 right 04:29:24 fucking 04:29:26 now 04:29:32 pikhq_: i wish smalltalk caught on rather than c. 04:29:34 :( 04:29:46 i should really get an fpga set-up. xeightsix sucks. 04:29:46 bye 04:29:47 jesus 04:29:48 fucking 04:29:48 bye 04:29:48 bye 04:29:49 bye 04:29:51 bye 04:30:41 gnight 04:32:02 pikhq_: you critique too! 04:34:15 -!- elliott_ has quit (Ping timeout: 255 seconds). 04:42:41 -!- SimonRC has joined. 04:44:47 -!- lifthrasiir has joined. 04:56:49 -!- zzo38 has joined. 05:00:02 -!- pikhq_ has quit (Ping timeout: 240 seconds). 05:00:12 -!- pikhq has joined. 06:08:13 I looked at my C codes of Deadfish for a few minutes and failed to see how to make any more shorter. I already have three question marks and adding more will just make it even longer! 06:17:22 -!- hagb4rd has quit (Ping timeout: 246 seconds). 06:18:42 !c 1^256&&1 06:19:24 !help languages 06:19:25 ​languages: Esoteric: 1l 2l adjust asm axo bch befunge befunge98 bf bf8 bf16 bf32 boolfuck cintercal clcintercal dimensifuck glass glypho haskell kipple lambda lazyk linguine malbolge pbrain perl qbf rail rhotor sadol sceql trigger udage01 underload unlambda whirl. Competitive: bfjoust fyb. Other: asm c cxx forth sh. 06:19:34 hm 06:19:47 !c printf("%d",1^256&&1); 06:19:49 ​1 06:19:55 !c printf("%d",256^256&&1); 06:19:57 ​0 06:20:10 er 06:20:14 !c printf("%d",256^256&&256); 06:20:16 ​0 06:25:57 Y'know, it is really weird to think that once upon a time, typing was a relatively uncommon skill. 06:26:21 s/typing/touch typing/ 06:26:30 especially before the 19th century 06:27:06 Well, yes, it was obviously an uncommon skill before the invention of the typewriter. 06:27:19 Only time travellers could really be said to have the skill, after all. 06:29:22 -!- lament has quit (Ping timeout: 240 seconds). 06:43:32 What kind of C code are you trying to write? 06:44:08 who, me? 06:44:30 i just had an idea for doing the deadfish 256 test 06:44:33 oerjan: ^ does not mean what you think it means. :) 06:44:47 pikhq: want to bet? 06:45:06 Oh, wait, yeah, you probably are meaning to use it as bitwise xor. 06:45:18 Kinda glossed over the boolean ands for some reason. 06:45:51 Perhaps I should sleep. 06:46:24 -!- FireFly has joined. 06:56:22 oerjan: Remember the order of precedence of operators in C, and how && and || works. Whatever you wrote is clearly not working I just used something like (~x&&x-256) which results in zero when you are -1 or 256 or in one otherwise. 06:56:51 um what's not working about it? 06:57:07 !c printf("%d",3^256&&3); 06:57:10 ​1 06:57:16 huh 06:57:55 gah, && doesn't return its last argument :( 06:58:27 or wait 06:58:36 Yes, it is not the same as in JavaScript. && converts to boolean and || also converts to boolean. If you want like || in JavaScript you can use ?: in GNU C. 06:58:36 isn't that what ? is for? 06:58:55 (With nothing else in between the ? and : signs) 06:59:00 ...i was trying to find something shorter than using ? 06:59:21 you could still use ^ for the test though, it's shorter than != 06:59:22 Maybe we need to make GoC 06:59:25 Along the lines of Goruby 07:00:25 !c printf("%d",3^256?:3); 07:00:31 ​259 07:00:46 wow, I didn't know about that one 07:00:48 well that didn't help 07:00:52 so if the argument tests to true, it prints the argument? 07:01:11 huh? it prints always 07:01:15 x?:y is the same as x?x:y in GNU C. 07:01:30 If the left size of ?: is nonzero then that is the result and it will skip the part afterward, if zero then it does evaluate the right part. 07:01:34 what pikhq said 07:01:34 didn't know about that shortcut 07:01:42 but apparently ?: doesn't have low enough precedence 07:01:42 Just a GNU-ism. 07:01:47 pikhq: Yes, except that it will not cause "x" evalulated twice 07:02:28 zzo38: Well, yeah, it more accurately maps out to ({ typeof (x) _x = x; _x?_x:y; }) 07:02:38 (more GNU!) 07:03:30 iirc x?:y is also different than x?x:y in that it doesn't evaluate x twice 07:03:34 oh wait ?: is like ||, not && 07:03:45 that's what's wrong 07:03:50 monqy: As we discussed. 07:03:58 oh I must have missed it 07:04:16 !c printf("%d",3^256?3:); 07:04:17 is there an official golf challenge for a deadfish interp? 07:04:17 ​Does not compile. 07:04:23 bah 07:04:31 quintopia: yes 07:04:36 !c printf("%d",!3^256?:3); 07:04:36 if anarchy golf is official 07:04:38 ​256 07:04:42 BAH 07:04:46 !c printf("%d",(!3^256)?:3); 07:04:48 ​256 07:05:06 pikhq: no point in that 07:05:06 ... Fuck me, I should sleep. 07:05:15 anagol is the MOST official 07:05:26 !c printf("%d",3^256?3:0); 07:05:28 ​3 07:06:05 that one works 07:06:30 It's a pity golf.shinh.org doesn't let you have arbitrarily many input/output pairs 07:07:49 Yes it is true, but it is also true that you can post multiple submissions by () after your name 07:08:27 well I mean, you can't pose problems that force you to implement the entirety of a problem, just whatever you can fit in three tests 07:08:37 unless you multiplex tests together but that's just a pain to have to do for every entry 07:09:22 it would be okay even if they just allowed ten tests 07:09:23 for deadfish it is easy to add more than one test, since each input can have many commands. however the deadfish golf still manages to miss some important tests 07:09:40 no real golf challenge should require more than ten tests 07:09:40 yeah 10 would be good for most things 07:09:43 oerjan: Which ones? Be specific? 07:09:58 zzo38: it never increments or decrements to 256 07:10:25 so some of the submissions cheat by only testing on squaring 07:11:00 but everyone knows they are CHEATERS 07:11:06 heh 07:11:07 Patashu: I know... but, at least, you can set the deadline, so it can be checked afterward. You can also put like (cheat) (embed) (genuine) (exec) and whatever else afterward. 07:11:38 oerjan: O, yes, so that's how they made their C codes shorter... I don't mind posting such solutions as long as they are clearly marked as such... 07:12:21 i spy one such cheater IN THIS VERY CHANNEL 07:12:37 PROTECT US FROM THE CHEATER OERJAN 07:13:45 ...i haven't made any submissions. 07:14:11 oh wait, grammar ambiguity 07:14:45 zzo38: well, they might. i haven't seen the c code 07:15:11 You will see on Friday, then. 07:15:15 but i've seen python code which did use that cheat 07:15:27 and haskell 07:16:40 i considered making a haskell one but when my first try was much much longer than the leader... 07:17:15 Send it anyways, maybe... 07:17:20 Is haskell good for golfing? 07:17:35 well my first try was a modification of someone else's 07:17:42 Patashu: well it's better than java :D 07:18:11 oerjan: O, in that case, you don't need to. 07:18:12 lol 07:19:26 I think haskell is not-very-good for golfing ... maybe not bad, but many useful function names are long and lots of stuff require import statements 07:19:45 -!- BeholdMyGlory has joined. 07:19:57 olsner: All that means is you have different challenge to determine to use function with short name or whatever else it is. 07:20:17 I mean obviously haskell's never going to beat golfscript 07:20:21 The way my C code is written, it seems that trying to make it check only on squaring doesn't make it shorter anyways. (If I could make it shorter using cheats, I would use separate submissions using () after my name, but I can't.) 07:20:22 But is it entertaining to minimize 07:20:57 sure 07:21:01 Patashu: Yes, but each language is a separate challenge. 07:21:38 haskellers like to make point-free code, which sometimes is more compact. but not always, it tends to add parentheses. 07:22:21 of course some of that suffers from the import problem, which is not an issue with lambdabot 07:22:48 (except that lambdabot has too many imports sometimes so inconsistent things crash) 07:23:39 (for really compact point-free code, you want to import at least Control.Applicative, Control.Arrow and Control.Monad.Instances) 07:24:55 and probably Data.List and Control.Monad for higher order combinators 07:27:02 As far as I know, the tags you put after your name in anarchy golf, can be two kinds, cheats and exotics. 07:27:27 (embed) is a common one 07:27:54 Yes, it is a common one falling into the "cheats" category in my categorization of two kinds. 07:30:12 > nubBy(((>1).).gcd)[2..] 07:30:13 [2,3,5,7,11,13,17,19,23,29,31,37,41,43,47,53,59,61,67,71,73,79,83,89,97,101... 07:30:14 Common tags for "exotics" include (bin) (nobin) (alnum) (sym) while some others in the "cheats" category might include (embed) (cheat) (exec) (noexec) (genuine) (luck) (rand) note sometimes tags indicate presence of cheating and sometimes absence of cheating 07:30:30 Patashu: ^ 07:30:35 ooo, clever 07:32:28 there's one with scanl for fibonacci which i have forgotten 07:33:11 * oerjan googles 07:34:17 > fix ((1:).scanl(+)1) -- This? 07:34:21 [1,1,2,3,5,8,13,21,34,55,89,144,233,377,610,987,1597,2584,4181,6765,10946,1... 07:34:26 what does fix too? 07:34:27 that it was 07:34:43 it's the fixpoint combinator 07:34:44 Patashu: Fixed-point combinator. fix f = f (fix f) 07:35:35 oerjan: What ar eyou doing in this channel instead of #haskell? :-) 07:35:58 i was here before i was in #haskell 07:40:08 morn 07:40:26 morndu 07:41:12 it seems while my eyes were closed, some time passed. 07:41:33 i dislike this 07:46:27 so oerjan how about the fact that if you have two exact sequences of length 5 and you have an onto homomorphism between the first nodes, an isomorphism between the second and fourth nodes and a one-to-one homomorphism between the fifth ones, then if the middle is a homomorphism it is an isomorphism 07:46:58 isn't that just something 07:47:14 a very nice fact which i used in the first article i published 07:47:19 :D 07:47:21 seriously? 07:47:24 yes 07:47:27 :D 07:47:53 and i recall the referee wanted us to remove it, so we had to clarify that it was actually needed 07:48:32 is it used very often or do we just have very a similar aesthetic sense 07:49:10 because i just grabbed that from a list of 50 random "preliminary" definitions and lemmas 07:49:31 not sure why i quoted preliminary, did some heavy editing on that sentence 07:51:27 i should just read all your articles 07:51:34 SHOULDN'T TAKE TOO LONG HUH :DDDDDDDDDDDDDDDDD 07:51:52 well we were doing k-theory, which is a sort of homological algebra, and we had homomorphisms going between two exact sequences and needed to show the important one was an isomorphism - it was pretty much a perfect fit. 07:51:55 actually i probably still wouldn't get any of it 07:52:22 i have a paper on a k-theoretic approach to symbolic dynamics on my desk 07:52:59 ah, is anyone of Giordano, Putnam or Skau involved? 07:53:03 but i don't know what k-theory is, except that it's a sort of homological algebra apparently 07:53:19 putnam sounds very familiar, he must be in at least one of the papers i'm reading 07:53:37 i can check today if i manage to drag myself out of the armchair 07:53:57 oklopol: well there's also the putnam contest which is something entirely unrelated 07:54:08 never heard of that 07:54:13 (except for also being math) 07:54:43 iirc i've heard "the putnam article" at some point at uni 07:54:46 who's putnam? 07:55:27 ian putnam, a professor who collaborated with my advisor on a breakthrough paper 07:55:43 (my advisor being Skau) 07:55:45 * oklopol read "i am putnam" 07:56:08 what was the breakthrough paper abourt 07:56:10 *about 07:57:15 about classifying cantor set minimal dynamic systems up to orbit equivalence by their k-theory 07:59:03 so umm, cantor set minimal dynamic systems, what's that in symbolic dynamics lingo, minimal closed shift-invariant subset of the bi-infinite sequences + shift? 07:59:26 that first article btw was proving that this classification broke down if you went slightly outside cantor sets 07:59:37 ah yes i recall us talking about this 07:59:43 apparently i didn't quite get it since i have to ask 07:59:44 :P 08:00:50 i believe the things you mention from shifts are a subset of minimal dynamic systems, the expansive ones 08:01:57 which have the property that there exists an epsilon > 0 such that for any distinct points x and y you can apply the dynamics some number of times to them and get them further away than epsilon 08:02:02 well at least shifts are expansive on minimal systems 08:02:06 yeah yeah 08:02:20 but umm 08:03:27 hmm the cantor set is actually more like the one-way infinite sequences? 08:03:28 I Can't Believe It's Not Hitler! 08:04:10 well one-way or two-way are homeomorphic, you just need two-way to put a dynamics on it 08:04:54 but so can you somehow connect CA and this stuff? 08:05:11 but the bratteli diagrams we used are more like one-sided ones 08:05:44 zzo38: maybe it is hitler 08:06:01 well CAs have their dynamics sort of in the orthogonal direction, don't they 08:06:15 intuitively 08:06:30 i have no idea, do you mean since you think of the dynamics in the cantor set case as a *shift*? 08:06:41 it's shift invariant though 08:07:00 CAs are the continuous shift invariant functions yes 08:07:19 while minimal systems are shift invariant _sets_ 08:07:38 oh so shift-invariant? so the dynamics is the shift? 08:07:54 well those that are shift systems are so 08:08:11 yeah that's what i wanted to hear 08:08:25 otherwise, invariant wrt whatever is the dynamical transformation 08:08:51 anyhow so this is the same thing symbolic dynamics does really, except symbolic dynamics usually restricts to SFTs and sofic shfits so that you can play with matrices 08:09:21 and transition graphs 08:09:25 well we played with bratteli diagrams, and an infinite tower of matrices 08:09:49 (bratteli-vershik diagrams, to be precise) 08:09:51 wow your dad must be rich 08:10:03 >_> 08:10:12 because your toys are so much cooler 08:10:38 that was my advisor and my co-author, silly 08:10:55 " well we played with bratteli diagrams, and an infinite tower of matrices" 08:11:08 we 08:11:40 (i never got the impression he was particularly rich. the main algebra professor supposedly was rather well off, though.) 08:12:10 for reasons not related to her work, i think 08:13:47 also the dynamics invariance idea is also played with in cellular automata 08:14:23 because you often let your CA run forever first and then look at the dynamics on the set you get are a result 08:14:25 more precisely 08:15:15 If you have any feature request and/or ideas for TeXnicard, please tell me; I would like to know. 08:15:16 a clopen set U is inward if G(U) \sub U, and U's attractor is \cap_i G^i(U) 08:15:56 often enough we take the attractor of S^Z because that's what the CA looks like "eventually" 08:16:07 What is a clopen set? 08:16:09 hm 08:16:14 zzo38: closed and open set 08:16:48 ah and the attractor of S^Z is called the limit set, didn't remember that so had to define attractors :P 08:16:54 consistency is important 08:17:18 there are interesting results on the limit set 08:17:26 our dynamics was always invertible so we never had that 08:17:42 for instance my supervisor proved a rice's theorem for limit sets some years ago 08:17:52 of course, this is very different from the kind of stuff you ppl do 08:18:12 rice's theorem is the one that says you can't say anything about a turing machine's language 08:18:14 which is trivial 08:18:40 yeah 08:18:57 for CA, it means for any property P (nontrivial subset of limit sets that *actually occur*), you can't tell if your CA has that property in general 08:19:10 and this is much harder to prove 08:19:21 (no connection with the TM case) 08:22:26 but anyhow there are less computational questions as well and for the 1D case, a lot of fun little stuff is known about the structure of the attractor "tree" 08:22:48 minimal systems are the leaves i guess 08:22:59 i should go to work at some point i think :D 08:27:58 zzo38: in 0-dimensional spaces, clopen sets form a base for the topology 08:28:03 so they are very important 08:29:13 U is open <==> clopen set around each point of U within U 08:29:40 and yes, i'm just fucking with you, i don't except this to be a very useful lesson 08:31:11 oerjan: what's "the" definition of dimension according to you? 08:32:00 well for zero dimensions they tend to agree don't they 08:32:34 that's exactly what another guy who studies zero dimensional spaces said when i asked 08:32:39 but in the later work on topological measures, the important one was about refinement of coverings 08:32:44 ah 08:32:46 don't tell 08:32:48 it's umm 08:33:16 you can always find a refinement such that at most dimension + 1 sets intersect 08:33:17 ? 08:33:29 refinement being a.... subset of the cover? :P 08:33:33 although our spaces were compact, which i think makes some variations of _that_ identical again 08:33:44 no actually subset + smallening? 08:33:58 yeah 08:34:04 or actually 08:34:48 i'm not sure if you might have to take more than one subset of each original open set 08:34:55 sometimes 08:34:59 hmm right 08:38:21 i thought for about a day that i had a proof that if a space had dimension >= 2, then there existed nontrivial (non-lebesgue derived) topological measures, but that proof had a fatal flaw. 08:38:42 I don't really like the changes they made to Magic: the Gathering, except for the change of the name of the remove from game zone. That zone is still used in the game so it makes sense to change it. Other than that I do not like their changes. 08:38:54 say you have 2n balls on a circle, each connected through the middle of the circle to the ball on the other side, have each of them contain a point contained in none of the other balls, and so that the whole circle gets covered, then you have to cut some of the connections in two or you have a huge number of collisions in the middle 08:38:59 HMM 08:39:06 actually that's not really true 08:39:14 yeah kind of hard to come up with an example 08:39:51 oerjan: would that be a huge thing? 08:40:09 that thing you thought you did 08:40:40 well it would have determined precisely which compact spaces have nontrivial top. measures 08:40:58 does every compact space have a dimension? 08:41:12 hmm 08:41:21 makes sense with the refinement thing at least 08:41:22 it can be infinite of course 08:41:46 what can't these days 08:42:32 oh yeah topological measures, that was again something you talked about but i don't remember at all 08:42:43 and since the spaces are compact, you can never find a particular cover which needs an infinite subcover 08:43:04 so it makes no sense to go beyond infinity 08:43:13 yeah 08:43:56 beyond the first infinity 08:43:57 :P 08:46:27 can you explain topological measures again? :) 08:46:34 argh 08:47:03 well you have a compact hausdorff space, and its families of closed and open sets 08:47:14 yes 08:48:28 and then you have a set function from the union of open and closed sets to some [0, a] where a may usually be assumed = 1 08:49:05 ok 08:49:34 call it m, and the space X, so m(X) = a 08:49:45 The default Magic Word, "Abracadabra", actually is a corruption of the Hebrew phrase "ha-Bracha dab'ra" 08:49:49 -!- zzo38 has quit (Quit: zzo38). 08:49:56 :D 08:49:58 bye zzo 08:50:24 if A is an open or closed subset of B, then m(A) <= m(B). 08:51:12 what, that's CRAZY!! (i'm being sarcastic btw) 08:51:20 if A, B and C are open or closed subsets, A and B are disjoint and C is their union, then m(A) + m(B) = m(C). 08:51:37 so it's additive 08:52:10 but yeah i guess it's useful to open up the definition because of our crazy domain constraints 08:52:36 yes, but note that if A in B are open or closed subsets, then B \ A need not be, and in that case m(B \ A) does _not exist_, and there is no additivity. 08:52:56 yes, that's why it made sense to open the definition 08:53:05 but actually the way you usually define it 08:53:24 is exactly what you said, measurable disjoint things and measurable union => additive on those 08:53:39 do continue, or are we done 08:53:51 yeah it is possible to formulate the definition exactly like for ordinary measures 08:53:55 not quite 08:54:49 if C is a closed set and epsilon > 0, then there exists an open set O containing C such that m(O) <= m(C) + epsilon 08:55:05 (regularity) 08:55:22 yeah 08:55:22 i think that may be it 08:56:05 well that's pretty neat, why do we even bother with borel sets 08:57:11 does it follow that you can approximate open from inside by compact? 08:57:25 an important theorem is that if m(A union B) <= m(A) + m(B) for every open or closed A and B such that A union B is open or closed (subadditivity), then m actually comes from a borel measure. 08:57:26 (or closed) 08:57:32 yes 08:57:54 we have m(C) + m(X \ C) = m(X), always, which gives a duality 08:58:28 indeed 08:58:54 okay it's noon, i have to go :P 08:58:55 -> 08:59:04 bye 09:02:56 -!- monqy has quit (Quit: hello). 09:12:15 -!- copumpkin has quit (Ping timeout: 250 seconds). 09:12:42 -!- copumpkin has joined. 09:18:33 -!- oerjan has quit (Quit: Good night). 09:44:44 -!- Vorpal has joined. 09:51:49 I need a C or C++ library for reading graphics files 09:51:51 What's good? 09:57:01 hmm, but if I use a library I have to supply the library with my assignment too 09:57:06 how are we expected to do it again ?_? 09:57:10 darn university assignments 10:26:56 You could possibly just use one of the trivial formats, maybe. 10:30:53 pikhq: ping 11:05:27 -!- Lymia has joined. 11:06:16 -!- Phantom_Hoover has joined. 11:06:44 Hello everyone. 11:06:44 Phantom_Hoover: You have 5 new messages. '/msg lambdabot @messages' to read them. 11:07:55 Let's send Phantom_Hoover 42 messages. 11:08:04 Noooooooooo 11:35:50 Yeah like .raw 11:35:54 Except I don't know how to convert into it 11:35:56 Otherwise that'd be good 11:36:07 Hmm 11:36:29 http://www.sydneyarchitecture.com/ROC/ROC006-Sydney_Opera_House_Sails.jpg I need a .raw of the kind of striped texture of the brown surface on this picture 12:18:46 nvm, PSP can save as raw. sweet 13:00:33 There are also other uncompressed formats with trivial headers. 13:01:05 I seem to recall Gimp has raw-writing capabilities too. (Or maybe it was just reading.) 13:06:21 -!- Lymia has quit (Ping timeout: 244 seconds). 13:08:31 -!- Patashu has quit (Quit: MSN: Patashu@hotmail.com , Gmail: Patashu0@gmail.com , AIM: Patashu0 , YIM: patashu2 .). 13:53:41 -!- sebbu has quit (Read error: Connection reset by peer). 13:54:10 -!- sebbu has joined. 13:54:13 -!- sebbu has quit (Changing host). 13:54:13 -!- sebbu has joined. 13:57:28 -!- Lymia has joined. 14:01:04 -!- Lymia has quit (Client Quit). 14:01:21 -!- Lymia has joined. 14:07:16 -!- Lymia has quit (Quit: 1... 2... 3... HUGS! :D). 14:07:34 -!- Lymia has joined. 14:08:14 There are also other uncompressed formats with trivial headers. <-- I don't know what .raw is, but if it is at all related to raw camera images it will be quite a pain to work with 14:09:11 -!- Lymia has quit (Client Quit). 14:09:31 -!- Lymia has joined. 14:13:19 -!- Lymia has quit (Read error: Connection reset by peer). 14:24:35 -!- copumpkin has quit (Quit: Computer has gone to sleep.). 14:39:45 -!- ais523 has joined. 14:53:03 -!- copumpkin has joined. 15:14:37 -!- pikhq has quit (Ping timeout: 255 seconds). 15:14:40 -!- pikhq_ has joined. 15:18:05 -!- sebbu2 has joined. 15:18:05 -!- sebbu2 has quit (Changing host). 15:18:05 -!- sebbu2 has joined. 15:21:38 -!- sebbu has quit (Ping timeout: 260 seconds). 15:38:43 -!- lament has joined. 15:55:07 -!- FireFly has quit (Quit: swatted to death). 15:56:21 -!- EgoBot has quit (Remote host closed the connection). 15:56:33 -!- EgoBot has joined. 16:20:47 -!- lament has quit (Ping timeout: 260 seconds). 16:33:27 -!- Lymia has joined. 16:33:27 -!- Lymia has quit (Changing host). 16:33:27 -!- Lymia has joined. 16:48:12 "Normally water and electricity don't mix." 16:48:16 Seriously, Cracked? 16:48:18 Seriously? 16:49:17 I can't tell whether you're making a comment at the obviousness of that, or that it's technically not distilled water that conducts electricity (iirc) 16:51:02 I'm saying that water and electricity mix extremely well. 16:51:19 Ah, lol 16:52:43 I could even explain the logic behind their statement but it's so stupid I... can't. 16:54:31 And distilled water does conduct electricity a little, just not nearly as much as water with dissolved ions does. 16:54:55 actually, distilled water only conducts electricity because of dissolved ions 16:55:01 it turns out water actually can dissolve in itself 16:55:05 although it saturates quickly 16:56:09 Yes, OK, but I didn't really feel like explaining dissociation to be technically correct. 16:56:29 how much water can you dissolve in 1L of water? 16:58:31 I can't remember offhand 16:58:41 but I know it's used as the basis for the definition of pH 16:59:12 pH = -log_10 [H+] where [H+] is the concentration of H+ ions. 17:00:18 -!- TOGoS has joined. 17:00:38 How much water can you dissolve in 1L of dissolved water? 17:00:49 It doesn't work that way. 17:01:16 It's a joke. 17:01:49 There's often not much difference between jokes and interesting questions. 17:05:31 (I suppose you could say that a litre of water will have 10^-7 mol of dissolved water in it.) 17:05:48 -!- monqy has joined. 17:07:40 http://en.wikipedia.org/wiki/Experiment 17:07:49 I applaud whoever got that image onto the article. 17:09:09 * pikhq_ is still confused by Firefox's absolutely stupid decision to make a self-signed SSL certificate scream loudly 17:09:40 Which of the following is more secure, cryptographically: an unencrypted message, or an encrypted message with no assurance of who sent it? 17:09:42 Which image? 17:10:00 If you said "the latter", congrats, you're not a complete, blithering moron! 17:10:33 pikhq_, it does become a problem if you're expecting the certificate to verify the owner... 17:10:43 I don't think that is "very often" 17:10:57 Phantom_Hoover, which image? 17:11:01 The first one. 17:11:52 Lymia: It could at least make a distinction between "this is encrypted" and "the CAs we trust have verified this". 17:12:37 Also. 17:12:56 Don't self-signed certs still have the power to stop man in the middle attacks, or phishing? 17:12:56 Well, beyond "ZOMG THIS IS ENCRYPTED WITH A SELF-SIGNED CERTIFICATE! IT WILL RAPE YOU, GIVE YOUR COMPUTER A VIRUS, AND KICK PUPPIES!" 17:13:40 Lymia: The *only* difference between a self-signed cert and a CA-signed cert is that there is no verification of the identity of the cert holder with a self-signed cert. 17:14:11 Which also mean it doesn't help in man-in-the-middle attacks, because the MITM can just present his own self-signed cert and pretend to be the destination server. 17:14:21 * pikhq_ is still confused by Firefox's absolutely stupid decision to make a self-signed SSL certificate scream loudly 17:14:29 I am confused by your not using Chrome. 17:14:41 Chrome Adblock sucks balls. 17:14:42 fizzie, if you already have a copy of the cert, you can stop it, can't you? 17:14:46 I am going to be confused if Chrome allows self-signed certs any more silently. 17:14:47 It used cert A before. 17:14:49 It now uses cert B. 17:15:06 Lymia: Yes, but no browser complains on unexpected certificate changes. 17:15:18 It should be cause for concern, shouldn't it? 17:15:46 Not in the CA-driven (broken) trust model. :p 17:16:14 SSH host key checking works like that, though. 17:16:35 True, the CA-driven trust model is pretty fundamentally broken. 17:16:40 I am going to be confused if Chrome allows self-signed certs any more silently. 17:16:42 I don't know many people who get the initial host key in any very secure manner most of the time. 17:17:09 IIRC they display an untrusted certificate by striking through their SHTTP thing. 17:17:22 Rather than opening an alert to tell you. 17:17:43 Phantom_Hoover: Huge difference between "WARNING WARNING WARNING THIS IS UNSIGNED. IT WILL RAPE PUPPIES. WOULD YOU LIKE TO MAKE AN EXCEPTION?" 17:17:52 "DOING SO WILL RAPE TWICE AS MANY PUPPIES." 17:17:56 Yeah, that's what I mean. 17:18:18 What's a bit strange is that even though it opts for that stance, it then makes the exception permanent by default. 17:18:30 Because that makes sense? 17:18:48 -!- copumpkin has quit (Ping timeout: 276 seconds). 17:19:11 If they went for a truly secure stance, it'd warn you on non-HTTPS. 17:19:37 fizzie, "Hey, none of our CAs recognize this cert." 17:19:44 As, after all, self-signed HTTPS is strictly more secure than HTTP. 17:19:47 "Let's notify the user if it decides to change" 17:20:03 Sure, but if they want to make sure no-one misses the self-signed cert warning, they could show it by default every time. Though I guess with a permanent exception it does sort-of do the change notification, that's true. 17:20:53 fizzie: Except that HTTP < self-signed HTTPS. They warn for the more secure of the two. 17:20:59 Which is moronic. 17:21:19 HTTP is too common to warn on. 17:21:27 This is unfortunate. 17:23:16 And if they wanted to go for "true" security, they'd just do "Warning: your computer has not been turned into a plasma. This may cause some of your personal data to be stolen. Click here to turn your computer into a ball of incandescent gas, a gigantic nuclear furnace." 17:24:48 -!- copumpkin has joined. 17:28:35 -!- sebbu has joined. 17:28:35 -!- sebbu has quit (Changing host). 17:28:35 -!- sebbu has joined. 17:31:51 -!- sebbu2 has quit (Ping timeout: 258 seconds). 17:53:28 has anyone here used the Oberon TUI? 17:53:44 apparently it's said that it doesn't have a prompt. how does someone run commands in it? 17:56:53 -!- sebbu2 has joined. 17:59:41 -!- CakeProphet has joined. 17:59:52 -!- sebbu has quit (Ping timeout: 248 seconds). 18:08:19 * Sgeo_ wants a nice emulator for it 18:08:50 Apparently there are penty 18:10:53 Not sure whether to emulate Oberon in VirtualBox or install it on Windows 18:10:58 Going to do the former first 18:12:23 Actually, what's Bluebottle? 18:12:40 -!- TOGoS has left. 18:13:26 Sgeo_, are you installing it now? 18:13:44 I'm downloading Bluebottle, not really sure what it is 18:15:21 it's based on AOL (active oberon language) 18:16:49 Sgeo_, i was specifically talking about this TUI: http://upload.wikimedia.org/wikipedia/commons/c/cc/OberonScreen.PNG 18:18:08 Oberon has a text user interface (TUI). It combines the point-and-click convenience of a graphical user interface (GUI) with the linguistic strength of a command line interface (CLI) and is closely tied to naming conventions of the Oberon language. Any text appearing on the screen can be edited and used as command input. Nothing like a prompt is required. Although radically different from a commandline, the TUI is very efficient and 18:18:09 powerful[4]. A steep ascend in the early learning curve makes it difficult to start with. Its usage and programming interface is documented in Martin Reiser's book "The Oberon System." 18:28:51 nice clown 18:29:22 A CLI means that when you press "ENTER" the command is executed. With a TUI you can do pretty much the same thing, except that the command isn't executed until you click on it. 18:29:25 a-ha 18:30:56 Sgeo_, it's fairly nifty like that 18:31:06 except i bet it has no vim support 18:31:11 Bluebottle in VirtualBox isn't working for me 18:31:11 so it inherently sucks 18:31:24 maybe vb isn't bluebottle certified :X 18:36:41 http://www-old.oberon.ethz.ch/cli.html < that argument sucks. 18:36:59 he wrote it in 1999? it's mostly void due to things which already existed in 79. 18:40:35 00:53:24: huh, someone renamed the despotic language section in http://en.wikipedia.org/wiki/Esoteric_programming_language to turning tarpit 18:40:40 That was me, FWIW. 18:41:07 -!- CakeProphet has quit (Ping timeout: 260 seconds). 18:43:07 brainfuck's minimality borders on elegant and pure language design; in fact it is related to the P'' family of Turing machines. 18:43:13 "related" :D 18:47:26 Brainfuck vs Java. 18:47:26 Go. 18:47:58 -!- CakeProphet has joined. 18:47:58 -!- CakeProphet has quit (Changing host). 18:47:58 -!- CakeProphet has joined. 18:51:16 21:45:56: zzo38: not really. i only have opinions on math nowadays :\ 18:51:23 I am sorry oko this is clearly false. 18:51:35 You have opinions on which soft drink is best to have sex in, for instance. 18:51:40 oh right 18:51:43 well that and math 18:51:53 oklopol: well, P'' is a bit of a tarpitisation of BF 18:52:03 ais523: i thought it's the exact same thing 18:52:03 although the causality goes the other way 18:52:13 P'' combines < and flip-bit, I think 18:52:16 also has no I/O 18:52:42 oh it combines those. i guess also the binary thing is a bit of a tarpitisation. 18:52:58 (well, more so) 18:53:32 -!- oerjan has joined. 18:54:32 That was me, FWIW. 18:54:40 -!- copumpkin has quit (Ping timeout: 250 seconds). 18:54:51 oerjan, are you watching us in Stalker Mode? 18:54:54 well the definition of turning tarpit is _not_ what wikipedia now says it is 18:55:07 tur*n*ing tarpit? 18:55:11 (well last i checked) 18:55:13 is that a typo or different word? 18:55:18 olsner: it's a pun 18:55:38 ok. I don't get it. 18:55:50 is wheel a turning tarpit? 18:55:56 olsner: also a different word 18:56:17 oklopol: Whirl is 18:56:25 and a couple others 18:56:25 erm yeah that. 18:57:01 anyway a despotic language did not have to have its command arranged in a wheel-like fashion, so need not be a turning tarpit 18:57:21 *commands 19:02:58 how much water can you dissolve in 1L of water? 19:03:15 yes, how much? 19:03:23 ais523 said water is water-soluble. 19:03:34 10^(-14) something, which is why equal amounts of positive and negative ions gives pH = 7 by what Phantom_Hoover said 19:03:46 (aka neutral pH) 19:03:56 oerjan: yep, that sounds about right 19:04:07 what 19:04:18 how could that possibly have to do with 7 19:04:20 It's 10^-7 l/mol of both H+ and OH- ions, so it's effectively 10^-7mol/l of dissolved water. 19:04:28 oklopol: 09:40:35 actually, distilled water only conducts electricity because of dissolved ions 19:04:31 09:40:41 it turns out water actually can dissolve in itself 19:04:37 or wait 19:04:38 oklopol, pH is defined as -log_10 concentration(H+). 19:05:15 it's the _product_ of amount of positive and negative ions which is 10^(-14) 19:05:18 i'm just wondering why a square root is taken when there's an equal amount of A and B 19:05:25 * oerjan isn't actually looking this up, btw 19:05:34 oklopol: ^ 19:05:35 -!- elliott_ has joined. 19:05:52 oerjan: it's because the product stays approximately constant even if you add ions to the mix by hand 19:05:54 *oklopol: 19:05:59 by default you get equal amounts 19:06:04 " It's 10^-7 l/mol of both H+ and OH- ions, so it's effectively 10^-7mol/l of dissolved water." <<< and how does this have anything to do with the number 10^-14 :D 19:06:14 oklopol, it's just the dissociation constant. 19:06:16 but say, if you add enough H+ to get 10^-6 H+, you get approximately 10^-8 OH- 19:06:19 derp 19:06:20 so it multiplies to 10^-14 always 19:06:42 reading backlog 19:06:46 Water molecules will pull each other apart until the concentrations are at 10^-7. 19:08:02 interesting, that's pretty cool 19:08:10 does the solution of water in water conduct electricity? 19:08:31 cheater__: well that's what ais523 said which started this 19:08:48 cheater__: yes, although not too well because not a lot of water self-dissolves 19:08:53 ok, but i thought he meant some other ions 19:09:00 not water ions. 19:09:15 it was distilled water, it doesn't have any other kind 19:09:32 ais523, "self-dissolves"? 19:09:51 oerjan, distilled water is never 100% pure. i thought he was referring to that. 19:10:04 cheater__: no, I'm referring to the way that distilled water nonetheless contains dissolved water 19:10:10 ok 19:10:11 cheater__: some of the water molecules split into H+ and OH- ions 19:10:15 being pure, it couldn't contain dissolved anything else, right? 19:10:17 that's fairly revealing 19:10:39 in typical distilled water, what's the percentage of water ions to all ions? 19:10:53 There's no such thing as a water ion. 19:11:05 i mean ions that come from water dissolving in water. 19:11:14 presumably you can it least get it pure enough that the water ions dominate, or this wouldn't be much of a subject 19:11:19 it's just a shorthand 19:11:28 10^-7 l/mol is the concentration. 19:11:48 Phantom_Hoover: he's a german and i'm norwegian, we're predisposed to making up new words 19:11:49 concentration in the solution, ok 19:12:13 but if you take the mass of all ions in distilled water, and call that 100%, then what part of that is the ions that come from water? 19:12:21 -!- copumpkin has joined. 19:12:22 and what part is impurities? 19:12:26 ("what do you mean it's not a word, i just made it up") 19:12:37 > (10^-7 * 18)/1000 19:12:38 Not in scope: `^-' 19:12:42 cheater__: all of them come from ions 19:12:44 > (10^(-7) * 18)/1000 19:12:44 *from water 19:12:44 *Exception: Negative exponent 19:12:48 because it's pure by definition 19:12:50 -_- 19:12:54 > (10**(-7) * 18)/1000 19:12:54 so they couldn't come from anywhere else 19:12:54 1.8e-9 19:12:59 ais523, i mean in the real world, not on paper. 19:13:01 cheater__, there's your percentage. 19:13:26 ais523, hence i asked about "typical" distilled water 19:13:35 Phantom_Hoover, i think we're misunderstanding eachother 19:14:06 the number you posted is just the percentage of the mass of ions in the complete mass of the solution, yes? 19:14:08 |at 19:14:11 Yes. 19:14:18 ** 19:14:21 cheater__: you mean water that isn't completely pure, but is more pure than tap water? 19:14:32 i want the percentage of the mass of impurities in the mass of all ions. 19:14:51 ais523, i mean the kind of stuff i get when i buy "distilled water" in a chemist's shop. 19:15:14 cheater__: I wouldn't know offhand, and it would probably depend on the chemist 19:15:23 i bet 19:15:55 i guess you could ask "how much can you purify water before the process does not improve" 19:16:07 oerjan: in which case it would depend on the process 19:16:09 that might actually depend on technological progress... 19:17:12 like in the future you can do it much better, but you have to shoot individual ions into a hard vacuum to do it... 19:17:25 oerjan, but that's not going to accurately reflect the typicality of that process in the production of distilled water. 19:17:31 Honestly, cheater__, can you talk about nothing without trolling? 19:17:43 wtf? 19:18:12 Phantom_Hoover: "CG: DO I GET ON YOUR CASE FOR ALL THE TERRIBLE HUMANNING YOU DO?" 19:18:15 cheater__: well the point is what process is typical depends on what is available 19:18:29 oerjan, hmm, yes 19:18:42 Today's "poster title that caught the eye" entry: "Horror Video Scene Recognition via Multiple-Instance Learning" 19:18:51 say in five years they discover a much better process based on sieving through graphene, or something 19:18:57 "Along with the ever-growing Web comes the proliferation of objectionable content, such as pornography, violence, horror information, etc. Horror videos, whose threat to children’s health is no less than pornographic video, are sometimes neglected by existing Web filtering tools. Consequently, an effective horror video filtering tool is necessary for preventing children from accessing these harmful horror videos." 19:19:08 (From the abstract.) 19:19:25 just don't go to stileproject 19:19:25 fizzie: Could they be any more political about it? 19:19:45 elliott_, yes 19:20:04 fizzie: I thought these types of papers were usually all academic about it and pretended to be oblivious of the real-world applications. :p 19:20:05 oerjan, this reminds me of this one quote that said you shouldn't be writing your programs as well as you can, only relatively well 19:20:12 * oerjan notes that "no less" doesn't actually imply that either value is actually large 19:20:14 because to debug them you need more skill than to write them 19:20:14 Authors: Jianchao Wang, Bing Li, Weiming Hu, Ou Wu, National Laboratory of Pattern Recognition / Institute of Automation / Chinese Academy of Sciences, China 19:20:32 I guess they know about filtering content in China. 19:21:31 i like how an "academy of sciences" proliferates information hampering, whereas science stands for information propagation. 19:21:54 that's like the Ministry of Truth. 19:22:49 It's based on "color emotion and color harmony theories". 19:23:25 This is possibly the one context those phrases could be used without immediately discrediting the paper they're in. 19:26:35 On the "no immediately obvious applications" front, there was a rather nice talk about parametrizing "audio textures" (natural-ish sounds like rain, wind, fire, applause, etc., that are in some sense stationary), and then synthesizing "equivalent" new sounds. It did sound quite nice. 19:27:02 fizzie 19:27:08 please link me up to the relevant paper and demos 19:27:16 i'm 100% interested in this 19:27:18 They had invented a soundtrack classification application for it, but it really didn't work all that well, and the whole task felt quite forced. 19:28:14 The synthesis paper is "Sound texture synthesis via filter statistics", Proc IEEE WASPAA, 2009. Googling for the title finds the paper, and also a project that cites it. 19:28:26 http://www.cs.cornell.edu/projects/sound/fire/ 19:28:35 "Coming soon: source code for .. audio texture synthesis". 19:28:49 Resynthesizer... for SOUND. 19:28:51 That would be cool. 19:29:11 It might use the same thing; the paper for the fire thing is 11M so I haven't bothered to load it over this slow net. 19:29:27 At least they cite the texture-synthesis page. 19:29:37 What happens if you try and stretch out a terrible pop song. 19:29:41 Can I produce the next big hit in this manner. 19:30:11 It's about synthesizing fire sounds that are synchronized with a physical fire simulation video thing. 19:30:34 fizzie, where are the demos? 19:30:44 I guess they know about filtering content in China. <-- i briefly had the approximate thought process here: "hm, they're wanting to censor _both_ violence _and_ sex? hm that fits neither with americans nor europeans. oh right, chinese." 19:31:01 The link has some videos of the fire thing. 19:31:10 thanks 19:31:17 are there any other demos that you have heard? 19:31:20 -!- augur has quit (Remote host closed the connection). 19:31:20 If the US censored violence, our culture might... 19:31:25 Be less screwed up. 19:31:31 Maybe. 19:31:41 btw, on the front of sound resynthesis i think hartmann neuron is very interesting. 19:31:56 Lymia, Ted Kaczynski has some interesting views on that. 19:32:05 did you actually just say that 19:32:05 ahahahaha 19:32:19 The talk contained examples of rain, fire, applause, stream, bubbles, insects and wind; but I don't know where (or if) those would be in the interwebs. 19:32:32 elliott_, say what? 19:32:33 :( 19:32:40 Lymia: i was talking to cheater__. you may have him on ignore. 19:33:11 I am mentally filtering him out. Does that count? 19:33:15 I think ICASSP didn't do the "include multimedia files in your submission" system, so the proceedings memory stick just has the papers. 19:33:18 Lymia, no. 19:33:19 Lymia: Yes. 19:33:19 elliott_: hey it was a pretty funny comment i say 19:33:42 OK my breath is fogging up. 19:33:47 In late May. 19:33:49 Phantom_Hoover: Ha ha Scotland. 19:33:50 Indoors. 19:33:53 oerjan: /msg 19:37:29 fizzie, was the talk for the paper you told me to google or for something else? 19:38:21 cheater__: It was for a newer paper, but that was just taking the same parametrization they used for synthesis and applying that for soundtrack classification, possibly not as interesting. 19:38:31 cheater__: But I did find the texture synthesis sound samples: http://www.cns.nyu.edu/~jhm/texture_examples/ 19:38:47 splendid! thanks 19:39:30 -!- CakeProphet has quit (Ping timeout: 258 seconds). 19:41:18 -!- CakeProphet has joined. 19:41:18 -!- CakeProphet has quit (Changing host). 19:41:18 -!- CakeProphet has joined. 19:44:40 -!- pikhq has joined. 19:44:48 -!- pikhq_ has quit (Ping timeout: 252 seconds). 19:46:39 fizzie, just downloaded the demos, listening now.. 19:47:52 10:00:45 Lymia: Yes, but no browser complains on unexpected certificate changes. 19:48:07 although putty does 19:48:27 oerjan: SSH clients in general; I mention SSH host key checking later on. 19:49:19 I know one person who refuses to SSH into systems without first acquiring the host key through some trustworthy means; everyone else seems to just trust their luck on the first connection attempt. 19:50:18 ah 19:51:26 Depending on the configuration, OpenSSH may be obstinate enough to completely refuse to connect until you manually edit ~/.ssh/known_hosts and remove the offending key. 19:52:47 I know one person who refuses to SSH into systems without first acquiring the host key through some trustworthy means; everyone else seems to just trust their luck on the first connection attempt. 19:52:48 i know vorpal too 19:52:52 :P 19:52:56 Oh; then I know two. 19:53:40 `addquote And if they wanted to go for "true" security, they'd just do "Warning: your computer has not been turned into a plasma. This may cause some of your personal data to be stolen. Click here to turn your computer into a ball of incandescent gas, a gigantic nuclear furnace." 19:53:41 fizzie, what are the marginals? 19:53:43 ​429) And if they wanted to go for "true" security, they'd just do "Warning: your computer has not been turned into a plasma. This may cause some of your personal data to be stolen. Click here to turn your computer into a ball of incandescent gas, a gigantic nuclear furnace." 19:54:20 [asterisk]miasma of incandescent plasma 19:55:27 -!- CakeProphet has quit (Ping timeout: 276 seconds). 19:56:32 cheater__: The parameters they use are based on a cochlea-simulating filterbank; the "marginals" (IIRC) were simple time-average statistics (mean, variance, kurtosis) over the marginal distributions (in practice, histograms) of the amplitudes of each filterbank signal. 19:57:47 -!- Lymia has quit (Remote host closed the connection). 19:57:49 The "full set" included something like cross-correlations between low-pass envelopes of those signals. 19:58:35 And he had some statistics where they just used the spectrum (well, spectrogram) to synthesize things, purely to show that the spectral shape is not enough to capture the texture. 19:58:42 s/statistics/samples/ 19:58:50 The "spectrum" examples might be those. 19:59:38 I haven't listened to the ones on the page, the multimedia didn't work out right (it never does) and then I lost interest before actually downloading them. 20:00:07 what are statistics "over marginal distributions"? 20:00:56 -!- CakeProphet has joined. 20:00:56 -!- CakeProphet has quit (Changing host). 20:00:56 -!- CakeProphet has joined. 20:01:20 "amplitudes of each filterbank signal" - do you mean "amplitudes of each filterbank band"? 20:01:32 Right. Output signal. 20:02:31 -!- augur has joined. 20:02:52 so what are these "marginal distributions"? 20:03:43 I know one person who refuses to SSH into systems without first acquiring the host key through some trustworthy means; everyone else seems to just trust their luck on the first connection attempt. <-- I can't fathom why most people don't bother with it 20:04:16 fizzie: wow I was even joking. 20:05:00 cheater__: In this case I'd say it refers to just taking the amplitude histogram of the signal. I mean, that's a bit like a marginal distribution p(amplitude) if you consider the sample as the joint distribution p(time, amplitude). 20:05:11 fizzie, i just did this to download: 1. wget the index file 2. do this: cat texture_examples3.html | sed -e "s/.*\"\(.*wav\)\".*/\1/" | grep wav | while read i; do wget "http://www.cns.nyu.edu/~jhm/texture_examples/$i" & done 20:05:23 i bet this sed/grep could be optimized but i suck at sed. 20:05:26 elliott_, wait, you were joking that I would verify host key? Of course I do! 20:05:42 I wonder if wget has a "fetch embeds" recursion mode. 20:06:01 fizzie, what are marginal and joint distributions? 20:06:04 fizzie: This shit up: you can't make it. 20:06:42 elliott_, however, when trusted channel for it is hard to come by I might settle for some other sort of independent channel to verify it by. 20:07:13 oh it combines those. i guess also the binary thing is a bit of a tarpitisation. 20:07:16 Vorpal, such as /dev/urandom ? 20:07:17 elliott_, this means I will accept verifying it in a discussion over IRC, though I'd prefer a more secure mean. 20:07:22 P'' isn't necessarily binary 20:07:31 fizzie: this shit up. 20:07:34 fizzie: make it. 20:07:36 fizzie: you can't. 20:07:48 that's what i thought initially 20:07:50 fizzie, wget can fetch embedded images in html yes 20:08:12 then why would it combine the two commands 20:08:20 he meant media mentioned in tags. 20:08:21 and even if it were binary, why would it combine them :\ 20:08:24 Depending on the configuration, OpenSSH may be obstinate enough to completely refuse to connect until you manually edit ~/.ssh/known_hosts and remove the offending key. <-- that is the default config afaik 20:08:42 Vorpal, that's what i've found it to be 20:09:03 it's a bit annoying for some hosts. 20:09:19 Vorpal: Could be. I do know that HashKnownHosts officially defaults (or at least did) to 'no', but some distributions make it 'yes'. 20:09:35 What is the prize for implementing the interpreter and Pong game? —ehird 19:58, 24 May 2011 (UTC) 20:09:35 fizzie, ubuntu does 20:09:35 Currently, no finally defined, but I think it will be free (ad-free) webspace. --80.139.110.66 19:52, 26 May 2011 (UTC) 20:09:36 hahahahaahahahahaa 20:09:46 here I was thinking it'd be something of value 20:09:57 fizzie, actually I think that depends on which sort of match it is. If it is a non-perfect match it will allow you to go on. As in, the host is not found in the config, but the ip of that host is found in it, with a different key 20:10:03 you could store the complete knowledge of humanity on said webspace 20:10:17 but if the host name is found with a different key, it won't allow you to continue 20:10:19 -!- pikhq_ has joined. 20:10:22 iirc something like that 20:10:22 -!- pikhq has quit (Ping timeout: 260 seconds). 20:10:54 -!- Lymia has joined. 20:10:54 -!- Lymia has quit (Changing host). 20:10:54 -!- Lymia has joined. 20:11:09 fizzie, as for HashKnownHosts that is somewhat different from the question of allowing connection on host key mismatch 20:11:12 If you want to win $300, there's a logo contest on some new IEEE Signal Processing Society conference they advertised at the start of the day today. (Though I don't think they're going to accept anything too... esoteric, or #esoteric.) 20:11:21 HashKnownHosts simply hashes the host part in the config 20:11:36 so someone can't check ~/.ssh/known_hosts to find out all systems you have access to 20:11:37 What kind of sane person distributes an animation as a GIF? 20:11:45 Phantom_Hoover: Hussie? 20:11:47 A large, detailed one, at least. 20:11:56 I'm talking about stuff like http://jila.colorado.edu/~ajsh/insidebh/engulf_640x480.gif 20:12:01 Phantom_Hoover, uh, that is quite common on wikipedia, though for short ones. Like showing how a motor works 20:12:13 That would take far, far less time to load if you just stuck it on YouTube. 20:12:43 Phantom_Hoover, that image doesn't take long to load. Just about 3 seconds 20:12:56 I wish browsers would delay showing animated GIFs until they load completely. 20:12:56 "Warning: If you fall into a black hole, you will die. You will not go through a wormhole to another time and place." 20:13:00 Best warning. 20:13:07 oklopol: i assume P'' was meant to be minimalized. i also read that the author later defined and used combined commands that were identical to the brainfuck ones 20:13:28 elliott_, good idea 20:13:28 i thought P'' was meant to prove you can have loops 20:13:39 why show you can have structured computation and then fucking obfuscate it 20:13:39 :D 20:13:46 elliott_, or at least until there was a buffer loaded 20:14:06 s/can have/only need to have/ 20:14:27 http://www.ieee-espa.org/LogoContest.asp 20:15:01 "Here are the conference parameters:" 20:15:05 nice way to put it 20:15:19 They certainly emphasized the "for practicing engineers" part. 20:15:28 They're doing talks without papers and all. 20:15:32 fizzie, also Industry with capital I 20:15:37 And it's been co-located with CES. 20:15:37 crazy shit 20:15:42 fizzie, CES? 20:15:48 Consumer Electronics Show. 20:15:50 Or some-such. 20:15:52 heh 20:16:56 practical assholes 20:17:05 There was a Microsoft Research guy doing today's plenary talk; quite a lot of Kinect advertisement included. 20:17:15 I guess they're releasing their own non-commercial SDK for it soon. 20:17:34 He mentioned that the open-source driver lacks the funky microphone array DSP stuff they have. 20:17:42 fizzie, what is the point? Others already done all the job 20:17:45 When do people use it for perverted purposes? 20:17:59 Lymia, eh? 20:18:05 oklopol: well the obfuscation isn't on the loop commands 20:18:10 Bets are open. 20:18:24 Vorpal: Well, there was one point right there. 20:18:40 Lymia: Surely it has already happened. 20:18:43 Vorpal, how long until people start using the Kinect for less than work safe purposes. 20:18:44 Also as I understood it it lacked something else too than just the mic-array stuff. 20:18:45 fizzie, yeah I wrote it before you entered your line. Had quite a lag spike there. Weird. 20:19:06 Lymia: http://kinecthacks.net/kinect-sex/ + 20:19:11 s/\+/?/ 20:19:16 oh... 20:19:18 Lymia, are you invoking rule 34 on it? 20:19:36 "brainfuck: based on breakthrough research in structured programming" 20:19:39 No. 20:19:52 I'm asking how long until somebody does perverted things wiht it. 20:19:54 :D 20:19:54 with* 20:20:05 Microsoft's also pushing out some sort of "Kinect Avatar" that does the whole virtual-reality chat stuff with facial expression and gesture tracking. 20:20:15 Lymia: you're saying that as if they haven't already 20:20:21 olsner, right. 20:20:45 Lymia: http://kinecthacks.net/kinect-sex/ + 20:20:50 Worryingly quick response :P 20:20:53 (Also for the record: there was quite a lot of content in the talk that wasn't advertising their own stuff.) 20:21:19 elliott_: It was right there in the google quick-linkery thing when I googled for "kinecthacks" -- couldn't remember the tld. 20:21:19 "But well, NF is NOT a programming-language. It is an esoteric language which extends BF." 20:21:22 this guy is an idiot 20:21:38 elliott_, you should know that fizzie is good at finding stuff of all sorts. Everything from log graphing to finding weird things on google quickly 20:21:44 fizzie: suuure 20:21:45 Vorpal: Suuuuuuuuuuuuuuuuuuuuuuuuuuure. 20:21:48 oerjan: hifive 20:21:51 ... 20:21:51 erm 20:21:53 fizzie: 20:21:54 asterisk 20:21:56 :D 20:22:38 elliott_: well that fit with Vorpal too 20:23:00 "All in all, NF is just a notation of a finite state machine," 20:23:08 elliott_, err why did you change your hifive to target fizzie? 20:23:20 That does not work with my definition of FSM. 20:23:30 Phantom_Hoover, does it have a bounded tape? 20:23:34 Vorpal: i changed the suure 20:23:45 elliott_, oh okay 20:23:50 elliott_, it wasn't terribly clear 20:23:54 "brainfuck: based on breakthrough research in structured programming" 20:24:03 hey cool i wanna touch virtual boobs 20:24:07 You should write a fake advertisement for that. 20:24:27 also you rarely get to put your hands completely inside a boob so that's even better? 20:24:52 oklopol: The guy(tm) said it'll still be quite a while before haptics really take off, that it's the least understood area basically. So it'll be a while before the boob will touch you back. 20:25:34 elliott_, I bet if you asked fizzie about almost any googlable topic he would be able to find something about it very quickly. He might however not be interested in doing that I guess. 20:25:35 `addquote [...] So it'll be a while before the boob will touch you back. 20:25:37 ​430) [...] So it'll be a while before the boob will touch you back. 20:25:45 elliott_: hey i was just about to! 20:25:46 elliott_, come on. That was evil. 20:25:48 fizzie: please find me some information on whole-hand boob immersion 20:25:50 thx 20:26:08 you just beat me because you didn't include the whole quote 20:26:11 fizzie: maybe i'll just buy a sex doll 20:26:18 fizzie, what haptics? Does kinect have that?! 20:26:33 Vorpal: No, that's why it'll be a while. 20:26:58 Also the smell thing, it lacks that too. 20:27:02 fizzie, ah. You said "least understood area" I thought you meant "least understood area of the kinect hardware" (from the point of view of open source drivers) 20:27:28 Oh; no, it was in the sense of "least understood area of immersive(tm) communications(buzzword)". 20:28:11 (The repetition of "immersive" was almost as bad as some other people are with "ubiquitous".) 20:28:30 ah 20:28:39 well immersing in immersive ubiquity is ubiquitous 20:31:08 He did tell an amusing anecdote about attending some VRML conference (way back when that was relevant) at Stanford, where they had VRMLized the whole university; and when the guy demonstrating the thing navigated to the meeting room door, the speaker said he got a sudden feeling of dread that the (physical) door's going to open and a some sort of a huge eyeball is going to peek in, with all the people in the meeting room peering out from behind it. 20:32:26 fizzie, heh :P 20:33:33 virtual reality ...? 20:33:46 nobody ping Sg_o 20:34:34 $ ./esotope -f text -t minus -v 20:34:34 Found a path with 5 processors (weight=45): stream --(10)--> text --(10)--> brainfuck --(5)--> spoon --(10)--> minus --(10)--> buffer 20:34:37 ;) 20:34:42 I was about to ping him just to annoy you but then I looked at the context and realised the dire consequences that would have. 20:34:52 lifthrasiir: You frighten me. 20:34:58 elliott_: you should get a doctor to look at your right eye 20:34:59 fizzie, i would love to see an application of this resynthesis paper to data provided with BSS 20:35:09 pikhq_: ugh, really? 20:35:45 lifthrasiir: Cause that's that's clever and frightening, simultaneously. 20:35:54 http://jila.colorado.edu/~ajsh/insidebh/realistic.html 20:35:55 anyway now it is capable for parsing Text code (yes, a plain text) and translating it into Brainfuck, Spoon and finally Minus 20:36:20 My mind is rarely blown, but that has blown it. 20:36:30 -!- copumpkin has changed nick to Philippa_. 20:36:31 Minus does not directly implement a Brainfuck transformer since Spoon is a superset of it (with an additional exit command). 20:36:35 -!- Philippa_ has changed nick to copumpkin. 20:36:44 but it is just fine in esotope. yeah. 20:37:10 now gotta have some sleep.. 20:37:47 mind being blown: ......... 20:37:51 $ ./esotope -f text -t minus -v <-- eh, what does that -f and -t mean? 20:38:02 "from" and "to". 20:38:19 yes, from and to. 20:38:21 you mean this is an automatic tool to find chains of esolang interpreters? 20:38:24 or what? 20:38:24 -v for verbose, obviously. 20:38:40 Vorpal: it is a unified architecture to support such thing 20:38:54 lifthrasiir, heh. Does it try to find the shortest or the longest one? 20:39:10 so everything has to be reimplemented (d'oh!) but once done it is quite powerful 20:39:17 Vorpal: shortest. 20:39:38 lifthrasiir, what if you want to find the longest one without cycles? 20:39:46 since the longest path would be infinite (due to cycles) 20:39:47 (quite tricky I imagine) 20:39:52 -!- CakeProphet has quit (Ping timeout: 248 seconds). 20:39:56 lifthrasiir, longest one without cycles I meant 20:40:03 ah, that seems interesting, but for now no. 20:40:13 ah 20:41:02 -!- CakeProphet has joined. 20:41:06 i also plan to add some optimization flag, so every processor has two kinds of weights; the complexity weight and optimization weight. 20:41:55 lifthrasiir, heh 20:42:23 lifthrasiir, so you don't look for shortest path in number of languages you pass through, but best path? 20:42:27 so that the driver will find the path with the minimal sum of complexity AND the maximal sum of optimization (which can be bounded by the option) 20:42:43 basically, yes. i'm still figuring out the details. 20:43:25 lifthrasiir, presumably it would be useful to have an option that goes for fewest number of languages as well 20:43:26 and i have lots of languages to implement (i think 20 is enough for architectural testing though). 20:43:40 lifthrasiir, what language is esotope in? 20:43:59 Vorpal: not always, since some transformation is trivial (e.g. Ook and BF) 20:44:01 Ocaml. 20:44:07 lifthrasiir, and couldn't it call out to external interpreters possibly? 20:44:36 lifthrasiir, why ocaml? 20:44:40 not yet, but definitely in the plan. 20:45:08 well, the prime reason is that i wanted to learn Ocaml in a hard way... 20:45:13 (i.e. trial and error) 20:45:37 lifthrasiir, is there any other way? 20:45:39 and other reasons include i can mix the imperative and functional programming styles 20:45:54 (though many told me that Haskell will do ;) 20:45:57 Vorpal: it's compilers 20:45:58 not interpreters 20:46:05 Whiiich is not a major consideration of language choice. 20:46:13 pikhq_: oh shut up he can use whatever language he likes 20:46:16 (as long as it's haskell) 20:46:20 lol 20:46:26 As you can do imperative and functional programming styles in just about any language younger than 50. 20:46:34 elliott_, hey, you let me write gravity.lisp in CL. 20:46:36 (some more sanely than others) 20:46:48 lifthrasiir, ah compilers only? So you won't ever be able to do befunge98 then, unless you "compile" to a bundled interpreter 20:47:07 Vorpal: it includes both the interpreter and compiler (i.e. transformer). 20:47:18 -!- cheater__ has quit (Ping timeout: 252 seconds). 20:47:27 the interpreter is simply a transformer that gives nothing (or, in Haskell, IO ()). 20:47:56 hm 20:49:16 some time later, esotope will have (for example) brainfuck-optimized kind which represents an optimized Brainfuck code, which can be used to generate an efficient C code (as the current esotope-bfc does)ff 20:49:33 but for now i'm going to keep implementing important languages 20:49:41 and the transformers for them 20:50:58 hm compiling from a language which has a self-interpreter gives an interpreter automatically 20:52:21 or if B has an interpreter for A, and you compile B to C, then you get an interpreter for A in C 20:52:53 yes, that's why i didn't implemented an interpreter for Ook!. 20:52:57 implement* 20:52:59 oerjan, what about a self-compiler, err wait that would be damn useless. never mind 20:53:14 Vorpal: :D 20:53:57 technically for interpreters there are up to two languages involved, while for compilers there are three 20:53:57 oerjan, the only case where a self-compiler could be useful would be if it was optimising. Like say one for bf that turned ++- into + 20:54:16 heh 20:55:26 oerjan, hm *tries to imagine some way to run programs that involve more than three languages* 20:57:15 Vorpal: well when bootstrapping a language compiler, you might write the compiler for B in A, compiling into C (the last one literally) 20:57:41 oerjan, that is three so far 20:57:43 no D 20:57:51 oh you said _more_ than three 20:57:56 yes 20:58:03 I had no success so far 20:58:31 oerjan, three is easy. Just imagine a bf->C compiler written in something else than bf or C 20:58:31 well obviously there are cases where one stage contains more than one language, but that may not count... 20:58:40 oerjan, hm... 20:58:58 elliott_, you mean like the language that is mostly not Murphy? 20:58:59 like web stuff, you have both client and server languages 20:59:06 ...what is that 20:59:13 I forgot the first part of it 20:59:17 -!- comex_ has joined. 20:59:19 something that cpressy made 20:59:23 oh 20:59:30 Ozzlybob or something iirc? 20:59:34 yeah 20:59:52 there is another pair of esolangs, one of which is Portia 20:59:58 oerjan, well it said that most parts of the language was the thing on O, the rest of it was however Murphy :P 21:00:11 heh 21:00:46 the fact that it's not called Oolzybub is called Murphy 21:00:46 portia is supposed to go with 2iota right? 21:00:48 to be precise 21:00:52 or beta-juliet 21:00:58 http://esolangs.org/wiki/Beta-Juliet_and_Portia 21:00:59 once of those car languages 21:01:21 http://esolangs.org/wiki/Oozlybub_and_Murphy 21:01:28 oh hm 21:01:34 oerjan, it says preprocessor 21:01:42 but isn't preprocessor another compiling step 21:02:10 Vorpal: yeah that's why i'm asking if splitting stages counts 21:02:38 oerjan, like CPP takes C source from C to C-preprocessed which a C compiler (as it is confusingly named, should be called a C-preprocessed-compiler!) then compiles to asm, which the assembler then compiles to an object file. And lets not discuss the linker. 21:02:39 although web stuff with client/server split feels more relevant 21:02:46 oerjan, yes kind of 21:03:19 hm what about that supercompilation stuff i've sometimes seen mentioned here 21:03:35 although that may be mainly one language applying to itself 21:03:58 oerjan, isn't supercompilation simply a way to optimize by brute forcing ways to compile an expression to 21:04:14 like testing all ways you could possibly zero out a register to find the shortest one or whatever 21:04:20 no. 21:04:36 elliott_, I heard supercompilation used to describe that however. Hm 21:04:42 i don't recall anything about "brute forcing", i'm unclear about it but i thought it was related to partial evaluation 21:04:57 so what is super-compilation then 21:05:24 specialisation. 21:05:25 well i think elliott_ may be who mentioned it previously 21:05:47 elliott_, hm that is still three languages at most right? 21:05:53 yes. 21:05:59 hm 21:06:16 in compilation, you have three languages involved in a single stage 21:06:43 which cannot be logically split up without some part having the same property, i think 21:06:55 or wait 21:08:43 if A compiles B to C, then you can split it by compiling the compiler from A to C first, then running the resulting C program to compile from B to C 21:09:54 so it's not quite clear that there are intrinsically units involving more that 2 languages 21:10:09 oerjan, you could do that, but then you need a new compiler that compiles from A to C. Involving an extra software 21:10:42 well yeah 21:10:46 and if you mention hand compilation I would argue that counts as a "software" in this context 21:12:15 it's a bit neat to think that the entire compiler/interpreter infrastructure we have today all builds necessarily on an ancient, forgotten layer of hand compilation 21:12:35 well maybe someone still does it occasionally for fun 21:12:50 or for efficiency 21:14:30 oerjan, I would say it is a form of hand compilation when turning pseudo-code into 21:14:38 the hand bootstrapping doesn't need to be done particularly efficiently either. like that inefficient "unregistered C" target for ghc, which is just to get an initial port 21:15:07 oerjan, why do you need unregistered C for ghc at all. Why not simply cross compile full blown ghc 21:15:08 Vorpal: hm yeah. and on the other side even assembly is not machine code 21:15:28 Vorpal: oh. current ghc does not support cross compilation. 21:15:38 oerjan, people started out writing machine code before they invented asm 21:15:41 oerjan, yeah why 21:16:13 the new upcoming code generator is supposed to though (i saw something to the effect that it always essentially cross compiles) 21:16:51 oerjan, hm, you could say that when turning your idea of what you want to do into code in any programming language, you are hand-compiling your thoughts to said language. 21:17:13 Vorpal: although it might still be nice to have easy port version to be able to get things started before you have a full implementation of all the target architecture's quirks 21:17:20 oerjan, what new upcomming code generator? 21:17:22 *easily ported 21:17:42 Vorpal: well i've seen it mentioned, they're completely rewriting that part 21:17:49 (in ghc) 21:18:14 i think the old one was getting so full of cruft that they had trouble adding new things to it 21:19:51 http://hackage.haskell.org/trac/ghc/wiki/Commentary/Compiler/NewCodeGen 21:20:13 heh, neat. I had a pencil lying on a shelf. (Yes my desk and the nearby area is quite messy.) I bumped the desk with my foot by mistake. This caused the pen to fall off. It landed right way up in the pencil jar (is that the right English word?) right below. 21:20:29 -!- CakeProphet has quit (Ping timeout: 260 seconds). 21:20:50 I think that "new" code gen has been the "current" code gen for a while now 21:20:53 oerjan, hm the unregistered C one uses the evil mangler? Or is that some other one? 21:21:10 -!- CakeProphet has joined. 21:21:16 olsner: no, see the current status report section 21:22:04 oerjan: meh! 21:22:07 Vorpal: the evil mangler is used by the _registered_ C backend 21:22:13 oerjan, oh 21:22:25 oerjan, why the name "(un)registered"? 21:22:33 the unregistered avoids those tricks to be completely portable, at the cost of efficiency 21:22:53 it does not attempt to control which machine registers are used for things 21:23:00 oerjan, how does the unregistered one do gc? 21:23:24 simpler, i assume? 21:23:26 oerjan, does it allocate a large array to use as backend for gced stuff? 21:23:32 I mean that is really the only portable way 21:23:35 i don't know such details :) 21:24:21 so i don't know _how_ portable it is 21:24:39 just that it doesn't depend on that many deep gcc specialties 21:25:55 ah 21:26:54 -!- CakeProphet has quit (Ping timeout: 276 seconds). 21:26:59 and architecture dependent details 21:28:35 -fvia-c is deprecated 21:28:36 by the way 21:28:41 i.e. registered C builds 21:28:43 so the mangler is deprecated too 21:28:59 BlueBottle works just fine in VMware Player :/ 21:29:10 elliott_, but the mangler is fun. Since it is evil. 21:31:50 Evil mangler? 21:32:00 yes 21:32:41 "In this respect, the A2 System GUI surpasses many existing windowing implementations. Windows can moved, rescaled etc. while the programs "behind" the windows kep running." 21:32:44 This sounds old 21:32:59 oh seems it is completely gone 21:33:06 Lymia: it's a perl script which ghc -fvia-c runs on the assembly produced by gcc, mangling it 21:33:13 Lymia, check this historical version of the page for more info http://hackage.haskell.org/trac/ghc/wiki/Commentary/EvilMangler?version=2 21:33:14 Vorpal: there's also the Satanic Splitter 21:33:17 Which does? 21:33:57 olsner, oh? link? 21:34:56 don't remember *exactly* what it does, but something like splitting assembly sources into several files to allow the linker to make smaller programs 21:35:07 Evil mangler, eh? 21:35:18 It's not that evil! 21:35:23 olsner: i think it also ensures proper tail calls 21:36:32 can't be bothered doing more than googling, but there is this: http://www.haskell.org/pipermail/glasgow-haskell-users/2005-February/008006.html 21:37:46 olsner: i don't think the mangler is what splits object files, at least that isn't its main purpose, see that link by Vorpal 21:38:01 I'm talking about the satanic splitter, not the evil mangler 21:38:29 olsner, confirmed: http://hackage.haskell.org/trac/ghc/wiki/Building/Porting#Thesplitter 21:38:41 olsner: oh right 21:38:50 "GHC no longer has an evil mangler." 21:39:00 oerjan, I said that 21:39:18 the splitter is still there: http://hackage.haskell.org/trac/ghc/browser/driver/split/ghc-split.lprl 21:39:39 Vorpal: i'm having trouble getting all the irc conversation _and_ checking out your links at the same time :P 21:39:44 oerjan: that's because the registered C backend is deprecated :P 21:40:09 oerjan, you are too slow 21:40:15 yes. 21:40:29 oerjan, overclock yourself (first ensure proper cooling) 21:40:53 unfortunately overclocking has severe side effects on me. 21:41:05 maybe the Oerjan Enhanced Speedstep Technology has kicked in already 21:42:12 elliott_: yeah but i didn't know it was deprecated to the point they had stopped keeping it working efficiently... 21:42:31 oerjan, the evil mangler file is gone even 21:42:37 oerjan: They started doing that because the LLVM backend performs better in all cases, IIRC. 21:42:39 figures 21:42:43 pikhq_: ah. 21:42:48 makes sense 21:43:35 oerjan: i think NFA = FNFA with any amount of dimensions 21:43:54 good, good 21:44:17 i've spent hundreds of hours trying to solve that, so yes :P 21:44:54 well a proper counterexample would of course have been good too 21:45:06 reading http://hackage.haskell.org/trac/ghc/wiki/Commentary/Compiler/Backends/LLVM/WIP : what *is* this elsusive STG register in haskell? 21:45:09 anyone knows? 21:45:12 * Vorpal looks at elliott_ 21:45:24 thing 21:45:28 yes, although it didn't take me long to convince myself 100% that they are equal in n-D 21:45:30 stg is for spineless tagless g-machine 21:45:39 oerjan, err, what 21:45:45 oerjan, joke right? 21:46:03 A language like Haskell cannot exist without nasty dirty hack under it's skin. 21:46:04 of course, this means i probably cannot sleep tonight. 21:46:14 oerjan, or did what you say actually mean something? 21:46:17 Vorpal: no. well it may contain a pun, but it's the intermediate representation stage between core and c-- 21:46:38 oerjan, ah, so what is the STG register for, and why is it pinned to a machine register 21:46:49 Lymia: why not 21:46:54 coq is dirty-hack free :P 21:47:02 oh 21:47:19 elliott_, wait a second. coq is written in ocaml. ocaml is not dirty-hack free 21:47:26 i mean if my solution is correct, this is bigger than P != NP for the picture-walking automata people! (that is, me and maybe that one japanese guy although he's probably dead already.) 21:47:36 Vorpal: i think there is more than one, it's the registers that result from something in the stg representation i assume 21:47:48 oerjan, hm 21:47:59 and the stg running model presumably requires them to be at a known place 21:48:14 my memory on stg is too vague 21:48:35 that representation is old i think, perhaps from the 80s 21:48:43 although ghc has modified it 21:49:24 (they added tags back to pointers at one stage for efficiency, iirc) 21:49:43 so it may not actually fit its name any longer :D 21:50:10 (argh, i knew i'd find a problem with the solution the second i announce it) 21:50:41 oklopol: ah yes. a known problem. 21:51:37 oklopol: is it big 21:51:40 actually when that happened for the 2-D problem, i managed to solve the problem the same day anyway. so i still have hope. 21:52:18 it's not at all big. the 2d question was left open in at least 2 articles and a book, but the n-dimensional case has only been asked by me. 21:52:21 oklopol: happened to me with that wrong >= 2d proof for topological measures i mentioned yesterday. 21:52:31 you solved it? 21:52:43 or you mean you announced it before realizing 21:52:49 no, it fell through when i was explaining it on the blackboard 21:52:55 oh dear 21:53:17 ouch 21:53:22 strictly speaking i had already announced it, sort of (i SMS'ed "Eureka" to my collaborator) 21:54:31 oerjan: nerd :D 21:54:38 i suppose "dork" is more appropriate 21:54:47 elliott_: you wouldn't understand 21:55:21 oklopol: I meant the "Eureka" 21:55:34 err right 21:55:34 :P 21:55:38 Vorpal: "spineless" iirc is for how that representation did away with having an internal representation of the reduction graph for functional languages, using stack arguments instead. 21:55:48 ah 21:56:16 argh if the solution doesn't work, then i now have three proofs for the 2d case but none for the general case 21:57:24 * Vorpal tries to think the most useless multiclassing in nwn. 21:57:26 That's one more than two. 21:57:49 "tagless" iirc is for how it used return destination to branch directly to the code that wanted the result, instead of allocating constructors that might never be used since the destination might just pick them apart 21:57:51 possibly multiclassing sorcerer and wizard? That one must come close at least 21:58:42 oerjan, ah nice 21:59:00 oerjan, is this like optimising a tuple into an unboxed tuple? 21:59:26 yes, except it works for data types with more than one constructor, such as Just x vs. Nothing 21:59:44 *it also works 21:59:51 oerjan, oh, nifty. But this would only work within a single module surely? 21:59:55 not across modules 22:00:34 no, the return address would be in a known spot on the stack or in a table pointed to from the stack, i think 22:01:05 and making that work across modules _might_ be a reason for pinning those registers 22:01:12 hm 22:01:26 why is everything slow suddenly 22:01:43 weird, I'm swap trashing for no obvious reason 22:05:01 oerjan: nerd :D <-- ARE YOU CALLING ARCHIMEDES A NERD? 22:05:06 $ git clone http://darcs.haskell.org/ghc.git/ <-- wait what, 1) ghc went git!? 2) why call it darcs.haskell.org if it isn't darcs? 22:05:13 oerjan: He was Greek :P 22:05:23 Vorpal: (one) Alas, yes. (I didn't realise they had finished transitioning though.) 22:05:28 Vorpal: (two) It's a separate server. 22:05:47 No point in fiddling with DNS, since they still host a lot of darcs repos too. 22:05:53 hm okay 22:06:06 elliott_, sad, darcs was better than git 22:06:16 oerjan: He was Greek :P 22:06:16 With foresight, it would have probably been named repo.haskell.org. :p 22:06:21 Vorpal: In fairness, GHC is huge, and darcs was quite slow for them. 22:06:22 elliott_, indeed 22:06:26 hm 22:06:28 elliott_, really? 22:06:32 elliott_: i guess i didn't do it properly, i should have ran naked in the streets as well 22:06:32 Yeah, the Greeks were the ultimate nerds. 22:06:34 Vorpal: Really what? 22:06:37 *run 22:06:40 elliott_, really it being slow 22:06:44 oerjan: Yes, this was definitely something you should have done. 22:06:45 They invented steam engines and just saw them as a cool toy. 22:06:47 Classic nerd. 22:06:48 elliott_, darcs never came across as slow to me 22:06:51 Vorpal: darcs has known efficiency issues for large projects /shrug 22:06:56 darcs one was absolutely glacial. 22:06:58 darcs two is faster, buuut. 22:07:12 elliott_, but what 22:07:43 Well, "fool me once, shame on you; fool me twice, can't get fooled again". 22:08:07 ah 22:08:27 elliott_, not all docs are updated to saying they use git however. Some still suggest darcs. 22:08:31 Heh, "I don't know why you're talking about Sweden. They're the neutral one. They don't have an army." 22:08:39 (Found when Googling to make sure I had the Bush quote right; http://en.wikiquote.org/wiki/George_W._Bush.) 22:09:08 elliott_, our army is just an excuse for not having one really 22:09:27 Don't you have a fairly good army, but just not use it for anything? :p 22:09:50 elliott_, we used to have that. But budget cuts has been going on for well over 15 years now, so not much left. 22:10:08 -!- cheater__ has joined. 22:10:15 elliott_: mind you the ghc guys refused to change to Darcs 2 format (or something like that) because it didn't work with something they used to do, thus not using some of darcs's later improvements, i think. 22:10:22 Vorpal: Well, hey, at least you're cutting from the military. That's something America hasn't quite figured out how to do yet. 22:10:32 oerjan: ah 22:10:33 elliott_, :D 22:10:47 oerjan, did it work with git however? 22:11:03 Vorpal: presumably. 22:11:27 oerjan, it would be somewhat funny if it didn't 22:11:51 which might mean that ghc chose git over darcs because they insisted on using git workflow... 22:12:02 (speculating here) 22:12:07 and always had 22:12:39 -!- CakeProphet has joined. 22:12:39 -!- CakeProphet has quit (Changing host). 22:12:39 -!- CakeProphet has joined. 22:13:16 oerjan, rebase *shudder* 22:13:57 Don't you have a fairly good army, but just not use it for anything? :p <-- i vaguely think that sweden, unlike norway, having a pretty good army at the start of ww2 was how that managed to _stay_ neutral then 22:14:19 (Found when Googling to make sure I had the Bush quote right; http://en.wikiquote.org/wiki/George_W._Bush.) <-- I glanced at that page and the stupidity ratio is utterly outstanding 22:14:39 *how they 22:14:43 -!- pikhq_ has quit (Ping timeout: 240 seconds). 22:14:47 -!- pikhq has joined. 22:15:21 Vorpal: Surely you've heard all the Bushisms by now. 22:15:29 oerjan, well... no we didn't have a good army at the start of ww2. However the politicians claimed we did and then did a mad rush to make it true. Meanwhile allowing the germans to transport troups through Sweden for the first few years of the war 22:15:38 Vorpal: aha 22:16:01 elliott_, I know he is stupid yes. But I haven't heard all those quotes no. Quite a few yes. 22:16:36 i just recently learned from wikipedia that britain invaded iceland shortly after germany invaded norway and denmark. the icelanders protested, but not overly much 22:16:45 oerjan, oh and we sold them iron ore too. Since they invaded you guys and thus controlled Narvik. 22:16:59 Vorpal: "En svensk tiger", eh? ;D 22:17:01 oerjan, really? that is news to me 22:17:05 oerjan, oh that yeah 22:17:21 oerjan, was that invented around then or after? 22:17:51 they then passed control of iceland over to us troops. this was _before_ the us formally joined the war. 22:18:14 Vorpal: um you should know, you're the swedish one. 22:18:31 oerjan, hey you are Norwegian and you told me you haven't seen Peer Gynt! 22:18:32 -!- sebbu has joined. 22:18:32 -!- sebbu has quit (Changing host). 22:18:32 -!- sebbu has joined. 22:18:40 oerjan, så kasta inte sten i glashus 22:18:49 (I have no idea what that idiom would be in English) 22:19:09 (literal translation would be "don't throw stones in houses made of glass") 22:19:26 The same, basically. 22:19:28 i think "those who live in glass houses shouldn't throw stones" 22:19:34 ah 22:19:53 although "the pot calling the kettle black" also fits the theme 22:20:18 yes indeed 22:20:28 oerjan, aren't they basically the same? 22:20:38 I mean, what is the difference in meaning between the two idioms 22:21:24 oerjan, rebase *shudder* <-- i am not entirely sure, but it may be close to the thing ghc insisted on using 22:21:44 -!- sebbu2 has quit (Ping timeout: 260 seconds). 22:21:47 oerjan, rebase is Modifying History. Which is Wrong. 22:22:02 (which means I hate the git workflow) 22:22:24 -!- BeholdMyGlory has quit (Remote host closed the connection). 22:22:50 > let divs 1 _ = False; divs 0 _ = True; divs n d = divs (n-d) d in divs 9 2 22:22:50 False 22:22:57 > let divs 1 _ = False; divs 0 _ = True; divs n d = divs (n-d) d in divs 8 2 22:22:57 True 22:22:59 yay 22:23:24 elliott_, shouldn't you use where instead of let, iirc you told me so some time ago? ;P 22:23:35 um not in this context, no 22:23:50 elliott_, okay, for the purposes of learning idiomatic style, why is that? 22:24:00 Vorpal: cannot use where on an expression alone 22:24:02 well this is ghci, it's hardly idiomatic 22:24:04 but what oerjan said 22:24:05 > x where x = 9 22:24:05 : parse error on input `where' 22:24:07 oerjan, oh right 22:24:12 besides, i wrote the function first ;) 22:24:14 elliott_: um lambdabot isn't ghci 22:24:20 oerjan: close enough :) 22:24:24 same "workflow" 22:24:28 just vaguely similar to a few subcommands of ghci 22:24:40 > :t thiswontwork 22:24:40 : parse error on input `:' 22:24:43 indeed 22:24:47 not ghci 22:24:52 :t thiswontwork 22:24:53 Not in scope: `thiswontwork' 22:25:04 oerjan, indeed. But > is not ghci 22:25:08 that was my main point 22:25:11 yeah 22:25:20 -!- sebbu2 has joined. 22:25:21 -!- sebbu2 has quit (Changing host). 22:25:21 -!- sebbu2 has joined. 22:25:39 Variable occurs more often in a constraint than in the instance head 22:25:39 in the constraint: Divisible n' d r 22:25:39 (Use -XUndecidableInstances to permit this) 22:25:40 gah 22:25:54 you need to remember when to add > or @, and you still only get non-IO expressions, let, :t and :k 22:25:56 elliott_, err, what does that mean 22:26:21 oerjan, what is @ for in lambdabot? 22:26:30 > let sub n 0 = n; sub n m = sub (n-1) (m-1) in sub 9 2 22:26:31 7 22:26:53 class Sub a b r | a b -> r, a r -> b 22:26:53 instance Sub n Z n 22:26:53 instance (Sub n m r) => Sub (S n) (S m) r 22:26:56 the last instance fails :( 22:26:58 elliott_: um your definition above fails on divs 1 1 22:27:05 oerjan: ugh does it 22:27:10 thankfully i never use one 22:27:12 but what mistake did i make 22:27:22 or wait 22:27:24 -!- sebbu has quit (Ping timeout: 258 seconds). 22:27:31 > let divs 1 _ = False; divs 0 _ = True; divs n d = divs (n-d) d in divs 1 1 22:27:32 False 22:27:34 yes 22:27:36 it does 22:27:41 > let divs 1 _ = False; divs 0 _ = True; divs n d = divs (n-d) d in divs 2 1 22:27:42 False 22:27:45 > let divs 1 _ = False; divs 0 _ = True; divs n d = divs (n-d) d in divs 1 2 22:27:45 False 22:27:47 um 22:28:09 shut up, my main problem is this undecidable instance :) 22:28:11 whichever order that is in, that is wrong (while 2 doesn't divide 1, 1 definitively divides 2) 22:28:27 > let divs 1 _ = False; divs 0 _ = True; divs n d = divs (n-d) d in divs 8 8 22:28:27 True 22:28:31 > let divs 1 _ = False; divs 0 _ = True; divs n d = divs (n-d) d in divs 2 2 22:28:31 True 22:28:33 hm 22:28:36 elliott_: more seriously, it fails on divs 8 3 22:28:41 > let divs 1 _ = False; divs 0 _ = True; divs n d = divs (n-d) d in divs 2 3 22:28:44 mueval-core: Time limit exceeded 22:28:48 whaaaat 22:28:48 oerjan: ok ok ok but i need to fix this first 22:28:56 > let divs 1 _ = False; divs 0 _ = True; divs n d = divs (n-d) d in divs 8 3 22:29:00 mueval-core: Time limit exceeded 22:29:00 yes 22:29:02 yes it does 22:29:14 > wat language is dis 22:29:14 Not in scope: `wat'Not in scope: `language'Not in scope: `is'Not in scope: ... 22:29:16 Oh. 22:29:23 Vorpal: same as your 2 3, actually. leaky recursion base. 22:29:27 oerjan, indeed 22:29:43 > let divs 1 _ = False; divs 0 _ = True; divs n d = divs (n-d) d in divs 8 5 22:29:45 Terminated 22:29:51 okay that one is new 22:29:53 oh wai 22:29:53 t 22:29:56 is it different or same? 22:29:57 the add definition works 22:29:59 in this case 22:30:56 Lymia, haskell 22:31:02 gah wtf... 22:31:06 do you need undecidable instances to do this 22:33:43 elliott_: i don't know why that claims it's got a greater context, maybe the functional dependencies are involved somehow? 22:33:58 probably 22:34:04 oh well, undecidabled up the wazoo 22:34:16 now to fix that divisible algo :D 22:34:20 as in, aren't they rewritten to type functions these days? 22:34:27 what does undecidable instances do? 22:34:37 or something like that 22:35:55 Vorpal: hm i recall seeing Terminated before, although i don't recall if there's a difference with that mueval-core thing 22:36:04 maybe there are two different competing timeouts 22:37:12 ah 22:37:28 i am wondering, is having both a b -> r and a r -> b causing unbounded growth in itself 22:37:39 I wonder if it is repeatable: 22:37:42 > let divs 1 _ = False; divs 0 _ = True; divs n d = divs (n-d) d in divs 8 3 22:37:45 Terminated 22:37:47 > let divs 1 _ = False; divs 0 _ = True; divs n d = divs (n-d) d in divs 8 3 22:37:51 mueval-core: Time limit exceeded 22:37:54 hm nope 22:38:04 oerjan, looks like two competing timeouts indeed 22:38:30 Vorpal: btw the 2 1 thing is equivalent to my first comment on 1 1, he doesn't handle d = 1 22:39:21 -!- zzo38 has joined. 22:39:28 in fact i think the mueval-core thing is newer, they may have restructured lambdabot at some point 22:40:05 -!- ntides has joined. 22:40:22 I got FizzBuzz C code down to 87 bytes now I should try for mathematical shortcuts, I think 22:40:25 oerjan, hm don't you usually write x divides y as x | y, rather than y | x 22:40:28 -!- ntides has left. 22:40:32 elliott_: does it work if you remove a r -> b ? 22:40:37 oerjan, in which case elliott_ has the parameters backwards 22:40:49 -!- ais523 has quit (Remote host closed the connection). 22:40:49 oerjan, which means 8 `divs` 3 will be all backwards 22:41:00 well yeah 22:41:01 -!- elliott_ has quit (Ping timeout: 255 seconds). 22:41:06 well he left 22:41:12 although it's the same order as div and mod 22:41:20 (which aren't predicates) 22:41:27 oerjan, yes but | is usually written the other way around 22:42:00 whatever 22:43:16 I saw on some periodic table of Perl6 that the % operator is "iffy" which means it can be ! (logical not), but when I tried using !% operator it didn't work but it did say to use %% instead, so I did, and it worked 22:43:24 Vorpal: "The famous poster for the propaganda campaign was created by Bertil Almqvist in 1941" 22:43:29 "periodic table of Perl6"? 22:43:30 (en svensk tiger) 22:43:35 -!- CakeProphet has quit (Ping timeout: 240 seconds). 22:43:36 oerjan, indeed 22:43:44 oerjan, I googled it several minutes ago 22:43:49 Vorpal: Yes I found the Periodic Table of Perl6 Operators 22:44:04 huh 22:44:23 The Table is quite old; maybe it hasn't been kept current? 22:44:24 Everything is categorized, similar to, but not quite like, the periodic table of elements. 22:44:50 fizzie: Maybe that is why it did that, then 22:45:30 Although at first when I read that on the table, I immediately thought that making % to be "iffy" is a good idea, and that it had other good ideas too from what I could see on that table 22:45:39 -!- CakeProphet has joined. 22:45:39 -!- CakeProphet has quit (Changing host). 22:45:39 -!- CakeProphet has joined. 23:00:39 -!- h[a]gb4rd has joined. 23:18:10 -!- CakeProphet has quit (Ping timeout: 264 seconds). 23:23:03 Apparently non-Americans think root beer tastes like medicine. Personally, I wonder how they get good-tasting medicine. 23:23:46 * Phantom_Hoover → sleep 23:23:53 -!- Phantom_Hoover has quit (Quit: Leaving). 23:24:54 pikhq, EZ Tab Tylenol or whatever tastes great! 23:25:09 I wish I could have medicine flavoring without the medicine 23:25:31 You are probably a freak. 23:25:53 pikhq, I don't think I mean raw medicine 23:26:01 They add flavoring to medications I think 23:26:06 I think that's what I like 23:26:11 But yes, I am a freak, so 23:26:12 Yes, you are probably a freak. 23:26:30 Well, some medicine, anyway 23:26:38 Ugh, the taste of Wellbutrin was horrible 23:26:52 Although I tended to chew it despite it not supposed to be chewed 23:27:26 man 23:27:30 Same with Ritalin. Oh wait, I didn't chew Wellbutrin, the taste was _too_ horrible 23:27:31 plural logic has never made sense until now 23:27:42 and now i know its because noone knows how to explain it properly except boolos himself 23:28:34 Dosn't this apply to a lot of things? 23:29:01 Lymia: yes 23:29:13 Even basic mathematics.... 23:29:33 Well, I guess that starts making sense once you get into Algebra, etc. 23:30:36 -!- variable has left ("I found 1 in /dev/zero"). 23:47:46 I shortened FizzBuzz C code to 63 23:49:16 Sorry, I didn't do that. I shorted "google" to 63 23:52:03 "FizzBuzz" is 86 23:55:35 Grats. 23:55:42 Now do it in Perl which you can actually golf worth shit. 23:59:43 -!- CakeProphet has joined. 23:59:43 -!- CakeProphet has quit (Changing host). 23:59:43 -!- CakeProphet has joined.