00:00:17 * FireFly lost :( 00:00:32 a 00:01:11 22:29 MizardX: http://sovietrussia.org/f/src/tetoris.swf 00:01:14 my god 00:01:16 it's full of empty space 00:02:38 A lab-mate dared any of us to fill-and-explosify a non-zero number of lines in that Tetris variant during work-time; I spent approximately seven minutes filling about half of a row, then misplaced a S leaving an unfillable hole, and finally closed the browser tab in disgust. 00:03:24 (Might be one of the reasons I'm still awake trying to write this must-be-ready-by-tomorrow text.) 00:07:59 fizzie: if by better you mean "easier", i cannot imagine one 00:09:00 I mean "as easy but not limited to collaboration on mathy subjects". 00:09:29 no idea 00:10:38 There ought to be some sort of a searchable GiganticSciencePublicationDatabaseGraphThing. 00:12:19 An erdos-bacon number of 0. 00:14:04 mm, bacon 00:14:10 Mm, Erdos. 00:14:10 Er. 00:14:15 (i can safely say this since AnMaster has left) 00:14:17 Actually I am eating bacon right now. 00:14:26 oerjan: he still responds to pings 00:14:30 i think he set his client to wake him up 00:14:36 Actually I am eating peanut butter right now. 00:14:42 That is not bacon oerjan. 00:14:47 indeed 00:15:06 i do not believe i have bacon, unless some has sneaked into the liver pate 00:16:01 since bacon pate is a separate product, i doubt it 00:20:31 * oerjan notes that US "jelly" seems to include "jam" as a subset 00:20:37 -!- FireFly has quit ("Later"). 00:20:37 really? 00:20:38 weird 00:21:00 so in fact i have all the ingredients for that famous peanut butter and jelly sandwich 00:21:15 :D 00:21:37 i merely failed at merging the slices into a sandwich 00:21:59 and now it is too late for this meal, since i have already eaten the peanut butter slice 00:22:39 but next time - FOR SCIENCE 00:26:44 hi Blipi, you new? 00:27:04 Yeah 00:27:50 welcome 00:27:56 what brings you here? 00:28:15 it's just a blip in existence 00:28:43 My friend ;D 00:28:50 wait wait ;D? 00:28:52 Must be FireFly 00:28:58 or MigoMipo, or BeholdMyGlory 00:29:00 One of the swedes 00:29:09 There is a large number of nicknames on the list; lots of alternatives. 00:29:11 whom just left 00:29:13 telia.com redirects to .se 00:29:14 Yep 00:29:16 yep yep, FireFly 00:29:18 FireFly 00:29:19 D:D 00:29:26 My investigation skills are unparalleled. 00:29:33 ehird: Do you have some psychic powers or something? 00:29:56 No, just deduction. 00:30:13 The initcaps of the nick; then ;D -- I've only seen the swedes use ;D, but not AnMaster, he's far too seriousf or that 00:30:20 So, FireFly, BeholdMyGlory or MigoMipo 00:30:28 I /whois'd, checked the ISP, redirects to .se. 00:30:32 And thus. 00:30:58 ehird: wait, are you saying Blipi _is_ FireFly? 00:31:04 No. 00:31:10 He said he was referred by "My friend ;D". 00:31:16 ok. so i don't have to swat him then. 00:31:21 So I tried to figure out who that was. 00:31:22 oerjan: heh 00:31:28 ; 00:31:32 ; 00:31:38 . 00:31:53 * Blipi lost the Game 00:32:02 I can't lose the game. Thank gawd for Not the Game. 00:32:14 i think telia may have some norwegian presence too, at least they used to 00:32:22 ehird: No, just deduction. <<< it was induction 00:32:45 telia and telenor tried to merge once 00:32:47 Blipi: 00:32:48 Here are the new rules to Not the Game: 00:32:49 { 00:32:51 1. You can start playing or stop playing Not the Game by announcing you do. 00:32:53 2. If you are playing Not the Game, you are not playing The Game. 00:32:55 3. Not the Game takes precedence over every other game, including games (apart 00:32:57 from Not the Game) that specify other rules of precedence. 00:32:59 } 00:33:01 s/new //, that was from 2007 00:33:03 er 00:33:05 2008 00:33:07 wow, just 2008-10 00:33:09 later than I thought 00:33:51 * Blipi is playing Not the Game 00:33:52 ehird: you cannot remove the new now, the old new rules take precedence ;D 00:34:00 (: 00:34:03 oerjan: heh 00:34:49 well, i guess you could stop playing first 00:35:20 or wait, that "new" is not technically part of the rules 00:35:30 or is it 00:35:56 Well, I'm off 00:35:57 Later 00:36:04 -!- Blipi has quit ("- nbs-irc 2.39 - www.nbs-irc.net -"). 00:36:05 bye 00:45:08 -!- tromp_ has joined. 00:45:30 question: what's the smallest known brainfuck interoreter? 00:45:41 tromp_: the one in the BrainfuckInterpreter language 00:45:45 it goes like this: x 00:45:56 but, seriously, well, that's not a simple question 00:46:22 ok, smallest non-cheating interpreter:) 00:46:31 define cheating 00:46:33 (We're pedants...) 00:46:55 tromp_: well, there are interps that fit in 4 80-character lines 00:46:57 (sig sized) 00:47:06 There's that 240-byte compiler, which is remarkable in that it is a compiler. And the redcode brainfuck interpreter, wasn't that pretty damn short? 00:47:07 tromp_: are you _that_ john tromp? 00:47:09 cheating is necessaarily ill defined:( but often you know it when you see it. like with the BrainfuckInterpreter language 00:47:23 yes, that's me 00:47:25 oerjan: ooh, I knew that name was familiar! 00:47:30 hi! 00:47:44 http://impomatic.blogspot.com/2009/01/brainf-interpreter-in-redcode.html has the redcode thing. For an assembly language program, it's rather short. 00:47:45 i just wrote one that seems pretty short 00:47:56 in 117 bytes 00:48:03 nice 00:48:04 what language? 00:48:18 in lambda calculus 00:48:29 binary lambda calculus to be precise 00:48:30 nice 00:48:37 -!- olsner has joined. 00:48:44 that's really concise, for LC 00:49:03 well, has anyone tried in LC before? 00:49:07 not that I know of 00:49:23 it was something of a challenge:) 00:49:28 tromp_: yours is probably the shortest interp 00:50:01 nice testament to the power of LC 00:50:05 yes 00:50:13 Lambda calculus is the preferred language for enterprise deployment. 00:50:30 :-) 00:50:40 with a little sugar on top:) 00:50:46 Shortest C brainfuck I've seen has been 201 bytes, but I'm not sure if it's the shortest C one there is. 00:51:14 imine is actually more like 936 bits 00:51:16 mine 00:51:26 but i just divide by 8 for convenience 00:51:36 that's ridiculously small. 00:51:45 show it? :) 00:51:57 it's on my BLC wikipedia page 00:52:21 http://en.wikipedia.org/wiki/Binary_lambda_calculus#Turing_machines_in_BLC 00:52:29 My own C interp seems to be 265 bytes; the bloatedness. 00:52:40 tromp_: very nice 00:52:44 I have no idae how that works at all :) 00:53:13 maybe i should show the haskell code i used as a guideline 00:53:26 that'd be neat :) 00:53:30 that's infinitely more legible:) 00:58:58 hmm, does wikipedia have a verbatim environment? 00:58:59 The redcode interpreter is 13 instructions, but it's a bit difficult to count the bits, since the instruction values are abstract sort of numbers, and the operand size is configurable; and in any case you might consider it a cheating one, as each brainfuck commands needs to be transformed to a single specific redcode instruction and appended to the interpreter program. 00:59:13 23:58 tromp_: hmm, does wikipedia have a verbatim environment? <- hm? 00:59:30 (12, not 13.) 01:01:01 And Wikipedia can do
...
for does-not-collapse-whitespace does-not-interpret-wiki-markup verbatimness. 01:01:03 tromp_: try
...
01:01:15 or wait, was that backwards 01:01:33 According to http://en.wikipedia.org/wiki/Wikipedia:How_to_edit_a_page the
 tag already ignores Wiki markup.
01:01:48  well, that or 
...
01:01:56 fizzie: not all, i think? 01:02:22 I'm no MediaWiki expert; the descriptions are identical, though. 01:02:48 seems just indenting works to 01:02:50 too 01:03:05 Leading spaces won't stop Wiki markup parsing, though. Just the text reformatting. 01:03:34 hm right 01:04:05 well, it looks ok to me. see http://en.wikipedia.org/wiki/Binary_lambda_calculus with the haskell version added 01:05:27 tromp_: you should be aware that article may risk being deleted for non-notability, though 01:06:45 our wiki has no such requirement, though; your binary combinatory logic already has an article there 01:07:12 Indeed http://esolangs.org/wiki/Binary_combinatory_logic 01:07:55 well though so does wikipedia... 01:08:08 i'd be happy to copy the blc article onto there as well 01:08:45 :-) 01:08:46 note the licenses are different though (we are public domain) 01:09:01 shouldn't matter as long as you're the author, of course 01:13:05 ouch; the esolang wiki doesn't do TeX? 01:13:15 it does doesn't it? 01:13:16 um right 01:13:20 always there is something 01:13:35 hmm 01:13:37 * ehird looks at the WP article 01:13:45 you could replicate that with HTML pretty easily 01:14:12 any volunteers:-? 01:14:25 ''λx0 .'λx1.x0'' 01:14:30 with the double single quotes 01:14:32 would be the True 01:14:46 if you put it on the esolang wiki I'll try to convert it 01:14:54 it's onm there 01:14:58 * oerjan sweeps a "my" under the carpet ^U^U changes a "my" to "this" 01:15:21 ? 01:15:30 in the wp article 01:15:37 oops, where did it go? 01:15:41 http://esolangs.org/wiki/Binary_lambda_calculus <-- I don't see it 01:16:08 "This interpreter is a rough translation of the following version written in Haskell" was "My ..." 01:16:44 tromp_: I'll copy it for you 01:17:06 ok, thanks 01:17:19 i said save page, answered a question, and then it wasnt there:( 01:18:19 hm do we have captchas for non-registered users? i forget. 01:18:24 ah, you got it alrd 01:29:13 man 01:29:15 grad school rocks 01:29:16 :o 01:29:32 now the problem is that i have to update two copies of the page:( 01:31:06 also, i wonder what sorts of computations can be computed using strictly binary lambdas. 01:31:33 ...all of them 01:31:33 ? 01:31:35 what are binary lambdas? 01:31:42 what tromp_ asked 01:32:42 yeah yeah. like.. \x,y -> ... with no pseudo binary higher orderedness 01:32:57 no \x -> \y -> ... 01:33:04 strictly binary functions. 01:33:21 erm just pass tuples? 01:33:31 well can't you just use dummy parameters to mimic unary functions 01:33:33 then you're kind of grounded in identity functions:( 01:33:38 hmm 01:33:44 wait 01:33:45 no no ehird 01:33:55 im curious about the computational power of such things 01:34:06 i was thinking about not even allowing binary lambdas 01:34:33 psygnisfive: please falsify mine 01:34:44 and yes, you could i suppose 01:34:45 like 01:35:29 once you have combinator S then you have everything 01:35:33 (\x,y -> y) (\x,y -> y) 1 01:35:35 or whatever 01:35:46 is the identity function, surely 01:35:46 hmm 01:35:59 well, a binary application 01:36:13 tromp_: you also need k 01:36:14 that produces an identity. 01:36:40 k is what you call binary lambda 01:36:53 oh 01:36:55 so i suppose because you can mimic unary lambdas with binary lambdas, they must be computationally equivalent 01:37:26 tromp_: psygnisfive didn't mean k by it 01:37:37 err 01:37:39 no sorry 01:37:45 i just misparsed your sentence 01:38:02 so you can represent true and false, but you can't compose functions, or make pairs? 01:38:20 but still, it does not seem overly difficult to just use dummies like oklofok says 01:38:26 you can define church numerals 01:38:37 but not the plus function? 01:38:58 or times? 01:39:03 tromp_ i do believe i just said they seem to be computationally equivalent to normal lambdas 01:39:09 note that there is still nothing preventing you from doing things like \x,y -> \z,w -> ... 01:39:27 you just need to write things down correctly :p 01:39:33 function composition: 01:39:38 basically applying a to x would be `x(a,_), and \x->... would be \x,_->... 01:40:08 *x to a 01:40:15 assuming you can't curry, and f(a,b) is used for calling f with a and b 01:40:23 err yes 01:40:31 (i did some variable name switcharoo) 01:41:05 you can fake \x \y \z with \x \y (\a a) \z 01:41:06 supposing f and g are pseudounary functions, and compose = \x,y -> \x',y' -> xx(yyy') 01:41:47 then f.g should be 01:41:51 compose f g 01:42:15 and then this would be applied to a dummy and a value 01:42:29 but you can also disallow any nesting of more than 2 lambdas, which would forbid that 01:42:37 what? 01:43:40 you can ask what functions can be written with de bruijn indices <= 2 01:43:48 what 01:43:56 :P 01:44:06 err what is unclear about what tromp_ is saying? 01:44:39 everything 01:44:52 im not sure if he's even talking about the same things i a 01:45:03 so you can only the two variables bond by the 2 directly enclosing lambdass 01:45:09 only use 01:45:17 what? 01:45:19 hmm 01:45:21 are you speaking english? 01:45:32 tromp_: well there are other universal combinator sets 01:45:38 maybe one only needs two 01:45:44 psygnisfive: no it's finnish 01:45:49 could be! 01:46:01 but theres too few vowels and no umlauts 01:46:01 they all have a combinator on >= 3 arguments 01:46:10 hmm interesting 01:46:31 more interesting than that quantifier thing psygnisfive told me about yesterday 01:46:32 i suspect that has been proven necessary 01:46:43 probably, if it's necessary 01:46:57 but it's not like it would be a millenium problem even if it was not solved 01:47:04 so, in essence, no closures, i think 01:47:10 what the fuck are you people talking about, jesus. 01:47:13 :D 01:47:20 are you all on drugs, god 01:47:49 oerjan: elaborate 01:47:54 psygnisfive: if a lambda expression can only mention the variables of the innermost lambda 01:48:02 oerjan: two innermost 01:48:15 oerjan: yes? go on? 01:48:17 so you have no direct access to outer ones 01:48:25 i didnt say that at all 01:48:37 i didnt say that the binary lambdas couldnt do that 01:48:42 i dont think anyone did, actually 01:48:46 psygnisfive: i don't think tromp_'s had anything to do with yours 01:48:48 psygnisfive: we are trying to make a new restriction so it actually becomes interesting, duh 01:49:06 ok so its got nothing to do with what i mentioned. 01:49:07 since we all agree yours is too easy to circumvent 01:49:16 psygnisfive: it has LC and the number 2. 01:49:21 :p 01:49:42 now since you have no access to outer lambdas, you in essence don't need full closures for your functions 01:49:52 or so i think 01:49:59 sorry so whats the formalism you're proposing, oerjan? 01:50:05 since you only need access to the arguments 01:50:08 oerjan: but you have access to the outer function, or am i misunderstanding something? 01:50:55 right "directly enclosing", so it's basically psygnisfive's but with a stronger limit 01:51:17 examples? 01:51:18 hmm sleep. 01:51:22 psygnisfive: never! 01:51:23 -> 01:51:29 night oklo 01:51:30 <3 01:51:34 sleep well darling 01:51:54 \x,y -> (\a,b -> a b) (\c,d -> c) (x y) would be an example 01:52:02 er wait 01:52:25 that seems like it would do nothing interesting 01:52:40 \x,y -> (\a,b -> a b b) (\c,d -> c) (x y y) 01:52:48 need to apply to pairs as well 01:53:00 \x,y -> (\a,b -> a b b) (\c,d -> c) (x y y) huh 01:53:04 thats 01:53:25 just an example of what's allowed 01:53:26 \x,y -> (\c,d -> c) (x y y) (x y y) 01:53:32 \x,y -> x y y 01:53:58 indeed. i think reductions may preserve this property? 01:54:21 well, what i mean is, that function you described is just "apply" :P 01:54:49 for pseudo-unary functions. 01:54:57 psygnisfive: i'm just showing the syntax, don't expect it to be interesting yet. 01:55:06 what syntax? 01:55:09 maybe there are no interesting functions definable this way 01:55:40 x and y can only be mentioned on the top level inside \x,y 01:55:48 -!- Sgeo has joined. 01:55:51 oh i see what you mean. 01:57:00 so in essence _all_ lambda expressions are closed terms 01:57:15 so that \x,y -> E, where, if E contains a lambda, that lambda cannot contain x or y inside it. 01:57:33 right 01:58:40 indeed i think this removes closures. 01:58:43 and if an expression is applied, it must be to an even number of arguments 01:58:45 but i think that's it, no? 01:59:49 do you need closures? 02:00:03 no. SKI doesnt have closures. 02:00:05 does it? 02:00:09 ah yes, for K 02:00:59 so i guess in some sense you're limiting the calculus to whatever you can do with I and S 02:01:12 actually S too, when you partially apply it 02:01:20 true enough. 02:01:23 hmm. 02:04:34 (\x,y -> x x y)(\x,y -> x x y)(\x,y -> x x y) <-- at least we have nontermination 02:05:14 or just \x,y -> x x x will do it for any second argument 02:05:36 (\x,y -> x x x) (\x,y -> x x x) _ 02:05:38 why not just (\x x x)(\x x x) ? 02:05:47 because that has no right hand side! 02:05:54 (\x,y -> x x y)(\x,y -> x y x)(\x,y -> y x x) 02:06:01 :o 02:06:37 tromp_: even number of arguments required 02:06:49 right hand sides even more so :P 02:07:59 oh right D 02:08:02 : 02:08:05 D: 02:08:06 :D 02:08:16 D:D:D:D:D:D:D:D:D:D 02:08:25 MAKE UP YOUR MIND 02:09:02 the Escher Smiley 02:09:32 so what kinds of computations can we have if we only can only have bound variables at the top levels of lambdas? 02:09:37 now we just need to draw it on a mobius strip 02:11:03 can such a system be TC? 02:11:26 can we fake closure? 02:11:47 the requirement that applications must be full is also important, i think 02:12:38 or wait is it 02:12:47 well lets just look at normal LC 02:12:50 with top-level only 02:13:47 actually i think it is still preserved under reductions even if partially applied, at least leftmost ones? 02:14:04 er 02:14:36 ]hm hm 02:16:06 lessee we want top-level variables only but with arbitrary number (or at least two) 02:16:24 so \x y z -> x y z would be allowed 02:16:56 sure why not 02:17:09 \a b c -> 1 2 3 02:17:09 :o 02:17:33 now if our reductions are outermost, then whatever we apply to will be closed 02:17:59 ey? 02:18:40 um i mean if we evaluate very lazily, then there are no free vars at that level 02:19:05 i think we need some more experience with formal proofs really to determine the power of this system 02:19:05 :p 02:19:05 hm actually SKI behaves that way 02:19:12 alas, indeed 02:19:29 but lets see if we can reformulate it as a formal grammar 02:19:35 because that we can get closer to 02:19:46 formal language theory is easy to prove over i think maybe 02:22:34 * oerjan lets the official formal linguist ponder this 02:32:34 any hardcore brainfuck programmers out here? 02:33:35 http://en.wikipedia.org/wiki/Binary_lambda_calculus#BLC8:_byte_sized_I.2FO has a nice challenge for you.... 02:34:37 make that section http://en.wikipedia.org/wiki/Binary_lambda_calculus#Brainfuck 02:40:00 oh thank god 02:40:04 did you clean up that page? 02:40:34 clean up how? 02:40:50 well before it was all ... stuff 02:41:06 also, whats this lambda stuff without variables? 02:41:33 pls follow the link to read up on De Bruijn indices 02:42:16 you replace variables by a count of how far out it's binding lambda is 02:42:22 its 02:42:49 oh i see 02:43:02 so \x\y x becomes \ \ 2 02:43:11 so \x.x = \1, \x.\y.x = \\2 02:43:13 etc 02:43:16 i see i see 02:43:17 interesting 02:43:36 indeed 02:43:42 and quite useful 02:44:20 very interesting indeed 02:46:27 i wonder if theres any insights that can be gained from this.. HMM 02:53:29 going home. g'night folks 02:53:37 night 04:34:24 hmmm 04:34:40 hmmmm 04:34:53 i wonder if you could write something for FUSE that can mount tarballs 04:38:54 tar is not meant to be random access, so you would keep a cache of offsets in the tar 04:55:41 my ~ has 62,000 files in it, so an 64 bit offset will take .5MB of ram 04:57:48 I'm sure it could. 04:58:14 HURD has tarfs... 04:58:41 compressed tarball would be more interesting 04:59:11 Stick the bzip2 translator on a tar file and then mount it with tarfs. 04:59:13 Done. 04:59:25 that's cheating 04:59:41 you don't have enough disk space to store the uncompressed version 05:00:12 ... And the bzip2 translator doesn't store the uncompressed version on disk. 05:00:27 ...but it's also not natively seekable 05:00:33 It's not very fast. 05:00:37 ;) 05:02:27 do you have a better idea than the quadratic time algorithm? 05:03:34 Don't use bzip2, dummy? :p 05:04:08 gzip is the same way... 05:06:21 i imagine you can cache compression contexts 05:06:35 Well... Seekable compression of some sort... Not exactly common for typical lossless compression formats. 05:07:03 but the would be on the order of ~32kb 05:10:35 doesn't need to be truly seekable, because we can make one pass over it 05:10:50 to build some index information 05:15:39 for something like simple huffman coding, you just have to store offsets 05:48:34 -!- Sgeo has quit (Read error: 110 (Connection timed out)). 06:27:51 -!- psygnisfive has quit (Read error: 110 (Connection timed out)). 07:13:17 -!- neldoreth has quit (No route to host). 07:56:13 -!- MigoMipo has joined. 07:57:57 FUSE has got avfs (which mounts "tar and gzip files, zip, bzip2, ar and rar files"; fuse.gunzip for transparent decompression of gzip; archivemount which allows "mounting of cpio, .tar.gz, .tar.bz2 archives. Reading and writing supported." 07:58:05 All of them might be horribly slow, though. 07:58:46 I've seen one seekable bzip2 variant, done in the obvious way (reset of the bzip2 context every N bytes, plus an index of offsets to beginning of blocks). 07:59:59 -!- clog has quit (ended). 08:00:00 -!- clog has joined. 08:02:39 Not very fast if you seek around a lot and read x << N (much-less-than, not bitshift) bytes here and there. 08:14:50 hmm 08:15:15 uuy 08:15:25 er, wwy 08:27:54 -!- MigoMipo has quit (Read error: 110 (Connection timed out)). 08:36:01 -!- oerjan has quit ("Optimis dentibus"). 09:44:25 -!- tombom has joined. 09:47:05 -!- ais523 has joined. 09:56:36 hi ais523 09:56:40 hi 09:58:38 ais523, awesome project name: http://code.google.com/p/google-thingbrowser/ 09:58:53 a thingbrowser? 09:58:53 not sure what it actually does 09:59:00 browses things, obviously 09:59:13 ais523, "The Thing Browser is a framework for loosely coupled components -- eventually, distributed objects -- that can be intuitively and securely composed by end-users." 09:59:29 oh dear, that's marketingspeak 09:59:32 any idea what it does in English? 09:59:36 ais523, no 09:59:53 ais523, I also think it is written in java 10:00:00 but anyway I like the project name 10:00:13 being written in Java doesn't automatically make something bad... 10:00:22 ais523, true 10:01:25 While being written in an esoteric language automatically *does* make something good. 10:01:56 or better than it would be otherwise, at least 10:01:58 ais523, there is a lot more text on http://code.google.com/p/google-thingbrowser/ about it 10:02:14 * AnMaster hopes either fizzie or ais523 can decode it 10:02:31 decoding marketingspeak hurts my brain, I'm not even sure if I dare look at it right now 10:02:55 ais523, I can't speak marketing. 10:03:08 I only can with great effort, well read it not speak it 10:03:12 reading marketing is like reading machine code 10:03:26 ais523, without a disassembler? 10:03:38 sort-of with a disassembler, but it's buggy 10:03:51 ah 10:04:28 I can read disassembled x86 pretty well. At least if it is in AT&T syntax. 10:04:33 I can't read intel syntax 10:04:46 I don't read marketing very well either, and I have that presentation in three hours and still haven't designed that homework for the listeners. :p 10:04:56 I can read both asm syntaxes, although I get confused between them 10:05:15 fizzie, go do it then? 10:06:00 fizzie, wait? I thought you were a student? 10:06:30 Yes, well, this is a seminar course, and each presenter must design (and grade) a short homework-style thing for the others. 10:06:43 ah 10:06:44 I assume it's a clever ploy by the course organizers to get by with less work. 10:06:48 fizzie: how many people are on it? 10:06:58 doing one homework for each other person in your class must be pretty awful 10:07:19 ais523: Just 12, if I counted right. And the homework should be something that "does not take more than half an hour to answer". 10:07:41 ok, so you'll be spending a little under 5 and a half hours doing it then 10:08:42 Yes. Well, in theory. In fact I haven't actually answered the homework thing given by last week's presentation-doer yet either; should do that too. 10:09:39 At least the ten-page article and the presentation slides are sort-of finished, so I'm not completely... what is the idiom? Screwed? 10:10:20 screwed is one possibility, yes 10:20:16 -!- neldoreth has joined. 10:24:36 -!- ais523 has quit (Read error: 104 (Connection reset by peer)). 10:27:43 http://code.google.com/p/es-operating-system/ <-- wow 10:30:11 5 and a half hours of homework is less than pretty much any of our courses have 10:30:16 except the ones that don't have homework 10:30:38 The homework in this one is supposed to be a pretty small part of the overall work-to-do. 10:30:52 usually it's more like 2-4 per week for the 6 weeks (or something) the courses last 10:30:59 ah 10:33:27 well. gotta go read my booker -> 10:36:22 Your bookie. 10:50:56 The homework, it is done. (I mean the questions for others, not the one given by last week's guy.) And with something like two hours to spare. Now I should print dead-tree copies for everyone of both my slides and the actual article; that's something like 15*(10+4) pages, even if I print the slides with four-slides-per-page. 10:56:07 Oh; the instructiomotions have been changed, and it's now only the slides. I guess that makes more sense. 11:05:18 hmm 11:05:26 what was the last guy's question? 11:05:36 i can probably solve it with a glance 11:05:40 cuz i know my signals. 11:05:50 (*at) 11:07:13 It's about microphone arrays for speech recognition, and it's a pretty free-form question; we're meant to pick one application that actually uses that stuff, and then answer (a) how many microphones, (b) in what sort of array and (c) what are they used for. 11:08:16 -!- Slereah has joined. 11:08:46 well this is just a hunch but (a) 7 (b) an array of 7 (c) speech recognition 11:09:17 Thanks for the help, but I think I'll write a slightly different answer here. 11:09:38 well yeah i guess you don't learn if you cheat 11:12:07 yay coffees are done, maybe some more readings, of the bookie -> 11:12:12 bookance 11:15:02 wow 11:15:20 -!- olsner has quit ("Leaving"). 11:15:26 i spilled all of my coffee on a pile of papers 11:15:33 so now i get to make more coffee! 11:15:35 :D 11:15:38 life is so awesome 11:18:36 -!- ais523 has joined. 11:20:51 -!- Slereah_ has quit (Read error: 110 (Connection timed out)). 11:43:33 ais523, http://code.google.com/p/es-operating-system/ 13:00:36 -!- MizardX has quit ("Proclamation of invalidity!"). 14:02:25 -!- Mony has joined. 14:07:34 plop 14:07:39 hi Mony 14:07:52 how are you ? 14:08:03 slightly tired, slightly overworked 14:08:40 Phew, presentation done; now to listen to the other speaker for today. 14:08:42 heh 14:46:00 Yay, I had an almost exactly 30 minutes (the time given in the specification) long talk even though I did slides for it half-asleep at 03am and didn't prepare the actual presentationary part even once. 14:47:08 huge http://www.geekologie.com/2009/03/real_life_spiderman_paralyzed.php 14:47:32 well you're a scientsist now, get used to it 14:48:23 The URL seems to suggest that a spiderman got paralyzed, but it was in fact almost the opposite. 14:48:27 hmm... it seems from the Slashdot discussion that the spider was probably irrelevant in that, but it makes a good story 14:48:37 lol fizzie 14:49:00 the URL are misleading 14:52:13 "great now everybody's gonna try it" :D 14:52:32 lol 14:52:50 Next: a three-page report in Swedish, delivered tomorrow. Maybe I could subcontract AnMaster or someone to do it... 14:53:47 well i have about 500 pages to read until monday! 14:54:11 i'm assuming this is a topping contest 14:54:30 fizzie, are you swedish or are you just learning swedish ? 14:55:06 i don't think either 14:56:28 Well, I mean, technically speaking I guess I should be learning it. There is an obligatory Swedish exam part of our study curriculum (it's in the law, even), and I thought I'd get it done easier by doing it in course form. 14:56:48 fizzie: what nationality are you? 14:56:53 Finnish. 14:57:06 This is a bilingual country. 14:57:13 sweet 14:57:56 5.6 % actually speak Swedish, but everyone's supposed to be able to. 14:59:52 ime most people are 14:59:57 (to some extent) 15:00:23 -!- BeholdMyGlory has joined. 15:01:37 Yes, I guess so. 15:01:56 i went to Sweden during February, it's a great place 15:05:39 [translated from the assignment] "Before you start writing you should make sure you understand the *purpose* for your report. *Why* are you writing it and *to whom* is it meant for?" I have a hunch for most the answers are "because it is required of us" and "to the teacher". 15:05:55 yep, typical teachers 15:06:36 I think we were also instructed to avoid writing it so that it looks like it's written for the teacher to read. 15:07:02 Do this work, and also come up with a plausible reason for doing it unrelated to this course! 15:11:10 -!- ehird has left (?). 15:11:19 -!- Hiato has joined. 15:11:40 -!- ehird has joined. 15:14:20 -!- FireFly has joined. 15:28:02 08:59 AnMaster: ais523, "The Thing Browser is a framework for loosely coupled components -- eventually, distributed objects -- that can be intuitively and securely composed by end-users." 15:28:11 err that sort of makes sense… 15:28:22 well yes, that doesn't prevent it being marketingspeak though 15:28:34 i mean, composing objects is like feeding the result of a function into another, pretty much, like composing "addition", "text form field" and "label" to make an addition-calculator app thingy 15:28:50 and the distributed thing means the components'll go over the interwebs eventually 15:29:00 and the intuitive part means, uh, you can use a gooey interface to do it 15:30:45 the google- part means google own it :P 15:33:23 -!- tromp_ has quit (Remote closed the connection). 15:37:04 ehird, so what is the thing useful for? It doesn't say really 15:37:15 what do you mean? 15:37:21 it does certain things, I just described what they are 15:37:31 what to do with them is, presumably, your problem. 15:37:46 heh 15:37:50 newsflash: Google develop complicated thingbrowser technology, are waiting for AnMaster to tell them what to do with it 15:37:57 well I was looking for something else when I ran into that project 15:37:58 :D 15:38:04 -!- Hiato has quit ("Leaving."). 15:38:10 http://www.bloomberg.com/apps/news?pid=20601087&sid=a_0CHRWCFx4w&refer=home 15:38:11 wut 15:38:16 ibm buying sun 15:38:26 * AnMaster checks date 15:38:27 not finalised yet 15:38:28 no not yet... 15:38:35 but yes, it's recent news 15:38:42 did I say bought 15:38:55 also, Dell have announced a laptop that's suspiciously similar to a macbook air, but worse 15:39:13 always the innovator 15:39:13 s 15:39:18 wow 15:39:18 that is ugly 15:39:18 ehird, also my "checks date" was a reference to 1 April 15:39:24 AnMaster: ah :P 15:40:10 see, I didn't think there could be a stupider laptop than the macbook air 15:40:13 now I've been proven wrong. 15:40:28 by the way, it's also more expensive and less powerfu 15:40:30 *powerful 15:40:37 hmm... power-fu sounds kind-of impressive 15:40:59 it's also probably heavier 15:41:13 yep 15:41:27 wow, why on earth :P 15:41:40 well, it does have a few things that the macbook air doesn't, like bluray drives 15:41:55 blu-ray? you mean that thing nobody uses? 15:42:07 although the macbook air doesn't even have a cd drive... 15:42:09 well, it is starting to catch on, at least more people use it than its rivals 15:42:17 but that's mostly for watching films 15:42:23 that's because it only had one rival and they dropped out :P 15:42:27 well it is better than no cd drive indeed. 15:42:30 and a macbook-air-like thing strikes me as not being ideal for filmwatching 15:42:39 ehird, how would you upgrade the OS on macbook air? 15:42:41 ais523: You're meant to buy the films from itunes 15:42:44 or install another one? 15:43:01 does it support network booting or such? 15:43:04 AnMaster: Plug it via usb into another computer, put in upgrade CD. Or, use software update to install a minor upgrade. 15:43:08 AnMaster: probably it supports Parallels 15:43:19 and you can get USB CD drives 15:43:24 or just have the OS on a USB stick 15:43:30 yes, you can buy an external cd drive 15:43:52 oh yes I remember now, the one you couldn't use with anything else and didn't work if you used an usb hub 15:43:53 right? 15:43:59 ? 15:44:10 uh, dunno about the hub hting 15:44:12 but probably the former 15:44:25 I don't think the Macbook air's target market is tinkerers 15:44:49 ehird, I remember there was some external usb based CD drive that apple released a month or so after the macbook air. IIRC it didn't work if you did: air - hub - cd drive, only like: air - cd drive 15:45:04 because apple did something weird to make it unusable with anything but air 15:45:08 the superdrive; that's quite possible about the hub, odd though 15:45:20 is a superdrive like a mighty mouse? 15:45:25 -!- Hiato has joined. 15:45:37 ais523: that name is really old; it used to mean the floppy drives 15:45:45 "SuperDrive is a trademark used by Apple Inc. for two different storage drives: from 1988–1999 to refer to a high-density floppy disk drive capable of reading all major 3.5" disk formats" 15:45:54 http://en.wikipedia.org/wiki/File:Applesuperdrive.png <- Macbook air thang 15:46:01 oddly big 15:46:04 there is only one major 3.5 inch disk format nowadays 15:46:11 although there were more back then 15:46:22 ais523: there's no major floppy formats nowadays 15:46:29 they're completely dead 15:47:02 I don't know, I've been known to use floppies to transfer files from my windows 95 computer to my windows XP computer, or vice versa 15:47:08 ais523: "windows 95" 15:47:11 that's also dead :P 15:47:12 oh, iphone os 3.0 will finally offer system-wide cut, copy and paste. how *revolutionary*. 15:47:15 A SuperDrive was an option for this iBook I didn't get, since I already had a DVD burninator. Now it's a completely superless drive. 15:47:27 although for really big transfers I use XPDT, it's a program I wrote specifically for the task of transferring files between windows XP and windows 95 15:47:36 fizzie: the only cd drives apple sells now are superdrives 15:47:43 ais523, does the dell one have ethernet? 15:47:44 ais523: what does it do? 15:47:58 it is fairly useful, like when you are configuring your wireless access point 15:48:08 ehird: just splurges the data through a serial port as fast as possible 15:48:14 heh 15:48:15 with minimal protocol involved 15:48:24 AnMaster: why on earth would you use wireless on a desktop machine 15:48:25 it works about 2/3 of the time 15:48:26 it's so slow for that 15:48:37 the rest of the time, packets get dropped and it doesn't notice 15:48:38 i mean, compared to ethernet the speed is excruciating 15:48:39 ehird, err talking about the dell laptop 15:48:42 oh 15:48:43 not everyone has more than one computer 15:48:51 I don't either 15:48:52 but 15:48:55 anyway what if a person only has a macbook air 15:48:58 you don't. 15:49:06 if it's your only computer you're not in the target market 15:49:13 Yes, the Dell laptop's got the etherweb. 15:49:14 ah 15:49:45 the target market mainly consists of, well, people who go to starbucks, order a latte while reading pitchfork reviews. As far as I can tell. 15:50:12 I assumed it was aimed at people who like to think they're stylish and have a lot of disposable income 15:50:18 There was that one nifty proggie (magelink?) that did file transfer over IPX networks. Useful since TCP/IP networking in DOS is always a bit iffy. 15:50:45 ais523: that's a superset of what I said 15:50:56 I wrote a remote binary diff program that could be used even when the two files you were diffing were on different computers that weren't connected in any way at all 15:51:15 the macbook air lets you choose your own colours; I'm going to go configure one like Hot Dog Stand. 15:51:21 take that, design 15:51:24 Did LapLink do serial lines too? (The parallel port laplink connection was faster, at least.) 15:52:08 hm wait, that's a third party case 15:52:26 how silly of me to think that apple would let a user tarnish their design! :P 15:52:37 Oh, oh, and SMODEM, now *that* was a revolutionary idea. You could do BBS-chatting *while downloading a file*, instead of just looking at the ZMODEM file download dialog for two hours. 15:53:21 But then your download would take four hours so you preferred to just stare anyway 15:53:44 I didn't; was it really a lot slower? 15:54:19 I was referring more to the fact that you're using up bandwidth for the chatting 15:54:40 I never actually used BBSs back in the day (only a few times late in the day) so I don't know 15:54:46 Oh. 15:55:14 Well, the chatting really didn't use that much bandwidth. 15:55:30 Bandwidth of a human/keyboard combination is pretty low, after all. 15:55:42 -!- MizardX has joined. 15:55:49 There's always protocol overhead. 15:56:04 Not much if it's a simple one. 15:56:38 I wrote a remote binary diff program that could be used even when the two files you were diffing were on different computers that weren't connected in any way at all <-- how did it work? Checksum of blocks? 15:57:14 yep, checksum 15:57:19 Smodem is also (according to Wikipedia) more efficient than Zmodem anyway, so maybe the chatting overhead just evens the scales. 15:57:32 ais523, that you had to enter manually on the other computer or something? 15:57:50 no, both computers displayed checksums, you merely had to compare them 15:57:55 as in, same or different 15:58:02 ah 15:58:43 ais523, hm did it use binary search? 15:58:53 if not I got an interesting idea just now... 15:59:03 yep, binary search 15:59:03 ... 15:59:04 what 15:59:07 ais523, aha 15:59:08 what's the idea anyway? 15:59:13 oh, misread 15:59:19 -!- Hiato has quit ("Leaving."). 15:59:41 ais523, "checksum file", ask user if same, if not display checksum for each half of the file, ask users which parts are different 15:59:51 you may end up with more than one different block of course 16:00:03 so not exactly binary search, but a related concept 16:00:13 err 16:00:16 related in that it's a search? 16:00:18 you can do the search from both ends to find where the different block starts and end 16:00:18 :| 16:00:19 *ends 16:00:27 Holy shit 16:00:29 Parrot 1.0 is out 16:00:33 .................................................................. 16:00:34 : | 16:00:34 ehird: I was going to mention that to you 16:00:39 : | 16:00:39 but you mentioned it to me first 16:00:43 | : 16:00:44 ehird, same idea as git bisect 16:00:56 (this is me stretching out in time, then going faster than light thus warping backwards) 16:00:58 ais523, parrot? as in perl6!? 16:00:59 | ::::::::::::::::::::::::::::: 16:01:01 WHAT?!! 16:01:01 > >>>>><<<<< 16:01:03 **BAM** 16:01:03 AnMaster: no, just the VM 16:01:06 . 16:01:06 oh 16:01:07 . 16:01:07 -!- M0ny has joined. 16:01:08 . 16:01:09 . 16:01:09 was scared there 16:01:11 . 16:01:13 . 16:01:15 ...... 16:01:15 . 16:01:16 although it still scores quite highly on hell-freezes-over stakes 16:01:17 . 16:01:22 * ehird dies 16:01:22 ais523, yes indeed. 16:01:27 Metadies, rather; I'm still a ghost. 16:01:32 hah 16:01:34 -!- ehird has changed nick to ehirdghostghost. 16:01:39 Metawhooooh 16:01:41 ehirdghostghost, a ghost with bad memory :P 16:01:54 Shut up or I'll metawalkrightthroughyou. 16:02:51 also, it seems that a bat managed to get into space 16:02:58 by hitching a lift on the outside of a space shuttle 16:03:07 ais523, it died? 16:03:10 I assume 16:03:13 ... 16:03:14 nooooooooooooooo 16:03:16 it survived fine 16:03:18 what a ridiculous question 16:03:24 i mean seriously. 16:03:27 AnMaster: nobody's quite sure 16:03:36 I suppose there's an offchance it's dormant, cryogenically preserved 16:03:50 Err, frozen != cryogenically preserved 16:03:52 but death does seem like a likely option for an animal catching a ride on the outside of a space shuttle 16:04:03 Unless there's cryopreservant in space 16:04:05 ais523, yeah *thinks about re-entry* 16:05:05 30 feet off the pad the engines gave out and the bat carried them into orbit. 16:05:26 :D 16:05:32 so guys 16:05:40 will MATT WRIGHT'S CGI ARCHIVE scripts work in perl6/parrot? 16:05:47 ? ?? 16:05:52 my business depends on formmail 16:06:00 I don't know 16:06:06 :D 16:06:10 you'd probably have to run them through 5to6 first 16:06:15 I was joking ;_; 16:06:36 ais523, hah 16:07:04 ais523, there is a 5to6? It sounds like a parody of the Python 2to3... 16:07:11 ... parody? 16:07:11 yes, of course there is 16:07:17 gee, to! hahahahaahaha! 16:07:21 so funny 16:07:31 ehirdghostghost, well I wonder what will happen when python 6 is released 16:07:35 5to6 came first IIRC, although obviously it hasn't been released 16:07:39 will there be two 5to6 then 16:07:49 AnMaster: by then we'll have invented something better than numbers 16:07:50 AnMaster: that'll never happen 16:07:53 and numbers will be obsolete 16:08:02 ehirdghostghost, oh? 16:08:16 Blue moons get tired of waiting for python's major releases 16:08:39 ehirdghostghost, it is still way shorter than waiting for perl major releases! 16:08:52 no, it's just 6 in particular that's slow 16:08:59 I see 16:09:01 because it's such an insane project 16:09:05 no, perl major releases are fast until 6 16:09:16 s/are/were/? 16:09:17 consider, it got to 5 in, what, 10 years? 16:09:26 python's been around for 10 years 16:09:28 it's only got to 3 16:09:40 hm 16:09:48 ehirdghostghost, perl must be older than that? 16:09:58 python's that old? 16:10:02 perl: 1988 16:10:05 python: 1990 16:10:17 I'm surprised 16:10:21 hmm, python 1991 16:10:27 so it's 18 16:10:30 also, perl caught on a lot faster than python did, it seems 16:10:31 ehirdghostghost, that is nowhere near 10 years indeed 16:10:37 AnMaster: ... yes it is? 16:10:39 it's 18 years 16:10:43 o_O 16:10:44 ehirdghostghost, that is closer to 20 16:10:51 well, fine 16:10:56 so, 18 years and python's up to 3 16:11:06 22 (it's 1987) years and perl's up to 5 16:11:15 and, well, perl 5 came out in the 90s 16:11:34 perl hit v5 at 6 years old 16:11:39 so, anyway, python's major releases are slow. 16:11:42 ehirdghostghost, perl show a non-linear release interval 16:11:57 someone should plot both perl and python releases 16:11:59 by the time 6 is out development will have been dropped due to the singularity causing the best programming language conceivable to be created :P 16:12:02 in some interesting way 16:12:05 * AnMaster looks at fizzie 16:12:19 ehirdghostghost: but what if that language /is/ Perl6? 16:12:26 ais523: holy shit, you're on to something 16:12:34 I bet one of perl6's features is so advanced it needs strong AI to implement 16:12:38 hm 16:12:42 well, they made some changes 16:12:44 and I bet perl6's first complete implementation will be released in 2012 16:12:49 IIRC, the syntax is no longer TC 16:12:51 ehirdghostghost, no. It isn't strong AI. It is something worse 16:12:52 ... 16:12:53 shit, guys, somebody kill larry wall, quickly! :P 16:12:57 AnMaster: strong AI isn't good or bad 16:13:03 it's just "real" AI 16:13:21 ehirdghostghost, Perl 5 == TC grammar. And ais523 said "the syntax is no longer TC"... Right. It is Super-TC 16:13:25 that is why it takes so long 16:13:32 *shit* 16:13:32 no, they made it actually parsable 16:13:37 ais523: stop ruining our fun 16:13:40 not only that, they're maintaining a regexp that parses the whole thing 16:13:41 that's AnMaster's job 16:13:44 which is kind-of insane in itself 16:13:59 ais523: well, perl6 regexps are really full contextual grammars 16:14:03 yep 16:14:03 with a verboser syntax, right/ 16:14:07 and the // is just shorthand 16:14:09 ehirdghostghost, and I heard some rumours about trunk containing a non-euclidean parser to manage it 16:14:11 whilst simultaneously still being regexps 16:14:26 ais523: well, they're certainly not regular (they're far more powerful), and since they have the full syntax they're not "expressions" either 16:14:33 so // is a regex, what you're talking about is a perl6 parser 16:14:40 it just so happens that regexs turn into perl6 parsers on evaluation 16:14:45 hmm... they still work much the same way as, say, cyclexa 16:14:54 or PCRE 16:14:56 ais523, also I agree with ehirdghostghost, stop ruining the fun 16:15:05 15:14 AnMaster: ais523, also I agree with ehirdghostghost, stop ruining the fun 16:15:09 I will treasure this forever. 16:15:20 hmm, and on the same day as parrot's released too 16:15:23 maybe hell really has frozen over 16:15:28 ais523, or worse 16:15:38 * ais523 awaits someone coming in and announcing a BF interp written in Malbolge 16:15:44 I'd better start being a good christian; I was fine with Hell because hey, it's warm right? 16:16:03 ehirdghostghost, yes, all computers overheat there. 16:16:09 anyway heaven is warmer. 16:16:09 ooh 16:16:17 maybe Hell, newly-frozen, is the ultimate way to overclock 16:16:40 * AnMaster remember reading some joke proof based on combining quotations from the bible with physical laws to calculate the temperature in heaven and hell 16:16:42 seventy core 60ghz? no problem, just stick a 286 in Hell 16:16:51 only on #esoteric could people try to figure out ways to take advantage of hell freezing over 16:16:58 :-D 16:16:59 ais523, heh 16:18:50 -!- Mony has quit (Connection timed out). 16:19:00 btw http://www.lhup.edu/~dsimanek/hell.htm (it also has another proof showing that hell is even hotter) 16:19:23 -!- ais523 has quit ("going home, bye everyone"). 16:19:41 hm 16:21:41 So, Duke Nukem Forever will be written in Perl6. 16:21:48 With some parts in raw Parrot for speed. 16:21:59 hah 16:23:53 man, I hope duke nukem forever is really really good. it better be. 16:26:48 -!- Asztal_ has quit ("."). 16:33:17 hm... 16:33:41 hmmmmmmmmmmm. 16:34:24 to me it sounds like perl 6 is suffering second system syndrome 16:34:31 yes, very 16:34:58 it may be the issue with DNF too I guess. 16:35:12 can you do that to a game? I'm not sure 16:35:55 ehirdghostghost, iirc DNF was a non-first game in a series of duke nukem? 16:36:05 well that was awkward 16:36:09 s/was/is/ 16:36:14 but, yes, except 16:36:24 I don't see how it applies to a game 16:36:45 ehirdghostghost, so you want to make a better game next time. That looks even better and with even more levels/maps/items/enemies/whatever 16:36:57 Well, true 16:37:12 Except that, second system effect for games makes them better 16:37:16 For software it makes it worse 16:37:18 sounds like potential SSS there... 16:37:30 For games you get rabidly perfected gameplay; for software you get bloat 16:37:34 err.. wut? 16:37:38 I don't think I've ever seen a game I could call bloated 16:37:44 In games, bloat = more expansiveness 16:37:45 game is not a subgroup of software? 16:37:51 Yes... 16:37:56 I meant non-game software. 16:37:58 ah 16:38:22 ehirdghostghost, I think a game that doesn't fit on a single DVD is bloated. 16:38:38 See, that's entirely irrelevant to the actual gameplay 16:38:45 In gameplay, bloat is meaningless 16:38:50 It's not bloat, it's expansiveness 16:38:57 ehirdghostghost, well depends. You need a very powerful computer to handle a bloated game 16:39:10 That's not related to gameplay 16:39:16 remember those demos back when DOS was the common OS? 16:39:34 some of them seemed to have better graphics than games released a few years later... 16:39:50 and they were way smaller 16:39:59 sure a demo is more limited in scope yes 16:40:04 but even so... 16:40:22 mm 16:40:25 demos are very impressive 16:40:40 The demoscene is just awesome 16:41:02 yep, I remember seeing quite fast real time software ray tracing. Low resolution compared to what we are used to these days, but still 16:41:12 AnMaster: have you watched any of the modern (post-2000) demos? They have the best 3d graphics I've ever seen; really realistic 16:41:21 It's crazy 16:41:32 ehirdghostghost, no I haven't really. 16:41:36 Any specific ones you recommend? 16:41:49 Hmm, I'll try and dig up one I really liked 16:41:52 Gimme a few minutes 16:42:02 ehirdghostghost, do you need dosbox or something? Or is there a video of it 16:42:14 I was going to link to a video site; it uses flash but you could just extract the link 16:42:24 Also, this is way above DOS's capabilities 16:42:33 yeah unless it was that one you used where it was embedded inside the swf 16:42:44 ehirdghostghost, so what hardware do they use then? 16:42:53 Windows, mostly 16:42:59 The modern-3d ones 16:43:08 there were two errors in that 16:43:25 ? 16:43:29 neither windows nor dos are hardware 16:43:41 Hardware was what you said 16:43:43 I assume it was a typo 16:43:47 and you said hardware after I said DOS 16:44:53 ehirdghostghost, well the "then" should have been removed indeed 16:45:03 well 16:45:05 regular hardware :P 16:45:32 ehirdghostghost, well, regular for today or for 2 years ago? 16:45:46 anyway I'll let you dig up that link 16:45:48 They have minimal system requirements, almost always 16:46:15 atm I'm trying to find the playback site 16:46:20 Maybe pouet.net has something 16:47:24 * ehirdghostghost listens to a .MOD while doing it for authenticity 16:47:29 ehirdghostghost, well I know of ways how to tune programs to run very fast on amd64 but slow on intel core 2. And vice verse. So a demo tuned on an amd cpu could potentially be way slower on intel. And the reverse too 16:47:32 Dear God, Just ignore the fact that I'm using OS X please. 16:47:49 AnMaster: yes, but you don't understand; the demoscene tries to do the most with so little 16:47:56 so system requirements are basically never an issue 16:47:59 16.47:31 Dear God, Just ignore the fact that I'm using OS X please. <-- I will treasure this forever. 16:48:05 :-D 16:49:22 ehirdghostghost, a pitty you need a dos/windows feel rather than oldstyle-unix feel. Otherwise I would ask you if it was possible to compile mosaic on OS X 16:49:36 It probably is possible. 16:49:41 With X11, of course. 16:49:47 ehirdghostghost, it uses motif 16:49:52 Yes. 16:50:44 Eh, I'll grep my old Adium logs for it; I linked someone else to i 16:50:45 t 16:50:49 k 16:52:35 % grep -i 'demo' **/*.chatlog 16:52:35 zsh: argument list too long: grep 16:52:40 >_< 16:53:56 Erm, I wonder what to do now. 16:54:50 AnMaster: thoughts? 16:58:48 heh 16:59:09 ehirdghostghost, find . -name '*.chatlog' -exec grep -i demo {} + 16:59:19 that won't print out the name :P 16:59:20 ehirdghostghost, you may need to quote something in that for zsh 16:59:21 not sure 16:59:24 ehirdghostghost, it will 16:59:29 oh, it will? 16:59:29 okay 16:59:33 ehirdghostghost, the + there? see it? 16:59:36 ah 17:00:06 ehirdghostghost, it tells find to "pass as many files in each invocation of the command as you can without getting argument list too long" 17:00:17 ha 17:00:43 ehirdghostghost, so it is possible if you just get one file in the last chunk that there will be no filename 17:00:48 but most likely not 17:01:37 hrm 17:01:44 ehirdghostghost, also that won't happen on linux btw. Linux dynamically decides the size that can be used for the argument list. 17:01:52 there is some ulimit setting for it iirc 17:01:58 Surely it could go into the shell 17:02:04 ehirdghostghost, ? 17:02:08 Or is it an OS-wide limitation? 17:02:12 If so that's stupid 17:02:35 ehirdghostghost, it is a limit you can check the value of with sysconf() 17:02:44 yikes. 17:02:53 $ getconf ARG_MAX 17:02:53 2097152 17:02:55 % ulimit 17:02:55 unlimited 17:02:59 That makes me feel reassured. 17:03:05 I don't think I'll search further; I like being unlimited. :P 17:03:08 ehirdghostghost, unlimited what? 17:03:13 AnMaster: Unlimited. 17:03:18 ulimit handle several things 17:03:19 remember 17:03:24 Yes; they're unlimited. 17:03:26 Just look! 17:03:27 not all no 17:03:31 anyway 17:03:32 16:03 ehirdghostghost: I don't think I'll search further; I like being unlimited. :P 17:03:36 STOP DESTROYING MY REASSURANCE :| 17:04:47 I can't find the argument count in ulimit here, *greps kernel sources* 17:06:09 hmm, finding this link is hard 17:07:35 ah 17:07:38 the stack size limit 17:07:44 Since Linux 2.6.23, this limit also determines the amount of space used for the process's command-line arguments and environment variables; for details, see 17:07:44 execve(2). 17:07:54 RLIMIT_STACK 17:08:22 "rootkit code to exploit major Intel chip flaw to be posted 3/19/09" 17:08:27 Oh, fuck. 17:08:33 ehirdghostghost, where? 17:08:34 Anyone got a spare AMD chip? 17:08:41 AnMaster: http://www.networkworld.com/community/node/39825 17:09:07 "The heart-stopping thing about this particular exploit is that it hides itself in the SMM space. To put that into perspective, SMM is more privileged than a hypervisor is and it's not controllable by any Operating System." 17:09:27 ehirdghostghost, you still need to be ring 0 to do it it seems 17:10:11 bad yes, but you need to be able to get your code to execute in ring 0. 17:10:41 but SMM... very nasty and hard to detect indeed 17:12:11 i am so tired of typing 'tar xzf' and dealing with zipbombs and blarrrgh; I think I'll write a program that lets me do 'unwrap ' and it figures everything out for me. 17:12:26 actually, OS X handles tarbombs 17:12:32 ehirdghostghost, hm I think I saw such a program 17:12:39 if there's one file, it puts it in the same dir 17:12:43 otherwise, it makes a folder for them 17:12:57 AnMaster: can it handle stuffit expander files? i hate them so fucking much 17:13:12 ehirdghostghost, probably not, I don't even remember the name of the program... 17:13:17 * AnMaster searches 17:13:56 [N] app-arch/unp (1.0.14): Script for unpacking various file formats 17:13:56 http://packages.qa.debian.org/u/unp.html 17:13:59 maybe 17:14:02 bbs 17:25:10 back 17:26:35 I guess I'll write it. 17:27:02 Need to find open source expanders to wrap, though. 17:28:09 ehirdghostghost, apart from stuffit that should be rather simple 17:28:23 Well. Yes. 17:28:36 But Stuffit support is quite important to me; it's an irritatingly common format. 17:28:40 Well not too common; still. 17:29:21 yay, sourceforge is down. 17:29:59 pax/tar + bzip2/gzip/lzma, zip/unzip, p7zip (7z for *nix), cpio, unrar 17:30:04 I suppose you don't care about shar 17:30:15 AnMaster: what's that from? 17:30:19 oh, right 17:30:24 outdated format 17:30:28 Problems: 17:30:31 Unzip/unrar aren't open source 17:30:35 ehirdghostghost, oh hm true 17:30:50 I mean, I could just depend on them anyway, but it'd be nice to have them open source 17:31:18 ehirdghostghost, http://search.cpan.org/dist/Archive-Zip ? 17:31:41 that would be usable; It'd be nice to have something a bit less perl though 17:31:52 * ehirdghostghost googles 17:32:00 ehirdghostghost, http://zziplib.sourceforge.net/ ? 17:32:15 http://common-lisp.net/project/zip/ ? 17:32:25 stop it 17:32:26 http://www.nih.at/libzip/ ? 17:32:27 I've found something 17:32:27 ok 17:32:30 wow at that url 17:32:33 nih... 17:32:37 heh 17:33:00 ehirdghostghost, this is bzip2. In parallel. http://compression.ca/pbzip2/ ? 17:33:09 …parallel? 17:33:11 * AnMaster just found it in his package manager 17:33:23 ehirdghostghost, multi-core it seems 17:33:32 sj. 17:33:33 *ah 17:34:11 yjod od omytrdyomh 17:34:47 this is not interesting :P 17:34:47 idomh sm pggdry pg pmr 17:34:57 Pggdry pg pmr. Sounds jabberwocky. 17:35:23 ehirdghostghost, you mean the author was a time traveler?! 17:35:31 wait yes, didn't IWC say so a while back? 17:35:45 or? I forgot the details 17:35:56 ? 17:36:29 ehirdghostghost, the pirates and the english sailors used that phone booth to time travel before the end of the universe. Remember? 17:36:41 I don't read IWC. 17:36:41 Info-ZIP supports hardware from microcomputers all the way up to Cray supercomputers, running on almost all versions of Unix, VMS, OS/2, Windows 9x/NT/etc. (a.k.a. Win32), Windows 3.x, Windows CE, MS-DOS, AmigaDOS, Atari TOS, Acorn RISC OS, BeOS, Mac OS, SMS/QDOS, MVS and OS/390 OE, VM/CMS, FlexOS, Tandem NSK and Human68K (Japanese). There is also some (old) support for LynxOS, TOPS-20, AOS/VS and Novell NLMs. Shared libraries (DLLs) are availa 17:36:45 ble for Unix, OS/2, Win32 and Win16, and graphical interfaces are available for Win32, Win16, WinCE and Mac OS. 17:36:48 ^ good lord what the fuck 17:36:51 didn't they run into Lewis Carroll 17:36:53 ? 17:37:16 oh ok 17:37:21 lewis caroll was awesome. 17:37:47 anyway info-zip seems to be open source, maybe 17:38:01 ehirdghostghost, yes, but he can't possibly have known about qwerty layout and used offset of one on the layout to create jabberwocky. 17:38:05 i guess I'll ask Phil Katz about the license :-D 17:38:06 unless he was a time traveler 17:38:11 /bad-taste 17:38:50 -!- tromp has joined. 17:40:53 As of 2007, the latest sources and binaries for Zip, UnZip, WiZ and MacZip (including encryption code) are available only at Info-ZIP's SourceForge site. 17:40:59 hmm, so unzip _is_ open source 17:41:24 oooh xkcd today. I had a dream like that a few weeks after finishing high school... 17:41:31 hi tromp 17:41:48 maybe it was used on itself and it like... opened it 17:42:08 sorry my brain has gas 17:42:10 What would this place be without oklofok attempting to match oerjan's puns 17:45:26 probably just a ghost town 17:46:44 hi , ehird 17:54:14 About file names, GNU grep (and I assume at least some others) has the -H (or --with-filename) flag for always printing the filename even given a single file argument. 18:20:20 -!- Hiato has joined. 18:22:41 bbl in a few hours 19:13:40 -!- Asztal_ has joined. 19:40:11 -!- BeholdMyGlory has quit (Remote closed the connection). 19:40:33 Is "Örjan Ekeberg" the same thing as "our" oerjan? 19:40:47 No. 19:40:51 He's oerjan johannsen or something 19:41:11 fizzie: Øerjan Johansen. 19:41:15 err 19:41:17 Ørjan 19:41:17 ofc 19:41:40 Right, this one is a .se person anyway. Never-mind. 19:43:23 Should've realized from the Ö/Ø thing anyway. You just never know with computar scientsists. 19:57:39 AnMaster: you there? 20:00:48 -!- tombom has quit ("Peace and Protection 4.22.2"). 20:24:41 -!- kar8nga has joined. 20:39:14 The binart LC article is all fucked up for the wiki code 20:39:33 -!- k has joined. 20:39:53 -!- kar8nga has quit (Nick collision from services.). 20:39:55 -!- k has changed nick to kar8nga. 20:42:23 -!- Hiato has quit ("Leaving."). 20:42:25 yes, better read the Wikipedia version 20:42:43 until ehird gets around to converting it 20:54:16 i added a NOTE linking to the wikipedia version 21:07:31 -!- xTC- has joined. 21:08:58 -!- tombom has joined. 21:33:17 -!- calamari has joined. 21:33:28 -!- neldoret1 has joined. 21:33:29 -!- neldoreth has quit ("leaving"). 21:40:10 -!- Mony has joined. 21:43:24 -!- tromp has left (?). 21:44:38 -!- Mony has quit (Client Quit). 21:46:19 -!- M0ny has quit (Read error: 60 (Operation timed out)). 21:54:20 -!- Slereah_ has joined. 21:57:58 -!- BeholdMyGlory has joined. 22:04:16 -!- Slereah has quit (Read error: 110 (Connection timed out)). 22:04:42 ehirdghostghost, hello 22:15:56 -!- olsner has joined. 22:21:03 ehirdghostghost, night 22:40:28 -!- xTC- has quit. 22:54:27 -!- kar8nga has quit (Remote closed the connection). 23:17:45 -!- tombom has quit ("Peace and Protection 4.22.2"). 23:28:58 -!- FireFly has quit ("Later"). 23:28:59 -!- BeholdMyGlory has quit (Remote closed the connection). 23:43:18 -!- comex has quit ("Caught sigterm, terminating...").