00:03:06 -!- MDude has joined. 00:06:37 -!- MSleep has quit (Ping timeout: 248 seconds). 00:18:27 -!- pumpkin has joined. 00:20:29 -!- pumpkin has changed nick to copumpkin_. 00:21:10 -!- copumpkin has quit (Ping timeout: 252 seconds). 00:21:15 -!- copumpkin_ has changed nick to copumpkin. 00:31:41 -!- azaq23 has quit (Ping timeout: 248 seconds). 00:39:19 * elliott sends an email that pikhq will see without sending an email to pikhq. 00:42:17 Is it a public mailing list? 00:44:27 Yup. 00:44:34 I saw said email. 00:48:49 -!- azaq23 has joined. 00:48:57 -!- azaq23 has quit (Max SendQ exceeded). 00:49:30 -!- azaq23 has joined. 00:53:32 -!- pikhq_ has joined. 00:53:37 -!- pikhq has quit (Ping timeout: 252 seconds). 00:59:54 pikhq_: ENJOY HAVING YOUR EYEBALLS ACCOSTED WITH WORDS 01:01:19 kallisti: How do I run a subcommand in Perl safely, without running into shell interpolation problems 01:23:52 -!- azaq23 has quit (Quit: Leaving.). 01:26:34 Wow, someone asked to become admin on the wiki who ISN'T Timwi :P 01:52:57 Vorpal: wow, skyrim depends on steam even when bought in-store? 01:54:51 Huh. I thought Valve was the only group that pulled that. 01:54:58 s/group/company/ 02:23:44 http://memegenerator.net/cache/instances/400x/11/11536/11813483.jpg XD (translit: "hajimemasite"tte? NIHONGO JŌZU translat: "Nice to meet you"? *You're good at Japanese!*) 02:30:50 -!- allanpes12 has joined. 02:31:01 -!- allanpes12 has quit (Client Quit). 02:34:20 -!- hagb4rd has joined. 02:34:53 -!- pikhq_ has quit (Quit: Reconnecting). 02:34:59 -!- pikhq has joined. 02:41:41 -!- Gregor has quit (Ping timeout: 260 seconds). 03:31:25 -!- esowiki has joined. 03:31:25 -!- glogbot has joined. 03:31:26 -!- HackEgo has joined. 03:31:28 -!- EgoBot has joined. 03:31:29 -!- esowiki has joined. 03:31:30 -!- esowiki has joined. 03:31:38 -!- Gregor has joined. 03:31:51 @tell Gregor Makes me kinda nervous that glogbackup parts before glogbot joins. 03:31:52 Consider it noted. 03:32:51 elliott: glogbackup parts once glogbot joins #glogbot, which is the first one it joins. 03:32:51 Gregor: You have 1 new message. '/msg lambdabot @messages' to read it. 03:33:12 Gregor: OK, but people could still talk before glogbot rejoins every channel :P 03:33:13 @messages 03:33:14 elliott said 1m 22s ago: Makes me kinda nervous that glogbackup parts before glogbot joins. 03:33:21 elliott: Fair 'nuff *shrugs* 03:34:41 "Let's call the flag -Ewarn, by analogy with -Werror." 03:35:25 Uhhhh, is this some imaginary flag that converts /errors/ into /warnings/ (a concept which doesn't even make sense)? 03:36:09 It also solves the halting problem and summons bacon. 03:36:35 Gregor: Yes. A flag that is actually being implemented in GHC :P 03:36:45 More specifically, it turns certain kinds of type errors into warnings + runtime errors. 03:36:49 For debugging. 03:37:15 I don't think that quote was even a joke, and it was by one of the Simons (I forget which by now), so it might even be called that :P 03:38:59 Ah, I see. 03:39:06 I just assumed GCC due to the "by analogy to -Werror" 03:39:15 It makes more sense in many other contexts. 03:39:26 GHC nabs gcc's -Wall, -W and -Werror. 03:39:28 The only errors there are in C are "I have no clue how to compile this shit, dude" 03:39:39 Except -Wall means "a lot of warnings" and -W means "literally every warning" 03:40:11 Similarly it copies gcc's -On options, where -O0 does nothing, -O does most optimisations, and -O2 slows compilation down massively and doesn't help in the majority of cases. 03:40:15 So... they shifted the numbers a bit :P 03:40:21 Gregor: Not QUITE. 03:40:34 Gregor: You could feasibly bypass C's type checker. 03:40:43 Gregor: Possibly other errors are possible in C as well, such as lack of address space for declared variables 03:41:11 zzo38: Sounds like "I have no clue how to compile this" to me :P 03:41:16 elliott: But virtually any case where the type checker fails, it fails because it doesn't know how to compile it. 03:41:28 elliott: OK, maybe that is what it is, then. 03:41:42 Gregor: Not really? Pointer casting, f'rinstance. 03:41:50 Which is a looooooot of errors in C prorgams. 03:41:58 Well, at least a fair amount :P 03:42:07 Are there even any other type errors?? 03:42:46 I know of no pointer cast that would give you an error (as opposed to a warning) in default GCC. I suppose int x; int y; y = *x; wouldn't compile, but it also doesn't have enough information to. 03:42:59 Cast x to (int *) :P 03:43:06 Then there's no error. 03:43:19 Then it "merely" hits UB. 03:43:33 Gregor: I mean automatically. 03:43:48 *groan* 03:43:56 Gregor: The machine has no notion of a pointer type, soooo :P 03:44:09 elliott: But how do you know it wanted an /int/ * as opposed to a /char/ * or a /wtf/ *? That's the "doesn't know how to compile it" part. 03:44:22 Gregor: Because y is an int. 03:44:24 Duh. 03:44:29 It's like you're not even trying! 03:44:34 Heh, fair 'nuff. 03:44:38 Perhaps you want (int*)(char*) though? 03:44:44 But that's not so much bypassing the type checker as doing nonsense type inference :P 03:44:50 pikhq: Well fuck you, you're compiling an invalid program, be happy it works :P 03:44:55 FSVO work 03:45:10 Gregor: Now I want a list of gcc's errors... 03:45:18 :P 03:45:22 I think in C, it ought to be error to use something that isn't a pointer where a pointer is required unless you have an explicit cast (possibly with some exceptions if it would help to do so) 03:45:40 zzo38: In C, it is. 03:45:50 pikhq: Mmmmm, not quite. 03:46:01 Well, I think except for confusing circumstances. 03:46:03 pikhq: int a(int *x); void foo() { a(42); } 03:46:04 pikhq: In GCC it is usually a warning but not an error when I do that 03:46:16 Most notably, I *think* using an integer literal as a pointer is a warning. 03:46:20 Dear C: WHY DON'T YOU HAVE LAMBDAS 03:46:24 pikhq: 'tis, 'tis. 03:46:29 #define TRANSFORM_RGB(expr) \ 03:46:29 do { \ 03:46:29 uint8_t x; \ 03:46:29 x = rgba.r; rgba.r = (expr); \ 03:46:29 x = rgba.g; rgba.g = (expr); \ 03:46:30 x = rgba.b; rgba.b = (expr); \ 03:46:31 } while (0) 03:46:33 I WANT TO REPLACE THIS HORRIBLE THING 03:46:48 pikhq: But dereferencing an integer literal is a no-go as it doesn't know the pointer type. 03:46:57 elliott: https://github.com/pikhq/clambda-demo/blob/master/lambda.h 03:47:07 Gregor: See, B just only had words. 03:47:09 So there was no problem. 03:47:14 You'd read a word, obviously. 03:47:22 :) 03:47:29 B: better than C? 03:47:33 It is due to the C preprocessor having a few thing missing. Some things I made in Enhanced CWEB allow you to add your own compile-time codes 03:47:53 pikhq: I can't tell you how INTENSELY willing I am to use this. 03:47:57 It's giong into mcmap RIGHT NOW. 03:48:00 *going 03:48:09 There is another programming language BLISS which has far more powerful macro capability and record types than C 03:48:22 pikhq: What happened to town, btw (GITHUB STALKING IS BEST) 03:48:38 elliott: I didn't do much with it? 03:48:47 Bah :P 03:48:51 I've spent the past few months basically doing fuck-all on github. :P 03:49:39 pikhq: And it mixes tabs and spaces too :'( 03:52:15 -!- MDude has quit (Ping timeout: 248 seconds). 03:53:46 The STRUCTURE command in BLISS allows you to do a lot of things; the structure can contain arbitrary commands (it does not have to contain only field declarations) 04:03:33 OWN X; LITERAL MARK = 4; MACRO M = MARK + %UNQUOTE MARK %; BEGIN LITERAL MARK = 5; X = M; END What will the value of X be in this program? It will be 9 04:07:44 int *a = alloca(sizeof(*a)); /* this works, but feels so wrong */ 04:08:36 -!- kmc has quit (Quit: Leaving). 04:09:10 Sure 'nough. 04:09:31 Gregor: Which part of it 04:09:32 It also involves like 20 preprocessor directives to use it "portably". 04:16:50 elliott: The fact that that the definition of a refers to a :P 04:17:02 Gregor: Uhh, but that's a standard idiom with malloc... 04:17:11 To avoid repeating yourself in case you e.g. change a's type. 04:17:32 Oh? I don't think I've seen that ... usually I've seen type a = malloc(sizeof(type)); 04:17:38 Err, type *a of course 04:17:46 Gregor: ...type *a? 04:17:55 Oh, right. 04:18:05 Gregor: Yeah, that's more common, but *a is fairly common too. 04:18:12 Hm *shrugs* 04:18:26 * elliott usually embds type because he finds refactoring the expression to be a more compelling example than retyping a variable. 04:18:30 Long story short, I'm at 2037 code bytes :P 04:18:41 Of course I'd prefer to be able to just omit the "type " on the LHS :P 04:18:46 Bring back auto, dudes! 04:18:54 C++ did it! 04:19:10 Gregor: For what this time? 04:19:30 elliott: ... same as always? 04:19:47 Gregor: Well, OK, you talked about doing something else at one point though :P 04:20:53 I considered it, couldn't think of a way to make it truly obtuse that wasn't also lame. 04:21:52 ALTHOUGH WE COULD DO A COLLAB 04:23:43 Gregor: You mean the MIDI thing? 04:24:21 Yuh 04:24:44 Gregor: I dunno, I don't think it'd have to be very obfuscated if it produced something nice-sounding given untweaked input... 04:25:15 But "nice" is relative, I'm sure the demoscene could synthesise semi-realistic piano and strings within IOCCC limits :P 04:25:15 Exactly. 04:25:41 That was the issue, I couldn't think of a way to make it obtuse. 04:25:50 elliott: ... If they don't already. 04:25:57 Also, piano: Sure. Pianos are easy. Strings: lolno. 04:25:57 Gregor: Are we using different definitions of obtuse? 04:26:13 I don't think obtuseness really matters if the result is impressive... the golfing makes code pretty hard to read to start with :P 04:26:18 Strings was facetious. 04:27:19 Idonno, I just don't feel like the result would be very obfuscated. 04:27:32 Nor do I think it would be sufficiently impressive, though maybe it would be. 04:28:44 I was thinking maybe a MOD player instead... those sound better, but that loses the synthesis element, and I dunno if the effects stuff that MOD players has are hard enough to implement that it'd be impressive :P 04:29:34 Hmmmmmmmmmmmm 04:29:38 Idonno much about MOD. 04:29:57 http://www.scp-wiki.net/clef101 04:30:52 Gregor: All I know is it has samples and channels of notes :P 04:31:39 Sgeo: six paragraphs in and this is the worst writing ever 04:31:51 seven paragraphs and it's worse 04:32:03 wow this is bad i'm not going to read it any more 04:32:34 elliott: That's all I know toooooo 8-D 04:32:37 http://scp-wiki.wikidot.com/skybluesky 04:33:42 Is there any MML compiler to MOD? 04:34:15 Gregor: So there's at least hacky resampling involved (in case /dev/audio doesn't have the right Hz), modifying the tone of samples, and I think applying various effects to a sample. 04:41:18 Hmmers. 04:44:23 Accessing memory-mapped files is just as efficient as accessing ordinary memory, as long as I don't do anything that actually requires disk access. Right? 04:47:02 tswett: Are you asking whether mmap() caches? 04:47:08 Because, yes, your OS has disk caches. 04:47:21 Excellent. 04:47:32 * tswett replaces malloc with mmap. Save all the things. 04:47:45 tswett: malloc is implemented with mmap for large enough allocations in all common libcs. 04:47:48 Something like >4k. 04:48:10 Huh, neat. 04:48:24 @where #haskell 04:48:24 Right here, silly! 04:48:27 Did you know? 04:48:48 @where #esoteric 04:48:49 I know nothing about #esoteric. 04:48:52 @where+ #esoteric Right here, silly! 04:48:53 It is stored. 04:54:27 -!- Jafet has joined. 04:57:47 http://www.scp-wiki.net/scp-1025 04:58:02 elliott, read that completely, don't stop reading just because 04:58:19 But I like stop reading just because! 05:01:52 Sgeo: OK that was amusing. 05:26:41 @tell kallisti By the way, (const undefined) is also strict, but doesn't evaluate its argument. 05:26:41 Consider it noted. 05:27:01 @tell kallisti Proof: const undefined _|_ = undefined = _|_. 05:27:02 Consider it noted. 05:32:36 -!- jix has quit (Ping timeout: 240 seconds). 05:33:18 -!- jix has joined. 05:34:36 -!- Zuu has quit (Read error: Connection reset by peer). 05:39:29 -!- Zuu has joined. 05:53:57 -!- GreaseMonkey has quit (Quit: The Other Game). 06:29:47 @ping 06:29:48 pong 06:32:02 took you a while 06:35:21 -!- elliott has quit (Quit: Leaving). 06:35:25 -!- elliott_ has joined. 06:38:29 What is the way to select a $2n$ by $2n$ matrix of booleans such that there is exactly $n$ true in each row and in each column, from all possible ones, uniformly? 06:40:43 It's impossible. 06:41:07 (Note: I am lying.) 06:45:52 -!- tytythetyty has joined. 06:45:59 hi 06:46:50 hi 06:47:48 i am having trouble with stochastic cellular automata, and i stumbled across this link (http://www.conwaylife.com/forums/viewtopic.php?f=3&t=398) do you guys talk about that in here?? 06:48:21 sure 06:48:27 Ask the question; they sometimes talk about various things in here and possibly someone knows 06:48:38 oh, ##gameoflife, heh 06:48:42 that channel lasted uh... a few days 06:48:51 lol sad 06:48:58 Phantom_Hoover is still a regular here though 06:49:06 I can't say we get CA talk very often but when we do it tends to be lively 06:49:18 although it's quiet hours right now 06:49:25 my question is to design two different stochastic cellular automata: 06:49:36 1. Design a 1D stochastic cellular automata that uses on rules involving two adjacent cells at at time that (with high probability) do the following tasks: 06:49:40 a. If the majority of cells are initially 0, the final state is all zeros. If the majority of cells is initial 1, the final state is 1. 06:49:45 b. Starting with all cells at 0, reach a final state in which one cell is a 1 and the rest are 2. 06:50:11 hmm, cool 06:50:17 this isn't homework, is it? :p 06:50:32 ABSOLUTELY NOT :P 06:51:11 oklopol works in CA, although you may find his answers rather unwantedly vague for a question he considers too trivial :P 06:51:37 and I don't think he's here right now 06:51:45 lol, that is typically the case for IRC channels 06:52:10 not us! we're helpful and cuddly and nice. well... occasionally 06:52:19 haha nice 06:52:30 do you have any idea on this?? 06:52:47 not personally, but there are like three people off the top of my head who might be able to help who aren't here right now :P 06:52:55 sux 06:53:31 ask monqy; you probably won't get an answer but the non-answer might be entertaining 06:53:43 lol kk 06:53:57 wouldn't he just see this?? how should i ask?? 06:54:06 smoke signals 06:54:20 he responds by sending "hi"s in morse code 06:55:06 I SUMMON THY MONQY 06:55:08 .... .. .... .. .... .. 06:55:25 hi 06:55:27 @google majority problem 06:55:29 http://en.wikipedia.org/wiki/Majority_problem_(cellular_automaton) 06:55:29 Title: Majority problem (cellular automaton) - Wikipedia, the free encyclopedia 06:56:47 I don't know enough about cellular automata 06:57:24 me neither :P 06:59:53 Do you know how I could figure out the answer to my question? 07:01:01 ask someone who knows and is willing and able to answer appropriately 07:02:08 lol 07:02:16 -!- DCliche has quit (Quit: You are now graced with my absence.). 07:08:52 elliott_, update 07:44:25 I think I might have figured out what contramonads and contracomonads is supposed to be: contrareturn :: (a -> m ()) -> m a; contrajoin :: m (m a) -> m (a -> m ()); contraextract :: w a -> a -> w (); contraduplicate :: w (a -> w ()) -> w (w a); I don't know how good this is, though. Maybe I made a few mistakes? 08:10:31 elliott_: yes, there's a setting for it 08:12:34 elliott_: :help swapfile 08:35:26 http://www.scp-wiki.net/sandrewswann-s-proposal 08:36:08 Does "But your thoughts are nothing except hallucination!" have anything to do with "greedy reductionism"? 08:39:51 -!- monqy has quit (Quit: hello). 08:47:49 -!- hagb4rd has quit (Quit: Nettalk6 - www.ntalk.de). 08:53:59 I wonder what research on cellular automatons look like 08:54:17 How much is math and how much is just trying stuff on it 08:59:51 -!- zzo38 has quit (Remote host closed the connection). 09:28:08 elliott_: hi 09:28:09 kallisti: You have 2 new messages. '/msg lambdabot @messages' to read them. 09:28:14 qx'blah blah blah' 09:30:46 elliott_: is how you prevent Perl from interpolating in a shell command 09:31:06 if you're using system or whatever else you would just pass a single quoted string obviously. 09:33:35 > (const undefined undefined) `seq` "hi" 09:33:38 "*Exception: Prelude.undefined 09:36:22 elliott_: similarly, using single quotes as a delimiter for any other interpolating quoke-like operator (except for qq, which always interpolates) will turn off interpolation 09:37:29 Slereah: there is a lot of bullshit being done with cellular automata, which is why we're not taken very seriously 09:37:42 s'blah'blah', m'blah', qr'blah' 09:37:51 ...I... don't remember if qw interpolations. 09:38:16 !perl my ($x,$y,$z) = 1..3; print qw($x $y $z) 09:38:19 ​$x$y$z 09:38:21 oklopol: clearly you should work for a prestigious research group like Wolfram 09:38:22 nope 09:38:36 :D 09:38:54 I hear they get great press coverage 09:39:01 !print ?bahahahahaha? 09:39:22 !print $_="hi"; print ?hi? 09:39:32 !print $_="hi"; print /hi/ 09:39:35 actually wolfram invented one of the most important concepts of CA on which i too work every day 09:39:37 the limit set 09:39:43 well he didn't actually define it 09:39:48 because he's a retard 09:40:04 but still, he had quite an influence there 09:40:42 or maybe he actually gave a definition, but you know for the sake of this story. 09:40:59 at least he doesn't have any actual results 09:41:39 Slereah: anyhow, searching for counterexamples is, as in any branch of math, rather ad hoc 09:42:54 otherwise, we use measure theory, compactness arguments, symbolic dynamics and ergodic theory on a daily basis. but any sort of deep results are needed rarely. 09:43:40 just the basic theory 09:44:54 for instance, i've needed a point whose ergodic fibre is the uniform bernoulli measure a few times when studying the besicovitch space, such a point is not that easy to construct, but it follows from ergodic theory that pretty much all points have this property 09:46:48 symbolic dynamics is really where most of the math happens, so if CA wasn't so easy to write, i'd always talk about endomorphisms of the shift just to keep the CA people out 09:47:39 there was this guy in a conference who said he's a computer scientiest but also does a lot of math. he said he's working on cellular automata, and said that HE HAD HEARD THAT YOU CAN GIVE A TOPOLOGY TO THE FULL SHIFT 09:47:43 :DDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDD 09:48:12 oklopol: whoosh 09:48:16 there went the joke 09:48:18 over my head. 09:48:30 kallisti: you can't do anything with CA without the cantor topology on the space 09:48:43 ah. 09:48:54 see I thought you were implying it was impossible or something which... sounded unlikely to me. 09:49:23 he studied something like gliders in elementary CA 09:49:29 what the fuck kind of research is that 09:49:33 ...heh 09:49:36 DUDE THEY LIKE 09:49:37 MOVE 09:49:39 IN A LINE 09:49:46 ITERATING IN A CYCLE 09:49:46 that's so coooool 09:50:10 I guess that's the non-mathematical side of CA research 09:51:09 Slereah: if you want links to the good kind of research, i can show some good representatives of the techniques 09:51:39 if you want to know more about the bad kind, you could always read nkos 09:53:42 > sum . map (\(n,x) -> 2*x / 3^n ) $ zip (cycle [0,1]) [1..] 09:53:46 mueval-core: Time limit exceeded 09:54:14 oklopol: wow shocking 09:55:01 this is a pretty good representative: http://www2.math.umd.edu/~mboyle/papers/automata20oct98.pdf 09:55:42 I wonder how one would convert points in a cantor space to real numbers in Haskell. :> 09:55:47 it seems... difficult. 09:59:59 tytythetyty: that's an easy consequence of http://www.cs.bu.edu/~gacs/papers/long-ca-ms.pdf 10:00:21 sorry, i didn't read any log 10:00:56 kk thanks! i will look at this :) 10:01:34 lol fault tolerant cellular automata 10:02:10 CAs can be used to model networks 10:02:19 So yes, you might ask whether they're fault tolerant 10:02:49 tytythetyty: but apart from being a consequence of that, stochastic automata are very hard to work with, and i don't know if anyone but gacs has really ever succeeded 10:03:09 tytythetyty: err actually 10:03:42 i think if i told someone at work that i just linked gacs to some random guy on irc to read, they would punch me in the face 10:04:06 see, apart from gacs, only one person has ever managed to read all of that 10:04:13 lol i was typing this 10:04:14 phosphoglycerate kinase 10:04:17 dammmit! 10:04:18 nevermind 10:04:35 i was gonna ask if there is a specific section that is applicable?? 10:04:36 in that pdf 10:04:38 why did you ask your question in the first place? 10:05:20 "So there are as many points in the Cantor set as there are in [0, 1], and the Cantor set is uncountable " 10:05:23 well you probably need only a part of the construction. i can link you the reader's guide 10:05:27 * kallisti mind blown 10:05:30 i am doing some work in a Synthetic Biology seminar 10:05:36 and it was a challenge 10:05:44 http://www.cs.bu.edu/~gacs/papers/long-ca-ms.pdf 10:05:50 to people in the seminar 10:05:59 who gave it? 10:06:32 do you know random people in synth bio?? 10:06:33 :P 10:07:09 well no but i mean did they know what they were asking... of course if you just want to have good approximations, you can just try stuff out 10:07:49 oh i see haha 10:07:58 gacs' automaton, afaiu, works with any probability, and simulates any CA you like reliably with high probability 10:08:00 yeah, he likes this stuff 10:08:39 synth bio... is that similar to mathematical biology? 10:08:40 cool, i will look at these, thanks! 10:08:44 errr 10:08:45 not really 10:08:53 i should ask what the exact results are at the university, i'm more into symbolic dynamics myself 10:09:07 so i don't really wank to gacs' paper unlike most CA ppl 10:09:42 since it solved like every problem ever 10:10:06 math bio is more modeling bio systems with math, synth bio is more engineering genetic pathways 10:10:13 well incidentally it didn't solve the problem of finding a uniquely ergodic CA which my colleague did this week, awesome right 10:10:47 i think so :P not sure what a uniquely ergodic CA is 10:17:24 tytythetyty: wow, I didn't realize how far we've come with genetic engineering. 10:17:30 * kallisti is reading about gene networks. 10:18:51 kallisti: if you have access to this somehow (i.e. an academic proxy), or an actual subscription, this is a good overview 10:18:52 http://www.sciencemag.org/site/special/syntheticbio/ 10:19:10 not currently no. 10:19:27 well, hmmm, actually I may be able to. 10:21:18 tytythetyty: uniquely ergodic means there's only one dynamics-invariant measure for your dynamical system (the dynamics being the CA). this measure is then automatically ergodic. 10:21:25 but in the case of CA 10:21:26 kallisti: this is interesting too, this is the undergraudate synth bio team at my school, who recently won the international undergrad competition (http://2011.igem.org/Team:Washington) 10:21:46 it means that in every column, the density of some symbol 0 always gets bigger and bigger 10:21:49 oklopol: ahhh i see 10:23:36 but umm i applied for this grant and they said they'd notify recipients by end of march. i heard rumors that i got it but there's another ville who applied and now no one just knows anything. except that i wasn't notified, so i probably didn't get it. well, today, i got an email that says something like dear recipient, please fucking register for our party already. 10:24:35 and i'm confused, did they just send that to everyone who applied and filter out non-recipients by addressing it to recipients (non-recipients have not been notified) 10:25:38 i should probably go talk to someone at the university... we already changed where i'm getting my next year's funding based on me not getting the grant :D 10:25:52 erm i mean my whole 4 year plan 10:27:15 Go to the party and make so many friends that they can't kick you out 10:27:44 good idea 10:27:46 bye 10:28:12 @source nDerivs 10:28:12 nDerivs not available 10:28:16 :t nDerivs 10:28:17 forall a i. (Num i, Num a) => (Dif Expr -> Dif a) -> i 10:28:24 > nDerivs sin 10:28:28 mueval-core: Time limit exceeded 10:28:31 > nDerivs log 10:28:37 mueval-core: Time limit exceeded 10:29:15 grants are a bitch sometimes :/ 10:29:19 and interviews 10:29:34 i had a microsoft interview today, didn't go as well as it should have 10:35:51 -!- hagb4rd has joined. 10:37:42 kallisti and oklopol: good talking to you guys! thanks for the help 10:37:45 have a good night 10:38:22 night 10:38:36 -!- tytythetyty has quit (Quit: Page closed). 10:57:25 -!- itidus21 has quit (Ping timeout: 240 seconds). 11:17:51 hail eris 11:21:32 hail 11:38:41 -!- oerjan has joined. 12:03:57 http://en.wikipedia.org/wiki/Mystical_Ninja_Starring_Goemon 12:04:01 has anyone else played this game? 12:04:45 -!- ais523 has joined. 12:13:23 -!- Phantom_Hoover has joined. 12:28:42 -!- Patashu has quit (Quit: MSN: Patashu@hotmail.com , Gmail: Patashu0@gmail.com , AIM: Patashu0 , YIM: patashu2 , Skype: patashu0 .). 12:32:55 ais523: hi 12:33:04 hi kallisti 12:33:10 http://en.wikipedia.org/wiki/Mystical_Ninja_Starring_Goemon 12:33:13 have you played this game? 12:33:17 N64 game. 12:33:32 no 12:33:35 I don't own an N64 12:34:04 (I /have/ played a very few N64 games on other people's N64s, but a small selection) 12:34:28 ah 12:34:53 some weird Japanese game I played as a kid. I just rediscovered it and my nostalgia glands kicked in. 12:34:57 Vorpal: wow, skyrim depends on steam even when bought in-store? <-- yes 12:35:33 "A cellular automaton is said to be reversible if for every current configuration of the cellular automaton there is exactly one past configuration (preimage)." 12:35:49 psh, come on. you can totally non-deterministically reverse to multiple past images. 12:35:50 elliott_: I'm surprised that you're surprised at that; it's become more and more common recently 12:35:55 I'm upset by it, but not surprised 12:36:12 what's wrong with requiring Steam? it's free right? 12:36:40 kallisti: assumes you're going to have a network connection when you play the game 12:36:47 this is almost always not the case for me, I typically play games offline 12:37:09 -!- elliott_ has quit (Ping timeout: 245 seconds). 12:37:33 I was under the impression that Steam had an offline mode as well. 12:37:52 * kallisti doesn't actually play PC video games often. 12:38:20 but when I do, I prefer Dos Equis. 12:38:30 Only pirates, criminals and communists play offline 12:38:55 And people who might not live in the first world, but they're mostly pirates, criminals and communists anyway 12:39:09 bleh, I prefer "anarchists, lunatics, and terrorists" 12:39:20 Steam has an "offline mode" if you check the "save login details" box, or something like that. 12:39:37 but you'd still need an Internet connection to install the game 12:39:58 Yes, well, it's a Internet-based delivery system, after all. But not when you play it. 12:39:58 it used to be that you could go to a shop, buy a game, take it home, and install it on a non-networked computer 12:40:15 and the only DRM would be requiring the disk to stay in the drive, together with measures to make the disk harder to copy 12:41:01 ais523: it also used to be that there was this thing called "dial-up" 12:41:04 meanwhile, "Scaffolding" seems like an interesting name for an esolang 12:41:11 kallisti: it still exists, believe it or not 12:41:16 well, yeah.. 12:41:35 at least one ISP gave free dial-up internet access to Egypt when the government cut off all the ISPs 12:41:35 it much the same way that floppy drives still exist. 12:41:45 on the basis that they had all this dial-up capacity that was hardly being used 12:41:50 kallisti: I actually have a USB floppy drive 12:42:00 .......why 12:42:30 are you a digital archeologist? 12:42:31 At one point in time multiple laptops came out bundled with USB floppy drives. 12:42:37 unearthing ancient tombs filled with floppy disks? 12:42:49 kallisti: because I used to back things up to floppy disks, back before CD burners were common 12:43:24 also, because floppy disk is quite a convenient way to quickly transfer files from one computer to another; it's pretty much as fast as doing it via USB stick, just with a lower capacity 12:43:51 I have this thing 12:44:09 where I can put up to 5 GBs of information onto a server probably hundreds of miles away. 12:44:17 and then log into that server on another computer 12:44:23 and download the information 12:44:41 I guess it's not as fast as a floppy 12:44:48 when you're like, physically next to both computers 12:45:09 why would you send data hundreds of miles to transfer it across the room? 12:45:11 According to Wikipedia, "the average sequential read speed is 30–70 kB/s". 12:45:34 -!- elliott_ has joined. 12:45:49 if both computers are on the same network, I'd just use that network to send it 12:46:03 ais523: it's okay it's traveling at light speed. 12:46:15 and yeah 12:46:18 the only time bouncing off an external server would make sense for a same-room transfer would be if there was some sort of firewall between them 12:46:20 it would be silly. 12:46:45 ais523: well I'm referring to Dropbox. if you have it installed on both computers it's even easier. 12:46:50 but it's also possible to login from any computer. 12:46:59 via a web interface 12:47:05 (extreme example: the wireless connection here in my office is outside the department's firewall, the wired connection is inside, so I'm careful not to use both at once; and the wired connection is really heavily monitored, to the extent that Chrome refuses to access Google, because I think it's MITMing the https) 12:47:10 that's ridiculous, i demand heavy speed! 12:47:23 We used to use MageLink for transferring files between the computers at the computer classroom at school. It's the spiffiest-looking IPX file transfer thing there is. 12:47:38 haha, now I remember XPDT 12:47:48 Sadly an image search for 'MageLink' is not being very helpful. 12:48:14 it seems that none of the file-transfer stuff that comes with Windows allows transferring over a serial link from Windows 95 to Windows XP, or the other way round 12:48:14 so I wrote my own program to do that 12:48:54 Do they still bundle a LapLink-like thing in modern Windowses? I suppose not. 12:49:24 The "null-printer" cable, the silliest name. 12:49:54 (By analogy with null modem cables.) 12:53:21 Debian's installation manual has a PLIP-based installation method described. 12:54:43 -!- ais523 has quit (Ping timeout: 252 seconds). 12:56:33 "Windows Vista drops support for the Direct cable connection feature [4] as ethernet, Wi-Fi and Bluetooth have become ubiquitous on current generation computers. To transfer files and settings, Windows Vista includes Windows Easy Transfer, which uses a proprietary USB-to-USB bridge cable known as the Easy Transfer Cable." <- right, they've given it up. 12:59:57 -!- ais523 has joined. 13:02:36 On Talk:Afterparty: This sure doesn't answer anything. Also, it's "after party", not one word. It's also a lame concept. (Wow, I wanted to go on a pro-Communist rant there, but stopped.) More needs to be written, otherwise, it just seems like a myth of some sort. Besides (stopping myself again from going totally anti-suburbs here). Apple8800 (talk) 17:26, 29 March 2011 (UTC) 13:03:43 fucking capitalists and their afterparties (one word) 13:06:47 -!- itidus21 has joined. 13:09:50 itidus21: hi 13:32:10 -!- kallisti has quit (Quit: Lost terminal). 13:36:28 also, because floppy disk is quite a convenient way to quickly transfer files from one computer to another; it's pretty much as fast as doing it via USB stick, just with a lower capacity <-- not really, floppies are really slow 13:36:53 fast enough when you're only transferring a few tens of kilobytes, which is typical 13:37:38 kallisti: I actually have a USB floppy drive <-- I have two of them I think. One pure USB floppy drive and one that doubles as a "ultrabay-for-old-dell floppy device" and USB one 13:38:06 ais523: Hm I seldom transfer less than a mb or so 13:38:21 ais523: and quite often something like 10 GB 13:38:30 you can't fit 10 GB on a floppy 13:38:33 indeed 13:38:47 last time I needed to transfer that sort of data, I put the two computers physically next to each other and connected them with an Ethernet crossover cable 13:38:49 then used rsync 13:38:52 I used to use ethernet over firewire back when my desktop only had 100 mbit connection. Because firewire allowed 400 mbit 13:38:52 *that sort of amount of data 13:39:15 now I have gbit ethernet on both my desktop and laptop 13:39:24 so generally the disk speed in the laptop is the bottleneck 13:40:16 -!- hagb4rd has quit (Quit: Nettalk6 - www.ntalk.de). 13:41:18 hm PSU at the bottom seems to be getting more and more popular in high end chassis these days. 13:41:29 what is actually the point of PSU at the bottom 13:42:16 my computer has that, but it just messes up with the cables as far as I'm concerned (since the PSU uses flat cables with connectors on the side of the cable. And it is designed for mounting at the top. 13:42:54 ais523: maybe you would know what the advantages of PSU at the bottom of the computer case is? 13:42:59 Vorpal: it's so that when YouTube upgrade their video service, it still works if you turn the entire computer case upside-down 13:43:08 ... 13:43:08 which gives even better performance than just inverting the monitor 13:43:11 what? 13:43:26 inverting monitor? I don't get the joke. 13:43:42 it was YouTube's April Fools thing this year (or maybe last year?) 13:43:47 ah 13:43:50 where they turned the entire site layout upside-down, also the videos 13:43:55 ah okay 13:43:59 ais523: but seriously, any idea? 13:44:04 on the basis that they'd discovered that videos looked better if you turned the monitor upside-down 13:44:08 and no, not offhand 13:44:19 it wouldn't surprise me if it was something to do with cooling, but I don't see how it would help offhand 13:44:23 *if it were 13:45:32 ah 13:47:53 1.1.1 The droll business is, whether you be convinced active it, the cares you should get are each familiar meaning! How does it effort? Besides solely, individual moves any earful most you, possibly on-line or conceivably eve via any scrap send that you tossed in the crank without eve trigger-happy it up. 13:49:28 Sounds fungotty. 13:49:28 fizzie: ' cheers,' said lu-tze. 13:50:13 -!- itidus21 has left ("Leaving"). 13:50:26 -!- kallisti has joined. 13:50:28 fungot: go on 13:50:28 ais523: " no we ain't," said esk. " granny always says that to women, for the arms, two twigs. 13:50:38 -!- kallisti has quit (Changing host). 13:50:38 -!- kallisti has joined. 13:50:44 `word 25 13:50:51 eszykkakemed obatleue raieffenisteinges paghters ac gan roleyum appard lvatoulphonicae hendfelletharbrogg ratereur bee tocitz bratous aeurrosly pred pic sten reanclan tlnhek bie un dodus blertrigentmousalgilemanatimend ro 13:51:44 -!- itidus21 has joined. 13:52:16 huh, for some reason youtube is saying "missing plugin" for every video. 13:52:34 kallisti: perhaps you don't have a Flash plugin? 13:52:44 I most definitely do 13:52:46 "bee" is a real word, out of that list 13:52:53 I just restarted after some updates though. 13:53:09 and quite a few of them sound like they could be real words 13:53:11 ais523: it happens. especially with short words. only so many possibilities in a markov model. 13:53:14 ais523: good 13:53:15 indeed 13:53:19 `word 25 13:53:22 fuermarabev as gcommenfer dumallyley wychafr mam zat tb tuproassitara man apitenins jurg autory cler hydrouzhdu aflutz flits ine boni met soirodestsovegeble bolia formagged lopui ch 13:53:52 aflutz :) 13:54:01 * kallisti has flits and aflutz 13:54:08 man 13:54:16 I've seen people seriously attempt to use "boni" as the plural of "bonus" 13:54:22 .. 13:54:22 "as" is also a word, as is "man" 13:54:27 met 13:54:32 "pic" and "pred" out of that first list were also rather common abbreviations. 13:54:36 actually, "formagged" is possibly the best nonword there 13:54:59 no soirodestsovegeble? 13:55:14 ais523: at least boni is correct latin, i think 13:55:33 unlike some other examples 13:55:47 yep, I think so 13:55:59 formagged obviously means turned into cheese 13:56:47 heh 13:57:03 `word 50 13:57:05 monsumizehutand senselees atticheler jnaal tans ber ozamoloupne dovolowx hici unthiries subs auroo zu pribacc obikanba vity iscu lammenleres ken meduratctrae wd troasakic venderpugaryszus alierectracist mesi bouraps bbizosinizaria fandombori obed uienyonoseranottinscrenzy wander rheetess syraibia te tro ruardentiminesta hus sesined thl metion sch brica inuce howmrech narremietty baarimptinfon bnue boletadvaligny cluvra kancr 13:57:18 monsumizehutand -- ancient Aztech ruler 13:57:30 "fandombori" 13:57:37 heh 13:57:56 What a senselees list of words. 13:58:43 Is an 'alierectracist' a person who won't submit to being probed by extraterrestials, or what? 13:58:52 ....... 13:58:55 I think so? 13:59:41 `word 1 13:59:43 urganaidonoropedeechl 14:00:26 :t zip`ap`tail 14:00:29 forall b. [b] -> [(b, b)] 14:01:06 err-gah-nah-ee-doh-noh-ropey-dee-ch-L 14:01:32 yep 14:01:46 god thats difficult 14:01:47 * kallisti has gotten the hang of pronouncing these things. 14:01:56 a little creative discretion is allowed 14:02:09 in english you must remember that pronunciation does not follow from spelling :P 14:02:10 as it's "pseudoEnglish" in nature 14:02:19 which already has a wide variety of different ways to pronounce combinations of letters 14:02:49 itidus21: no it does, it's just based on a wide variety of linguistic influenced 14:03:01 *influences 14:03:20 it seems to me like it simply borrows several pronunciation systems 14:03:45 everything is "borrowed" and slightly reinvented in natural language 14:03:47 and, well, a lot of things. 14:05:12 Speech synthesis thingies often have (in addition to a large pronunciation dictionary) some sort of a rule-based device to generate plausible phonemes for OOV words. 14:05:22 perhaps the idea of english is to acquire the shibbaleths of their eneies 14:05:34 ^enemies 14:05:58 music, art, religions, holidays, sports, mythologies 14:06:17 someones probably already done a phd on that possibility i suggested and found it's a dead-end 14:06:20 echo 'monsumizehutand senselees ... cluvra kancr' | festival --tts # the best babble ever. 14:06:23 if you look at all of these things you'll see a (mostly) linear progression of linear through history. 14:06:38 `run festival --help 14:06:40 bash: festival: command not found 14:06:55 It's so fast I can't make anything out of it, and I don't quite recall how to control speaking speed. 14:07:20 `word 5 14:07:22 parne euwessoly coms coaulanquicolve tra 14:07:25 yes a linear progression of linear. 14:07:32 itidus21: 25 is a good number 14:07:36 bound to find something interesting. 14:07:48 Sometimes it heuristicizes into pronouncing things as lettersims; like "sch" and "wd" it does like that. 14:08:30 The prosody for that "sentence" is... somewhat arbitrary too. 14:08:54 `word 30 14:08:57 imerapposan ccyptyrs strcd plitaspet plation cxlere decocs vie chya schth sap cont eldenernefl chron diss ophyphofe symo imbee hoyoj dro einee dion eouciircloromplam wehenss remed nifteng jeeppertia sa sch dimely 14:09:14 strcd sounds like a string.h function 14:09:14 There's a short break between uienyonoseranottinscrenzy and wander, like a comma, for some reason. 14:09:35 schth -- best word 14:11:05 eldenernefl - el-den-nur-neh-ful 14:11:16 fizzie: unfortunately I believe googles data contains roman numerals 14:11:29 my pseudo-intelligence leads me to say, ("ais523") = "523". ("523") = '5', '2', '3'. ('5') = 5. ('2') = 2. ('3') = 3. 5+2+3 = 10 14:11:29 which sometimes leads to ridiculous things like triple i's 14:11:38 `word 10 14:11:40 tris bayions au minis tolotti ital coneu via tchliplanosleociot exillesta 14:12:19 Festival pronounces eldenernefl with pretty much a silent "den"; el-ner-neh-ful. 14:12:33 i quite like tchliplanosleociot 14:12:33 Sorry, 'de'. 14:12:59 as for exillesta... thats just a kick ass word 14:12:59 itidus21: it's squishy in my mouth 14:13:11 In fact, I can't hear any difference between elnernefl and eldenernefl. Maybe the latter is just archaic spelling for the former? 14:13:29 most likely... 14:14:34 `word 14:14:37 sen 14:14:51 That's valid Finnish. 14:15:13 And I suppose quite a few other languages too. 14:15:30 Japanese and Swedish, at least. 14:15:37 yay i got 23000 euros today 14:15:47 Don't spend it all at once now. 14:15:47 `word 14:15:49 progyribure 14:15:54 oklopol: grant? 14:15:56 yes 14:16:04 but that sounds less cool 14:16:13 oklopol: you have to like spend it a certain way don't you 14:16:17 ? 14:16:26 it's for living expenses 14:16:30 ah cool. 14:16:41 -!- Slereah_ has joined. 14:17:14 -!- Slereah has quit (Ping timeout: 252 seconds). 14:17:14 oklopol: you should buy a nice TV and a PS3 and play Demons' Souls 14:17:27 And maybe some food too. 14:17:37 If there's any money left after the essentials. 14:17:44 Like those mentioned above. 14:17:45 oklopol: my grant comes more gradually 14:17:46 eh 23000 euros is plenty for food. poor American white trash could live off of that for a year at least. 14:17:54 I just look at my bank account and find that there's more money in there than I remember 14:18:09 ais523: mine too, probably. i'm just polishing the facts a bit. 14:18:25 i have to live off it for a year. it's actually just my normal salary. 14:18:35 My grant just somehow goes into some (rather large, I think maybe 80% or so?) percentage of my regular monthly salary. 14:18:38 perhaps even slightly less since i just got a raise. 14:18:40 I don't really know the details. 14:18:50 yeah maybe mine does too 14:18:55 it's from väisälä, you may know it 14:19:01 oklopol: you could probably get like, what, 92 high-class escorts (read: prostitutes)? 14:19:14 i don't know where to get those in finland 14:19:28 Also the rest of the paycheck comes from some place, but I don't know where. Maybe it's the... department? I don't know, the bureaucracy is confusing. 14:19:51 kallisti: I'm amused that you know the price that accurately 14:19:55 all of you guys have grants? weird. 14:20:01 I just get paid. 14:20:13 fizzie: my grant and paycheck are added separately 14:20:18 my paycheck currently comes from the "project" of my supervisor (in math, project just means... nothing) 14:20:18 ais523: well it was an estimate. 14:20:19 i don't have an income. i get 'supported' by my family 14:20:25 amusingly, the paycheck has many times more bureaucracy 14:20:31 not to complain. i do nothing to deserve an income 14:20:43 itidus21: that was the case for me before I got a job, too 14:20:52 and I imagine it's the case for pretty much everyone too young to have a job 14:21:24 * kallisti is currently supported by his family /and/ making money. 14:21:25 im not on a dole either.. and im relatively happy really 14:21:28 kallisti: up to now, i just had a paycheck, but my supervisor asked me to apply for a grant since he ran out of money for next year because he desperately needed to fish a good student for himself. 14:21:33 at least for now. I intend to move out sometime next year. 14:21:46 oklopol: academia is weird. 14:22:08 since the student might have gone with a different professor if he'd waited. 14:22:15 i chat with a lot of people more intelligent than me 14:22:36 it's not good for my relative sense of intelligence :D 14:22:52 itidus21: I think most people do, actually; people tend to gravitate towards people with similar levels of intelligence 14:22:56 ais523: Well... I have a regular salary selected from our salary tables, but I have this four-year "Doctoral Programme" position from http://www.cs.helsinki.fi/hecse/ too, so that money gets somehow funneled in as a "funding source" into whatever percentage of my salary it happens to cover, and the department makes up for the difference from some other project/funding/whatever. 14:23:02 and thus will tend to meet people more intelligent than themselves, no matter how intelligent they are 14:23:08 fizzie: hmm, how complex 14:23:10 ais523: oh.. thats a positive 14:23:22 * kallisti makes terrible web apps for a living. 14:23:24 i gravitated towards #esoteric 14:23:27 (exception: the most intelligent few people in the world) 14:23:52 itidus21: I'm certainly not as intelligent as most people on this channel, but I still consider myself intelligent. do you know why? 14:23:56 because: real world 14:23:58 people are stupid 14:24:05 fizzie: it's much simpler for me; I have a 75% part-time PhD (which I'm being funded by the department to work on, at the suggestion of my supervisor), and am paid directly for a 25% part time teaching job 14:24:29 kallisti: if i had to judge, my question would be, do you realize there are inherent contradictions to such statements? :D 14:24:31 I've considered going getting a Phd at some point, but... I don't know if it's worth it. 14:24:50 itidus21: there aren't 14:25:17 taking oneself too seriously leads to a holistic decrease in intelligence 14:25:25 hm? 14:25:35 * kallisti takes everything SERIOUSLY AAAAAAH 14:25:38 .. :) 14:26:35 i mean... eh.. nevermind.. i need to be in another mood for that silly topic of mine 14:26:43 Also Condor, this "let's use our idle desktops as a computing grid" thingie, has a confusing -help for some commands: 14:26:44 ais523: oklopol: fizzie: the problem with me getting a Ph.D in Computer Science is that I don't really think I will be very /good/ at research. 14:26:44 $ condor_hold -help |& grep addr 14:26:44 -addr Connect directly to the given "sinful string" 14:26:48 or pretending that I'm researching. 14:27:19 What makes an "ip:port" sinful is unclear to me. Maybe it refers to struct sin_addr. 14:27:21 hmm, that could be a problem 14:27:33 it helps to have an obvious thing to be working on already 14:27:38 i thought i might just be a good student and a horrible researcher, but this seems to be going well 14:27:40 like the hardware compiler, in my case 14:27:44 i have something like 50 theorems now 14:27:51 wow 14:27:56 ais523: I work on side projects sure, but nothing that's groundbreaking 14:28:02 meanwhile, I have 3 papers, which contain an average of less than one theorem each 14:28:03 and three publications 14:28:08 I wouldn't write a paper on how I made a program that randomly generates words. :P 14:28:15 indeed, it's known techniques 14:28:21 oklopol: I hope you are naming them with a numbering scheme, so that you can have people referring to "oklopol's 37th theorem". 14:28:25 the general point of PhDs is that you're expanding the boundaries of knowledge 14:28:30 right. 14:28:43 oklopol is proving new results, whereas what I'm doing is basically programming 14:28:52 kallisti: uhhh.... this is a difficult topic. in the end intelligence as a measurement crumbles under it's own weight. 14:28:53 it's coming up with new algorithms, I guess, looked at from the mathematical view 14:29:12 and when we prove theorems, it's either to prove that they produce the right results, or that they always terminate 14:29:12 i don't think many of them will ever be referred to, this is the number of theorems that aren't trivial to prove, the number of useful results is way less. 14:29:21 ais523: maybe as I focus on new side-projects I'll come across something somewhat new. 14:29:26 so i guess what i am saying is there is a mild sarcasm for me whenever i use that word intelligent 14:29:27 *something 14:29:28 Programming can be a research topic 14:29:38 But results tend to become dated 14:29:44 kallisti: we couldn't believe that what we were doing was new, in some cases, but it turned out that it was 14:30:17 we have some computational results, decidability and semidecidability stuff 14:30:25 ais523: maybe I should focus on what interests me outside of computing. I have a pretty strong grasp of signal processing as it relates to music. I could probably find something new there. 14:30:26 kallisti: uhhh.. like.. you know.. theres tangible intelligence and intangible intelligence *pulls hair out* 14:30:40 mainly on zero entropy sofic shifts, since we're trying to get to a CS conference that emphasizes this 14:31:21 itidus21: I generally don't think that the many different kinds of intelligence are truly quantifiable. IQ has statistical importance but there are other ways to think of what intelligence means. 14:31:27 maybe not 50, the number was 37 last i checked, but that was after summer and we have 3 new results this week i think 14:31:34 so it might be about 50 but dunno 14:31:39 "Zero entropy sofic shift" sounds like something straight out of some new age crackpot website. 14:31:40 Phantom_Hoover: You have 12 new messages. '/msg lambdabot @messages' to read them. 14:31:50 ;_; the topic of intelligence itself is indeed one that requires intelligence to traverse 14:31:57 i'm counting mine and my colleague's, there's a couple that he proved and about half are joint work 14:32:07 ais523: in particular I've been considering that there are a number of combinators that you can apply to form rhythmic patterns. It may be under some existing generalization though. 14:32:58 @tell elliott FFS, find a way of messaging me that doesn't overflow so easily. 14:32:59 Consider it noted. 14:33:09 the simplest would be parallel and serial combinations. 14:33:17 but perhaps there are more. 14:33:19 a sofic shift is just a shift space defined by a labeled directed graph, they are the closure under factor maps of subshifts of finite type, which on the other hand are exactly the subshifts defined by a clopen set 14:33:25 in a natural sense 14:33:52 And time has four corners 14:33:55 and zero entropy means the topological entropy of the dynamical system where the left shift is the dynamics 14:34:31 Phantom_Hoover: does lambdabot overflow? 14:34:37 when messaging 14:34:41 but there's a nice characterization for these 14:35:10 i have no idea who i'm talking to 14:35:11 :D 14:35:33 hmm also you can shrink and expand rhythmic patterns. 14:36:03 so you could combine two rhythms serial (basically end to end) and also have the result be the same duration as the originally, basically doubling the speed of both. 14:36:49 but I'm pretty sure these notions of serial and parallel exist elsewhere. If you model rhythms as a linked list, then the serial combinator is just (++) in Haskell. 14:37:16 If you work hard on this, you might successfully get hired by whoever puts those bland pop tunes on the radios 14:37:25 ha 14:37:33 to algorithmically generate pop? sounds... uh... good 14:37:50 * kallisti would change pop forever by NOT USING FUCKING 4/4 TIME SIGNATURE FOR EVERYTHING 14:37:52 Sounds profitable. And Orwellian 14:38:19 yeah why is 4/4 so popular. Nothing wrong with 2/4, 3/4, 2/3 and so on IMO 14:38:27 Taylor Swift's latest single will be 5/4 7/4 5/4 14:38:33 They don't even use time signatures any more! They steal time signatures from older tunes. 14:38:46 (well, just 5-7-5, the denominator is irrelevant to the actual meter) 14:38:47 Look at breakbeats 14:39:26 Fibonacci time. 1-1-2-3-5 14:39:49 well you could have like, i,j-Fibonacci time. 14:40:04 i and j being the interval you want to use. 14:40:19 or something. :P 14:40:43 You could use nothing but concatenated permutations of a set of tones. Oh wait. 14:41:01 oklopol: i think sofic shift spaces are like the two-sided infinite generalization of regular languages 14:41:29 The Tool song "Lateralus" has one part where the syllables in each measure of lyrics follow the pattern: 1 1 2 3 5 8 5 3 14:41:45 you have a finite automaton, but it never stops nor has it ever started 14:42:28 yeah they are exactly the subshifts whose language is regular, and a regular language that's factor closed and extendable gives a sofic shift 14:42:41 Jafet: I think I'll use subsequences selected from an infinite continuum of sinusoids. 14:42:55 oerjan: You mean a duracell-powered finite automaton? 14:42:59 but my definition makes more sense in symbolic dynamics 14:43:02 fizzie: pretty much 14:43:03 imo 14:43:10 that's how it's mostly used 14:43:25 that it's the closure of sft's under factors 14:44:38 Festival pronounces "Phantom_Hoover" as "phantom-underscore-hoover". 14:45:51 isn't that how everyone pronounces Phantom_Hoover? 14:46:01 shockingly, no 14:46:12 I pronounce it as "Wally" 14:46:31 -!- oerjan has quit (Quit: ORLY?). 14:46:36 Jafet: I pronounce "Wally" as "[;[8]]&*]6" 14:46:51 "Some pronounce it 'asshole'." No, I mean, I wouldn't spell out the underscore if someone told me to read these IRC logs out loud. 14:47:12 !perl [;[8]]&*]6 14:47:13 Number found where operator expected at /tmp/input.7581 line 1, near "*]6" 14:47:20 !perl [;[8]]&*] 14:47:20 syntax error at /tmp/input.7641 line 1, near "[;" 14:47:24 !perl [[8]]&*] 14:47:28 there we go. 14:48:24 it was /almost/ valid Perl. :P 14:48:55 !perl print *] 14:48:55 ​*main::] 14:49:20 !perl print *]{SCALAR} 14:49:21 SCALAR(0x7f7155eaebf8) 14:49:28 !perl print ${*]{SCALAR} 14:49:28 Missing right curly or square bracket at /tmp/input.7956 line 1, at end of line 14:49:29 er 14:49:56 !perl print ${*]{SCALAR}} 14:49:56 5.010001 14:50:43 fizzie: ais523: have you ever found a use for typeglobs now that Perl 5 introduced refs? 14:51:11 kallisti: injecting variables into other package's symbol tables; this makes the most sense in a library designed to generate code 14:51:30 !perl print []|*] 14:51:30 ksi::}7f625ec6ed48) 14:51:55 I haven't, but I haven't been doing very "deep" Perl either. 14:51:56 e.g. you write a library which adds extra functions to packages that reference it 14:52:04 fizzie: wait, how does that work? 14:52:17 It just bitwise-ors the bytes. 14:52:19 aha, it's a reference bitwise-ored with a symbol table entry 14:52:30 but, err, what? 14:52:31 Of "*main::]" and "ARRAY(0xwhatever)". 14:52:41 !perl print [[8]]&*] 14:52:48 >_> 14:52:57 !perl print [[8]]|*] 14:52:57 ksi::}7fdd767b69e8) 14:53:12 The anded version starts with a null byte, sadly, so EgoBot won't print it. 14:53:24 stupid C. 14:54:36 -!- copumpkin has quit (Quit: Leaving...). 14:54:57 fizzie: is there anyway to enfroce a bit width of 8 on Perl scalar values? 14:55:16 I'd like to play around with a simple 8-bit synthesis in Perl. I could just use C but.... why not use Perl instead if I can. :P 14:55:46 -!- copumpkin has joined. 14:55:50 Don't know; they *could've* made "use integer" take a bit-width (like "use integer 8;"), but it doesn't. 14:56:02 -!- derdon has joined. 14:56:11 You can just &0xff everywhere, of course. 14:57:24 that's a possibility 14:57:46 What I think is a bit weird is how "use integer" makes the always-integral bitwise things (&, |, ^, ~, <<, >>) use signed integers instead of the unsigned ones they usually use. 14:58:26 !perl $a = ~0; { use integer; $b = ~0; } print "a $a, b $b"; 14:58:26 a 18446744073709551615, b -1 14:58:53 I suppose it makes some amount of sense, I just think unsigned integers are somehow more... integery. 14:59:00 Okay, *that* probably doesn't make any sense. 14:59:03 lol 14:59:19 fizzie and his arbitrary notions of integeriness 14:59:44 mayb "use integer" is like saying "use /all/ the integers" 15:00:22 even those negative ones. 15:00:46 what I find strange about bitwise operators in high-level languages is that normally those languages don't specify anything about the bit composition of values. 15:00:53 it just seems out of place. 15:02:51 negative integers? someone still believes in those?? :D 15:03:05 ..? 15:03:15 by the way, that article that supposedly proved peano axioms are inconsistent, it was retracted and there's no trace of it anywhere :D 15:03:25 heh. 15:03:39 the revelation was simply too shocking 15:03:50 it had to be concealed. 15:03:51 kallisti: if you multiply two negative integers, you get a positive one, that's fucking ridiculous 15:03:57 how can anyone think that's true 15:04:02 ..... 15:04:08 therefore there are no negative integers 15:04:14 bahaha 15:04:24 > bitSize (1 :: Integer) 15:04:25 I mean it would make sense if the result continued to be negative.. 15:04:25 *Exception: Data.Bits.bitSize(Integer) 15:04:29 that was actually an argument used when these were introduced 15:04:29 Aw, it has no size. :/ 15:04:56 but then who knows what would happen when you tried to multiply a negative and a positive.. 15:05:00 if negative * negative = negative 15:05:08 + * - = ..... +-? 15:05:13 there was really no concept of a mathematical object back then 15:05:23 yep 15:05:25 another proof 15:05:29 that they don't exist 15:05:35 well no you could totally do it that way 15:05:42 so that + * - = +- 15:05:44 anyhow, i have to go to a party 8Z 15:05:48 :D 15:05:55 bye 15:05:59 oklopol THE SOCIALITE 15:06:01 bye. 15:06:33 oklopol the SOCIALIST 15:06:50 A "socialite" is the no-sugar version of a "socialist". 15:06:56 fizzie: ais523: conjecture with me what would happen if you made multiplication of a positive number and a negative number have two possible results 15:07:45 + * - * - 15:08:26 anyhow, i have to go to a party 8Z 15:08:27 is 15:08:32 is that a running man smiley 15:08:40 well, it's still commutative I think.. 15:09:14 Can smileys be commutative? 15:09:21 if they're running yes 15:09:23 they're commuting somewhere 15:09:30 but that's not what I was talking about obviously :P 15:10:34 er.... no maybe it's not commutative 15:10:56 2 * -3 * -4 evaluating left to right 15:11:05 +-6 * -4 15:11:18 that's still just... +-24 15:11:44 Are you sure it's not +--24. 15:11:46 even though there are three results two of them are the same 15:12:07 but if you did it that way then it would result in it being non-comutative I think 15:12:11 because then 15:12:20 -4 * -3 * 2 15:12:24 I prefer the Copenhagen interpretation of arithmetic 15:12:37 evaluating left to right, would produce just +-24 15:12:50 instead of +--24 15:12:54 UNLESS 15:12:58 - * - = -- 15:13:03 but... okay nevermind 15:13:09 screw this direction of thought 15:13:17 you can only have +, -, and +- 15:13:18 LO 15:13:22 s/LO/:P/ 15:13:45 Lo, there are many planets in the archipelago of worlds. 15:13:50 Low colon pee 15:14:33 help I affiliate myself with madmen. 15:15:03 s/affiliate myself/confederate/ 15:15:09 AWWW YEAH +1 WORD CHOICE 15:15:45 inb4 "not +-1 word choice" 15:15:50 or something similar 15:15:52 fizzie, and, in turn, many archipelagos on the planets. 15:16:11 Phantom_Hoover: dude what if there are universe archipelagos 15:16:25 Univarchipelagos. 15:16:45 and... like, univunivarchipelago 15:17:03 WHAT IF EVERYTHING SMALL IS A SMALL VERSION OF SOMETHING BIG. 15:17:49 There's an island in the sea around here, on which there's a lake, in which there's a small islet, on which there's a puddle. (Didn't someone make a comic out of this already?) 15:18:23 That island must make great target practice for bomber pilots 15:18:25 kallisti, careful now, too much of that kind of thinking and you're ruining Minecraft's terrain gen. 15:19:07 #esoteric is an peoplarchipelago 15:19:23 and people are CELL ARCHIPELAGOS 15:19:24 zomg 15:19:27 zaaaaah 15:25:26 are there any four-signed number systems out there? 15:28:55 I think a four-signed number would mess up 1 as the multiplicative identity 15:29:01 0 would still be the additive identity though 15:30:48 unless you just kept 1 as the multiplicative identity and made multiplication asymmetric 15:42:34 -!- myndzi has quit (Ping timeout: 245 seconds). 15:46:35 -!- MSleep has joined. 15:56:12 -!- MSleep has changed nick to MDude. 16:05:53 -!- calamari has joined. 16:06:37 Wolfram blog on the fold function: "It shows unusual mastery of functional programming constructs to achieve a beautiful graphic result." 16:09:12 -!- Ngevd has joined. 16:15:04 Hello! 16:17:19 hi 16:17:46 ais523, you're in my top two people who are, or I think are, in Birmingham 16:17:51 Holloo. 16:18:10 Ngevd: there's quite a lot of evidence that I'm usually in Birmingham 16:18:14 Phantom_Hoover, you're my favourite person in Edinburgh. By a long way 16:18:19 although that doesn't imply I'm in Birmingham right now, it makes it quite a bit more likely 16:18:46 I can only conclude that you either a) hate someone else in Edinburgh or b) aren't very good at being pedantic. 16:18:59 I don't know many Edinburghians 16:19:07 Pretty much only you 16:19:57 In fact, only you 16:19:57 And possibly Alexander McCall Smith 16:20:33 I recognise the name, and I'm not entirely sure it's as an author. 16:21:09 It's probably just as an author, actually. 16:21:43 Yes 16:21:44 Did the 44 Scotland Street and Number 1 Lady's Decective Agency books 16:21:47 two of those names are kind-of familiar to me 16:22:17 I would be extremely surprised if neither 'Alexander' nor 'Smith' were familiar to you. 16:23:21 -!- itidus21 has quit (Ping timeout: 252 seconds). 16:23:36 My mum met him once when he came to Hexham 16:23:36 ais523, the other person who lives in Birmingham who I have heard of is possibly the author of Gunnerkrigg Court 16:23:50 hmm 16:24:01 I think there are quite a lot of famous people from Birmingham, because it's quite large 16:24:11 on the other hand, they're not generally famous /for/ being from Birmingham 16:24:42 hmmm, I'm confused. 16:24:53 I change the source of mueval to import some modules 16:24:59 and then I run the build script to install it 16:25:04 and.... nothing changed? 16:25:22 oh possibly the wrong mueval. 16:26:49 hmmm, no 16:28:31 -!- Ngevd has quit (Ping timeout: 248 seconds). 16:28:40 ah there we go 16:28:46 had to do Setup copy instead of install for some reason? 16:35:18 -!- Ngevd has joined. 17:00:36 -!- Klisz has joined. 17:04:52 -!- Ngevd has quit (Quit: Leaving). 17:25:56 -!- monqy has joined. 17:28:46 -!- elliott_ has changed nick to elliott. 17:28:49 -!- elliott has quit (Changing host). 17:28:49 -!- elliott has joined. 17:36:48 09:30:46: elliott_: is how you prevent Perl from interpolating in a shell command 17:36:48 09:31:06: if you're using system or whatever else you would just pass a single quoted string obviously. 17:36:51 kallisti: that is not the problem 17:37:01 kallisti: how do i pass args in a list like python subprocess 17:38:34 @ping 17:38:35 pong 17:39:18 elliott: um... 17:39:26 like how? 17:39:28 I don't remember subprocess. 17:39:39 but 17:39:39 what is the hardest language to debug? 17:39:44 'ls', '-l', 'filename with spaces' 17:39:50 this is necessary to avoid shell injection 17:39:55 I think you want system? 17:40:01 or exec, depending on if you want to wait or not. 17:40:11 system waits for the child to finish, exec doesn't 17:40:17 quintopia: Malbolge? 17:40:29 yes i guess i agree 17:40:29 a good IDE would help, though 17:40:33 yeah 17:40:46 ais523: how do I do perl's `...` but with an argv instead of a string or do i have to emulaet it myself 17:40:46 showing both original and normalized views of memory at once, and probably encryption chains too 17:40:51 kallisti: it's not system 17:41:02 it can be done with exec but that's a painfully low-level interface 17:41:09 elliott: there's no operator for doing that straight off, I don't think 17:41:16 there's almost certainly a library, probably a standard one 17:41:24 or, hmm, what about piped open? 17:41:32 I've never used it, but I think that's what you need 17:41:33 doesn't that look like '|foo bar' 17:41:37 which is not an argv 17:41:57 elliott: there's a four-or-more arg version of open 17:42:09 which does take an argv as the extra arguments at the end 17:42:33 ah 17:42:35 oh you want to pass the argv directly.... 17:42:39 open my $fh, '-|', 'somecommand', @argv; 17:42:52 then $fh is its stdout and you can just read it via the normal means 17:42:57 right 17:43:01 that sounds doable 17:45:06 * elliott does the regular Chrome restart 17:45:26 hmmm, when would you use CReal? 17:45:40 when you want computable reals 17:45:51 if you don't know whether you do or not, you don't 17:46:04 so it would basically be to avoid floating point errors? 17:46:09 no 17:46:13 that's Rational 17:46:37 so it would basically be to avoid floating point errors when you're not dealing with rational numbers? 17:46:48 you are dealing with rational numbers 17:46:51 floats are an approximation of rationals 17:46:55 kallisti: if it's referring to all computable reals, it only work properly on irrational numbers 17:46:58 *works 17:47:07 ais523: huh? 17:47:10 computable reals that happen to be equal to integers can't actually be converted to decimal expansion 17:47:17 you can still compute with them 17:47:24 ais523: err, I don't know what you're talking about 17:47:27 but you can't do < or > on computable reals unless the numbers happen to actually be different 17:47:29 CReal can output approximate decimals just fine 17:47:33 elliott: what about pi? 17:47:37 that's got nothing to do with decimal expansion 17:47:39 kallisti: if your question involves "avoid", "errors", you don't want CReal 17:47:40 that's irrational and approximated by float. 17:47:47 kallisti: unless you're a mathematician, you don't want CReal 17:47:48 elliott: I'm thinking of infinite-precision real numbres 17:47:52 *numbers 17:47:54 or a theoretical CSist 17:48:02 can I pretend to be one of those? 17:48:09 ais523: nobody uses the infinite-digit representation 17:48:10 kallisti: no. 17:48:10 I mean 17:48:14 I actually have no use case at the moment. 17:48:19 kallisti: here's some things CReal can't do: 17:48:19 I was merely curious 17:48:22 kallisti: terminate when you do (a == a) 17:48:30 kallisti: terminate when you do (a > a) 17:48:34 elliott: I've been to seminars where infinite-digit representation was involved 17:48:35 kallisti: terminate when you do (a < a) 17:48:49 and the main difficulty, as you're mentioning there, is comparing two numbers that happen to be equal 17:48:55 ais523: the main difficulty is arithmetic 17:48:57 elliott: you mean /always/ terminate right? 17:49:01 that's just a property of the computable reals 17:49:01 > let a = 1.23 :: CReal in a == a 17:49:01 kallisti: no 17:49:02 True 17:49:06 Deewiant: that's a cheat 17:49:09 kallisti: no, you get an infinite loop whenever you compare a number to itself 17:49:14 CReal's (==) instance just does it to an approximation 17:49:20 also 17:49:27 elliott: oh, I guess it's approximating with the decimal expansions too? 17:49:28 I'm not sure that's Few Digits' CReal 17:49:34 as in, there's always a chance that the last digit is wrong? 17:49:37 that makes sense 17:50:03 ais523: well, um, you can always take a CReal to within a given precision 17:50:07 i.e. "pi to 0.0000001" 17:50:28 elliott: yes, and your result will have more digits than you asked for, with the last potentially being wrong 17:50:36 you can't say "pi to 8 decimal places", though 17:50:52 elliott: they should be renamed to "undecidably equal reals" 17:51:00 well, you can because pi is irrational, but you couldn't if there was a chance that the number was actually accurate to 8 decimal places 17:51:10 kallisti: you act like you were expecting computable reals to be useful for computation 17:51:35 yeah.. 17:51:41 > pi ::CReal 17:51:42 3.1415926535897932384626433832795028841972 17:51:45 delicious cereal 17:52:07 http://sourcereal.com/ 17:52:15 elliott: such a good site 17:52:40 elliott: so CReal isn't good for like... high precision math because it doesn't terminate often? 17:52:56 kallisti: even if computable reals were useful for computation, Few Digits is incredibly slow 17:53:02 as are all the implementatinos, because... they're not useful for computation 17:53:13 it's theory. unless you're a theorist, you don't care. 17:53:22 "Few Digits is not fast. Few Digits is part of my Ph. D. research. My goal is to implement an exact real arithmetic package in Coq that is proven correct (with respect to C-CoRN) and is sufficiently fast. The goal is to be fast enough to prove the inequalities required by Hales’s proof of Kepler’s Conjecture." 17:54:03 so I only want to use CReal if I'm proving the inequalities required by Hale's proof of Kepler's Conjecture 17:54:07 got it. 17:54:27 elliott: I've been to a seminar that was using computable reals to calculate pi to infinitely many decimal places 17:54:34 and we've been using the resulting program as a test of the hardware compiler 17:54:46 on the basis that it's the sort of program that makes no sense to typical hardware compilers 17:54:48 ais523: that's, err, noteworthy enough for a seminar? 17:54:55 I can do that in five lines of Haskell 17:55:07 elliott: it was about computable reals, and just an example 17:55:11 fair enough 17:55:43 I ran the resulting VHDL for almost a week in a simulator on my laptop, it output the first 4 balanced binary digits 17:55:53 @quote pi 17:55:54 quicksilver says: overlapping actually shatters the language into tiny inconsistent pieces, and incoherent files off the edges of the pieces so they don't even fit together any more. 17:55:56 erm 17:55:58 `quote pi 17:56:00 9) Lil`Cube: you had cavity searches? not yet trying to thou, just so I can check it off on my list of things to expirence \ 14) First, invent the direct mind-computer interface. Second, you know the rest. \ 15) IN AN ALTERNATE UNIVERSE: First, invent the direct mind-computer interface. Second, learn the rest with your NEW MIND-COMPUTER INTERFACE. \ 30) 17:56:02 gah 17:56:04 `quote digit 17:56:06 427) meanwhile, I've been running a program for over 24 hours (getting close to 48 now) which is calculating digits of pi, in binary so far, it has found four digits I hope it will find the fifth some time this week \ 513) I actually had a Neopets account. I later gained a second digit in my age. \ 643) sadhu: it's been said that boole is the crowning jewel perched 17:56:09 427 17:56:18 yep 17:56:22 I knew it was in there somewhere 17:57:03 part of the slowness is all the interpretation between paradigms 17:58:10 simulating FPGA behaviour on a CPU is slow, as they're rather different arches; doing single-threaded recursion on an FPGA isn't really faster than doing it on a CPU; and there was no memoization, in a program designed to run in a call-by-need language 17:58:22 lol, this spam is from "Google Incorporation. (info.google@msn.com)" 17:58:29 (OK, so Haskell isn't /technically required/ call-by-need, but what sane interp doesn't implement it like that?) 17:58:36 Gregor: that's pretty good 17:58:54 (OK, so Haskell isn't /technically required/ call-by-need, but what sane interp doesn't implement it like that?) 17:59:00 ais523: speculative evaluation is pretty sane! 17:59:25 elliott: hmm, what's that, and how is it different from call-by-need? 17:59:43 ais523: it isn't a complete strategy itself, it just refers to evaluating thunks even when they're not demanded 17:59:56 and looking away sheepishly and pretending nothing happened if it ends up _|_ 18:00:09 the idea is that you predict which thunks are going to be used in the future and evaluate them ahead of time in another thread 18:00:11 thus saving time later on 18:00:21 ah, I see 18:00:55 presumably that ends up somewhat concurrent if done well? 18:01:15 well, you could even not do it in a separate thread 18:01:20 you'd just have to bound the number of steps you take 18:01:34 to make sure you don't accidentally make a terminating program with a non-terminating subterm less terminating than you'd like 18:02:23 elliott: if it's not in a separate thread, there's no benefit to doing it out of order, is there? 18:03:44 elliott: what was your Perl question about, btw? 18:03:56 ais523: well, you could change [high activity][long pause][result] into [halved activity][result] 18:04:07 by using every other cycle to reduce another complicated thunk 18:04:13 ah, hmm 18:04:14 that will be forced afterwards 18:04:20 you don't get the result any faster, but it might look nicer to the user? 18:04:29 ais523: well, incremental results are a useful thing :) 18:04:40 ais523: you wouldn't have wanted to get no output until the program found every digit of pi, right? 18:04:58 but it wouldn't give you any output any earlier 18:05:06 it'd just delay some of the outputs 18:05:11 ais523: well, you could change [high activity][long pause][result] into [halved activity][result] 18:05:15 the high activity outputs constantly 18:05:15 and you can do that just as easily with a postprocessor 18:05:17 as does the result 18:05:37 elliott: yep, so you can get a postprocessor that just hides the results until it's ready to feed one to the user and make it look like they're coming constantly 18:05:50 heh 18:05:53 well, OK 18:06:07 ais523: your argument seems to imply threads are useless on a uniprocessor 18:06:11 I imagine most people wouldn't find that useful, but in case it's required… 18:06:19 -!- Klisz has quit (Ping timeout: 244 seconds). 18:06:28 ais523: his question was about why Perl is so awesome and why he hasn't been using it since he was 8. 18:06:35 elliott: no, we were specifically discussing the single-threaded case 18:06:51 I can agree with a concurrent but uniprocessor case, potentially, although I think any performance gains will be marginal 18:10:27 it would be nice if strictness analysis worked perfectly always. 18:10:55 I guess it can't because doing so would be uncomputable? 18:11:18 ais523: but, umm, what's the difference between doing A, B, A, B, and doing A and B in threads on a uniprocessor machine? 18:12:17 because doing A and B in threads can just do "A then B", or it can interleave if necessary to do something like benefit from pipeline stalls 18:12:26 ais523: I think you can get pretty close to perfect.. 18:12:48 ais523: that's a microoptimisation 18:12:50 kallisti: no you can't 18:12:52 kallisti: indeed; you can write infinite loop checkers in practice that catch the majority of accidental infinite loops 18:13:04 strictness analysis is harder than detecting infinite loops imo 18:13:10 yeah.. 18:13:11 but you're not going to be able to use them for solving the world's big fundamental problems 18:13:17 it's incredibly easy to change semantics 18:13:57 elliott: this sort of thing's what my PhD is about, transforming programs without changing their semantics 18:14:14 as long as you have a really good type system for the transformation (which needn't be the same as that of the underlying language), it's not too bad 18:14:20 ais523: "this sort of thing" -- you took a large leap in generality there :P 18:14:28 elliott: indeed :) 18:14:37 strictness analysis results in rather simple transformations, they're just really tricky to find 18:17:08 12:35:49: psh, come on. you can totally non-deterministically reverse to multiple past images. 18:17:08 12:35:50: elliott_: I'm surprised that you're surprised at that; it's become more and more common recently 18:17:09 12:35:55: I'm upset by it, but not surprised 18:17:11 I don't keep up with PC gaming 18:17:14 12:36:12: what's wrong with requiring Steam? it's free right? 18:17:28 kallisti: must be running to play the game, requires network connection to play the game, forced updates 18:17:32 kallisti: DRM 18:18:15 Network connection isn't required 18:18:34 Deewiant: I haven't heard good things about the "offline mode" 18:18:35 Nor, in all cases, is keeping Steam running; you can often run the game exe directly 18:18:43 (Mostly what I've heard is "it doesn't work") 18:18:43 elliott: It's worked for me the few times I've used it 18:19:22 Deewiant: Well, my other objections are still relevant :P 18:19:37 Forced updates is the only relevant one 18:19:55 existence of DRM is often relevant 18:20:12 I took "DRM" as a summary of the previous points 18:20:38 nah, the fundamental concept of DRM is that it prevents you copying the game to a different system 18:20:46 the previous points are consistent with that, but don't imply it 18:20:48 Which Steam doesn't 18:20:57 it does if you don't own it 18:21:24 Deewiant: It was a summary, yes 18:21:30 Another objection: You're not allowed to sell your copy of the game 18:21:36 (Or, more generally, share it) 18:22:14 True enough 18:22:35 I don't think is an objection to "Steam has DRM" as much as it is "games are copyrighted and closed source" 18:22:47 Deewiant: Even more generally, I object to buying a physical box containing what is essentially a right to rent the game out :P 18:22:58 kallisti: Uhh, no. 18:23:02 kallisti: Selling games is not illegal. 18:23:17 I can buy a non-Steam game box and sell it and the recipient can play the game. 18:23:19 elliott: well, I did just that in the case of Neverwinter Nights; went to a shop and bought the Windows version for a license to download the Linux version (from the manufacturers) 18:23:22 I can buy a non-Steam game box and give it to someone else and the recipient can play the game. 18:23:30 None of these things are possible with a Steam box. 18:24:07 ais523: At least you could move those bits around and have them still work without them being tied to an account 18:24:10 I see. 18:24:19 ais523: Although selling them would be illegal because [copyright law] 18:24:28 elliott: indeed 18:24:51 kallisti: It's pretty impressive that copyright has managed to convince people that wanting to sell something you bought is unreasonable because it's INFRINGEMENT, though 18:25:07 I liked Borland's licenses; they were pretty much "please treat this software as a physical object; it's just fine for you to give or sell it to someone else but you must delete all your copies in the process" 18:25:45 ais523: well, it's more reasonable than most licenses, but it smells of lawyers demanding that bits be coerced into being like boxes 18:25:59 although they generally had a clause that you couldn't use the product to make another product that competed with itself 18:26:14 * elliott wonders what would happen if you just gave people a license to do absolutely anything with the downloaded binaries 18:26:15 so, say, you couldn't produce a commercial C compiler using Borland C 18:26:20 sure, it'd make it legal to put it up on a torrent site 18:26:24 but it'll end up on a torrent site /anyway/ 18:26:37 so if it doesn't noticeably increase piracy and the like, it sounds like a good idea 18:26:44 because of all the hassle it eliminates 18:26:45 elliott: I'm upset that game companies don't sell legal ROMs 18:26:53 I know quite a few people who'd buy them; I probably would 18:27:01 (that is, console game companies) 18:27:08 heh 18:27:21 unfortunately that runs into the problem that ROMs have tons of formats 18:27:24 ais523: many years ago atari licensed some of their arcade roms to a company called star roms.. I bought a few of them 18:27:24 the ROM is going to get dumped anyway, so why not give a legal access route for it rather than force people who want it to get it illegally? 18:27:30 and many of them are bad (= can't accurately represent the source media) 18:27:41 and the chances of the game company picking the right one are very small :P 18:27:48 *games 18:27:49 people would write converters 18:27:58 ais523: you can't write a converter from a broken format to a working one 18:28:04 well, OK 18:28:06 yeah that's what happened.. star roms handled the formatting of it for mame or whatever 18:28:17 I'd assume they'd pick a format that contained all necessary info 18:28:20 elliott: for modern roms you just need a memory dump. Much harder for snes era and older of course. 18:28:23 ais523: consider that the vastly most common SNES ROM format is broken 18:28:27 well, *formats 18:28:38 to my understanding 18:28:42 elliott: hmm, what info does it miss? 18:28:42 indeed 18:28:51 ais523: ask pikhq for details 18:29:09 http://bos.github.com/criterion/ oh my god this is so pretty 18:29:43 Valve Corporation President Gabe Newell also stated "most DRM strategies are just dumb" because they only decrease the value of a game in the consumer's eyes. 18:29:46 heh 18:30:00 wow, is that using something like 200-space indents? 18:30:10 ais523: ? 18:30:20 let me take a screenshot 18:31:06 Nuance's "Dragon Dictation" for iPhone has funny license terms: "You may not: -- (i) use the Service for purposes of comparison with or benchmarking against products or services made available by third parties." So you can't compare it with other iDevice speech recognition things in order to pick one of them to use. 18:31:29 fizzie: ... wtf 18:31:30 elliott: what's up with all the letters blending together at the bottom... 18:31:42 kallisti: where? 18:31:46 kallisti: that's a standard linux/chromium font rendering problem 18:31:53 try Ctrl + Ctrl - 18:32:00 kallisti: the link? works fine in firefox 18:32:14 elliott: oh weird. 18:32:24 elliott: I thought that was going to do... nothing, but it fixed it. 18:32:29 And the "no reverse engineering" clause is particularly broad too: "You may not -- decompile, disassemble, reverse engineer or otherwise attempt to derive, reconstruct, identify or discover any source code, underlying ideas, or algorithms, of the Software or Service by any means". If you're literal enough, you can't even just idly wonder what they've done, because it would be "identifying underlying ideas". 18:32:34 bleh, imgur got stuck at 79% 18:32:37 speaking of linux.. I switched to lubuntu (lxde) and I like it 18:32:40 ais523: ompldr.org? 18:32:43 fizzie: Identifying would be getting it correct 18:32:47 elliott: what's that site about? 18:32:57 fizzie: You can wonder as long as you wonder about the wrong things 18:32:57 Deewiant: Okay, "attempting to identify". 18:33:01 ais523: hardcore pornography and warez, obviously 18:33:10 fizzie: Oh, true. Heh. 18:33:12 hmm 18:33:16 I'll just describe it 18:33:22 also, file hosting 18:33:28 and not hardcore pornography and warez 18:33:31 but close enough 18:33:55 it looks like JavaScript, or a similar language; it starts beyond the left edge of the screen, and the first line is ','","");this.element_.insertAdjacentHTML("BeforeEnd",AU.join(""))};M.stroke=function(AM){var m=10;var AN=10;var AE=5000;var AG={x:null,y:null};var AL={x:null,y:null};for(var AH=0;AHAL.x){AL.x 18:33:58 Deewiant: Someone at work had tried to use it (they added the Finnish option just recently), got some bad results for complicated words with lots of suffixes, and wondered about the language model; felt tempted to say "no, stop! you're attempting to identify ideas!" 18:34:07 ais523: where is that from? 18:34:07 haha top files.. 2 are ponies 18:34:08 then the second line, and all the lines on the first screen, start level with the H in AHAL 18:34:12 Vorpal: elliott's link 18:34:15 ais523: no? 18:34:21 ais523: what browser are you using 18:34:22 ais523: s.getCoords_(AH+AJ,AF+AV);AS.x=z.max(AS.x,AR.x,AP.x,AL.x);AS.y=z.max(AS.y,AR.y,AP.y,AL.y);AU.push("padding:0 ",K(AS.x/D),"px ",K(AS.y/D),"px 0;filter:progid:DXImageTransform.Microsoft.Matrix(",p.join(""),", sizingmethod='clip');")}else{AU.push("top:",K(AW.y/D),"px;left:",K(AW.x/D),"px;")}AU.push(' ">',' G-AM-AQ)/AG,'"',' cropbottom="',(AT-AK-AX)/AT,'"'," />","");this.element_.insertAdjacentHTML(" 18:34:28 ais523: looks like your browser misparses html 18:34:29 Vorpal: Firefox 18:34:44 ais523: that stuff doesn't show up in my firefox window. Nor in view source 18:34:50 function t(Z){switch(Z){case"butt":return"flat";case"round":return"round";case"square":default:return"square"}} 18:34:52 elliott: then a screenful down from that, it starts even /further/ to the right, about one and a half screenfuls of horizontal scorlling 18:34:52 what a good function 18:35:01 ais523: run it in a non-broken browser 18:35:05 ais523: the whole page is a script though 18:35:07 *load 18:35:13 Vorpal: no it's not 18:35:18 well, the chart-drawing is 18:35:29 The Chromium "let's mangle letters together" thing is slightly annoying. (I've been using Chromium lately because my Firefox got so full of tabs I haven't dared to restore-session it, and I don't want to be all ineiros about it.) 18:35:30 right, I just right clicked and went "view source" 18:35:37 below that, about two screenfulls scrolling to the right, are the words "criterion performance measurements" in really large font 18:35:44 ais523: you can stop describing 18:35:56 and then various information, with a rather worrying amount of vertical spacing 18:36:01 elliott: I'm just amused that you found it all beautiful 18:36:02 ais523: which firefox version? 18:36:04 fizzie: it never used to happen to me before i switched to arch; I think Chrome might not have that problem somehow 18:36:08 ais523: because it doesn't look like that 18:36:10 Vorpal: 3.6 18:36:20 ais523: ha ha, I'm seeing something different to you and I think it looks nice 18:36:20 ais523: that is ancient! 18:36:24 amusing! 18:36:25 ais523: I'm on 8.0.1 18:36:25 elliott: in that case, why won't you believe me when I say github is broken? 18:36:28 seemed to look okay for me.. ff 8 in linux 18:36:33 -!- elliott has left ("ragepat"). 18:36:36 -!- elliott has joined. 18:36:37 -!- elliott has left ("ragepart"). 18:36:50 ais523: let me try on whatever luicd has 18:37:16 @tell elliott fuckfacer 18:37:16 Consider it noted. 18:37:20 -!- hagb4rd has joined. 18:37:35 @ignore kallisti 18:37:35 Not enough privileges 18:37:39 hmm 18:37:45 how do I tell it not to let kallisti @tell me? 18:37:49 @help tell 18:37:50 tell . When shows activity, tell them . 18:37:59 @help commands 18:38:00 help . Ask for help for . Try 'list' for all commands 18:38:03 @tell elliott -shooshpap- 18:38:04 Consider it noted. 18:38:05 @list 18:38:05 http://code.haskell.org/lambdabot/COMMANDS 18:38:07 what elliott just did reminds me of when I am trying to get off the phone with someone and saying the usual crap and the phone disconnects 18:38:09 ais523: yeah it is broken on firefox 3.6 18:38:19 do I call back to say "bye", or just leave it at that? 18:38:20 ais523: just upgrade already 18:38:28 hm html5 gepaart mit ein wenig krimineller energie.. enjoy ;) 18:38:32 http://homepage.alice.de/hagbard/disintegration.html 18:38:38 never call to say bye 18:39:12 yeah I don't,, it's just awkward .. so rejoining irc to correct your /quit typo seems weird lol 18:39:13 ais523: in a modern browser it really has a very nice design 18:39:25 Vorpal: isn't being broken on a reasonably recent browser (Firefox 3.6 is reasonably recent, even if the version numbers have been going stratospheric recently to hide it) a sign of a badly-designed website? 18:39:59 I can't think of a clever way to format my code for my IOCCC entry. 18:40:00 Firefox 3.6 is two years old 18:40:02 ais523: maybe. But then it might be a bug in firefox 3.6 18:40:13 Deewiant: right, it's only two years old 18:40:14 didn't work in ie 6 18:40:14 I'm thinking I'll just indent it nicely and say "look: even indented nicely it's incomprehensible" 18:40:18 (kidding, didn't test that) 18:40:20 ais523: I don't think you should design for anything but the official specs. 18:40:30 -!- elliott has joined. 18:40:31 ais523: if a browser is buggy is it not the problem of the web designer 18:40:33 calamari: It was a /part typo! 18:40:34 elliott: You have 2 new messages. '/msg lambdabot @messages' to read them. 18:40:43 -!- elliott has quit (Quit: this is a quit typo!). 18:40:55 ais523: it is the problem of the browser programmers 18:40:57 Deewiant: But latest 3.6.x is less than a month old. 18:41:05 ais523: do you disagree about that? 18:41:07 (3.6.24, Nov 08.) 18:41:13 Gregor: I've had the best indentation idea ever for my entry 18:41:14 it's managed to confuse every prettypogram I've tried yet 18:41:17 Vorpal: perhaps; I'm not convinced that the website is following the specs, though 18:41:32 fizzie: Since < 3.6.10 they've all been only security updates 18:41:36 ais523: I could throw it into the w3c validator. Not sure it does js 18:41:56 well, it shouldn't be displaying scripts it doesn't understand 18:42:04 -!- elliott has joined. 18:42:04 well, it isn't 18:42:05 ais523: It is; your browser misparses comments. 18:42:08 Deewiant: An update is an update is an update. 18:42:09 and it should gracefully degrade for people with scripts turned off (I have scripts turned off) 18:42:16 elliott: glad you found my analysis amusing tho 18:42:19 elliott: hmm, aha, I think I might know what the problem is 18:42:23 "118 Errors, 5 warning(s) " 18:42:38 does the page have a "comment" starting