00:01:28 -!- oklofok has quit (Client Quit). 00:06:31 uh 00:11:55 -!- FireFly has quit (Read error: Connection reset by peer). 00:12:33 -!- FireFly has joined. 00:21:39 nooga: uh? 00:21:45 tswett: I'm waaiiiitiiiing. :p 00:24:34 Please write a C program, at most 415241 characters long, that calculates pi to one million decimal places. 00:25:44 tswett: That's not difficult, is it. 00:26:17 No, too lenient. It must be at most 165373 characters long. 00:26:26 tswett: Can it print in hex as long as the result is 1 million decimal places when converted to decimal? 00:26:42 my gf went to a girl's party 00:27:01 Yes, if it's accurate to one million decimal places. 00:28:21 and got back completely stoned 00:28:24 gerat 00:28:26 great 00:29:08 tswett: Just a minute then. 00:30:04 tswett: Or a few minutes. 00:31:06 tswett: Is it alright if I, uh, steal some code from elsewhere and adapt some of it? 00:31:07 http://www.spoj.pl/ranks/PIVAL/lang=C here's some that are limited to 4096 bytes (no code, though). 00:31:20 They presumably could go to 1 million given enough time 00:31:26 elliott: only if that code is written in plain English. 00:31:40 tswett: What if it's written in mathematicalformulas? 00:31:47 *mathematical formulas 00:32:13 Those are fine. 00:32:52 -!- poiuy_qwert has joined. 00:35:54 tswett: Oh fer gods sake, this is fiddly; I was just going to do http://en.wikipedia.org/wiki/Bailey%E2%80%93Borwein%E2%80%93Plouffe_formula in an infinite loop. 00:36:04 tswett: That's an easy way to print the entirety of pi with a simple program. 00:36:07 In hex, no less. 00:36:28 Since you don't need any arbitary-precision this or whatever. 00:36:39 tswett: Aha, wait, I think I've got it. 00:36:50 tswett: Gimme a minute. 00:39:16 tswett: Why do you want this, anywy? 00:39:18 *anyway 00:41:33 No good reason. 00:46:32 tswett: this is SO NEARLY ALMOST SORT OF WORKING 00:49:37 tswett: are you trying to kolgomorov complexitise pi? :P 00:50:37 tswett: btw i cheated *slightly* (stole one single expression that is just a basic arithmetical calculation restated in python) 00:50:40 just to be a bit lazy 00:53:13 -!- FireFly has quit (Quit: swatted to death). 00:55:34 -!- sftp has quit (Remote host closed the connection). 00:59:10 tswett: Think I've got it. 00:59:56 tswett: 245 foo.c 00:59:58 tswett: 00:59:59 #include 00:59:59 #include 00:59:59 int main(void){printf("3.");for(long double n=1,r,l=0;n<1000000;n++){printf("%x",(int)floorl(16*(l=(modfl((16*l)+(((120*n-89)*n+16) / ((((512*n-1024)*n+712)*n-206)*n+21)),&r)))));fflush(stdout);}printf("\n");} 01:00:04 tswett: Valid C99, or something in that vicinity. 01:00:13 Verification of the one million hex digits is left as an exercise to the reader. 01:00:19 Oh, wait, that can be even shorter. 01:00:51 243 foo.c 01:00:53 #include 01:00:53 #include 01:00:53 int main(void){printf("3.");for(long double n=1,r,l=0;n<1000000;n++){printf("%x",(int)floorl(16*(l=(modfl((16*l)+(((120*n-89)*n+16)/((((512*n-1024)*n+712)*n-206)*n+21)),&r)))));fflush(stdout);}printf("\n");} 01:01:11 -!- poiuy_qwert has quit (Read error: Connection reset by peer). 01:01:15 tswett: Are you happy? :p 01:01:33 int main(void){printf("3.");for(long double n=1,r,l=0;n<1000000;n++){printf("%x",(int)floorl(16*(l=(modfl((16*l)+(((120*n-89)*n+16)/((((512*n-1024)*n+712)*n-206)*n+21)),&r)))));fflush(stdout);}puts("");} 01:01:36 One shorter! Oh the joy 01:01:39 *joy! 01:02:28 #include 01:02:28 #include 01:02:28 int main(void){printf("3.");for(long double n=1,r,l=0;n<1<<20;n++){printf("%x",(int)floorl(16*(l=(modfl((16*l)+(((120*n-89)*n+16)/((((512*n-1024)*n+712)*n-206)*n+21)),&r)))));fflush(stdout);}puts("");} 01:02:30 Shorter AND prints more digits! 01:02:45 Although the results of that one are fishy. 01:03:09 Actually the end results are fishy in general but that might just be my eyes. 01:03:35 tswett: praise plz 01:04:54 Ilari: http://marc.info/?l=openbsd-tech&m=129236621626462&w=2 01:05:04 Ilari: Alleged FBI backdoors in widely-copies OpenBSD IPSEC code. 01:06:32 -!- poiuy_qwert has joined. 01:12:28 tswett: Do you have the power to summon oerjan? 01:16:12 Ilari: http://marc.info/?l=openbsd-tech&m=129236621626462&w=2 <-- ouch 01:16:35 "The FBI-backdoors-OpenBSD story is like the NSA assassinations stories: Believable except for the minor detail that it's the wrong TLA." --cperciva 01:17:13 elliott, TLA? 01:17:16 three letter acronym 01:17:24 elliott, indeed 01:17:24 one would expect it to be the NSA wanting such backdoors, not the FBI 01:17:27 Vorpal: can i have servers+funding please 01:17:34 elliott, for? 01:17:43 Vorpal: project! 01:17:47 also /me is trolling #openbsd now >_> 01:17:51 elliott, what project 01:17:57 (okay, opaquely enough that it doesn't actually count as trolling) 01:17:58 and trolling is *never* nice 01:18:03 it's joke-trolling 01:18:21 elliott, anyway, going to sleep as I packed the backpack 01:18:28 Vorpal: the project project! parade! 01:18:30 9.21M addresses allocated from APNIC this calender month (based on delegated file dated 2010-12-14)... At this rate, IANA pool will deplete in a month (mid-January). 01:18:40 Ilari: http://marc.info/?l=openbsd-tech&m=129236621626462&w=2 :P 01:18:47 elliott, answer: no, you are being annoying vague too 01:18:53 annoyingly* 01:18:57 Ilari: Freow. 01:18:58 Vorpal: do you really /want/ me to be specific 01:19:00 pikhq: http://marc.info/?l=openbsd-tech&m=129236621626462&w=2 01:19:05 amusing if nothing else 01:19:10 elliott, sure 01:19:17 elliott, is it for kitten? 01:19:26 Vorpal: nope, it's an entirely separate project 01:19:28 Vorpal: but an excellent one! 01:19:32 elliott, what project 01:19:39 Vorpal: i want to obsolete VPSes. well, mostly. >:) 01:19:46 Vorpal: i even know how to market it! 01:19:48 elliott, details 01:20:16 elliott: Aaaah, the joys of the spy agencies spying on each other. 01:21:06 Vorpal: basically, you pay me $N/mo. where by me I mean $company. put "float" in front of a command, and it executes on $server, where you don't care what $server is. you can manage this process in various ways from the command-line -- putting it in the "background", monitoring it, killing it, sending stuff to its stdin, etc. -- and also view its output, log files (you have to tell it what those are, though -- pretty much looking through the / it 01:21:06 's in), etc. 01:21:33 Vorpal: this process is seamlessly migrated to other servers as the server load demands, and given a Xen-or-whatever allocation according to how much you're paying us, and how much you're already using, globally 01:21:49 Vorpal: you can start processes in more involved ways to make them integrate better with the system, by specifying more about how they act 01:22:08 Vorpal: the migration doesn't lose internet connections and the like, because sockets and the like go through "fd proxy" machines 01:22:13 (tl;dr custom patched libc) 01:22:23 elliott, sooo. some application level virtualisation along google apps lines? 01:22:28 Vorpal: not really, it'd be actual linux 01:22:31 Xen-based or whatever 01:22:38 elliott, but per process? 01:22:48 Vorpal: pretty much. obviously you can spawn multiple processes at a time, but yes, per "task" 01:23:03 Vorpal: the basic use-case is... you've made an IRC bot, and you've been running it locally, but now you want it up permanently 01:23:05 elliott, and how do you count fork() 01:23:06 put "float" in front of the command 01:23:08 forget about it 01:23:09 tada 01:23:11 Vorpal: "count"? 01:23:14 it goes in the same namespace 01:23:19 elliott, same or different task 01:23:19 as the original process 01:23:22 same. duh 01:23:28 "task" is up to you 01:23:28 elliott, okay, fork() + execv() ? 01:23:31 ... 01:23:31 ah 01:23:32 okay 01:23:34 right 01:23:35 YOU ARE SUCH A FUCKING NITPICKER >_< 01:23:38 that makes sense 01:23:49 Vorpal: anyway, it'd also scale to things above IRC bots 01:24:02 elliott, such as minecraft? 01:24:07 sure :D 01:24:15 for instance, the float command would try to automatically figure out what paths and files you've mentioned, and upload them (you are, of course, able to do this manually if the magic isn't working out) 01:24:21 so e.g. "mybot myconffile mydatadir/" 01:24:30 elliott, do you think you could implement it actually 01:24:34 you could also access the filesystem from afar 01:24:37 e.g., you start a web server 01:24:40 and want to upload stuff there 01:24:43 so you scp/sftp in 01:24:49 Vorpal: sure, it's all technically feasible 01:25:07 Vorpal: migrating processes w/ the patched libc for keeping FDs it'd require would be the hardest part, really 01:25:08 elliott, and the file system, this is shared in the "cluster"? 01:25:10 -!- poiuy_qwert has quit (Ping timeout: 240 seconds). 01:25:16 Vorpal: no, it's Xen-esque 01:25:23 every process gets its own root filesystem 01:25:30 (of course you could share them, by setting it up manually, I suppose) 01:25:44 elliott, so how do you properly migrate that. The fd-proxy would work fine for, say, sockets 01:25:52 elliott, but not so great if you try to mmap() files 01:26:09 Vorpal: well it'd pretty much re-mmap() them on the other side 01:26:25 elliott, can you ensure proper mmap() semantics between threads and forks? 01:26:27 Vorpal: it'll either have to be non-Linux-based at the migration level, or use some kind of kernel patch to be able to freeze and restore processes, I'll have to see about it 01:26:35 it's likely that you'll have to specifically enable migrating to another server 01:26:35 elliott, I'm thinking "database server" here 01:26:41 Vorpal: let's see if i can! 01:26:47 if not, fine, your DB server isn't migratable 01:26:56 you get shitty performance if the server gets bogged down, but what can you do? you suffer that with VPSes anyway 01:27:01 the performance of my solution is always >=VPS 01:27:07 elliott, there is linux kernel support for freezing processes 01:27:09 iirc 01:27:11 basically, this is all based on the observation that VPSes generally run a bunch of services, but they rarely *interact* directly with each other 01:27:18 and it's much easier to be able to manage them separately 01:27:36 also, the cost of entry is way too high; setting up an IRC bot involves scp, shelling in, probably setting up a screen instance if it has a useful stdio interface 01:27:36 etc. 01:27:42 as opposed to just saying "here, run this somewhere on the interwebs" 01:27:50 elliott, you need to ensure proper handling of shared memory between a db server and something using it 01:27:52 marketing strategy: can't you guess?! Cloud! Run your programs from the cloud! 01:27:55 which is annoying 01:27:56 Total cloud virtualised environment cloud! 01:28:11 Vorpal: well, SQL servers are often used over sockets :-P 01:28:15 usually, even 01:28:31 elliott, yes true, but they are over shm if local often 01:28:49 Vorpal: yeah, true ... I mean there'd be no limit to how "big" a task can be really 01:28:56 since it's the sum of your resource usage that's considered 01:29:00 rigjt 01:29:02 right* 01:29:17 it'd certainly be a technically interesting project to work on 01:29:27 the only problem is that it needs a bunch of dedicated servers to get started :) 01:30:23 elliott, resources I don't have 01:30:33 Vorpal: buy them! 01:30:58 elliott, ah, but the resource I lack here is money 01:31:11 XD 01:31:13 elliott, do like xkcd suggested and report it to the idea agency or whatever it was when someone else implement this. ;P 01:31:17 Vorpal: it's okay, you can have 50% of the profits!!!!!!!! 01:31:21 i'm an IDEA MAN 01:31:33 Halliburton is offering to pay Nigeria $250 million to drop *bribery* charges against Cheney. 01:31:49 They are bribing someone to get away with bribery. 01:32:00 pikhq, Halliburton? 01:32:10 pikhq, also, are they doing that in the open? 01:32:18 is it a bribe then technically 01:32:38 Vorpal: They do oil. 01:32:44 Vorpal: Also, no, not in the open. 01:33:20 They've also been a major military construction contractor. 01:33:40 Dick Cheney was their CEO. 01:33:46 more like cheney the dick LOL 01:34:51 elliott: I praise you, O great elliott. 01:35:37 He most certainly is a dick. 01:35:39 tswett: thank you. so why did you want that program 01:36:14 #include 01:36:14 #include 01:36:15 int main(){printf("3.");for(long double n=1,r,l=0;n<1000000;n++){printf("%x",(int)floorl(16*(l=(modfl((16*l)+(((120*n-89)*n+16)/((((512*n-1024)*n+712)*n-206)*n+21)),&r)))));}puts("");} 01:36:19 doesn't flush output after every digit, but is smaller 01:36:23 (220 chars) 01:36:27 anyone feel like golfing it? pikhq? :P 01:36:39 night → 01:39:24 elliott: Meh. 01:39:49 pikhq: but it's 1 million hex digits of pi! 01:39:58 probably, there might be an overflow in there :) 01:40:25 elliott: like I said, no good reason. 01:41:10 tswett: C'mon, I need to feel good about having written it :P 01:41:14 Since it was utterly pointless. 01:43:51 *facepalm*... 01:44:09 Gawker stored passwords. But only the first 8 characters. 01:44:14 They actually truncated passwords. 01:44:16 pikhq: That's because of DES. 01:44:57 elliott: ... 01:45:08 What, did they use *1 block* of DES? 01:45:34 SO MUCH FAIL. 01:45:49 pikhq: Abcdefgh, Joy of Man's DESiring. 01:45:56 (worst pun ever) 01:52:37 elliott: I think it's amazing how you're able to write working C code at your age. 01:53:16 You clearly understand computing better than the vast majority of people in the world. 01:54:10 tswett: Yes. Also, I am god. 01:54:17 Furthermore, I invented sandwiches. 01:54:55 elliott: which kind? 01:55:10 Mathnerd314: ALL KINDS 01:56:38 I have a language construct, which theoretically can do everything in existence. does anyone care? 01:57:01 maybe 01:57:33 Mathnerd314: there are lots of those drifting around. :) But sure, what is it? 01:58:05 tswett: I'd imagine many of us could write working C code at elliott's age. 01:58:05 quick description: "fixed-point ambiguous lazy definition" 01:58:22 But, then, we're people sitting in a channel *all about* programming. Sooo. :) 01:58:25 I have a language construct, which theoretically can do everything in existence. does anyone care? 01:58:27 I do too. 01:58:29 it's called lambda. 01:58:38 elliott: this is more powerful. 01:58:46 Mathnerd314: ...no, no it's not. 01:58:50 it can do types 01:58:52 unless it's uncomputable. 01:58:54 in which case sure. 01:59:03 but no, your construct is not more powerful than lambda. by definition. 01:59:17 It may be easier to use than lambda. 01:59:29 Mathnerd314: Lambda is Turing-complete. Therefore your construct is not more powerful than lambda. QVOD ERAT DEMONSTRANDVM 01:59:37 Qvod. 01:59:45 But if it's computable, lambda calculus can compute it. 01:59:49 tswett: That's very subjective and arguable, still. 01:59:53 I find lambda pretty easy. :p 01:59:54 elliott: theoretically, its value is undecidable in some cases. 02:00:05 s/its/my construct's/ 02:00:14 Mathnerd314: lambda too :P 02:00:26 (\x. _|_) \equiv _|_ 02:01:07 I'll ignore you and give examples 02:01:09 Mathnerd314: okay, then maybe it is more powerful than lambda. 02:01:19 Ignoring elliott is often a good plan. :P 02:01:23 a = 2 -- simple definition 02:01:31 tswett: I'm right, though. 02:01:34 a = 1; b = a -- simple assignment 02:01:41 tswett: const _|_ = _|_ 02:01:59 Mathnerd314: ok so that's two constructs (definition and assignment) already 02:02:30 elliott: no, just examples. the type is [(LHS,RHS)] 02:02:42 very well. 02:02:46 show something more interesting? :-P 02:03:39 a = 1 : a -- infinite data structure 02:04:08 ok, so right now it's haskell :P 02:04:17 or is : not part of the control structure? 02:04:26 i.e. is that "a = f x a" for arbitrary function f and value x? 02:04:47 no, (:) is just an operator 02:05:04 Mathnerd314: is it part of the construct? 02:05:05 or not? 02:05:08 a = 0 + a -- a is all numbers 02:05:31 (all elements of rings, to be precise) 02:05:38 Mathnerd314: erm, "a = 0 + a" -- a is 0 02:05:42 presumably you mean 1 + a? 02:05:48 are : and + part of the construct? 02:06:03 no, they're just operators. 02:06:10 very well. 02:06:15 more examples? show some actual computation 02:06:17 the construct finds the fixed points 02:06:30 Mathnerd314: btw, this thing is not more powerful than lambda because it /depends/ on lambda or something equally powerful 02:06:33 that is, it depends on functions 02:08:06 Mathnerd314: am i being too nitpicky :) 02:08:18 well, it does simultaneous equation solving too: a = b^2 + 1; b = a - 1 -- a is {-2,-1}, b is {-3,-2} 02:08:44 or maybe I have that wrong 02:08:46 ok ... do you have any actual implementation strategy for this? :) 02:09:01 elliott: no. but I have lots of special cases. 02:09:12 Mathnerd314: so it's not really a construct so much as a heap of special cases :) 02:10:05 elliott: no, I can describe it simply: fixed-points of the given equations 02:10:21 Mathnerd314: ok, but it's not really much of a construct if it doesn't have an implementation that isn't a heap of special cases 02:10:49 right... it's too powerful to have one implementation that works all the time 02:11:10 i think we have different definitions of power 02:11:21 but hey, it'll fit right in to mathematica :P 02:11:54 that's the main point: integrate math and programming so well you don't notice the difference 02:12:35 Mathnerd314: have you ever used mathematica? 02:12:38 you'll love it. 02:12:43 well at least until you realise how hateful it is. 02:13:00 a = 1 + a -- various types of infinity (-infinity and infinity for reals, set {0,1,2,3} for Z_4, etc.) 02:13:12 -infinity and infinity aren't reals 02:13:31 well, reals extended to limits 02:13:52 -!- poiuy_qwert has joined. 02:14:30 deewiant floating http://imgur.com/5fLbZ.png 02:14:46 * Mathnerd314 used mathematica until he met Stephan Wolfram 02:14:53 +1 02:15:05 I have not met him 02:15:08 but he seems errrrgh 02:15:56 he's funny :) 02:16:28 "my brilliant mind was correct! it's universal! i'm a genius! a genius! i tell you! p.s. thanks to ais523 who helped fellate my ego by proving it but who cares about that" 02:18:05 Mathnerd314: I'd like to see a formal definition of this construct. 02:18:34 Speaking of constructs, I've been pondering languages that are based as much as possible on a single thing. 02:18:52 Like a language where the only thing there is is the associative array, or the semaphore. 02:19:02 Or the continuation. 02:19:33 tswett: {variable assignments s.t. all equalities are satisfies} 02:19:49 s/s}/d}/ 02:20:11 Mathnerd314: ok, so it's... an equation! 02:20:33 elliott: right, I guess the key is in which types of values you have 02:20:48 what? 02:21:25 elliott: the variables can be assigned *sets* of values 02:21:44 e.g. a = {1,2,3,4,...} is a legitimate result 02:21:44 Mathnerd314: ok... so it's an equation. some equations have multiple solutions. 02:21:54 that's... a rather old mathematical result, one would think 02:21:58 elliott: but it has more solutions 02:22:04 Mathnerd314: more solutions than what 02:22:10 Mathnerd314: so it's a function that takes a set of equations, and returns the set of all values that satisfy the equations? 02:23:13 elliott: than normal equation solving. e.g. a = !a has solution {True,False} 02:23:17 c = root(n, a^n + b^n), n = 3+x 02:23:22 fuck yeah! 02:23:34 Mathnerd314: but True = !True isn't true at all 02:23:37 and False = !False isn't true, either 02:23:40 so that's a meaningless result 02:23:49 and I don't think you can define results like that even *slightly* in the general case 02:24:07 elliott: but !{True,False}={!True,!False} = {False,True} = {True,False} 02:24:24 that's why it's weirdly powerful 02:24:43 because it will *always* assign values 02:24:46 Mathnerd314: ok, but you can define that with normal equations too, more or less. 02:24:58 I mean, solving "a = !a" over sets of booleans 02:25:00 and you get that result 02:25:12 Mathnerd314: if so, then what constants and operators are allowed in the equations, and what values are considered to be possible solutions? 02:25:17 right... that was what I was saying 02:25:59 Mathnerd314: so it's an equation. :P 02:26:08 tswett: it's defined on general functions, so you can really extend it to any domain 02:26:24 elliott: but it always has solutions, which is not a property of most things 02:26:33 Mathnerd314: c = root(n, a^n + b^n), n = 3+x 02:26:35 Mathnerd314: what's the solution? 02:26:55 elliott: an equation set dependent on x 02:26:58 anyway the function that takes a set of equations and returns a set of bindings that satisfy those equations is... mathematics in a function 02:27:04 Mathematica already has your heap of special cases 02:27:06 it's called Solve 02:27:20 Mathnerd314: erm you /do/ realise it has no solutions, right? 02:27:23 it's fermat 02:27:31 Mathnerd314: well, if you say "all functions are allowed", then it simply inherits the power of all the functions that you can stick into it. 02:27:45 -!- zzo38 has joined. 02:27:54 tswett: right... so it's definitely at least as powerful as lambda 02:28:09 Mathnerd314: that's misleading. 02:28:15 It isn't "at least as powerful as"; it's all-powerful. 02:28:19 elliott: no solutions over integers, maybe. but I can construct lots of others 02:28:20 It can solve any theorem. 02:28:29 elliott: probably :p 02:28:40 Mathnerd314: c = root(n, a^n + b^n), n = 3+x, a = floor(a), b = floor(b), n = floor(b) 02:28:45 Mathnerd314: Go on, gimme a solution. 02:29:03 Mathnerd314: what you have is a bit like HQ9+B. It's a Turing-complete programming language with five instructions: H, Q, 9, +, and B. 02:29:04 Anyway, it isn't "at least as powerful as" for another reason: it doesn't replace lambda, it depends on it -- it depends on functions to work. 02:29:08 So it isn't really a control structure at all. 02:29:15 tswett: What does B stand for? :-P 02:29:22 Mathnerd314: but the instruction B simply executes the remainder of the program as Brainfuck, and that's where all the power lies. 02:29:33 elliott: everything set to infinity 02:29:40 :p 02:30:05 Mathnerd314: c = root(n, a^n + b^n), n = 3+x, a = floor(a), b = floor(b), n = floor(b), a+1 != a, b+1 != b, n+1 != b 02:30:12 Mathnerd314: your construct allows arbitrary things from set theory to be included. But set theory is already Turing-complete and more. 02:30:16 Presumably you allow !=, since it's all-powerful /anyway/ you might as well make it even more fun. 02:30:44 tswett: What /is/ the computational class of a machine that can work out every solution to a given equation, or tell you if there are none, anyway? 02:30:45 Mathnerd314: you could just say "x = the result of running the following Brainfuck program: ..." 02:30:53 Apart from "the highest computational class even theoretically possible". 02:31:06 elliott: essentially, this takes questions in set theory and answers them. 02:31:18 tswett: Yes, yes, I'm just asking a side-question. 02:31:21 Which... is a pretty high computational class. :P 02:31:24 Do you know anything about weather systems? 02:31:25 tswett: Is the computational class of a machine that can do that even /named/? 02:31:27 It's pretty insane :P 02:31:36 Yeah. I wonder how well-defined it is. 02:31:44 tswett: don't forget: it gives multiple answers too 02:32:04 it's *ambiguous*, so it's always well-defined 02:32:13 tswett: I think it's well-defined but *almost certainly* not physically possible :P 02:32:39 tswett: I mean, you can verify any theorem in O(1) time (well, if the machine can do it in O(1)). 02:32:46 More importantly: You can verify any theorem haltingly. 02:32:51 Any theorem! Any! 02:32:58 Well, any expressable in ZFC or whatever. 02:33:01 Which is an awful lot. 02:34:35 elliott: right, it can "do" the halting problem, by giving {true,false} as well as true and false 02:34:49 Mathnerd314: that... 02:34:54 Mathnerd314: no, you can't do that. 02:35:01 that doesn't make sense even in the framework of your definition 02:35:36 Do you know anything about weather systems that you can make some weather in a computer game and that the captain can sometimes predict the weather from it? 02:37:06 elliott: hmm, I'm missing the inconsistency 02:37:51 Mathnerd314: can it always return {true,false}, or only sometimes? 02:37:53 Mathnerd314: {True,False} = Halts(P,P) 02:37:58 that makes no real sense, at all 02:41:22 -!- elliott has quit (Quit: Leaving). 02:41:55 Does anyone know anything about weather in here at all? 02:47:11 Do you know which channel and on which network can help? 02:50:39 tswett: only when the input depends on the output. it returns the *least* fixed point. 02:50:51 that's not the empty set 02:52:50 hmm, I detect a distinct silence due to lack of elliott 03:24:07 " !!!" 03:52:45 -!- wareya_ has joined. 03:55:36 -!- wareya has quit (Ping timeout: 260 seconds). 04:08:00 -!- poiuy_qwert has left (?). 04:35:58