00:00:03 <olsner> oerjan: hmm, I think you sent me a message earlier... I may have intended to answer
00:00:41 <olsner> maybe it was not important?
00:00:55 <shachaf> kmc: what if you want rust not to do it
00:01:04 <shachaf> imo doubleplusunsafe { ... }
00:02:06 <oerjan> `pastelogs oerjan> @.* olsner
00:02:41 <kmc> danger(10000) { ... }
00:02:54 <HackEgo> http://codu.org/projects/hackbot/fshg/index.cgi/raw-file/tip/paste/paste.32727
00:04:13 <olsner> oerjan: ah! then I would guess "yes"
00:04:20 <kmc> pwned boxes
00:04:46 <olsner> though I'm not sure how much asbestos contamination does to fireproof a spider
00:05:02 <oerjan> olsner: we are all doomed hth
00:07:36 <olsner> oerjan: hopefully only wales is doomed hth
00:08:11 <oerjan> hm well as long as they cannot swim
00:08:34 <olsner> p. sure wales can swim
00:10:29 <oerjan> it's hard to argue with a pun based on a misspelling.
00:15:59 <kmc> but if they wash up on the beach they're property of the queen
00:16:36 -!- Sgeo has joined.
00:20:07 -!- Sprocklem has joined.
00:21:11 -!- mroman_ has quit (Read error: Operation timed out).
00:21:25 -!- mroman_ has joined.
00:21:26 <oerjan> kmc: so that's how england got wales in the first place?
00:21:40 -!- ter2 has joined.
00:21:45 -!- tertu has quit (Ping timeout: 265 seconds).
00:22:15 <oerjan> thx it all makes sense now
00:36:24 -!- muskrat has quit (Ping timeout: 246 seconds).
00:42:58 -!- Phantom_Hoover has joined.
00:43:01 -!- Zuu has joined.
00:55:28 -!- Zuu_ has joined.
00:56:42 -!- Zuu_ has quit (Remote host closed the connection).
00:58:38 <kmc> ruddy: are you enjoying finnish independence day
00:58:39 <ruddy> because i'm `thanks independence day also it's independence day
00:58:46 <kmc> ruddy: have you declared yourself independent from finland
00:58:47 <ruddy> i think that is true independent of number of cells
00:58:55 <ruddy> No clue what you mean. What do you think, fungot?
00:59:02 -!- Zuu_ has joined.
01:01:19 -!- drlemon has quit (Read error: Connection reset by peer).
01:02:50 -!- drlemon has joined.
01:06:29 -!- MindlessDrone has quit (Quit: MindlessDrone).
01:07:07 <nooodl> `thanks independence day also it's independence day
01:07:09 <HackEgo> Thanks, independence day also it's independence day. Thindependence day also it's independence day.
01:16:33 -!- Halite has quit (Ping timeout: 250 seconds).
01:19:46 <coppro> why is there no dramatic reading of the call of cthulhu for beginning readers on the internet
01:21:19 -!- conehead has joined.
01:21:56 <oerjan> coppro: they keep mysteriously disappearing together with the readers
01:22:42 <coppro> oerjan: but I can find the *regular* version fine
01:22:46 <coppro> I just want the seussized one
01:24:11 <oerjan> coppro: just read through the iwc christmas comics hth http://www.irregularwebcomic.net/archive.html
01:25:18 <int-e> One cthulhu, two cthulhu, red cthulhu, blue cthulhu.
01:26:12 <zzo38> The reason must be that dramatic readers who like to post stuff on internet don't like that book or do not know how to read it for some reason. Either that or they are hiding it.
01:27:04 <shachaf> happy independence day, nooodl
01:27:28 <nooodl> sinterklaas "basically that"
01:27:57 <oerjan> sinterklaas = basically cthulhu
01:34:04 <HackEgo> Semi-automatic text generation.
01:34:15 <oerjan> oh come on, that doesn't even backronym.
01:34:31 <zzo38> I notice that too.
01:35:07 <HackEgo> zzo38 is not actually the next version of fungot, much as it may seem.
01:35:12 <HackEgo> fungot cannot be stopped by that sword alone.
01:35:17 <ruddy> `? shachaf `? shachaf
01:35:18 <ruddy> HackEgo? ¯\(°_o)/¯
01:35:34 <ruddy> eille! no violence against my peaceful bot!
01:35:47 <shachaf> ruddy: today is the day i am going to sleep
01:35:59 <shachaf> ruddy: you gotta do what you gotta do
01:36:54 <oerjan> `run echo "HackEgo? ¯\(°_o)/¯" >wisdom/ruddy
01:39:17 -!- Zuu_ has quit (Remote host closed the connection).
01:41:25 <oerjan> * int-e blames lambdabot <-- ok how did lambdabot get you here
01:42:06 -!- Zuu_ has joined.
01:43:25 -!- Zuu_ has quit (Remote host closed the connection).
01:43:38 -!- Zuu_ has joined.
01:44:18 <int-e> oerjan: I looked at the list of channels that it joins automatically
01:44:39 -!- yorick_ has joined.
01:44:49 -!- yorick has quit (Ping timeout: 246 seconds).
01:45:15 -!- Zuu_ has quit (Remote host closed the connection).
01:46:08 -!- Zuu_ has joined.
01:48:16 -!- Zuu_ has quit (Remote host closed the connection).
01:48:36 -!- Zuu_ has joined.
01:49:02 -!- Zuu_ has quit (Remote host closed the connection).
01:49:27 -!- Zuu_ has joined.
01:50:08 -!- Zuu_ has quit (Remote host closed the connection).
01:50:26 -!- Zuu_ has joined.
01:51:23 <oerjan> Zuu_: fix your connection thx
01:51:50 -!- elliott_ has changed nick to elliott.
01:53:18 <Zuu> oerjan: tis fine, i was just configuring something :/
01:53:47 <Bike> is there some canonical algorithm for assigning binary codes to each node of a state diagram
01:54:16 <oerjan> Bike: it's called "counting" hth
01:54:51 <Bike> i mean, that minimizes the logic
01:55:57 <Bike> like in a circuit implementing the state diagram.
01:56:10 <Bike> minimizing hamming distance between connected nodes for instance
01:56:14 <oerjan> there's an algorithm for minimizing the number of states iirc
01:56:22 -!- Zuu_ has quit (Remote host closed the connection).
01:56:24 <Bike> not what i mean
01:57:06 -!- Zuu has quit (Read error: Connection reset by peer).
01:57:53 <oerjan> enough that i expect the answer is "no"
01:58:30 <Bike> well there are multiple constraints you could optimize for, like size of the state
01:59:01 -!- Zuu has joined.
01:59:01 <oerjan> yes, but it sounds like something NP-complete or worse.
01:59:29 <oerjan> so whatever algorithm you find won't be "canonical".
01:59:43 <int-e> size of state is easy, but I expect that often you'll want to use more bits than are strictly required, to encode some features of states.
01:59:47 <oerjan> unless it happens to be for surprising reasons.
01:59:58 <int-e> --> it's a hard problem.
02:04:13 -!- yorick_ has quit (Remote host closed the connection).
02:04:13 <Bike> i'm wondering this because my homework gave constraining hamming distance to one as a rule of thumb but also said it was usually impossible
02:04:16 <Bike> ~engineering~ i guess
02:04:16 <int-e> (a nice example of such features comes up when your automaton is the result of a powerset construction; it may be better to implement the nondeterministic automaton and use one bit per original state)
02:04:39 <Bike> What's a powerset construction.
02:05:03 <int-e> https://en.wikipedia.org/wiki/Powerset_construction
02:19:53 -!- Phantom_Hoover has quit (Read error: Connection reset by peer).
02:21:53 <zzo38> In a 6502 code, if I have the accumulator being zero, what is an efficient way to make the zero flag cleared if some value in the zero-page is zero, and set if it is nonzero?
02:26:04 <Bike> what instructions alter the zero flag again
02:27:21 <fizzie> Loads, transfers, arithmetic.
02:28:27 <zzo38> ADC, AND, ASL, BIT, CMP, CPX, CPY, DEC, DEX, DEY, EOR, INC, INX, INY, LDA, LDX, LDY, LSR, ORA, PLP, ROL, ROR, SBC, TAX, TAY, TSX, TXA, TXS, TYA.
02:29:30 <oerjan> obviously LDA whatever gives the opposite flag setting of what you want
02:29:47 <zzo38> oerjan: Yes I can see that
02:31:09 <Bike> has everyone used the 6502...
02:31:24 <Bike> anyway i have no idea.
02:31:52 <oerjan> way back. although the manual for the assembly was missing instructions...
02:32:17 <zzo38> I wonder if just pushing the flags onto the stack and then testing the zero flag with BIT or AND, would be best way.
02:32:25 <fizzie> oerjan: The opposite of PHP.
02:32:31 <zzo38> oerjan: PLP is pop flags from stack
02:35:24 <Bike> hey zzo38, i'm planning on making some CPUs with my FPGA. do you have any fun ISAs to recommend
02:36:08 <Bike> i guess that would make sense
02:36:32 <Bike> though i don't remember anything about it other than that it drops the crazy decimal stuff and has a zillion registers.
02:39:05 -!- doesthiswork has joined.
02:39:44 <zzo38> Bike: Yes, MMIX is one possibility; GCC supports it, even.
02:40:14 <Bike> i'm not actually sure how i should go about getting code onto the thing, there's no serial port. guess i'll have to get another cable and figure out uart
02:41:12 <zzo38> You can also try 6502 if you want to program it in assembly language (and if it is the variant used in the Famicom, then there is no decimal arithmetic either)
02:42:07 <Bike> any more esoteric ideas?
02:42:29 <Bike> i was thinking of trying imlementing the processor from 'lambda the ultimate opcode'. no alu
02:55:37 <Bike> apparently there was an asynchronous RISC called "AMULET", i'm thinking of finding some specs
02:55:49 -!- doesthiswork has quit (Quit: Leaving.).
03:30:09 <kmc> <kmc> I guess to formalize a statement like "C is Turing complete" you would talk about a family of languages C_i where sizeof(void *) = i, and require that the languages have a uniform description
03:30:13 <kmc> so I think that doesn't work
03:30:39 <kmc> because any given TM (that you would try to implement in C_i for some i) can take arbitrarily large inputs and use arbitrarily much tape, even a vastly more than superexponential function of the input length
03:31:26 <Bike> intptr_t rather puts a damper on things, eh
03:31:59 -!- tertu3 has joined.
03:32:10 <zzo38> What if a "char" cell can contain an arbitrary integer, and it is C89 rather than C99?
03:32:28 <Bike> can you have a C where malloc never fails and just keeps giving you new pointers
03:33:30 <kmc> zzo38: what's the relevant difference between C89 and C99?
03:34:14 <zzo38> kmc: I think C99 defines a lot of stuff that tells you how many bits everything is, isn't it?
03:34:45 <Bike> c89 doesn't have char_bit?
03:35:26 -!- ter2 has quit (Ping timeout: 240 seconds).
03:56:27 -!- pikhq has quit (Ping timeout: 260 seconds).
03:56:35 -!- pikhq has joined.
04:09:33 -!- pikhq has quit (Ping timeout: 252 seconds).
04:09:43 -!- pikhq has joined.
04:14:50 -!- pikhq has quit (Ping timeout: 245 seconds).
04:15:53 -!- pikhq has joined.
04:27:33 -!- Sorella has quit (Remote host closed the connection).
04:46:02 <zzo38> Is there an editor for 6502 assembly language files (even if the file includes macros and other stuff) that will also display the generated hex codes and the bytes and cycles of each line, and the bytes and cycles of any highlighted block, and the addresses, etc?
04:49:56 -!- Slereah_ has quit (Read error: Connection reset by peer).
04:50:13 -!- Slereah has joined.
04:58:53 -!- nooodl_ has joined.
05:01:51 -!- nooodl has quit (Ping timeout: 240 seconds).
05:08:47 <Sgeo> http://slashdot.org/story/13/11/06/1323223/mozilla-backtracks-on-third-party-cookie-blocking
05:08:54 <Sgeo> Said as though third-party cookies are evil
05:11:13 <Bike> i think i would be happier if i hung out in a channel where when people said stuff like that they were talking about politics or such
05:11:39 <Sgeo> I have legitimate use cases for third-party cookies :/
05:11:51 <Sgeo> (non-advertising uses, I mean)
05:12:01 <Bike> i have legitimate use cases for mah diiiiick
05:15:19 <Sgeo> http://kingjamesprogramming.tumblr.com/
05:17:19 -!- oerjan has quit (Quit: leaving).
05:33:12 -!- pikhq has quit (Ping timeout: 246 seconds).
05:35:08 -!- pikhq has joined.
06:12:43 <zzo38> Did you see what Famicom Z-machine programming I have made up so far? http://wiki.nesdev.com/w/index.php/User:Zzo38/Famicom_Z-machine
06:13:23 -!- Sgeo_ has joined.
06:14:15 <Bike> kmc: i don't know this notation. does it mean that you as well have legitimate use cases for mah dick
06:15:10 <kmc> it means "high five"
06:15:15 <kmc> so, kinda?
06:16:15 -!- Sgeo has quit (Ping timeout: 246 seconds).
06:17:07 -!- Sgeo has joined.
06:17:34 <kmc> diiiiickdiiiiickdiiiiickdiiiiickdiiiiickdiiiiickdiiiiickdiiiiickdiiiiickdiiiiickdiiiiickdiiiiickdiiiiickdiiiiickdiiiiickdiiiiick
06:17:38 -!- Sgeo_ has quit (Ping timeout: 246 seconds).
06:18:18 <quintopia> zzo38: why did you choose famicom as the machine you want to develop this for
06:23:04 -!- SingingBoyo has joined.
06:23:40 <zzo38> quintopia: It is one computer that doesn't have Z-machine interpreter on it, is one reason.
06:24:01 <quintopia> zzo38: you think every computer should be able to interpret Z-machine?
06:25:40 <quintopia> @tell ais523 it looks like James Parsons has decided not to make a FLARP! language, you should delete the page.
06:26:54 <zzo38> quintopia: It may be intended. Infocom didn't do it since Famicom is by Nintendo of Japanese, but now other people can make it too. Also, to practice to write such programs.
06:27:42 <quintopia> zzo38: what other kinds of programs like that do you plan on writing in the future
06:28:50 <zzo38> quintopia: I am not exactly sure by now.
06:29:06 <zzo38> Did you find any mistakes in this program or any other comment/questions?
06:35:59 <quintopia> i don't have the assembler or the emulator or test programs. in short, i am a poor alpha tester. sorry.
06:36:41 <zzo38> It isn't complete enough for testing the program; I mean if you noticed anything wrong with the program just by looking at it.
06:38:12 <quintopia> i'm not smart enough to do that either
06:40:10 <zzo38> Do you know anything about Z-machine programming, Famicom programming, and/or 6502 programming?
06:40:47 <quintopia> you've found yourself a very esoteric niche
06:41:08 <zzo38> That wiki contains a lot of information about programming NES and Famicom.
06:41:31 <zzo38> (and other related systems, such as Vs.Unisystem)
06:44:21 <zzo38> One thing a bit unusual about this Z-machine interpreter is that several things are decided at compile-time; this may make the program more efficient in some ways.
06:49:48 <zzo38> The CPU chip used in that computer is the 2A03, which contains two cores which is the CPU core and APU core. The CPU core is a 6502 but with BCD arithmetic disabled (the logic is still there, it is just disconnected from the rest of the circuit). the APU core is used for audio (two square waves, one triangle wave, one noise, one DPCM) but also I/O ports and a DMA transfer.
06:50:11 <Bike> how is the apu controlled
06:50:38 <zzo38> The APU is controlled by memory-mapped registers $4000-$4020.
06:53:03 <zzo38> The address of the DMA transfer is hard-coded and cannot be changed; you can change the source address to any 256-byte boundary, and when activated it reads the data from all 256 addresses in that page and writes the results all to a single address (not to 256 separate addresses).
06:58:29 <kmc> why disable BCD
06:58:47 -!- tertu3 has changed nick to tertu.
06:59:24 <tertu> don't most z-machine games expect a somewhat wider screen than the famicom has
07:01:05 <zzo38> tertu: Yes, but I do with what we have. (Also, this program is only version 1 to 3 Z-machine)
07:02:03 <kmc> http://en.wikipedia.org/wiki/File:Z-machine480.jpg
07:02:22 <Bike> iirc wasn't the bcd part of the chip still on it, just with the wires cut so you couldn't access it
07:02:34 <zzo38> Bike: Yes, it is like that.
07:04:11 -!- L8D has quit (Ping timeout: 252 seconds).
07:04:49 <zzo38> Due to overscan, the visible area is only 30x26, but if no words are longer than 30 then it would fit. However you probably will need to scroll the screen a lot.
07:07:27 <zzo38> Also, for simplicity, everything is displayed in uppercase (which is OK in version 1-3 only, as long as it can be distinguished internally in all cases that matter)
07:09:50 -!- pikhq has quit (Ping timeout: 240 seconds).
07:10:57 <zzo38> kmc: Apparently due to patents on the 6502 CPU design.
07:11:36 -!- pikhq has joined.
07:13:24 <zzo38> In addition, it doesn't implement the status line (there is a bit in the header to indicate this), and the SPLIT and SCREEN operations are not supported (there is also the bit in the header to indicate this too).
07:16:48 -!- nooodl_ has quit (Ping timeout: 246 seconds).
07:38:30 -!- ^v has quit (Remote host closed the connection).
07:39:29 -!- conehead has quit (Quit: Computer has gone to sleep.).
08:01:28 -!- Taneb has joined.
08:03:50 -!- upgrayeddd has joined.
08:35:43 <kmc> zzo38: heh
09:54:30 -!- zzo38 has quit (Remote host closed the connection).
10:19:15 -!- MindlessDrone has joined.
10:20:43 -!- Sprocklem has quit (Ping timeout: 246 seconds).
10:49:27 -!- Sorella has joined.
10:50:00 -!- Sorella has quit (Changing host).
10:50:00 -!- Sorella has joined.
11:05:04 -!- SingingBoyo has quit (Ping timeout: 250 seconds).
11:14:38 -!- carado has joined.
11:44:06 -!- Phantom_Hoover has joined.
12:11:47 <fizzie> Huh, there's really no NES Z-machine interpreter? That is quite surprising.
12:20:46 -!- int-e has quit (Ping timeout: 246 seconds).
12:21:43 -!- int-e has joined.
12:30:11 -!- nooodl has joined.
12:40:36 -!- ais523 has joined.
13:16:15 -!- carado has quit (Ping timeout: 252 seconds).
13:20:42 -!- nooodl has quit (Quit: Ik ga weg).
13:21:03 -!- nooodl has joined.
14:02:56 -!- Taneb has quit (Quit: Leaving).
14:05:25 -!- L8D has joined.
14:09:26 -!- pikhq has quit (Ping timeout: 240 seconds).
14:10:05 -!- pikhq has joined.
14:24:24 -!- MindlessDrone has quit (Ping timeout: 250 seconds).
14:39:43 -!- MindlessDrone has joined.
15:42:43 <ais523> anyone here have advice on how to read a line from a file without using stdio?
15:43:49 <ais523> like, I can think of a few ways, was wondering what the best way is
15:43:58 <ais523> byte-at-a-time is pretty slow, I'd hope to avoid that
15:45:15 -!- pikhq has quit (Ping timeout: 245 seconds).
15:45:29 <olsner> is it ok to read more than the first line? can you seek?
15:45:47 -!- pikhq has joined.
15:45:58 <ais523> olsner: I can seek, and I can over-read
15:46:14 <ais523> also I have an advisory read lock on the file and everything that touches the file respects advisory locks
15:46:21 <ais523> so I can guarantee it won't change while I'm reading
15:46:24 <olsner> hmm, how about mmap and strchr/memchr?
15:46:39 <ais523> huh, I hadn't thought of that at all
15:46:47 <ais523> although, this should work on Windows too
15:46:58 <ais523> I believe there's some sort of compatibility layer somewhere translating lseek, read, etc
15:47:07 <ais523> but it almost certainly doesn't translate mmap
15:47:36 <HackEgo> olist 934: shachaf oerjan Sgeo FireFly
15:48:36 <olsner> Windows certainly can mmap files, but it probably doesn't have a function called mmap
15:48:55 <ais523> yeah, it's not a case of "can Windows do that", but "can I do that without writing a bunch of Windows-specific code"
15:50:48 <olsner> hmm, why can't you just use stdio though? (not that stdio has a portable and safe getline function that I know of)
15:52:07 <ais523> olsner: because there's a huge amount of jumping around seeking and locking and unlocking and so on, stdio buffering would screw this up
15:54:41 <Gregor> Just read into a growing buffer and stop when strchr('\n') isn't -1.
15:56:20 <ais523> yeah, that's what I was planning
16:00:24 <Gregor> Also, surely fgets on Windows has correct behavior?
16:00:43 <Gregor> Oh, without using stdio.
16:00:51 <Gregor> So, reimplementing stdio for no good reason.
16:01:16 <ais523> Gregor: yeah, pretty much
16:01:34 <ais523> or, "reimplementing stdio because it doesn't like files unexpectedly shrinking while you're writing to them"
16:02:07 * Gregor thinks for a second...
16:02:26 <Gregor> If a file unexpectedly shrinks while you're writing to it, stdio's correct behavior should be to re-expand the file with zeroes.
16:03:07 <ais523> Gregor: re-expanding the file with zeroes
16:03:22 <ais523> whereas the behaviour that I /want/ is to abort the write
16:03:25 <ais523> which implies no buffering
16:03:33 <ais523> because I need to know what has and hasn't been written at any given point
16:08:15 <ais523> or put it this way: the previous code used stdio, it kept failing in bizarre ways because trying to keep track of buffering on top of everything else was just too much
16:19:11 <mroman_> Does the halting problem imply, that there actually is a program where it is undecidable if it halts or not?
16:19:20 <mroman_> Or does it just state, that no algorithm can decide it
16:19:47 <mroman_> or are those statements equivalent?
16:21:44 <int-e> usually one proofs that for any would be decision procedure for the halting program there is some input which it gets wrong (either giving a wrong answer or failing to terminate).
16:23:12 <Gregor> mroman_: In a sense, whether those statements are equivalent is philosophical. If you believe that humans are super-Turing, then they're not. If humans are not super-Turing, then they are.
16:24:47 <ais523> mroman_: Gödel's Theorem actually does explicitly construct an undecidable program
16:37:14 <mroman_> I'm searching for a proof of the halting problem that is actually understandable :)
16:37:27 <mroman_> and I haven't found a single one
16:38:36 <mroman_> The ones that don't make use of crazy math feed the program to itself
16:39:19 <mroman_> L(Q,x) halts, if Q(x) does not and does not halt, if Q(x) halts
16:39:29 <mroman_> then L(L,L) and stuff goes boom
16:39:48 <mroman_> and my brain just keeps yelling: How the fuck would that work
16:40:18 <myname> where's the problem with it
16:40:29 <myname> if L(L) halts, L(L) does not halt
16:40:33 <mroman_> L computes whether Q halts for Input X
16:40:51 <mroman_> so L(L,L) computes whether L(L) halts
16:41:04 <mroman_> but what's the input of the second L?
16:41:38 <myname> the second L is actually the description of L
16:41:46 <myname> so it does not need any input
16:43:00 <mroman_> Yeah. But what are you actually deciding then?
16:43:19 <mroman_> Assuming f(x) is a function
16:43:35 <int-e> L(Q,x) should test whether Q(x,x) halts.
16:43:36 <mroman_> and L(f,x) tries to state whether f(x) can actually be computed
16:43:58 <int-e> then L(L,L) does test termination of L(L,L) and all works out as claimed.
16:45:02 -!- Sgeo has quit (Read error: Connection reset by peer).
16:45:13 <myname> it is not actually f(f), it is more like f(written down definition of f)
16:45:32 <myname> as i.e. gödel number or the like
16:45:35 <ais523> yeah, the "how do you write down the definition of f" is the main mathematical content of the proof
16:45:36 <mroman_> I have no idea what that definition is supposed to be
16:45:39 <ais523> everything else is pretty trivial
16:45:47 <mroman_> The definition does not have any input
16:46:06 <myname> the defintion is just a number or a string or whatever
16:46:15 <mroman_> so for what input are you actually trying to calculate whether it halts or not
16:46:16 <myname> and as such does not need any kind of input
16:47:06 <myname> at L(Q,x) you test Q with input x,x as int-e stated
16:47:16 <myname> so it will never ever come to f(f) there
16:48:15 <myname> so, if L(L,L) halts, L(L,L) does not halt
16:48:21 <myname> therefore, L cannot exist
16:48:40 <FreeFull> What if L is like universes in type theory?
16:49:20 <FreeFull> So you would actually have L_i(L_i-1,L_i-1)
16:49:27 <FreeFull> I don't know if that would help
16:49:27 <ais523> bleh… actually what I really want is a scanf that gives pointers into the original string and decodes base 64 and decompresses zlib
16:50:08 <mroman_> What if I assume that there are two turing machines that are able to decide the halting problem
16:50:15 <ais523> I don't know what i Want
16:50:17 <FreeFull> ais523: What you actually want is a program that writes all your software for you, as you want it
16:50:24 <FreeFull> Even if you don't know how you want it
16:50:25 <mroman_> and then do L(G,G) and G(L,L)
16:50:28 <ais523> FreeFull: yeah, that'd be nice, but it'd take too long to impl
16:50:53 <FreeFull> Well, you wouldn't be the one to impl it
16:51:25 <myname> L(G,G) would by definition be G(G,G)
16:52:32 <myname> L(G,L) would be G(L,L) would be L(L,L)
16:53:16 <int-e> mroman_: you can play such tricks, but you need more arguments, L(G,L) and G(G,L), where L(G,L) tests whether G(G,L) terminates and G(G,L) tests where L(G,L) terminates.
16:53:59 <int-e> and these things become tedious soon.
16:55:43 -!- Sgeo has joined.
17:27:13 -!- oerjan has joined.
17:38:13 <mroman_> http://codepad.org/Z6Z4ZT3b <- you can also argue that way
17:38:26 <mroman_> but it leaves unclear whether P(i) actually exists or not
17:38:53 <mroman_> but we know, that if it does, that no program can decide that stuff
17:41:10 <myname> "i put something undecideable into something that does decides everything, therefore it can't exists" sounds pretty strange
17:46:34 -!- doesthiswork has joined.
17:48:17 <oerjan> elan's dad is starting to get annoying.
17:50:07 <oerjan> ais523: you managed to make an `olist that wasn't duplicate, this is _highly_ irregular hth
17:51:15 -!- SingingBoyo has joined.
17:51:45 <FireFly> luckily they seem to be on their way of his subplot
17:52:04 <oerjan> FireFly: um did you even _read_ today's update.
17:52:36 <int-e> what are you reading?
17:54:31 <int-e> I have a bookmark on #726, haven't read it since.
17:56:14 <ais523> oerjan: well the forum thread was only at 4 pages, and I'd checked like half an hour before
18:01:12 <oerjan> <mroman_> but what's the input of the second L? <-- i think you are a bit confused. L should only take one argument, and L(Q) should decide whether Q(Q) halts.
18:02:04 <oerjan> (and do the opposite.)
18:02:44 -!- doesthiswork has quit (Quit: Leaving.).
18:02:50 <oerjan> it's the initially assumed halting function which takes two arguments, not L.
18:04:29 <oerjan> it's quite possible your source was confused about this, wouldn't be the first time.
18:13:48 <oerjan> <FreeFull> What if L is like universes in type theory? <-- then you get lots of nice proofs about separation in complexity hierarchies.
18:14:06 <Slereah> I wanted to ask a question
18:14:18 <Slereah> Can we find my question using some sort of programming trick
18:14:28 <oerjan> good, good, as i should be leaving now.
18:14:31 -!- oerjan has quit (Quit: leaving).
18:14:48 <FreeFull> oerjan: And of course you always need a higher hierarchy to prove something about the one right below
18:17:38 <ais523> found on Reddit: someone's trying to translate Haskell's Prelude into pure GNU Make (i.e. no shelling out or scripting): https://github.com/PiPeep/prelude.mk/blob/master/prelude.mk
18:18:06 <ais523> although it seems pretty partial atm
18:21:44 <ais523> I'd probably upvote it if it were more complete
18:21:59 <ais523> like, how does PiPeep plan to handle lists of lists, for instance?
18:36:15 -!- Phantom__Hoover has joined.
18:38:19 -!- conehead has joined.
18:39:39 -!- Phantom_Hoover has quit (Ping timeout: 246 seconds).
18:43:06 -!- muskrat has joined.
19:17:02 -!- conehead has quit (Ping timeout: 240 seconds).
19:26:07 -!- Taneb has joined.
19:32:36 -!- ^v has joined.
19:43:38 -!- doesthiswork has joined.
19:45:00 -!- muskrat has quit (Quit: Leaving).
19:51:51 -!- Phantom_Hoover has joined.
19:52:58 -!- Phantom__Hoover has quit (Ping timeout: 246 seconds).
20:03:54 <Slereah> Is there a language that is directly based on set theory?
20:04:09 <Slereah> LISP is pretty close but it is not really
20:04:23 <Slereah> And there are a lot of languages based on axiomatic systems
20:04:32 <Slereah> But I can't think of any from ZFC
20:06:07 <Bike> most sets being unrepresentable doesn't hel.
20:07:03 <Slereah> Well all sets are representable in ZF-NC
20:07:47 <Slereah> But well, µ-recursive functions are based on some axiomatic system of integers
20:07:58 <Slereah> Logical combinators as well, and lambda calculus
20:08:09 <Slereah> So I'm wondering if there's any for some flavor of set theory
20:08:31 <Bike> http://en.wikipedia.org/wiki/Axiom_of_choice#Statements_consistent_with_the_negation_of_AC iunno
20:11:02 <Slereah> I think constructive set theory requires both the negation of choice and of the excluded middle
20:11:04 <Bike> zfc isn't constructive, i guess is what i mean.
20:12:24 <Bike> http://plato.stanford.edu/entries/set-theory-constructive/ oh good.
20:12:57 -!- carado has joined.
20:13:28 <Slereah> A lot of people feel strongly about axiomatic systems, which I always find weird
20:13:40 <Slereah> Like one of them is truer than another
20:13:48 <Sgeo> Is there any reason I shouldn't treat Red Bull as slightly weaker coffee that tastes much, much better?
20:14:33 <Slereah> People basically seem to treat it like DRUUUUGS
20:14:41 <Slereah> Also the taurine is kind of a boogey man
20:14:44 -!- keoni29 has joined.
20:14:57 <Sgeo> Does the taurine even do anything? Also, I treat coffee as a drug, so
20:15:13 <Sgeo> I drink coffee to get a specific effect
20:16:23 <Bike> well it is a drug.
20:16:44 <Sgeo> Also, I want something that has the taste of Red Bull without the caffeine, so I could drink it constantly without using it in drug mode. Technically still a drug I guess, but less dangerous to overconsume
20:17:11 <Slereah> Yeah, but I mean drug as in
20:21:44 <ais523> I always assumed Red Bull was stronger than either
20:22:15 <ais523> (in fact, I haven't drunk caffeinated drinks for like 8 years now)
20:22:30 -!- conehead has joined.
20:22:51 <Slereah> Redbull has about the same amount of caffeine as regular coffee
20:23:14 <FireFly> I'll stick to tea & occasional soft drinks
20:33:17 <keoni29> Does anyone here have experience with Forth?
20:34:00 <keoni29> I am building a microcomputer based on an eZ8 microcontroller and I have to decide what it boots into
20:34:13 <keoni29> We were thinking about Forth.
20:34:29 <keoni29> At the moment I just have this shell from which you can run programs:
20:35:38 <keoni29> http://www.youtube.com/watch?v=4DyPKRJV9PI&feature=player_embedded
20:35:56 <keoni29> This is an old version of the current firmwae
20:37:32 -!- nisstyre has quit (Quit: Leaving).
20:51:59 <ais523> keoni29: yeah, Forth seems vaguely appropriate for that
20:52:12 <ais523> its main uses are as a bootstrap language, and on embedded systems, and for making DSLs
20:52:20 <ais523> and your use seems to fit into two out of three of those categories
20:57:31 -!- conehead has quit (Ping timeout: 254 seconds).
20:57:39 <keoni29> I already wrote routines for comparing strings with a list of tokens
20:57:52 <keoni29> So writing an interpreter would not be a huge deal
20:58:04 -!- impomatic has quit (Quit: http://corewar.co.uk).
20:58:09 <keoni29> It's all written in Z8 assembler
20:58:16 <Taneb> ais523, one of the people at my university seems to know you from b nomic
21:05:28 <ais523> Taneb: hmm, interesting
21:05:32 <ais523> there aren't that many B players
21:05:41 <ais523> what's their name? or do you have a reason to keep it secret?
21:06:23 <ais523> actually, considering the list of people who would say B rather than Agora
21:06:26 <ais523> it's a very short list indeed
21:21:15 -!- keoni29 has quit (Quit: Omnimaga.org).
21:29:59 <ais523> oh, must be James Baxter
21:30:34 <ais523> they've only posted to B 6 times, I had to search to see who it was
21:30:40 <ais523> so I'm impressed that they can remember who I am
21:31:40 <Taneb> I think he said that he had a debate with you on an interpretation of a rule?
21:34:47 <ais523> that happens a lot in nomics
21:36:02 <olsner> isn't that the only thing that happens in nomics?
21:36:12 <ais523> olsner: well atm, nothing happens at all :-(
21:36:18 <b_jonas> olsner: anything can happen in a nomic
21:36:20 <ais523> but you should see BlogNomic for a nomic that's normally based on using the rules
21:36:24 <ais523> rather than debating over them
21:36:43 -!- prooftechnique has joined.
21:40:49 <Taneb> Someone is trying to log in as me on IRC
21:41:18 <Taneb> They are not succeeding
21:41:24 <Taneb> I am pretty sure that I am actually me
21:41:43 <Bike> sock yourself just to be sure.
21:43:34 -!- L8D has quit (Ping timeout: 246 seconds).
21:45:06 -!- MindlessDrone has quit (Quit: MindlessDrone).
21:45:31 -!- muskrat has joined.
21:52:03 <b_jonas> Taneb: no, it's just that temple ghosts often have your name
21:52:41 <FireFly> I've had that happen to me occasionally
21:52:51 <ais523> b_jonas: temple? I thought it was a 1 in 7 chance for ghosts generally
21:53:03 <ais523> (or have I got the context completely wrong?)
21:53:21 <ais523> I remember because I refactored that code earlier today due to making the player's name not a global variable
21:53:28 <ais523> err, yesterday, probably
21:53:30 <b_jonas> ais523: dunno, possible, I just spend lots of time in temples going in and out and get free ghosts for it
21:53:38 <b_jonas> I don't meet many ghosts in other places
21:53:42 <ais523> b_jonas: why? is this some new farming method I've never heard of?
21:53:58 <b_jonas> no, it's just that I have to leave and enter the temple for other reasons
21:54:05 <b_jonas> the ghosts themselves are useless
21:54:11 <ais523> right, but why did you kill the priest? for conversion purposes?
21:54:49 <b_jonas> I didn't kill her! I was trying to save her! It was an accident and nothing to do with me!
21:54:58 <b_jonas> I'm not responsible for where dragons breathe!
21:55:50 <b_jonas> and it was a mistake to kill her so early, I should have donated more to her and THEN kill and sacrifice her
21:56:17 -!- muskrat has quit (Ping timeout: 252 seconds).
21:57:04 <ais523> now you're going to blame it on me, in 3.4.3 the dragon would have ripped her apart with tooth and claw rather than using the breath attack
21:57:06 <b_jonas> two mistakes, because I killed both the minetown priest and the dlvl15 priest too early, but at least the dlvl15 priest was totally necessary eventually because she was cross-aligned
21:57:21 <ais523> you still have the Valley priest left
21:57:39 <b_jonas> I'll kill her only right before the asc run
21:57:41 <ais523> now I wonder what happens if you StF a statue of a priest in the Orcustown temple
21:57:53 <b_jonas> I should have donated all the gold to all three priests
21:58:01 <b_jonas> but I messed it up, I can get the value of my gold only once now
21:58:04 <ais523> also this playstyle is alien to me, I typically don't kill priests
21:58:23 <b_jonas> also, um, can we move to #nethack4?
21:58:23 <ais523> I'm (almost always) lawful
21:58:37 <ais523> the discussion's only funny /because/ it's in the wrong channel
21:59:15 <olsner> it was way better before you managed to mention what the context was
22:10:57 <HackEgo> 112) <CakeProphet> how does a "DNA computer" work. <CakeProphet> von neumann machines? <Phantom_Hoover> CakeProphet, that's boring in the context of DNA. <Phantom_Hoover> It's just stealing the universe's work and passing it off as our own. \ 715) <elliott> then they edited their own talk page comments after someone replied to it, and edited /th
22:13:56 <quintopia> ais523: hi. got anything interesting in the works
22:15:10 <ais523> quintopia: I wanted to get a NetHack 4 release out by tomorrow
22:15:14 <ais523> but no way it's going to be ready on time
22:15:26 <ais523> unless I recruit the entire population of people who are familiar with the codebase to help out, and probably not even then
22:15:53 <HackEgo> 715) <elliott> then they edited their own talk page comments after someone replied to it, and edited /the replier's comment/ so that it made sense in context
22:16:04 <ais523> shachaf: 10th anniversary of NetHack 3.4.3
22:16:10 <quintopia> ais523: you should get zzo38 to make nethack 4 for famicom :D
22:16:30 <shachaf> ais523: Maybe aim for Apr 1.
22:17:04 <ais523> quintopia: I actually joked about NetHack for NES in the release message last April 1, although I wasn't thinking about zzo38 at the time
22:17:09 <ais523> just trying to find a platform that it obviously didn't wok on
22:20:54 <olsner> how big is nethack? how hard could it be?
22:22:54 <ais523> NetHack 3.4.3 is 206429 lines for the entire distribution; NetHack 4 (latest development version) is 150529
22:23:16 <ais523> partly due to a lot of refactoring, partly due to dropping support for various obscure platforms
22:25:21 <olsner> apparently cartridges ranged up to 1MB, so it should be reasonably possible
22:25:59 <ais523> from what I remember, it's impossible to fit the NetHack binaries and a save file onto the same (1.44MB) floppy disk
22:26:58 <Taneb> Could you make a ridiculously NES-optimized version
22:27:08 <ais523> me personally? probably not
22:27:13 <ais523> in general, it might be possible
22:27:25 <ais523> you could start with 3.1 or so and then turn all the optional features off, that'd help a bit
22:28:35 <Taneb> Also, I just realised that if homeopathy actually worked I'd be immune to caffeine by now
22:30:57 <kmc> if homeopathy worked then all the world's oceans would be incredibly powerful medicine
22:33:25 <Bike> the strongest shark poop solution in history
22:34:47 <kmc> "I don't drink water. Fish fuck in it."
22:35:37 <olsner> "I drink water. Sometimes I get thirsty."
22:36:05 <Bike> a lot more than fish fuck in it if you know what i mean *seductive wink*
22:37:54 <kmc> don't think i could manage that
22:38:16 -!- zzo38 has joined.
22:38:18 <Bike> i'm saying i'm going to fuck a jellyfish
22:41:10 <kmc> because it's so easy?
22:41:14 <shachaf> although i've been drinking lots of non-water things for the past few months
22:41:22 <kmc> (double entendre on the word "easy")
22:41:23 <shachaf> because, yes, it's so easy
22:41:46 <kmc> shachaf: which things have you been drinking?
22:42:17 <shachaf> i think mostly various things that claim to be tea
22:43:06 <shachaf> i should drink proper tea, though
22:43:32 <quintopia> Bike: don't fuck a cuttle fish tho
22:44:01 <olsner> Bike: cuddle a cuddlefish tho
22:44:19 <Bike> except it will probably vomit up ink and mucus on me
22:44:26 <kmc> they're so cute
22:44:39 <Bike> small price to pay i suppose
22:45:18 <shachaf> can i be a cuddlefish when i grow up
22:45:29 <kmc> they don't live long enough to use their smart brains :/
22:45:47 <Bike> i don't live long enough to use my smart brain either :(
22:46:01 <quintopia> kmc: you mean how they have brains that are bigger than their bodies, including their brains?
22:46:20 <shachaf> does anyone live long enough to use their smart brains
22:46:35 <Bike> a paleontologist posted that video yesterday. it made me happy
22:46:42 <Taneb> On that note, does anyone know how to use pietcreator?
22:52:09 <olsner> cuddlefish have three hearts
22:53:51 <ion> http://youtu.be/1bmrm_8Y9Bc
22:55:48 <quintopia> https://gist.github.com/Janiczek/7849921
23:00:13 <Bike> if you start with the infinite radical set equal to zero you get the same value of x.
23:00:23 <Bike> so i'm guessing the substitution is somehow invalid.
23:01:37 <Bike> like if there's no value of x that makes the radical 1, or somethin
23:06:04 <Bike> wait, what are you talking about, if x equals one than the substitution gives the square root of two.
23:08:03 <Bike> might wanna get that checked out
23:08:58 <Phantom_Hoover> as far as "How can you get 1 from nested square roots of 0?
23:08:58 <Phantom_Hoover> There is nothing accumulating that would converge to a limit or something..." you're not taking a limit here
23:09:02 <Bike> you can call the problem √(x+y) = y, in which case you end up with x = y² - y, and again y = 0 and y = 1 have the same value for x, so like whatever man.
23:09:59 <quintopia> the solution is phi (and the other negative one, eta or w/e)
23:12:43 <Bike> Phantom_Hoover: can't you make √x, √(x+√x), √(x+√(x+√x)) bla bla an actual series somehow
23:13:27 <Phantom_Hoover> sure, but there's nothing saying the limit of that series will be the same as whatever you get by twiddling the infinite expression
23:16:54 <Taneb> I kind of want to enter one of the BBC's quiz shows for some reason
23:22:27 <b_jonas> Taneb: which kind of quiz show?
23:25:33 -!- doesthiswork has quit (Read error: Connection reset by peer).
23:25:58 <kmc> i almost learned why φ is good for integer hashing but didn't learn really :/
23:25:58 <kmc> http://brpreiss.com/books/opus4/html/page214.html
23:26:33 -!- oerjan has joined.
23:33:37 <Taneb> b_jonas, I was thinking more Only Connect
23:36:46 <kmc> Dragons' Den
23:37:00 <Taneb> Phantom_Hoover, they aren't auditioning for Weakest Link!
23:37:39 -!- conehead has joined.
23:39:27 -!- pikhq has quit (Ping timeout: 260 seconds).
23:41:01 <Taneb> Phantom_Hoover, no
23:41:11 -!- pikhq has joined.
23:41:36 <Phantom_Hoover> Taneb, yes (ok no you're not enough of a twat to go on the apprentice)
23:41:59 <Taneb> (also they aren't auditioning for it at the moment)
23:42:12 <Bike> the apprentice is the show where that one rich douche yells at you right
23:42:26 <Bike> i wanna say tom cruise but that's not him
23:44:12 <kmc> is there something which is like hackathons but instead of pretending to make something useful, you just see who can write the fastest HTTP server or w/e in 24 hours
23:45:08 <oerjan> i would not think tom cruise and donald trump were very confusable
23:45:39 <oerjan> admittedly i mainly just know what they look like.
23:46:09 <elliott> Phantom_Hoover is thinking of alan sugar, presumably
23:47:45 <Bike> just another face of the platonic rich douche in the heavens
23:54:50 -!- ais523 has quit.
23:57:36 <Phantom_Hoover> Bike, it retains some entertainment value in that the people he's yelling at are deliberately selected to be wannabe rich douches who are completely unbearable, but it still wears thin fast.
23:58:35 <Bike> that just sounds really sad.
23:58:39 -!- carado has quit (Ping timeout: 252 seconds).