00:00:03 <Bike> Sgeo: as long as those other foods aren't mounds of salt
00:00:10 <zzo38> Yes make up such a Elo rankings
00:00:15 <Sgeo> Bike: does such food exist?
00:00:35 <Bike> oh, right, you usually eat incredible garbage
00:00:50 <Bike> in that case i'm gonig to say it's unsafe for you to eat anything because everything you eat is terrible
00:00:54 <Bike> just photosynthesize
00:01:22 <pikhq> Well, a potato chip is potato-flavored salt.
00:01:33 <Sgeo> On the plus side, this is not convenient food for me to make
00:01:49 <Sgeo> I don't have much in the way of clean microwave-safe containers. Boiling water in a pot now
00:02:07 <pikhq> http://sprunge.us/AjPf Unrelatedly, Lex is "clearly" the best way of producing compilers.
00:02:10 <Sgeo> So I don't think I'm going to be tempted to eat this frequently
00:02:14 <pikhq> As evidenced by this trivial case.
00:02:21 <Sgeo> Is pizza mounds of salt?
00:02:29 <Sgeo> If this isn't filling I may just go for pizza
00:02:35 <pikhq> Depends on the pizza.
00:02:41 <Sgeo> Plain cheese pizza
00:03:02 <Sgeo> Also I had a footlong Subway earlier, and two Lucky Charms bars
00:03:31 <pikhq> You worry about your nutrition a lot for someone who seems to eat random thingsw.
00:03:54 <fizzie> pikhq: Couldn't the ignore regex just be a "."?
00:04:16 <Bike> golly, it even optimizes
00:04:17 <fizzie> (For two matches of same length, the rule listed first is chosen.)
00:04:25 <pikhq> fizzie: Quite likely.
00:04:47 <fizzie> It's very possibly more efficient when it chomps more in one go, though.
00:04:49 <pikhq> This resulted from me sitting down and realizing I had literally done nothing with lex or yacc before.
00:05:39 <fizzie> http://sprunge.us/ABVC heh
00:06:00 <ais523> pikhq: there's less salt in a packet of crisps (potato chips) than there is in a slice of bread
00:06:04 <ais523> which is kind-of shocking, really
00:06:37 <pikhq> ais523: Might be true of UK crisps, but...
00:07:00 <pikhq> Though, US bread probably uses about as much salt and sugar as flour, so. :P
00:12:20 -!- Slereah__ has joined.
00:13:59 <Sgeo> Why do lactaid chewable pills taste so good?
00:14:42 <Bike> what about rocks i've painted pictures of tigers onto
00:15:13 -!- Slereah_ has quit (Ping timeout: 264 seconds).
00:19:57 -!- nooodl has left ("Leaving").
00:20:43 <Sgeo> Wonder if there exist devices that let me pretend to have a second monitor
00:21:21 <Sgeo> Plug into VGA and USB, and it will somehow let you switch between "screens" with keypresses
00:23:31 <fizzie> Sounds more like a job for a piece of software.
00:23:44 <Sgeo> Was about to say, googled it and found some software results
00:29:13 <Sgeo> This stuff is kind of soupy
00:29:20 <Sgeo> I don't like soup, maybe I should learn to liek it
00:33:39 -!- nooodl has joined.
00:35:17 <Taneb> Why are macros called macros
00:36:21 <Slereah__> Blargh, transferring things from the VM to the computer is not fun
00:37:13 <Slereah__> "FATAL: No bootable medium found! System halted."
00:37:21 <Slereah__> Well apparently I didn't do it that well
00:37:27 <lexande> Taneb: something about mapping short input sequences to longer ones?
00:37:37 <Taneb> lexande, seems likely
00:38:09 -!- JZTech101 has quit (Remote host closed the connection).
00:38:53 <Bike> still better than macrophotography
00:49:16 <Slereah__> Dang, that bootloader thing still doesn't work.
00:50:56 <Slereah__> If the 4 bytes at 0x158 are 00 00 20 00, does that mean that the boot starts at 0x200000?
00:55:01 <fizzie> In x86, assuming little-endian (so 00 00 20 00 for 0x00200000) is usually the safer bet.
00:55:22 <fizzie> (Not sure what kind of boot loader that is. It's clearly not a regular BIOS-compatible boot sector.)
00:55:45 <Slereah__> Let's try one from the osdev people
00:57:30 <Sgeo> Ugh still hungry
00:57:42 <Sgeo> After having a ball of salt (not a literal ball of salt0
00:58:52 <pikhq> Try a literal ball of salt.
00:59:00 <fizzie> Google seems to suggest that "4 bytes at 0x158" refers to VirtualBox disk image format, which sounds not all that terribly relevant for simple bootloader things, since you can use raw floppy images with VirtualBox.
01:00:06 <Slereah__> I am just following this dude's tutorial, though
01:00:45 <Slereah__> I guess I could make some fake floppies though
01:03:21 <fizzie> As for transferring things between the VM to the computer, VirtualBox's "shared folders" are quite okay.
01:03:35 <Slereah__> I couldn't make it work so far though
01:03:41 <Slereah__> i don't know how to access it inside the VM
01:05:41 <fizzie> "In a Linux guest, use the following command: mount -t vboxsf [-o OPTIONS] sharename mountpoint". I think it worked pretty much out-of-the-box after installing the guest additions. Though this was a while ago.
01:06:02 <fizzie> (Just plain ssh daemon in the guest and winscp or something in the host is reasonable too.)
01:08:44 <zzo38> I have idea of a computer game, that you would fly from the center of Pluto to the center of the Earth, without crashing into any Gas Giant planets. Such thing as software library to calculate rotations of Mars and Pluto would be needed; the program "Swiss Ephemeris" can already calculate rotations of Earth.
01:10:05 <zzo38> It can make up a real date, perhaps even a date in the future (although the present date or a date in the past works too; ephemeris data can calculate for far past and far future), and figure out what date would result in a better (most interesting/challenging) game, since that would affect position of planets.
01:13:26 -!- Tritonio has joined.
01:16:09 <Bike> that sounds boring to play, but people play Eve, so
01:16:34 -!- impomatic has quit (Quit: ChatZilla 0.9.90.1 [Firefox 27.0.1/20140212131424]).
01:16:40 <fizzie> People play Desert Bus.
01:16:43 -!- Phantom_Hoover has quit (Remote host closed the connection).
01:16:44 <zzo38> Probably it does, but it would be made more interesting by various ways
01:16:57 <zzo38> But you are correct in general
01:17:05 <zzo38> But I didn't write all about it of course!
01:17:22 <zzo38> Since it would take a long time, you could use warp drives and magical spells and so on to help
01:17:32 <zzo38> But it would complicate with a lot more things too.
01:17:45 <zzo38> Such as, going through space isn't the entire game
01:18:04 <zzo38> Since the surface and inside of planets is also important
01:18:14 <Slereah__> Welp putting it on a floppy didn't help
01:18:25 <Slereah__> Let's investigate this loading of boot more closely
01:18:42 <zzo38> Slereah__: What operating system are you using?
01:19:11 -!- Phantom_Hoover has joined.
01:19:22 <Slereah__> I guess the answer would be "none"?
01:19:31 <Slereah__> The virtual machine is on the windows though
01:19:47 -!- JZTech101 has joined.
01:19:56 <zzo38> What kind of file system is it?
01:20:52 <Slereah__> i was trying to follow this tutorial : http://www.dreamincode.net/forums/blog/150/entry-2999-using-virtualbox-as-a-bootloader-testing-environment/
01:22:36 <fizzie> Copy-pasting a boot sector into a .vdi file with a hex editor seems like a really circuitous way, compared to the conventional floppy approach.
01:22:59 <Slereah__> I tried with a floppy, but I guess it's not a very good idea to use that bootloader for it
01:23:21 <zzo38> What exactly are you trying to copy?
01:23:23 <fizzie> The bootloader should be just fine for a floppy, since it just prints a string and hangs.
01:23:59 <Slereah__> I put the binary of said bootloader on a floppy and then booted up
01:24:21 <Slereah__> But it just tells me "No systemdisk. Booting from harddisk."
01:24:44 <fizzie> I don't know/remember what VirtualBox expects from a floppy image; possibly it wants a floppy-sized file with that at the beginning.
01:25:22 <Slereah__> It does also say "Disk formatted with MagicISO 4.70 (c) 2001-04 MagicISO, Inc."
01:25:22 <zzo38> It is possible to copy disk images on Windows by using the "DefineDosDevice" API.
01:25:24 <fizzie> (Not that I know how you made the floppy image, and how you stuck the boot sector binary in either.)
01:25:35 <Slereah__> So the problem might be that MagicISO added a bunch of stuff
01:26:01 <fizzie> Did you just copy the binary as a file on the floppy?
01:26:10 <Slereah__> Yeah that's probably not a good idea
01:26:24 <fizzie> Yes, no. It needs to be the first sector on the floppy.
01:26:45 <Slereah__> I suppose it might be "load boot image"
01:28:42 <fizzie> I wouldn't know anything about MagicISO; I've usually just used dd. Quick googling suggests that the "load boot image" might be inteded for making bootable CDs (with boot floppy images), but who knows.
01:29:57 <fizzie> It's a command-line tool for copying bytes.
01:31:00 <Bike> Slereah__: you can understand why having a filesystem isn't good here, right?
01:31:17 <Slereah__> that is not a frequent feeling but I wish I had a floppy drive right now
01:31:30 <fizzie> It'd be dd if=/dev/zero of=floppy.img bs=1024 count=1440; dd if=boot.bin of=floppy.img conv=notrunc in a Linux system to make a floppy-sized image of zeros, and then put boot.bin in front.
01:31:58 <fizzie> (Or any out of a number of alternative ways, of course.)
01:32:11 -!- tromp has joined.
01:32:24 <Bike> this reminds me that i should probably understand uefi better than i do
01:33:22 <Slereah__> Well there is most certainly a file now
01:33:33 <fizzie> (cat boot.bin; dd if=/dev/zero bs=512 count=2879) > floppy.img # another reasonable way
01:33:43 <Slereah__> Isn't there supposed to be more things other than the program and 0's for a virtual floppy?
01:34:12 <fizzie> If you want to put something on the floppy, sure, but there doesn't need to.
01:34:30 <Slereah__> I mean, aren't there data around the file?
01:34:44 <Slereah__> to tell the computer "hey that's a virtual floppy file"
01:35:01 <fizzie> There can be, but I'm reasonably sure VirtualBox doesn't mind raw images.
01:35:56 <fizzie> Bike: It's kind of boring, to have an actual executable file format for things you can run at boot.
01:36:25 <Bike> heh well, i liked it more than a bios when setting up this computer
01:36:38 <fizzie> (A variant of PE, even.)
01:36:52 <Bike> the large number of efi variables was more interesting to me, though
01:37:20 <fizzie> Bah; OpenBoot is where it's at.
01:37:30 <fizzie> There's not even a Forth interpreter in the UEFI standard.
01:37:42 <Bike> "Open Firmware may be accessed through its Forth language shell interface" i see your point
01:38:18 <zzo38> Having a Forth interpreter built-in to the BIOS is a good idea.
01:38:40 <fizzie> I think I tried out a Fibonacci or some other such standard simple test program in the boot prompt of my SparcStation.
01:39:48 <fizzie> Now you can start figuring out how to go from real mode to 64-bit mode.
01:40:28 <fizzie> Sure, but you did say you want to do x86-64.
01:40:41 <Slereah__> Sure, but gotta learn to walk before I can run
01:40:45 <Slereah__> Is real mode in x64 the same as the old timey real mode?
01:40:56 <Slereah__> That is, can I use the extra registers
01:41:33 <fizzie> You can't use the extra registers. There's no way to encode them in the opcode.
01:41:56 <fizzie> Since the prefix bytes used for that are repurposed actual instructions.
01:42:22 <fizzie> You can use the 32-bit wide registers in real mode, though.
01:42:27 <Slereah__> do you still have access to the interrupts in long mode by the way?
01:43:13 <fizzie> BIOS services? No, I think those pretty much all go out of the window when you ascend to protected mode. Which you will need to do before switching to long mode.
01:44:28 <Slereah__> Is there an equivalent of interrupts?
01:44:34 <Slereah__> Or do you have to do it all yourself
01:45:07 <fizzie> Incidentally, a slightly less painful way to write bare-metal code (compared to a 512-byte bootloader that loads more code and does mode switching) can be to write what you want to run in the form of a multiboot-compatible image (which doesn't have so many size restrictions), and have something like GRUB load that.
01:47:44 <Slereah__> Like how would you go about to print a letter in protected mode
01:47:53 <zzo38> They almost always worsen the PC design instead of making improvements. The original PC design was fine, and had a better keyboard too (although it is loud, it works better in general)
01:48:13 <Slereah__> Other than loading the sprite of the letter to the memory of the VGA
01:48:34 <zzo38> You should switch it to text mode and write to the corresponding tile address, if that still works.
01:48:55 -!- yorick has quit (Remote host closed the connection).
01:48:59 <fizzie> You can also just not switch away from text mode.
01:49:26 <fizzie> (But you would write into the video memory, and have to track e.g. cursor position manually.)
01:49:51 <Slereah__> But how can I use the text mode without interrupts, is what I'm asking
01:50:20 <Slereah__> http://courses.engr.illinois.edu/ECE390/books/labmanual/realprot-diff-interrupts.html
01:50:24 <zzo38> I don't think you need interrupts to use the text mode.
01:50:53 <zzo38> You just need to write into the proper memory address.
01:50:54 <Slereah__> Usually I load the letter into ax and then call an interrupt for such a thing
01:51:17 <zzo38> You can read/write it even during rendering and at any time, and even execute code from video memory, unlike the Famicom.
01:52:08 <fizzie> The link you provided is specific to code running under DOS, with a DPMI host available, for the record.
01:53:16 <Slereah__> Now what to do with all that power
01:53:37 <Slereah__> Maybe I should just paste my pong program in there, see if it works!
01:59:12 <Slereah__> Is $ just a reference to the beginning of the program?
01:59:17 <fizzie> The same as "foo: jmp foo".
01:59:24 <fizzie> $ is the current position in NASM.
02:00:08 <fizzie> (It's just "stop here", basically.)
02:00:13 -!- kmc has left.
02:00:40 <ais523> do processors optimize a jump-to-self into a halt?
02:03:11 <Bike> Hey, has anyone seen 3d-modeled objects that cast weird shadows?
02:03:13 <fizzie> I don't think they do. Are the semantics not slightly different? (In terms of the jump instruction being modified by a DMA transfer or whatnot.)
02:03:20 <Bike> Like, I remember a cube that casts an "A" in one direction, a "B" in another...
02:04:30 <zzo38> Bike: Hofstadter's book "Godel, Escher, Bach" has picture of a object that casts shadow "G", "E", "B" in different directions.
02:04:40 <fizzie> Bike: http://www.shapeways.com/blog/archives/554-3d-printed-shadow-cloud.html is kind of like that.
02:04:46 <Bike> zzo38: oh shit, i totally forgot!
02:05:11 <zzo38> ais523: If you do it, it is wrong, depending on what the memory that it is executing from is mapped to.
02:05:28 <zzo38> It might trigger on reads or writes, or have bank switching.
02:05:29 <fizzie> (It seems to refer the GEB thing in the description.)
02:05:49 <ais523> zzo38: the processor needs to know about that nowadays, though, because modern processors do their own caching
02:07:21 <zzo38> ais523: Implicit caching is a bad idea, though.
02:08:29 <ais523> zzo38: perhaps, but it's necessary for modern computers to run at the speed they do, because most programs don't contain explicit cache control commands, and the processor is much faster than the memory it accesses
02:09:04 <zzo38> ais523: Yes, but it is a bad idea to design computers like that.
02:09:41 <zzo38> It overcomplicates and confuses too many things, and makes it unpredictable.
02:09:52 -!- augur has joined.
02:10:04 <ais523> I guess there's a constant supply of tradeoffs between complexity and speed
02:10:19 <ais523> doing optimizations in one place (like caching) may cause optimizations in other places (like self-modifying programs) to fail
02:10:48 <Slereah__> You can unroll all your fucking loops~
02:10:51 -!- augur has quit (Remote host closed the connection).
02:11:02 -!- augur has joined.
02:11:21 <zzo38> Better would be to run it at a slow speed (also saving power!) until the instruction is told to load into cache/microcode RAM, and then switch to it at a faster speed, which would make it faster than the memory, and can then access instruction and data simultaneously.
02:11:33 <ais523> loop unrolling often slows a system down nowadays, because you can fit less program in the cache if it's unrolled
02:11:43 <ais523> and correctly predicted jumps are pretty fast
02:11:51 <ais523> although I imagine it still often speeds things up
02:12:04 <zzo38> ais523: And that is a part of the bad idea too. It causes such problems.
02:12:28 <ais523> zzo38: GPUs do something like you suggested with respect to memory accesses
02:12:58 <ais523> they load into block memory (sort of like a cache) explicitly, and while waiting for the memory to load, the GPU's used to run other blocks instead
02:13:04 <ais523> so the memory access and the code run entirely in parallel
02:19:19 <zzo38> Yes, that would work well
02:20:00 -!- nisstyre has quit (Quit: WeeChat 0.4.3).
02:27:49 -!- Phantom_Hoover has quit (Read error: Connection reset by peer).
02:29:33 <ais523> however, GPUs /also/ have implicitly managed caches, like CPUs do
02:29:40 <ais523> just they're only allowed to be used for read-only data
02:29:55 <ais523> presumably to prevent synchronization problems
02:31:59 <zzo38> Well, my idea is different, make one that has no implicitly managed cache, and that the cache is only explicit.
02:32:17 -!- Sellyme has quit (Excess Flood).
02:33:43 -!- Sellyme has joined.
02:34:32 <zzo38> Same with pipelining, parallization, etc, are all only working explicitly.
02:34:55 <ais523> zzo38: are you familiar with what happened to the Itanium?
02:35:12 <ais523> it required explicit specification several things on every instruction
02:35:18 <ais523> and this probably made it fail, because compilers couldn't keep up
02:36:25 <zzo38> Then just tell it don't use such things. You can hand-code those things in if you need it.
02:39:37 <ais523> well, then it would be really slow on programs that didn't have those indications, by modern standards
02:40:08 <ais523> also, you'd be able to fit less in an explicitly managed cache than an implicit one
02:40:17 <ais523> because the cache management instructions need to be stored in the cache
02:40:51 <zzo38> Yes, but also make the microcode reprogrammable and use microcode RAM as cache.
02:41:28 <ais523> that sounds like an esolang idea
02:41:43 <ais523> a processor that loads things into cache implicitly
02:41:47 <ais523> using a certain algorithm
02:41:53 <ais523> and what happens to be in cache is interpreted as microcode
02:42:59 <zzo38> You can post them in esolang list of ideas
03:08:13 -!- conehead has quit (Quit: Computer has gone to sleep.).
03:13:12 -!- Sellyme has quit (Excess Flood).
03:13:42 -!- Sellyme has joined.
03:13:54 -!- Sellyme has quit (Excess Flood).
03:15:13 -!- Sellyme has joined.
03:19:50 -!- Sellyme has quit (Excess Flood).
03:20:44 <Taneb> Things I have done in the past few hours:
03:20:53 <Taneb> Walked to a cinema and watched the Lego movie
03:21:01 -!- Sellyme has joined.
03:21:01 -!- Sellyme has quit (Excess Flood).
03:21:12 <Taneb> Went to the tail end of a masquerade ball
03:22:13 -!- Sellyme has joined.
03:24:25 -!- Tritonio has quit (Ping timeout: 240 seconds).
03:37:26 <Taneb> Then I did some other stuff
03:48:11 -!- conehead has joined.
03:57:39 <zzo38> What kind of other stuff?
04:00:27 <Taneb> Mainly, wandering into places while wearing a suit
04:21:24 -!- pikhq has quit (Ping timeout: 245 seconds).
04:21:42 -!- pikhq has joined.
04:26:43 <zzo38> You have said that GUST OF WIND has been renamed as POKEMON CATCHER and is a very good card in the new format. Actually, it is very good card in the old format too.
04:32:52 <zzo38> It can be useful if opponent's cards have high retreat cost, and might need a lots of energy for attacking, or if you are also resisted to them. It can also be used if they are resisted to you, and if your cards attack both active cards and bench cards, then you can avoid to attack active pokemon cards in this way.
04:33:25 <zzo38> It may also be used, if you have one side card, to switch into the card that you can knock out and win right away.
04:42:46 <zzo38> This game uses extremely bad AI; they keep doing things which are advantageous to me.
04:50:50 -!- conehead has quit (Quit: Computer has gone to sleep.).
04:55:41 -!- nooodl has quit (Quit: Ik ga weg).
04:56:45 -!- nisstyre has joined.
05:04:42 -!- quintopia has quit (Remote host closed the connection).
05:04:49 -!- quintopia has joined.
05:33:54 -!- copumpkin has joined.
05:58:07 -!- tromp has quit (Remote host closed the connection).
05:58:39 -!- tromp has joined.
06:02:49 -!- tromp has quit (Ping timeout: 240 seconds).
06:13:11 -!- shikhin has joined.
07:04:42 -!- oerjan has joined.
07:40:17 -!- conehead has joined.
07:47:35 <zzo38> Or even if you have three side cards and then you can use GUST OF WIND and SWITCH and possibly win on one turn using that!
07:53:58 -!- prooftechnique has joined.
08:00:51 -!- prooftechnique has quit (Quit: leaving).
08:05:12 -!- prooftechnique has joined.
08:06:12 -!- prooftechnique has quit (Client Quit).
08:09:08 -!- prooftechnique has joined.
08:15:10 -!- MoALTz has joined.
08:20:36 -!- prooftechnique has quit (Quit: leaving).
08:26:33 -!- MindlessDrone has joined.
08:34:44 -!- itsy has quit (Ping timeout: 252 seconds).
08:53:52 <oerjan> all this gnu assembler talk i the logs made me wonder if there's an assembler called "plasma". apparently there is a 6502 VM/Assembler by that name
08:56:49 <oerjan> no particularly promising hits for liquid assembler. and solid assembler is something completely different.
08:58:31 <oerjan> and the solid graphics wiki looks completely dead.
09:00:03 -!- tromp has joined.
09:01:57 -!- tromp__ has joined.
09:04:25 -!- tromp has quit (Ping timeout: 264 seconds).
09:06:01 -!- tromp__ has quit (Ping timeout: 240 seconds).
09:57:08 -!- conehead has quit (Quit: Computer has gone to sleep.).
09:57:59 <fizzie> The Bose-Einstein condensate assembler.
10:36:36 <Jafet> http://web.cs.dal.ca/~boardman/viz/proj/EnthusiASM.pdf
10:37:20 <oerjan> that's a state of mind, not matter, Jafet
10:38:26 <Jafet> Does the mind not matter?
10:42:57 <oerjan> `addquote <Sgeo> Ugh still hungry <Sgeo> After having a ball of salt (not a literal ball of salt0 <pikhq> Try a literal ball of salt.
10:43:00 <HackEgo> 1173) <Sgeo> Ugh still hungry <Sgeo> After having a ball of salt (not a literal ball of salt0 <pikhq> Try a literal ball of salt.
11:08:06 -!- shikhout has joined.
11:11:02 -!- shikhin has quit (Ping timeout: 264 seconds).
11:11:04 -!- shikhout has changed nick to shikhin.
11:46:10 -!- nooodl has joined.
11:48:02 -!- Sellyme has quit (Excess Flood).
11:49:42 -!- Sellyme has joined.
11:57:55 -!- oerjan has quit (Quit: leaving).
12:11:55 * ais523 observes someone on Reddit claim that the Church-Turing thesis implies that typical computers can run without electricity
12:12:40 <olsner> run? most computers don't even have legs!
12:14:10 -!- Phantom_Hoover has joined.
12:16:58 <olsner> I suddenly thought I should check out Eiffel a bit, but apparently most of the tutorial is about a new process for making software the eiffel way, not about the language
13:01:37 -!- Phantom__Hoover has joined.
13:03:49 -!- Phantom_Hoover has quit (Ping timeout: 264 seconds).
13:22:26 <FreeFull> ais523: Technically, you could make an entirely mechanical computer =P
13:22:47 <FreeFull> It probably wouldn't be very efficient
13:22:57 <ais523> FreeFull: I put the "typical" in there because I knew someone was going to make that particular, inapplicable-in-context, nitpick
13:23:16 <ais523> also, is the Analytical Engine entirely mechanical? IIRC, although Babbage never built it, someone else later built one from his plans, and it did indeed work
13:23:44 <FreeFull> It is, but I think there is enough resistance that you need an electric motor to power it
13:24:27 <FreeFull> Although I guess an engine with a flywheel should work too
13:25:00 <ais523> ah no, it was the Difference Engine that got built
13:25:09 <ais523> couldn't you just use a counterweight?
13:25:35 <FreeFull> Flywheel would be better if you want to provide continuous power
13:26:55 <ais523> flywheels that powerful are really hard to make
13:26:59 <ais523> and also kind-of hard to extract energy from
13:27:13 <ais523> they have a tendency to explode if you try to store too much energy in them
13:27:24 <ais523> I'd rather have a weight attached to a spool of rope, dangling over a cliff
13:28:39 <FreeFull> A powerful electric motor would probably be best
13:31:21 <ais523> what about a petrol (gasoline) engine?
13:31:28 <ais523> actually, better, diesel
13:31:33 <ais523> so you don't need spark plugs
13:35:26 <Jafet> (Masterwork obsidian watermills, if you like)
13:36:12 -!- boily has joined.
14:12:30 -!- ais523 has quit.
14:18:27 -!- atrapado has joined.
14:31:36 -!- password2 has joined.
14:31:37 <Phantom__Hoover> 12:11:55: * ais523 observes someone on Reddit claim that the Church-Turing thesis implies that typical computers can run without electricity
14:32:01 <Phantom__Hoover> i've argued with people who seriously think the human brain is turing complete
14:32:20 <Slereah__> Well it's as complete as a regular computer
14:32:36 <Slereah__> Complete up to memory space and functioning time
14:33:36 <fizzie> Maybe they mean "Turing complete" in the sense that Turing had one too.
14:34:09 <Slereah__> Then again no physical thing is Turing complete
14:34:34 <Slereah__> Because of the bloody cosmological expansion
14:36:14 <Jafet> Murray was Turing-hard.
14:36:30 <Slereah__> Do you mean Murray the satanic skull?
14:37:07 <Slereah__> http://www.youtube.com/watch?v=8FJSGe6MVpU
14:39:58 -!- tromp has joined.
14:53:03 -!- boily has quit (Quit: SHUFFLED CHICKEN).
15:05:11 <int-e> `ddg anagram chicken
15:05:12 <HackEgo> /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: ddg: not found
15:05:28 <int-e> `duck anagram chicken
15:05:28 <HackEgo> /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: duck: not found
15:06:05 <int-e> oh, wait. that must be the wrong bot, since it doesn't have easy internet access. But which was it?
15:06:10 <int-e> ~duck anagram chicken
15:09:45 <int-e> The bots are abandoning us! What do they know that we don't?
15:13:21 -!- yorick has joined.
15:13:24 <int-e> Wow, the April fool's day jokes are starting already? http://www.theinquirer.net/inquirer/news/2331400/the-paper-printer-will-be-dead-in-four-years-says-ieee
15:43:44 <fizzie> fungot: duck anagram chicken
15:43:45 <fungot> fizzie: i turned my flashlight to the corner where the downward slant met the inward slant. she seemed to crystallize at a point where the curwen farm, where they had never seen the man, but had difficulty. then in the end the officials took no action, but carefully set down the new york tribune, in which mr. merritt in curwen's farmhouse more than a week.
16:00:14 -!- copumpkin has quit (Quit: My MacBook Pro has gone to sleep. ZZZzzz…).
16:01:34 <olsner> ah, you tried that already... anyway, metasepia is only here when boily is
16:17:14 <fungot> Jafet: " he built him a house in the hills had sounded, and all round on the north rim of the plateau near the crest whence i had first reached the central chamber as well, he might show me something rather unusual something a bit hysterical in his occasional exhilaration, while his moods of fright and depression were altogether too frequent. the old bookseller had told him about the fright of the dunwich people by giving any h
16:18:30 <int-e> fungot should write a novel
16:18:30 <fungot> int-e: unseen fingers grip me... ghost fingers that lack the physical strength to force me to
16:19:09 <int-e> which we can then give to children in order to train them in fungot's style.
16:19:09 <fungot> int-e: the fnord attempt was a fnord cut of a hideous cult of nocturnal worshippers whose strange customs descended from fnord and laboratory supply houses.
16:19:15 <Jafet> fungot's style of writing seems familiar.
16:19:15 <fungot> Jafet: document modified:
16:19:38 <Jafet> But now a fungot twist
16:19:38 <fungot> Jafet: " to my ancestor, gilbert de la poer, and i answered. he remembered one particular village of the creatures seemed not quite usual in their aspects and motions, and making it more and more
16:20:11 <int-e> wait, who did s/Cthulhu/fnord/g there?
16:20:33 <int-e> (That's my association at least.)
16:22:12 <olsner> fungot knows better than to mention fnord by its true name(s)
16:22:12 <fungot> olsner: but as the weeks passed without further disclosures there began to float a picture of what had disappeared, whilst there had formed on the chest a very peculiar specimen. the proportions of its body seemed slightly altered in a queer antarctic haze such a haze, perhaps, which the maniac ( if maniac he were) yearned to avenge.
16:24:36 <Slereah__> Gonna put some COOL GRAPHICS in my boot man
16:25:05 <Slereah__> But for now, I just put it in SUPER VGA mode and filled the screen white
16:32:26 <Slereah__> Now to fill it with amazing graphics
16:40:23 -!- blotter has joined.
16:43:31 -!- blotter_ has joined.
16:44:11 -!- blotter_ has quit (Client Quit).
16:44:24 -!- blotter has quit (Client Quit).
16:44:42 -!- blitter64 has joined.
16:48:52 <int-e> so "check elfish fund" is an anagram of "shuffled chicken".
16:50:23 <int-e> so is "filched hen fucks", and that ends my session with the internet anagram finder.
16:51:53 <Slereah__> Rand Paul is an anagram of PANDA URL
16:53:34 -!- variable has changed nick to constant.
16:54:24 <int-e> http://wordsmith.org/anagram/hof.html has a couple of cute ones
17:06:41 -!- conehead has joined.
17:08:17 -!- shikhout has joined.
17:10:48 -!- oerjan has joined.
17:11:38 -!- shikhin has quit (Ping timeout: 264 seconds).
17:11:38 -!- shikhout has changed nick to shikhin.
17:15:21 <oerjan> <FreeFull> A powerful electric motor would probably be best <-- clearly only a steam engine would be proper hth
17:22:15 -!- password2 has quit (Read error: Connection reset by peer).
17:32:32 -!- tromp has quit (Remote host closed the connection).
17:41:37 -!- blitter64 has quit (Ping timeout: 240 seconds).
17:43:17 <Sgeo> Is it unethical to buy a VIP for WorldsPlayer? The company is a bit of a patent troll
17:44:55 <oerjan> itt Sgeo ethically gets sued
17:45:14 <fizzie> Internet Relay Ethics.
17:45:22 <Taneb> I am feeling a bit down today
17:45:51 * oerjan ties helium balloons to Taneb
17:46:03 <Taneb> Now I am stuck to the ceiling
17:46:08 <Taneb> Typing is somewhat difficult
17:46:15 <oerjan> but you're no longer down, right?
17:46:31 <Taneb> Well, I am no longer literally down
17:47:20 <Taneb> Metaphorically I still feel somewhat down
17:47:22 <int-e> A little tied up there?
17:47:27 <Taneb> Although less so than before
17:48:45 -!- Phantom__Hoover has quit (Ping timeout: 252 seconds).
18:02:12 -!- Phantom__Hoover has joined.
18:02:12 -!- Phantom__Hoover has quit (Changing host).
18:02:12 -!- Phantom__Hoover has joined.
18:09:09 -!- atslash has joined.
18:13:08 <quintopia> the internet is moving too fast today
18:15:10 <oerjan> i realized resPairate-exactly-2 has a solvable halting problem
18:15:22 <oerjan> that is, if you use only (n,2) pairs
18:15:42 -!- Phantom__Hoover has quit (Ping timeout: 252 seconds).
18:16:04 <quintopia> really? i cannot fathom what that means in the context of an ever-growing thing
18:16:07 <Sgeo> http://www.reddit.com/r/SubredditDrama/comments/1zc63k/who_is_to_blame_for_mtgox_collapsing_and_losing/
18:16:42 <oerjan> quintopia: it's not _necesarily_ ever-growing. you can have (0,2) pairs.
18:17:22 <quintopia> right, but all the programs that do interesting things will clearly be the ones that get ever longing, constantly dragging around and duplicating useless chaff
18:18:11 <Sgeo> Will Haskell ever be fully dependently-typed?
18:18:17 <Sgeo> I can hardly look at Haskell when I see Idris
18:18:22 <Sgeo> Even for more trivial things
18:18:42 <Sgeo> Like instead of type synonyms, just a simple definition
18:18:49 <oerjan> Will Sgeo ever find a language which doesn't disappoint him?
18:19:07 <Sgeo> No. Even if I made my own, the lack of libaries would disappoint me.
18:19:49 <quintopia> Sgeo: try opening up the wolfram language and fixing the things you don't like. probably easier than writing all those libraries yourself :D
18:20:26 <Sgeo> Hadn't heard of it until now
18:20:29 <olsner> but rename it to "sgeo language" first (or ego language?)
18:21:05 <oerjan> quintopia: anyway the thing is, if (n,2) is the largest pair in your queue, then if it survives one cycle there will be two (n,0) at most n apart, and then they can never both be deleted again.
18:21:45 <oerjan> (and the one which isn't deleted will be duplicated for the next cycle)
18:22:55 <oerjan> and i think this also means the number of such close pairs can never shrink
18:23:27 <quintopia> ah, so something like conway's argument for look-and-say
18:24:15 <oerjan> well the idea of comparing those occured to me, but i haven't thought enough to say if a similar argument can work
18:25:03 <oerjan> because a lot clearly depends on alignment with the execution, unlike look-and-say
18:26:24 <oerjan> anyway also with ever-growing things you get into the trouble like with ais523's TM proof that turing-completeness is rather fuzzily defined
18:26:38 -!- shikhin has quit (Ping timeout: 264 seconds).
18:27:35 <quintopia> indeed. if computation is possible, you have to extract the answer manually. so it be called weak universality if it's there.
18:27:38 -!- shikhin has joined.
18:35:35 <oerjan> one thing you need to avoid is that any part of the queue starts growing exponentially so information cannot cross it.
18:37:27 <oerjan> which is similar to what happens to roman numeral look and say - there are no atom boundaries in the ordinary sense, but things grow faster than they can communicate so you end up with an analogous process
18:38:45 <oerjan> hm i suppose with resPairate=2 you'd still have the question of whether you could communicate using the alignments
18:39:19 <oerjan> which _might_ work even across something exponentially growing.
18:40:22 <Sgeo> quintopia: why would Wolfram Language have a lot of libraries?
18:40:50 <oerjan> Sgeo: the libraries are called "Mathematica"
18:41:22 <Sgeo> Does Mathematica have libraries for processing WSDLs? IRC clients?
18:42:20 <oerjan> it has everything, don't you know
18:43:42 <fizzie> Did Wikipedia(/MediaWiki) have a "git blame" style feature where it annotates the lines of the current version of the page and shows who last edited those lines?
18:45:05 <fizzie> (I'm wondering who added a barely disguised SatelliteDirect ad -- http://www.oscarslive2014.com/ -- on top of the "Official websites" external links section.)
18:45:52 <oerjan> i just use binary search to find such things
18:46:12 <oerjan> (although i suppose i'd be interested to know the answer, too)
18:46:46 <fizzie> http://en.wikipedia.org/wiki/86th_Academy_Awards
18:47:37 <fizzie> I've found a couple of edits changing it so far. (It used to have "Watch Oscars Live On Your PC" as the link title, which sounds a lot more ad-y than the current.)
18:48:25 <fizzie> And it's come and gone too.
18:49:00 <oerjan> eek, i've not used binary search on pages with _that_ large a history before.
18:49:10 <fizzie> Seems to be different people adding it with different URLs every now and then.
18:49:17 <fizzie> The one I saw was already reverted, too.
18:49:47 <fizzie> (It was added by the second-latest edit, anyway.)
18:51:37 <fizzie> (By Mr. 176.42.136.235, who coincidentally enough is from Turkey, like whoever registered the domain the link was to.)
18:52:32 <fizzie> Should probably go ahead and press that undo button myself. I mean, it is the encyclopedia *everyone* can edit.
18:54:08 -!- JZTech101 has quit (Quit: Now what?).
18:55:14 <fizzie> That was a good thing.
18:55:25 <fizzie> Because I have no idea what my Wikipedia user account is.
18:55:38 <fizzie> And appearing as an IP address is just so pedestrian, you know.
18:56:44 <oerjan> e also added the link to "Academy Awards", but that was already reverted.
18:59:35 <oerjan> and the same ip did a completely unrelated, but very dubious set of edits to Samuel Holmén
18:59:54 <oerjan> but that was months ago
19:02:44 <oerjan> surely you must be https://en.wikipedia.org/wiki/User:Fizzi830
19:02:56 <fizzie> I have no idea what the "(streamable)" in their link in the "Academy Awards" meant.
19:03:30 <fizzie> I like the talk page of that user page.
19:03:37 -!- Guest73663 has joined.
19:04:33 <fizzie> Or, I mean... I am like, SO sure it's my page. UGH! TTYL!
19:06:54 -!- Guest73663 has changed nick to JZTech101.
19:06:56 <oerjan> well if you have a wikipedia account, you've not made it under any obvious name
19:07:12 -!- JZTech101 has quit (Changing host).
19:07:12 -!- JZTech101 has joined.
19:07:29 <fizzie> Are you sure I'm not Fizzie-zombie23, editing "Talk:Mayday (Taiwanese band)" and "Sokka"?
19:07:41 <fizzie> (Perhaps I didn't have an account at all.)
19:09:06 <oerjan> or that you never made a user page
19:09:35 <fizzie> I was looking at Special:ListUsers for myself.
19:11:03 -!- Phantom_Hoover has joined.
19:11:13 <fizzie> Perhaps I was Fizzieissie, whose only contribution was to change the Portuguese translation of Locomotive from "Locomotiva" to "vaginia", which does sound slightly suspicious.
19:13:20 -!- JZTech101 has quit (Quit: Now what?).
19:13:30 <Sgeo> Gardening ad playing. "Wipe out this" depicting a dandelion
19:13:34 <Sgeo> But it looks so pretty :(
19:14:21 -!- JZTech102 has joined.
19:16:08 <Sgeo> Which I keep spelling dandeloin
19:16:21 <Sgeo> Unless I try to type dandeloin in which case I type dandelion
19:18:04 <Sgeo> Wikipedia says it's a "beneficial weed", which makes me more confused. I assume the ad had them removing it because it's a weed (I knew that much), but if it's beneficial..
19:18:34 <oerjan> hmph, changing ListUsers to sort by creation date makes giving a username prefix completely useless.
19:20:25 <fizzie> I presume it's all about what you want in your garden.
19:21:42 -!- JZTech102 has changed nick to JZTech101.
19:21:42 -!- JZTech101 has quit (Changing host).
19:21:42 -!- JZTech101 has joined.
19:22:20 <pikhq> Ah, it's a nitrogen fixer.
19:22:27 <pikhq> Among other things.
19:22:52 <pikhq> Also, it's edible. Bitter, but hey.
19:34:58 -!- shikhin has changed nick to sad.
19:36:17 -!- sad has changed nick to shikhin.
19:41:07 <Sgeo> "A phone touted as the first to put privacy and security ahead of all other considerations launched at a packed event at Mobile World Congress in Barcelona, Spain, today."
19:41:17 <Sgeo> So, a paperweight?
19:44:23 <pikhq> I dunno man, could be bugged.
19:45:42 <oerjan> `addquote <Sgeo> "A phone touted as the first to put privacy and security ahead of all other considerations launched at a packed event at Mobile World Congress in Barcelona, Spain, today." <Sgeo> So, a paperweight? <FreeFull> Yes <pikhq> I dunno man, could be bugged.
19:45:44 <HackEgo> 1174) <Sgeo> "A phone touted as the first to put privacy and security ahead of all other considerations launched at a packed event at Mobile World Congress in Barcelona, Spain, today." <Sgeo> So, a paperweight? <FreeFull> Yes <pikhq> I dunno man, could be bugged.
20:14:10 <zzo38> In the computer game I described (which could be called "Flying to the center of the Earth"), of course you would need to go through a lot of passages just getting through the center of the planets to the outside and to the inside. So, those are three phases; and then once you reach the center, it still isn't finished yet so it is a fourth phase. Each phase can affect the future phases too in various ways.
20:16:39 <zzo38> And in the final phase, once you are in the center of the Earth, if you could optionally enable listening on a TCP port so that it can be multi-player, other players can try to stop you, to help you, or to be a double-agent.
20:18:41 -!- atrapado has quit (Quit: Leaving).
20:18:43 <zzo38> So, it is like four games in one that interfere with each other, I suppose like how Ultimate Gipf does something like that
20:28:19 <ion> "You just don't in the 21st century behave in 19th century fashion by invading another country on completely trumped up pre-text," Kerry told the CBS program "Face the Nation." http://www.reuters.com/article/2014/03/02/us-ukraine-crisis-usa-kerry-idUSBREA210DG20140302
20:29:02 <Bike> what are you talking about, defending ethnic russians is a totally sensible reason
20:32:02 <Taneb> I keep forgetting that the 19th century was two centuries ago
20:32:03 <int-e> Oh did Russia forget to tell the US about the proof that the Ukraine has weapons of mass destruction?
20:34:46 <zzo38> What kind of weapons of mass destructions?
20:34:57 <pikhq> Politically convenient ones.
20:35:09 <int-e> like the US had in 2001.
20:35:25 <int-e> (Which was in the 21st century. That's why I thought of it.)
20:43:09 <ion> This article documents a current event. Information may change rapidly as the event progresses. (March 2014) http://en.wikipedia.org/wiki/World_War_III
20:45:16 <fizzie> "Russia invaded Ukraine meaning WW3 has begun, prepare yourself for war."
20:45:23 <fizzie> From a since-reverted version of that page.
20:46:09 <Bike> if we're going for dumb wikivandalism i prefer the one on G8 "The Group of Eight (G8) is a forum for the governments of seven leading industrialized democracies, plus Russia."
20:46:27 <fizzie> Welcome to Wikipedia, the free encyclopedia that anyone can edit.
20:47:40 -!- yorick has quit (Remote host closed the connection).
20:50:33 -!- yorick has joined.
20:51:06 <fizzie> (Whoever reverted it apparently didn't notice the current-event template.)
20:58:38 <oerjan> well i'm _not_ going to revert it.
20:59:56 -!- MindlessDrone has quit (Quit: MindlessDrone).
21:01:20 <oerjan> well the G8 one was reverted, naturally by a russian IP.
21:15:03 -!- blitter64 has joined.
21:23:42 <Bike> "unsigned char y = -1" nooooo
21:26:23 <quintopia> is there a language where ÿ is a legal identifier
21:35:04 <Sgeo> > :t [Vect 5 String, String, Type]
21:35:05 <lambdabot> <hint>:1:1: parse error on input `:'
21:35:36 <Sgeo> ^^ doesn't work with the Idris bot either
21:35:42 <Sgeo> <Sgeo> > :t with Vect [Vect 5 String, String, Type]
21:35:42 <Sgeo> <idris-ircslave> [Vect (fromInteger 5) String, String, Type] : Vect 3 Type
21:38:11 -!- Slereah_ has joined.
21:38:12 <Sgeo> Aww, the colors didn't copy
21:38:25 -!- atslash has quit (Ping timeout: 240 seconds).
21:38:27 <Sgeo> And Quassel isn't showing me the mode of the channel
21:38:37 -!- atslash has joined.
21:39:49 -!- Slereah__ has quit (Ping timeout: 252 seconds).
21:44:01 -!- Phantom_Hoover has quit (Ping timeout: 264 seconds).
21:49:57 <Sgeo> So no color stripping
21:50:38 <Sgeo> I may turn into a language bot collector
21:50:50 <Bike> that sounds like something you would do.
21:51:23 <Sgeo> I think I'll store my collection in #esoteric
21:51:50 <quintopia> everyone here wants access to all languages all the time
21:52:37 <Sgeo> Wonder if it would just be easier to install an Idris interpreter in HackEgo
21:54:44 <quintopia> but then how will we make the population of this channel over 50% bots in the longrun?
21:55:04 <Bike> yeah sgeo, think this through. take your position as bot steward seriously.
22:00:00 -!- Phantom_Hoover has joined.
22:00:51 -!- MoALTz has quit (Quit: Leaving).
22:02:36 -!- Phantom__Hoover has joined.
22:04:42 -!- Phantom_Hoover has quit (Ping timeout: 244 seconds).
22:12:04 -!- eighteighteighte has joined.
22:13:58 -!- eighteighteighte has left.
22:20:13 -!- nisstyre has quit (Quit: WeeChat 0.4.3).
22:20:37 -!- Bicyclidine has joined.
22:29:43 <Sgeo> So, person who runs Idris bot says e can add #esoteric to the list
22:29:46 <Sgeo> Should I do it?
22:32:47 <Sgeo> Maybe not, the prefix (> ) is hardcoded
22:32:50 <ion> The channel traditionally has a very anti-bot sentiment, that would be groundbreaking.
22:33:00 <Sgeo> Could lead to some hilarity with lambdabot
22:33:15 <lambdabot> "*Exception: *Exception: *Exception: *Exception: *Exception: *Exception: *Ex...
22:33:28 <int-e> I can make lambdabot leave. *shrugs*
22:33:39 <ion> Everyone simply needs to start using the intersection of Haskell and Idris.
22:33:46 <Sgeo> int-e: I think people value a Haskell bot over Idris bot
22:34:25 <int-e> @bf +++++[>+++++++<-].
22:34:37 <HackEgo> 23coin attocoin tricoin subtlecoin eccecoin cvhcoin iyurduebagcoin thromacoin brailitcoin doubcoin acrowcoin cencoin rhyoldcoin combacoin tendcoin glycoin whirancoin stoplicoin duplearmynacoin hanifiecoin
22:34:49 <lambdabot> bf <expr>. Evaluate a brainf*ck expression
22:35:21 <ion> brainfasteriskck
22:35:51 -!- reynir has joined.
22:36:06 <int-e> @bf +++++[>+++++++<-]>.
22:36:11 <Sgeo> `welcome reynir
22:36:12 <HackEgo> reynir: Welcome to the international hub for esoteric programming language design and deployment! For more information, check out our wiki: <http://esolangs.org/wiki/Main_Page>. (For the other kind of esoterica, try #esoteric on irc.dal.net.)
22:36:16 <int-e> ok, that was stupid. so @bf works :)
22:36:20 <blitter64> bf is a great demonstration of how a languages popularity is tied to it's name
22:36:45 -!- Melvar has joined.
22:36:49 <int-e> but it filters control characters.
22:37:02 <int-e> and I'm not sure what it does with , ...
22:37:12 <ion> @bf print "Done."
22:37:29 -!- idris-ircslave has joined.
22:37:51 <Sgeo> > :t with Vect [Int, Bool, Vect 3 Type]
22:37:52 <lambdabot> <hint>:1:1: parse error on input `:'
22:38:50 <int-e> Unfortunately, I don't think that lambdabot's prefix can be configured on a per channel basis.
22:39:13 <Sgeo> idris-ircslave's prefix is hardcoded too. If I ran it myself, could change it
22:39:41 <Sgeo> Melvar: can idris-ircslave do IO?
22:40:27 <Sgeo> > the Bool "Hello"
22:40:28 <lambdabot> Not in scope: `the'Not in scope: data constructor `Bool'
22:41:10 <Sgeo> > the (so True) oh
22:41:11 <lambdabot> Not in scope: `the'Not in scope: `so'
22:41:11 <lambdabot> `s' (imported from Debug.SimpleReflect),
22:41:12 <lambdabot> `o' (imported from Debug.SimpleReflect),
22:41:14 <lambdabot> `iso' (imported from Control.Lens)Not in scope: `oh'
22:41:21 <ion> idris-ircslave seems to be faster than lambdabot. Idris is faster than Haskell. QED
22:41:27 <Melvar> I filter commands, only a few I deem safe are allowed.
22:41:29 <reynir> what is lambdabot doing
22:41:46 <Melvar> ion: Both Idris itself and the Bot are written in Haskell.
22:41:47 <Sgeo> s and o are... things, I guess?
22:42:03 <Sgeo> > the (so False) oh
22:42:04 <lambdabot> Not in scope: `the'Not in scope: `so'
22:42:04 <lambdabot> `s' (imported from Debug.SimpleReflect),
22:42:04 <lambdabot> `o' (imported from Debug.SimpleReflect),
22:42:04 <lambdabot> `iso' (imported from Control.Lens)Not in scope: `oh'
22:42:18 <oerjan> so, you write a bot in haskell and make its prefix collide with _the_ haskellbot? sounds smart.
22:42:20 <int-e> ok, @bf does not support ,.
22:42:22 <ion> melvar: Re: Idris, i know. :-)
22:42:54 <Melvar> oerjan: I started out with ‣ , but people complained.
22:43:45 <ion> melvar: You shouldn’t listen to people, they tend to be wrong.
22:43:48 <int-e> heh, how about 'I ' (short for 'idris') *whistles innocently*
22:44:47 <HackEgo> [U+2023 TRIANGULAR BULLET]
22:45:02 <Sgeo> Doesn't look very triangular to me
22:45:09 <Sgeo> Looks like a thick |
22:45:16 <oerjan> int-e: technically + is taken by thutubot, but it's never here so...
22:46:05 <oerjan> idris-ircslave: > True
22:46:40 <oerjan> that's a little inconvenient
22:47:41 <reynir> > putStrLn "Hello, world! This is IO"
22:47:42 <idris-ircslave> MkIO (\w => prim_io_bind (mkForeignPrim (FFun "putStr" [FString] FUnit) "Hello, world! This is IO\n" w) (\x => prim__IO x)) : IO ()
22:49:00 <Sgeo> > :t mkForeignPrim
22:49:01 <lambdabot> <hint>:1:1: parse error on input `:'
22:49:09 <lambdabot> <hint>:1:1: parse error on input `:'
22:49:45 <Sgeo> I know the Idris tutorial has a chapter on FFI, I just haven't read it
22:50:20 <Sgeo> Also, I wish standard IO just used the neweffects library
22:50:39 <Melvar> I have the feeling you guys are going to crash the bot before long …
22:50:50 <Sgeo> It's crashable?
22:51:42 <Sgeo> If we break something, what's the chance of it being an Idris bug vs. bot bug?
22:53:13 <oerjan> > unsafePerformCrashing
22:53:14 <lambdabot> Not in scope: `unsafePerformCrashing'
22:53:32 <lambdabot> "*Exception: *Exception: *Exception: *Exception: *Exception: *Exception: *Ex...
22:53:41 -!- oerjan has quit (Quit: We'll see how many bots are alive tomorrow.).
22:53:45 <Melvar> Well, it depends. The easiest way is to ask for something that diverges. I have no way available to filter that, so the idris process hangs or stackoverflows or whatever, and the bot has no real way to deal with that except terminating.
22:54:45 <Sgeo> Just have a timeout, and if Idris doesn't respond within the timeout, restart the Idris process
22:54:54 <Melvar> (It started as a weekend hack, don’t expect too much.)
22:55:07 <Sgeo> Although some terminating functions will inevitably be terminated early
22:55:20 <Melvar> It has a timeout. The timeout terminates the bot, because I have no way to reset its state.
23:05:07 <lambdabot> <hint>:1:1: parse error on input `:'
23:05:17 <lambdabot> <hint>:1:1: parse error on input `:'
23:05:52 <Sgeo> What does casesplit do?
23:06:04 <lambdabot> <hint>:1:1: parse error on input `:'
23:06:10 <lambdabot> <hint>:1:1: parse error on input `:'
23:06:37 <Melvar> Splits cases on a pattern variable, used by editor integration.
23:06:44 <fungot> Bot prefixes: fungot ^, HackEgo `, EgoBot !, lambdabot @ or ?, thutubot +, metasepia ~, jconn ) , blsqbot !
23:07:43 <Sgeo> > :casesplit 5 xs
23:07:43 <idris-ircslave> : openFile: does not exist (No such file or directory)
23:07:44 <lambdabot> <hint>:1:1: parse error on input `:'
23:07:51 <fizzie> Look at that greedy lambdabot, grabbing two prefixes.
23:07:53 <Taneb> ?pl \x -> not (not x)
23:08:14 <int-e> @pl id `asTypeOf` not
23:08:28 <fizzie> (The opening parenthesis is free also.)
23:08:30 <Sgeo> @djinn Void -> Bool
23:08:36 -!- shikherr has joined.
23:09:02 <int-e> void has type Void -> a
23:09:20 <Sgeo> How would void be written in Idris?
23:09:31 <lambdabot> <hint>:1:1: parse error on input `:'
23:09:47 <Sgeo> Where is it defined?
23:10:19 <Melvar> It’s builtin, I believe.
23:10:44 <int-e> > iterate ('>':) "> "
23:10:45 <lambdabot> ["> ",">> ",">>> ",">>>> ",">>>>> ",">>>>>> ",">>>>>>> ",">>>>>>>> ",">>>>>>...
23:11:38 -!- shikhin has quit (Ping timeout: 264 seconds).
23:11:38 -!- shikherr has changed nick to shikhin.
23:11:42 <Melvar> > take 10 (iterate (strCons '>') "> ")
23:11:42 <idris-ircslave> ["> ", ">> ", ">>> ", ">>>> ", ">>>>> ", ">>>>>> ", ">>>>>>> ", ">>>>>>>> ", ">>>>>>>>> ", ">>>>>>>>>> "] : Vect 10 String
23:12:01 <lambdabot> <hint>:1:1: parse error on input `:'
23:12:30 <idris-ircslave> Prelude.Stream.take : (n : Nat) -> (Stream a) -> Vect n a
23:12:30 <idris-ircslave> Prelude.Vect.take : (m : Fin (S n)) -> (Vect n a) -> Vect (cast m) a
23:12:30 <lambdabot> <hint>:1:1: parse error on input `:'
23:13:22 <fizzie> I'm feeling a bit sad for poor lambdabot hitting its head on all those parse errors on input `:'.
23:13:37 * Sgeo considers reading http://www.andres-loeh.de/LambdaPi/LambdaPi.pdf
23:13:54 <Sgeo> Maybe then I'll fully understand it
23:14:43 <Sgeo> Not that I actually understand the simply-typed lambda calculus either
23:22:25 -!- tromp has joined.
23:26:00 -!- blitter64 has quit (Ping timeout: 252 seconds).
23:32:20 -!- Bicyclidine has quit (Quit: Page closed).
23:34:54 -!- Bicyclidine has joined.
23:56:32 <zzo38> I am trying to figure out how to pattern match on a list of Z-machine instructions, and on an 'inner block' of instructions. I have already partially figured it out: There is a pseudo 'branch target' instruction to help, and each opcode has flags associated with it, for one thing. (For example, IGRTR? has flags (IF_in_place|IF_pure|IF_predicate).)
23:59:37 <zzo38> Do you have advice?