00:00:23 -!- calamari has quit (Quit: Leaving).
00:03:25 <david_werecat> !bfjoust quirtle (>)*8(>[([++[-[--[-[(-)*120(-.)*16>]]]]]>)*21])*21
00:03:28 <EgoBot> Score for david_werecat_quirtle: 37.5
00:04:05 <quintopia> i'm surprised it still works so well
00:06:12 <quintopia> it shouldn't. bumps its flag by enough to beat it
00:06:28 <david_werecat> !bfjoust quirtle (>)*8(>[++[-[--[-[(-)*120(-.)*16>]]]]])*21
00:06:31 <EgoBot> Score for david_werecat_quirtle: 37.7
00:07:12 <david_werecat> Another example is david_werecat_ill_take_the_stairs
00:07:32 <quintopia> brachiation also bumped its flag, but i deleted it :P
00:10:19 <quintopia> this is why poke is such a great strategy: you can change your strategy for every tape length
00:11:25 <quintopia> (i was only bumping my flag for slow rushes, not fast rushes)
00:13:24 <david_werecat> I have an alternate strategy for ill_take_the_stairs when you change it :D
00:13:56 <quintopia> i haven't looked at i'll_take_the_stairs
00:13:59 -!- oerjan has joined.
00:15:05 <quintopia> of course, bumping flag only works on one polarity when the opponent is already on the flag
00:20:21 <david_werecat> !bfjoust quirtle (>)*8(>[++[-[--[-[(-)*122(-.)*12>]]]]])*21
00:20:23 <EgoBot> Score for david_werecat_quirtle: 39.0
00:20:41 * oerjan sporks quintopia ----(Z
00:21:13 <oerjan> by taking a couple of long walks
00:25:05 <oerjan> there is a lot of asphalt, but there is also a foresty coastal path passing almost next to this house
00:28:01 <david_werecat> !bfjoust quirtle --(>)*8(>[++[-[--[-[(-)*122(-.)*11>]]]]])*21
00:28:03 <EgoBot> Score for david_werecat_quirtle: 40.1
00:28:26 <oerjan> david_werecat: HEY NO POKEMONS ON OUR HILL
00:29:07 <EgoBot> Use: !bfjoust <program name> <program> . Scoreboard, programs, and a description of score calculation are at http://codu.org/eso/bfjoust/
00:30:10 -!- azaq23 has joined.
00:30:27 -!- azaq23 has quit (Max SendQ exceeded).
00:30:30 <oerjan> bah i don't know enough pokemon to recognize any ais523 ones
00:30:55 -!- azaq23 has joined.
00:32:33 <quintopia> slow poke is a slow version of a poke strategy
00:32:51 <quintopia> and has sprunge started working again?
00:34:01 <oerjan> quintopia: no but it's just one letter off one :P
00:34:48 <oerjan> you should use hpaste just to get any random onlookers to wonder
00:35:22 * oerjan wonders if hpaste has brainfuck syntax coloring :P
00:35:38 <quintopia> !bfjoust space_elevator http://pastebin.com/raw.php?i=P3ptiKeC
00:35:44 <EgoBot> Score for quintopia_space_elevator: 57.5
00:35:45 -!- Phantom_Hoover has quit (Quit: Leaving).
00:36:16 <oerjan> quintopia: it's the main haskell pastebin, but it supports several languages
00:36:20 <quintopia> david_werecat: okay quirtle loses now
00:36:35 <oerjan> (also gives hlints for haskell programs)
00:37:08 <david_werecat> quintopia: Note that ill_take_the_stairs still beats space_elevator
00:37:15 <oerjan> it seems you cannot write in other languages in that box
00:37:16 <lambdabot> *** "14" wn "WordNet (r) 3.0 (2006)"
00:37:17 <lambdabot> adj 1: being one more than thirteen [syn: {fourteen}, {14},
00:37:17 <lambdabot> n 1: the cardinal number that is the sum of thirteen and one
00:37:29 <quintopia> david_werecat: i haven't even looked at ill_take_the_stairs
00:38:27 <david_werecat> It's essentially quirtle with hardcoded offsets to beat space_elevator
00:38:46 <quintopia> yeah there's obviously no way to simultaneously beat quirtle and ill_take_the_stairs
00:39:26 <quintopia> except maybe something like shudderlock?
00:40:16 <quintopia> yeah looks like shudderlock beats it
00:41:03 <oerjan> it would be interesting to have a set of programs that were proved not to be simultaneously beatable
00:41:49 <quintopia> something tells me that we'll never achieve it
00:42:33 <EgoBot> Score for david_werecat_ties: 8.0
00:42:42 <oerjan> ais523 seemed to claim defence was broken, but is it broken completely or just in a rock/paper/scissors way...
00:42:43 <EgoBot> Score for david_werecat_ties: 9.8
00:43:33 <quintopia> oerjan: i'm pretty sure it's just "any defense strategy can be defeated by a rush, any rush can be defeated by a defense, and so on for any mix of strategies"
00:44:03 <oerjan> well that is not broken, i though he meant something more than that
00:44:28 <oerjan> like if there was a program which could beat any defense
00:44:53 <quintopia> that it was possible to construct a clear that would break any lock
00:45:27 <quintopia> breaking all locks does not imply defense can't be useful
00:46:04 <oerjan> any feature valuable against one opponent might weaken you against another
00:46:14 <elliott> <oerjan> ais523 seemed to claim defence was broken, but is it broken completely or just in a rock/paper/scissors way...
00:46:18 <elliott> ais523 no longer claims defence is broken
00:46:34 <elliott> coppro: nice lambdabot invocation
00:46:34 <oerjan> (which is the essence of r/p/s phenomena i think)
00:47:05 * Sgeo attempts to remember that thing that he and ais523 proved
00:47:42 <Sgeo> I think I remember the claim but not the proof
00:51:38 <quintopia> !bfjoust brachiation2 >->+++++(>[(<)*3(+)*10((<+[{}(+)*50(>--[(+)*50(>)*8(([----[+[+[+[+[+[+[+[+[+[(+)*38[-][+][+--]]>{}]>]]]]]]]]]>)%21)*21](+)*50>--[(-)*50(>)*8(([----[+[+[+[+[+[+[+[+[+[(+)*38[-][+][+--]]>{}]>]]]]]]]]]>)%21)*21](-)*50)*15])%28)*28]+)*27
00:51:42 <EgoBot> Score for quintopia_brachiation2: 16.2
00:54:47 -!- Vorpal has quit (Ping timeout: 244 seconds).
00:54:50 <quintopia> !bfjoust brachiation2 >->+++++(>[(<)*3(+)*10((<+[{}(-)*50>(+)*44(>--[(+)*50(>)*8(([----[+[+[+[+[+[+[+[+[+[(+)*38[-][+][+--]]>{}]>]]]]]]]]]>)%21)*21](+)*50>--[(-)*50(>)*8(([----[+[+[+[+[+[+[+[+[+[(+)*38[-][+][+--]]>{}]>]]]]]]]]]>)%21)*21](-)*50)*15])%28)*28]+)*27
00:54:53 <EgoBot> Score for quintopia_brachiation2: 46.7
00:56:06 <quintopia> !bfjoust brachiation >->(-)*5(>)*4(>[(<)*3(+)*7<<(+[<{}>(-)*50<(-)*50<(-)*10>>(>-[(-)*50(>)*7(>[(-)*5([+{[(+)*25[-]]}[+]])%10[-]][+][-])*27](+)*50)*3 (>-[(-)*50(>)*7(>[(-)*5([+{[(+)*25[-]]}[+]])%10[-]][+][-])*27](-)*50)*2(>)*7 (>[(-)*5([+{[(+)*25[-]]}[+]])%10[-]][+][-])*22])%28]++)*22>(-)*115[-][+][-]
00:56:09 <EgoBot> Score for quintopia_brachiation: 46.3
00:56:47 <EgoBot> Score for quintopia_brachiation: 0.0
00:57:15 <quintopia> !bfjoust brachiation >->(-)*5(>)*4(>[(<)*3(+)*7<<(+[<{}>(-)*50<(-)*50<(-)*10>>(>-[(-)*50(>)*7(>[(-)*5([+{[(+)*25[-]]}[+]])%10[-]][+][-])*27](+)*50)*3 (>-[(-)*50(>)*7(>[(-)*5([+{[(+)*25[-]]}[+]])%10[-]][+][-])*27](-)*50)*2(>)*7 (>[(-)*5([+{[(+)*25[-]]}[+]])%10[-]][+][-])*22])%28]++)*22>(-)*115[-][+][-]
00:57:17 <EgoBot> Score for quintopia_brachiation: 46.2
00:57:47 <EgoBot> Score for quintopia_brachiation2: 0.0
01:24:11 -!- zzo38 has joined.
01:33:51 <zzo38> I think the category (->) has Zero as initial object and () as final object, while (Kleisli Maybe) has Zero as both initial and final, and that (Kleisli IO) has no final object at all. In (Kleisli Finalize) all objects are final. Is this correct?
01:46:54 <oerjan> zzo38: assuming you are ignoring bottoms, that seems right except i don't know Finalize. also Zero is initial for any Kleisli m
01:48:24 <oerjan> assuming Finalize ~ Proxy, that would seem right too
01:49:06 <zzo38> oerjan: Yes I know Zero is initial for any Kleisli category too, and yes Finalize is the same as Proxy
01:50:27 <zzo38> And where Zero is uninhabited type
01:50:55 <zzo38> And, yes, I do mean ignoring bottoms
01:51:33 -!- nortti_ has quit (Quit: AndroIRC - Android IRC Client ( http://www.androirc.com )).
01:59:28 <zzo38> I called it Finalize monad and its dual is Initialize comonad, which can be made with any category having final objects/initial objects. They are not specific to Haskell.
02:10:53 -!- copumpkin has quit (Ping timeout: 250 seconds).
02:11:25 -!- copumpkin has joined.
02:14:46 <david_werecat> !bfjoust quirtle --(>)*8((>[++[-[--[-[(-)*122(-.)*11>{}]]]]])%21)*21
02:14:49 <EgoBot> Score for david_werecat_quirtle: 41.5
02:15:00 <david_werecat> !bfjoust lightning --(>)*8((>[++[-[--[-[(+)*14[-]>{}]]]]])%20)*20
02:15:02 <EgoBot> Score for david_werecat_lightning: 37.4
02:27:33 <david_werecat> Well, I do need to go now, but before I do: the clear takes advantage of the fact that most programs leave an even number of decoys.
02:31:41 -!- david_werecat has quit (Ping timeout: 245 seconds).
03:32:15 <tswett> Sgeo: did you see the potato?
03:32:42 <tswett> It's been out for about half an hour.
03:34:01 <Sgeo> Nope, thank yiu
04:01:47 -!- pikhq_ has quit (Quit: Reconnecting).
04:02:00 -!- pikhq has joined.
04:04:53 -!- itidus22 has joined.
04:06:24 -!- itidus20 has quit (Ping timeout: 245 seconds).
04:11:52 -!- asiekierka has joined.
04:12:51 -!- zzo38 has quit (Remote host closed the connection).
04:58:14 -!- pikhq has quit (Ping timeout: 246 seconds).
04:58:16 -!- pikhq_ has joined.
05:06:40 -!- asiekierka has quit (Quit: Wychodzi).
05:54:10 -!- john_metcalf has quit (Quit: john_metcalf).
05:58:47 -!- pikhq has joined.
05:58:54 -!- pikhq_ has quit (Ping timeout: 255 seconds).
06:19:43 -!- zzo38 has joined.
06:26:20 -!- john_metcalf has joined.
06:28:16 <Sgeo> tswett, elliott monqy ta ng update
06:35:46 <oerjan> <Sgeo> tswett, elliott monqy ta ng update <-- out of context that could be an interesting esolang...
06:36:25 <oerjan> not that those aren't esoteric.
06:42:54 <zzo38> What monads/comonads are possible in a thin category? Is it only the identity monad/comonad, Finalize monad, and Initialize comonad? Are any others possible?
06:43:05 <fizzie> What, a WordNet interface? So neat.
06:43:20 <lambdabot> *** "esoteric" wn "WordNet (r) 3.0 (2006)"
06:43:22 <lambdabot> adj 1: confined to and understandable by only an enlightened
06:43:24 <lambdabot> inner circle; "a compilation of esoteric philosophical
06:43:46 <fizzie> (The *** line seems a bit superfluous.)
06:44:12 -!- itidus22 has changed nick to itidus21.
06:58:04 <zzo38> As far as I can tell there would not be possible to have any monads on a thin category where distinct objects are unequal other than the identity monad and the Finalize monad.
07:05:29 <zzo38> Is there a proof, disproof, or counterexample?
07:06:55 <oerjan> what do you mean by "unequal"?
07:07:16 <zzo38> I mean as in a partial ordering; in case of categories, they would be not isomorphic.
07:08:25 <oerjan> well M x and M (M x) would always have to be the same object then, i think
07:08:55 <zzo38> oerjan: Yes it is what I thought.
07:09:59 <zzo38> Therefore, the only way it can make a monad is if M is the identity functor or Finalize, I think.
07:10:19 <zzo38> (And with comonads, also it must be the identity functor or Initialize)
07:10:27 <oerjan> i think you can have disjoint components behaving independently
07:10:52 <kmc> i made my polyglot program also a .COM file and it worked on the first try
07:11:08 <kmc> this is naked bragging but I am very pleased :)
07:12:02 <zzo38> oerjan: O, yes, perhaps you can have a locally final object or something like that, is that what you mean?
07:13:16 <oerjan> what if you take an arbitrary partial order and take the product with (0,1); then can't you make a monad which sends (x,i) to (x,1) ?
07:15:13 <zzo38> What is a product with {0,1}?
07:15:53 <oerjan> a partial order with underlying set {(x,i) | x in original partial order, i in {0,1}}
07:16:15 <oerjan> and (x,i) <= (y,j) means x <= y and i <= j
07:16:53 <zzo38> Yes, I can see that now
07:17:59 <zzo38> So in this case you have a product of monads; in the other case with disjoint components behaving independently it was a sum of monads, I guess
07:20:08 <zzo38> Since it is the product of the identity monad on the original partial order and the Finalize monad on {0,1}
07:21:37 -!- oerjan has quit (Quit: May need to do shopping).
07:30:57 -!- sebbu2 has joined.
07:30:58 -!- sebbu2 has quit (Changing host).
07:30:58 -!- sebbu2 has joined.
07:31:11 -!- sebbu has quit (Ping timeout: 246 seconds).
07:41:20 -!- augur_ has quit (Read error: Operation timed out).
07:42:44 -!- augur has joined.
07:45:05 -!- azaq23 has quit (Read error: Connection reset by peer).
07:56:30 -!- MDoze has changed nick to MSleep.
08:24:38 <kmc> why are system calls in DOS done through interrupts rather than an ordinary (far) call?
08:24:57 <kmc> (ditto BIOS)
08:25:12 <fizzie> You can at least hook them better when they've done like that, though I don't know if that has been the reason.
08:25:22 <fizzie> Also, no changing addresses.
08:25:36 <fizzie> Admittedly there could be a list of indirect-call addresses at some fixed location.
08:25:48 <fizzie> But that's pretty much what the interrupt vector table is.
08:27:27 <kmc> and the code to invoke an interrupt is shorter, maybe
08:28:00 <shachaf> Don't interrupts have a performance penalty compared to just a far call?
08:32:02 <fizzie> I don't know if there's much of a difference in real mode. It certainly doesn't do much more than an absolute-indirect far call/return, except for saving/restoring flags in addition to cs/ip.
08:37:32 <fizzie> Apparently it's possible to do pushf + far call to the start of an interrupt handler and things will work out right, if you for some reason want to.
08:41:29 <fizzie> Oh, and I guess "INT x" also automagically disables interrupts (clears the trap and interrupt flags) which might be useful.
08:56:03 <pikhq> kmc: I think so that the functions don't have to be at specific memory addresses.
08:56:40 <pikhq> The extra layer of indirection lets you change memory layout.
08:57:46 <itidus21> this is quite possibly the best individual work of super mario bros. fan art ever: https://sites.google.com/site/jinsdepositry/home/jp_mariofin05fnsm.jpg
09:00:36 <fizzie> pikhq: You could do that equally well with an indirect far call. (But admittedly software interrupts and the IVT pretty much pre-implement that sort of thing.)
09:04:14 <zzo38> In a MBR code I wrote once, to call the disk interrupt it did not use INT; instead it had the almost equivalent set of instructions, so that the return address would be different
09:06:39 <pikhq> fizzie: And it's CISC.
09:06:45 <pikhq> Why use 2 instructions when 1 will do?
09:07:21 -!- Phantom_Hoover has joined.
09:07:36 <zzo38> I gave one example; such as adjusting the return address
09:08:10 <pikhq> zzo38: My *point* was just that real mode x86 coding style prefers to use as few instructions as humanly possible. :)
09:08:47 <pikhq> Obviously, if you're doing that sort of hack, you want to manually call the IVT.
09:09:18 <zzo38> pikhq: Yes, and I agree that too. However, there are sometimes you want to do things slightly different than the standard instructions
09:11:18 <zzo38> This MBR code first fills the screen with "p" black on gray (the color is actually the reason why it is "p"), and then loads the kernel.
09:14:00 <fizzie> I guess the indirect far call also means you'd have to clobber a segment register to hold the location where system calls would be stored.
09:27:42 <itidus21> correction: i concede there is no such thing as best art
09:32:06 -!- zzo38 has quit (Remote host closed the connection).
09:57:02 -!- sebbu has joined.
09:57:03 -!- sebbu has quit (Changing host).
09:57:03 -!- sebbu has joined.
10:00:17 -!- sebbu2 has quit (Ping timeout: 246 seconds).
10:01:49 -!- derdon has joined.
10:06:21 <itidus21> this next link is a wtf moment :-D
10:08:17 <itidus21> http://oi50.tinypic.com/2cgbj0k.jpg
10:11:05 <kmc> yeah windows XP, what the fuck
10:29:09 <fizzie> "My name is Johny, what the F**K?????"
10:47:52 -!- sebbu2 has joined.
10:47:53 -!- sebbu2 has quit (Changing host).
10:47:53 -!- sebbu2 has joined.
10:48:56 -!- sebbu has quit (Ping timeout: 246 seconds).
10:51:08 -!- ion has quit (Ping timeout: 240 seconds).
10:52:08 -!- ion has joined.
10:57:11 -!- nooga has quit (Read error: Connection reset by peer).
10:57:27 -!- nooga has joined.
11:02:29 -!- nooga_ has joined.
11:03:49 -!- nooga has quit (Ping timeout: 248 seconds).
11:27:17 -!- elliott has quit (Ping timeout: 248 seconds).
11:34:21 <kmc> h<|XP- {P_X(%GGG(%GGWZ- sh LI!XI!Hello, DOS!$
11:42:50 -!- DHeadshot has quit (Ping timeout: 246 seconds).
11:48:47 -!- Slereah has joined.
11:51:05 -!- Slereah_ has quit (Ping timeout: 250 seconds).
12:58:53 <Phantom_Hoover> OK so the connection from my computer to my headphones is so noisy that I can hear scrolling.
12:59:44 -!- Vorpal has joined.
13:00:01 <Phantom_Hoover> I don't even know how this is happening; it's like the headphone cord is picking up induction currents from every wire in the computer/
13:05:19 <fizzie> Hearing scrolling is not very unusual, really.
13:05:41 <fizzie> I can hear mouse movements on one box if I use the integrated sound chip.
13:05:57 <fizzie> I got my external USB audio stick partially because of that.
13:06:35 <fizzie> Though I wouldn't immediately expect it to be the headphone cord's fault.
13:07:45 <Phantom_Hoover> Yeah I had to switch to using a USB adapter because I think the front headphone socket is broken,.
13:08:07 -!- david_werecat has joined.
13:08:35 -!- kmc_ has joined.
13:08:39 -!- kmc_ has quit (Remote host closed the connection).
13:14:23 <fizzie> "I can hear all sorts of things. the most notable is dragging windows around or when web pages load. I can also hear transferring files. The variation on timbre per click is kind of interesting. Copying files will produce a higher pitched tone than loading a web page, for example. Dragging scrollbars will produce more of a staticy sound than a click."
13:14:35 <fizzie> I see some people devote time to learning how their computer sounds like.
13:19:06 <fizzie> I have a vague recollection of one noise problem solved by muting the CD audio input; presumably the analog CD audio wire from the optical drive was picking up those noises. But that'd probably come through non-headphone playback at the same relative volume if it were the case. (Also not sure if modern optical drives even bother including analog CD-audio.)
13:20:31 <fizzie> I can see through my OpenOffice Writer window. :/
13:21:23 <fizzie> Also a picture embedded on one page is upside-down, but that might be the document author's fault.
13:26:22 -!- sebbu2 has changed nick to sebbu.
13:31:27 <Gregor> Why does the Purdue Bursar's Office send me a billing statement of $0 ...
13:31:37 <Gregor> I have to pay this $0 before July, too.
13:36:44 <fizzie> You'd better act fast, then.
13:37:28 <fizzie> I got a $0 bill the other day, too. It had just a row of stars in place of the reference number, but it did have recipient bank account info and a sum.
13:37:32 <Gregor> If I don't pay in time, they'll charge me a late fee...
13:37:50 <fizzie> Gregor: Sadly, the late fee is probably a fixed sum, and not a fraction of the $0.
13:38:25 <Gregor> Or "¼ the owed fee, minus $50"
13:38:34 <itidus21> I don't know which is which between debit and credit
13:39:00 <itidus21> but sounds like you have $0 credit with them, whereas i have $0 debit with them
13:39:06 <Gregor> But to your credit, you haven't any debit.
13:39:53 <Vorpal> (Also not sure if modern optical drives even bother including analog CD-audio.) <-- mine doesn't have that
13:40:15 -!- MSleep has changed nick to MDude.
13:41:03 <itidus21> anyway it sounds like the -0 +0 problem
13:41:05 <Vorpal> fizzie, so yeah analogue CD-audio is dying
13:41:53 <Vorpal> or rather, it makes no difference if it is
13:41:57 <fizzie> I have a feeling that my most recent (SATA) optical drive didn't have CD audio either.
13:42:06 <Vorpal> two-complement and unsigned both works for brainfuck
13:42:24 <Vorpal> fizzie, yeah, neither of my more modern DVD drives have that
13:42:32 <itidus21> if [] was skipped over for +0 but not -0
13:42:40 <Vorpal> (though one of them don't really count, it is ultrabay)
13:42:44 -!- copumpkin has quit (Quit: Computer has gone to sleep.).
13:42:59 <Vorpal> itidus21, what are you talking about? There is just one zero in two-complement
13:43:08 <Vorpal> or are you talking about sign bit?
13:43:18 <Vorpal> which nobody uses due to the dual zero issue!
13:43:18 <fizzie> Ultraboy, a superhero that lives in an Ultrabay slot.
13:43:31 <itidus21> it's my analogy to gregor's debts
13:43:55 <Vorpal> itidus21, right, which doesn't work in two-complement. Which is what everybody uses
13:44:12 <fizzie> Certainly people use a "sign bit", just not sign-magnitude. (Well, except if you count floats.)
13:44:42 <Vorpal> fizzie, floating point is irrelevant to brainfuck though
13:44:59 <Vorpal> Gregor, I guess you should call them and ask wtf they are high on :P
13:45:47 <fizzie> I see C11 still obeys the traditions and allows one's-complement and sign-and-magnitude.
13:45:55 <itidus21> i think thats what i mean anyway
13:46:07 <Vorpal> itidus21, I have no idea what you mean by that though
13:46:09 <itidus21> where it is waiting for +1 as an input
13:46:32 <itidus21> but if the cell is 1 .. then the - operation reduces it down to -0
13:46:42 <Gregor> fizzie: Technically speaking, C11 probably allows you to implement ints as floats that you always truncate.
13:46:47 <Vorpal> itidus21, do you even understand what two-complement is?
13:46:51 <itidus21> im out of my mind, some would call it trolling
13:46:53 <Gregor> (i.e. it doesn't give two shits about the representation)
13:47:38 <itidus21> maybe i should try and talk logically
13:48:53 <fizzie> Gregor: No, it doesn't. Neither did C99.
13:49:19 <itidus21> ha ha. that was an amusing bureaucratic error they pulled. this has no rational relationship with number polarity or brainfuck.
13:49:21 <Gregor> fizzie: Is that no, it doesn't allow you to represent them that way, or no, it doesn't give two shits about the representation?
13:50:07 <fizzie> No, it doesn't allow them. It does give shits about the representation. Except in that particular case you might be able to make it work as long as you consider the exponent parts of the float padding bits.
13:50:26 <Gregor> quintopia: You notice how I never "hi" anyone, I just leap into conversations?
13:50:59 <Vorpal> Gregor, hello there! greetings!
13:51:11 <fizzie> "For signed integer types, the bits of the object representation shall be divided into three groups: value bits, padding bits, and the sign bit. -- Each bit that is a value bit shall have the same value as the same bit in the object representation of the corresponding unsigned type. -- If the sign bit is one, the value shall be modified in one of the following ways: [list of ...
13:51:17 <itidus21> i do that. i think a lot of bandwidth is wasted on greetings
13:51:17 <fizzie> ... sign-and-magnitude, two's-complement and one's-complement]"
13:51:27 <Gregor> Vorpal: That pain in your stomach is the blade. Don't think too much about it, the pain will be gone soon. Just breaaaathe.
13:51:59 <fizzie> And as for the value bits of unsigned integer types, "each bit shall represent a different power of 2 between 1 and 2^(N-1)".
13:52:13 <Vorpal> itidus21, only 3 packets: SYN, SYN-ACK, ACK (in TCP)
13:52:19 <itidus21> worst case scenario is a mandatory "i'm back!" every time someone joins a channel, and mandatory "i'm leaving" before quitting or lefting
13:52:47 <Gregor> quintopia: Where's the damned cooooode!
13:52:48 <Vorpal> Gregor, well I will continue greeting people.
13:53:22 <quintopia> Gregor: are you going to install it RIGHT THE FUCK NOW?
13:53:32 <Gregor> Probably not, but it's not impossible.
13:53:43 <Vorpal> quintopia, code for what?
13:54:06 <quintopia> then i'll send it when i get home from work rather than stressing about it now
13:54:36 <Vorpal> quintopia, mail a copy on a CD :P
13:54:44 <quintopia> then i'll send it to my private server and not give you access
13:54:55 <Gregor> quintopia: Can't you just ... put it online somewhere and give me a URL? Email is so clunky.
13:55:22 <Vorpal> quintopia, Gregor: what is the code for?
13:55:34 <Gregor> Vorpal: New BF Joust hill scoring system.
13:56:02 <Gregor> In almost every conceivable way... but you'd have to ask quintopia for the actual algorithmic details.
13:56:25 <Gregor> As I recall, it's a proper fixed point. But we talked about it MONTHS ago.
13:56:29 <quintopia> well, it gives Gregor a hand job every time a new program is submitted. this is the most important feature.
13:56:32 <Vorpal> quintopia, what are the algorithmic differences?
13:56:40 -!- sebbu2 has joined.
13:57:04 <itidus21> I think bfjoust really exemplifies programmable robot tournaments
13:57:09 -!- sebbu has quit (Ping timeout: 248 seconds).
13:58:04 <quintopia> i think it really exemplifies "you're nuts"
13:59:12 <itidus21> i once had a plan for a game which exemplified the serious business trope, where there would be remote control robot tournaments in an arena
13:59:22 <itidus21> and the whole society around it would be simulated
13:59:59 <itidus21> but i never quite felt that it was a strong enough basis for a society
14:01:32 <itidus21> but battling robots by remote control, no
14:01:38 -!- copumpkin has joined.
14:01:52 <itidus21> battling robots by pre-programming would be too geek for me
14:03:11 <itidus21> well i think it is implied in things like bfjoust that actual systems could compete along similar lines
14:04:35 <fizzie> Are any of those TV series robot battles doing it right i.e. no remote controls? I haven't been paying attention.
14:04:38 <quintopia> as far as robots go, bfjoust warriors have the poorest sensors and the weakest attacks :P
14:05:15 <quintopia> fizzie: battle bots had an "autonomous class" once, but it wasn't very good.
14:05:40 <quintopia> robocup is the only serious competition along those lines these days
14:06:59 <itidus21> i think bfjoust is like an aggressive form of golf
14:07:24 <fizzie> I think the aggressive form of golf is where you hit other people with the clubs.
14:07:43 <itidus21> like suppose there was a golf course where the ball could trigger walls to go up by landing on certain zones
14:08:36 <itidus21> yes, i have imagination and am proud of it >.<
14:09:03 <quintopia> there is a certain amount of golfing of programs required
14:09:21 <itidus21> in minigolf it could be that there was some pressure censor which could make the doors close
14:09:25 -!- asiekierka has joined.
14:09:35 <itidus21> on one of those buildings on the courses
14:09:36 -!- david_werecat has quit (Ping timeout: 245 seconds).
14:09:50 <fizzie> Computer minigolf games always have that kind of fluff, I'm not sure how feasible that is in real reality.
14:09:57 <Vorpal> itidus21 you mean sensor?
14:10:02 <Vorpal> itidus21, rather than censor surely
14:10:26 -!- david_werecat has joined.
14:10:27 <itidus21> yeah i fnord meant sensor, but i don't know fnord why i typed censor
14:11:03 <itidus21> because typing sensor as censor is creepy
14:12:33 <itidus21> anyway.. fizzie i didn't know that about computer golf games
14:12:49 <itidus21> fizzie: of course it's feasible though :P
14:13:40 <itidus21> the course would need some way to know when to do the sensing though
14:13:59 <itidus21> it could be an informal rule not to step on the sensors
14:14:19 <itidus21> but it only really matters during a swing
14:14:51 <itidus21> and you could have little fences emerge from the field
14:14:55 <fizzie> I mean "feasible" in the "makes business sense" sense.
14:15:00 <fizzie> Of course it's physically possible.
14:15:20 <fizzie> Sounds like in a mostly-unattended course that sort of stuff would just forever be getting stuck or whatnot.
14:15:32 <itidus21> i was also referring to code golfing too indirectly
14:16:25 <itidus21> since bfjoust against a static unchanging opponent amounts to code golf
14:16:58 <quintopia> not really. there is no rewardd for shorter code in bfjoust
14:17:19 <quintopia> code golfmight mean something different than you thik
14:21:08 <itidus21> i arrived at this point by pondering what would happen if the jousters in bfjoust didn't compete. and i didn't think it through. i guess that it would simply be a case of >>>>>>>>>
14:24:04 -!- MDude has quit (Ping timeout: 276 seconds).
14:47:12 -!- nortti has quit (Quit: Found terminal).
14:48:54 -!- oonbotti has quit (Remote host closed the connection).
14:51:03 -!- nortti has joined.
15:06:17 -!- sebbu2 has changed nick to sebbu.
15:27:31 -!- Frooxius|TabletP has joined.
15:41:09 -!- Frooxius|TabletP has quit (Ping timeout: 248 seconds).
15:49:10 <Vorpal> who thought that abbreviating "Low self-discharge nickel-metal hybrid battery" to "LSD NiMH" was a good idea...
15:50:13 <ion> Sounds like a good idea to me.
15:55:01 <itidus21> one thing i seem to do is put into words things which wouldn't normally be put into words
15:57:28 <itidus21> i have these strange insights which seem to often be incomplete versions of theories and ideas known to the wise
15:58:05 <itidus21> and then i make it perverse by trying to use them primarily for creating entertainment
15:59:19 <itidus21> like i once had this thought about a turtle shell from super mario bros bouncing against bricks being almost like a program running
15:59:48 <Vorpal> it is a program running. The game simulating the event you just described
16:00:16 <itidus21> i mean as if bricks and turtle shells could almost form a turing machine
16:00:51 <itidus21> i assume to begin with i would need a non-breaking brick, a breaking brick, and a turtle shell bouncing
16:00:52 <Vorpal> there is no flow control
16:01:22 <itidus21> the shell goes either left or right, hits a brick, bounces back the other way
16:01:31 <Vorpal> also program output would e
16:01:41 <itidus21> then when it bounces back again
16:01:46 <Vorpal> be* limited based on the number of bricks you have available
16:01:54 <Vorpal> you couldn't create more bricks as required
16:02:23 <Vorpal> so I'm pretty sure it isn't TC, though don't quote me on that
16:02:42 <Gregor> So, Subway has introduced a guacamole bacon lettuce and tomato sandwich.
16:02:51 <itidus21> theres this scene near the end of one of the levels where you knock one brick down and watch it destroy row after row of bricks
16:02:55 <Gregor> Tragically, they don't abbreviate it to GBLT, and hence do not attract only a niche audience.
16:03:10 <itidus21> and i guess that is the most inspiring part of t
16:03:34 <itidus21> well.. there could be a thing where a shell can fall down a hole
16:03:57 <itidus21> which would mean the program has stopped running perhaps
16:05:07 <itidus21> only now at the end do you see how truely perverse i am about this stuff
16:07:38 -!- azaq23 has joined.
16:08:57 <itidus21> Vorpal: if i figured out a way to make more bricks though, that could be a cool thing
16:10:20 <Phantom_Hoover> I'm not sure the brain is the put-upon one in the iti/iti's brain relationship.
16:10:39 <itidus21> i would actually have to understand TCs to take this any further
16:11:04 <oklopol> i only understand like half of the TCs
16:11:54 <itidus21> the shell proceeds along resting on bricks
16:12:20 <itidus21> and when it encounters a brick in it's path it erases the brick and then changes direction
16:19:11 -!- MDude has joined.
16:24:02 <itidus21> ok so mario gives us the easy part.. a tape
16:24:33 <itidus21> if i suppose that an unbounded strip of tiles either containing a blank or a brick is a tape
16:24:44 <itidus21> and the turtleshell is the head
16:25:11 <itidus21> i think my lack of understanding begins by assuming that tells me much on it's own
16:26:28 <itidus21> so the head(the turtle) can erase a brick.. but it needs the capacity to write a brick.. which can be added too
16:26:43 -!- john_metcalf has quit (Quit: john_metcalf).
16:27:06 <itidus21> just a bit weird having the shell writing bricks but i guess it is necessary
16:28:49 <itidus21> so i can say that it is currently running such a program that
16:32:02 <itidus21> if the shell is on a blank it will proceed to the right until it encounters a brick. if the shell is on a brick it will erase the brick, then proceed to the left. if the shell is on a blank it will proceed to the left until it encounters a brick. if the shell is on a brick it will erase the brick, then proceed to the right. and repeat.
16:33:58 <itidus21> but it is that part, the defining that "program" which is the aspect of turing machines i am most uneducated about
16:34:50 <itidus21> anyway, have i defined an instance of a finite state machine? :D
16:35:05 <fizzie> Phantom_Hoover: You could equally well say "Eurg argh aaargh" to the phrase "LOTR WRITING"; there are quite a few fictional scripts in the Middle-earth legendarium.
16:35:40 <Phantom_Hoover> He's imitating the English-but-with-Tengwar-vowel-signs script that's used for maps and the like.
16:35:45 <itidus21> i will call that program mario
16:38:11 <fizzie> I don't recall anything like that in the maps that are in my copy. But that's a Finnish printing.
16:38:50 <Phantom_Hoover> But it's definitely used somewhere for some window-dressing text.
16:56:59 <itidus21> while(1) { while (n[i] == 0) i++; n[i] = 0; i--; while (n[i] == 0) i--; n[i] = 0; i++; }
16:58:23 <itidus21> one more time to be a bit cleaner: while(1){ while(!n[i]) i++; n[i]=0; i--; while(!n[i]) i--; n[i]=0; i++; }
16:58:53 <quintopia> it looks like a tape clear to the right that never comes back
16:59:14 <itidus21> i actually did spend far too long doing that
16:59:18 <quintopia> it clears an equal number of cells to the right
16:59:47 <quintopia> and then when one direction is completely cleared, it heads off to infinity
16:59:53 <itidus21> it can't compile of course since the size of n is unspecified and nor is the type of i
17:00:30 <itidus21> i applied brainfuck logic to it in C
17:01:09 <itidus21> im yet to write a bf version of this though
17:02:31 <itidus21> while (n == 0) is certainly not what bf was made for :D
17:02:31 <quintopia> since you need to keep something on the tape to use as your while loop flag
17:02:47 <quintopia> much easier to set the whole tape to 1s
17:03:18 -!- ais523 has joined.
17:03:24 <itidus21> i mean... i could inverse the logic
17:03:44 -!- boily has joined.
17:05:10 <itidus21> perhaps if i say 1 is a blank, then i can do this: while(1){ while(n[i]) i++; n[i]=1; i--; while(n[i]) i--; n[i]=1; i++; }
17:06:00 <quintopia> if the tape is all bits that would work
17:08:35 <quintopia> here's what i'm thinking: [-][[>]>[-]<+[<]<[-]>+]
17:09:05 <itidus21> while(1){ while(n[i]==1) i++; n[i]=1; i--; while(n[i]==1) i--; n[i]=1; i++; }
17:09:09 <quintopia> boily: did i ever tell you that i <3 aubergine
17:14:25 <boily> quintopia: thanks!
17:20:20 -!- Frooxius|TabletP has joined.
17:23:26 -!- TeruFSX_ has quit (Ping timeout: 246 seconds).
17:31:13 -!- DHeadshot has joined.
17:54:54 <itidus21> fark that took more work than i expected
18:05:01 <fizzie> Purely going on substrings, ...][-] looks dubious, because immediately after passing through ] the cell can only be 0.
18:05:07 -!- atrapado has joined.
18:07:35 <itidus21> this interpreter best demonstrates it's behaviour (in the debug box) http://www.lordalcol.com/brainfuckjs/
18:11:00 <fizzie> That's a very confusing debug box.
18:11:19 <fizzie> Took me a while to realize that it doesn't skip jumps, instead it walks through then with "SKIP WHILE: true" in the box.
18:12:07 <fizzie> Anyway, I'm reasonably sure it's not physically possible for those [-]s that come immediately after a ] to actually have an effect on anything.
18:16:25 <itidus21> though i can't figure out what it's doing when it stops
18:16:28 <fizzie> The [-] immediately at the start is I guess technically redundant too, assuming the usual start of a tape full of zeros.
18:18:05 <fizzie> Incidentally, I suppose there's some reason why the program is not just +[[>]+<[<]>] ? I haven't really been following the conversation.
18:18:56 <itidus21> because i'm the one who wrote it >.<
18:21:37 <itidus21> fizzie: well that program, which you have written properly, it's a simulation of a turtle shell from super mario bros.
18:23:10 <fizzie> Oh, I see. Except the way I wrote it, it only grows to one direction. But +[[>]+<[<]+] should be it.
18:23:42 -!- Taneb has joined.
18:23:46 <fizzie> Sorry, I mean +[[>]+[<]+] -- that one < is obviously unnecessary.
18:25:15 <itidus21> fizzie: so... like.. what you have basically got here is a way to tell a super mario turtle shell how to bounce
18:27:11 <fizzie> In a hypothetical brainfuck where < and > were replaced with ^ (either > or < depending on the current direction) and | (flip the current direction), that would be the simpler +[[^]+|].
18:28:06 <itidus21> i think the reason i made such a mess of hte program was i was trying to base it on C
18:30:55 <itidus21> so [-]+[[[-]+>][-]+<[[-]+<][-]+] was supposed to be:
18:31:03 <itidus21> n[i]=1;while(n[i]!=0){while(n[i]!=0) { n[i]=1;i++;}n[i]=1;i--;while(n[i]!=0){n[i]=1;i--;}n[i]=1;}
18:34:19 -!- Frooxius|TabletP has quit (Ping timeout: 245 seconds).
18:35:27 -!- MDude has quit (Read error: Connection reset by peer).
18:35:47 -!- MDude has joined.
18:37:36 <itidus21> yeah... thats over. i can try to learn from this
18:44:09 <fizzie> Yes, I guess a direct translation of "=1" does easily end up with some [-]+'s. Anyway, +[[>]+[<]+] translates to n[i]++; while (n[i]) { while (n[i]) i++; n[i]++; while (n[i]) i--; n[i]++; }, it's just that all n[i]++s are performed only on cells that are zero, therefore they are equivalent to n[i] = 1s.
18:45:10 <fizzie> As for something like while (n[i] != 0) { n[i] = 1; i++; }, it's not really necessary to have n[i] = 1 in there, because n[i] is already 1, otherwise it wouldn't be looping any more.
18:45:17 <itidus21> well i have developed a sound system of translating brainfuck to C as part of a way to read it
18:45:31 <itidus21> i think only now have i got it all sorted
18:49:05 <itidus21> this is the translating: http://pastebin.com/sXRNnvVr
18:54:02 <fizzie> Wikipedia Brainfuck article has a C translation too, unless I misremember.
18:54:20 <fizzie> Yes, they have. Though it's the pointer version.
18:55:23 -!- john_metcalf has joined.
18:57:13 <itidus21> i personally like mine better.. but blah
18:59:46 <itidus21> my knowledge of C is limited also.. i don't really consider deeply what ++ before and after means
19:00:18 <itidus21> nor do i know whether a pointer can be trusted to address an array
19:05:30 <fizzie> That's pretty much the only thing a pointer can really do.
19:06:21 <fizzie> As for pre/postfix ++/--, it doesn't much matter if the value of the expression is not being used for anything.
19:07:55 <pikhq> Some (idiots) will tell you one of them is more efficient (I honestly don't remember which it is) when you're not using the result.
19:08:06 <pikhq> ... This may have been true of pathologically un-optimizing compilers.
19:08:12 <itidus21> a mario shell is a strange sort of tapehead
19:09:24 <itidus21> rather, it can perhaps read the before and after
19:09:43 <fizzie> It's the prefix version that's usually considered efficient, though mostly in the context of C++ and user-defined types.
19:10:36 <fizzie> Since the postfix version of the operator has to create a temporary (and hope it's optimized away), while the prefix operator can just return a self-reference.
19:12:01 <fizzie> I've seen C++ books advocate it's good form to have a habit of using the prefix version in all cases when the value is not used, though.
19:14:15 <fizzie> http://www.parashift.com/c++-faq-lite/operator-overloading.html#faq-13.15 from the much-maligned FAQ. The FQA just pokes fun at how the compiler's optimization powers in this answer aren't limitless.
19:14:38 <pikhq> fizzie: Yes, well, C++ ought to be ignored.
19:15:38 <fizzie> Perhaps they should be calling it ++C, so that it'd be efficient.
19:19:07 <fizzie> C++ trivia (well, it's maybe not that trivia): to show that you intend to overload the postfix ++ instead of the prefix ++, you give the "operator++" method a dummy int parameter.
19:20:26 <fizzie> I vaguely think it's defined to have the value 0, but using it would be really bizarre.
19:20:27 <ais523> gah, I had "/me wonders" unsent in my #esoteric tab
19:20:32 <ais523> and have forgotten what I was wondering
19:20:52 <ais523> oh right; I was wondering if anyone received emails sent to abuse@example.com or webmaster@example.com
19:20:58 * boily suffers from repressed bad memories from his freshman programming introductory course...
19:21:18 <fizzie> ais523: Isn't that just a MX record check?
19:23:12 <fizzie> ais523: example.com doesn't seem to have a MX record, but it *does* have A/AAAA records, and I think those are used as fallbacks by MTAs.
19:23:49 <fizzie> (It also has dnssec signatures, interestingly enough.)
19:26:34 <fizzie> Nobody seems to be home at port 25 though.
19:29:48 -!- elliott has joined.
19:35:08 -!- MoALTz has quit (Quit: brb).
19:38:34 -!- MoALTz has joined.
19:46:26 -!- asiekierka has quit (Remote host closed the connection).
20:00:21 -!- Taneb has quit (Quit: battery).
20:09:09 <Gregor> Purdue University Board of Trustees, which are appointed by the governor of Indiana, elects the governor of Indiana as president of Purdue.
20:09:58 <itidus21> why are you so interested in purdue? one might think you lived there :-j
20:10:15 <Gregor> And/or am a student there.
20:11:01 <lambdabot> Local time for itidus21 is Fri Jun 22 06:09:29
20:12:47 <itidus21> the first paragraph on this page is quite enlightening http://www.iwriteiam.nl/Ha_bf_Turing.html
20:19:37 -!- MDude has quit (Ping timeout: 276 seconds).
20:33:01 -!- boily has quit (Quit: WeeChat 0.3.8).
20:33:03 <itidus21> ok! i have come up with a way to think of an FSM in C
20:35:52 <itidus21> that took quite a while to arrive at actually..
20:42:44 <itidus21> and, even better: http://pastebin.com/xqP9wrxi
20:44:20 <itidus21> it seems what i am missing is a null operation
20:52:03 <nortti> wasn't microvax vax in desktop case?
20:54:41 -!- david_werecat has quit (Quit: Page closed).
20:56:43 <Vorpal> nortti, are you in Vax channels!? :D
20:57:12 <Phantom_Hoover> I really want to make a clever joke about my computer's name, but it's just not coming.
20:57:45 <nortti> Phantom_Hoover: wasn't your computer named vax?
20:58:07 <Phantom_Hoover> This reminds me, I really need to get some compressed air and clean it out.
21:02:13 <itidus21> i don't know if there is any actual end to my edits to this, but im wondering if this is a sound comprehension of turing machines http://pastebin.com/xpNcw1jh
21:02:15 <Vorpal> nortti, ah right that sounds familiar
21:02:30 <itidus21> with lots of my notorious custom notations
21:02:36 <Vorpal> Phantom_Hoover, do you have dust filters?
21:02:49 <Vorpal> I find that while they are not perfect (no surprise there) they help a lot
21:03:11 <Vorpal> rather than cleaning every other month I only need to clean maybe 1.5 times / year
21:03:14 <Phantom_Hoover> I guess all the dust in this room is because it's carpeted or something.
21:03:29 <Vorpal> ah I only have a small carpet in my room
21:03:35 <Vorpal> and it isn't a very dusty one
21:03:52 <Vorpal> according to google translate it is a "rag rug"
21:04:00 <Vorpal> no idea if that was accurately translated
21:04:34 <Vorpal> anyway it doesn't cause that much dust
21:04:56 <Vorpal> Phantom_Hoover, do you have fans pushing air in or sucking air out?
21:05:30 <Vorpal> dust filters wouldn't help that much with the sucking (negative air pressure) design, since then a lot of air will be sucked in elsewhere as well
21:06:54 <Vorpal> huh I was given an address for tomorrow, except when I check google maps and google street view it is just forest there
21:07:29 <Vorpal> I guess it must be newly built
21:08:00 <Vorpal> hm there is a building site on the map (not on street view) some way further down the road, so I guess that is possible
21:19:48 -!- Slereah has quit (Ping timeout: 244 seconds).
21:23:05 <itidus21> ok i think i have done something worth looking at now :D
21:23:27 <itidus21> so my question is, is this a valid turing machine to simulate the bouncing mario bros shell? http://pastebin.com/iu2atVC3
21:29:35 -!- Nisstyre has joined.
21:30:11 -!- Nisstyre has quit (Read error: Connection reset by peer).
21:33:27 <Phantom_Hoover> <Vorpal> Phantom_Hoover, do you have fans pushing air in or sucking air out?
21:46:04 -!- oerjan has joined.
21:46:34 -!- copumpkin has quit (Quit: Computer has gone to sleep.).
21:56:02 -!- Vorpal has quit (Ping timeout: 272 seconds).
22:15:00 -!- copumpkin has joined.
22:27:57 -!- derdon has quit (Remote host closed the connection).
22:28:24 -!- derdon has joined.
22:29:51 -!- nortti_ has joined.
22:33:07 -!- derdon has quit (Ping timeout: 264 seconds).
22:35:28 -!- nortti_ has quit (Quit: AndroIRC - Android IRC Client ( http://www.androirc.com )).
22:36:07 -!- nortti_ has joined.
22:37:10 -!- oonbotti has joined.
22:37:12 <olsner> hmm, so nothing interesting here since last night?
22:37:16 <olsner> see you next week then
22:39:12 <olsner> "The dilemma is that I want something fine-tuned for parsing binary protocols, hence I can't really externalize the tokenizer."
22:39:21 <olsner> externalize .... the tokenizer?
22:39:24 <pikhq> olsner: omosiroi koto ha huhituyounntà
22:40:12 <olsner> pikhq: admit that you gave up after the ha? the thing after that looks like gibberish
22:40:44 <olsner> besides, à in japanese? can you even do that?
22:40:47 <nortti_> elliott: why did you block Heliumint
22:41:10 <olsner> "... a parser that uses a trie structure to determine the parse rules before the actual extraction. The problem of course is that it will not solve the problem with left recursion."
22:42:05 <olsner> maybe I should suggest that this guy read a book and/or get a clue?
22:54:33 <Sgeo> Should I try Bastion
22:57:02 <pikhq> olsner: I use a really absurd romanisation scheme.
22:57:24 <pikhq> olsner: 面白い事は不必要んだ。 in normal Japanese writing.
22:58:26 <pikhq> olsner: "tà" indicates "ta" with a dakuten. i.e. た + ﾞ.
22:59:00 <pikhq> "Omoshiroi koto wa fuhitsuyônda" in a more ordinary romanisation scheme.
23:00:13 <elliott> <nortti_> elliott: why did you block Heliumint
23:00:24 <elliott> The reason is right there on the page.
23:02:54 <itidus21> if i didn't know better i'd think you all speak japanese in addition to the other skills
23:03:10 <pikhq> That's mostly my fault. :P
23:04:33 <itidus21> so what you're doing is romanizing japanese for a finnish audience?
23:05:03 <pikhq> itidus21: Romanizing Japanese by way of a strict encoding of kana orthography.
23:05:44 <itidus21> i'm contented with the word romanizing.. though i used to throw the word transliteration around a lot once
23:06:11 * oerjan wonders if there is any japanese word which would be clearly illegal phonetically in finnish
23:06:49 <oerjan> oh well, there's vowel harmony, but you can make compounds breaking that
23:06:53 <itidus21> theres one way to get a positive proof
23:07:24 <itidus21> if surely if all of the syllables are ok then combining them is ok
23:07:48 <pikhq> oerjan: Hmm. Trying to think of a weird edge case in Japanese.
23:07:51 <oerjan> itidus21: actually no, that's what vowel harmony thwarts
23:08:34 <pikhq> Also, whether or not it looks clearly illegal possibly depends on the chosen romanization.
23:08:35 <oerjan> you cannot have a,o in the same non-compound word as ä,ö,y
23:08:51 <oerjan> hm, add u the first group
23:09:01 <pikhq> "zutsū" and "sùtuu" look rather different, for example.
23:09:41 <oerjan> hm right z is not a finnish letter i think
23:12:59 <pikhq> As do "wikipedia" and "uīkihe̊tèīa"...
23:13:44 <oerjan> ...i hope the japanese at least pronounce it with a p
23:14:46 <pikhq> It's denoted as "he" with a handakuten, i.e. へ + ﾟ, though.
23:16:36 <pikhq> Likewise, "wi" is written as "u" with a small "i", ウィ, because there's no glyph for "wi" (... in modern usage, that anybody actually *knows*)
23:16:56 <pikhq> Similar for "di", ディ.
23:19:53 -!- sebbu2 has joined.
23:19:53 -!- sebbu2 has quit (Changing host).
23:19:53 -!- sebbu2 has joined.
23:20:11 -!- sebbu has quit (Ping timeout: 244 seconds).
23:25:17 <itidus21> {g,a,n,d,l,f}\{t,h,e,g,r,a,y} = {n,d,l,f} !
23:25:50 <itidus21> i wonder if it's gray or grey though
23:29:04 <fizzie> It's "Grey" in my electronic version, at least.
23:31:19 <itidus21> given that gray/grey is a word upon which england(?) and usa have chosen to differ, and the signifigance of LOTR coming from jsdlkjdklsjdkls shit why don't i even know what to call that place
23:39:28 <fizzie> oerjan: Comparing wiki's phonology articles, we don't have a superset of their phonemes. Though it's not *terribly* far. (We're missing IPA /z/, /w/, /N/ and on the vowel side they've got /a/ in place of our /ɑ/.)
23:45:59 <fizzie> http://en.wikipedia.org/wiki/Uvular_nasal
23:48:18 <fizzie> We do have /ŋ/ which I guess is not horribly far, but not the same thing.
23:50:31 <fizzie> And regarding /z/, even though we don't have it natively, it's not as if we were especially unfamiliar with it, what with English and all.
23:54:48 <fizzie> (There might be a tendency to unvoice it to a [t͡s] kind of sound though.)
23:57:01 <fizzie> (Possibly German influence? I think their 'z' letter is pronounced like that.)